openthoughts3_30k / trainer_state.json
gsmyrnis's picture
End of training
f963fd8 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 500,
"global_step": 1235,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004048582995951417,
"grad_norm": 8.333611650880178,
"learning_rate": 3.2258064516129035e-07,
"loss": 1.4026,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 8.335202390181609,
"learning_rate": 6.451612903225807e-07,
"loss": 1.4218,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 8.383779657806848,
"learning_rate": 9.67741935483871e-07,
"loss": 1.4688,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 8.282442667394012,
"learning_rate": 1.2903225806451614e-06,
"loss": 1.4035,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 8.002029086183455,
"learning_rate": 1.6129032258064516e-06,
"loss": 1.4101,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 7.667931888899882,
"learning_rate": 1.935483870967742e-06,
"loss": 1.392,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 6.118244268193469,
"learning_rate": 2.2580645161290324e-06,
"loss": 1.3717,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 6.003565574801474,
"learning_rate": 2.580645161290323e-06,
"loss": 1.3358,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 3.360669569247667,
"learning_rate": 2.903225806451613e-06,
"loss": 1.3331,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 2.827106394564498,
"learning_rate": 3.225806451612903e-06,
"loss": 1.2998,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 2.5370299317289016,
"learning_rate": 3.548387096774194e-06,
"loss": 1.3225,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 6.172918079391466,
"learning_rate": 3.870967741935484e-06,
"loss": 1.2971,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 6.461343411490263,
"learning_rate": 4.193548387096774e-06,
"loss": 1.3197,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 6.622499850130837,
"learning_rate": 4.516129032258065e-06,
"loss": 1.2861,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 6.404456957643055,
"learning_rate": 4.838709677419355e-06,
"loss": 1.293,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 4.9651758430298,
"learning_rate": 5.161290322580646e-06,
"loss": 1.316,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 4.478414421487632,
"learning_rate": 5.483870967741935e-06,
"loss": 1.2683,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 3.7951693311150665,
"learning_rate": 5.806451612903226e-06,
"loss": 1.2207,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 2.5057052044604182,
"learning_rate": 6.129032258064517e-06,
"loss": 1.202,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.7506197721980563,
"learning_rate": 6.451612903225806e-06,
"loss": 1.2082,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 1.656118910612408,
"learning_rate": 6.774193548387097e-06,
"loss": 1.2172,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.706371566824014,
"learning_rate": 7.096774193548388e-06,
"loss": 1.1875,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.7082994733808483,
"learning_rate": 7.4193548387096784e-06,
"loss": 1.1825,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 1.5408332218731293,
"learning_rate": 7.741935483870968e-06,
"loss": 1.1705,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 1.300450953626318,
"learning_rate": 8.064516129032258e-06,
"loss": 1.1537,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 1.0023576637851004,
"learning_rate": 8.387096774193549e-06,
"loss": 1.1542,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 0.9548200730084797,
"learning_rate": 8.70967741935484e-06,
"loss": 1.1428,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 1.0695001347854036,
"learning_rate": 9.03225806451613e-06,
"loss": 1.1337,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 0.9760337849992999,
"learning_rate": 9.35483870967742e-06,
"loss": 1.121,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 0.9483511292086398,
"learning_rate": 9.67741935483871e-06,
"loss": 1.1368,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 0.7775980902569146,
"learning_rate": 1e-05,
"loss": 1.144,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 0.8765763149687843,
"learning_rate": 1.0322580645161291e-05,
"loss": 1.1134,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 0.8401481059294742,
"learning_rate": 1.0645161290322582e-05,
"loss": 1.1462,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 0.6968097804160561,
"learning_rate": 1.096774193548387e-05,
"loss": 1.0706,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 0.7479464146384982,
"learning_rate": 1.1290322580645164e-05,
"loss": 1.1244,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.7554864191839974,
"learning_rate": 1.1612903225806453e-05,
"loss": 1.111,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.6532863655111179,
"learning_rate": 1.1935483870967743e-05,
"loss": 1.0847,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.7787872537996289,
"learning_rate": 1.2258064516129034e-05,
"loss": 1.0979,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.7973334872378441,
"learning_rate": 1.2580645161290324e-05,
"loss": 1.074,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 0.6564288812040109,
"learning_rate": 1.2903225806451613e-05,
"loss": 1.0678,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.6953655783486145,
"learning_rate": 1.3225806451612903e-05,
"loss": 1.0592,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.6953926206212766,
"learning_rate": 1.3548387096774194e-05,
"loss": 1.0274,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.5345948970774378,
"learning_rate": 1.3870967741935486e-05,
"loss": 1.0603,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.6068566650113759,
"learning_rate": 1.4193548387096776e-05,
"loss": 1.009,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.607028512734249,
"learning_rate": 1.4516129032258066e-05,
"loss": 1.06,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 0.6072967769765082,
"learning_rate": 1.4838709677419357e-05,
"loss": 1.037,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.5700661067443111,
"learning_rate": 1.5161290322580646e-05,
"loss": 1.0821,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.5964142991449629,
"learning_rate": 1.5483870967741936e-05,
"loss": 1.0944,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 0.6875820960259476,
"learning_rate": 1.5806451612903226e-05,
"loss": 1.0664,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 0.6326611153613333,
"learning_rate": 1.6129032258064517e-05,
"loss": 1.0433,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 0.5778970826190213,
"learning_rate": 1.6451612903225807e-05,
"loss": 1.061,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.5111818595826042,
"learning_rate": 1.6774193548387098e-05,
"loss": 1.042,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 0.5785729328183238,
"learning_rate": 1.7096774193548388e-05,
"loss": 1.0316,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.6424098072736746,
"learning_rate": 1.741935483870968e-05,
"loss": 1.0336,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.7211050202555828,
"learning_rate": 1.774193548387097e-05,
"loss": 1.0535,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 0.7362807276654936,
"learning_rate": 1.806451612903226e-05,
"loss": 1.0478,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.6976460094856115,
"learning_rate": 1.838709677419355e-05,
"loss": 1.0409,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 0.7397332376075727,
"learning_rate": 1.870967741935484e-05,
"loss": 1.1007,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 1.0209477848347843,
"learning_rate": 1.903225806451613e-05,
"loss": 1.0045,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 1.2007562850058973,
"learning_rate": 1.935483870967742e-05,
"loss": 1.066,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 0.8243761967220639,
"learning_rate": 1.967741935483871e-05,
"loss": 1.0235,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 0.7246634423685037,
"learning_rate": 2e-05,
"loss": 1.0878,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 0.8119262038539811,
"learning_rate": 2.0322580645161292e-05,
"loss": 1.0605,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 1.1451426091976258,
"learning_rate": 2.0645161290322582e-05,
"loss": 1.0199,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 0.9591609308622439,
"learning_rate": 2.0967741935483873e-05,
"loss": 1.0015,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 1.3613490323938195,
"learning_rate": 2.1290322580645163e-05,
"loss": 1.0111,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 0.87814332223975,
"learning_rate": 2.161290322580645e-05,
"loss": 1.0653,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 0.6698211552531007,
"learning_rate": 2.193548387096774e-05,
"loss": 1.0509,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 1.5063066335260737,
"learning_rate": 2.225806451612903e-05,
"loss": 1.041,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 1.0023606386332933,
"learning_rate": 2.2580645161290328e-05,
"loss": 1.0305,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 1.402424207858731,
"learning_rate": 2.2903225806451615e-05,
"loss": 1.015,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 1.0061724615405323,
"learning_rate": 2.3225806451612906e-05,
"loss": 1.0393,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 1.9203485465896635,
"learning_rate": 2.3548387096774196e-05,
"loss": 1.0159,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 1.4847768899532463,
"learning_rate": 2.3870967741935486e-05,
"loss": 1.0405,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 1.7684841696117786,
"learning_rate": 2.4193548387096777e-05,
"loss": 1.0184,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 1.5629633842050954,
"learning_rate": 2.4516129032258067e-05,
"loss": 1.0463,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 1.0395122455979373,
"learning_rate": 2.4838709677419358e-05,
"loss": 1.0342,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 1.7084724204261312,
"learning_rate": 2.5161290322580648e-05,
"loss": 1.0386,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 1.1351959999390049,
"learning_rate": 2.548387096774194e-05,
"loss": 1.0203,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 1.848741361388483,
"learning_rate": 2.5806451612903226e-05,
"loss": 1.0322,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 1.5083474274120405,
"learning_rate": 2.6129032258064516e-05,
"loss": 1.0192,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 1.0152160238330197,
"learning_rate": 2.6451612903225806e-05,
"loss": 1.0787,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 1.972104462393819,
"learning_rate": 2.6774193548387097e-05,
"loss": 1.0265,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 1.2861933894946407,
"learning_rate": 2.7096774193548387e-05,
"loss": 1.0013,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 1.6068414743825579,
"learning_rate": 2.7419354838709678e-05,
"loss": 1.004,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 1.3562887583130363,
"learning_rate": 2.774193548387097e-05,
"loss": 1.0205,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 1.5006539838895685,
"learning_rate": 2.8064516129032262e-05,
"loss": 1.0261,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 1.0897142546318912,
"learning_rate": 2.8387096774193552e-05,
"loss": 1.008,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 1.6917273637647807,
"learning_rate": 2.8709677419354843e-05,
"loss": 1.0241,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 1.1320199895891088,
"learning_rate": 2.9032258064516133e-05,
"loss": 1.0411,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 1.661747418333821,
"learning_rate": 2.9354838709677423e-05,
"loss": 1.0033,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 1.354872262821002,
"learning_rate": 2.9677419354838714e-05,
"loss": 1.0446,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 1.559346703618332,
"learning_rate": 3.0000000000000004e-05,
"loss": 1.0584,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 1.5510591107105858,
"learning_rate": 3.032258064516129e-05,
"loss": 1.0475,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 1.2734258086139396,
"learning_rate": 3.0645161290322585e-05,
"loss": 1.0523,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 1.5106454665433178,
"learning_rate": 3.096774193548387e-05,
"loss": 1.0294,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 1.0832330128471188,
"learning_rate": 3.1290322580645166e-05,
"loss": 1.021,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 1.3224963916772494,
"learning_rate": 3.161290322580645e-05,
"loss": 1.0087,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 1.2168661751992864,
"learning_rate": 3.193548387096774e-05,
"loss": 1.022,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 1.5902291307681693,
"learning_rate": 3.2258064516129034e-05,
"loss": 1.048,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 1.634756350470521,
"learning_rate": 3.258064516129033e-05,
"loss": 1.0517,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 1.4872072427535652,
"learning_rate": 3.2903225806451614e-05,
"loss": 1.0549,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 0.9882535304036589,
"learning_rate": 3.322580645161291e-05,
"loss": 0.9822,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 1.665151299128665,
"learning_rate": 3.3548387096774195e-05,
"loss": 1.007,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 1.24081796924225,
"learning_rate": 3.387096774193549e-05,
"loss": 0.9784,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 1.7778824043725363,
"learning_rate": 3.4193548387096776e-05,
"loss": 1.0257,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 1.6711346572057322,
"learning_rate": 3.451612903225807e-05,
"loss": 1.0049,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 1.2843435475171947,
"learning_rate": 3.483870967741936e-05,
"loss": 1.0062,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 2.5497409225953733,
"learning_rate": 3.516129032258065e-05,
"loss": 1.0234,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 1.4689125600048174,
"learning_rate": 3.548387096774194e-05,
"loss": 1.0579,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 2.504097336577243,
"learning_rate": 3.5806451612903225e-05,
"loss": 1.014,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 1.6962378631193007,
"learning_rate": 3.612903225806452e-05,
"loss": 1.043,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 2.5007850998546433,
"learning_rate": 3.6451612903225805e-05,
"loss": 1.0207,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 1.9505088166227071,
"learning_rate": 3.67741935483871e-05,
"loss": 1.0209,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 2.097523920616982,
"learning_rate": 3.7096774193548386e-05,
"loss": 1.0356,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 1.775300258831302,
"learning_rate": 3.741935483870968e-05,
"loss": 1.0337,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 1.6917948583815896,
"learning_rate": 3.7741935483870974e-05,
"loss": 1.0181,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 1.821493804802333,
"learning_rate": 3.806451612903226e-05,
"loss": 0.9577,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 1.9576193861595281,
"learning_rate": 3.8387096774193555e-05,
"loss": 0.9971,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 1.5037016773252536,
"learning_rate": 3.870967741935484e-05,
"loss": 1.0209,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 1.9725736369658753,
"learning_rate": 3.9032258064516135e-05,
"loss": 1.0352,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 1.6715055687088236,
"learning_rate": 3.935483870967742e-05,
"loss": 1.0183,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 1.5506151727407969,
"learning_rate": 3.9677419354838716e-05,
"loss": 1.0161,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 1.4441046559639465,
"learning_rate": 4e-05,
"loss": 1.03,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 1.3543203951580498,
"learning_rate": 3.9999920040266474e-05,
"loss": 0.9728,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 1.7120678225487613,
"learning_rate": 3.999968016170525e-05,
"loss": 1.0125,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 1.3576900482151706,
"learning_rate": 3.9999280366234394e-05,
"loss": 1.0581,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 1.652650058858321,
"learning_rate": 3.999872065705065e-05,
"loss": 0.99,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 1.0745111946202068,
"learning_rate": 3.999800103862945e-05,
"loss": 1.002,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 1.6487904659894923,
"learning_rate": 3.9997121516724834e-05,
"loss": 1.0652,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 1.2924704198553028,
"learning_rate": 3.999608209836944e-05,
"loss": 1.0403,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 1.3721383141158674,
"learning_rate": 3.999488279187443e-05,
"loss": 1.0275,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 1.2102548503708643,
"learning_rate": 3.9993523606829425e-05,
"loss": 0.9704,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 1.3591612104900874,
"learning_rate": 3.9992004554102435e-05,
"loss": 1.0347,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 1.481602897646584,
"learning_rate": 3.999032564583976e-05,
"loss": 0.961,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 1.3557926423516333,
"learning_rate": 3.998848689546591e-05,
"loss": 1.0167,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 1.3448026028482167,
"learning_rate": 3.9986488317683496e-05,
"loss": 0.9678,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 1.7138068870190226,
"learning_rate": 3.9984329928473067e-05,
"loss": 1.004,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 1.13145059563698,
"learning_rate": 3.998201174509306e-05,
"loss": 0.9504,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 1.4779536961378035,
"learning_rate": 3.997953378607961e-05,
"loss": 1.0271,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 1.2973598593706228,
"learning_rate": 3.9976896071246403e-05,
"loss": 1.0032,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 1.5826896984789913,
"learning_rate": 3.997409862168454e-05,
"loss": 1.0543,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 1.872373896173996,
"learning_rate": 3.997114145976237e-05,
"loss": 0.9852,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 0.9637207832315541,
"learning_rate": 3.9968024609125245e-05,
"loss": 1.0023,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 1.6775446749665166,
"learning_rate": 3.996474809469545e-05,
"loss": 0.9375,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 1.29817239024934,
"learning_rate": 3.996131194267189e-05,
"loss": 1.0106,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 0.9991699522020473,
"learning_rate": 3.995771618052995e-05,
"loss": 1.0221,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 1.523373015273603,
"learning_rate": 3.9953960837021256e-05,
"loss": 1.0069,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 1.2505305335508663,
"learning_rate": 3.995004594217343e-05,
"loss": 1.0167,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 1.4022133159946786,
"learning_rate": 3.994597152728987e-05,
"loss": 0.9948,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 1.0126881203730231,
"learning_rate": 3.994173762494947e-05,
"loss": 1.0351,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 1.5512190313414809,
"learning_rate": 3.9937344269006424e-05,
"loss": 0.991,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 1.0366220595295745,
"learning_rate": 3.993279149458988e-05,
"loss": 1.0552,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 1.0220737113414442,
"learning_rate": 3.9928079338103695e-05,
"loss": 1.0195,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 1.3906356894859822,
"learning_rate": 3.992320783722616e-05,
"loss": 0.9792,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 1.2027424821589094,
"learning_rate": 3.991817703090965e-05,
"loss": 0.9527,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 1.228183988242521,
"learning_rate": 3.991298695938038e-05,
"loss": 0.9756,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 1.0334843507560314,
"learning_rate": 3.9907637664138e-05,
"loss": 0.9969,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 2.089374249379887,
"learning_rate": 3.990212918795534e-05,
"loss": 1.0343,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 1.5009505262801923,
"learning_rate": 3.9896461574878036e-05,
"loss": 1.0053,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 1.8845076146021211,
"learning_rate": 3.989063487022416e-05,
"loss": 1.0117,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 2.1277108822937185,
"learning_rate": 3.9884649120583903e-05,
"loss": 0.9606,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 1.0360161479031147,
"learning_rate": 3.987850437381915e-05,
"loss": 0.9762,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 2.911799243339387,
"learning_rate": 3.987220067906313e-05,
"loss": 0.9895,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 2.052306469665553,
"learning_rate": 3.986573808672002e-05,
"loss": 1.0542,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 3.18767914479431,
"learning_rate": 3.9859116648464534e-05,
"loss": 1.0292,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 2.8528151744481303,
"learning_rate": 3.985233641724152e-05,
"loss": 1.0265,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 2.023298306434441,
"learning_rate": 3.984539744726554e-05,
"loss": 0.9979,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 2.1284509947537322,
"learning_rate": 3.983829979402038e-05,
"loss": 0.9929,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 2.095676586026605,
"learning_rate": 3.983104351425871e-05,
"loss": 1.0382,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 1.8908991289290957,
"learning_rate": 3.982362866600154e-05,
"loss": 1.0164,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 2.059751080231314,
"learning_rate": 3.9816055308537806e-05,
"loss": 1.0184,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 1.4193103541667949,
"learning_rate": 3.980832350242388e-05,
"loss": 0.9925,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 2.4117723110956732,
"learning_rate": 3.980043330948306e-05,
"loss": 0.9818,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 1.7199590101023223,
"learning_rate": 3.9792384792805124e-05,
"loss": 1.0331,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 2.3772230592925583,
"learning_rate": 3.978417801674579e-05,
"loss": 0.9818,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 1.8145485010626468,
"learning_rate": 3.9775813046926244e-05,
"loss": 1.0052,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 2.208938139917684,
"learning_rate": 3.976728995023254e-05,
"loss": 1.0107,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 1.7834695206954263,
"learning_rate": 3.975860879481514e-05,
"loss": 0.9941,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 2.130947588431107,
"learning_rate": 3.974976965008833e-05,
"loss": 1.0218,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 2.013193145114115,
"learning_rate": 3.9740772586729674e-05,
"loss": 1.0222,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 1.744594874552037,
"learning_rate": 3.973161767667946e-05,
"loss": 1.0132,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 1.5888411062054084,
"learning_rate": 3.972230499314009e-05,
"loss": 1.0005,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 2.001874361601905,
"learning_rate": 3.9712834610575545e-05,
"loss": 0.9497,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 1.37551113139702,
"learning_rate": 3.9703206604710746e-05,
"loss": 1.0255,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 2.4972211206798542,
"learning_rate": 3.969342105253098e-05,
"loss": 0.9811,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 2.212977688404549,
"learning_rate": 3.9683478032281246e-05,
"loss": 1.0259,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 1.536025677786751,
"learning_rate": 3.967337762346568e-05,
"loss": 0.9766,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 1.4996687009689824,
"learning_rate": 3.9663119906846885e-05,
"loss": 0.9905,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 1.5015230955356587,
"learning_rate": 3.965270496444528e-05,
"loss": 0.9598,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 0.9979703228945389,
"learning_rate": 3.9642132879538465e-05,
"loss": 0.996,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 2.092664652213499,
"learning_rate": 3.9631403736660566e-05,
"loss": 0.9925,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 1.7123470723716483,
"learning_rate": 3.962051762160151e-05,
"loss": 1.0115,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 1.7193047500365453,
"learning_rate": 3.960947462140638e-05,
"loss": 0.9501,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 1.5949462676532407,
"learning_rate": 3.959827482437472e-05,
"loss": 0.9638,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 1.5297896392351247,
"learning_rate": 3.958691832005981e-05,
"loss": 1.0463,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 1.2761002880438659,
"learning_rate": 3.9575405199267944e-05,
"loss": 0.9993,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 1.895660719034219,
"learning_rate": 3.956373555405774e-05,
"loss": 1.0143,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 1.5291307782000485,
"learning_rate": 3.955190947773937e-05,
"loss": 1.0034,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 1.6853775319724205,
"learning_rate": 3.9539927064873824e-05,
"loss": 1.0283,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 1.2918944325748385,
"learning_rate": 3.952778841127214e-05,
"loss": 1.0088,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 1.5812442641548923,
"learning_rate": 3.95154936139947e-05,
"loss": 0.9878,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 1.2397220236213669,
"learning_rate": 3.950304277135034e-05,
"loss": 0.9722,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 1.660203486054776,
"learning_rate": 3.94904359828957e-05,
"loss": 1.0355,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 1.308041216923771,
"learning_rate": 3.94776733494343e-05,
"loss": 0.9959,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 1.4231035653703161,
"learning_rate": 3.9464754973015826e-05,
"loss": 0.9954,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 1.2976555594265085,
"learning_rate": 3.9451680956935275e-05,
"loss": 0.9597,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 1.2795331529492957,
"learning_rate": 3.943845140573213e-05,
"loss": 0.983,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.9252498773802087,
"learning_rate": 3.942506642518952e-05,
"loss": 0.9791,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 1.2150292576923796,
"learning_rate": 3.9411526122333396e-05,
"loss": 0.981,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 0.9668366293168936,
"learning_rate": 3.939783060543167e-05,
"loss": 0.9859,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 1.0475216575243056,
"learning_rate": 3.938397998399333e-05,
"loss": 0.9511,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.8923120657520823,
"learning_rate": 3.936997436876756e-05,
"loss": 0.9724,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.9852291883606946,
"learning_rate": 3.935581387174289e-05,
"loss": 0.987,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 0.7430208365876981,
"learning_rate": 3.9341498606146295e-05,
"loss": 1.0261,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 1.0632774246644026,
"learning_rate": 3.932702868644224e-05,
"loss": 0.9575,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.8081448119098288,
"learning_rate": 3.931240422833181e-05,
"loss": 1.0013,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.8608837434856927,
"learning_rate": 3.9297625348751806e-05,
"loss": 0.9745,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 0.8076299889945348,
"learning_rate": 3.9282692165873744e-05,
"loss": 0.9718,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 1.1394961830415011,
"learning_rate": 3.9267604799102945e-05,
"loss": 1.022,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.7025948269096001,
"learning_rate": 3.9252363369077615e-05,
"loss": 0.9799,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 1.0855674075004202,
"learning_rate": 3.9236967997667796e-05,
"loss": 1.0015,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 0.9395315655364539,
"learning_rate": 3.9221418807974496e-05,
"loss": 1.0136,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 1.037846043091017,
"learning_rate": 3.92057159243286e-05,
"loss": 0.967,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 0.8182627993928366,
"learning_rate": 3.9189859472289956e-05,
"loss": 0.9712,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 0.8404138557681315,
"learning_rate": 3.9173849578646324e-05,
"loss": 0.9736,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 0.7572641797152988,
"learning_rate": 3.9157686371412395e-05,
"loss": 0.9578,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.7542541350885996,
"learning_rate": 3.914136997982874e-05,
"loss": 0.9762,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 0.7168581617870907,
"learning_rate": 3.912490053436079e-05,
"loss": 0.9759,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 0.7344191715407573,
"learning_rate": 3.9108278166697805e-05,
"loss": 1.0115,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.7793812316335148,
"learning_rate": 3.909150300975178e-05,
"loss": 0.9327,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 0.7840536731009324,
"learning_rate": 3.907457519765642e-05,
"loss": 1.0415,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.6083886460867812,
"learning_rate": 3.905749486576607e-05,
"loss": 0.9628,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.8622071632195134,
"learning_rate": 3.90402621506546e-05,
"loss": 1.0182,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 0.7999413471516166,
"learning_rate": 3.902287719011435e-05,
"loss": 0.997,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.8199447531675075,
"learning_rate": 3.900534012315499e-05,
"loss": 0.9861,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 0.5664365069506606,
"learning_rate": 3.8987651090002454e-05,
"loss": 0.9892,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 0.6560669056248714,
"learning_rate": 3.896981023209777e-05,
"loss": 1.0041,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 0.8601316425317141,
"learning_rate": 3.8951817692095966e-05,
"loss": 0.9798,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 1.0906654512463587,
"learning_rate": 3.893367361386491e-05,
"loss": 0.9677,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 1.1257104673247718,
"learning_rate": 3.891537814248417e-05,
"loss": 0.9909,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 1.167219734931071,
"learning_rate": 3.889693142424384e-05,
"loss": 0.967,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 0.7551885902089175,
"learning_rate": 3.8878333606643405e-05,
"loss": 1.0042,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 0.8626080979327223,
"learning_rate": 3.885958483839049e-05,
"loss": 0.9995,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 1.3209162224606796,
"learning_rate": 3.8840685269399786e-05,
"loss": 1.0144,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 0.6927783494983766,
"learning_rate": 3.882163505079171e-05,
"loss": 1.0269,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 0.8288208580545586,
"learning_rate": 3.880243433489132e-05,
"loss": 0.974,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 0.7707134052568022,
"learning_rate": 3.878308327522702e-05,
"loss": 0.9515,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 0.9785093422598659,
"learning_rate": 3.876358202652936e-05,
"loss": 0.989,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 1.2329269658821855,
"learning_rate": 3.8743930744729835e-05,
"loss": 0.9602,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.8299985769833725,
"learning_rate": 3.8724129586959535e-05,
"loss": 0.9502,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.9087458932347574,
"learning_rate": 3.870417871154801e-05,
"loss": 0.9212,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 1.202510070838854,
"learning_rate": 3.868407827802192e-05,
"loss": 0.9742,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.9625316466537471,
"learning_rate": 3.8663828447103804e-05,
"loss": 0.912,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.8761450857944901,
"learning_rate": 3.8643429380710765e-05,
"loss": 0.9024,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.9235878681660622,
"learning_rate": 3.862288124195319e-05,
"loss": 0.988,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.7792870622033949,
"learning_rate": 3.8602184195133464e-05,
"loss": 0.9137,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 0.877491414666308,
"learning_rate": 3.8581338405744604e-05,
"loss": 0.8957,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 0.9155750850928364,
"learning_rate": 3.8560344040468996e-05,
"loss": 0.9405,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 1.0417538225951872,
"learning_rate": 3.853920126717703e-05,
"loss": 0.9184,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 0.8214206222112403,
"learning_rate": 3.851791025492574e-05,
"loss": 0.9328,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.6675103441058624,
"learning_rate": 3.8496471173957515e-05,
"loss": 0.9357,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.6338943371771394,
"learning_rate": 3.847488419569866e-05,
"loss": 0.9468,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.5959515152702208,
"learning_rate": 3.8453149492758077e-05,
"loss": 0.9456,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 0.5603904043326104,
"learning_rate": 3.8431267238925886e-05,
"loss": 0.9109,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.6139880737135063,
"learning_rate": 3.840923760917199e-05,
"loss": 0.9892,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 0.6891456012339401,
"learning_rate": 3.838706077964473e-05,
"loss": 0.927,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 0.7778322635724332,
"learning_rate": 3.836473692766944e-05,
"loss": 0.9641,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 0.5897795433144808,
"learning_rate": 3.834226623174705e-05,
"loss": 0.9514,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.6484530192750916,
"learning_rate": 3.831964887155264e-05,
"loss": 0.9436,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.7197586125846916,
"learning_rate": 3.829688502793403e-05,
"loss": 0.9322,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.6460224883147329,
"learning_rate": 3.827397488291029e-05,
"loss": 1.0125,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.5881132996019722,
"learning_rate": 3.8250918619670343e-05,
"loss": 0.9444,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.5820796658632065,
"learning_rate": 3.822771642257145e-05,
"loss": 0.9391,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.6740922865725226,
"learning_rate": 3.820436847713776e-05,
"loss": 0.9651,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 0.6547561298863256,
"learning_rate": 3.8180874970058826e-05,
"loss": 0.9202,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.6554313412571405,
"learning_rate": 3.8157236089188103e-05,
"loss": 0.927,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.5969943096653779,
"learning_rate": 3.813345202354145e-05,
"loss": 0.9255,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 0.740558476494132,
"learning_rate": 3.810952296329563e-05,
"loss": 0.9348,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.7970232566423644,
"learning_rate": 3.808544909978676e-05,
"loss": 0.9523,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.5230555044955315,
"learning_rate": 3.806123062550882e-05,
"loss": 0.9567,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.6889580216935847,
"learning_rate": 3.803686773411208e-05,
"loss": 0.9297,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.6778480868869965,
"learning_rate": 3.801236062040158e-05,
"loss": 0.9286,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 0.7656430734698791,
"learning_rate": 3.798770948033553e-05,
"loss": 0.948,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 1.0030487533107928,
"learning_rate": 3.796291451102381e-05,
"loss": 0.954,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.84110559034925,
"learning_rate": 3.793797591072631e-05,
"loss": 0.9404,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.6308100475281114,
"learning_rate": 3.791289387885144e-05,
"loss": 0.9464,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.7811589204844055,
"learning_rate": 3.788766861595443e-05,
"loss": 0.9012,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 1.0041453513618113,
"learning_rate": 3.7862300323735835e-05,
"loss": 0.9098,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 1.425263121320758,
"learning_rate": 3.783678920503983e-05,
"loss": 0.9218,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 1.0271014833832348,
"learning_rate": 3.781113546385265e-05,
"loss": 0.9386,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 1.214414861039099,
"learning_rate": 3.7785339305300916e-05,
"loss": 0.9517,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.6774360047245315,
"learning_rate": 3.7759400935650025e-05,
"loss": 0.9288,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.8686761599444099,
"learning_rate": 3.77333205623025e-05,
"loss": 0.9182,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 1.0157912900851507,
"learning_rate": 3.7707098393796295e-05,
"loss": 0.9565,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.9399893954382302,
"learning_rate": 3.7680734639803185e-05,
"loss": 0.9209,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 0.8749138875267188,
"learning_rate": 3.765422951112704e-05,
"loss": 0.9445,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 0.7809939492480434,
"learning_rate": 3.762758321970216e-05,
"loss": 0.9797,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.6070806184540536,
"learning_rate": 3.7600795978591584e-05,
"loss": 0.9465,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.589878094877523,
"learning_rate": 3.757386800198538e-05,
"loss": 0.9342,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.7016241067341403,
"learning_rate": 3.7546799505198925e-05,
"loss": 0.9403,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 1.0038120264644248,
"learning_rate": 3.7519590704671197e-05,
"loss": 0.939,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.8723500255577337,
"learning_rate": 3.749224181796305e-05,
"loss": 0.9598,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.8357878687800793,
"learning_rate": 3.746475306375545e-05,
"loss": 0.9628,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.8841452832727392,
"learning_rate": 3.743712466184774e-05,
"loss": 0.9501,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.6805302522947746,
"learning_rate": 3.7409356833155885e-05,
"loss": 0.9672,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.857822312332312,
"learning_rate": 3.73814497997107e-05,
"loss": 0.9471,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.7964669273495139,
"learning_rate": 3.7353403784656095e-05,
"loss": 0.9304,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.6032723806277464,
"learning_rate": 3.732521901224724e-05,
"loss": 0.9685,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.8498038189474576,
"learning_rate": 3.7296895707848844e-05,
"loss": 0.9575,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.7645695983098254,
"learning_rate": 3.7268434097933275e-05,
"loss": 0.912,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.6090599838764823,
"learning_rate": 3.723983441007882e-05,
"loss": 0.9648,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.8613059985812866,
"learning_rate": 3.721109687296781e-05,
"loss": 0.9116,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.8062880217571453,
"learning_rate": 3.718222171638484e-05,
"loss": 0.9613,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 0.7791129277239077,
"learning_rate": 3.715320917121488e-05,
"loss": 0.9348,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.6633377698689099,
"learning_rate": 3.712405946944149e-05,
"loss": 0.9516,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.6312818452167418,
"learning_rate": 3.709477284414488e-05,
"loss": 0.9486,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.5177138110969999,
"learning_rate": 3.7065349529500136e-05,
"loss": 0.9156,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 0.5358183743012603,
"learning_rate": 3.7035789760775305e-05,
"loss": 0.8798,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.6107003514872626,
"learning_rate": 3.70060937743295e-05,
"loss": 0.9406,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.6637062397186996,
"learning_rate": 3.6976261807611045e-05,
"loss": 0.9398,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.5578547072929093,
"learning_rate": 3.694629409915555e-05,
"loss": 0.9549,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.6026108989684528,
"learning_rate": 3.691619088858401e-05,
"loss": 0.9462,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.7015136709727977,
"learning_rate": 3.68859524166009e-05,
"loss": 0.9359,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.7714660650059075,
"learning_rate": 3.6855578924992226e-05,
"loss": 0.9492,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.8688632597898773,
"learning_rate": 3.6825070656623626e-05,
"loss": 0.9423,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.8512839427219354,
"learning_rate": 3.67944278554384e-05,
"loss": 0.934,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 0.6979990841173546,
"learning_rate": 3.676365076645557e-05,
"loss": 0.978,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.6342429010042794,
"learning_rate": 3.673273963576791e-05,
"loss": 0.9386,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.5610588294694288,
"learning_rate": 3.670169471054001e-05,
"loss": 0.9096,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.7987794739128546,
"learning_rate": 3.6670516239006254e-05,
"loss": 0.9585,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 1.0524757963305331,
"learning_rate": 3.663920447046888e-05,
"loss": 0.9905,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.5798961961524499,
"learning_rate": 3.6607759655295954e-05,
"loss": 0.9661,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.731867655544751,
"learning_rate": 3.657618204491937e-05,
"loss": 0.9569,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.9830406045324748,
"learning_rate": 3.6544471891832875e-05,
"loss": 0.967,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.9671558102261097,
"learning_rate": 3.6512629449589983e-05,
"loss": 0.962,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.8582188273120706,
"learning_rate": 3.6480654972802045e-05,
"loss": 0.9557,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.7871550480936079,
"learning_rate": 3.644854871713611e-05,
"loss": 0.9375,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.7345449527379003,
"learning_rate": 3.6416310939312924e-05,
"loss": 0.9821,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.5925830649339158,
"learning_rate": 3.638394189710493e-05,
"loss": 0.9238,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.4671462088410286,
"learning_rate": 3.635144184933412e-05,
"loss": 0.952,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.5097527009486488,
"learning_rate": 3.6318811055869994e-05,
"loss": 0.939,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.6729349582045147,
"learning_rate": 3.628604977762752e-05,
"loss": 0.9315,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.8126701034730582,
"learning_rate": 3.6253158276565004e-05,
"loss": 0.9992,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.5028349090229165,
"learning_rate": 3.6220136815682016e-05,
"loss": 0.9328,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.5670497826768459,
"learning_rate": 3.618698565901727e-05,
"loss": 0.9626,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.6223503318753266,
"learning_rate": 3.615370507164654e-05,
"loss": 0.9362,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 1.0066488449481343,
"learning_rate": 3.61202953196805e-05,
"loss": 0.9504,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 0.7869429882687041,
"learning_rate": 3.6086756670262654e-05,
"loss": 0.9456,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.8251382779481367,
"learning_rate": 3.6053089391567135e-05,
"loss": 0.9425,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.5614309447473174,
"learning_rate": 3.601929375279662e-05,
"loss": 0.9138,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.7811588481244207,
"learning_rate": 3.598537002418012e-05,
"loss": 0.9793,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.6175907134968016,
"learning_rate": 3.5951318476970885e-05,
"loss": 0.906,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.5108279187519688,
"learning_rate": 3.591713938344416e-05,
"loss": 0.949,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.6241297505468675,
"learning_rate": 3.588283301689507e-05,
"loss": 0.9236,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 0.506407967104974,
"learning_rate": 3.5848399651636424e-05,
"loss": 0.9764,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.4704842489151783,
"learning_rate": 3.5813839562996473e-05,
"loss": 0.9532,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.522332034451328,
"learning_rate": 3.5779153027316775e-05,
"loss": 0.9664,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.5402712736741025,
"learning_rate": 3.574434032194994e-05,
"loss": 0.967,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.5325330496186494,
"learning_rate": 3.5709401725257436e-05,
"loss": 0.9461,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.42868428222660976,
"learning_rate": 3.5674337516607346e-05,
"loss": 0.9246,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.5347719805102209,
"learning_rate": 3.563914797637216e-05,
"loss": 0.9103,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.49392690503155795,
"learning_rate": 3.560383338592649e-05,
"loss": 0.9347,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.7795269884712429,
"learning_rate": 3.556839402764487e-05,
"loss": 0.9394,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.8928562131244747,
"learning_rate": 3.553283018489946e-05,
"loss": 0.9458,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.7233083330762156,
"learning_rate": 3.54971421420578e-05,
"loss": 0.931,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.488545012334917,
"learning_rate": 3.5461330184480524e-05,
"loss": 0.9649,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 0.5796677349970902,
"learning_rate": 3.54253945985191e-05,
"loss": 0.9367,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.6445648825144046,
"learning_rate": 3.5389335671513524e-05,
"loss": 0.918,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.48411310724676127,
"learning_rate": 3.5353153691789986e-05,
"loss": 0.9419,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.7389606295742934,
"learning_rate": 3.531684894865866e-05,
"loss": 0.9382,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.7184553491864023,
"learning_rate": 3.528042173241128e-05,
"loss": 0.9091,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.7395311616102086,
"learning_rate": 3.524387233431892e-05,
"loss": 0.9402,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.7169619574729609,
"learning_rate": 3.520720104662958e-05,
"loss": 0.8763,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.5842168634769135,
"learning_rate": 3.5170408162565904e-05,
"loss": 0.9378,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.6275248803059679,
"learning_rate": 3.513349397632281e-05,
"loss": 0.9466,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.6620635556179478,
"learning_rate": 3.5096458783065145e-05,
"loss": 0.956,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.5691828290522688,
"learning_rate": 3.505930287892533e-05,
"loss": 0.8941,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.5239737874683912,
"learning_rate": 3.502202656100099e-05,
"loss": 0.9581,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.480757092943525,
"learning_rate": 3.498463012735256e-05,
"loss": 0.9162,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.5768795887465396,
"learning_rate": 3.494711387700094e-05,
"loss": 0.9846,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.6369810752949592,
"learning_rate": 3.490947810992505e-05,
"loss": 0.933,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 0.5478647163758876,
"learning_rate": 3.48717231270595e-05,
"loss": 0.9352,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.4730302087080864,
"learning_rate": 3.483384923029211e-05,
"loss": 0.9561,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.6062773705426638,
"learning_rate": 3.479585672246156e-05,
"loss": 0.9975,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.5895584218628388,
"learning_rate": 3.4757745907354924e-05,
"loss": 0.9442,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.5407427268594074,
"learning_rate": 3.471951708970527e-05,
"loss": 0.9156,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.6249589004614358,
"learning_rate": 3.468117057518921e-05,
"loss": 0.8794,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.8793939030443602,
"learning_rate": 3.4642706670424426e-05,
"loss": 0.9207,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.988513271859189,
"learning_rate": 3.460412568296731e-05,
"loss": 0.8881,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 1.0937100398598385,
"learning_rate": 3.456542792131039e-05,
"loss": 0.8875,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.7608425419674375,
"learning_rate": 3.4526613694879936e-05,
"loss": 0.9592,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.48054222928587437,
"learning_rate": 3.448768331403348e-05,
"loss": 0.9401,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.6528068753476125,
"learning_rate": 3.44486370900573e-05,
"loss": 0.9222,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 1.0988673152509985,
"learning_rate": 3.440947533516396e-05,
"loss": 0.9552,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.7965924413505715,
"learning_rate": 3.4370198362489816e-05,
"loss": 0.9241,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 0.6071078048880951,
"learning_rate": 3.43308064860925e-05,
"loss": 0.9433,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.47788138882050907,
"learning_rate": 3.4291300020948393e-05,
"loss": 0.934,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.7651881685237016,
"learning_rate": 3.425167928295015e-05,
"loss": 0.9582,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.8220115666147271,
"learning_rate": 3.421194458890411e-05,
"loss": 0.9297,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.6500924688484875,
"learning_rate": 3.4172096256527865e-05,
"loss": 0.8576,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.7251314763913599,
"learning_rate": 3.4132134604447595e-05,
"loss": 0.9421,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.6202108091850395,
"learning_rate": 3.409205995219561e-05,
"loss": 0.9256,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.6918553366024746,
"learning_rate": 3.4051872620207765e-05,
"loss": 0.9377,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.6895195017590655,
"learning_rate": 3.40115729298209e-05,
"loss": 0.9896,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.6389081481622437,
"learning_rate": 3.397116120327025e-05,
"loss": 0.9171,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.4456199240024166,
"learning_rate": 3.3930637763686916e-05,
"loss": 0.9519,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.615379469316467,
"learning_rate": 3.389000293509524e-05,
"loss": 0.9555,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.8553647926612643,
"learning_rate": 3.3849257042410225e-05,
"loss": 0.9304,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.9732817995767903,
"learning_rate": 3.380840041143494e-05,
"loss": 0.9012,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.7531535692708544,
"learning_rate": 3.376743336885792e-05,
"loss": 0.9201,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.3796828452019743,
"learning_rate": 3.372635624225055e-05,
"loss": 0.9419,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.6191225508473505,
"learning_rate": 3.3685169360064445e-05,
"loss": 0.9339,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.5182982505536988,
"learning_rate": 3.3643873051628804e-05,
"loss": 0.9374,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.47931953202498045,
"learning_rate": 3.3602467647147814e-05,
"loss": 0.9515,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.7279203871446972,
"learning_rate": 3.356095347769799e-05,
"loss": 0.9455,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.8202845875434519,
"learning_rate": 3.351933087522552e-05,
"loss": 0.9255,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.620658834022427,
"learning_rate": 3.3477600172543626e-05,
"loss": 0.9153,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.5871828036755316,
"learning_rate": 3.3435761703329894e-05,
"loss": 0.9231,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.7726748948532435,
"learning_rate": 3.3393815802123606e-05,
"loss": 0.9298,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.780316277952043,
"learning_rate": 3.335176280432308e-05,
"loss": 0.896,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.5401333932431787,
"learning_rate": 3.330960304618295e-05,
"loss": 0.9333,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.4981252309689352,
"learning_rate": 3.326733686481153e-05,
"loss": 0.8693,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.7072418928226705,
"learning_rate": 3.3224964598168084e-05,
"loss": 0.9535,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.7771657901727996,
"learning_rate": 3.318248658506012e-05,
"loss": 0.9085,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.7102317626123384,
"learning_rate": 3.313990316514069e-05,
"loss": 0.9449,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.4600661677258215,
"learning_rate": 3.309721467890571e-05,
"loss": 0.9629,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.46438513669248255,
"learning_rate": 3.305442146769115e-05,
"loss": 0.9475,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.5816238835032491,
"learning_rate": 3.30115238736704e-05,
"loss": 0.9696,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.6069206553561647,
"learning_rate": 3.296852223985148e-05,
"loss": 0.8958,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 0.49201877807968575,
"learning_rate": 3.2925416910074315e-05,
"loss": 0.9287,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.4967257479145969,
"learning_rate": 3.288220822900796e-05,
"loss": 0.9235,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.6040076251092747,
"learning_rate": 3.283889654214788e-05,
"loss": 0.9097,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.6684882122939547,
"learning_rate": 3.279548219581317e-05,
"loss": 0.9116,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.7078860153877542,
"learning_rate": 3.275196553714379e-05,
"loss": 0.9589,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.6151116375626962,
"learning_rate": 3.270834691409778e-05,
"loss": 0.9366,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.5466608966361449,
"learning_rate": 3.266462667544849e-05,
"loss": 0.9425,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.5913202916300112,
"learning_rate": 3.262080517078178e-05,
"loss": 0.939,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.5464254730077653,
"learning_rate": 3.257688275049323e-05,
"loss": 0.95,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.7007608711249035,
"learning_rate": 3.253285976578535e-05,
"loss": 0.9317,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.6885018954214152,
"learning_rate": 3.2488736568664756e-05,
"loss": 0.9409,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.6507491645212887,
"learning_rate": 3.244451351193935e-05,
"loss": 0.948,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 0.7050280679947862,
"learning_rate": 3.240019094921551e-05,
"loss": 0.9194,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.747811574331153,
"learning_rate": 3.235576923489527e-05,
"loss": 0.919,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.4965536798272971,
"learning_rate": 3.2311248724173476e-05,
"loss": 0.9683,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.5990728209704519,
"learning_rate": 3.226662977303494e-05,
"loss": 0.9145,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.7705509715485763,
"learning_rate": 3.2221912738251626e-05,
"loss": 0.9299,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.5122456374324237,
"learning_rate": 3.2177097977379736e-05,
"loss": 0.9686,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.6052484241008574,
"learning_rate": 3.21321858487569e-05,
"loss": 0.926,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.6656184671216725,
"learning_rate": 3.208717671149931e-05,
"loss": 0.9555,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.5502642714890511,
"learning_rate": 3.204207092549882e-05,
"loss": 0.8803,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 0.4890924142749746,
"learning_rate": 3.19968688514201e-05,
"loss": 0.8995,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.5774566600561619,
"learning_rate": 3.195157085069774e-05,
"loss": 0.9564,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.6371902221994836,
"learning_rate": 3.190617728553332e-05,
"loss": 0.998,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.5055178005407451,
"learning_rate": 3.1860688518892606e-05,
"loss": 0.8722,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.48778106074774974,
"learning_rate": 3.181510491450254e-05,
"loss": 0.9332,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.556247296326318,
"learning_rate": 3.176942683684842e-05,
"loss": 0.9356,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.48495590205820405,
"learning_rate": 3.1723654651170934e-05,
"loss": 0.958,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.47976958221901034,
"learning_rate": 3.167778872346327e-05,
"loss": 0.9238,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.528877328001153,
"learning_rate": 3.163182942046815e-05,
"loss": 0.9255,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.5710100119217286,
"learning_rate": 3.158577710967494e-05,
"loss": 0.9272,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.4441583994496748,
"learning_rate": 3.153963215931669e-05,
"loss": 0.9362,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.6583639077628991,
"learning_rate": 3.14933949383672e-05,
"loss": 0.9498,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.7023733623786277,
"learning_rate": 3.144706581653805e-05,
"loss": 0.9404,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.44692764508140387,
"learning_rate": 3.140064516427566e-05,
"loss": 0.9648,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.7014988327999577,
"learning_rate": 3.135413335275833e-05,
"loss": 0.9359,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.6511186061736828,
"learning_rate": 3.130753075389327e-05,
"loss": 0.9257,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.6593415383841089,
"learning_rate": 3.1260837740313625e-05,
"loss": 0.9522,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.6144800330042101,
"learning_rate": 3.121405468537547e-05,
"loss": 0.9318,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.7091831963927984,
"learning_rate": 3.116718196315488e-05,
"loss": 0.8888,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.5620402720021436,
"learning_rate": 3.1120219948444894e-05,
"loss": 0.9148,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.5459260392436435,
"learning_rate": 3.107316901675252e-05,
"loss": 0.9676,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.516910550724978,
"learning_rate": 3.1026029544295745e-05,
"loss": 0.9082,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.5977012549010818,
"learning_rate": 3.097880190800057e-05,
"loss": 0.9444,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.5998128952994659,
"learning_rate": 3.093148648549788e-05,
"loss": 0.9773,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.5422747566388798,
"learning_rate": 3.088408365512055e-05,
"loss": 0.9083,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.5588974109995464,
"learning_rate": 3.083659379590034e-05,
"loss": 0.9501,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.5530431196880292,
"learning_rate": 3.0789017287564913e-05,
"loss": 0.9624,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.5178788062468911,
"learning_rate": 3.074135451053475e-05,
"loss": 0.8809,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.5718483885892799,
"learning_rate": 3.069360584592015e-05,
"loss": 0.9683,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.5164685571647706,
"learning_rate": 3.064577167551816e-05,
"loss": 0.9102,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.563685741331889,
"learning_rate": 3.059785238180953e-05,
"loss": 0.9725,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.5126301549008246,
"learning_rate": 3.054984834795566e-05,
"loss": 0.9305,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.515774259766013,
"learning_rate": 3.0501759957795537e-05,
"loss": 0.9455,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.6481473197350509,
"learning_rate": 3.0453587595842627e-05,
"loss": 0.9463,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.4981163439558594,
"learning_rate": 3.0405331647281856e-05,
"loss": 0.9131,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.42798436159337433,
"learning_rate": 3.0356992497966508e-05,
"loss": 0.8989,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.5764892310416084,
"learning_rate": 3.0308570534415137e-05,
"loss": 0.8892,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.6922136060579253,
"learning_rate": 3.0260066143808467e-05,
"loss": 0.9207,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.3825115232082585,
"learning_rate": 3.021147971398632e-05,
"loss": 0.9723,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.5850649444218881,
"learning_rate": 3.0162811633444478e-05,
"loss": 0.9838,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.5258489180556746,
"learning_rate": 3.0114062291331635e-05,
"loss": 0.8966,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.5826697289220425,
"learning_rate": 3.0065232077446217e-05,
"loss": 0.9258,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.5036111426264699,
"learning_rate": 3.001632138223332e-05,
"loss": 0.9458,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.6088038539922859,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.8702,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 1.1145401433686555,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.8551,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 0.7576986542448317,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.8432,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.7771090755324548,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.8888,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.6785818211297726,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.8633,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.6980804972135541,
"learning_rate": 2.972118904869376e-05,
"loss": 0.8663,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.7246142452234492,
"learning_rate": 2.9671725970312563e-05,
"loss": 0.9107,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.5054501114136327,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.8359,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.6238233972727982,
"learning_rate": 2.95725682050846e-05,
"loss": 0.8575,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.635121864762875,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.8908,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.5322092738484685,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.8109,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.5981479540082791,
"learning_rate": 2.942325848714505e-05,
"loss": 0.8641,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.5490518256457761,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.9051,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.5070135830146454,
"learning_rate": 2.932334127129686e-05,
"loss": 0.9043,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.5967429448746376,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.8283,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.4733216638746647,
"learning_rate": 2.922312585929131e-05,
"loss": 0.841,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.6102151545194943,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.8367,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.4997409565705392,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.8184,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.6030107258602593,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.8266,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.4955303994901396,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.8387,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.52334529251283,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.8854,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.45666091455687774,
"learning_rate": 2.8920722546128863e-05,
"loss": 0.8619,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.512997976787232,
"learning_rate": 2.887006998351372e-05,
"loss": 0.884,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.45979455982134115,
"learning_rate": 2.881934649605535e-05,
"loss": 0.8534,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.5308912444249018,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.8747,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.49251652032110765,
"learning_rate": 2.8717688369507425e-05,
"loss": 0.8562,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.46236690076695725,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.8889,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.4546150215318691,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.8591,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.40768418783990806,
"learning_rate": 2.856467940121025e-05,
"loss": 0.8683,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.45679827687644925,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.8763,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.3892105910408107,
"learning_rate": 2.846233032790149e-05,
"loss": 0.8707,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.43945062191173445,
"learning_rate": 2.84110540896642e-05,
"loss": 0.8275,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.445310247140667,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.8823,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.3985750890024574,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.8809,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.45822696037125565,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.8249,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.5235436384265286,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.8599,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.44589759991521355,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.8848,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.4792334379217217,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.8479,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.46706296110474255,
"learning_rate": 2.805026032173773e-05,
"loss": 0.8731,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.6044137417595182,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.8504,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.4899773160524196,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.8885,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.5142374603104041,
"learning_rate": 2.7894660620408e-05,
"loss": 0.8736,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.42349030908397356,
"learning_rate": 2.784266724883173e-05,
"loss": 0.8357,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.49537299327998,
"learning_rate": 2.7790611167497128e-05,
"loss": 0.8526,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.42579271768206883,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.8593,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.45266425652070635,
"learning_rate": 2.768631254100716e-05,
"loss": 0.8953,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.484346201803559,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.847,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.45984772829511955,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.8503,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.4172251761745271,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.8429,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.47363725183604694,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.839,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.49657639554787253,
"learning_rate": 2.742449775134085e-05,
"loss": 0.8968,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.4778147737258869,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.8829,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.4901480004172421,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.8879,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.47953642410729996,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.8587,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.5105764055880445,
"learning_rate": 2.7213974824975785e-05,
"loss": 0.8298,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.46222824532310697,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.8473,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.39922824248058186,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.8627,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.4088680664468038,
"learning_rate": 2.705547548999996e-05,
"loss": 0.8708,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.564375635359649,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.8706,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.5163667548869662,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.825,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.4040502655194167,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.8866,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.47925768097512095,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.8537,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.36715204560568815,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.8858,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.44830340333449403,
"learning_rate": 2.67369650552155e-05,
"loss": 0.8784,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.3998909332096338,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.8958,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.3733146808085352,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.8587,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.3696797323854807,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.8422,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.37915622564341384,
"learning_rate": 2.652354173593614e-05,
"loss": 0.8692,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.41152723752006093,
"learning_rate": 2.647005443331317e-05,
"loss": 0.8813,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.3833833321513675,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.8797,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.4475315087301215,
"learning_rate": 2.636292505301542e-05,
"loss": 0.8761,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.4521256096114029,
"learning_rate": 2.630928383194431e-05,
"loss": 0.8395,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.4850151238969147,
"learning_rate": 2.625559216200781e-05,
"loss": 0.9031,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.3668617902442007,
"learning_rate": 2.620185047252308e-05,
"loss": 0.879,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.485381839863662,
"learning_rate": 2.614805919320723e-05,
"loss": 0.8646,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.4466302360022263,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.8407,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.35826634359085163,
"learning_rate": 2.60403295859298e-05,
"loss": 0.8492,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.49058642815291514,
"learning_rate": 2.5986392119371297e-05,
"loss": 0.8792,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.4244702272416615,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.8746,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.4122450053695481,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.8644,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.4466330711943622,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.8342,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.4606297389036521,
"learning_rate": 2.577016790136453e-05,
"loss": 0.8523,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.43552756884279253,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.899,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.4105756986012028,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.8524,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.45232803241052294,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.8684,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.436402225223323,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.8728,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.4642451330065423,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.8758,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.4471745224844517,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.8942,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.41894842068002774,
"learning_rate": 2.539001591449889e-05,
"loss": 0.8814,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.37363239449303687,
"learning_rate": 2.53355326151989e-05,
"loss": 0.8646,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.4216068126730836,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.8955,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.41397277319467385,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.8605,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.3668385482347605,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.8684,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.43156073331137373,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.8496,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.4784140965116139,
"learning_rate": 2.506248490185627e-05,
"loss": 0.8443,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.4443556329464633,
"learning_rate": 2.500775217241923e-05,
"loss": 0.9258,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.48841728470014717,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.8556,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.49808278218070196,
"learning_rate": 2.4898167025948e-05,
"loss": 0.8817,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.5343411879521545,
"learning_rate": 2.4843315485153703e-05,
"loss": 0.8843,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 1.6968158438715424,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.9017,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.48873835914509894,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.8837,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.40669080587550666,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.8867,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.41028096805541087,
"learning_rate": 2.46235264422994e-05,
"loss": 0.8766,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.45897932641483297,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.8753,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.42915735256607773,
"learning_rate": 2.451340834511592e-05,
"loss": 0.855,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.4248117674660586,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.8875,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.41584647490362914,
"learning_rate": 2.440314589184958e-05,
"loss": 0.8524,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.4272728400798265,
"learning_rate": 2.434796163357541e-05,
"loss": 0.8794,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.4279536964141046,
"learning_rate": 2.429274260911588e-05,
"loss": 0.8841,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.5250292276835243,
"learning_rate": 2.423748926000084e-05,
"loss": 0.8723,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.5446181109375564,
"learning_rate": 2.418220202803459e-05,
"loss": 0.8321,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.4613788235005681,
"learning_rate": 2.412688135529237e-05,
"loss": 0.9177,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.4765463223980114,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.8823,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 3.527909423312136,
"learning_rate": 2.4016141457114372e-05,
"loss": 0.832,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 84.9648969808189,
"learning_rate": 2.396072311715187e-05,
"loss": 0.9345,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 6.008727854358243,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.8598,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 2.4447343298828015,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.9498,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.591372777826392,
"learning_rate": 2.379427985200223e-05,
"loss": 0.8506,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.7804818399802602,
"learning_rate": 2.373873749394972e-05,
"loss": 0.8668,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.990540742500588,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.8923,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.5693613578359268,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.9058,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.8377117919039451,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.902,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.7984335827972527,
"learning_rate": 2.351627355800233e-05,
"loss": 0.8476,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.5076688204393986,
"learning_rate": 2.346058617160256e-05,
"loss": 0.8467,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.7251225475563663,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.928,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.7054368884279382,
"learning_rate": 2.334912883203469e-05,
"loss": 0.8714,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.4529534435935972,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.9169,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.8581411261637781,
"learning_rate": 2.323756437450136e-05,
"loss": 0.8406,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.6169773772785235,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.8594,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.5825743818870219,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.8721,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.7581524891351568,
"learning_rate": 2.307002464849052e-05,
"loss": 0.8323,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.4830380672145711,
"learning_rate": 2.30141283818846e-05,
"loss": 0.8707,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.5586238996404341,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.9235,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.5893506951963766,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.862,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.3862039250581151,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.8807,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.5166695710775248,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.8859,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.4739674401475522,
"learning_rate": 2.273429448107257e-05,
"loss": 0.8758,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.4079727614686886,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.8267,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.5243545710649422,
"learning_rate": 2.262220474347102e-05,
"loss": 0.8437,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.3758299999169489,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.8783,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.5613927450881825,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.8651,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.41081160229560254,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.8682,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.4043904742795726,
"learning_rate": 2.2397777251561302e-05,
"loss": 0.8535,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.46554003022126544,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.9062,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.370596537667868,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.8501,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.4229229628457843,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.8742,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.4227112134875075,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.9153,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.5798778816263762,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.8743,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.3854394951917942,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.8544,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.45520769593855775,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.8625,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.3440776185544304,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.8716,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.3846306129608798,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.8659,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.44582597749662506,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.8501,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.4407268608350623,
"learning_rate": 2.17791064266051e-05,
"loss": 0.8931,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.3963227716604555,
"learning_rate": 2.172276926861741e-05,
"loss": 0.8603,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.4560411934683307,
"learning_rate": 2.166641833541255e-05,
"loss": 0.8736,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.4693813334626481,
"learning_rate": 2.161005407757109e-05,
"loss": 0.883,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.42496220918147676,
"learning_rate": 2.155367694578013e-05,
"loss": 0.8574,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.4253369504467861,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.8837,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.37882949203406296,
"learning_rate": 2.1440885863609208e-05,
"loss": 0.874,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.45822183883345413,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.9086,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.3996532439303831,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.8303,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.46882932376780656,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.8563,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.4314597200251593,
"learning_rate": 2.1215169053088825e-05,
"loss": 0.8288,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.4537339496022339,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.8623,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.4503875665094128,
"learning_rate": 2.110225054402733e-05,
"loss": 0.8622,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.4200962613164072,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.8409,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.4694559981803835,
"learning_rate": 2.09892967807724e-05,
"loss": 0.8503,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.3874539946732516,
"learning_rate": 2.093280780774835e-05,
"loss": 0.8739,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.43727445735616366,
"learning_rate": 2.087631137601793e-05,
"loss": 0.8237,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.4534851933239706,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.8627,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.44433954624212074,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.8871,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.43062516556971403,
"learning_rate": 2.070678184630458e-05,
"loss": 0.9148,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.43740548638086324,
"learning_rate": 2.06502600977305e-05,
"loss": 0.8359,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.8398447843125578,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.9215,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.4365046139934525,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.8594,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.4389118929598701,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.8429,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.4214807569808412,
"learning_rate": 2.042412562887594e-05,
"loss": 0.8643,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.4410663862055793,
"learning_rate": 2.0367582403231033e-05,
"loss": 0.858,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.3926066235959198,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.8664,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.5200296848632224,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.8764,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.5438650683504072,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.8846,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.4089375396444277,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.8527,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.5303610962532241,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.8911,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.4426269954125942,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.8987,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.47105177163707046,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.8589,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.5296391254481522,
"learning_rate": 1.99151687695709e-05,
"loss": 0.8553,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.3498200202951205,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.8406,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.5589477083946122,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.8453,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.445253696048184,
"learning_rate": 1.974551241345447e-05,
"loss": 0.9027,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.43555985518476126,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.8991,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.47824435189976156,
"learning_rate": 1.963241759676897e-05,
"loss": 0.8323,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.3717219657854215,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.8472,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.5204535007119139,
"learning_rate": 1.951933453677638e-05,
"loss": 0.8797,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.4463766441635691,
"learning_rate": 1.9462798545816932e-05,
"loss": 0.8741,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.3874614899283417,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.8496,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.3285144539881533,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.8968,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.43185901989211806,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.8782,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.3929147441719251,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.8814,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.3993424089428593,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.8977,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.4006789800561976,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.8827,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.393922807740093,
"learning_rate": 1.906719219225165e-05,
"loss": 0.8426,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.348839341884872,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.8419,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.48457953832776407,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.869,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.400620389301169,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.8986,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.4514800961869469,
"learning_rate": 1.884128556891707e-05,
"loss": 0.8665,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.3455311978828433,
"learning_rate": 1.878483094691119e-05,
"loss": 0.8819,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.4360971100944007,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.8779,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.34975995913313396,
"learning_rate": 1.867195130360947e-05,
"loss": 0.8622,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.33569848417300807,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.8768,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.46331967436839405,
"learning_rate": 1.8559114136390795e-05,
"loss": 0.8689,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.31827351026534395,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.8733,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.47076394860603665,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.8583,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.42908098939197387,
"learning_rate": 1.838994592242891e-05,
"loss": 0.8627,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.3810831668752224,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.8576,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.3836440529357289,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.8598,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.47322243767552596,
"learning_rate": 1.82208935733949e-05,
"loss": 0.8891,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.37235675422691333,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.8719,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.32382002926157305,
"learning_rate": 1.810826238483892e-05,
"loss": 0.8782,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.3742658890511117,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.8793,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.30942812049633706,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.881,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.36346861833341,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.8857,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.34577334504830515,
"learning_rate": 1.788318512320838e-05,
"loss": 0.8968,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.3365388529020734,
"learning_rate": 1.782695699829051e-05,
"loss": 0.9195,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.3460855253287032,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.8484,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.3516278818691024,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.8439,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.3195079802798312,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.8468,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.3731348423982012,
"learning_rate": 1.76022227484387e-05,
"loss": 0.8475,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.3434348763357526,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.8921,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.38843466781072433,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.8675,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.3508577322960591,
"learning_rate": 1.743387180005809e-05,
"loss": 0.8539,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.3539583957643248,
"learning_rate": 1.737779525652899e-05,
"loss": 0.864,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.36404187223169987,
"learning_rate": 1.732173968007914e-05,
"loss": 0.8604,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.4012205371817755,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.8611,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.4065347452508291,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.8515,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.4107283529573038,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.8803,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.4338732393531828,
"learning_rate": 1.709773600686015e-05,
"loss": 0.889,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.37274938569488386,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.8765,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.3386102731142736,
"learning_rate": 1.69858716181154e-05,
"loss": 0.8734,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.4035251284136471,
"learning_rate": 1.692997535150948e-05,
"loss": 0.8913,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.3303630831084321,
"learning_rate": 1.687410363273884e-05,
"loss": 0.8483,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.38965993467195464,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.8411,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.49055156185353976,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.8858,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.35147157124968775,
"learning_rate": 1.67066402299235e-05,
"loss": 0.8321,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.4460955241441598,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.8956,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.36407667610113204,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.8675,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.30675296121247536,
"learning_rate": 1.653941382839745e-05,
"loss": 0.8594,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.4336983130748123,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.8157,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.3956534524960194,
"learning_rate": 1.642806717162757e-05,
"loss": 0.8153,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.49394415648960627,
"learning_rate": 1.637243646233718e-05,
"loss": 0.8,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.4747012466731327,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.8116,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.4722046939711531,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.8102,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.46716079893906093,
"learning_rate": 1.620572014799777e-05,
"loss": 0.8322,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.4729447213812195,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.8092,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.4437121783571126,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.8002,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.4704631365527792,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.7947,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.4804119994978665,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.8342,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.4527857394117859,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.7762,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.4290690968368398,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.8193,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.42409935837434015,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.7825,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.3882435815161408,
"learning_rate": 1.576251073999917e-05,
"loss": 0.7938,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.39086224080788556,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.8113,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.37018627344222277,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.8152,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.3519393604172108,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.8073,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.3957022849995132,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.7902,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.3668192492727328,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.8248,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.37521432983242586,
"learning_rate": 1.543151434154755e-05,
"loss": 0.8179,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.3737420857033,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.82,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.3035518377792026,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.7949,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.35819536834831356,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.8032,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.32741355551315066,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.789,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.34880348659414284,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.8267,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.39738418601050884,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.8586,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.3570979884541014,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.7544,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.35289345113083787,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.8388,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.8118646593691414,
"learning_rate": 1.493751509814374e-05,
"loss": 0.8007,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.38511878113774095,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.8115,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.8103989162155566,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.8336,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.39858715740511896,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.7876,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.3200225689706896,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.8399,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.32128036505625873,
"learning_rate": 1.466446738480111e-05,
"loss": 0.7795,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.3307397177617547,
"learning_rate": 1.460998408550112e-05,
"loss": 0.7907,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.3211169599210664,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.8458,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.30742913711965814,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.8405,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.2967776253669858,
"learning_rate": 1.444679451900437e-05,
"loss": 0.8059,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.3382486143098926,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.8006,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.31180519495122955,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.8155,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.2998122721080791,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.734,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.3172581605271774,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.8192,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.5089219937406695,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.8325,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.38648890769783845,
"learning_rate": 1.412162598317602e-05,
"loss": 0.7796,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.3056055701303507,
"learning_rate": 1.406759321421907e-05,
"loss": 0.8062,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.36179235985766917,
"learning_rate": 1.401360788062871e-05,
"loss": 0.7867,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.3059750882730509,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.8354,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.3781098559316415,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.7811,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.31472395803129444,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.8167,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.32337480206342756,
"learning_rate": 1.379814952747693e-05,
"loss": 0.8344,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.3577556537353298,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.8125,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.3339473512085623,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.7711,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.325095202174718,
"learning_rate": 1.363707494698459e-05,
"loss": 0.7977,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.2999223033098914,
"learning_rate": 1.358348460369265e-05,
"loss": 0.8102,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.36049427339504575,
"learning_rate": 1.352994556668684e-05,
"loss": 0.7873,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.35414252419172726,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.788,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.32838872902539074,
"learning_rate": 1.342302312350678e-05,
"loss": 0.7775,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.31505081832142745,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.7972,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.36871148242433627,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.8012,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.2969825806372329,
"learning_rate": 1.326303494478451e-05,
"loss": 0.8207,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.3545659280273189,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.8447,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.36882643917750196,
"learning_rate": 1.315664479122898e-05,
"loss": 0.845,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.3396436601778794,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.8427,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.3621830354801158,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.8145,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.4326998623468316,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.8082,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.3112656820687157,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.8442,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.3699880719608834,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.8215,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.2933490803641997,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.8141,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.3427809826766174,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.7992,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.32595850772822976,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.7745,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.31414802282308296,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.8042,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.3271963217383293,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.8159,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.34561636028904963,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.8354,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.298728889459734,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.7861,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.34511646751062736,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.7943,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.3216497256867882,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.8218,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.3159738397852705,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.8253,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.3499769842543922,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.7755,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.28544411832268296,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.812,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.35150542056748596,
"learning_rate": 1.220938883250288e-05,
"loss": 0.7842,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.30116857098501304,
"learning_rate": 1.215733275116827e-05,
"loss": 0.8071,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.3140554543020025,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.7947,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.3395569700878606,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.8313,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.27088614793369004,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.8272,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.3061791831222483,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.8567,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.3117985519653742,
"learning_rate": 1.189800129803139e-05,
"loss": 0.8579,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.30195559899769303,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.801,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.28410807483076156,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.8134,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.29830505174833716,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.8147,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.3112880822057123,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.8227,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.514730683259108,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.8257,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.28225108067872073,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.8049,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.3301223493807353,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.7988,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.3293692214878606,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.8103,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.2834770849912893,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.8089,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.3225035687597856,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.7664,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.285549870011786,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.7795,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.27909045932538434,
"learning_rate": 1.128231163049258e-05,
"loss": 0.842,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.3046728741875823,
"learning_rate": 1.123144751066259e-05,
"loss": 0.7915,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.32132977817242153,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.7919,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.3200772117196991,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.7846,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.3382748456122188,
"learning_rate": 1.107927745387114e-05,
"loss": 0.8032,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.34901667041706114,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.7989,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.32279908598920387,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.7983,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.31703853895023,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.8003,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.29173132618197073,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.8501,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.3103799827903322,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.8457,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.2994948947212055,
"learning_rate": 1.077687414070869e-05,
"loss": 0.8131,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.2886793937516848,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.8623,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.2975745378084057,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.8364,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.31162600199018203,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.8309,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.28037439379877677,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.8165,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.2679842967927236,
"learning_rate": 1.052689572753247e-05,
"loss": 0.7947,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.2940815055690916,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.7839,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.29128926569047847,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.8148,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.2577209597422716,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.8235,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.28319802599822175,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.7916,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.2951030412795322,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.7933,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.27298612116821386,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.7554,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.282620740204104,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.842,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.27760397170946866,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.7699,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.27845143598638566,
"learning_rate": 1.008173988718005e-05,
"loss": 0.7758,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.27324279807432245,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.8004,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.2910496076802125,
"learning_rate": 9.983678617766677e-06,
"loss": 0.8147,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.2743786363856544,
"learning_rate": 9.934767922553788e-06,
"loss": 0.7974,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.30117418508730726,
"learning_rate": 9.885937708668371e-06,
"loss": 0.8213,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.29456167680769263,
"learning_rate": 9.837188366555524e-06,
"loss": 0.7956,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.29396635575922664,
"learning_rate": 9.788520286013691e-06,
"loss": 0.8057,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.3086011460193805,
"learning_rate": 9.739933856191537e-06,
"loss": 0.7804,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.2733101950345138,
"learning_rate": 9.691429465584866e-06,
"loss": 0.8447,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.33904757216996023,
"learning_rate": 9.643007502033492e-06,
"loss": 0.8013,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.3054391337887812,
"learning_rate": 9.594668352718152e-06,
"loss": 0.7743,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.274452493477198,
"learning_rate": 9.546412404157383e-06,
"loss": 0.7979,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.2837788485607618,
"learning_rate": 9.498240042204471e-06,
"loss": 0.81,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.287440819604087,
"learning_rate": 9.450151652044339e-06,
"loss": 0.7765,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.3178015557494268,
"learning_rate": 9.402147618190472e-06,
"loss": 0.7807,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.2760910566363394,
"learning_rate": 9.354228324481841e-06,
"loss": 0.8065,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.31798016003869767,
"learning_rate": 9.306394154079852e-06,
"loss": 0.7949,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.27069413839887674,
"learning_rate": 9.258645489465254e-06,
"loss": 0.8048,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.29089483087777446,
"learning_rate": 9.21098271243509e-06,
"loss": 0.8094,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.33374990201538673,
"learning_rate": 9.163406204099659e-06,
"loss": 0.8229,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.2756856546511513,
"learning_rate": 9.115916344879452e-06,
"loss": 0.8126,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.3070543890062776,
"learning_rate": 9.068513514502126e-06,
"loss": 0.7974,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.28863993078018446,
"learning_rate": 9.021198091999439e-06,
"loss": 0.8226,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.2932028249302029,
"learning_rate": 8.973970455704251e-06,
"loss": 0.827,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.26938711717068625,
"learning_rate": 8.926830983247495e-06,
"loss": 0.792,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.31654618888822855,
"learning_rate": 8.879780051555116e-06,
"loss": 0.809,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.2818313411599093,
"learning_rate": 8.83281803684512e-06,
"loss": 0.8442,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.29245470641419863,
"learning_rate": 8.785945314624532e-06,
"loss": 0.8182,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.253818140722221,
"learning_rate": 8.739162259686385e-06,
"loss": 0.8294,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.27723647702268384,
"learning_rate": 8.692469246106734e-06,
"loss": 0.7755,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.27136263240671926,
"learning_rate": 8.645866647241677e-06,
"loss": 0.8076,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.43808075633764515,
"learning_rate": 8.599354835724348e-06,
"loss": 0.8909,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.2774419898338891,
"learning_rate": 8.552934183461958e-06,
"loss": 0.8135,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.29382593624255215,
"learning_rate": 8.506605061632805e-06,
"loss": 0.8326,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.272808468475545,
"learning_rate": 8.460367840683308e-06,
"loss": 0.7767,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.32521980164174086,
"learning_rate": 8.414222890325065e-06,
"loss": 0.8015,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.26066981754603463,
"learning_rate": 8.368170579531855e-06,
"loss": 0.8085,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.2751503361723927,
"learning_rate": 8.322211276536733e-06,
"loss": 0.8093,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.3246446381623262,
"learning_rate": 8.27634534882907e-06,
"loss": 0.7899,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.27881260081226045,
"learning_rate": 8.230573163151585e-06,
"loss": 0.8011,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.3047124828820495,
"learning_rate": 8.184895085497466e-06,
"loss": 0.7715,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.26603976687709646,
"learning_rate": 8.139311481107401e-06,
"loss": 0.8209,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.30711773504291623,
"learning_rate": 8.093822714466686e-06,
"loss": 0.8072,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.3035363659761062,
"learning_rate": 8.048429149302271e-06,
"loss": 0.7936,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.2915375790048304,
"learning_rate": 8.003131148579902e-06,
"loss": 0.8006,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.32992864568285435,
"learning_rate": 7.957929074501182e-06,
"loss": 0.814,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.2957682659796153,
"learning_rate": 7.912823288500695e-06,
"loss": 0.8239,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.3277878379742811,
"learning_rate": 7.867814151243102e-06,
"loss": 0.7708,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.34381667551087824,
"learning_rate": 7.822902022620267e-06,
"loss": 0.7873,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.2937327889669594,
"learning_rate": 7.778087261748377e-06,
"loss": 0.8298,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.3252368028571613,
"learning_rate": 7.733370226965056e-06,
"loss": 0.7936,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.3387086227332782,
"learning_rate": 7.688751275826527e-06,
"loss": 0.8075,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.28822942735086554,
"learning_rate": 7.644230765104735e-06,
"loss": 0.7794,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.34379222396999976,
"learning_rate": 7.599809050784502e-06,
"loss": 0.8663,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.35601955811665675,
"learning_rate": 7.555486488060661e-06,
"loss": 0.8304,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.29759406129944616,
"learning_rate": 7.511263431335249e-06,
"loss": 0.7791,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.32320981597311116,
"learning_rate": 7.467140234214656e-06,
"loss": 0.8207,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.3089039547429421,
"learning_rate": 7.423117249506777e-06,
"loss": 0.7844,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.3126499324768706,
"learning_rate": 7.379194829218231e-06,
"loss": 0.7943,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.27796482944656853,
"learning_rate": 7.335373324551518e-06,
"loss": 0.7938,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.3062478054574462,
"learning_rate": 7.291653085902224e-06,
"loss": 0.8321,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.3658596596230191,
"learning_rate": 7.248034462856213e-06,
"loss": 0.8013,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.2795380758124018,
"learning_rate": 7.20451780418683e-06,
"loss": 0.8376,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.2943149540663175,
"learning_rate": 7.161103457852128e-06,
"loss": 0.8388,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.3190619269679789,
"learning_rate": 7.117791770992046e-06,
"loss": 0.8198,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.34387802796928374,
"learning_rate": 7.074583089925688e-06,
"loss": 0.8116,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.28342993817129153,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.764,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.32472794124752347,
"learning_rate": 6.988476126329604e-06,
"loss": 0.785,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.29178892135717394,
"learning_rate": 6.945578532308857e-06,
"loss": 0.8353,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.26193616107454065,
"learning_rate": 6.902785321094301e-06,
"loss": 0.8384,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.2545739033224156,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.8011,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.3068517441417269,
"learning_rate": 6.817513414939892e-06,
"loss": 0.8603,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.25841524028915336,
"learning_rate": 6.775035401831924e-06,
"loss": 0.7974,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.2647024613781748,
"learning_rate": 6.732663135188473e-06,
"loss": 0.7915,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.2680615153830016,
"learning_rate": 6.690396953817055e-06,
"loss": 0.8064,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.27713447562814236,
"learning_rate": 6.648237195676927e-06,
"loss": 0.7849,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.2850822339574058,
"learning_rate": 6.606184197876395e-06,
"loss": 0.781,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.28654970048902473,
"learning_rate": 6.564238296670109e-06,
"loss": 0.8128,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.2554094190211513,
"learning_rate": 6.52239982745638e-06,
"loss": 0.8364,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.3067878700852697,
"learning_rate": 6.480669124774484e-06,
"loss": 0.8141,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.2876227733117812,
"learning_rate": 6.439046522302013e-06,
"loss": 0.815,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.26026314474018175,
"learning_rate": 6.397532352852191e-06,
"loss": 0.7848,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.3005193551116843,
"learning_rate": 6.356126948371202e-06,
"loss": 0.7765,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.2737827804689421,
"learning_rate": 6.314830639935559e-06,
"loss": 0.7978,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.28400076765426874,
"learning_rate": 6.273643757749448e-06,
"loss": 0.8012,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.2914811698360687,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.801,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.29029000099632957,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.773,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.27571209693345666,
"learning_rate": 6.150742957589784e-06,
"loss": 0.797,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.28551875422837386,
"learning_rate": 6.109997064904774e-06,
"loss": 0.8238,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.2740726837116015,
"learning_rate": 6.069362236313092e-06,
"loss": 0.8062,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.2857755867871268,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.7509,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.27840865624350686,
"learning_rate": 5.988427070179108e-06,
"loss": 0.8155,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.2935281468526863,
"learning_rate": 5.948127379792234e-06,
"loss": 0.7968,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.25658711669805173,
"learning_rate": 5.90794004780439e-06,
"loss": 0.8183,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.24817304130875997,
"learning_rate": 5.867865395552405e-06,
"loss": 0.7981,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.2655245195319302,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.8234,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.273495060226351,
"learning_rate": 5.788055411095892e-06,
"loss": 0.8003,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.26067212180981225,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.7913,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.25918313594309683,
"learning_rate": 5.708699979051613e-06,
"loss": 0.8175,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.2823534469944724,
"learning_rate": 5.66919351390751e-06,
"loss": 0.8137,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.2468744906895734,
"learning_rate": 5.629801637510188e-06,
"loss": 0.7901,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.28630569938647443,
"learning_rate": 5.590524664836043e-06,
"loss": 0.7806,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.26827562582979886,
"learning_rate": 5.551362909942708e-06,
"loss": 0.7822,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.6531887686073221,
"learning_rate": 5.51231668596653e-06,
"loss": 0.7924,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.29857257734464976,
"learning_rate": 5.47338630512007e-06,
"loss": 0.8121,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.27346479035193094,
"learning_rate": 5.434572078689617e-06,
"loss": 0.7966,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.24566072890458684,
"learning_rate": 5.395874317032694e-06,
"loss": 0.8226,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.31774173002832723,
"learning_rate": 5.357293329575573e-06,
"loss": 0.754,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.28276694092433147,
"learning_rate": 5.318829424810797e-06,
"loss": 0.7675,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.2488670640807651,
"learning_rate": 5.280482910294728e-06,
"loss": 0.8274,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.25186866932654106,
"learning_rate": 5.242254092645078e-06,
"loss": 0.8309,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.29296555133607916,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.7905,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.2414044127178353,
"learning_rate": 5.166150769707892e-06,
"loss": 0.8335,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.23189586111646932,
"learning_rate": 5.128276872940503e-06,
"loss": 0.809,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.2675067359730285,
"learning_rate": 5.090521890074953e-06,
"loss": 0.8158,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.27435139430460725,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.7666,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.25283759468342715,
"learning_rate": 5.015369872647442e-06,
"loss": 0.8197,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.26812949677459386,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.803,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.2512862487944523,
"learning_rate": 4.940697121074675e-06,
"loss": 0.8138,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.25900366683738785,
"learning_rate": 4.903541216934864e-06,
"loss": 0.8119,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.2706355432230592,
"learning_rate": 4.8665060236772e-06,
"loss": 0.8279,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.27435359548926397,
"learning_rate": 4.829591837434102e-06,
"loss": 0.7892,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.2463824849771203,
"learning_rate": 4.792798953370424e-06,
"loss": 0.8559,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.28137155106345146,
"learning_rate": 4.756127665681083e-06,
"loss": 0.7708,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.2562380692844913,
"learning_rate": 4.719578267588725e-06,
"loss": 0.7865,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.23790026219471172,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.7853,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.3528386554797099,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.773,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.2880943953740471,
"learning_rate": 4.610664328486485e-06,
"loss": 0.7699,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.2542239910668536,
"learning_rate": 4.5746054014809e-06,
"loss": 0.7934,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.33410598020066956,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.7603,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.3035043599845133,
"learning_rate": 4.502857857942204e-06,
"loss": 0.8086,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.27676070788995316,
"learning_rate": 4.467169815100549e-06,
"loss": 0.7819,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.3196948111545434,
"learning_rate": 4.431605972355137e-06,
"loss": 0.7796,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.32493801707434405,
"learning_rate": 4.396166614073516e-06,
"loss": 0.7735,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.2849926571692678,
"learning_rate": 4.360852023627846e-06,
"loss": 0.7421,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.29253496412695135,
"learning_rate": 4.325662483392656e-06,
"loss": 0.7239,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.29671963374119925,
"learning_rate": 4.29059827474257e-06,
"loss": 0.7446,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.31537270168502485,
"learning_rate": 4.255659678050063e-06,
"loss": 0.7655,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.3074293265257969,
"learning_rate": 4.22084697268323e-06,
"loss": 0.7782,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.2695583524190623,
"learning_rate": 4.186160437003535e-06,
"loss": 0.7528,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.2773912566502896,
"learning_rate": 4.151600348363582e-06,
"loss": 0.7662,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.300961168531863,
"learning_rate": 4.11716698310493e-06,
"loss": 0.7734,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.30789131214076243,
"learning_rate": 4.082860616555848e-06,
"loss": 0.8157,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.2596455433399703,
"learning_rate": 4.048681523029125e-06,
"loss": 0.7544,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.264656440091299,
"learning_rate": 4.014629975819882e-06,
"loss": 0.7264,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.31308721873750106,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.7439,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.273535664137524,
"learning_rate": 3.94691060843287e-06,
"loss": 0.7607,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.23500590052953546,
"learning_rate": 3.913243329737353e-06,
"loss": 0.7127,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.23362110455120244,
"learning_rate": 3.879704680319503e-06,
"loss": 0.7326,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.257328373182299,
"learning_rate": 3.846294928353471e-06,
"loss": 0.8024,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.26109983051657637,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.7757,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.23993526358767162,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.7318,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.2559807303797682,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.729,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.268318947198849,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.749,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.25757238870576843,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.7236,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.24315280053916125,
"learning_rate": 3.648558150665886e-06,
"loss": 0.8013,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.24894614061794978,
"learning_rate": 3.616058102895075e-06,
"loss": 0.7584,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.28560150609408225,
"learning_rate": 3.583689060687079e-06,
"loss": 0.7981,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.2558629452712428,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.7852,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.2358469899270662,
"learning_rate": 3.51934502719796e-06,
"loss": 0.7647,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.22775523450771237,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.7724,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.23264424725572666,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.7836,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.22966909929931387,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.7962,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.24062105145904447,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.7846,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.2327693706707547,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.7417,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.23552564915491453,
"learning_rate": 3.32948376099375e-06,
"loss": 0.7713,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.2426412150375587,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.771,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.22688159523270066,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.7454,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.22141315226471656,
"learning_rate": 3.236349233544438e-06,
"loss": 0.7599,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.2276589924475581,
"learning_rate": 3.205572144561604e-06,
"loss": 0.7566,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.22254590274235478,
"learning_rate": 3.174929343376374e-06,
"loss": 0.7439,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.23849436695010276,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.7497,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.22961439523462168,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.7496,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.22104754979891833,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.7517,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.22256580095934148,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.7832,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.226277523091713,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.7812,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.24082860271088144,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.7674,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.23713066718197856,
"learning_rate": 2.964210239224696e-06,
"loss": 0.792,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.23442006340992685,
"learning_rate": 2.934650470499869e-06,
"loss": 0.8204,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.22535308170879312,
"learning_rate": 2.905227155855128e-06,
"loss": 0.7534,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.25375625056114454,
"learning_rate": 2.875940530558519e-06,
"loss": 0.7324,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.26429273780250334,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.7706,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.23270584378685108,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.7615,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.23747455728916228,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.7852,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.224669183682322,
"learning_rate": 2.760165589921184e-06,
"loss": 0.7623,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.251360376902023,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.7386,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.22218208652187155,
"learning_rate": 2.703104292151164e-06,
"loss": 0.8058,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.22621120711263176,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.7593,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.21856706542483917,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.7618,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.22707020973604847,
"learning_rate": 2.618550200289303e-06,
"loss": 0.8065,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.2075542182238665,
"learning_rate": 2.59064316684412e-06,
"loss": 0.7054,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.22104570156191547,
"learning_rate": 2.562875338152264e-06,
"loss": 0.7616,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.2291227443166333,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.7448,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.2235119947455408,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.7347,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.23058393592269638,
"learning_rate": 2.480409295328803e-06,
"loss": 0.7283,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.2218690549301708,
"learning_rate": 2.45320049480108e-06,
"loss": 0.8257,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.23420069524974224,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.7569,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.2294422940360824,
"learning_rate": 2.399204021408421e-06,
"loss": 0.7784,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.22206138223470856,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.7913,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.2271816271521197,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.757,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.22863865686034238,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.7666,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.23479329318541448,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.7712,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.21701287663491525,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.7777,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.21741315657660187,
"learning_rate": 2.24059906434998e-06,
"loss": 0.7568,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.23250193407806102,
"learning_rate": 2.214660694699091e-06,
"loss": 0.7446,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.22190849925679584,
"learning_rate": 2.188864536147355e-06,
"loss": 0.7853,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.22212035203910915,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.7942,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.23005701496917166,
"learning_rate": 2.137699676264171e-06,
"loss": 0.7523,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.2241601531500833,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.7639,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.23448212554772785,
"learning_rate": 2.08710612114857e-06,
"loss": 0.7338,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.21767378148948394,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.7448,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.24275182159239891,
"learning_rate": 2.0370854889762e-06,
"loss": 0.7598,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.24580754546896819,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.7814,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.2192619978078122,
"learning_rate": 1.987639379598427e-06,
"loss": 0.789,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.21251901046748736,
"learning_rate": 1.963132265887919e-06,
"loss": 0.8031,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.22260692413583158,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.7559,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.22434199622803985,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.7589,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.23493290688117596,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.7887,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.22648430613176157,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.83,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.22926528212567324,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.7597,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.2308014291608183,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.761,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.23028520664701946,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.7918,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.24021330208724154,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.7733,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.21855177527455663,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.7743,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.21715327955288594,
"learning_rate": 1.726025117089718e-06,
"loss": 0.7421,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.24474736226603097,
"learning_rate": 1.70311497206598e-06,
"loss": 0.7789,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.2168517795738684,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.7859,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.2214687036899852,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.7706,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.22098484112468086,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.7683,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.22115797170266013,
"learning_rate": 1.612939220355272e-06,
"loss": 0.7598,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.22523992772471793,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.7814,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.22467955300478742,
"learning_rate": 1.568732761074121e-06,
"loss": 0.7469,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.21442496552225274,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.7664,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.23698058355697918,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.7444,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.209287841961101,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.7607,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.21407630422308385,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.773,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.22602422113118412,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.745,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.22073649129988615,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.775,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.2039528147985424,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.7382,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.22595113254082527,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.7486,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.2185499832161193,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.7604,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.21661616123810445,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.8132,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.23346127442448894,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.7497,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.23343936697519763,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.7758,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.2228623825871588,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.7897,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.221110549856945,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.7675,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.2220738127388567,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.744,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.22341376885676986,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.7704,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.21307907486664873,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.7624,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.20014419432432712,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.7791,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.21412403851538817,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.7908,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.21809889255923365,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.755,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.2297226018799783,
"learning_rate": 1.140415161609505e-06,
"loss": 0.7446,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.2230540987850742,
"learning_rate": 1.121666393356604e-06,
"loss": 0.7675,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.21772350083638903,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.7734,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.2127229005332477,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.7468,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.21009146195311207,
"learning_rate": 1.066326386135097e-06,
"loss": 0.7657,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.23365326352802537,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.7535,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.21821776943101456,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.7708,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.2187740466464381,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.7352,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.216176381169966,
"learning_rate": 9.9465987684501e-07,
"loss": 0.756,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.21264832683538226,
"learning_rate": 9.771228098856534e-07,
"loss": 0.7708,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.2035101681408127,
"learning_rate": 9.597378493454013e-07,
"loss": 0.7626,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.2005393497739708,
"learning_rate": 9.425051342339353e-07,
"loss": 0.7547,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.22970254806405646,
"learning_rate": 9.254248023435864e-07,
"loss": 0.7961,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.23627707881161109,
"learning_rate": 9.084969902482288e-07,
"loss": 0.7494,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.2106254113240315,
"learning_rate": 8.917218333021993e-07,
"loss": 0.7941,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.23309484616096285,
"learning_rate": 8.750994656392109e-07,
"loss": 0.7655,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.2110955224845181,
"learning_rate": 8.58630020171265e-07,
"loss": 0.7429,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.20481015804759214,
"learning_rate": 8.423136285876099e-07,
"loss": 0.7484,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.21471192173588538,
"learning_rate": 8.26150421353682e-07,
"loss": 0.7456,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.22161536823313857,
"learning_rate": 8.101405277100549e-07,
"loss": 0.7618,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.21127797436314774,
"learning_rate": 7.942840756714077e-07,
"loss": 0.8241,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.21865161308812633,
"learning_rate": 7.785811920255093e-07,
"loss": 0.7423,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.2164359431887903,
"learning_rate": 7.630320023322024e-07,
"loss": 0.7748,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.21795440396219637,
"learning_rate": 7.476366309223903e-07,
"loss": 0.7903,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.2099612677299921,
"learning_rate": 7.323952008970537e-07,
"loss": 0.7524,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.21291147909330635,
"learning_rate": 7.173078341262618e-07,
"loss": 0.7646,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.2105336630872568,
"learning_rate": 7.023746512481989e-07,
"loss": 0.8111,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.20240482336220475,
"learning_rate": 6.875957716681902e-07,
"loss": 0.7504,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.20299722216774324,
"learning_rate": 6.729713135577687e-07,
"loss": 0.7699,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.21475496026037175,
"learning_rate": 6.585013938537143e-07,
"loss": 0.805,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.19823201553083122,
"learning_rate": 6.441861282571115e-07,
"loss": 0.7762,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.2769746936303445,
"learning_rate": 6.300256312324471e-07,
"loss": 0.7684,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.22063548850811254,
"learning_rate": 6.160200160066776e-07,
"loss": 0.7729,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.21046036090139308,
"learning_rate": 6.021693945683327e-07,
"loss": 0.7708,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.2035853469708454,
"learning_rate": 5.884738776666044e-07,
"loss": 0.7664,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.21479696942072585,
"learning_rate": 5.749335748104878e-07,
"loss": 0.7717,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.21873705954742842,
"learning_rate": 5.615485942678778e-07,
"loss": 0.7636,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.19725207972560133,
"learning_rate": 5.483190430647289e-07,
"loss": 0.7454,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.20776695411084328,
"learning_rate": 5.352450269841747e-07,
"loss": 0.7633,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.20845291172382463,
"learning_rate": 5.22326650565701e-07,
"loss": 0.7353,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.22125444777152833,
"learning_rate": 5.095640171043048e-07,
"loss": 0.7783,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.19829260485101222,
"learning_rate": 4.969572286496594e-07,
"loss": 0.7545,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.21286623519185408,
"learning_rate": 4.845063860053079e-07,
"loss": 0.7798,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.20118694228147233,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.7486,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.21378240571066986,
"learning_rate": 4.600729351261857e-07,
"loss": 0.7442,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.21346746352974638,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.789,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.2037426699237324,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.7675,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.20630981940176854,
"learning_rate": 4.245948007320588e-07,
"loss": 0.7742,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.2148145501918614,
"learning_rate": 4.130816799401971e-07,
"loss": 0.7528,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.21125691503047225,
"learning_rate": 4.017251756252827e-07,
"loss": 0.7791,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.20794076599213185,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.7606,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.21274394053186277,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.7731,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.20956753109840612,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.7791,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.2180453193174492,
"learning_rate": 3.578671204615347e-07,
"loss": 0.7625,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.20141084880380766,
"learning_rate": 3.472950355547267e-07,
"loss": 0.7707,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.2113235194605102,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.797,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.21535742901829025,
"learning_rate": 3.266223765343224e-07,
"loss": 0.7983,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.21089038419300465,
"learning_rate": 3.16521967718757e-07,
"loss": 0.7967,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.20168034314361435,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.7668,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.2013048064402292,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.7525,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.21195308750325187,
"learning_rate": 2.871653894244597e-07,
"loss": 0.7735,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.21197318098859574,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.7638,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.21100361731238956,
"learning_rate": 2.683823233205485e-07,
"loss": 0.7643,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.21653082355665865,
"learning_rate": 2.592274132703287e-07,
"loss": 0.7561,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.21741893492813272,
"learning_rate": 2.502303499116754e-07,
"loss": 0.8047,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.20712481204529617,
"learning_rate": 2.413912051848666e-07,
"loss": 0.7728,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.31842896477100047,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.7543,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.21068655063846,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.7407,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.2081437574889192,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.7477,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.21774080531961498,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.7428,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.23224843772076656,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.7532,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.21057968263443644,
"learning_rate": 1.916764975761254e-07,
"loss": 0.7457,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.20385332075397183,
"learning_rate": 1.839446914621923e-07,
"loss": 0.7365,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.20097500556421552,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.7266,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.21587965577841253,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.7654,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.20479238345712386,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.7694,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.21505541585805182,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.7533,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.21323907129107672,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.7403,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.22429320621571175,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.7683,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.2018501404284287,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.7817,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.21121031002995433,
"learning_rate": 1.277993209368744e-07,
"loss": 0.7246,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.20545501918340892,
"learning_rate": 1.214956261808542e-07,
"loss": 0.7818,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.20514527585397851,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.7718,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.2052330557480646,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.7617,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.20608398549533918,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.7569,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.2174762377713197,
"learning_rate": 9.787081204466253e-08,
"loss": 0.769,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.2114754825439525,
"learning_rate": 9.236233586200627e-08,
"loss": 0.7665,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.19757094542388715,
"learning_rate": 8.70130406196279e-08,
"loss": 0.7455,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.22630254853385967,
"learning_rate": 8.182296909035047e-08,
"loss": 0.7653,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.21068459696274558,
"learning_rate": 7.679216277384439e-08,
"loss": 0.7877,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.201201375190972,
"learning_rate": 7.19206618963053e-08,
"loss": 0.7543,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.3736229222807489,
"learning_rate": 6.720850541012347e-08,
"loss": 0.7061,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.20682141500926307,
"learning_rate": 6.265573099357714e-08,
"loss": 0.7355,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.19737775760575357,
"learning_rate": 5.826237505053067e-08,
"loss": 0.7389,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.21243020264977075,
"learning_rate": 5.4028472710137e-08,
"loss": 0.7521,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.22467996141647575,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.7711,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.23728245030318545,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.7798,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.36415477771834737,
"learning_rate": 4.228381947004812e-08,
"loss": 0.782,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.20088628332507866,
"learning_rate": 3.868805732811343e-08,
"loss": 0.7337,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.20153619360495895,
"learning_rate": 3.525190530455636e-08,
"loss": 0.8121,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.2095322920175952,
"learning_rate": 3.197539087475887e-08,
"loss": 0.7366,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.21687243231381007,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.7643,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.2034799967603498,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.7838,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.22784786021530912,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.7679,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.28842838876837756,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.7397,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.20016314736624777,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.7218,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.20578468603137848,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.7547,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.21211574180888526,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.7479,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.20755558154137452,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.7425,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.2015715445573485,
"learning_rate": 9.67435416023843e-09,
"loss": 0.7634,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.206237255839792,
"learning_rate": 7.995445897566178e-09,
"loss": 0.7775,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.20796526681747296,
"learning_rate": 6.476393170578021e-09,
"loss": 0.7727,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.20485668666376902,
"learning_rate": 5.11720812557348e-09,
"loss": 0.7739,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.21199218948013782,
"learning_rate": 3.917901630563492e-09,
"loss": 0.785,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.21258611985348014,
"learning_rate": 2.878483275170485e-09,
"loss": 0.7608,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.20262204210525867,
"learning_rate": 1.998961370555108e-09,
"loss": 0.7359,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.1946759806198077,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.7509,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.19568735700844325,
"learning_rate": 7.196337656112384e-10,
"loss": 0.7932,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.19849594124345216,
"learning_rate": 3.198382947511292e-10,
"loss": 0.7706,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.21714357337660564,
"learning_rate": 7.995973352770137e-11,
"loss": 0.8184,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.2117794745829479,
"learning_rate": 0.0,
"loss": 0.766,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 5.755190914611937e+18,
"train_loss": 0.8852730185396758,
"train_runtime": 28797.5187,
"train_samples_per_second": 5.487,
"train_steps_per_second": 0.043
}
],
"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": 5.755190914611937e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}