neginr's picture
End of training
29e285c verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 500,
"global_step": 1235,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004048582995951417,
"grad_norm": 6.9423529105640265,
"learning_rate": 3.2258064516129035e-07,
"loss": 0.8135,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 6.874176081638632,
"learning_rate": 6.451612903225807e-07,
"loss": 0.8219,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 7.1262821846254525,
"learning_rate": 9.67741935483871e-07,
"loss": 0.8462,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 6.87762250079229,
"learning_rate": 1.2903225806451614e-06,
"loss": 0.8436,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 6.404661928728426,
"learning_rate": 1.6129032258064516e-06,
"loss": 0.7837,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 6.310391518436217,
"learning_rate": 1.935483870967742e-06,
"loss": 0.7932,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 5.087796637320357,
"learning_rate": 2.2580645161290324e-06,
"loss": 0.7653,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 4.882979336830973,
"learning_rate": 2.580645161290323e-06,
"loss": 0.7457,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 2.7986285316821697,
"learning_rate": 2.903225806451613e-06,
"loss": 0.6816,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 2.6889631937722482,
"learning_rate": 3.225806451612903e-06,
"loss": 0.6906,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 2.2987431551599378,
"learning_rate": 3.548387096774194e-06,
"loss": 0.6831,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 3.521503547034523,
"learning_rate": 3.870967741935484e-06,
"loss": 0.6562,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 3.6145118210270244,
"learning_rate": 4.193548387096774e-06,
"loss": 0.6492,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 3.6918446695777765,
"learning_rate": 4.516129032258065e-06,
"loss": 0.6398,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 3.5478354225143094,
"learning_rate": 4.838709677419355e-06,
"loss": 0.6339,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 2.905464917971483,
"learning_rate": 5.161290322580646e-06,
"loss": 0.5645,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 2.814623578844521,
"learning_rate": 5.483870967741935e-06,
"loss": 0.5728,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 2.432004781427902,
"learning_rate": 5.806451612903226e-06,
"loss": 0.5737,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 1.7970225616876496,
"learning_rate": 6.129032258064517e-06,
"loss": 0.5744,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.2578754686616365,
"learning_rate": 6.451612903225806e-06,
"loss": 0.5339,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 1.4969764134083872,
"learning_rate": 6.774193548387097e-06,
"loss": 0.5436,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.5287301750631048,
"learning_rate": 7.096774193548388e-06,
"loss": 0.5167,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.3195660270162832,
"learning_rate": 7.4193548387096784e-06,
"loss": 0.5148,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 0.8910283306193092,
"learning_rate": 7.741935483870968e-06,
"loss": 0.5136,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 0.8258027047577567,
"learning_rate": 8.064516129032258e-06,
"loss": 0.4985,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 0.9824042332170089,
"learning_rate": 8.387096774193549e-06,
"loss": 0.4938,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 0.960991534879826,
"learning_rate": 8.70967741935484e-06,
"loss": 0.4965,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 0.7489501575041454,
"learning_rate": 9.03225806451613e-06,
"loss": 0.4873,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 0.6970476627901383,
"learning_rate": 9.35483870967742e-06,
"loss": 0.4795,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 0.8556671564569807,
"learning_rate": 9.67741935483871e-06,
"loss": 0.4587,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 0.9249533325910695,
"learning_rate": 1e-05,
"loss": 0.4595,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 0.6703689904345914,
"learning_rate": 1.0322580645161291e-05,
"loss": 0.4766,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 0.6388803704276681,
"learning_rate": 1.0645161290322582e-05,
"loss": 0.463,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 0.7399792807480949,
"learning_rate": 1.096774193548387e-05,
"loss": 0.4791,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 0.7155040326551289,
"learning_rate": 1.1290322580645164e-05,
"loss": 0.4613,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.6844259619574156,
"learning_rate": 1.1612903225806453e-05,
"loss": 0.4568,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.6624955067236954,
"learning_rate": 1.1935483870967743e-05,
"loss": 0.4684,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.5953407791819502,
"learning_rate": 1.2258064516129034e-05,
"loss": 0.4305,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.6330257306967051,
"learning_rate": 1.2580645161290324e-05,
"loss": 0.4508,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 0.6618885258878658,
"learning_rate": 1.2903225806451613e-05,
"loss": 0.4433,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.5661746587557192,
"learning_rate": 1.3225806451612903e-05,
"loss": 0.44,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.607568977075495,
"learning_rate": 1.3548387096774194e-05,
"loss": 0.4565,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.5323407560073551,
"learning_rate": 1.3870967741935486e-05,
"loss": 0.4444,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.5891041819755839,
"learning_rate": 1.4193548387096776e-05,
"loss": 0.4596,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.5807717663839018,
"learning_rate": 1.4516129032258066e-05,
"loss": 0.4442,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 0.5802693482508338,
"learning_rate": 1.4838709677419357e-05,
"loss": 0.4561,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.5338274205758565,
"learning_rate": 1.5161290322580646e-05,
"loss": 0.4392,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.5100479056041157,
"learning_rate": 1.5483870967741936e-05,
"loss": 0.4345,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 0.6122075382043705,
"learning_rate": 1.5806451612903226e-05,
"loss": 0.4348,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 0.6351899222081645,
"learning_rate": 1.6129032258064517e-05,
"loss": 0.4296,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 0.5803260526767975,
"learning_rate": 1.6451612903225807e-05,
"loss": 0.4366,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.6108191526209804,
"learning_rate": 1.6774193548387098e-05,
"loss": 0.4192,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 0.5916673672213443,
"learning_rate": 1.7096774193548388e-05,
"loss": 0.4451,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.5612721679972342,
"learning_rate": 1.741935483870968e-05,
"loss": 0.4303,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.6146126049022449,
"learning_rate": 1.774193548387097e-05,
"loss": 0.4297,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 0.5079716247782361,
"learning_rate": 1.806451612903226e-05,
"loss": 0.4213,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.5946270438758549,
"learning_rate": 1.838709677419355e-05,
"loss": 0.4183,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 0.5976514508771535,
"learning_rate": 1.870967741935484e-05,
"loss": 0.4084,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 0.5372546486682969,
"learning_rate": 1.903225806451613e-05,
"loss": 0.4163,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 0.7627328610871474,
"learning_rate": 1.935483870967742e-05,
"loss": 0.4087,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 0.5195797089591795,
"learning_rate": 1.967741935483871e-05,
"loss": 0.4259,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 0.6642849326433669,
"learning_rate": 2e-05,
"loss": 0.4178,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 0.521629851738115,
"learning_rate": 2.0322580645161292e-05,
"loss": 0.4137,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 0.5477851566502988,
"learning_rate": 2.0645161290322582e-05,
"loss": 0.4391,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 0.5961238249678555,
"learning_rate": 2.0967741935483873e-05,
"loss": 0.4221,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 0.5469553192488475,
"learning_rate": 2.1290322580645163e-05,
"loss": 0.4232,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 0.5939741880623763,
"learning_rate": 2.161290322580645e-05,
"loss": 0.4252,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 0.5986765490004146,
"learning_rate": 2.193548387096774e-05,
"loss": 0.4241,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 0.47874498259704157,
"learning_rate": 2.225806451612903e-05,
"loss": 0.3995,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 0.5543738462575027,
"learning_rate": 2.2580645161290328e-05,
"loss": 0.4125,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 0.5633658928636445,
"learning_rate": 2.2903225806451615e-05,
"loss": 0.411,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 0.6181515058576877,
"learning_rate": 2.3225806451612906e-05,
"loss": 0.4139,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 0.5361509083963997,
"learning_rate": 2.3548387096774196e-05,
"loss": 0.4061,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 0.7508299448425834,
"learning_rate": 2.3870967741935486e-05,
"loss": 0.4193,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 0.5438558910411969,
"learning_rate": 2.4193548387096777e-05,
"loss": 0.3867,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 0.738084229781646,
"learning_rate": 2.4516129032258067e-05,
"loss": 0.4141,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 0.5843373813480383,
"learning_rate": 2.4838709677419358e-05,
"loss": 0.4166,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 0.8515237115235735,
"learning_rate": 2.5161290322580648e-05,
"loss": 0.4044,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 0.6873514594574375,
"learning_rate": 2.548387096774194e-05,
"loss": 0.4103,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 0.6852492732273774,
"learning_rate": 2.5806451612903226e-05,
"loss": 0.4093,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 0.8500970736233115,
"learning_rate": 2.6129032258064516e-05,
"loss": 0.4074,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 0.5414173765350795,
"learning_rate": 2.6451612903225806e-05,
"loss": 0.3926,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 0.7247686716762975,
"learning_rate": 2.6774193548387097e-05,
"loss": 0.3933,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 0.593124588453728,
"learning_rate": 2.7096774193548387e-05,
"loss": 0.4123,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 0.8222344568800566,
"learning_rate": 2.7419354838709678e-05,
"loss": 0.4314,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 0.7978692059141226,
"learning_rate": 2.774193548387097e-05,
"loss": 0.419,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 0.7772540138532303,
"learning_rate": 2.8064516129032262e-05,
"loss": 0.4037,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 0.6993864172653979,
"learning_rate": 2.8387096774193552e-05,
"loss": 0.4083,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 0.7066150502511073,
"learning_rate": 2.8709677419354843e-05,
"loss": 0.406,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 0.8598904954298298,
"learning_rate": 2.9032258064516133e-05,
"loss": 0.4046,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 0.7699334578334588,
"learning_rate": 2.9354838709677423e-05,
"loss": 0.4152,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 0.8190482565295801,
"learning_rate": 2.9677419354838714e-05,
"loss": 0.4052,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 0.8190211642019075,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.407,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 0.9013599919853816,
"learning_rate": 3.032258064516129e-05,
"loss": 0.4057,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 0.8132753189807838,
"learning_rate": 3.0645161290322585e-05,
"loss": 0.4036,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 0.6660409669364636,
"learning_rate": 3.096774193548387e-05,
"loss": 0.4175,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 0.9090303850328008,
"learning_rate": 3.1290322580645166e-05,
"loss": 0.4122,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 0.7662720817225777,
"learning_rate": 3.161290322580645e-05,
"loss": 0.4142,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 0.7723881079938062,
"learning_rate": 3.193548387096774e-05,
"loss": 0.3959,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 0.6866132944430656,
"learning_rate": 3.2258064516129034e-05,
"loss": 0.4101,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 0.7022692267459147,
"learning_rate": 3.258064516129033e-05,
"loss": 0.3898,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 0.6551838303549807,
"learning_rate": 3.2903225806451614e-05,
"loss": 0.4133,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 0.6749514803799332,
"learning_rate": 3.322580645161291e-05,
"loss": 0.3976,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 0.6406213671311237,
"learning_rate": 3.3548387096774195e-05,
"loss": 0.4061,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 0.6470453830717576,
"learning_rate": 3.387096774193549e-05,
"loss": 0.4115,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 0.6325648394099841,
"learning_rate": 3.4193548387096776e-05,
"loss": 0.4016,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 0.5356480802929214,
"learning_rate": 3.451612903225807e-05,
"loss": 0.3985,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 0.5634489106546214,
"learning_rate": 3.483870967741936e-05,
"loss": 0.4007,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 0.5330730140378126,
"learning_rate": 3.516129032258065e-05,
"loss": 0.3868,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 0.6246373654686972,
"learning_rate": 3.548387096774194e-05,
"loss": 0.4168,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 0.5805644913095783,
"learning_rate": 3.5806451612903225e-05,
"loss": 0.4016,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 0.7165014836795914,
"learning_rate": 3.612903225806452e-05,
"loss": 0.4043,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 0.8226013313368671,
"learning_rate": 3.6451612903225805e-05,
"loss": 0.407,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 0.5348330753552054,
"learning_rate": 3.67741935483871e-05,
"loss": 0.3923,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 0.8660452912024521,
"learning_rate": 3.7096774193548386e-05,
"loss": 0.4054,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 0.7101976351383762,
"learning_rate": 3.741935483870968e-05,
"loss": 0.3978,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 0.6494467920421529,
"learning_rate": 3.7741935483870974e-05,
"loss": 0.4033,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 0.6517186858466721,
"learning_rate": 3.806451612903226e-05,
"loss": 0.4072,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 0.6584695854085189,
"learning_rate": 3.8387096774193555e-05,
"loss": 0.3972,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 0.5383165020429066,
"learning_rate": 3.870967741935484e-05,
"loss": 0.4045,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 0.7046885175611489,
"learning_rate": 3.9032258064516135e-05,
"loss": 0.4032,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 0.6996163097743429,
"learning_rate": 3.935483870967742e-05,
"loss": 0.3919,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 0.512241831109312,
"learning_rate": 3.9677419354838716e-05,
"loss": 0.389,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 0.604027331657842,
"learning_rate": 4e-05,
"loss": 0.396,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 0.5199660212184166,
"learning_rate": 3.9999920040266474e-05,
"loss": 0.3901,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 0.7646398209731203,
"learning_rate": 3.999968016170525e-05,
"loss": 0.39,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 0.5910973996634473,
"learning_rate": 3.9999280366234394e-05,
"loss": 0.3825,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 0.5964496273039139,
"learning_rate": 3.999872065705065e-05,
"loss": 0.4054,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 0.5725946434226825,
"learning_rate": 3.999800103862945e-05,
"loss": 0.397,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 0.6716353201840248,
"learning_rate": 3.9997121516724834e-05,
"loss": 0.407,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 0.6596729186939215,
"learning_rate": 3.999608209836944e-05,
"loss": 0.4046,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 0.5921079029529798,
"learning_rate": 3.999488279187443e-05,
"loss": 0.4001,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 0.5130118729316494,
"learning_rate": 3.9993523606829425e-05,
"loss": 0.3914,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 0.6227492801936535,
"learning_rate": 3.9992004554102435e-05,
"loss": 0.4047,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 0.5875139941035956,
"learning_rate": 3.999032564583976e-05,
"loss": 0.4072,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 0.8889434791572968,
"learning_rate": 3.998848689546591e-05,
"loss": 0.4184,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 0.8125577207060808,
"learning_rate": 3.9986488317683496e-05,
"loss": 0.3888,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 0.641154813028446,
"learning_rate": 3.9984329928473067e-05,
"loss": 0.4075,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 0.5504131096743263,
"learning_rate": 3.998201174509306e-05,
"loss": 0.411,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 0.6750532304336851,
"learning_rate": 3.997953378607961e-05,
"loss": 0.3765,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 0.5906136817859419,
"learning_rate": 3.9976896071246403e-05,
"loss": 0.3862,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 0.5563534286323905,
"learning_rate": 3.997409862168454e-05,
"loss": 0.3839,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 0.5009193306531106,
"learning_rate": 3.997114145976237e-05,
"loss": 0.3985,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 0.46822707619103193,
"learning_rate": 3.9968024609125245e-05,
"loss": 0.3881,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 0.5272976961294604,
"learning_rate": 3.996474809469545e-05,
"loss": 0.3929,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 0.4613969090618867,
"learning_rate": 3.996131194267189e-05,
"loss": 0.3888,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 0.46380491216602954,
"learning_rate": 3.995771618052995e-05,
"loss": 0.3827,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 0.4689094685697794,
"learning_rate": 3.9953960837021256e-05,
"loss": 0.3695,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 0.5028404095185447,
"learning_rate": 3.995004594217343e-05,
"loss": 0.4078,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 0.4366179108075408,
"learning_rate": 3.994597152728987e-05,
"loss": 0.3819,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 0.5495463750381202,
"learning_rate": 3.994173762494947e-05,
"loss": 0.3786,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 0.5621236550524967,
"learning_rate": 3.9937344269006424e-05,
"loss": 0.4173,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 0.447295896876076,
"learning_rate": 3.993279149458988e-05,
"loss": 0.4036,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 0.7325310918201142,
"learning_rate": 3.9928079338103695e-05,
"loss": 0.4096,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 0.41695013489647104,
"learning_rate": 3.992320783722616e-05,
"loss": 0.3872,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.6396317656837619,
"learning_rate": 3.991817703090965e-05,
"loss": 0.384,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 0.4716520744413679,
"learning_rate": 3.991298695938038e-05,
"loss": 0.403,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 0.449783097555441,
"learning_rate": 3.9907637664138e-05,
"loss": 0.3789,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 0.4519686997361641,
"learning_rate": 3.990212918795534e-05,
"loss": 0.3901,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 0.4140862141289775,
"learning_rate": 3.9896461574878036e-05,
"loss": 0.3878,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 0.4863823804767403,
"learning_rate": 3.989063487022416e-05,
"loss": 0.4029,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 0.421397442103566,
"learning_rate": 3.9884649120583903e-05,
"loss": 0.379,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 0.45431182051553315,
"learning_rate": 3.987850437381915e-05,
"loss": 0.3925,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 0.475117080955099,
"learning_rate": 3.987220067906313e-05,
"loss": 0.4066,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 0.5672685608663984,
"learning_rate": 3.986573808672002e-05,
"loss": 0.393,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 0.46782324583240087,
"learning_rate": 3.9859116648464534e-05,
"loss": 0.3995,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 0.7388060329665525,
"learning_rate": 3.985233641724152e-05,
"loss": 0.3991,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 0.4888558494512745,
"learning_rate": 3.984539744726554e-05,
"loss": 0.4053,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 0.641820549064771,
"learning_rate": 3.983829979402038e-05,
"loss": 0.3748,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 0.555143097922395,
"learning_rate": 3.983104351425871e-05,
"loss": 0.3869,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 0.6049049895608304,
"learning_rate": 3.982362866600154e-05,
"loss": 0.387,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 0.6041445503204619,
"learning_rate": 3.9816055308537806e-05,
"loss": 0.4087,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 0.5462343690972074,
"learning_rate": 3.980832350242388e-05,
"loss": 0.4003,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 0.5115841640854688,
"learning_rate": 3.980043330948306e-05,
"loss": 0.3812,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 0.47902600637222165,
"learning_rate": 3.9792384792805124e-05,
"loss": 0.4029,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 0.6028529546625387,
"learning_rate": 3.978417801674579e-05,
"loss": 0.3728,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 0.6590979388605386,
"learning_rate": 3.9775813046926244e-05,
"loss": 0.3963,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 0.6280338680882992,
"learning_rate": 3.976728995023254e-05,
"loss": 0.4199,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 0.5287834428059499,
"learning_rate": 3.975860879481514e-05,
"loss": 0.3914,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 0.5119036328708853,
"learning_rate": 3.974976965008833e-05,
"loss": 0.4133,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 0.5639408564986021,
"learning_rate": 3.9740772586729674e-05,
"loss": 0.3982,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.6116906560502448,
"learning_rate": 3.973161767667946e-05,
"loss": 0.4017,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 0.4845154006034919,
"learning_rate": 3.972230499314009e-05,
"loss": 0.3619,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 0.5938557891230456,
"learning_rate": 3.9712834610575545e-05,
"loss": 0.4003,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 0.5117303320503606,
"learning_rate": 3.9703206604710746e-05,
"loss": 0.3781,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 0.4932374027707179,
"learning_rate": 3.969342105253098e-05,
"loss": 0.3885,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 0.5511061469697187,
"learning_rate": 3.9683478032281246e-05,
"loss": 0.3931,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 0.7361673566970764,
"learning_rate": 3.967337762346568e-05,
"loss": 0.3847,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 0.5362328463639989,
"learning_rate": 3.9663119906846885e-05,
"loss": 0.4056,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.7171649374368779,
"learning_rate": 3.965270496444528e-05,
"loss": 0.4035,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 0.5895331281913747,
"learning_rate": 3.9642132879538465e-05,
"loss": 0.377,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 0.5931715796963459,
"learning_rate": 3.9631403736660566e-05,
"loss": 0.3842,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 0.5683430923898121,
"learning_rate": 3.962051762160151e-05,
"loss": 0.3854,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 0.8219353482394286,
"learning_rate": 3.960947462140638e-05,
"loss": 0.3974,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 0.561542750798209,
"learning_rate": 3.959827482437472e-05,
"loss": 0.3808,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 0.6164265962421799,
"learning_rate": 3.958691832005981e-05,
"loss": 0.3943,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 0.6542739095300001,
"learning_rate": 3.9575405199267944e-05,
"loss": 0.3937,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 0.6247084242131644,
"learning_rate": 3.956373555405774e-05,
"loss": 0.3913,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 0.6256475467265312,
"learning_rate": 3.955190947773937e-05,
"loss": 0.4026,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 0.6532128918807941,
"learning_rate": 3.9539927064873824e-05,
"loss": 0.406,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 0.4893633724398487,
"learning_rate": 3.952778841127214e-05,
"loss": 0.4016,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 0.6454630753339586,
"learning_rate": 3.95154936139947e-05,
"loss": 0.3974,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 0.5294753628972007,
"learning_rate": 3.950304277135034e-05,
"loss": 0.3809,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 0.4412408193097422,
"learning_rate": 3.94904359828957e-05,
"loss": 0.3851,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 0.5395207326888909,
"learning_rate": 3.94776733494343e-05,
"loss": 0.3794,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 0.49658855197727053,
"learning_rate": 3.9464754973015826e-05,
"loss": 0.3875,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 0.49880849280553874,
"learning_rate": 3.9451680956935275e-05,
"loss": 0.3659,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.5100998352113061,
"learning_rate": 3.943845140573213e-05,
"loss": 0.3684,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.48571429641381275,
"learning_rate": 3.942506642518952e-05,
"loss": 0.3772,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 0.4963236552602622,
"learning_rate": 3.9411526122333396e-05,
"loss": 0.3769,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 0.4956158252850868,
"learning_rate": 3.939783060543167e-05,
"loss": 0.368,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 0.5711379489378268,
"learning_rate": 3.938397998399333e-05,
"loss": 0.3947,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.43126946312096914,
"learning_rate": 3.936997436876756e-05,
"loss": 0.382,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.5186378332364991,
"learning_rate": 3.935581387174289e-05,
"loss": 0.3846,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 0.425828421807542,
"learning_rate": 3.9341498606146295e-05,
"loss": 0.3942,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.4253739772254414,
"learning_rate": 3.932702868644224e-05,
"loss": 0.3995,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.4343308075651997,
"learning_rate": 3.931240422833181e-05,
"loss": 0.403,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.4006917280285471,
"learning_rate": 3.9297625348751806e-05,
"loss": 0.3705,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 0.4953662724267713,
"learning_rate": 3.9282692165873744e-05,
"loss": 0.3841,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 0.45554230164834886,
"learning_rate": 3.9267604799102945e-05,
"loss": 0.3932,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.45603887554368794,
"learning_rate": 3.9252363369077615e-05,
"loss": 0.3867,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 0.3933713573754479,
"learning_rate": 3.9236967997667796e-05,
"loss": 0.3753,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 0.49402477782523413,
"learning_rate": 3.9221418807974496e-05,
"loss": 0.3783,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 0.4480621598877441,
"learning_rate": 3.92057159243286e-05,
"loss": 0.3837,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 0.5171782907463587,
"learning_rate": 3.9189859472289956e-05,
"loss": 0.394,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 0.5135366222792062,
"learning_rate": 3.9173849578646324e-05,
"loss": 0.388,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 0.4855614284998899,
"learning_rate": 3.9157686371412395e-05,
"loss": 0.3714,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.4083568933367323,
"learning_rate": 3.914136997982874e-05,
"loss": 0.3835,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 0.4562832256547664,
"learning_rate": 3.912490053436079e-05,
"loss": 0.382,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 0.44648915100508174,
"learning_rate": 3.9108278166697805e-05,
"loss": 0.3728,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.40997384032003514,
"learning_rate": 3.909150300975178e-05,
"loss": 0.3565,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 0.4360998565657497,
"learning_rate": 3.907457519765642e-05,
"loss": 0.3663,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.39457399485147765,
"learning_rate": 3.905749486576607e-05,
"loss": 0.3683,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.4278313325432642,
"learning_rate": 3.90402621506546e-05,
"loss": 0.374,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 0.5448408372190856,
"learning_rate": 3.902287719011435e-05,
"loss": 0.3898,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.4177309545014541,
"learning_rate": 3.900534012315499e-05,
"loss": 0.3857,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 0.4957978652706148,
"learning_rate": 3.8987651090002454e-05,
"loss": 0.38,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 0.42783208700482345,
"learning_rate": 3.896981023209777e-05,
"loss": 0.3938,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 0.41751526806279604,
"learning_rate": 3.8951817692095966e-05,
"loss": 0.3814,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 0.3820850335900586,
"learning_rate": 3.893367361386491e-05,
"loss": 0.3621,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 0.43585704625066723,
"learning_rate": 3.891537814248417e-05,
"loss": 0.3633,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 0.3797344696762601,
"learning_rate": 3.889693142424384e-05,
"loss": 0.365,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 0.5633194811583935,
"learning_rate": 3.8878333606643405e-05,
"loss": 0.3974,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 0.3670354854947812,
"learning_rate": 3.885958483839049e-05,
"loss": 0.3703,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 0.6058084187035943,
"learning_rate": 3.8840685269399786e-05,
"loss": 0.3741,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 0.3738056144640129,
"learning_rate": 3.882163505079171e-05,
"loss": 0.3687,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 0.5875277222929692,
"learning_rate": 3.880243433489132e-05,
"loss": 0.3828,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 0.44503321763003084,
"learning_rate": 3.878308327522702e-05,
"loss": 0.3157,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 0.4268449256983812,
"learning_rate": 3.876358202652936e-05,
"loss": 0.3259,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 0.3857719967075647,
"learning_rate": 3.8743930744729835e-05,
"loss": 0.3311,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.4562832227520598,
"learning_rate": 3.8724129586959535e-05,
"loss": 0.3211,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.40614211386771343,
"learning_rate": 3.870417871154801e-05,
"loss": 0.3256,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 0.43949591243016556,
"learning_rate": 3.868407827802192e-05,
"loss": 0.3063,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.5231832502368552,
"learning_rate": 3.8663828447103804e-05,
"loss": 0.3214,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.4332700104932504,
"learning_rate": 3.8643429380710765e-05,
"loss": 0.3187,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.5451706133016572,
"learning_rate": 3.862288124195319e-05,
"loss": 0.3173,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.41426691577592906,
"learning_rate": 3.8602184195133464e-05,
"loss": 0.3283,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 0.4698257776538071,
"learning_rate": 3.8581338405744604e-05,
"loss": 0.3227,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 0.46773125479819383,
"learning_rate": 3.8560344040468996e-05,
"loss": 0.3402,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.42950850280751846,
"learning_rate": 3.853920126717703e-05,
"loss": 0.3293,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 0.5707096518312519,
"learning_rate": 3.851791025492574e-05,
"loss": 0.3252,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.39076985459655666,
"learning_rate": 3.8496471173957515e-05,
"loss": 0.3352,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.5579504156342322,
"learning_rate": 3.847488419569866e-05,
"loss": 0.3294,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.4848209869023167,
"learning_rate": 3.8453149492758077e-05,
"loss": 0.336,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 0.4429734517009479,
"learning_rate": 3.8431267238925886e-05,
"loss": 0.3415,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.4763261088988063,
"learning_rate": 3.840923760917199e-05,
"loss": 0.3174,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 0.4352857603092906,
"learning_rate": 3.838706077964473e-05,
"loss": 0.3146,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 0.5099445669644449,
"learning_rate": 3.836473692766944e-05,
"loss": 0.3195,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 0.41250709295743376,
"learning_rate": 3.834226623174705e-05,
"loss": 0.3241,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.5020092561351547,
"learning_rate": 3.831964887155264e-05,
"loss": 0.3219,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.4389472522796087,
"learning_rate": 3.829688502793403e-05,
"loss": 0.3204,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.4794016462290678,
"learning_rate": 3.827397488291029e-05,
"loss": 0.3296,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.4817025998250113,
"learning_rate": 3.8250918619670343e-05,
"loss": 0.3351,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.4619780171275328,
"learning_rate": 3.822771642257145e-05,
"loss": 0.3092,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.42715087204428803,
"learning_rate": 3.820436847713776e-05,
"loss": 0.3103,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 0.42761021049087533,
"learning_rate": 3.8180874970058826e-05,
"loss": 0.3017,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.5389688820703032,
"learning_rate": 3.8157236089188103e-05,
"loss": 0.3262,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.44861712424009487,
"learning_rate": 3.813345202354145e-05,
"loss": 0.3313,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 0.4813550886270817,
"learning_rate": 3.810952296329563e-05,
"loss": 0.3221,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.3961809163609869,
"learning_rate": 3.808544909978676e-05,
"loss": 0.3171,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.4623753812954366,
"learning_rate": 3.806123062550882e-05,
"loss": 0.3395,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.4154366837012121,
"learning_rate": 3.803686773411208e-05,
"loss": 0.3242,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.42287901993974575,
"learning_rate": 3.801236062040158e-05,
"loss": 0.3122,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 0.46734782694579874,
"learning_rate": 3.798770948033553e-05,
"loss": 0.3321,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.3835516216095152,
"learning_rate": 3.796291451102381e-05,
"loss": 0.3298,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.451470651217603,
"learning_rate": 3.793797591072631e-05,
"loss": 0.3134,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.461117179010244,
"learning_rate": 3.791289387885144e-05,
"loss": 0.3182,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.4113544614968738,
"learning_rate": 3.788766861595443e-05,
"loss": 0.3197,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 0.4473611006787302,
"learning_rate": 3.7862300323735835e-05,
"loss": 0.3319,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 0.44660905906752346,
"learning_rate": 3.783678920503983e-05,
"loss": 0.3108,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 0.4381938152434538,
"learning_rate": 3.781113546385265e-05,
"loss": 0.3114,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.4070155245541933,
"learning_rate": 3.7785339305300916e-05,
"loss": 0.322,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.4719852231376736,
"learning_rate": 3.7759400935650025e-05,
"loss": 0.3121,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.37875783704665145,
"learning_rate": 3.77333205623025e-05,
"loss": 0.3131,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.42770865263858,
"learning_rate": 3.7707098393796295e-05,
"loss": 0.3205,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.3925612835634973,
"learning_rate": 3.7680734639803185e-05,
"loss": 0.3247,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 0.49905796943986125,
"learning_rate": 3.765422951112704e-05,
"loss": 0.3504,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 0.35487316717264195,
"learning_rate": 3.762758321970216e-05,
"loss": 0.3051,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.40806623756272725,
"learning_rate": 3.7600795978591584e-05,
"loss": 0.3264,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.39373620101858947,
"learning_rate": 3.757386800198538e-05,
"loss": 0.3362,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.42351576708867605,
"learning_rate": 3.7546799505198925e-05,
"loss": 0.3056,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.3878832654748417,
"learning_rate": 3.7519590704671197e-05,
"loss": 0.3142,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.4739233437765857,
"learning_rate": 3.749224181796305e-05,
"loss": 0.331,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.3938120103091055,
"learning_rate": 3.746475306375545e-05,
"loss": 0.3515,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.4493895299711083,
"learning_rate": 3.743712466184774e-05,
"loss": 0.3311,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.4268756958890564,
"learning_rate": 3.7409356833155885e-05,
"loss": 0.3183,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.42638857832587346,
"learning_rate": 3.73814497997107e-05,
"loss": 0.3178,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.4034339255529194,
"learning_rate": 3.7353403784656095e-05,
"loss": 0.3146,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.444366989125983,
"learning_rate": 3.732521901224724e-05,
"loss": 0.3253,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.40724655776401414,
"learning_rate": 3.7296895707848844e-05,
"loss": 0.3228,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.40937143016836064,
"learning_rate": 3.7268434097933275e-05,
"loss": 0.3093,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.3481210890527789,
"learning_rate": 3.723983441007882e-05,
"loss": 0.3187,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.4483623196240404,
"learning_rate": 3.721109687296781e-05,
"loss": 0.3088,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.43128792554607154,
"learning_rate": 3.718222171638484e-05,
"loss": 0.323,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 0.4806712869801189,
"learning_rate": 3.715320917121488e-05,
"loss": 0.3088,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.49979069895089795,
"learning_rate": 3.712405946944149e-05,
"loss": 0.3428,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.3794618055620391,
"learning_rate": 3.709477284414488e-05,
"loss": 0.3239,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.4068089994265772,
"learning_rate": 3.7065349529500136e-05,
"loss": 0.3311,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 0.3804927681181612,
"learning_rate": 3.7035789760775305e-05,
"loss": 0.323,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.43300210099127584,
"learning_rate": 3.70060937743295e-05,
"loss": 0.3118,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.4411103622863515,
"learning_rate": 3.6976261807611045e-05,
"loss": 0.3275,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.36077489617659175,
"learning_rate": 3.694629409915555e-05,
"loss": 0.3289,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.3733140397784797,
"learning_rate": 3.691619088858401e-05,
"loss": 0.3135,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.3752558738666805,
"learning_rate": 3.68859524166009e-05,
"loss": 0.3398,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.3683931573848629,
"learning_rate": 3.6855578924992226e-05,
"loss": 0.32,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.3649193630025182,
"learning_rate": 3.6825070656623626e-05,
"loss": 0.3212,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.37225831066184795,
"learning_rate": 3.67944278554384e-05,
"loss": 0.3108,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 0.3746511464860324,
"learning_rate": 3.676365076645557e-05,
"loss": 0.3223,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.39795478520668087,
"learning_rate": 3.673273963576791e-05,
"loss": 0.3381,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.4032129683509813,
"learning_rate": 3.670169471054001e-05,
"loss": 0.316,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.4287172192096478,
"learning_rate": 3.6670516239006254e-05,
"loss": 0.331,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.4386165440315268,
"learning_rate": 3.663920447046888e-05,
"loss": 0.3114,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.43830874024885724,
"learning_rate": 3.6607759655295954e-05,
"loss": 0.3301,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.4594983314086936,
"learning_rate": 3.657618204491937e-05,
"loss": 0.3308,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.4470520088796891,
"learning_rate": 3.6544471891832875e-05,
"loss": 0.3157,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.4158224507311116,
"learning_rate": 3.6512629449589983e-05,
"loss": 0.3235,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.4285244008139653,
"learning_rate": 3.6480654972802045e-05,
"loss": 0.3215,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.4756740979853791,
"learning_rate": 3.644854871713611e-05,
"loss": 0.3143,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.3995157783326863,
"learning_rate": 3.6416310939312924e-05,
"loss": 0.3394,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.4016558241968382,
"learning_rate": 3.638394189710493e-05,
"loss": 0.3189,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.41369207085665544,
"learning_rate": 3.635144184933412e-05,
"loss": 0.321,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.3872214375477554,
"learning_rate": 3.6318811055869994e-05,
"loss": 0.3241,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.41672973059057494,
"learning_rate": 3.628604977762752e-05,
"loss": 0.3198,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.41721254986083683,
"learning_rate": 3.6253158276565004e-05,
"loss": 0.3251,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.3843826266278,
"learning_rate": 3.6220136815682016e-05,
"loss": 0.3265,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.4143952738968092,
"learning_rate": 3.618698565901727e-05,
"loss": 0.3209,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.40545407811862916,
"learning_rate": 3.615370507164654e-05,
"loss": 0.3294,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.4056235269034816,
"learning_rate": 3.61202953196805e-05,
"loss": 0.3316,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 0.38210310672160075,
"learning_rate": 3.6086756670262654e-05,
"loss": 0.3185,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.4037670159554757,
"learning_rate": 3.6053089391567135e-05,
"loss": 0.3318,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.3842163868104707,
"learning_rate": 3.601929375279662e-05,
"loss": 0.3254,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.38191153000710265,
"learning_rate": 3.598537002418012e-05,
"loss": 0.308,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.3569399328353084,
"learning_rate": 3.5951318476970885e-05,
"loss": 0.3319,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.44730902094813746,
"learning_rate": 3.591713938344416e-05,
"loss": 0.3217,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.3665304010680108,
"learning_rate": 3.588283301689507e-05,
"loss": 0.3138,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 0.4798542979702406,
"learning_rate": 3.5848399651636424e-05,
"loss": 0.3205,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.46885153534435714,
"learning_rate": 3.5813839562996473e-05,
"loss": 0.3247,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.40806838414036667,
"learning_rate": 3.5779153027316775e-05,
"loss": 0.3302,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.443739073603629,
"learning_rate": 3.574434032194994e-05,
"loss": 0.3234,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.32862980138618975,
"learning_rate": 3.5709401725257436e-05,
"loss": 0.3242,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.3747515375020223,
"learning_rate": 3.5674337516607346e-05,
"loss": 0.3145,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.40571962119100147,
"learning_rate": 3.563914797637216e-05,
"loss": 0.3236,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.38347374542753593,
"learning_rate": 3.560383338592649e-05,
"loss": 0.3214,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.40048077087346867,
"learning_rate": 3.556839402764487e-05,
"loss": 0.3295,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.38269881751326557,
"learning_rate": 3.553283018489946e-05,
"loss": 0.313,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.3943057218087236,
"learning_rate": 3.54971421420578e-05,
"loss": 0.3198,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.35118189626452795,
"learning_rate": 3.5461330184480524e-05,
"loss": 0.3169,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 0.4458983550058221,
"learning_rate": 3.54253945985191e-05,
"loss": 0.3133,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.39191681409264584,
"learning_rate": 3.5389335671513524e-05,
"loss": 0.3194,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.46541088511057066,
"learning_rate": 3.5353153691789986e-05,
"loss": 0.3372,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.3614645699692343,
"learning_rate": 3.531684894865866e-05,
"loss": 0.3197,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.4016776952278192,
"learning_rate": 3.528042173241128e-05,
"loss": 0.3285,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.3636411437660278,
"learning_rate": 3.524387233431892e-05,
"loss": 0.3033,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.33010412499153846,
"learning_rate": 3.520720104662958e-05,
"loss": 0.3472,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.4510021666832989,
"learning_rate": 3.5170408162565904e-05,
"loss": 0.3283,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.3520463940566836,
"learning_rate": 3.513349397632281e-05,
"loss": 0.3189,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.38292106791362257,
"learning_rate": 3.5096458783065145e-05,
"loss": 0.3177,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.429991022085268,
"learning_rate": 3.505930287892533e-05,
"loss": 0.3441,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.382513491094526,
"learning_rate": 3.502202656100099e-05,
"loss": 0.3219,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.3894906839557917,
"learning_rate": 3.498463012735256e-05,
"loss": 0.3273,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.32336219967463203,
"learning_rate": 3.494711387700094e-05,
"loss": 0.3253,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.3807459602100437,
"learning_rate": 3.490947810992505e-05,
"loss": 0.3151,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 0.3579692663705623,
"learning_rate": 3.48717231270595e-05,
"loss": 0.3207,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.3417319061833639,
"learning_rate": 3.483384923029211e-05,
"loss": 0.3227,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.39780495346053374,
"learning_rate": 3.479585672246156e-05,
"loss": 0.3222,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.3376192718806144,
"learning_rate": 3.4757745907354924e-05,
"loss": 0.317,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.38244279915669027,
"learning_rate": 3.471951708970527e-05,
"loss": 0.3112,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.431423052398115,
"learning_rate": 3.468117057518921e-05,
"loss": 0.3178,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.3930722374385204,
"learning_rate": 3.4642706670424426e-05,
"loss": 0.3371,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.44031799624379364,
"learning_rate": 3.460412568296731e-05,
"loss": 0.328,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.4270698575321784,
"learning_rate": 3.456542792131039e-05,
"loss": 0.3053,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.4888957828270502,
"learning_rate": 3.4526613694879936e-05,
"loss": 0.314,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.41869078940730103,
"learning_rate": 3.448768331403348e-05,
"loss": 0.3286,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.4075979122331733,
"learning_rate": 3.44486370900573e-05,
"loss": 0.3292,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.41021799267860254,
"learning_rate": 3.440947533516396e-05,
"loss": 0.3215,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.3465537257976065,
"learning_rate": 3.4370198362489816e-05,
"loss": 0.3243,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 0.4111783965538572,
"learning_rate": 3.43308064860925e-05,
"loss": 0.3379,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.3434774337300989,
"learning_rate": 3.4291300020948393e-05,
"loss": 0.3158,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.4032562484696663,
"learning_rate": 3.425167928295015e-05,
"loss": 0.3006,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.4076165903116418,
"learning_rate": 3.421194458890411e-05,
"loss": 0.3219,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.4383232856391409,
"learning_rate": 3.4172096256527865e-05,
"loss": 0.3313,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.42693517001777886,
"learning_rate": 3.4132134604447595e-05,
"loss": 0.318,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.4307329072443837,
"learning_rate": 3.409205995219561e-05,
"loss": 0.3211,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.4683373233503692,
"learning_rate": 3.4051872620207765e-05,
"loss": 0.3195,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.39295223469461504,
"learning_rate": 3.40115729298209e-05,
"loss": 0.3236,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.4290660188172711,
"learning_rate": 3.397116120327025e-05,
"loss": 0.3309,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.371659958930056,
"learning_rate": 3.3930637763686916e-05,
"loss": 0.3106,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.32035583347105206,
"learning_rate": 3.389000293509524e-05,
"loss": 0.3239,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.4295722102926204,
"learning_rate": 3.3849257042410225e-05,
"loss": 0.3221,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.3302409179742167,
"learning_rate": 3.380840041143494e-05,
"loss": 0.3303,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.35849789243737307,
"learning_rate": 3.376743336885792e-05,
"loss": 0.321,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.3746464201187395,
"learning_rate": 3.372635624225055e-05,
"loss": 0.3135,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.35966788806457317,
"learning_rate": 3.3685169360064445e-05,
"loss": 0.3244,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.4201746503760204,
"learning_rate": 3.3643873051628804e-05,
"loss": 0.3187,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.3696879516147183,
"learning_rate": 3.3602467647147814e-05,
"loss": 0.3299,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.3491927459583141,
"learning_rate": 3.356095347769799e-05,
"loss": 0.3258,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.3862374934405694,
"learning_rate": 3.351933087522552e-05,
"loss": 0.3397,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.3522659245056294,
"learning_rate": 3.3477600172543626e-05,
"loss": 0.325,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.3483040557848633,
"learning_rate": 3.3435761703329894e-05,
"loss": 0.3092,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.34806791589818753,
"learning_rate": 3.3393815802123606e-05,
"loss": 0.3174,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.36226912948766765,
"learning_rate": 3.335176280432308e-05,
"loss": 0.3268,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.33126183617131916,
"learning_rate": 3.330960304618295e-05,
"loss": 0.3155,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.3598522251049202,
"learning_rate": 3.326733686481153e-05,
"loss": 0.3042,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.358791988093143,
"learning_rate": 3.3224964598168084e-05,
"loss": 0.3301,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.4109621420518497,
"learning_rate": 3.318248658506012e-05,
"loss": 0.3244,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.3438606853722536,
"learning_rate": 3.313990316514069e-05,
"loss": 0.2983,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.3522268096506027,
"learning_rate": 3.309721467890571e-05,
"loss": 0.331,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.33114101127603274,
"learning_rate": 3.305442146769115e-05,
"loss": 0.31,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.3865514985365716,
"learning_rate": 3.30115238736704e-05,
"loss": 0.3198,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.3961181069427452,
"learning_rate": 3.296852223985148e-05,
"loss": 0.3259,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 0.36038901055779043,
"learning_rate": 3.2925416910074315e-05,
"loss": 0.3148,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.3860935111049238,
"learning_rate": 3.288220822900796e-05,
"loss": 0.314,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.3433755900966691,
"learning_rate": 3.283889654214788e-05,
"loss": 0.3087,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.4149325992785654,
"learning_rate": 3.279548219581317e-05,
"loss": 0.3144,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.3669573362001192,
"learning_rate": 3.275196553714379e-05,
"loss": 0.3298,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.3733371467779981,
"learning_rate": 3.270834691409778e-05,
"loss": 0.3244,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.33421291993289687,
"learning_rate": 3.266462667544849e-05,
"loss": 0.3081,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.38587563231094907,
"learning_rate": 3.262080517078178e-05,
"loss": 0.3368,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.3693844682156917,
"learning_rate": 3.257688275049323e-05,
"loss": 0.3154,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.330890350450153,
"learning_rate": 3.253285976578535e-05,
"loss": 0.3349,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.3422887412830603,
"learning_rate": 3.2488736568664756e-05,
"loss": 0.319,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.400662605554219,
"learning_rate": 3.244451351193935e-05,
"loss": 0.3298,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 0.3361764295732026,
"learning_rate": 3.240019094921551e-05,
"loss": 0.3206,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.3879445601391551,
"learning_rate": 3.235576923489527e-05,
"loss": 0.3148,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.348347953745778,
"learning_rate": 3.2311248724173476e-05,
"loss": 0.3238,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.3676661664789582,
"learning_rate": 3.226662977303494e-05,
"loss": 0.3166,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.333593585453171,
"learning_rate": 3.2221912738251626e-05,
"loss": 0.3106,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.35637420094317107,
"learning_rate": 3.2177097977379736e-05,
"loss": 0.3047,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.3304824146011177,
"learning_rate": 3.21321858487569e-05,
"loss": 0.3282,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.35254496500387533,
"learning_rate": 3.208717671149931e-05,
"loss": 0.3295,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.3443996927334605,
"learning_rate": 3.204207092549882e-05,
"loss": 0.3213,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 0.3510384368490265,
"learning_rate": 3.19968688514201e-05,
"loss": 0.317,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.34672446616046065,
"learning_rate": 3.195157085069774e-05,
"loss": 0.322,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.33932574878280486,
"learning_rate": 3.190617728553332e-05,
"loss": 0.3255,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.3569610637939083,
"learning_rate": 3.1860688518892606e-05,
"loss": 0.3287,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.34740563838708344,
"learning_rate": 3.181510491450254e-05,
"loss": 0.3187,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.31909462517469145,
"learning_rate": 3.176942683684842e-05,
"loss": 0.3199,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.32482918065528765,
"learning_rate": 3.1723654651170934e-05,
"loss": 0.324,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.3428251190832483,
"learning_rate": 3.167778872346327e-05,
"loss": 0.3033,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.3294736155912481,
"learning_rate": 3.163182942046815e-05,
"loss": 0.3217,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.354826695246014,
"learning_rate": 3.158577710967494e-05,
"loss": 0.31,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.33580581456799724,
"learning_rate": 3.153963215931669e-05,
"loss": 0.3052,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.3405322731369996,
"learning_rate": 3.14933949383672e-05,
"loss": 0.3243,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.394370261559824,
"learning_rate": 3.144706581653805e-05,
"loss": 0.3234,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.35768748037020576,
"learning_rate": 3.140064516427566e-05,
"loss": 0.3201,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.44170583312253997,
"learning_rate": 3.135413335275833e-05,
"loss": 0.3271,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.38789418500558404,
"learning_rate": 3.130753075389327e-05,
"loss": 0.3303,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.3419551775860493,
"learning_rate": 3.1260837740313625e-05,
"loss": 0.3209,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.34184002385967766,
"learning_rate": 3.121405468537547e-05,
"loss": 0.3039,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.3947134367157758,
"learning_rate": 3.116718196315488e-05,
"loss": 0.3197,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.34163667734031455,
"learning_rate": 3.1120219948444894e-05,
"loss": 0.3331,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.35888777697011237,
"learning_rate": 3.107316901675252e-05,
"loss": 0.3204,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.3458584929288499,
"learning_rate": 3.1026029544295745e-05,
"loss": 0.3233,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.33933006697485707,
"learning_rate": 3.097880190800057e-05,
"loss": 0.3269,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.37362203775158964,
"learning_rate": 3.093148648549788e-05,
"loss": 0.3217,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.3353487123450009,
"learning_rate": 3.088408365512055e-05,
"loss": 0.314,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.3494310088635901,
"learning_rate": 3.083659379590034e-05,
"loss": 0.3188,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.33545164910618397,
"learning_rate": 3.0789017287564913e-05,
"loss": 0.3372,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.3170888680092684,
"learning_rate": 3.074135451053475e-05,
"loss": 0.3164,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.33821438271981547,
"learning_rate": 3.069360584592015e-05,
"loss": 0.3177,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.36933096020533884,
"learning_rate": 3.064577167551816e-05,
"loss": 0.3137,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.3439264612810848,
"learning_rate": 3.059785238180953e-05,
"loss": 0.3391,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.38277385560157273,
"learning_rate": 3.054984834795566e-05,
"loss": 0.3182,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.38162808331011855,
"learning_rate": 3.0501759957795537e-05,
"loss": 0.3299,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.37392746293480195,
"learning_rate": 3.0453587595842627e-05,
"loss": 0.3224,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.37676303110315307,
"learning_rate": 3.0405331647281856e-05,
"loss": 0.3292,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.40927485911006267,
"learning_rate": 3.0356992497966508e-05,
"loss": 0.3194,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.36273429856851347,
"learning_rate": 3.0308570534415137e-05,
"loss": 0.3158,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.39225506690010387,
"learning_rate": 3.0260066143808467e-05,
"loss": 0.3215,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.3241473064137531,
"learning_rate": 3.021147971398632e-05,
"loss": 0.3026,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.32552698937408786,
"learning_rate": 3.0162811633444478e-05,
"loss": 0.3236,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.35299124173649155,
"learning_rate": 3.0114062291331635e-05,
"loss": 0.3099,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.3200929461709145,
"learning_rate": 3.0065232077446217e-05,
"loss": 0.3368,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.3802584253960338,
"learning_rate": 3.001632138223332e-05,
"loss": 0.332,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.4364795301804018,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.2487,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 0.39037194158471317,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.2413,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 0.6120338597138144,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.2638,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.47423727243371827,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.2502,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.37990030920008294,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.2716,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.34167093167513823,
"learning_rate": 2.972118904869376e-05,
"loss": 0.2442,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.4111356738933073,
"learning_rate": 2.9671725970312563e-05,
"loss": 0.2592,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.3963803923762786,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.2613,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.3621514641903223,
"learning_rate": 2.95725682050846e-05,
"loss": 0.236,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.3987793998147666,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.2501,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.38112739299015597,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.2605,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.3931002935177087,
"learning_rate": 2.942325848714505e-05,
"loss": 0.247,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.33947848974276856,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.2346,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.4239277494455557,
"learning_rate": 2.932334127129686e-05,
"loss": 0.2554,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.3849009039877778,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.2492,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.35298085491565523,
"learning_rate": 2.922312585929131e-05,
"loss": 0.2531,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.3358001600520372,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.2286,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.3488560352950848,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.2305,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.36780329003021595,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.2449,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.3459767671622799,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.2238,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.38425955898488195,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.2435,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.30083375406998086,
"learning_rate": 2.8920722546128863e-05,
"loss": 0.2572,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.39200083603602187,
"learning_rate": 2.887006998351372e-05,
"loss": 0.2443,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.3307438833560409,
"learning_rate": 2.881934649605535e-05,
"loss": 0.2573,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.40486610737358136,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.2465,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.3385540749822172,
"learning_rate": 2.8717688369507425e-05,
"loss": 0.2566,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.3798541325505264,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.2352,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.3556227170336656,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.232,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.37689356272188906,
"learning_rate": 2.856467940121025e-05,
"loss": 0.2251,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.35737000551541576,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.2506,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.35727052476782734,
"learning_rate": 2.846233032790149e-05,
"loss": 0.244,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.3294617096977523,
"learning_rate": 2.84110540896642e-05,
"loss": 0.2374,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.3660820255065659,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.244,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.3417596311995891,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.2209,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.3561518434614923,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.2517,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.32776297382644937,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.2454,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.3541901600109665,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.2411,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.31329481756392147,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.2555,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.31497824420881604,
"learning_rate": 2.805026032173773e-05,
"loss": 0.2574,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.3132104021829122,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.2381,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.32404053277635253,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.2526,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.3293796861621834,
"learning_rate": 2.7894660620408e-05,
"loss": 0.253,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.31930380540178827,
"learning_rate": 2.784266724883173e-05,
"loss": 0.236,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.3718522127793506,
"learning_rate": 2.7790611167497128e-05,
"loss": 0.239,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.3360865958245743,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.2435,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.39872689811932643,
"learning_rate": 2.768631254100716e-05,
"loss": 0.2385,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.3116282922526265,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.234,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.45075473961088963,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.2534,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.3244637454170248,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.2416,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.3924437614025676,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.2621,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.32291869138878826,
"learning_rate": 2.742449775134085e-05,
"loss": 0.2359,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.3542060437175637,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.2433,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.34569141972947653,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.246,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.3308208035722863,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.2534,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.3236496865121672,
"learning_rate": 2.7213974824975785e-05,
"loss": 0.2225,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.33315243779129033,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.24,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.31061566037409755,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.2433,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.32495098946118717,
"learning_rate": 2.705547548999996e-05,
"loss": 0.2503,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.3334380410368891,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.2594,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.33781300159289074,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.2436,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.3217617051184434,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.2374,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.3274064632964173,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.2395,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.3203098983660167,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.2432,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.33441731230694344,
"learning_rate": 2.67369650552155e-05,
"loss": 0.2564,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.3439551762828283,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.2323,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.3674327284885919,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.2556,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.3585935890123786,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.2458,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.3389736197715658,
"learning_rate": 2.652354173593614e-05,
"loss": 0.2452,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.3687533456054352,
"learning_rate": 2.647005443331317e-05,
"loss": 0.2423,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.3354693356153226,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.2658,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.38700071282898274,
"learning_rate": 2.636292505301542e-05,
"loss": 0.2489,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.33244490155449824,
"learning_rate": 2.630928383194431e-05,
"loss": 0.2498,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.36951187323094675,
"learning_rate": 2.625559216200781e-05,
"loss": 0.2363,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.3996546812831347,
"learning_rate": 2.620185047252308e-05,
"loss": 0.2398,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.3354585593343862,
"learning_rate": 2.614805919320723e-05,
"loss": 0.2456,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.3406262053434616,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.2338,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.3620462200140203,
"learning_rate": 2.60403295859298e-05,
"loss": 0.2541,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.35796800860907224,
"learning_rate": 2.5986392119371297e-05,
"loss": 0.2468,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.3738305037341401,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.234,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.3427293244409305,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.2286,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.3529801717492802,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.2371,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.31549203642000156,
"learning_rate": 2.577016790136453e-05,
"loss": 0.2318,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.33271963063875465,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.2251,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.3622578887333546,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.2384,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.32194777854474277,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.237,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.3781467089889032,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.2259,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.3344504773479469,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.2403,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.3515909740501302,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.232,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.3432940202250912,
"learning_rate": 2.539001591449889e-05,
"loss": 0.2477,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.3309366531963324,
"learning_rate": 2.53355326151989e-05,
"loss": 0.2452,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.355743172709585,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.2293,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.3346672001869296,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.2455,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.3813221137704225,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.2482,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.32237081485799995,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.26,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.3805602680838594,
"learning_rate": 2.506248490185627e-05,
"loss": 0.2422,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.3295735083418356,
"learning_rate": 2.500775217241923e-05,
"loss": 0.2595,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.35234520021060123,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.2551,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.32518334761801393,
"learning_rate": 2.4898167025948e-05,
"loss": 0.2507,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.3254901407274687,
"learning_rate": 2.4843315485153703e-05,
"loss": 0.2456,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.31852189706593054,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.2495,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.31772011411900586,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.2423,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.31453245262963125,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.2398,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.3039334976344167,
"learning_rate": 2.46235264422994e-05,
"loss": 0.2361,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.30878369233169306,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.2519,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.33404148186501365,
"learning_rate": 2.451340834511592e-05,
"loss": 0.2491,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.3135455578285616,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.2478,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.294735318044109,
"learning_rate": 2.440314589184958e-05,
"loss": 0.2304,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.3346655856272017,
"learning_rate": 2.434796163357541e-05,
"loss": 0.2453,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.30017618594122314,
"learning_rate": 2.429274260911588e-05,
"loss": 0.241,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.3081202146585755,
"learning_rate": 2.423748926000084e-05,
"loss": 0.2382,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.3325677784508012,
"learning_rate": 2.418220202803459e-05,
"loss": 0.2535,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.32898054814811556,
"learning_rate": 2.412688135529237e-05,
"loss": 0.2431,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.33330718983584884,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.2385,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.32389634059205485,
"learning_rate": 2.4016141457114372e-05,
"loss": 0.2399,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.3331607264365844,
"learning_rate": 2.396072311715187e-05,
"loss": 0.2385,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.3862655028952796,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.2508,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.3136883605072788,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.257,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.37826626525618395,
"learning_rate": 2.379427985200223e-05,
"loss": 0.2498,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.3228144188242653,
"learning_rate": 2.373873749394972e-05,
"loss": 0.225,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.34207987944535234,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.25,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.35949856032689337,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.2576,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.32773600719498197,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.2223,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.31983579813592494,
"learning_rate": 2.351627355800233e-05,
"loss": 0.2401,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.34705219836141904,
"learning_rate": 2.346058617160256e-05,
"loss": 0.2423,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.31248798944147643,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.2535,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.34183499189778954,
"learning_rate": 2.334912883203469e-05,
"loss": 0.2523,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.30626021073690746,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.255,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.3324078202842087,
"learning_rate": 2.323756437450136e-05,
"loss": 0.2444,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.328592947048003,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.2381,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.2971328968290161,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.2241,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.34977494974491763,
"learning_rate": 2.307002464849052e-05,
"loss": 0.2574,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.3075015357954314,
"learning_rate": 2.30141283818846e-05,
"loss": 0.2432,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.33531999628494286,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.2549,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.34953022877664786,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.2421,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.36984242226320324,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.2395,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.3589232506093638,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.2674,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.337462348110498,
"learning_rate": 2.273429448107257e-05,
"loss": 0.2334,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.35514949770372367,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.251,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.3485219301863608,
"learning_rate": 2.262220474347102e-05,
"loss": 0.2491,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.3535718548976911,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.2461,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.3493279166921077,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.232,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.2991374231699166,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.2492,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.31680503332807897,
"learning_rate": 2.2397777251561302e-05,
"loss": 0.2576,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.3189019915321672,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.2379,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.2857808135763465,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.2422,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.3093449098662856,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.2379,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.31672432687594737,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.2394,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.3132346320093893,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.2603,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.3130996715117827,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.2389,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.28983390382875696,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.2408,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.35921494666575154,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.2633,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.32308100663699985,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.2351,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.3274214419158542,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.2456,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.33236891263730356,
"learning_rate": 2.17791064266051e-05,
"loss": 0.2481,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.3180459935683039,
"learning_rate": 2.172276926861741e-05,
"loss": 0.2595,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.3456573286224441,
"learning_rate": 2.166641833541255e-05,
"loss": 0.2391,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.32193718211688754,
"learning_rate": 2.161005407757109e-05,
"loss": 0.24,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.3237885835382109,
"learning_rate": 2.155367694578013e-05,
"loss": 0.2517,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.32155127047959225,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.2458,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.3154193890694552,
"learning_rate": 2.1440885863609208e-05,
"loss": 0.237,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.3390851290942148,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.2339,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.29020191273713586,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.2414,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.30620557744547,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.229,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.3141164217945169,
"learning_rate": 2.1215169053088825e-05,
"loss": 0.2532,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.291984055490793,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.239,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.32767719426428166,
"learning_rate": 2.110225054402733e-05,
"loss": 0.2432,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.32934855031319016,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.2339,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.2947929071755304,
"learning_rate": 2.09892967807724e-05,
"loss": 0.2386,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.3912091516400769,
"learning_rate": 2.093280780774835e-05,
"loss": 0.2511,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.2822417581495003,
"learning_rate": 2.087631137601793e-05,
"loss": 0.2382,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.33535926696489193,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.2443,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.3094066012163447,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.2316,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.30214090534569454,
"learning_rate": 2.070678184630458e-05,
"loss": 0.2497,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.28204660453772834,
"learning_rate": 2.06502600977305e-05,
"loss": 0.2384,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.28925745610055115,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.2572,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.2983758340021091,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.2416,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.29478728917316527,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.2462,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.3049852882706982,
"learning_rate": 2.042412562887594e-05,
"loss": 0.2494,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.2890309028561928,
"learning_rate": 2.0367582403231033e-05,
"loss": 0.2426,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.31258031760086036,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.2636,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.2938378538343625,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.237,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.2722126864670681,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.2408,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.2985623922450035,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.2464,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.2855816640641325,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.2522,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.2994843943865835,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.2542,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.31084843194144934,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.2391,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.28271206011025446,
"learning_rate": 1.99151687695709e-05,
"loss": 0.2619,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.29770106783898054,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.24,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.31525190029039757,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.2195,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.3061138720181529,
"learning_rate": 1.974551241345447e-05,
"loss": 0.2314,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.30286361447224125,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.2401,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.34556250139604267,
"learning_rate": 1.963241759676897e-05,
"loss": 0.2375,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.2807472196969102,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.2314,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.3016042469304362,
"learning_rate": 1.951933453677638e-05,
"loss": 0.2513,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.29143915129168,
"learning_rate": 1.9462798545816932e-05,
"loss": 0.2422,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.31636763551559194,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.2328,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.30353264555252535,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.2351,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.32028810048117023,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.2577,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.322777775607389,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.2514,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.30246180434247066,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.2141,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.2926831247938379,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.2406,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.28946264915705155,
"learning_rate": 1.906719219225165e-05,
"loss": 0.2499,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.309661432400513,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.2394,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.3142441048370252,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.238,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.27790778905002567,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.2483,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.3270429004034868,
"learning_rate": 1.884128556891707e-05,
"loss": 0.2546,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.29754535470443605,
"learning_rate": 1.878483094691119e-05,
"loss": 0.2347,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.3032599621095821,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.2485,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.30253571995151396,
"learning_rate": 1.867195130360947e-05,
"loss": 0.2212,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.32354442746690193,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.2273,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.3355357424052192,
"learning_rate": 1.8559114136390795e-05,
"loss": 0.2352,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.31850714592505763,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.2425,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.3235203004699312,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.259,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.30176776971350594,
"learning_rate": 1.838994592242891e-05,
"loss": 0.2373,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.3182448757560293,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.2468,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.31741306906145383,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.2496,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.31401299412192646,
"learning_rate": 1.82208935733949e-05,
"loss": 0.25,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.3067692542701142,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.23,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.3167455100214344,
"learning_rate": 1.810826238483892e-05,
"loss": 0.2238,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.3059547805479491,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.2587,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.31987984844982914,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.2551,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.2842014146153195,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.2306,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.29484984353530513,
"learning_rate": 1.788318512320838e-05,
"loss": 0.2265,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.33480302164235565,
"learning_rate": 1.782695699829051e-05,
"loss": 0.2454,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.29039797585559896,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.2352,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.3183957334875611,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.2365,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.2948673833459462,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.2513,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.30180318709226556,
"learning_rate": 1.76022227484387e-05,
"loss": 0.2533,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.29223464326180204,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.2718,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.28735440353891706,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.2442,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.30490741978586217,
"learning_rate": 1.743387180005809e-05,
"loss": 0.2405,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.3154135904506339,
"learning_rate": 1.737779525652899e-05,
"loss": 0.2489,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.26141025299047194,
"learning_rate": 1.732173968007914e-05,
"loss": 0.2384,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.28586837926417424,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.2305,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.3469082181470152,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.2415,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.26086731869174573,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.2566,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.29153458338996474,
"learning_rate": 1.709773600686015e-05,
"loss": 0.2576,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.31000248612276327,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.2382,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.28274143873661695,
"learning_rate": 1.69858716181154e-05,
"loss": 0.2275,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.2991632488254796,
"learning_rate": 1.692997535150948e-05,
"loss": 0.2504,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.288582482633689,
"learning_rate": 1.687410363273884e-05,
"loss": 0.2222,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.29553520117743454,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.2527,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.3268012002900916,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.2553,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.289028387947957,
"learning_rate": 1.67066402299235e-05,
"loss": 0.2413,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.28880452694891035,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.2385,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.32830677335127745,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.2357,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.29406258304428967,
"learning_rate": 1.653941382839745e-05,
"loss": 0.2395,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.39894847815346185,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.1869,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.33015450916430245,
"learning_rate": 1.642806717162757e-05,
"loss": 0.1813,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.4372260705548211,
"learning_rate": 1.637243646233718e-05,
"loss": 0.1836,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.5087990842179527,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.1643,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.3160519371075031,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.1747,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.4185505449674389,
"learning_rate": 1.620572014799777e-05,
"loss": 0.1694,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.4216278009257383,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.1954,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.33654584874550203,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.1657,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.33307135438415914,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.1544,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.3978588387390767,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.1575,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.3463829742513479,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.1782,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.33192674907758,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.1827,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.32928196531484005,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.1648,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.3277380147682354,
"learning_rate": 1.576251073999917e-05,
"loss": 0.1802,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.3355151135158476,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.1718,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.3015622986137986,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.1915,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.31179964375306063,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.1699,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.3078478932084736,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.1657,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.2979312151662141,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.1707,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.3280154924438686,
"learning_rate": 1.543151434154755e-05,
"loss": 0.1852,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.2886963631668464,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.1856,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.3056467295734517,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.1775,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.30392000082058407,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.1673,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.2688161529061672,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.1642,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.3093110401656012,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.1757,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.304299171619081,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.164,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.2793040663093504,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.1781,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.28229805420690407,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.1672,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.30415885205907806,
"learning_rate": 1.493751509814374e-05,
"loss": 0.1719,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.2809719966129507,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.1631,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.281127644682037,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.1484,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.28813380598818844,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.1749,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.2897180936100407,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.1586,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.3064192929757688,
"learning_rate": 1.466446738480111e-05,
"loss": 0.1687,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.2899171173565801,
"learning_rate": 1.460998408550112e-05,
"loss": 0.1973,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.27865665321166616,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.1549,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.28492303073228387,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.1767,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.3649569574589832,
"learning_rate": 1.444679451900437e-05,
"loss": 0.1732,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.28730055654054093,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.1767,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.27772272879183285,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.1857,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.2809115447435534,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.1594,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.29636358180875993,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.157,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.2702887416892754,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.1753,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.28757626569344574,
"learning_rate": 1.412162598317602e-05,
"loss": 0.1513,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.2793620640494234,
"learning_rate": 1.406759321421907e-05,
"loss": 0.1642,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.26572746488495474,
"learning_rate": 1.401360788062871e-05,
"loss": 0.1876,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.2835305675995708,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.1842,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.2821833212728209,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.1567,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.27647401429173957,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.1619,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.27319053064683674,
"learning_rate": 1.379814952747693e-05,
"loss": 0.1789,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.26056026864748894,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.156,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.281551552589962,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.1837,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.283472831985056,
"learning_rate": 1.363707494698459e-05,
"loss": 0.1738,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.27186163827357845,
"learning_rate": 1.358348460369265e-05,
"loss": 0.1985,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.2869786975238748,
"learning_rate": 1.352994556668684e-05,
"loss": 0.1672,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.27776519848291487,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.1664,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.27117547089992355,
"learning_rate": 1.342302312350678e-05,
"loss": 0.188,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.27232762638637603,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.1749,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.28804026130296156,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.1835,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.26447070411854556,
"learning_rate": 1.326303494478451e-05,
"loss": 0.1775,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.2768927043081837,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.1627,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.29045858844819133,
"learning_rate": 1.315664479122898e-05,
"loss": 0.189,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.2750008940369048,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.1883,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.2740630448989081,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.17,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.2812845431907478,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.1655,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.2889184838661304,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.1949,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.2812349509492053,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.1674,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.27858524582948324,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.1921,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.2901209704433316,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.1927,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.27522947179608304,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.1583,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.29017961611439036,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.18,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.2778019831727299,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.1981,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.2879001793771752,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.1884,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.2878668111634125,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.187,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.2805369522383092,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.1848,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.2651262514757175,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.1685,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.2875658420852622,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.1653,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.2610178324206058,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.1822,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.2671775863982204,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.1615,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.2658172640243069,
"learning_rate": 1.220938883250288e-05,
"loss": 0.1788,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.2619580678210443,
"learning_rate": 1.215733275116827e-05,
"loss": 0.1631,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.2644580419771806,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.1734,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.28554490563254975,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.1615,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.2597329451967679,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.1844,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.29432373471394696,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.1601,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.2699153072967803,
"learning_rate": 1.189800129803139e-05,
"loss": 0.1814,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.2989199418663044,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.1761,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.27222222058893814,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.1659,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.2702800907359734,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.1749,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.27184360765582927,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.1701,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.2869023156888822,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.1765,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.2678485221955412,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.1821,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.2767505917130381,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.1749,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.25823435423602764,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.1899,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.27573220190168096,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.1718,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.28191169759729695,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.1653,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.2768649848890759,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.1761,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.268529904743165,
"learning_rate": 1.128231163049258e-05,
"loss": 0.1761,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.26416205256256137,
"learning_rate": 1.123144751066259e-05,
"loss": 0.181,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.25784925967981126,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.1765,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.26246997518671444,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.1785,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.251492843888469,
"learning_rate": 1.107927745387114e-05,
"loss": 0.167,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.26390209843238094,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.1579,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.26477953809318294,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.1886,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.2730722785320716,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.1736,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.2657153039892178,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.166,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.2620974587600631,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.1694,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.26895829207591126,
"learning_rate": 1.077687414070869e-05,
"loss": 0.1778,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.2725880156204729,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.1877,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.27433507439480675,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.1649,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.2513881742094647,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.1829,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.2750194264247469,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.1783,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.2727429459117772,
"learning_rate": 1.052689572753247e-05,
"loss": 0.1777,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.2637916529443554,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.16,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.25745107233180237,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.176,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.26583780890232395,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.1792,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.2547178976400879,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.1756,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.24990288936683813,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.1843,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.25071093130590827,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.2015,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.26832385468646647,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.182,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.25531712283447255,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.1831,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.261896140019981,
"learning_rate": 1.008173988718005e-05,
"loss": 0.1787,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.27208496018542544,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.1744,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.26341674146215854,
"learning_rate": 9.983678617766677e-06,
"loss": 0.183,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.2528439894613671,
"learning_rate": 9.934767922553788e-06,
"loss": 0.1959,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.26230533105889536,
"learning_rate": 9.885937708668371e-06,
"loss": 0.1719,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.2515852508549631,
"learning_rate": 9.837188366555524e-06,
"loss": 0.1706,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.24882373864779825,
"learning_rate": 9.788520286013691e-06,
"loss": 0.178,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.25371385717749223,
"learning_rate": 9.739933856191537e-06,
"loss": 0.1802,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.26314076025434724,
"learning_rate": 9.691429465584866e-06,
"loss": 0.1727,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.26386669347712705,
"learning_rate": 9.643007502033492e-06,
"loss": 0.1748,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.25708883077533085,
"learning_rate": 9.594668352718152e-06,
"loss": 0.1766,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.285129553845346,
"learning_rate": 9.546412404157383e-06,
"loss": 0.1892,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.23893955271983072,
"learning_rate": 9.498240042204471e-06,
"loss": 0.195,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.2635777596719944,
"learning_rate": 9.450151652044339e-06,
"loss": 0.1868,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.28751825642746887,
"learning_rate": 9.402147618190472e-06,
"loss": 0.1759,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.25362423784220234,
"learning_rate": 9.354228324481841e-06,
"loss": 0.1842,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.25125984871096024,
"learning_rate": 9.306394154079852e-06,
"loss": 0.1727,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.28245004269949453,
"learning_rate": 9.258645489465254e-06,
"loss": 0.1687,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.2584656744160311,
"learning_rate": 9.21098271243509e-06,
"loss": 0.1893,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.24702780220929924,
"learning_rate": 9.163406204099659e-06,
"loss": 0.1774,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.26270059325215417,
"learning_rate": 9.115916344879452e-06,
"loss": 0.1701,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.2622229506205746,
"learning_rate": 9.068513514502126e-06,
"loss": 0.1717,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.26181539853625163,
"learning_rate": 9.021198091999439e-06,
"loss": 0.174,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.26257663380955765,
"learning_rate": 8.973970455704251e-06,
"loss": 0.1635,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.2794834665129006,
"learning_rate": 8.926830983247495e-06,
"loss": 0.1727,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.2501388125450285,
"learning_rate": 8.879780051555116e-06,
"loss": 0.186,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.25775841635553576,
"learning_rate": 8.83281803684512e-06,
"loss": 0.1544,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.2621130812991917,
"learning_rate": 8.785945314624532e-06,
"loss": 0.195,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.2542111032220385,
"learning_rate": 8.739162259686385e-06,
"loss": 0.1688,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.2519721061203662,
"learning_rate": 8.692469246106734e-06,
"loss": 0.1813,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.25550847644155406,
"learning_rate": 8.645866647241677e-06,
"loss": 0.1625,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.2688843552679665,
"learning_rate": 8.599354835724348e-06,
"loss": 0.1538,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.26367415933906446,
"learning_rate": 8.552934183461958e-06,
"loss": 0.1713,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.24772761795865314,
"learning_rate": 8.506605061632805e-06,
"loss": 0.1745,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.255767185663579,
"learning_rate": 8.460367840683308e-06,
"loss": 0.1699,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.2565993696669278,
"learning_rate": 8.414222890325065e-06,
"loss": 0.1756,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.261834334327355,
"learning_rate": 8.368170579531855e-06,
"loss": 0.1679,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.2750171059347311,
"learning_rate": 8.322211276536733e-06,
"loss": 0.1623,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.25963806005330614,
"learning_rate": 8.27634534882907e-06,
"loss": 0.1555,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.2566500886620149,
"learning_rate": 8.230573163151585e-06,
"loss": 0.1705,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.25751353938046717,
"learning_rate": 8.184895085497466e-06,
"loss": 0.171,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.24250148256530543,
"learning_rate": 8.139311481107401e-06,
"loss": 0.1936,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.2595512532199503,
"learning_rate": 8.093822714466686e-06,
"loss": 0.1752,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.2584586110163031,
"learning_rate": 8.048429149302271e-06,
"loss": 0.14,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.2466539664608372,
"learning_rate": 8.003131148579902e-06,
"loss": 0.1844,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.26818634900674015,
"learning_rate": 7.957929074501182e-06,
"loss": 0.1554,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.26668096924982587,
"learning_rate": 7.912823288500695e-06,
"loss": 0.1866,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.25161265484585693,
"learning_rate": 7.867814151243102e-06,
"loss": 0.1744,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.2547040417107977,
"learning_rate": 7.822902022620267e-06,
"loss": 0.1659,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.24843822378407113,
"learning_rate": 7.778087261748377e-06,
"loss": 0.1728,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.2614415132367202,
"learning_rate": 7.733370226965056e-06,
"loss": 0.1719,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.2500731268155325,
"learning_rate": 7.688751275826527e-06,
"loss": 0.1783,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.24074215675887362,
"learning_rate": 7.644230765104735e-06,
"loss": 0.1644,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.2614418651124604,
"learning_rate": 7.599809050784502e-06,
"loss": 0.1703,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.25765416652222056,
"learning_rate": 7.555486488060661e-06,
"loss": 0.1737,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.2581837693742814,
"learning_rate": 7.511263431335249e-06,
"loss": 0.1501,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.2528180116175525,
"learning_rate": 7.467140234214656e-06,
"loss": 0.1903,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.26006165526408054,
"learning_rate": 7.423117249506777e-06,
"loss": 0.1834,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.2675416223748372,
"learning_rate": 7.379194829218231e-06,
"loss": 0.1891,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.25686460227353336,
"learning_rate": 7.335373324551518e-06,
"loss": 0.1762,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.2541264195565479,
"learning_rate": 7.291653085902224e-06,
"loss": 0.1747,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.25764737062584503,
"learning_rate": 7.248034462856213e-06,
"loss": 0.1826,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.26702778730129784,
"learning_rate": 7.20451780418683e-06,
"loss": 0.1701,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.2519390204372131,
"learning_rate": 7.161103457852128e-06,
"loss": 0.1714,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.264714308008474,
"learning_rate": 7.117791770992046e-06,
"loss": 0.1778,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.2549501936151392,
"learning_rate": 7.074583089925688e-06,
"loss": 0.1758,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.2650616304676507,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.1649,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.2522083123964776,
"learning_rate": 6.988476126329604e-06,
"loss": 0.1738,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.24792909208780967,
"learning_rate": 6.945578532308857e-06,
"loss": 0.1603,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.26739403246879,
"learning_rate": 6.902785321094301e-06,
"loss": 0.1653,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.2460395490279645,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.1806,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.25408115910532353,
"learning_rate": 6.817513414939892e-06,
"loss": 0.1917,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.2643246094617065,
"learning_rate": 6.775035401831924e-06,
"loss": 0.1818,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.25795207095698003,
"learning_rate": 6.732663135188473e-06,
"loss": 0.1793,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.2513476317337617,
"learning_rate": 6.690396953817055e-06,
"loss": 0.1885,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.25562306640987287,
"learning_rate": 6.648237195676927e-06,
"loss": 0.1562,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.2590601494968054,
"learning_rate": 6.606184197876395e-06,
"loss": 0.1781,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.25133219545304275,
"learning_rate": 6.564238296670109e-06,
"loss": 0.1711,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.25609238121721195,
"learning_rate": 6.52239982745638e-06,
"loss": 0.1464,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.2532174360708129,
"learning_rate": 6.480669124774484e-06,
"loss": 0.1547,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.26123586364897294,
"learning_rate": 6.439046522302013e-06,
"loss": 0.1505,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.2390455561352138,
"learning_rate": 6.397532352852191e-06,
"loss": 0.1879,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.24518542152093414,
"learning_rate": 6.356126948371202e-06,
"loss": 0.1658,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.26677580615682783,
"learning_rate": 6.314830639935559e-06,
"loss": 0.1774,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.2497947916520558,
"learning_rate": 6.273643757749448e-06,
"loss": 0.1607,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.24920621652167663,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.1642,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.24646750478099852,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.1627,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.25684697162488174,
"learning_rate": 6.150742957589784e-06,
"loss": 0.1769,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.26043432049350046,
"learning_rate": 6.109997064904774e-06,
"loss": 0.1719,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.25431162673209956,
"learning_rate": 6.069362236313092e-06,
"loss": 0.179,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.24633367168223758,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.1764,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.25740024657063526,
"learning_rate": 5.988427070179108e-06,
"loss": 0.1695,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.2656965909708146,
"learning_rate": 5.948127379792234e-06,
"loss": 0.1834,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.2517400186391963,
"learning_rate": 5.90794004780439e-06,
"loss": 0.164,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.24120707594368365,
"learning_rate": 5.867865395552405e-06,
"loss": 0.1855,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.2525563097449222,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.1683,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.2718311934228082,
"learning_rate": 5.788055411095892e-06,
"loss": 0.1815,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.2716192710386246,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.1669,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.2451438555548243,
"learning_rate": 5.708699979051613e-06,
"loss": 0.1676,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.2535389332804559,
"learning_rate": 5.66919351390751e-06,
"loss": 0.1653,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.2637574646671685,
"learning_rate": 5.629801637510188e-06,
"loss": 0.1501,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.2569886860696604,
"learning_rate": 5.590524664836043e-06,
"loss": 0.1624,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.2567873245257835,
"learning_rate": 5.551362909942708e-06,
"loss": 0.1548,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.23887767312075966,
"learning_rate": 5.51231668596653e-06,
"loss": 0.1672,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.24246786580935623,
"learning_rate": 5.47338630512007e-06,
"loss": 0.1621,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.2451876424032211,
"learning_rate": 5.434572078689617e-06,
"loss": 0.181,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.24814961827034318,
"learning_rate": 5.395874317032694e-06,
"loss": 0.1639,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.25315457721099865,
"learning_rate": 5.357293329575573e-06,
"loss": 0.1741,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.25815357861634436,
"learning_rate": 5.318829424810797e-06,
"loss": 0.1959,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.26196394122276984,
"learning_rate": 5.280482910294728e-06,
"loss": 0.147,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.246502811142539,
"learning_rate": 5.242254092645078e-06,
"loss": 0.1828,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.2543910004724456,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.1559,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.24981452884828667,
"learning_rate": 5.166150769707892e-06,
"loss": 0.1968,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.24750696026870264,
"learning_rate": 5.128276872940503e-06,
"loss": 0.1593,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.2477497356524705,
"learning_rate": 5.090521890074953e-06,
"loss": 0.1592,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.25818737156885896,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.156,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.24571773005148817,
"learning_rate": 5.015369872647442e-06,
"loss": 0.1678,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.25355834156372986,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.1828,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.25387629627792807,
"learning_rate": 4.940697121074675e-06,
"loss": 0.1767,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.2605472420636793,
"learning_rate": 4.903541216934864e-06,
"loss": 0.17,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.23652225682261438,
"learning_rate": 4.8665060236772e-06,
"loss": 0.1757,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.243583880760791,
"learning_rate": 4.829591837434102e-06,
"loss": 0.1819,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.2470314882304363,
"learning_rate": 4.792798953370424e-06,
"loss": 0.1743,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.24046236681731445,
"learning_rate": 4.756127665681083e-06,
"loss": 0.1759,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.24254143650856963,
"learning_rate": 4.719578267588725e-06,
"loss": 0.1557,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.24666292388871763,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.1842,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.3067726111046041,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.148,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.27106667842967486,
"learning_rate": 4.610664328486485e-06,
"loss": 0.13,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.22242524999098146,
"learning_rate": 4.5746054014809e-06,
"loss": 0.1385,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.2241949040877905,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.1257,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.29456689253344714,
"learning_rate": 4.502857857942204e-06,
"loss": 0.1348,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.3090322931906911,
"learning_rate": 4.467169815100549e-06,
"loss": 0.137,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.28619149135915556,
"learning_rate": 4.431605972355137e-06,
"loss": 0.1191,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.2509824481341111,
"learning_rate": 4.396166614073516e-06,
"loss": 0.1237,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.24627621170159336,
"learning_rate": 4.360852023627846e-06,
"loss": 0.1512,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.2618792408051273,
"learning_rate": 4.325662483392656e-06,
"loss": 0.1156,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.28899813951890163,
"learning_rate": 4.29059827474257e-06,
"loss": 0.1551,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.2518195357049507,
"learning_rate": 4.255659678050063e-06,
"loss": 0.1277,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.2229656027186048,
"learning_rate": 4.22084697268323e-06,
"loss": 0.1155,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.22041846149850478,
"learning_rate": 4.186160437003535e-06,
"loss": 0.1343,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.23433520357100937,
"learning_rate": 4.151600348363582e-06,
"loss": 0.1315,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.24146141324001502,
"learning_rate": 4.11716698310493e-06,
"loss": 0.1302,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.23369897829249237,
"learning_rate": 4.082860616555848e-06,
"loss": 0.1043,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.24144056060993405,
"learning_rate": 4.048681523029125e-06,
"loss": 0.1233,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.20830564793811288,
"learning_rate": 4.014629975819882e-06,
"loss": 0.1375,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.21429700058235934,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.1273,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.219397918253273,
"learning_rate": 3.94691060843287e-06,
"loss": 0.142,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.22200292370910737,
"learning_rate": 3.913243329737353e-06,
"loss": 0.1308,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.22859384509714725,
"learning_rate": 3.879704680319503e-06,
"loss": 0.1335,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.2321369367624093,
"learning_rate": 3.846294928353471e-06,
"loss": 0.1525,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.2211107081800096,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.1055,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.20881560816821498,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.1423,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.21134722671434347,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.1402,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.218153533652802,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.1163,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.21876403894215776,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.1333,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.22641805985783178,
"learning_rate": 3.648558150665886e-06,
"loss": 0.1385,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.20119632222848247,
"learning_rate": 3.616058102895075e-06,
"loss": 0.1354,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.21772313930611323,
"learning_rate": 3.583689060687079e-06,
"loss": 0.1588,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.22063310819201978,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.126,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.21440765344348275,
"learning_rate": 3.51934502719796e-06,
"loss": 0.1194,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.22115085532181958,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.125,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.21532936726988472,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.1128,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.2156754769197773,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.1363,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.21918519044633225,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.125,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.21981819297704586,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.1235,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.20139055403797848,
"learning_rate": 3.32948376099375e-06,
"loss": 0.1312,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.21055256190832966,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.1285,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.2144013242308829,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.1263,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.2180501304174244,
"learning_rate": 3.236349233544438e-06,
"loss": 0.1281,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.20712906836047787,
"learning_rate": 3.205572144561604e-06,
"loss": 0.1241,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.20796157766931914,
"learning_rate": 3.174929343376374e-06,
"loss": 0.1358,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.2031485303870768,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.1421,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.21349475425011155,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.1236,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.2178732423151195,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.1321,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.21615387518516108,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.1329,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.20959256382130292,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.1253,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.20718998141571277,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.121,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.20979013923501422,
"learning_rate": 2.964210239224696e-06,
"loss": 0.1427,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.20281527109620182,
"learning_rate": 2.934650470499869e-06,
"loss": 0.139,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.22077872592562314,
"learning_rate": 2.905227155855128e-06,
"loss": 0.1286,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.20422959627972134,
"learning_rate": 2.875940530558519e-06,
"loss": 0.1257,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.21744179286481635,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.1484,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.2050041417691314,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.1493,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.20726366883363287,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.1358,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.2086213692893822,
"learning_rate": 2.760165589921184e-06,
"loss": 0.134,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.19785819973554497,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.1517,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.2010410627505197,
"learning_rate": 2.703104292151164e-06,
"loss": 0.1247,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.2268127605399557,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.1367,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.21042375280358847,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.1447,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.2123815349900318,
"learning_rate": 2.618550200289303e-06,
"loss": 0.1401,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.20369980662679144,
"learning_rate": 2.59064316684412e-06,
"loss": 0.1405,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.20931524413781727,
"learning_rate": 2.562875338152264e-06,
"loss": 0.1308,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.20658631708640873,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.1277,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.20637581147951234,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.1228,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.20169082464782384,
"learning_rate": 2.480409295328803e-06,
"loss": 0.1276,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.2141917640505506,
"learning_rate": 2.45320049480108e-06,
"loss": 0.1252,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.22155114316185076,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.1156,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.21289767605314677,
"learning_rate": 2.399204021408421e-06,
"loss": 0.1244,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.2132888080886385,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.1291,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.20448842457794067,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.1303,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.21524403196975717,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.1259,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.2242795726697733,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.1131,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.21542980259706718,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.1202,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.20783013167814665,
"learning_rate": 2.24059906434998e-06,
"loss": 0.1449,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.21040848544992125,
"learning_rate": 2.214660694699091e-06,
"loss": 0.143,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.22300492470583266,
"learning_rate": 2.188864536147355e-06,
"loss": 0.1213,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.2144193839629037,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.1251,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.20298384329813932,
"learning_rate": 2.137699676264171e-06,
"loss": 0.11,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.2149277485968832,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.111,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.2057409372942773,
"learning_rate": 2.08710612114857e-06,
"loss": 0.1281,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.2018924223608103,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.14,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.2155638599825007,
"learning_rate": 2.0370854889762e-06,
"loss": 0.1424,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.20107629390766418,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.1335,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.20746861748904233,
"learning_rate": 1.987639379598427e-06,
"loss": 0.1148,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.21284510522114647,
"learning_rate": 1.963132265887919e-06,
"loss": 0.1374,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.20738148515929708,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.1322,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.21002865462250062,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.1296,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.19923432671352437,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.1432,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.20555321973506924,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.1327,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.21058668788847232,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.1145,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.22001742792936668,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.1257,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.20697087356173136,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.1235,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.20665022089293963,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.1159,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.19775482778124295,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.1314,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.21253596219392,
"learning_rate": 1.726025117089718e-06,
"loss": 0.1562,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.20099962295800133,
"learning_rate": 1.70311497206598e-06,
"loss": 0.1199,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.21408544821929879,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.1228,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.22178899360464568,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.1495,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.21326924544703796,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.136,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.2119104081509518,
"learning_rate": 1.612939220355272e-06,
"loss": 0.1274,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.2112686627252628,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.1344,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.21030190099400253,
"learning_rate": 1.568732761074121e-06,
"loss": 0.1249,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.20906158931172933,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.1404,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.21873892611726561,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.1284,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.2162400251291063,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.1414,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.2100659123402026,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.1228,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.20911868538258682,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.1252,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.2213995680038394,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.1354,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.2213770647711708,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.132,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.21339067254621238,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.129,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.21243148525430364,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.1282,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.20235427037179135,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.1257,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.21532642691995488,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.1231,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.21343252074695887,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.1232,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.20589527244572312,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.1191,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.22035403100771425,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.1169,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.21102210015509182,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.12,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.22394692924121043,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.1382,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.20944497865111877,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.1419,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.2089852634284172,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.1155,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.19969509311383876,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.1274,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.2126842439304377,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.1333,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.20778729615963917,
"learning_rate": 1.140415161609505e-06,
"loss": 0.1238,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.2147904933518008,
"learning_rate": 1.121666393356604e-06,
"loss": 0.1324,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.20494880711690122,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.1138,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.2034479887000389,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.1338,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.20025018409386994,
"learning_rate": 1.066326386135097e-06,
"loss": 0.1328,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.19477111689607968,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.1232,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.2048581176382372,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.1362,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.21307009841799224,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.1244,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.2159456172918717,
"learning_rate": 9.9465987684501e-07,
"loss": 0.1254,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.20850509947016965,
"learning_rate": 9.771228098856534e-07,
"loss": 0.1126,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.20891083052519824,
"learning_rate": 9.597378493454013e-07,
"loss": 0.1335,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.21292042264333974,
"learning_rate": 9.425051342339353e-07,
"loss": 0.1397,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.21226941285877907,
"learning_rate": 9.254248023435864e-07,
"loss": 0.1182,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.2048198304110971,
"learning_rate": 9.084969902482288e-07,
"loss": 0.1253,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.20347234087574834,
"learning_rate": 8.917218333021993e-07,
"loss": 0.1188,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.22070778658229018,
"learning_rate": 8.750994656392109e-07,
"loss": 0.1294,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.2163964670835818,
"learning_rate": 8.58630020171265e-07,
"loss": 0.1094,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.21653164494654226,
"learning_rate": 8.423136285876099e-07,
"loss": 0.1236,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.2042626960244209,
"learning_rate": 8.26150421353682e-07,
"loss": 0.1356,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.20574994420760223,
"learning_rate": 8.101405277100549e-07,
"loss": 0.1467,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.2195990780767827,
"learning_rate": 7.942840756714077e-07,
"loss": 0.1367,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.20036397835415864,
"learning_rate": 7.785811920255093e-07,
"loss": 0.1371,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.21237772338233707,
"learning_rate": 7.630320023322024e-07,
"loss": 0.113,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.20530682581046705,
"learning_rate": 7.476366309223903e-07,
"loss": 0.1385,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.20173472845258583,
"learning_rate": 7.323952008970537e-07,
"loss": 0.1173,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.20458975157561937,
"learning_rate": 7.173078341262618e-07,
"loss": 0.1323,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.20583023964759317,
"learning_rate": 7.023746512481989e-07,
"loss": 0.1231,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.19830546513462263,
"learning_rate": 6.875957716681902e-07,
"loss": 0.136,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.20320531685714727,
"learning_rate": 6.729713135577687e-07,
"loss": 0.1438,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.21303377331335224,
"learning_rate": 6.585013938537143e-07,
"loss": 0.1394,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.2021969867429079,
"learning_rate": 6.441861282571115e-07,
"loss": 0.1265,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.20202836771562804,
"learning_rate": 6.300256312324471e-07,
"loss": 0.124,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.19963905564169562,
"learning_rate": 6.160200160066776e-07,
"loss": 0.1187,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.20712658170051748,
"learning_rate": 6.021693945683327e-07,
"loss": 0.1329,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.20192292792042715,
"learning_rate": 5.884738776666044e-07,
"loss": 0.141,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.20641111961569814,
"learning_rate": 5.749335748104878e-07,
"loss": 0.125,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.20531978428199998,
"learning_rate": 5.615485942678778e-07,
"loss": 0.1165,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.21292630791591158,
"learning_rate": 5.483190430647289e-07,
"loss": 0.1263,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.21513220629476415,
"learning_rate": 5.352450269841747e-07,
"loss": 0.1358,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.2097235913434023,
"learning_rate": 5.22326650565701e-07,
"loss": 0.141,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.206483442713284,
"learning_rate": 5.095640171043048e-07,
"loss": 0.1335,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.20696384757997704,
"learning_rate": 4.969572286496594e-07,
"loss": 0.1156,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.20223061388414837,
"learning_rate": 4.845063860053079e-07,
"loss": 0.1276,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.20785859739978593,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.1184,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.2097139181203632,
"learning_rate": 4.600729351261857e-07,
"loss": 0.1395,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.19644416037702825,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.1283,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.2165025169412115,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.1197,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.21988926180470583,
"learning_rate": 4.245948007320588e-07,
"loss": 0.1314,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.2096813257431988,
"learning_rate": 4.130816799401971e-07,
"loss": 0.1288,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.20153451800250333,
"learning_rate": 4.017251756252827e-07,
"loss": 0.1267,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.20951741861939202,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.1308,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.21810898817902305,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.135,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.2067086970996633,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.1188,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.20424285543824722,
"learning_rate": 3.578671204615347e-07,
"loss": 0.1201,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.21121388674141237,
"learning_rate": 3.472950355547267e-07,
"loss": 0.1284,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.21074835923542062,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.1235,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.20548573922875124,
"learning_rate": 3.266223765343224e-07,
"loss": 0.1286,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.1995361735500216,
"learning_rate": 3.16521967718757e-07,
"loss": 0.1379,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.20217165630165593,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.1262,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.1979344250881828,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.1357,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.21291159505002402,
"learning_rate": 2.871653894244597e-07,
"loss": 0.1428,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.20400400873765873,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.1187,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.21107547738909332,
"learning_rate": 2.683823233205485e-07,
"loss": 0.1231,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.2023815160648257,
"learning_rate": 2.592274132703287e-07,
"loss": 0.1381,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.21643329000979,
"learning_rate": 2.502303499116754e-07,
"loss": 0.1326,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.21565398807414407,
"learning_rate": 2.413912051848666e-07,
"loss": 0.1185,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.20642029746701682,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.1532,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.2132804934654202,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.1197,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.2043246907128604,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.1567,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.21279513680500808,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.1277,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.21193274859032946,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.137,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.2150872426185109,
"learning_rate": 1.916764975761254e-07,
"loss": 0.151,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.2062815389075316,
"learning_rate": 1.839446914621923e-07,
"loss": 0.1376,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.21336819304205684,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.1331,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.2005712279781133,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.136,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.2035851399266462,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.1409,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.21942599575865115,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.1335,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.20861280928199802,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.1061,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.20496534413295348,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.1407,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.2043382646342811,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.1256,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.20465944916255432,
"learning_rate": 1.277993209368744e-07,
"loss": 0.1268,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.19780269829729877,
"learning_rate": 1.214956261808542e-07,
"loss": 0.1479,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.2203786896164272,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.1177,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.20899799864613136,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.1243,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.20116295457212505,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.1217,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.20148094653345403,
"learning_rate": 9.787081204466253e-08,
"loss": 0.1298,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.2078805118039584,
"learning_rate": 9.236233586200627e-08,
"loss": 0.1163,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.2159197495756097,
"learning_rate": 8.70130406196279e-08,
"loss": 0.1307,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.2048141750634525,
"learning_rate": 8.182296909035047e-08,
"loss": 0.1181,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.21433617430855034,
"learning_rate": 7.679216277384439e-08,
"loss": 0.1274,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.2040633647203904,
"learning_rate": 7.19206618963053e-08,
"loss": 0.1301,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.21728403471005397,
"learning_rate": 6.720850541012347e-08,
"loss": 0.1566,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.19912246087287658,
"learning_rate": 6.265573099357714e-08,
"loss": 0.1127,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.21014063302870345,
"learning_rate": 5.826237505053067e-08,
"loss": 0.1317,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.19344767520902945,
"learning_rate": 5.4028472710137e-08,
"loss": 0.1281,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.21245163776811932,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.1286,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.2043331925048127,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.148,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.20079977326526083,
"learning_rate": 4.228381947004812e-08,
"loss": 0.1156,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.19921102211996622,
"learning_rate": 3.868805732811343e-08,
"loss": 0.1335,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.20782501850386606,
"learning_rate": 3.525190530455636e-08,
"loss": 0.1238,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.20671683499019766,
"learning_rate": 3.197539087475887e-08,
"loss": 0.1422,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.20468391556061208,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.124,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.2155792906923114,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.1327,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.20099964551735922,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.1248,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.2149455570158737,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.1319,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.19991749949346227,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.1466,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.21333398863842742,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.1238,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.20950885841456912,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.1239,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.2108228800573318,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.1368,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.21123510833519893,
"learning_rate": 9.67435416023843e-09,
"loss": 0.1282,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.20792312323843337,
"learning_rate": 7.995445897566178e-09,
"loss": 0.1113,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.2006087585865384,
"learning_rate": 6.476393170578021e-09,
"loss": 0.1254,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.20884627697837693,
"learning_rate": 5.11720812557348e-09,
"loss": 0.1251,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.20937584574593368,
"learning_rate": 3.917901630563492e-09,
"loss": 0.1389,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.20071502786918013,
"learning_rate": 2.878483275170485e-09,
"loss": 0.1104,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.2057609500835199,
"learning_rate": 1.998961370555108e-09,
"loss": 0.1269,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.205106381725266,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.1256,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.20958813791491865,
"learning_rate": 7.196337656112384e-10,
"loss": 0.1599,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.20182280434435412,
"learning_rate": 3.198382947511292e-10,
"loss": 0.1232,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.20407058126275684,
"learning_rate": 7.995973352770137e-11,
"loss": 0.1632,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.20312204552630886,
"learning_rate": 0.0,
"loss": 0.1275,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 1.9419526307753492e+18,
"train_loss": 0.25979115677627956,
"train_runtime": 23668.0281,
"train_samples_per_second": 6.676,
"train_steps_per_second": 0.052
}
],
"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": 1.9419526307753492e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}