openthoughts3_30k_llama3 / trainer_state.json
sedrickkeh's picture
End of training
e1f3c3d 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": 3.752959438972416,
"learning_rate": 3.2258064516129035e-07,
"loss": 1.4061,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 3.7268491291976384,
"learning_rate": 6.451612903225807e-07,
"loss": 1.4222,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 3.6893588278681277,
"learning_rate": 9.67741935483871e-07,
"loss": 1.4736,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 3.679680752043151,
"learning_rate": 1.2903225806451614e-06,
"loss": 1.4146,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 3.244457313603604,
"learning_rate": 1.6129032258064516e-06,
"loss": 1.4248,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 2.9765108131867306,
"learning_rate": 1.935483870967742e-06,
"loss": 1.3999,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 1.9961014797172008,
"learning_rate": 2.2580645161290324e-06,
"loss": 1.3916,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 1.9734445628271227,
"learning_rate": 2.580645161290323e-06,
"loss": 1.3595,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 3.77922008935092,
"learning_rate": 2.903225806451613e-06,
"loss": 1.3799,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 3.549395957309696,
"learning_rate": 3.225806451612903e-06,
"loss": 1.3565,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 3.011219180261707,
"learning_rate": 3.548387096774194e-06,
"loss": 1.3621,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 3.817862899599031,
"learning_rate": 3.870967741935484e-06,
"loss": 1.3284,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 3.554286263593431,
"learning_rate": 4.193548387096774e-06,
"loss": 1.3404,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 2.4567430292536185,
"learning_rate": 4.516129032258065e-06,
"loss": 1.2887,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 1.603044782240477,
"learning_rate": 4.838709677419355e-06,
"loss": 1.2824,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 2.0171879913118747,
"learning_rate": 5.161290322580646e-06,
"loss": 1.3232,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 1.8403765997254888,
"learning_rate": 5.483870967741935e-06,
"loss": 1.2855,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 1.53933172241602,
"learning_rate": 5.806451612903226e-06,
"loss": 1.2399,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 1.4569601446528218,
"learning_rate": 6.129032258064517e-06,
"loss": 1.2208,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.3969148242602856,
"learning_rate": 6.451612903225806e-06,
"loss": 1.2461,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 1.1081935155883884,
"learning_rate": 6.774193548387097e-06,
"loss": 1.2343,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.233266011462352,
"learning_rate": 7.096774193548388e-06,
"loss": 1.2205,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.1038978581857255,
"learning_rate": 7.4193548387096784e-06,
"loss": 1.2287,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 0.9247251019540264,
"learning_rate": 7.741935483870968e-06,
"loss": 1.2128,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 0.8402539188559813,
"learning_rate": 8.064516129032258e-06,
"loss": 1.1724,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 0.9181247621881261,
"learning_rate": 8.387096774193549e-06,
"loss": 1.1828,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 0.787768929377028,
"learning_rate": 8.70967741935484e-06,
"loss": 1.1819,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 0.8007818391104617,
"learning_rate": 9.03225806451613e-06,
"loss": 1.1794,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 0.9595255904242201,
"learning_rate": 9.35483870967742e-06,
"loss": 1.1699,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 0.6518302792300165,
"learning_rate": 9.67741935483871e-06,
"loss": 1.1947,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 0.8414049724866954,
"learning_rate": 1e-05,
"loss": 1.2008,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 0.9386997095558097,
"learning_rate": 1.0322580645161291e-05,
"loss": 1.1648,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 0.7719293329344594,
"learning_rate": 1.0645161290322582e-05,
"loss": 1.1961,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 0.9044623214441246,
"learning_rate": 1.096774193548387e-05,
"loss": 1.1193,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 1.096377117424554,
"learning_rate": 1.1290322580645164e-05,
"loss": 1.1709,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.6361882876079592,
"learning_rate": 1.1612903225806453e-05,
"loss": 1.166,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.6701211569346524,
"learning_rate": 1.1935483870967743e-05,
"loss": 1.1431,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.901473511152083,
"learning_rate": 1.2258064516129034e-05,
"loss": 1.1474,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.9159789764090605,
"learning_rate": 1.2580645161290324e-05,
"loss": 1.1402,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 0.7401457992759276,
"learning_rate": 1.2903225806451613e-05,
"loss": 1.1269,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.7659089234583752,
"learning_rate": 1.3225806451612903e-05,
"loss": 1.1078,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.8311668810265789,
"learning_rate": 1.3548387096774194e-05,
"loss": 1.0863,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.6589217824754495,
"learning_rate": 1.3870967741935486e-05,
"loss": 1.1183,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.7399038409956371,
"learning_rate": 1.4193548387096776e-05,
"loss": 1.0588,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 1.2516152198370565,
"learning_rate": 1.4516129032258066e-05,
"loss": 1.1269,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 1.0184195726590284,
"learning_rate": 1.4838709677419357e-05,
"loss": 1.1,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.824815945745295,
"learning_rate": 1.5161290322580646e-05,
"loss": 1.1364,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.7343554559193773,
"learning_rate": 1.5483870967741936e-05,
"loss": 1.1515,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 0.9296155530197808,
"learning_rate": 1.5806451612903226e-05,
"loss": 1.1217,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 0.9837168735863181,
"learning_rate": 1.6129032258064517e-05,
"loss": 1.1028,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 1.2737062986145689,
"learning_rate": 1.6451612903225807e-05,
"loss": 1.1053,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 1.1840317227479622,
"learning_rate": 1.6774193548387098e-05,
"loss": 1.102,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 0.8420719558423372,
"learning_rate": 1.7096774193548388e-05,
"loss": 1.0957,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.8350767050699407,
"learning_rate": 1.741935483870968e-05,
"loss": 1.0796,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.9487742736488837,
"learning_rate": 1.774193548387097e-05,
"loss": 1.1065,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 1.55170444710603,
"learning_rate": 1.806451612903226e-05,
"loss": 1.1046,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.859596675646451,
"learning_rate": 1.838709677419355e-05,
"loss": 1.1028,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 0.9018672554626425,
"learning_rate": 1.870967741935484e-05,
"loss": 1.1517,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 1.1813868049633511,
"learning_rate": 1.903225806451613e-05,
"loss": 1.0563,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 1.4824316808237559,
"learning_rate": 1.935483870967742e-05,
"loss": 1.1076,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 0.9771632980758486,
"learning_rate": 1.967741935483871e-05,
"loss": 1.084,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 1.5016521411608943,
"learning_rate": 2e-05,
"loss": 1.1651,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 0.9383917998454525,
"learning_rate": 2.0322580645161292e-05,
"loss": 1.1153,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 1.3644476416956035,
"learning_rate": 2.0645161290322582e-05,
"loss": 1.0912,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 1.4786007891329551,
"learning_rate": 2.0967741935483873e-05,
"loss": 1.0715,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 1.3340211043026113,
"learning_rate": 2.1290322580645163e-05,
"loss": 1.0794,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 1.028927962815125,
"learning_rate": 2.161290322580645e-05,
"loss": 1.1244,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 1.5774174296614722,
"learning_rate": 2.193548387096774e-05,
"loss": 1.1195,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 1.0632215563491927,
"learning_rate": 2.225806451612903e-05,
"loss": 1.1032,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 1.39678179567259,
"learning_rate": 2.2580645161290328e-05,
"loss": 1.0998,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 1.2591817639334302,
"learning_rate": 2.2903225806451615e-05,
"loss": 1.0819,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 1.3075009500069308,
"learning_rate": 2.3225806451612906e-05,
"loss": 1.1142,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 1.1419086631348105,
"learning_rate": 2.3548387096774196e-05,
"loss": 1.0623,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 1.2875127903828805,
"learning_rate": 2.3870967741935486e-05,
"loss": 1.1017,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 1.696247598381462,
"learning_rate": 2.4193548387096777e-05,
"loss": 1.0761,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 0.8663339121997229,
"learning_rate": 2.4516129032258067e-05,
"loss": 1.1003,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 1.4893048888950842,
"learning_rate": 2.4838709677419358e-05,
"loss": 1.0919,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 1.179386416891394,
"learning_rate": 2.5161290322580648e-05,
"loss": 1.1014,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 1.6076133988724988,
"learning_rate": 2.548387096774194e-05,
"loss": 1.0859,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 1.047709420434774,
"learning_rate": 2.5806451612903226e-05,
"loss": 1.0935,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 1.5028556739713093,
"learning_rate": 2.6129032258064516e-05,
"loss": 1.0872,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 1.3547675132348425,
"learning_rate": 2.6451612903225806e-05,
"loss": 1.1309,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 1.132419982656028,
"learning_rate": 2.6774193548387097e-05,
"loss": 1.0779,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 1.1871317257540637,
"learning_rate": 2.7096774193548387e-05,
"loss": 1.061,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 1.7992248465050493,
"learning_rate": 2.7419354838709678e-05,
"loss": 1.0692,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 1.2014209799583615,
"learning_rate": 2.774193548387097e-05,
"loss": 1.0694,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 1.3416791567282729,
"learning_rate": 2.8064516129032262e-05,
"loss": 1.0823,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 1.1372325409739106,
"learning_rate": 2.8387096774193552e-05,
"loss": 1.0749,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 1.2973129940009573,
"learning_rate": 2.8709677419354843e-05,
"loss": 1.0827,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 1.315404849775449,
"learning_rate": 2.9032258064516133e-05,
"loss": 1.1118,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 1.3091311376897046,
"learning_rate": 2.9354838709677423e-05,
"loss": 1.0534,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 1.605670189648343,
"learning_rate": 2.9677419354838714e-05,
"loss": 1.106,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 0.9376399602095846,
"learning_rate": 3.0000000000000004e-05,
"loss": 1.109,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 1.2591918999969884,
"learning_rate": 3.032258064516129e-05,
"loss": 1.1041,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 1.8408037239151243,
"learning_rate": 3.0645161290322585e-05,
"loss": 1.1146,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 0.9791101045195995,
"learning_rate": 3.096774193548387e-05,
"loss": 1.0774,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 1.4251394349096158,
"learning_rate": 3.1290322580645166e-05,
"loss": 1.0902,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 1.240968341157319,
"learning_rate": 3.161290322580645e-05,
"loss": 1.0745,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 1.4220159780591295,
"learning_rate": 3.193548387096774e-05,
"loss": 1.09,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 1.5750519129671539,
"learning_rate": 3.2258064516129034e-05,
"loss": 1.111,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 1.0202423452321008,
"learning_rate": 3.258064516129033e-05,
"loss": 1.1037,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 1.3793767525858727,
"learning_rate": 3.2903225806451614e-05,
"loss": 1.1176,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 1.5320904447318824,
"learning_rate": 3.322580645161291e-05,
"loss": 1.0492,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 1.4154863562253057,
"learning_rate": 3.3548387096774195e-05,
"loss": 1.0644,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 1.2458089252435143,
"learning_rate": 3.387096774193549e-05,
"loss": 1.0407,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 0.9413869880441236,
"learning_rate": 3.4193548387096776e-05,
"loss": 1.0808,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 1.4099061182026424,
"learning_rate": 3.451612903225807e-05,
"loss": 1.0746,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 1.6120627178978781,
"learning_rate": 3.483870967741936e-05,
"loss": 1.0643,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 1.0236225829637595,
"learning_rate": 3.516129032258065e-05,
"loss": 1.0857,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 1.3989099282060649,
"learning_rate": 3.548387096774194e-05,
"loss": 1.1173,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 1.1653311517278075,
"learning_rate": 3.5806451612903225e-05,
"loss": 1.0823,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 1.6228884419239336,
"learning_rate": 3.612903225806452e-05,
"loss": 1.1073,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 1.2277689595896975,
"learning_rate": 3.6451612903225805e-05,
"loss": 1.0658,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 1.9378290610733515,
"learning_rate": 3.67741935483871e-05,
"loss": 1.0876,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 1.0693274549885663,
"learning_rate": 3.7096774193548386e-05,
"loss": 1.0829,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 1.4753220829385782,
"learning_rate": 3.741935483870968e-05,
"loss": 1.0885,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 1.2584652384713295,
"learning_rate": 3.7741935483870974e-05,
"loss": 1.0714,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 1.532884960026014,
"learning_rate": 3.806451612903226e-05,
"loss": 1.0267,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 1.1897373962246984,
"learning_rate": 3.8387096774193555e-05,
"loss": 1.0649,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 1.2539050964324556,
"learning_rate": 3.870967741935484e-05,
"loss": 1.0804,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 1.180401357379282,
"learning_rate": 3.9032258064516135e-05,
"loss": 1.093,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 1.3755169002485002,
"learning_rate": 3.935483870967742e-05,
"loss": 1.0635,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 1.676464556456474,
"learning_rate": 3.9677419354838716e-05,
"loss": 1.0765,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 1.3664128599952763,
"learning_rate": 4e-05,
"loss": 1.0919,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 1.065141619367559,
"learning_rate": 3.9999920040266474e-05,
"loss": 1.0337,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 1.5924995948904237,
"learning_rate": 3.999968016170525e-05,
"loss": 1.0862,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 1.016918464565714,
"learning_rate": 3.9999280366234394e-05,
"loss": 1.1125,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 1.4767928104223127,
"learning_rate": 3.999872065705065e-05,
"loss": 1.0495,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 1.5547492204763982,
"learning_rate": 3.999800103862945e-05,
"loss": 1.0713,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 1.1832398429539992,
"learning_rate": 3.9997121516724834e-05,
"loss": 1.1148,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 1.4276004721805138,
"learning_rate": 3.999608209836944e-05,
"loss": 1.1035,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 1.342906118097333,
"learning_rate": 3.999488279187443e-05,
"loss": 1.0894,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 1.057562567992674,
"learning_rate": 3.9993523606829425e-05,
"loss": 1.0324,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 1.261295606219989,
"learning_rate": 3.9992004554102435e-05,
"loss": 1.0854,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 1.1814047231217457,
"learning_rate": 3.999032564583976e-05,
"loss": 1.0152,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 1.2592954006845456,
"learning_rate": 3.998848689546591e-05,
"loss": 1.0816,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 1.5717285397610978,
"learning_rate": 3.9986488317683496e-05,
"loss": 1.0479,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 1.1346660906796935,
"learning_rate": 3.9984329928473067e-05,
"loss": 1.0726,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 1.3844548239538155,
"learning_rate": 3.998201174509306e-05,
"loss": 1.0202,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 1.0855860684961525,
"learning_rate": 3.997953378607961e-05,
"loss": 1.0936,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 1.852963759678287,
"learning_rate": 3.9976896071246403e-05,
"loss": 1.0702,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 1.429307414459608,
"learning_rate": 3.997409862168454e-05,
"loss": 1.1271,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 1.1870269384544023,
"learning_rate": 3.997114145976237e-05,
"loss": 1.0589,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 1.673652588197742,
"learning_rate": 3.9968024609125245e-05,
"loss": 1.0777,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 1.1941556854899844,
"learning_rate": 3.996474809469545e-05,
"loss": 1.0041,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 1.1472161731369859,
"learning_rate": 3.996131194267189e-05,
"loss": 1.0734,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 1.2970288188726777,
"learning_rate": 3.995771618052995e-05,
"loss": 1.0849,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 1.4842373021134827,
"learning_rate": 3.9953960837021256e-05,
"loss": 1.0548,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 2.2243486281448983,
"learning_rate": 3.995004594217343e-05,
"loss": 1.1068,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 1.22541422334223,
"learning_rate": 3.994597152728987e-05,
"loss": 1.0591,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 2.097994866744288,
"learning_rate": 3.994173762494947e-05,
"loss": 1.0954,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 1.1142882119471116,
"learning_rate": 3.9937344269006424e-05,
"loss": 1.0761,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 2.621774296931014,
"learning_rate": 3.993279149458988e-05,
"loss": 1.1198,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 1.717806185374537,
"learning_rate": 3.9928079338103695e-05,
"loss": 1.1106,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 2.0732189891398995,
"learning_rate": 3.992320783722616e-05,
"loss": 1.0586,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 1.4100278489135758,
"learning_rate": 3.991817703090965e-05,
"loss": 1.0269,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 1.732897474997514,
"learning_rate": 3.991298695938038e-05,
"loss": 1.0582,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 1.1999612838217735,
"learning_rate": 3.9907637664138e-05,
"loss": 1.0651,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 1.4862529345897726,
"learning_rate": 3.990212918795534e-05,
"loss": 1.1014,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 1.0406982619032485,
"learning_rate": 3.9896461574878036e-05,
"loss": 1.0712,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 1.2821890308008497,
"learning_rate": 3.989063487022416e-05,
"loss": 1.0838,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 1.1945730260729124,
"learning_rate": 3.9884649120583903e-05,
"loss": 1.0209,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 1.3044993043847728,
"learning_rate": 3.987850437381915e-05,
"loss": 1.0303,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 1.1306816102782196,
"learning_rate": 3.987220067906313e-05,
"loss": 1.0501,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 1.0527612006746976,
"learning_rate": 3.986573808672002e-05,
"loss": 1.1022,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 1.2314573563954718,
"learning_rate": 3.9859116648464534e-05,
"loss": 1.0904,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 1.0128941437936831,
"learning_rate": 3.985233641724152e-05,
"loss": 1.0697,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 0.9884003998055189,
"learning_rate": 3.984539744726554e-05,
"loss": 1.0481,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 1.1501690285322046,
"learning_rate": 3.983829979402038e-05,
"loss": 1.0516,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 1.5907518685950726,
"learning_rate": 3.983104351425871e-05,
"loss": 1.0924,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 0.6924924752597376,
"learning_rate": 3.982362866600154e-05,
"loss": 1.0554,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 0.915256688322165,
"learning_rate": 3.9816055308537806e-05,
"loss": 1.0757,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 1.5789482633749723,
"learning_rate": 3.980832350242388e-05,
"loss": 1.0538,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 0.9367348516878612,
"learning_rate": 3.980043330948306e-05,
"loss": 1.0317,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 1.24332020421418,
"learning_rate": 3.9792384792805124e-05,
"loss": 1.0862,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 0.8508632339671997,
"learning_rate": 3.978417801674579e-05,
"loss": 1.0379,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 0.9871567428269461,
"learning_rate": 3.9775813046926244e-05,
"loss": 1.0695,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 0.9370230852234833,
"learning_rate": 3.976728995023254e-05,
"loss": 1.0731,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 0.8268409926504391,
"learning_rate": 3.975860879481514e-05,
"loss": 1.0504,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 0.8927588560001212,
"learning_rate": 3.974976965008833e-05,
"loss": 1.0834,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 0.9622829330524794,
"learning_rate": 3.9740772586729674e-05,
"loss": 1.0695,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 1.1464161931943029,
"learning_rate": 3.973161767667946e-05,
"loss": 1.0652,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 1.3142897507435019,
"learning_rate": 3.972230499314009e-05,
"loss": 1.0716,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 1.3534527456044512,
"learning_rate": 3.9712834610575545e-05,
"loss": 1.0039,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 0.9426132547050626,
"learning_rate": 3.9703206604710746e-05,
"loss": 1.0918,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 1.1480765519017113,
"learning_rate": 3.969342105253098e-05,
"loss": 1.0452,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 1.2593571160115096,
"learning_rate": 3.9683478032281246e-05,
"loss": 1.0885,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 0.7059198902011251,
"learning_rate": 3.967337762346568e-05,
"loss": 1.0316,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 0.9135622758150101,
"learning_rate": 3.9663119906846885e-05,
"loss": 1.0549,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 1.2789379005722097,
"learning_rate": 3.965270496444528e-05,
"loss": 1.0141,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 0.7376614825354809,
"learning_rate": 3.9642132879538465e-05,
"loss": 1.0493,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 1.0480312019808358,
"learning_rate": 3.9631403736660566e-05,
"loss": 1.0473,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 1.1909566551727322,
"learning_rate": 3.962051762160151e-05,
"loss": 1.0603,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 1.0758104679832694,
"learning_rate": 3.960947462140638e-05,
"loss": 1.0009,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 1.1391724826428788,
"learning_rate": 3.959827482437472e-05,
"loss": 1.0278,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 0.8896313542810236,
"learning_rate": 3.958691832005981e-05,
"loss": 1.1015,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 0.8779803305533661,
"learning_rate": 3.9575405199267944e-05,
"loss": 1.0449,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 1.3835178459710449,
"learning_rate": 3.956373555405774e-05,
"loss": 1.0745,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 1.0976198095919676,
"learning_rate": 3.955190947773937e-05,
"loss": 1.0537,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 0.812079126258225,
"learning_rate": 3.9539927064873824e-05,
"loss": 1.1026,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 0.717601806254551,
"learning_rate": 3.952778841127214e-05,
"loss": 1.0643,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 0.8883792017326673,
"learning_rate": 3.95154936139947e-05,
"loss": 1.0376,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 0.983806784714299,
"learning_rate": 3.950304277135034e-05,
"loss": 1.0276,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 1.0575422933324445,
"learning_rate": 3.94904359828957e-05,
"loss": 1.0951,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 0.923096718807064,
"learning_rate": 3.94776733494343e-05,
"loss": 1.0574,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 0.8500938624156756,
"learning_rate": 3.9464754973015826e-05,
"loss": 1.0496,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 0.9559247420064049,
"learning_rate": 3.9451680956935275e-05,
"loss": 1.0196,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 1.1338343788710807,
"learning_rate": 3.943845140573213e-05,
"loss": 1.0349,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.9147938233096559,
"learning_rate": 3.942506642518952e-05,
"loss": 1.0368,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 0.8804933804091483,
"learning_rate": 3.9411526122333396e-05,
"loss": 1.022,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 1.0233823206858603,
"learning_rate": 3.939783060543167e-05,
"loss": 1.0247,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 1.3292597950800504,
"learning_rate": 3.938397998399333e-05,
"loss": 1.0063,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.9078670697193829,
"learning_rate": 3.936997436876756e-05,
"loss": 1.0378,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.6982593260580877,
"learning_rate": 3.935581387174289e-05,
"loss": 1.0359,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 0.6038786884349373,
"learning_rate": 3.9341498606146295e-05,
"loss": 1.0676,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.6513994092868409,
"learning_rate": 3.932702868644224e-05,
"loss": 1.0121,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.5676354741389011,
"learning_rate": 3.931240422833181e-05,
"loss": 1.0445,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.6730410251309107,
"learning_rate": 3.9297625348751806e-05,
"loss": 1.0392,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 1.14826866906872,
"learning_rate": 3.9282692165873744e-05,
"loss": 1.0306,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 1.069031340815696,
"learning_rate": 3.9267604799102945e-05,
"loss": 1.0665,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 1.0414448691352143,
"learning_rate": 3.9252363369077615e-05,
"loss": 1.0231,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 1.0002665725566804,
"learning_rate": 3.9236967997667796e-05,
"loss": 1.047,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 1.1562879719544041,
"learning_rate": 3.9221418807974496e-05,
"loss": 1.0632,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 0.9556154753121082,
"learning_rate": 3.92057159243286e-05,
"loss": 1.0204,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 0.8805943780044758,
"learning_rate": 3.9189859472289956e-05,
"loss": 1.0247,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 0.9022490198118436,
"learning_rate": 3.9173849578646324e-05,
"loss": 1.0285,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 1.1189045201959997,
"learning_rate": 3.9157686371412395e-05,
"loss": 1.016,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.9241330832721575,
"learning_rate": 3.914136997982874e-05,
"loss": 1.0409,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 0.6780335008088698,
"learning_rate": 3.912490053436079e-05,
"loss": 1.0349,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 0.7023697010942589,
"learning_rate": 3.9108278166697805e-05,
"loss": 1.065,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.8996229223522556,
"learning_rate": 3.909150300975178e-05,
"loss": 0.99,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 0.9742342579277666,
"learning_rate": 3.907457519765642e-05,
"loss": 1.0954,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.8223658612113403,
"learning_rate": 3.905749486576607e-05,
"loss": 1.0086,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.9434066766318637,
"learning_rate": 3.90402621506546e-05,
"loss": 1.077,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 1.2571391461353194,
"learning_rate": 3.902287719011435e-05,
"loss": 1.0528,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.7020413539724916,
"learning_rate": 3.900534012315499e-05,
"loss": 1.0295,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 0.7794096274140773,
"learning_rate": 3.8987651090002454e-05,
"loss": 1.0441,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 0.9452758489920909,
"learning_rate": 3.896981023209777e-05,
"loss": 1.058,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 0.9389254965518409,
"learning_rate": 3.8951817692095966e-05,
"loss": 1.0246,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 1.1981050278690553,
"learning_rate": 3.893367361386491e-05,
"loss": 1.0177,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 1.081457935932546,
"learning_rate": 3.891537814248417e-05,
"loss": 1.0298,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 0.9058293883719528,
"learning_rate": 3.889693142424384e-05,
"loss": 1.0242,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 1.0021187052136538,
"learning_rate": 3.8878333606643405e-05,
"loss": 1.0595,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 0.9849138798822306,
"learning_rate": 3.885958483839049e-05,
"loss": 1.0476,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 1.0244012356493093,
"learning_rate": 3.8840685269399786e-05,
"loss": 1.0693,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 1.0563174955384478,
"learning_rate": 3.882163505079171e-05,
"loss": 1.0901,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 0.9838609990653884,
"learning_rate": 3.880243433489132e-05,
"loss": 1.0265,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 0.9818256341978502,
"learning_rate": 3.878308327522702e-05,
"loss": 0.9993,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 1.0000011071329078,
"learning_rate": 3.876358202652936e-05,
"loss": 1.0417,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 1.0673925484799462,
"learning_rate": 3.8743930744729835e-05,
"loss": 1.0109,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.8349302290922611,
"learning_rate": 3.8724129586959535e-05,
"loss": 0.9955,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.7553470316352939,
"learning_rate": 3.870417871154801e-05,
"loss": 0.9682,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 0.666610841200986,
"learning_rate": 3.868407827802192e-05,
"loss": 1.0135,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.6125825745404474,
"learning_rate": 3.8663828447103804e-05,
"loss": 0.953,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.7305320555242723,
"learning_rate": 3.8643429380710765e-05,
"loss": 0.9404,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.6883469486476439,
"learning_rate": 3.862288124195319e-05,
"loss": 1.025,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.5643276222732962,
"learning_rate": 3.8602184195133464e-05,
"loss": 0.9609,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 0.5847250673632353,
"learning_rate": 3.8581338405744604e-05,
"loss": 0.9409,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 0.6799083064872957,
"learning_rate": 3.8560344040468996e-05,
"loss": 0.9818,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.7055951734146976,
"learning_rate": 3.853920126717703e-05,
"loss": 0.9711,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 0.6639580359073468,
"learning_rate": 3.851791025492574e-05,
"loss": 0.9826,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.6430720350018264,
"learning_rate": 3.8496471173957515e-05,
"loss": 0.9808,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.7298021853371872,
"learning_rate": 3.847488419569866e-05,
"loss": 0.9902,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.6912447982740161,
"learning_rate": 3.8453149492758077e-05,
"loss": 1.001,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 0.7010132406830067,
"learning_rate": 3.8431267238925886e-05,
"loss": 0.9601,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.9118395172203796,
"learning_rate": 3.840923760917199e-05,
"loss": 1.0337,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 1.1071917655952503,
"learning_rate": 3.838706077964473e-05,
"loss": 0.9713,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 1.060363103747742,
"learning_rate": 3.836473692766944e-05,
"loss": 1.0045,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 0.9458340493654479,
"learning_rate": 3.834226623174705e-05,
"loss": 0.9904,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.9561973742907937,
"learning_rate": 3.831964887155264e-05,
"loss": 0.9901,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.9681250146517684,
"learning_rate": 3.829688502793403e-05,
"loss": 0.987,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.9065155324153669,
"learning_rate": 3.827397488291029e-05,
"loss": 1.0608,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.8916646634333553,
"learning_rate": 3.8250918619670343e-05,
"loss": 0.9897,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.7700879349750972,
"learning_rate": 3.822771642257145e-05,
"loss": 0.9902,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.6818165111149527,
"learning_rate": 3.820436847713776e-05,
"loss": 1.0152,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 0.7221433158386124,
"learning_rate": 3.8180874970058826e-05,
"loss": 0.9664,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.8212574022250483,
"learning_rate": 3.8157236089188103e-05,
"loss": 0.9754,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 1.0053212072970001,
"learning_rate": 3.813345202354145e-05,
"loss": 0.966,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 1.1070859519297958,
"learning_rate": 3.810952296329563e-05,
"loss": 0.9889,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.8358166253941451,
"learning_rate": 3.808544909978676e-05,
"loss": 0.9868,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.6924774927875122,
"learning_rate": 3.806123062550882e-05,
"loss": 1.0032,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.6080332883770533,
"learning_rate": 3.803686773411208e-05,
"loss": 0.982,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.5680048355104554,
"learning_rate": 3.801236062040158e-05,
"loss": 0.9706,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 0.548368433870711,
"learning_rate": 3.798770948033553e-05,
"loss": 0.9907,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.7026437819198584,
"learning_rate": 3.796291451102381e-05,
"loss": 0.9954,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.8345070136240147,
"learning_rate": 3.793797591072631e-05,
"loss": 0.9868,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 1.0186881485545558,
"learning_rate": 3.791289387885144e-05,
"loss": 0.9931,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 1.0225920904516472,
"learning_rate": 3.788766861595443e-05,
"loss": 0.9486,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 0.8129488740679961,
"learning_rate": 3.7862300323735835e-05,
"loss": 0.9513,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 0.8043451187756199,
"learning_rate": 3.783678920503983e-05,
"loss": 0.9739,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 0.946235681453711,
"learning_rate": 3.781113546385265e-05,
"loss": 0.9865,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.5195756186922322,
"learning_rate": 3.7785339305300916e-05,
"loss": 0.992,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.6054290360004044,
"learning_rate": 3.7759400935650025e-05,
"loss": 0.9727,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.7085038277517843,
"learning_rate": 3.77333205623025e-05,
"loss": 0.9559,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.8176542533384601,
"learning_rate": 3.7707098393796295e-05,
"loss": 1.0134,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.9885935553012307,
"learning_rate": 3.7680734639803185e-05,
"loss": 0.9662,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 1.0548708413586283,
"learning_rate": 3.765422951112704e-05,
"loss": 0.9881,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 0.8155844161559442,
"learning_rate": 3.762758321970216e-05,
"loss": 1.0115,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.5435689607351407,
"learning_rate": 3.7600795978591584e-05,
"loss": 0.9885,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.5499916070990403,
"learning_rate": 3.757386800198538e-05,
"loss": 0.9781,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.7579013700158448,
"learning_rate": 3.7546799505198925e-05,
"loss": 0.9908,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.8316043080885248,
"learning_rate": 3.7519590704671197e-05,
"loss": 0.9756,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.7050873614097113,
"learning_rate": 3.749224181796305e-05,
"loss": 1.0039,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.5316797128336919,
"learning_rate": 3.746475306375545e-05,
"loss": 1.0037,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.48736264692453546,
"learning_rate": 3.743712466184774e-05,
"loss": 0.9935,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.56580671700632,
"learning_rate": 3.7409356833155885e-05,
"loss": 1.0196,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.6433855322255358,
"learning_rate": 3.73814497997107e-05,
"loss": 0.9961,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.6390112173503845,
"learning_rate": 3.7353403784656095e-05,
"loss": 0.981,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.5742867144667388,
"learning_rate": 3.732521901224724e-05,
"loss": 1.0052,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.550800986964586,
"learning_rate": 3.7296895707848844e-05,
"loss": 1.0017,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.661097442919019,
"learning_rate": 3.7268434097933275e-05,
"loss": 0.9571,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.7281501680976977,
"learning_rate": 3.723983441007882e-05,
"loss": 1.0033,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.6152643267332816,
"learning_rate": 3.721109687296781e-05,
"loss": 0.9595,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.6417006890707692,
"learning_rate": 3.718222171638484e-05,
"loss": 1.0079,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 1.9318617841976309,
"learning_rate": 3.715320917121488e-05,
"loss": 0.9972,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.8405434955272019,
"learning_rate": 3.712405946944149e-05,
"loss": 0.9997,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 1.5387650035579035,
"learning_rate": 3.709477284414488e-05,
"loss": 0.9957,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.8897235594055258,
"learning_rate": 3.7065349529500136e-05,
"loss": 0.9743,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 1.2550944998849483,
"learning_rate": 3.7035789760775305e-05,
"loss": 0.9515,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.6798468573635309,
"learning_rate": 3.70060937743295e-05,
"loss": 0.9905,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.9844010588228468,
"learning_rate": 3.6976261807611045e-05,
"loss": 0.9893,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.8446405882340238,
"learning_rate": 3.694629409915555e-05,
"loss": 1.0113,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 1.122513247391353,
"learning_rate": 3.691619088858401e-05,
"loss": 1.0164,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.8892579589334574,
"learning_rate": 3.68859524166009e-05,
"loss": 0.9917,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.6588099178476312,
"learning_rate": 3.6855578924992226e-05,
"loss": 0.9962,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.8017235039916527,
"learning_rate": 3.6825070656623626e-05,
"loss": 1.0013,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.9286155631134279,
"learning_rate": 3.67944278554384e-05,
"loss": 0.978,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 0.9825851905692686,
"learning_rate": 3.676365076645557e-05,
"loss": 1.0221,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.9537041782660471,
"learning_rate": 3.673273963576791e-05,
"loss": 0.9977,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.7948368343307717,
"learning_rate": 3.670169471054001e-05,
"loss": 0.9624,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.615165034407675,
"learning_rate": 3.6670516239006254e-05,
"loss": 1.0156,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.5953568998677415,
"learning_rate": 3.663920447046888e-05,
"loss": 1.0495,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.5570809898443418,
"learning_rate": 3.6607759655295954e-05,
"loss": 1.0219,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.5897657861238078,
"learning_rate": 3.657618204491937e-05,
"loss": 1.0093,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.8519852007876482,
"learning_rate": 3.6544471891832875e-05,
"loss": 1.0305,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.7280594720149118,
"learning_rate": 3.6512629449589983e-05,
"loss": 1.0185,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.6013111162170628,
"learning_rate": 3.6480654972802045e-05,
"loss": 1.0111,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.48802367624463455,
"learning_rate": 3.644854871713611e-05,
"loss": 0.9805,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.5536204831880953,
"learning_rate": 3.6416310939312924e-05,
"loss": 1.0261,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.5942295316408076,
"learning_rate": 3.638394189710493e-05,
"loss": 0.9737,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.6218301497467428,
"learning_rate": 3.635144184933412e-05,
"loss": 0.9991,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.5875056929701024,
"learning_rate": 3.6318811055869994e-05,
"loss": 0.9812,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.5094266564191617,
"learning_rate": 3.628604977762752e-05,
"loss": 0.9777,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.45066212207110246,
"learning_rate": 3.6253158276565004e-05,
"loss": 1.0425,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.4995474547803434,
"learning_rate": 3.6220136815682016e-05,
"loss": 0.9799,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.5115616146082715,
"learning_rate": 3.618698565901727e-05,
"loss": 1.0021,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.4785306122844305,
"learning_rate": 3.615370507164654e-05,
"loss": 0.9851,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.5088379495934384,
"learning_rate": 3.61202953196805e-05,
"loss": 0.9974,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 0.5674851639181496,
"learning_rate": 3.6086756670262654e-05,
"loss": 0.9905,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.563734330250393,
"learning_rate": 3.6053089391567135e-05,
"loss": 0.9842,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.546539740383158,
"learning_rate": 3.601929375279662e-05,
"loss": 0.9538,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.4570459524263264,
"learning_rate": 3.598537002418012e-05,
"loss": 1.0174,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.47314951109732656,
"learning_rate": 3.5951318476970885e-05,
"loss": 0.9566,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.47791440397433166,
"learning_rate": 3.591713938344416e-05,
"loss": 0.9814,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.5282003144217979,
"learning_rate": 3.588283301689507e-05,
"loss": 0.9902,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 0.5114972379633558,
"learning_rate": 3.5848399651636424e-05,
"loss": 1.0216,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.43452297670788237,
"learning_rate": 3.5813839562996473e-05,
"loss": 0.9931,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.5157321095837728,
"learning_rate": 3.5779153027316775e-05,
"loss": 1.02,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.5628341435074773,
"learning_rate": 3.574434032194994e-05,
"loss": 1.0087,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.4485207576155541,
"learning_rate": 3.5709401725257436e-05,
"loss": 0.9916,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.4970580992602166,
"learning_rate": 3.5674337516607346e-05,
"loss": 0.9732,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.5580638710455011,
"learning_rate": 3.563914797637216e-05,
"loss": 0.9609,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.5660142348026367,
"learning_rate": 3.560383338592649e-05,
"loss": 0.9887,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.6678901011684348,
"learning_rate": 3.556839402764487e-05,
"loss": 0.9822,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.7088245310723994,
"learning_rate": 3.553283018489946e-05,
"loss": 0.9909,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.7434661833008189,
"learning_rate": 3.54971421420578e-05,
"loss": 0.978,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.751354775893553,
"learning_rate": 3.5461330184480524e-05,
"loss": 1.0096,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 0.7447447693174901,
"learning_rate": 3.54253945985191e-05,
"loss": 0.9889,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.7038537732930645,
"learning_rate": 3.5389335671513524e-05,
"loss": 0.9651,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.7146274357578744,
"learning_rate": 3.5353153691789986e-05,
"loss": 0.9998,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.6931037543878953,
"learning_rate": 3.531684894865866e-05,
"loss": 0.9828,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.6183877484388208,
"learning_rate": 3.528042173241128e-05,
"loss": 0.9638,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.5707503419066859,
"learning_rate": 3.524387233431892e-05,
"loss": 0.9912,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.48584056705453255,
"learning_rate": 3.520720104662958e-05,
"loss": 0.9263,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.6102604428227488,
"learning_rate": 3.5170408162565904e-05,
"loss": 0.9782,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.8657561691981681,
"learning_rate": 3.513349397632281e-05,
"loss": 0.9831,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.9985502938365051,
"learning_rate": 3.5096458783065145e-05,
"loss": 0.9924,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.9258986047149581,
"learning_rate": 3.505930287892533e-05,
"loss": 0.9306,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.7660403754141846,
"learning_rate": 3.502202656100099e-05,
"loss": 0.996,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.5364759833894386,
"learning_rate": 3.498463012735256e-05,
"loss": 0.9511,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.4811808784508476,
"learning_rate": 3.494711387700094e-05,
"loss": 1.0344,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.7147131628475951,
"learning_rate": 3.490947810992505e-05,
"loss": 0.9644,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 0.9087602265324034,
"learning_rate": 3.48717231270595e-05,
"loss": 0.9872,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.8356958351123597,
"learning_rate": 3.483384923029211e-05,
"loss": 1.0099,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.6440333914914466,
"learning_rate": 3.479585672246156e-05,
"loss": 1.043,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.5712212765775649,
"learning_rate": 3.4757745907354924e-05,
"loss": 0.9937,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.5295352119002253,
"learning_rate": 3.471951708970527e-05,
"loss": 0.9727,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.658373251644469,
"learning_rate": 3.468117057518921e-05,
"loss": 0.9232,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.8208251162014409,
"learning_rate": 3.4642706670424426e-05,
"loss": 0.9537,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.8638296849611784,
"learning_rate": 3.460412568296731e-05,
"loss": 0.9462,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.7587662583059781,
"learning_rate": 3.456542792131039e-05,
"loss": 0.9345,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.6607056261962304,
"learning_rate": 3.4526613694879936e-05,
"loss": 1.0038,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.5820970015030855,
"learning_rate": 3.448768331403348e-05,
"loss": 0.986,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.5696532626943559,
"learning_rate": 3.44486370900573e-05,
"loss": 0.9619,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.6471665086434109,
"learning_rate": 3.440947533516396e-05,
"loss": 1.0037,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.6579297410069151,
"learning_rate": 3.4370198362489816e-05,
"loss": 0.9811,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 0.4769525670839716,
"learning_rate": 3.43308064860925e-05,
"loss": 0.9895,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.4393698285842185,
"learning_rate": 3.4291300020948393e-05,
"loss": 0.9862,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.4814416264709961,
"learning_rate": 3.425167928295015e-05,
"loss": 1.0055,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.5321128260455452,
"learning_rate": 3.421194458890411e-05,
"loss": 0.9635,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.45299924725159224,
"learning_rate": 3.4172096256527865e-05,
"loss": 0.8987,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.44888765302178546,
"learning_rate": 3.4132134604447595e-05,
"loss": 0.9921,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.4559788630984437,
"learning_rate": 3.409205995219561e-05,
"loss": 0.9825,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.41221344523271375,
"learning_rate": 3.4051872620207765e-05,
"loss": 0.9886,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.420330404236502,
"learning_rate": 3.40115729298209e-05,
"loss": 1.0313,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.4178005665369989,
"learning_rate": 3.397116120327025e-05,
"loss": 0.9622,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.4306697063229711,
"learning_rate": 3.3930637763686916e-05,
"loss": 0.9899,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.4764008133278496,
"learning_rate": 3.389000293509524e-05,
"loss": 1.0002,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.4860505898700285,
"learning_rate": 3.3849257042410225e-05,
"loss": 0.9806,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.4811642265956119,
"learning_rate": 3.380840041143494e-05,
"loss": 0.9401,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.5122231044077027,
"learning_rate": 3.376743336885792e-05,
"loss": 0.9697,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.5567599218530843,
"learning_rate": 3.372635624225055e-05,
"loss": 0.992,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.5775513640167085,
"learning_rate": 3.3685169360064445e-05,
"loss": 0.9819,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.6243146389051534,
"learning_rate": 3.3643873051628804e-05,
"loss": 0.9785,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.7373893520909308,
"learning_rate": 3.3602467647147814e-05,
"loss": 0.9895,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.7389679512953781,
"learning_rate": 3.356095347769799e-05,
"loss": 0.9989,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.6634665201843225,
"learning_rate": 3.351933087522552e-05,
"loss": 0.9645,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.5486697336661424,
"learning_rate": 3.3477600172543626e-05,
"loss": 0.9523,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.4537522562138538,
"learning_rate": 3.3435761703329894e-05,
"loss": 0.979,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.6172309567617814,
"learning_rate": 3.3393815802123606e-05,
"loss": 0.9828,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.6207251417985161,
"learning_rate": 3.335176280432308e-05,
"loss": 0.9512,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.5722930910273645,
"learning_rate": 3.330960304618295e-05,
"loss": 0.9824,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.5279078383806912,
"learning_rate": 3.326733686481153e-05,
"loss": 0.9202,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.5526727851424164,
"learning_rate": 3.3224964598168084e-05,
"loss": 1.0004,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.5004095096076575,
"learning_rate": 3.318248658506012e-05,
"loss": 0.9661,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.49058263745886743,
"learning_rate": 3.313990316514069e-05,
"loss": 0.9908,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.6226606934277205,
"learning_rate": 3.309721467890571e-05,
"loss": 1.0039,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.6544250330120043,
"learning_rate": 3.305442146769115e-05,
"loss": 0.998,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.5427012553780731,
"learning_rate": 3.30115238736704e-05,
"loss": 1.0074,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.44797210294950957,
"learning_rate": 3.296852223985148e-05,
"loss": 0.9472,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 0.5423397339225748,
"learning_rate": 3.2925416910074315e-05,
"loss": 0.9767,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.6120444122709845,
"learning_rate": 3.288220822900796e-05,
"loss": 0.9715,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.602657061428114,
"learning_rate": 3.283889654214788e-05,
"loss": 0.959,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.6896593974047599,
"learning_rate": 3.279548219581317e-05,
"loss": 0.9588,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.8116639110633531,
"learning_rate": 3.275196553714379e-05,
"loss": 0.9923,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.8479263174441652,
"learning_rate": 3.270834691409778e-05,
"loss": 0.9763,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.6479397979878955,
"learning_rate": 3.266462667544849e-05,
"loss": 0.9775,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.4693590132215904,
"learning_rate": 3.262080517078178e-05,
"loss": 0.9868,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.6990181949530473,
"learning_rate": 3.257688275049323e-05,
"loss": 0.989,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.6985550169270033,
"learning_rate": 3.253285976578535e-05,
"loss": 0.9848,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 1.8014863618307404,
"learning_rate": 3.2488736568664756e-05,
"loss": 1.0031,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.7011623249291962,
"learning_rate": 3.244451351193935e-05,
"loss": 0.9914,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 1.217271717197215,
"learning_rate": 3.240019094921551e-05,
"loss": 0.9635,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.5812997052917468,
"learning_rate": 3.235576923489527e-05,
"loss": 0.975,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.7907650424590001,
"learning_rate": 3.2311248724173476e-05,
"loss": 1.0198,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 1.0074184454501751,
"learning_rate": 3.226662977303494e-05,
"loss": 0.9695,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.8118487927074812,
"learning_rate": 3.2221912738251626e-05,
"loss": 0.9782,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.8657019304092993,
"learning_rate": 3.2177097977379736e-05,
"loss": 1.0188,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.7178445535131357,
"learning_rate": 3.21321858487569e-05,
"loss": 0.9766,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.8076278218621119,
"learning_rate": 3.208717671149931e-05,
"loss": 0.9928,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.8179791785329524,
"learning_rate": 3.204207092549882e-05,
"loss": 0.9366,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 0.7301536874751084,
"learning_rate": 3.19968688514201e-05,
"loss": 0.9468,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.5425496018685252,
"learning_rate": 3.195157085069774e-05,
"loss": 1.0047,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.5404078659749875,
"learning_rate": 3.190617728553332e-05,
"loss": 1.0406,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.6454294738138822,
"learning_rate": 3.1860688518892606e-05,
"loss": 0.9191,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.6967933002254915,
"learning_rate": 3.181510491450254e-05,
"loss": 0.9877,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.625039263839292,
"learning_rate": 3.176942683684842e-05,
"loss": 0.9809,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.5775225244410825,
"learning_rate": 3.1723654651170934e-05,
"loss": 1.005,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.485973937678743,
"learning_rate": 3.167778872346327e-05,
"loss": 0.9868,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.5762168467207328,
"learning_rate": 3.163182942046815e-05,
"loss": 0.9793,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.5557433333464239,
"learning_rate": 3.158577710967494e-05,
"loss": 0.974,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.458280124890063,
"learning_rate": 3.153963215931669e-05,
"loss": 0.9758,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.49370406017063617,
"learning_rate": 3.14933949383672e-05,
"loss": 0.991,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.4840670815057065,
"learning_rate": 3.144706581653805e-05,
"loss": 0.9954,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.503764823628867,
"learning_rate": 3.140064516427566e-05,
"loss": 0.9987,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.46958508986798375,
"learning_rate": 3.135413335275833e-05,
"loss": 0.998,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.5221528926695,
"learning_rate": 3.130753075389327e-05,
"loss": 0.9723,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.5729664331248164,
"learning_rate": 3.1260837740313625e-05,
"loss": 1.011,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.5767875147110665,
"learning_rate": 3.121405468537547e-05,
"loss": 0.983,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.6790597320326259,
"learning_rate": 3.116718196315488e-05,
"loss": 0.9318,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.6694654117173116,
"learning_rate": 3.1120219948444894e-05,
"loss": 0.9653,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.5329244250025319,
"learning_rate": 3.107316901675252e-05,
"loss": 1.0193,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.4275960900699499,
"learning_rate": 3.1026029544295745e-05,
"loss": 0.9471,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.4736510311036021,
"learning_rate": 3.097880190800057e-05,
"loss": 0.9951,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.5316771766079021,
"learning_rate": 3.093148648549788e-05,
"loss": 1.0277,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.46014070963696374,
"learning_rate": 3.088408365512055e-05,
"loss": 0.9416,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.417148452735237,
"learning_rate": 3.083659379590034e-05,
"loss": 0.991,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.42499863884077277,
"learning_rate": 3.0789017287564913e-05,
"loss": 0.9996,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.4497006375856358,
"learning_rate": 3.074135451053475e-05,
"loss": 0.9356,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.44386877326015617,
"learning_rate": 3.069360584592015e-05,
"loss": 1.0151,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.4396406267737719,
"learning_rate": 3.064577167551816e-05,
"loss": 0.9591,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.5099561131382756,
"learning_rate": 3.059785238180953e-05,
"loss": 1.0076,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.4643196651195298,
"learning_rate": 3.054984834795566e-05,
"loss": 0.9797,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.35826274240266187,
"learning_rate": 3.0501759957795537e-05,
"loss": 0.9837,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.5431260664341061,
"learning_rate": 3.0453587595842627e-05,
"loss": 0.9803,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.43748015909042837,
"learning_rate": 3.0405331647281856e-05,
"loss": 0.9478,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.3844108701372895,
"learning_rate": 3.0356992497966508e-05,
"loss": 0.9468,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.41583631527816306,
"learning_rate": 3.0308570534415137e-05,
"loss": 0.9536,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.4381298032649685,
"learning_rate": 3.0260066143808467e-05,
"loss": 0.9586,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.41244714217249645,
"learning_rate": 3.021147971398632e-05,
"loss": 1.0163,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.4212948285531601,
"learning_rate": 3.0162811633444478e-05,
"loss": 1.0316,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.39896300201325285,
"learning_rate": 3.0114062291331635e-05,
"loss": 0.9573,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.3768505670640723,
"learning_rate": 3.0065232077446217e-05,
"loss": 0.9619,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.5313582030436435,
"learning_rate": 3.001632138223332e-05,
"loss": 1.0003,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.6884163216505367,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.9194,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 3.037973168686491,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.9651,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 1.3281799845190698,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.8914,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 1.3645959499123963,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.9385,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.7048851257615878,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.9083,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 1.059546409998008,
"learning_rate": 2.972118904869376e-05,
"loss": 0.9088,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.9650439205638497,
"learning_rate": 2.9671725970312563e-05,
"loss": 0.9622,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.7121496809731486,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.8824,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.8666257196273133,
"learning_rate": 2.95725682050846e-05,
"loss": 0.9011,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.7158907394411279,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.9341,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.6780463440961704,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.8667,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.5732381690987731,
"learning_rate": 2.942325848714505e-05,
"loss": 0.8991,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.6115776410766879,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.9451,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.6243369645096739,
"learning_rate": 2.932334127129686e-05,
"loss": 0.9536,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.6301663376865223,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.876,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.521651436072975,
"learning_rate": 2.922312585929131e-05,
"loss": 0.8826,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.5605744986812924,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.8811,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.5835848210225023,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.8587,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.6871199816296313,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.875,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.7178557448492745,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.8868,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.6700182847476721,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.9288,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.5100325056056121,
"learning_rate": 2.8920722546128863e-05,
"loss": 0.9176,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.476257743633832,
"learning_rate": 2.887006998351372e-05,
"loss": 0.9265,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.47845896406646965,
"learning_rate": 2.881934649605535e-05,
"loss": 0.895,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.584637069831058,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.9262,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.5263692667189204,
"learning_rate": 2.8717688369507425e-05,
"loss": 0.8994,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.4322287233192028,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.9389,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.42393323646210773,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.8945,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.44198405682131336,
"learning_rate": 2.856467940121025e-05,
"loss": 0.9193,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.42092819550369964,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.917,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.40493918043405513,
"learning_rate": 2.846233032790149e-05,
"loss": 0.9073,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.40001623092399985,
"learning_rate": 2.84110540896642e-05,
"loss": 0.8767,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.4100524312263103,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.9239,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.4659166010919262,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.9363,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.38447991017698474,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.8662,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.3941643307975717,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.895,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.40709983488994955,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.93,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.3812379869776182,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.9014,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.45861419865586284,
"learning_rate": 2.805026032173773e-05,
"loss": 0.9198,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.4663238027759,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.8923,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.42961509117736263,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.9327,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.42805096410129556,
"learning_rate": 2.7894660620408e-05,
"loss": 0.9112,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.46282367703899474,
"learning_rate": 2.784266724883173e-05,
"loss": 0.891,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.49715856262416125,
"learning_rate": 2.7790611167497128e-05,
"loss": 0.8978,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.4282181617601264,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.908,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.48084484825161,
"learning_rate": 2.768631254100716e-05,
"loss": 0.9338,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.4657555244374793,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.9099,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.4637885942465652,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.8883,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.44979585818916923,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.8836,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.37188390211689587,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.8931,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.4232546920681972,
"learning_rate": 2.742449775134085e-05,
"loss": 0.9322,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.46436637920342944,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.916,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.4385624151088659,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.9358,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.3789409750614994,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.8956,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.41461193204990826,
"learning_rate": 2.7213974824975785e-05,
"loss": 0.8856,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.45486144995272615,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.8919,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.4968368874835962,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.916,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.3974466982263039,
"learning_rate": 2.705547548999996e-05,
"loss": 0.9133,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.3863272774634849,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.9079,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.7591464483941917,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.8746,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.5892980931598595,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.9194,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.542701442461637,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.8911,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.4371959045207093,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.9371,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.44561662763941234,
"learning_rate": 2.67369650552155e-05,
"loss": 0.9291,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.5419332903936236,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.9334,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.5318367625307961,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.8921,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.41796034344989846,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.8874,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.36998348910775614,
"learning_rate": 2.652354173593614e-05,
"loss": 0.911,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.40067637356798136,
"learning_rate": 2.647005443331317e-05,
"loss": 0.9253,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.4804851704791345,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.9222,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.4403539450259624,
"learning_rate": 2.636292505301542e-05,
"loss": 0.9257,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.43762306450587585,
"learning_rate": 2.630928383194431e-05,
"loss": 0.8821,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.43872434286480333,
"learning_rate": 2.625559216200781e-05,
"loss": 0.9472,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.38055587991801937,
"learning_rate": 2.620185047252308e-05,
"loss": 0.9248,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.453742617744559,
"learning_rate": 2.614805919320723e-05,
"loss": 0.907,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.425975112860076,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.8924,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.39443886876228373,
"learning_rate": 2.60403295859298e-05,
"loss": 0.8935,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.39676489046093155,
"learning_rate": 2.5986392119371297e-05,
"loss": 0.923,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.35476038220425654,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.9201,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.37707641861849817,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.9263,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.3817915090174972,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.8684,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.38305796897895683,
"learning_rate": 2.577016790136453e-05,
"loss": 0.8904,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.4235507153217303,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.9588,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.38413859118226,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.8989,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.385688407725466,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.9032,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.39491900188225976,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.9174,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.4085395737050213,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.9122,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.40380757306340564,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.9417,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.38641175636914227,
"learning_rate": 2.539001591449889e-05,
"loss": 0.9293,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.42998310596627193,
"learning_rate": 2.53355326151989e-05,
"loss": 0.9123,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.4162201882771508,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.9204,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.36987400120739566,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.9149,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.42750861244871435,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.9108,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.4355623278909556,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.8954,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.37326096132619163,
"learning_rate": 2.506248490185627e-05,
"loss": 0.8849,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.4256817672755169,
"learning_rate": 2.500775217241923e-05,
"loss": 0.9602,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.44409693493774427,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.894,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.4153056695439231,
"learning_rate": 2.4898167025948e-05,
"loss": 0.9223,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.39726559480429086,
"learning_rate": 2.4843315485153703e-05,
"loss": 0.9321,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.7396614215012653,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.9061,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.4045311617540596,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.9264,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.40547559813417194,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.9311,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.7874163575170446,
"learning_rate": 2.46235264422994e-05,
"loss": 0.9311,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.4008859637711039,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.9102,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.379110463290193,
"learning_rate": 2.451340834511592e-05,
"loss": 0.8995,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.41550017262461036,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.9367,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.43550444968965935,
"learning_rate": 2.440314589184958e-05,
"loss": 0.9025,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.40283003630128356,
"learning_rate": 2.434796163357541e-05,
"loss": 0.9291,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 1.4719367802366705,
"learning_rate": 2.429274260911588e-05,
"loss": 0.9313,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.43753635250374323,
"learning_rate": 2.423748926000084e-05,
"loss": 0.9151,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.41781270309563895,
"learning_rate": 2.418220202803459e-05,
"loss": 0.8728,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.41708365943846176,
"learning_rate": 2.412688135529237e-05,
"loss": 0.9432,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.4322936331155281,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.9276,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.3820874199485479,
"learning_rate": 2.4016141457114372e-05,
"loss": 0.8708,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.408336489884189,
"learning_rate": 2.396072311715187e-05,
"loss": 0.9023,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.3946649064243332,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.8821,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.385571400665883,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.9531,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.39335731665781143,
"learning_rate": 2.379427985200223e-05,
"loss": 0.899,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.5241078275986305,
"learning_rate": 2.373873749394972e-05,
"loss": 0.9202,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.40823479859382555,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.9229,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.3809661510698757,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.9388,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.4228804263527677,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.9358,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.4158426832051447,
"learning_rate": 2.351627355800233e-05,
"loss": 0.8872,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.37883796482056137,
"learning_rate": 2.346058617160256e-05,
"loss": 0.8872,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.38266796089158667,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.9521,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.3766597472360932,
"learning_rate": 2.334912883203469e-05,
"loss": 0.9109,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.3854499439107829,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.9622,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.4690604107699915,
"learning_rate": 2.323756437450136e-05,
"loss": 0.8895,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.3953634171025897,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.8911,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.3512368985135435,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.9208,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.3567020032382442,
"learning_rate": 2.307002464849052e-05,
"loss": 0.8867,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.3594340101857233,
"learning_rate": 2.30141283818846e-05,
"loss": 0.9137,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.35272659981987703,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.9691,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.33124026161677117,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.8988,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.3605849492642791,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.9138,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.33275363366471944,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.9373,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.36015997844307385,
"learning_rate": 2.273429448107257e-05,
"loss": 0.9091,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.3776529986861501,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.8678,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.41082285628956927,
"learning_rate": 2.262220474347102e-05,
"loss": 0.8967,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.3716586878270196,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.9209,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.4119028447776786,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.913,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.4014732993576428,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.9122,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 2.6136597960405985,
"learning_rate": 2.2397777251561302e-05,
"loss": 0.9389,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.51913086567641,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.9629,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.5194807734222375,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.891,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.4415937475486327,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.9152,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.5445256083571617,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.9633,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.4192595695600564,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.9124,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.49727495676506744,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.8907,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.4674547295522453,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.8992,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.42710333377921433,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.9012,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.4723442091672315,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.9077,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.39443415787582464,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.8936,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.47079569899670365,
"learning_rate": 2.17791064266051e-05,
"loss": 0.9329,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.3339031318565018,
"learning_rate": 2.172276926861741e-05,
"loss": 0.8988,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.4638276189863681,
"learning_rate": 2.166641833541255e-05,
"loss": 0.9166,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.40620420483121494,
"learning_rate": 2.161005407757109e-05,
"loss": 0.9264,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.40694883991829156,
"learning_rate": 2.155367694578013e-05,
"loss": 0.891,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.3577036276780823,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.9157,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.4792730233943608,
"learning_rate": 2.1440885863609208e-05,
"loss": 0.9062,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.39903637552788745,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.9496,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.4099034601601157,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.8746,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.5439656217128491,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.8974,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.44664340414051723,
"learning_rate": 2.1215169053088825e-05,
"loss": 0.8715,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.35680068154746963,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.8973,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.495610399240754,
"learning_rate": 2.110225054402733e-05,
"loss": 0.899,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.4479365689484607,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.8816,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.33149536899176496,
"learning_rate": 2.09892967807724e-05,
"loss": 0.9019,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.45178605362109386,
"learning_rate": 2.093280780774835e-05,
"loss": 0.9103,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.46208524567870013,
"learning_rate": 2.087631137601793e-05,
"loss": 0.8665,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.34667361779972816,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.9072,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.42393559202361986,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.9346,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.47780381004786093,
"learning_rate": 2.070678184630458e-05,
"loss": 0.948,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.363683275381115,
"learning_rate": 2.06502600977305e-05,
"loss": 0.8849,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.7265745280911478,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.95,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.45658950851070673,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.9095,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.3697889305714029,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.8933,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.3733304092948457,
"learning_rate": 2.042412562887594e-05,
"loss": 0.9188,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.43066010684709727,
"learning_rate": 2.0367582403231033e-05,
"loss": 0.9016,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.3773279990643832,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.9031,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.3961455154814621,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.9225,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.463211542124309,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.9214,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.3513792711932367,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.9,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.3740946202831882,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.933,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.4331281522690081,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.9475,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.5301601584482166,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.9041,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.38645944389408277,
"learning_rate": 1.99151687695709e-05,
"loss": 0.8935,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.38710700907692347,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.8843,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.36613758445557015,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.8842,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 2.4794308890361556,
"learning_rate": 1.974551241345447e-05,
"loss": 0.9525,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.5774754968762793,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.9349,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.5426009558937555,
"learning_rate": 1.963241759676897e-05,
"loss": 0.8824,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.4097396645796262,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.8856,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.5492251546563234,
"learning_rate": 1.951933453677638e-05,
"loss": 0.9063,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.5646718349624749,
"learning_rate": 1.9462798545816932e-05,
"loss": 0.9231,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.4381655282036284,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.885,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.4447020494191072,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.9545,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.7087366891854597,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.9386,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.3841987979664424,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.9277,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.4458828153406673,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.9318,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.4058574592543282,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.9171,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.3745566860242917,
"learning_rate": 1.906719219225165e-05,
"loss": 0.9097,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.4100297099314172,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.8994,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.3814876949216998,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.9225,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.45716979091495613,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.9367,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.5141635687464093,
"learning_rate": 1.884128556891707e-05,
"loss": 0.8967,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.3842885266982001,
"learning_rate": 1.878483094691119e-05,
"loss": 0.9303,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.3855888031790581,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.9259,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.3978247499373885,
"learning_rate": 1.867195130360947e-05,
"loss": 0.9016,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.35298351195260275,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.908,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.384911814587293,
"learning_rate": 1.8559114136390795e-05,
"loss": 0.9194,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.33377287741803685,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.9215,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.40093185428124933,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.9046,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.32147324354212825,
"learning_rate": 1.838994592242891e-05,
"loss": 0.8988,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.3957251771585145,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.899,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.38000989129478996,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.8974,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.6042177768631888,
"learning_rate": 1.82208935733949e-05,
"loss": 0.9478,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.4932512228659294,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.9239,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.343172265688248,
"learning_rate": 1.810826238483892e-05,
"loss": 0.9137,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.4280387711459951,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.9236,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.46099472595373026,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.9122,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.3786657018557926,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.9181,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.46192603368595864,
"learning_rate": 1.788318512320838e-05,
"loss": 0.9359,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.42021006699921964,
"learning_rate": 1.782695699829051e-05,
"loss": 0.977,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.3561112023913921,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.8998,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.44903903860982414,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.8861,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.3208396254672673,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.8804,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.39555014338183353,
"learning_rate": 1.76022227484387e-05,
"loss": 0.8863,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.372043546700203,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.9323,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.3299587519483892,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.9208,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.39474344017527,
"learning_rate": 1.743387180005809e-05,
"loss": 0.9019,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.35144030996154096,
"learning_rate": 1.737779525652899e-05,
"loss": 0.9183,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.35213286993828236,
"learning_rate": 1.732173968007914e-05,
"loss": 0.902,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.3670562417783027,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.891,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.3174998920724343,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.8996,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.3641208290601015,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.9268,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.3233396280289397,
"learning_rate": 1.709773600686015e-05,
"loss": 0.9294,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.35795594327159447,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.9173,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.3418697713676465,
"learning_rate": 1.69858716181154e-05,
"loss": 0.9253,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.3352669579450277,
"learning_rate": 1.692997535150948e-05,
"loss": 0.9343,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.3726885599672759,
"learning_rate": 1.687410363273884e-05,
"loss": 0.8986,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.3154847077538375,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.8852,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.3333530714820343,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.9143,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.29838995178762834,
"learning_rate": 1.67066402299235e-05,
"loss": 0.8663,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.3312277341923782,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.931,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.30073151614101673,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.9048,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.3295710185711007,
"learning_rate": 1.653941382839745e-05,
"loss": 0.8972,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.6297023693730435,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.84,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.747743093753634,
"learning_rate": 1.642806717162757e-05,
"loss": 0.8423,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.47673061568440156,
"learning_rate": 1.637243646233718e-05,
"loss": 0.8216,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.6935540047144473,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.8473,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.5576809564393509,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.8505,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.5037130058898189,
"learning_rate": 1.620572014799777e-05,
"loss": 0.855,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.4651161701556942,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.8419,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.4465283990264351,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.8324,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.41605564095837577,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.8182,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.45063595676876117,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.8588,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.45754287622066364,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.8142,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.40789778070336374,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.8411,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.42401709157294504,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.8062,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.392848998276422,
"learning_rate": 1.576251073999917e-05,
"loss": 0.8239,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.4522365928681546,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.8556,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.38861036584450304,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.8426,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.4519003641552602,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.8454,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.3840188072348076,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.8153,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.4282486328196753,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.8635,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.40108142397665053,
"learning_rate": 1.543151434154755e-05,
"loss": 0.8454,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.43678470585845275,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.831,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.36572812780773545,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.8149,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.3862341992496903,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.8377,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.342171016170547,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.8248,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.39094556613862846,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.8478,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.3756799924352633,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.8751,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.3537282339711265,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.8053,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.37058713677446803,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.8607,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.6618073206981072,
"learning_rate": 1.493751509814374e-05,
"loss": 0.8345,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.3882559815175935,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.8508,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.4702526187994361,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.8485,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.41157650228349746,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.8163,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.8969824909478641,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.8903,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.3983381147444987,
"learning_rate": 1.466446738480111e-05,
"loss": 0.8251,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.36536883150114685,
"learning_rate": 1.460998408550112e-05,
"loss": 0.8311,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.401081422220945,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.8713,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.3980544502021142,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.8721,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.3887693273749174,
"learning_rate": 1.444679451900437e-05,
"loss": 0.8558,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.37228580995231936,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.8385,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.38452469955083246,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.8419,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.3656141294064058,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.7756,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.3829533475271474,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.8342,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.3459406779089447,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.853,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.3785876523182128,
"learning_rate": 1.412162598317602e-05,
"loss": 0.8035,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.3447118335374823,
"learning_rate": 1.406759321421907e-05,
"loss": 0.8261,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.4057834510810352,
"learning_rate": 1.401360788062871e-05,
"loss": 0.823,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.3373170047837757,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.8548,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.37772431947195295,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.8178,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.3916489984796227,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.8431,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.36286261470699094,
"learning_rate": 1.379814952747693e-05,
"loss": 0.8636,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.3931852416455726,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.8394,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.3642103583610204,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.7999,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 1.108649100192136,
"learning_rate": 1.363707494698459e-05,
"loss": 0.845,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.34532738452831324,
"learning_rate": 1.358348460369265e-05,
"loss": 0.8268,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.4097116129174744,
"learning_rate": 1.352994556668684e-05,
"loss": 0.8356,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.3759015126678771,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.8411,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.3810773958810235,
"learning_rate": 1.342302312350678e-05,
"loss": 0.8071,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.4079365126873841,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.8365,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.3378288833933401,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.8164,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.37290275116639837,
"learning_rate": 1.326303494478451e-05,
"loss": 0.8628,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.33984092941373323,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.8565,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.36280903756953375,
"learning_rate": 1.315664479122898e-05,
"loss": 0.8782,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.35256305844590063,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.8678,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.36617183778991463,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.8478,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.3292861099797721,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.8344,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.3844037825864315,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.8827,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.3213088258688475,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.8412,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.34471968280588233,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.8421,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.3241581054418815,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.8307,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.3299590431900622,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.8111,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.3376591715501789,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.8336,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.3596931218065692,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.841,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.5118290020931913,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.8858,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.35293032508456024,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.8234,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.3476424312238666,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.827,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.3478986666991711,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.8422,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.3564589857535227,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.8553,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.33339589392403013,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.8243,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.34338144999311,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.8438,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.33140277668904067,
"learning_rate": 1.220938883250288e-05,
"loss": 0.8114,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.3374577492007716,
"learning_rate": 1.215733275116827e-05,
"loss": 0.8244,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.3433115169808026,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.8197,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.3563344810714113,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.8626,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.33115267747808325,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.8506,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.36498712969859776,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.8726,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.33509180899725827,
"learning_rate": 1.189800129803139e-05,
"loss": 0.8719,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.3808827337967661,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.8456,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.3308514086853438,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.8287,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.3629002669627735,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.8431,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.3478025058368733,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.8674,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.7812564719018479,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.8667,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.37778069399746517,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.8331,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.37558642465468095,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.8344,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.36402608276202847,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.852,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.3701248837861336,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.8411,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.3638261489473457,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.8161,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.32481849172776706,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.8077,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.3444791291211661,
"learning_rate": 1.128231163049258e-05,
"loss": 0.8626,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.3463455039948358,
"learning_rate": 1.123144751066259e-05,
"loss": 0.8502,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.35377158725026026,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.8276,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.3389524490077735,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.8129,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.3287604669208669,
"learning_rate": 1.107927745387114e-05,
"loss": 0.8368,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.357005297635601,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.8506,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.34720127710526233,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.8467,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.3357541951999004,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.8323,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.33089432200594054,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.8781,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.3515316223731354,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.8734,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.33864150540371196,
"learning_rate": 1.077687414070869e-05,
"loss": 0.8304,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.3536630256667887,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.8794,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.32761542232414115,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.8593,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.3562087219748188,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.8528,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.35016209026808887,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.8333,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.3327173769301368,
"learning_rate": 1.052689572753247e-05,
"loss": 0.8117,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.3577252543765181,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.8079,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.32966197253654445,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.8494,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.3542531297052944,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.8521,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.35794505474290084,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.8403,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.3602181628503177,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.8365,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.3282175600353841,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.7923,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.33673412292711746,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.8753,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.3421138015277738,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.8212,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.33258654946975874,
"learning_rate": 1.008173988718005e-05,
"loss": 0.8062,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.3317106044067729,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.8233,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.3391732748919098,
"learning_rate": 9.983678617766677e-06,
"loss": 0.8487,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.320672407271927,
"learning_rate": 9.934767922553788e-06,
"loss": 0.8147,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.3296205106347405,
"learning_rate": 9.885937708668371e-06,
"loss": 0.8528,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.31881775630016423,
"learning_rate": 9.837188366555524e-06,
"loss": 0.8271,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.3180477844789366,
"learning_rate": 9.788520286013691e-06,
"loss": 0.8318,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.3199057571004285,
"learning_rate": 9.739933856191537e-06,
"loss": 0.8099,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.32411339408465023,
"learning_rate": 9.691429465584866e-06,
"loss": 0.8591,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.3998249224364718,
"learning_rate": 9.643007502033492e-06,
"loss": 0.8199,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.332992551376687,
"learning_rate": 9.594668352718152e-06,
"loss": 0.8048,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.3633088930545061,
"learning_rate": 9.546412404157383e-06,
"loss": 0.8419,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.329647326708144,
"learning_rate": 9.498240042204471e-06,
"loss": 0.8455,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.33098474104496384,
"learning_rate": 9.450151652044339e-06,
"loss": 0.811,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.3085253677282711,
"learning_rate": 9.402147618190472e-06,
"loss": 0.8193,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.3556860631976794,
"learning_rate": 9.354228324481841e-06,
"loss": 0.8444,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.3259010631540704,
"learning_rate": 9.306394154079852e-06,
"loss": 0.8463,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.3449510287509729,
"learning_rate": 9.258645489465254e-06,
"loss": 0.8325,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.33428983563418524,
"learning_rate": 9.21098271243509e-06,
"loss": 0.8599,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.31916541321475916,
"learning_rate": 9.163406204099659e-06,
"loss": 0.861,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.3308188691238198,
"learning_rate": 9.115916344879452e-06,
"loss": 0.8279,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.3205259528322495,
"learning_rate": 9.068513514502126e-06,
"loss": 0.8314,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.310358593612008,
"learning_rate": 9.021198091999439e-06,
"loss": 0.851,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.3596875502576508,
"learning_rate": 8.973970455704251e-06,
"loss": 0.8574,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.319764545866084,
"learning_rate": 8.926830983247495e-06,
"loss": 0.8224,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.3367173077384067,
"learning_rate": 8.879780051555116e-06,
"loss": 0.8373,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.32703858177955736,
"learning_rate": 8.83281803684512e-06,
"loss": 0.8786,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.34301357704863855,
"learning_rate": 8.785945314624532e-06,
"loss": 0.8477,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.3401565055110522,
"learning_rate": 8.739162259686385e-06,
"loss": 0.8596,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.3046019495147914,
"learning_rate": 8.692469246106734e-06,
"loss": 0.8051,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.32166523944174125,
"learning_rate": 8.645866647241677e-06,
"loss": 0.8358,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.7351928664708427,
"learning_rate": 8.599354835724348e-06,
"loss": 0.9001,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.3343320515854786,
"learning_rate": 8.552934183461958e-06,
"loss": 0.8358,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.32703920180099594,
"learning_rate": 8.506605061632805e-06,
"loss": 0.8576,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.34198219724815776,
"learning_rate": 8.460367840683308e-06,
"loss": 0.8183,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.3373483123442002,
"learning_rate": 8.414222890325065e-06,
"loss": 0.8325,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.3336399297914752,
"learning_rate": 8.368170579531855e-06,
"loss": 0.8379,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.3413983071947952,
"learning_rate": 8.322211276536733e-06,
"loss": 0.8226,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.31394287787236397,
"learning_rate": 8.27634534882907e-06,
"loss": 0.813,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.34370060101771177,
"learning_rate": 8.230573163151585e-06,
"loss": 0.829,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.312770766720335,
"learning_rate": 8.184895085497466e-06,
"loss": 0.8026,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.34770891871279463,
"learning_rate": 8.139311481107401e-06,
"loss": 0.8483,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.32375187291459745,
"learning_rate": 8.093822714466686e-06,
"loss": 0.8454,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.3317015589728057,
"learning_rate": 8.048429149302271e-06,
"loss": 0.8405,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.3218910021164767,
"learning_rate": 8.003131148579902e-06,
"loss": 0.8278,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.3024880778769386,
"learning_rate": 7.957929074501182e-06,
"loss": 0.8468,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.3188459175411774,
"learning_rate": 7.912823288500695e-06,
"loss": 0.8549,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.31709319618993714,
"learning_rate": 7.867814151243102e-06,
"loss": 0.8161,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.3054536380601063,
"learning_rate": 7.822902022620267e-06,
"loss": 0.8295,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.3166001315553756,
"learning_rate": 7.778087261748377e-06,
"loss": 0.8606,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.31267216660763286,
"learning_rate": 7.733370226965056e-06,
"loss": 0.837,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.32312590746280956,
"learning_rate": 7.688751275826527e-06,
"loss": 0.8334,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.3029103307948962,
"learning_rate": 7.644230765104735e-06,
"loss": 0.8079,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.3097759490071008,
"learning_rate": 7.599809050784502e-06,
"loss": 0.8879,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.2918893257726039,
"learning_rate": 7.555486488060661e-06,
"loss": 0.8595,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.3185579020433405,
"learning_rate": 7.511263431335249e-06,
"loss": 0.8191,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.3008895467073797,
"learning_rate": 7.467140234214656e-06,
"loss": 0.8436,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.2831376419996222,
"learning_rate": 7.423117249506777e-06,
"loss": 0.8132,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.29459066850892207,
"learning_rate": 7.379194829218231e-06,
"loss": 0.8342,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.2984366543607027,
"learning_rate": 7.335373324551518e-06,
"loss": 0.8232,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.2993009401232747,
"learning_rate": 7.291653085902224e-06,
"loss": 0.8626,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.28636997392538494,
"learning_rate": 7.248034462856213e-06,
"loss": 0.8341,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.3006521071252211,
"learning_rate": 7.20451780418683e-06,
"loss": 0.8548,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.30849345215466917,
"learning_rate": 7.161103457852128e-06,
"loss": 0.8553,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.29941975232235396,
"learning_rate": 7.117791770992046e-06,
"loss": 0.8455,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.29727404360785553,
"learning_rate": 7.074583089925688e-06,
"loss": 0.8319,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.307280013609262,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.8065,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.31150003146458644,
"learning_rate": 6.988476126329604e-06,
"loss": 0.8304,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.30604945307033826,
"learning_rate": 6.945578532308857e-06,
"loss": 0.8594,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.3204030883002217,
"learning_rate": 6.902785321094301e-06,
"loss": 0.8694,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.3178007078155459,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.8388,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.2991356350939788,
"learning_rate": 6.817513414939892e-06,
"loss": 0.8862,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.30906765593687935,
"learning_rate": 6.775035401831924e-06,
"loss": 0.8258,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.3139040152253322,
"learning_rate": 6.732663135188473e-06,
"loss": 0.838,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.29359943237906816,
"learning_rate": 6.690396953817055e-06,
"loss": 0.8467,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.3016183289447883,
"learning_rate": 6.648237195676927e-06,
"loss": 0.8273,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.29412193063176945,
"learning_rate": 6.606184197876395e-06,
"loss": 0.8117,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.2976489243406478,
"learning_rate": 6.564238296670109e-06,
"loss": 0.8646,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.30250513721295125,
"learning_rate": 6.52239982745638e-06,
"loss": 0.86,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.3135466669553611,
"learning_rate": 6.480669124774484e-06,
"loss": 0.8327,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.2994863933423467,
"learning_rate": 6.439046522302013e-06,
"loss": 0.8534,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.31049850413349367,
"learning_rate": 6.397532352852191e-06,
"loss": 0.8248,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.30242752057531963,
"learning_rate": 6.356126948371202e-06,
"loss": 0.8194,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.2824853579398436,
"learning_rate": 6.314830639935559e-06,
"loss": 0.8178,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.3010893309941173,
"learning_rate": 6.273643757749448e-06,
"loss": 0.8236,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.30279800343692664,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.8189,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.3082644855909138,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.8149,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.30004667697302745,
"learning_rate": 6.150742957589784e-06,
"loss": 0.832,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.28768114847440607,
"learning_rate": 6.109997064904774e-06,
"loss": 0.8456,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.3040861391597255,
"learning_rate": 6.069362236313092e-06,
"loss": 0.8399,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.2926180365872843,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.7743,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.2858529095919846,
"learning_rate": 5.988427070179108e-06,
"loss": 0.8425,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.304856870209877,
"learning_rate": 5.948127379792234e-06,
"loss": 0.8389,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.29647021216039915,
"learning_rate": 5.90794004780439e-06,
"loss": 0.8533,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.2825699396374881,
"learning_rate": 5.867865395552405e-06,
"loss": 0.8154,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.2855535348190632,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.8636,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.2780965283120706,
"learning_rate": 5.788055411095892e-06,
"loss": 0.8152,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.2958353395073402,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.8279,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.2818642443188115,
"learning_rate": 5.708699979051613e-06,
"loss": 0.8395,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.30147533434668816,
"learning_rate": 5.66919351390751e-06,
"loss": 0.8381,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.28541193439957097,
"learning_rate": 5.629801637510188e-06,
"loss": 0.8158,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.278139427213066,
"learning_rate": 5.590524664836043e-06,
"loss": 0.8109,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.29665664737514263,
"learning_rate": 5.551362909942708e-06,
"loss": 0.8219,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.5592387822652736,
"learning_rate": 5.51231668596653e-06,
"loss": 0.8404,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.2952901983911588,
"learning_rate": 5.47338630512007e-06,
"loss": 0.8413,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.30260867420427046,
"learning_rate": 5.434572078689617e-06,
"loss": 0.8256,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.2950760441736342,
"learning_rate": 5.395874317032694e-06,
"loss": 0.8555,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.30512236837072587,
"learning_rate": 5.357293329575573e-06,
"loss": 0.7948,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.3073672840787437,
"learning_rate": 5.318829424810797e-06,
"loss": 0.8141,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.2896911660343535,
"learning_rate": 5.280482910294728e-06,
"loss": 0.85,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.28766258181692794,
"learning_rate": 5.242254092645078e-06,
"loss": 0.8456,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.30394517973857044,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.8202,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.293430332138288,
"learning_rate": 5.166150769707892e-06,
"loss": 0.8507,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.29142443647275273,
"learning_rate": 5.128276872940503e-06,
"loss": 0.824,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.31365247239545235,
"learning_rate": 5.090521890074953e-06,
"loss": 0.8625,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.28911866208108844,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.8142,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.28962603768354833,
"learning_rate": 5.015369872647442e-06,
"loss": 0.8374,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.2941875002572757,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.8259,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.2892345445443984,
"learning_rate": 4.940697121074675e-06,
"loss": 0.8352,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.2877492449550457,
"learning_rate": 4.903541216934864e-06,
"loss": 0.8303,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.32044604588248243,
"learning_rate": 4.8665060236772e-06,
"loss": 0.8569,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 1.1551302309302507,
"learning_rate": 4.829591837434102e-06,
"loss": 0.8426,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.28700832240816254,
"learning_rate": 4.792798953370424e-06,
"loss": 0.8719,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.301322279387943,
"learning_rate": 4.756127665681083e-06,
"loss": 0.813,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.30227745288944424,
"learning_rate": 4.719578267588725e-06,
"loss": 0.8288,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.30081949027731697,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.8095,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.7764683013877782,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.7847,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.5147670628800348,
"learning_rate": 4.610664328486485e-06,
"loss": 0.7741,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.585756271778584,
"learning_rate": 4.5746054014809e-06,
"loss": 0.7934,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.6888907628199232,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.7822,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.5234314059955437,
"learning_rate": 4.502857857942204e-06,
"loss": 0.7996,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.7054835102951591,
"learning_rate": 4.467169815100549e-06,
"loss": 0.8007,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.5592705720218287,
"learning_rate": 4.431605972355137e-06,
"loss": 0.7817,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.4328691328807524,
"learning_rate": 4.396166614073516e-06,
"loss": 0.7777,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.6312810321534622,
"learning_rate": 4.360852023627846e-06,
"loss": 0.7781,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.46576575871997133,
"learning_rate": 4.325662483392656e-06,
"loss": 0.7506,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.40797530989363534,
"learning_rate": 4.29059827474257e-06,
"loss": 0.7673,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.45526290260419944,
"learning_rate": 4.255659678050063e-06,
"loss": 0.7735,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.41212614650627577,
"learning_rate": 4.22084697268323e-06,
"loss": 0.7934,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.37778228213714726,
"learning_rate": 4.186160437003535e-06,
"loss": 0.7764,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.3861537143779034,
"learning_rate": 4.151600348363582e-06,
"loss": 0.7607,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.4447272558044648,
"learning_rate": 4.11716698310493e-06,
"loss": 0.78,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.35429637416083887,
"learning_rate": 4.082860616555848e-06,
"loss": 0.8222,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.3512268667048509,
"learning_rate": 4.048681523029125e-06,
"loss": 0.7675,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.37987388175740777,
"learning_rate": 4.014629975819882e-06,
"loss": 0.7337,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.35876990313490564,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.7538,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.3245671698364371,
"learning_rate": 3.94691060843287e-06,
"loss": 0.7795,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.3552762226233531,
"learning_rate": 3.913243329737353e-06,
"loss": 0.7347,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.36957300622334044,
"learning_rate": 3.879704680319503e-06,
"loss": 0.7445,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.33090408930353554,
"learning_rate": 3.846294928353471e-06,
"loss": 0.8162,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.3248724027077166,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.8007,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.3467472074769218,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.7463,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.34792776080452864,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.7698,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.33300705420523524,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.7794,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.34133100913407577,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.7455,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.3298813220015371,
"learning_rate": 3.648558150665886e-06,
"loss": 0.8085,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.32050006111121543,
"learning_rate": 3.616058102895075e-06,
"loss": 0.7856,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.32917287041912263,
"learning_rate": 3.583689060687079e-06,
"loss": 0.8098,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.32705771046999466,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.7883,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.32047746846354264,
"learning_rate": 3.51934502719796e-06,
"loss": 0.7759,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.325887521446907,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.7956,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.31994227005830406,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.7924,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.3242641774422554,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.8051,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.3112118681035547,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.8012,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.31967630710870604,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.7759,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.322540447119496,
"learning_rate": 3.32948376099375e-06,
"loss": 0.7865,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.3223588484287709,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.7977,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.30774203214722373,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.7605,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.32190604609345763,
"learning_rate": 3.236349233544438e-06,
"loss": 0.7671,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.30755667429810885,
"learning_rate": 3.205572144561604e-06,
"loss": 0.771,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.3114160840159848,
"learning_rate": 3.174929343376374e-06,
"loss": 0.7555,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.3183740834217468,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.7714,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.30195309748367855,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.7567,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.2990672277894764,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.7506,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.3191492521174232,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.7868,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.30622563934606545,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.8016,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.3120138069211633,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.7766,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.3222472211422519,
"learning_rate": 2.964210239224696e-06,
"loss": 0.7871,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.31222414944765053,
"learning_rate": 2.934650470499869e-06,
"loss": 0.8158,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.30420256192081724,
"learning_rate": 2.905227155855128e-06,
"loss": 0.7648,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.31255357722784927,
"learning_rate": 2.875940530558519e-06,
"loss": 0.7515,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.31340030881917696,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.7777,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.30312370052428905,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.7577,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.31154026451258343,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.7995,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.30657015456837117,
"learning_rate": 2.760165589921184e-06,
"loss": 0.7754,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.2996621291446761,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.7527,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.3127542366554723,
"learning_rate": 2.703104292151164e-06,
"loss": 0.8288,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.30303680472299477,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.7602,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.3097931422441811,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.7765,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.30054915364958856,
"learning_rate": 2.618550200289303e-06,
"loss": 0.8048,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.2950930014728783,
"learning_rate": 2.59064316684412e-06,
"loss": 0.7322,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.3031720469955428,
"learning_rate": 2.562875338152264e-06,
"loss": 0.7714,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.3224096411736726,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.7676,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.3079444815365176,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.7664,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.3125152648966555,
"learning_rate": 2.480409295328803e-06,
"loss": 0.7699,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.30352581275373947,
"learning_rate": 2.45320049480108e-06,
"loss": 0.8228,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.28864806959291056,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.7615,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.3140916473625961,
"learning_rate": 2.399204021408421e-06,
"loss": 0.7989,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.2937709585622314,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.7974,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.3037842952723279,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.7849,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.30373776040959644,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.7742,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.30208192803432393,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.78,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.2950435514306163,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.7922,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.3018487458152484,
"learning_rate": 2.24059906434998e-06,
"loss": 0.7634,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.31409646470492586,
"learning_rate": 2.214660694699091e-06,
"loss": 0.7736,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.2899324762832243,
"learning_rate": 2.188864536147355e-06,
"loss": 0.778,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.2968152187296897,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.7975,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.2928634600211876,
"learning_rate": 2.137699676264171e-06,
"loss": 0.7614,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.29237557373547773,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.764,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.28641168446718046,
"learning_rate": 2.08710612114857e-06,
"loss": 0.737,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.2901449890993085,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.7598,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.30618397712037587,
"learning_rate": 2.0370854889762e-06,
"loss": 0.7825,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.29651809623085096,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.7804,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.29985063388523264,
"learning_rate": 1.987639379598427e-06,
"loss": 0.7965,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.29972781604884846,
"learning_rate": 1.963132265887919e-06,
"loss": 0.8103,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.30456636287548683,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.7884,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.2963890795994842,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.7796,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.3036141887449457,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.8054,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.2917469997108405,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.8171,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.2980853145504803,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.784,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.288466249519229,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.7834,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.29935675545332247,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.7985,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.3001530759391249,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.79,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.2914860956833573,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.7851,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.2983285923112461,
"learning_rate": 1.726025117089718e-06,
"loss": 0.7616,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.30560826873024155,
"learning_rate": 1.70311497206598e-06,
"loss": 0.8045,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.293127584476821,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.7793,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.30173294354360625,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.7931,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.3007685904151683,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.7781,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.2988817298041022,
"learning_rate": 1.612939220355272e-06,
"loss": 0.7606,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.326982246247001,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.817,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.2928509155267131,
"learning_rate": 1.568732761074121e-06,
"loss": 0.7676,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.29203955544935617,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.7669,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.3231954950888484,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.7633,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.2904749818828555,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.7577,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.3024094687689471,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.7775,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.2957671996044935,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.7706,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.29830476363405894,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.799,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.292874190574043,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.7469,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.2998590914825815,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.7695,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.2995475761196305,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.7714,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.2980624050791187,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.8176,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.30409101965709706,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.7598,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.3063474512088334,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.7912,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.30552742005985817,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.7948,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.29390657972964146,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.7807,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.2925047665448151,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.7517,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.2954083889834795,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.783,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.2955765053661477,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.7717,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.2927937320067896,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.7724,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.3063158747813953,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.8055,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.2875332740942384,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.7681,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.29967403063392506,
"learning_rate": 1.140415161609505e-06,
"loss": 0.7679,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.29849697216295473,
"learning_rate": 1.121666393356604e-06,
"loss": 0.775,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.3002383756207161,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.7887,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.2951054833787778,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.7742,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.6562472795146026,
"learning_rate": 1.066326386135097e-06,
"loss": 0.7814,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.297967580136971,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.798,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.30234843302040476,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.8028,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.2897091831459045,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.7477,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.2880071458757294,
"learning_rate": 9.9465987684501e-07,
"loss": 0.7615,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.2831708189823934,
"learning_rate": 9.771228098856534e-07,
"loss": 0.7853,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.2828009783003648,
"learning_rate": 9.597378493454013e-07,
"loss": 0.7644,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.2776783241627069,
"learning_rate": 9.425051342339353e-07,
"loss": 0.7585,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.29331887289941505,
"learning_rate": 9.254248023435864e-07,
"loss": 0.8095,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.3087350633234457,
"learning_rate": 9.084969902482288e-07,
"loss": 0.8004,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.2837121556718219,
"learning_rate": 8.917218333021993e-07,
"loss": 0.7946,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.2980920570642939,
"learning_rate": 8.750994656392109e-07,
"loss": 0.7861,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.2985986224726157,
"learning_rate": 8.58630020171265e-07,
"loss": 0.7693,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.2910546581625392,
"learning_rate": 8.423136285876099e-07,
"loss": 0.7638,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.29344912507893717,
"learning_rate": 8.26150421353682e-07,
"loss": 0.7627,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.28904737577976575,
"learning_rate": 8.101405277100549e-07,
"loss": 0.7746,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.28948027447931846,
"learning_rate": 7.942840756714077e-07,
"loss": 0.8344,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.2974470160696474,
"learning_rate": 7.785811920255093e-07,
"loss": 0.7707,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.29215868610406304,
"learning_rate": 7.630320023322024e-07,
"loss": 0.7835,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.28243202319648514,
"learning_rate": 7.476366309223903e-07,
"loss": 0.7994,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.282020133265319,
"learning_rate": 7.323952008970537e-07,
"loss": 0.765,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.29352149612724077,
"learning_rate": 7.173078341262618e-07,
"loss": 0.7795,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.28951311738290614,
"learning_rate": 7.023746512481989e-07,
"loss": 0.8213,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.286719290008893,
"learning_rate": 6.875957716681902e-07,
"loss": 0.7752,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.28962063485793704,
"learning_rate": 6.729713135577687e-07,
"loss": 0.7746,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.2932850478612325,
"learning_rate": 6.585013938537143e-07,
"loss": 0.8158,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.2899045935513774,
"learning_rate": 6.441861282571115e-07,
"loss": 0.7785,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.29812710895368366,
"learning_rate": 6.300256312324471e-07,
"loss": 0.793,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.2917242274278519,
"learning_rate": 6.160200160066776e-07,
"loss": 0.7844,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.2902817224663875,
"learning_rate": 6.021693945683327e-07,
"loss": 0.7801,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.2838612920443065,
"learning_rate": 5.884738776666044e-07,
"loss": 0.7789,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.3186636307671861,
"learning_rate": 5.749335748104878e-07,
"loss": 0.7753,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.28254543425626455,
"learning_rate": 5.615485942678778e-07,
"loss": 0.7733,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.2760234181423924,
"learning_rate": 5.483190430647289e-07,
"loss": 0.7423,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.28272419197245513,
"learning_rate": 5.352450269841747e-07,
"loss": 0.7699,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.2934640746559737,
"learning_rate": 5.22326650565701e-07,
"loss": 0.7543,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.2877123688391272,
"learning_rate": 5.095640171043048e-07,
"loss": 0.7938,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.27979015182052785,
"learning_rate": 4.969572286496594e-07,
"loss": 0.7596,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.2881636687635836,
"learning_rate": 4.845063860053079e-07,
"loss": 0.7982,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.2918267531716322,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.7644,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.2863401502180594,
"learning_rate": 4.600729351261857e-07,
"loss": 0.7691,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.28763131822976457,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.792,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.28181677215660655,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.7729,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.2838903341902389,
"learning_rate": 4.245948007320588e-07,
"loss": 0.778,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.30143065667045665,
"learning_rate": 4.130816799401971e-07,
"loss": 0.777,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.28389079545851514,
"learning_rate": 4.017251756252827e-07,
"loss": 0.7968,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.2758296262351073,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.7596,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.28607245230998407,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.7957,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.29010901176390225,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.7929,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.29562757594362327,
"learning_rate": 3.578671204615347e-07,
"loss": 0.7854,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.28937786717166075,
"learning_rate": 3.472950355547267e-07,
"loss": 0.7862,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.2833789000101269,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.8042,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.27957492716390797,
"learning_rate": 3.266223765343224e-07,
"loss": 0.7994,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.2896596821942234,
"learning_rate": 3.16521967718757e-07,
"loss": 0.7945,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.29042515988120093,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.7713,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.2894606698303935,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.7707,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.6437800118269968,
"learning_rate": 2.871653894244597e-07,
"loss": 0.7803,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.30118816941597504,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.788,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.28334146778825814,
"learning_rate": 2.683823233205485e-07,
"loss": 0.7754,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.2971496616147492,
"learning_rate": 2.592274132703287e-07,
"loss": 0.7739,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.2911350125802112,
"learning_rate": 2.502303499116754e-07,
"loss": 0.8063,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.2827362065632661,
"learning_rate": 2.413912051848666e-07,
"loss": 0.783,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.5914984093390157,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.7751,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.2872307213340865,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.7639,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.2877280265798513,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.7531,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.2919493162910426,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.7552,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.2901415178797672,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.7879,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.2826963948132567,
"learning_rate": 1.916764975761254e-07,
"loss": 0.7567,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.29354203535888246,
"learning_rate": 1.839446914621923e-07,
"loss": 0.7801,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.2865961172164031,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.7563,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.29122414700001903,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.798,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.2875025769223873,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.7874,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.2897380095277583,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.7854,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.3018868836739345,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.7927,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.2834019583040241,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.7876,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.2835329322793742,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.7861,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.30875206046408193,
"learning_rate": 1.277993209368744e-07,
"loss": 0.7506,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.2740059108919849,
"learning_rate": 1.214956261808542e-07,
"loss": 0.7888,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.28924809205744567,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.7729,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.29160143998823623,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.781,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.2867878954861124,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.7682,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.2948562858652404,
"learning_rate": 9.787081204466253e-08,
"loss": 0.7963,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.2800587805098633,
"learning_rate": 9.236233586200627e-08,
"loss": 0.7755,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.2782359622180485,
"learning_rate": 8.70130406196279e-08,
"loss": 0.7604,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.2798005409545193,
"learning_rate": 8.182296909035047e-08,
"loss": 0.7754,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.2848157440727662,
"learning_rate": 7.679216277384439e-08,
"loss": 0.7967,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.3348364952503759,
"learning_rate": 7.19206618963053e-08,
"loss": 0.7515,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.39772707794460666,
"learning_rate": 6.720850541012347e-08,
"loss": 0.744,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.28360039220791705,
"learning_rate": 6.265573099357714e-08,
"loss": 0.7653,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.26967263768343686,
"learning_rate": 5.826237505053067e-08,
"loss": 0.7333,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.2867374717260544,
"learning_rate": 5.4028472710137e-08,
"loss": 0.7717,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.29258148840058457,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.8088,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.2892006130318691,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.7928,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.3090592278880841,
"learning_rate": 4.228381947004812e-08,
"loss": 0.7867,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.28259973678666667,
"learning_rate": 3.868805732811343e-08,
"loss": 0.7443,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.2804103730664775,
"learning_rate": 3.525190530455636e-08,
"loss": 0.8161,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.27728951291467424,
"learning_rate": 3.197539087475887e-08,
"loss": 0.7493,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.2914890076266428,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.7934,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.28062857558496107,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.7996,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.28785503976121163,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.798,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.2987889984140742,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.7569,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.28129881383616245,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.7257,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.2849459501978042,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.774,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.2820399692960616,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.7655,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.27850872810520694,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.7562,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.2772127510945601,
"learning_rate": 9.67435416023843e-09,
"loss": 0.7794,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.2788486978357787,
"learning_rate": 7.995445897566178e-09,
"loss": 0.7854,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.2750123060876264,
"learning_rate": 6.476393170578021e-09,
"loss": 0.779,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.2839729605770126,
"learning_rate": 5.11720812557348e-09,
"loss": 0.7839,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.28083224354809866,
"learning_rate": 3.917901630563492e-09,
"loss": 0.7866,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.2940454735730501,
"learning_rate": 2.878483275170485e-09,
"loss": 0.7877,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.27852160272967463,
"learning_rate": 1.998961370555108e-09,
"loss": 0.7396,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.2807950691795994,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.7673,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.2806296930417516,
"learning_rate": 7.196337656112384e-10,
"loss": 0.7976,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.27629947181036596,
"learning_rate": 3.198382947511292e-10,
"loss": 0.7787,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.28797953737616727,
"learning_rate": 7.995973352770137e-11,
"loss": 0.8238,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.2841731575479532,
"learning_rate": 0.0,
"loss": 0.7688,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 5.402617348857266e+18,
"train_loss": 0.9231608434727318,
"train_runtime": 74558.7534,
"train_samples_per_second": 2.119,
"train_steps_per_second": 0.017
}
],
"logging_steps": 1.0,
"max_steps": 1235,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 5.402617348857266e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}