e1_science_longest_qwq_together / trainer_state.json
sedrickkeh's picture
End of training
ee4f79c 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": 7.891869926100749,
"learning_rate": 3.2258064516129035e-07,
"loss": 1.5815,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 7.911094915424423,
"learning_rate": 6.451612903225807e-07,
"loss": 1.5739,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 7.883732918235941,
"learning_rate": 9.67741935483871e-07,
"loss": 1.585,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 8.01791615070707,
"learning_rate": 1.2903225806451614e-06,
"loss": 1.6184,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 7.3751345628305724,
"learning_rate": 1.6129032258064516e-06,
"loss": 1.572,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 7.323470663652246,
"learning_rate": 1.935483870967742e-06,
"loss": 1.5916,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 5.697066019921216,
"learning_rate": 2.2580645161290324e-06,
"loss": 1.5039,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 5.29773566382523,
"learning_rate": 2.580645161290323e-06,
"loss": 1.4628,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 2.8711821707252736,
"learning_rate": 2.903225806451613e-06,
"loss": 1.4825,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 2.7380120279280047,
"learning_rate": 3.225806451612903e-06,
"loss": 1.5016,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 2.4133592454018364,
"learning_rate": 3.548387096774194e-06,
"loss": 1.4416,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 5.227587106053472,
"learning_rate": 3.870967741935484e-06,
"loss": 1.4672,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 5.478527617647755,
"learning_rate": 4.193548387096774e-06,
"loss": 1.4785,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 5.397902507190127,
"learning_rate": 4.516129032258065e-06,
"loss": 1.4258,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 5.114035644561169,
"learning_rate": 4.838709677419355e-06,
"loss": 1.4694,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 3.887769194878138,
"learning_rate": 5.161290322580646e-06,
"loss": 1.3848,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 3.5845230298633575,
"learning_rate": 5.483870967741935e-06,
"loss": 1.3774,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 3.117309456536742,
"learning_rate": 5.806451612903226e-06,
"loss": 1.3695,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 2.4818032362763174,
"learning_rate": 6.129032258064517e-06,
"loss": 1.3908,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.8437322838705492,
"learning_rate": 6.451612903225806e-06,
"loss": 1.3683,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 1.4373613850013474,
"learning_rate": 6.774193548387097e-06,
"loss": 1.3439,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.4614571603376565,
"learning_rate": 7.096774193548388e-06,
"loss": 1.3114,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.6595357500364922,
"learning_rate": 7.4193548387096784e-06,
"loss": 1.3126,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 1.531486283752174,
"learning_rate": 7.741935483870968e-06,
"loss": 1.3127,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 1.188041862703481,
"learning_rate": 8.064516129032258e-06,
"loss": 1.3011,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 0.9622960804099122,
"learning_rate": 8.387096774193549e-06,
"loss": 1.3305,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 1.0552822226166378,
"learning_rate": 8.70967741935484e-06,
"loss": 1.2635,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 1.1313015687189192,
"learning_rate": 9.03225806451613e-06,
"loss": 1.2504,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 0.8869217450078676,
"learning_rate": 9.35483870967742e-06,
"loss": 1.278,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 0.8947225112815047,
"learning_rate": 9.67741935483871e-06,
"loss": 1.2495,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 0.9214597763706822,
"learning_rate": 1e-05,
"loss": 1.2229,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 0.9157165155101619,
"learning_rate": 1.0322580645161291e-05,
"loss": 1.2918,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 0.7683963391793106,
"learning_rate": 1.0645161290322582e-05,
"loss": 1.249,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 0.7465782797523662,
"learning_rate": 1.096774193548387e-05,
"loss": 1.2124,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 0.8773287884128909,
"learning_rate": 1.1290322580645164e-05,
"loss": 1.2869,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.6796635158668209,
"learning_rate": 1.1612903225806453e-05,
"loss": 1.2588,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.678759565594785,
"learning_rate": 1.1935483870967743e-05,
"loss": 1.196,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.6650153002155527,
"learning_rate": 1.2258064516129034e-05,
"loss": 1.2325,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.6698334628739859,
"learning_rate": 1.2580645161290324e-05,
"loss": 1.187,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 0.7237057233415337,
"learning_rate": 1.2903225806451613e-05,
"loss": 1.1808,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.6323308000038605,
"learning_rate": 1.3225806451612903e-05,
"loss": 1.2008,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.7334429422685188,
"learning_rate": 1.3548387096774194e-05,
"loss": 1.2155,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.7229361994533432,
"learning_rate": 1.3870967741935486e-05,
"loss": 1.1951,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.6858782175612425,
"learning_rate": 1.4193548387096776e-05,
"loss": 1.2129,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.7715919092469566,
"learning_rate": 1.4516129032258066e-05,
"loss": 1.1779,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 0.6385976315584758,
"learning_rate": 1.4838709677419357e-05,
"loss": 1.1845,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.8149736197637352,
"learning_rate": 1.5161290322580646e-05,
"loss": 1.1828,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.6464395506269729,
"learning_rate": 1.5483870967741936e-05,
"loss": 1.2207,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 0.7798720882168116,
"learning_rate": 1.5806451612903226e-05,
"loss": 1.2224,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 0.6821315039932511,
"learning_rate": 1.6129032258064517e-05,
"loss": 1.1751,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 0.8543888341494051,
"learning_rate": 1.6451612903225807e-05,
"loss": 1.2173,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.6747578720058202,
"learning_rate": 1.6774193548387098e-05,
"loss": 1.2175,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 0.8613172214063869,
"learning_rate": 1.7096774193548388e-05,
"loss": 1.1921,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.6798849897692147,
"learning_rate": 1.741935483870968e-05,
"loss": 1.1523,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.6912362843756867,
"learning_rate": 1.774193548387097e-05,
"loss": 1.1914,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 0.5906012413125366,
"learning_rate": 1.806451612903226e-05,
"loss": 1.1614,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.8278604429079938,
"learning_rate": 1.838709677419355e-05,
"loss": 1.1793,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 0.7042391815782569,
"learning_rate": 1.870967741935484e-05,
"loss": 1.1798,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 0.6431366267618435,
"learning_rate": 1.903225806451613e-05,
"loss": 1.1732,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 0.7651797336186761,
"learning_rate": 1.935483870967742e-05,
"loss": 1.1834,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 0.7758899755133839,
"learning_rate": 1.967741935483871e-05,
"loss": 1.221,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 0.6749278308008637,
"learning_rate": 2e-05,
"loss": 1.166,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 0.6898102632922138,
"learning_rate": 2.0322580645161292e-05,
"loss": 1.1658,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 0.8301911958908781,
"learning_rate": 2.0645161290322582e-05,
"loss": 1.1478,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 1.0635576448677542,
"learning_rate": 2.0967741935483873e-05,
"loss": 1.1943,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 0.9548316044674672,
"learning_rate": 2.1290322580645163e-05,
"loss": 1.1477,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 0.8913364823094778,
"learning_rate": 2.161290322580645e-05,
"loss": 1.1582,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 0.5943590120095265,
"learning_rate": 2.193548387096774e-05,
"loss": 1.1893,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 0.6854556203930099,
"learning_rate": 2.225806451612903e-05,
"loss": 1.1615,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 1.0617260089898972,
"learning_rate": 2.2580645161290328e-05,
"loss": 1.1825,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 1.180111808264413,
"learning_rate": 2.2903225806451615e-05,
"loss": 1.1202,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 0.8293465724215239,
"learning_rate": 2.3225806451612906e-05,
"loss": 1.2038,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 0.902885391849973,
"learning_rate": 2.3548387096774196e-05,
"loss": 1.1487,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 0.9726370883033926,
"learning_rate": 2.3870967741935486e-05,
"loss": 1.1504,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 1.2895328352688962,
"learning_rate": 2.4193548387096777e-05,
"loss": 1.1885,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 0.7662850450618113,
"learning_rate": 2.4516129032258067e-05,
"loss": 1.1534,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 0.9094750486345357,
"learning_rate": 2.4838709677419358e-05,
"loss": 1.2353,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 1.1451829088465062,
"learning_rate": 2.5161290322580648e-05,
"loss": 1.1616,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 1.1764111414760055,
"learning_rate": 2.548387096774194e-05,
"loss": 1.1605,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 1.0148708239554092,
"learning_rate": 2.5806451612903226e-05,
"loss": 1.1742,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 1.7893183898664278,
"learning_rate": 2.6129032258064516e-05,
"loss": 1.1795,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 0.8331771309200284,
"learning_rate": 2.6451612903225806e-05,
"loss": 1.1656,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 1.8903159387599977,
"learning_rate": 2.6774193548387097e-05,
"loss": 1.1848,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 1.0687635022363862,
"learning_rate": 2.7096774193548387e-05,
"loss": 1.1391,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 1.7158595664817728,
"learning_rate": 2.7419354838709678e-05,
"loss": 1.1024,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 1.5555651079871446,
"learning_rate": 2.774193548387097e-05,
"loss": 1.1927,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 1.1936634296787805,
"learning_rate": 2.8064516129032262e-05,
"loss": 1.1811,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 1.3981706245736538,
"learning_rate": 2.8387096774193552e-05,
"loss": 1.1623,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 1.4299016950535965,
"learning_rate": 2.8709677419354843e-05,
"loss": 1.1893,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 1.1932689384139745,
"learning_rate": 2.9032258064516133e-05,
"loss": 1.1245,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 1.274563957055977,
"learning_rate": 2.9354838709677423e-05,
"loss": 1.1625,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 1.2657392835368566,
"learning_rate": 2.9677419354838714e-05,
"loss": 1.1275,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 1.2434457983852805,
"learning_rate": 3.0000000000000004e-05,
"loss": 1.2122,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 1.0531166528946576,
"learning_rate": 3.032258064516129e-05,
"loss": 1.1708,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 1.4034019303162804,
"learning_rate": 3.0645161290322585e-05,
"loss": 1.1725,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 1.1212930593883823,
"learning_rate": 3.096774193548387e-05,
"loss": 1.1406,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 1.1238803909849895,
"learning_rate": 3.1290322580645166e-05,
"loss": 1.1578,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 1.566780331232182,
"learning_rate": 3.161290322580645e-05,
"loss": 1.1593,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 1.0406979738872832,
"learning_rate": 3.193548387096774e-05,
"loss": 1.1495,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 1.057005404358704,
"learning_rate": 3.2258064516129034e-05,
"loss": 1.1869,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 1.1510818786603494,
"learning_rate": 3.258064516129033e-05,
"loss": 1.1373,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 1.2838178643216045,
"learning_rate": 3.2903225806451614e-05,
"loss": 1.1483,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 1.4265071791176078,
"learning_rate": 3.322580645161291e-05,
"loss": 1.134,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 1.0696829596440223,
"learning_rate": 3.3548387096774195e-05,
"loss": 1.1472,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 1.0038366696615717,
"learning_rate": 3.387096774193549e-05,
"loss": 1.1265,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 0.812545683417207,
"learning_rate": 3.4193548387096776e-05,
"loss": 1.1966,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 1.0076948462959645,
"learning_rate": 3.451612903225807e-05,
"loss": 1.1602,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 2.0076372388890604,
"learning_rate": 3.483870967741936e-05,
"loss": 1.1653,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 1.2460655408119699,
"learning_rate": 3.516129032258065e-05,
"loss": 1.1633,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 1.6069846632365574,
"learning_rate": 3.548387096774194e-05,
"loss": 1.1825,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 1.469324764828521,
"learning_rate": 3.5806451612903225e-05,
"loss": 1.1473,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 1.0386201612574895,
"learning_rate": 3.612903225806452e-05,
"loss": 1.0853,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 1.89744445988627,
"learning_rate": 3.6451612903225805e-05,
"loss": 1.1558,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 1.186522703732291,
"learning_rate": 3.67741935483871e-05,
"loss": 1.1386,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 1.6577194029312314,
"learning_rate": 3.7096774193548386e-05,
"loss": 1.1895,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 1.1997292520798264,
"learning_rate": 3.741935483870968e-05,
"loss": 1.1804,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 1.577489433136177,
"learning_rate": 3.7741935483870974e-05,
"loss": 1.1301,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 1.3378965176861928,
"learning_rate": 3.806451612903226e-05,
"loss": 1.1559,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 1.3635872654196801,
"learning_rate": 3.8387096774193555e-05,
"loss": 1.1681,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 1.3947387615607538,
"learning_rate": 3.870967741935484e-05,
"loss": 1.1006,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 1.2507260953458528,
"learning_rate": 3.9032258064516135e-05,
"loss": 1.1639,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 1.3266573647609698,
"learning_rate": 3.935483870967742e-05,
"loss": 1.1873,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 1.1938656473271558,
"learning_rate": 3.9677419354838716e-05,
"loss": 1.1397,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 0.9708792126917662,
"learning_rate": 4e-05,
"loss": 1.1676,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 1.3252600614731247,
"learning_rate": 3.9999920040266474e-05,
"loss": 1.1449,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 0.8208428173556761,
"learning_rate": 3.999968016170525e-05,
"loss": 1.1305,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 1.5887420857029737,
"learning_rate": 3.9999280366234394e-05,
"loss": 1.1444,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 0.9633492095707706,
"learning_rate": 3.999872065705065e-05,
"loss": 1.1775,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 1.5312315161981043,
"learning_rate": 3.999800103862945e-05,
"loss": 1.1327,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 1.2402005591467629,
"learning_rate": 3.9997121516724834e-05,
"loss": 1.1702,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 1.3104916814834537,
"learning_rate": 3.999608209836944e-05,
"loss": 1.1386,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 1.2652417648048901,
"learning_rate": 3.999488279187443e-05,
"loss": 1.1586,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 1.6461084801773873,
"learning_rate": 3.9993523606829425e-05,
"loss": 1.1358,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 1.0999572574107668,
"learning_rate": 3.9992004554102435e-05,
"loss": 1.1343,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 1.3901809522627189,
"learning_rate": 3.999032564583976e-05,
"loss": 1.1324,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 1.154794483020958,
"learning_rate": 3.998848689546591e-05,
"loss": 1.1822,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 1.1850035141823765,
"learning_rate": 3.9986488317683496e-05,
"loss": 1.1657,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 1.4404400282873644,
"learning_rate": 3.9984329928473067e-05,
"loss": 1.1521,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 0.8891024490938398,
"learning_rate": 3.998201174509306e-05,
"loss": 1.1547,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 0.9034345318909436,
"learning_rate": 3.997953378607961e-05,
"loss": 1.1605,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 0.8610017058881397,
"learning_rate": 3.9976896071246403e-05,
"loss": 1.1478,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 1.1098443198793724,
"learning_rate": 3.997409862168454e-05,
"loss": 1.1266,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 1.550775643532458,
"learning_rate": 3.997114145976237e-05,
"loss": 1.15,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 1.4680501095279064,
"learning_rate": 3.9968024609125245e-05,
"loss": 1.1421,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 0.9699229899999807,
"learning_rate": 3.996474809469545e-05,
"loss": 1.172,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 1.9105900487430258,
"learning_rate": 3.996131194267189e-05,
"loss": 1.1609,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 1.1606959467873936,
"learning_rate": 3.995771618052995e-05,
"loss": 1.1489,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 1.4785190114275863,
"learning_rate": 3.9953960837021256e-05,
"loss": 1.1791,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 1.61985427916554,
"learning_rate": 3.995004594217343e-05,
"loss": 1.1422,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 0.9389223509859219,
"learning_rate": 3.994597152728987e-05,
"loss": 1.1789,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 1.320460873412239,
"learning_rate": 3.994173762494947e-05,
"loss": 1.1464,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 1.2613539136449081,
"learning_rate": 3.9937344269006424e-05,
"loss": 1.1471,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 0.9790004415089727,
"learning_rate": 3.993279149458988e-05,
"loss": 1.1447,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 0.9914685872454635,
"learning_rate": 3.9928079338103695e-05,
"loss": 1.1183,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 1.2556485312548251,
"learning_rate": 3.992320783722616e-05,
"loss": 1.1839,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 1.1516600460417548,
"learning_rate": 3.991817703090965e-05,
"loss": 1.105,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 1.3582918909449855,
"learning_rate": 3.991298695938038e-05,
"loss": 1.1481,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 1.109431020891737,
"learning_rate": 3.9907637664138e-05,
"loss": 1.1377,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 0.9193329183985031,
"learning_rate": 3.990212918795534e-05,
"loss": 1.1408,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 1.1633576960182752,
"learning_rate": 3.9896461574878036e-05,
"loss": 1.1407,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 0.9792601741469776,
"learning_rate": 3.989063487022416e-05,
"loss": 1.1578,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 0.9102740222765235,
"learning_rate": 3.9884649120583903e-05,
"loss": 1.1158,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 0.9113102233953218,
"learning_rate": 3.987850437381915e-05,
"loss": 1.159,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 0.7644008858213365,
"learning_rate": 3.987220067906313e-05,
"loss": 1.1408,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 1.0772141209110675,
"learning_rate": 3.986573808672002e-05,
"loss": 1.1252,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 0.9863302206248837,
"learning_rate": 3.9859116648464534e-05,
"loss": 1.1613,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 0.8914690256453375,
"learning_rate": 3.985233641724152e-05,
"loss": 1.1274,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 0.8773439906465553,
"learning_rate": 3.984539744726554e-05,
"loss": 1.1599,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 0.80009725486135,
"learning_rate": 3.983829979402038e-05,
"loss": 1.1354,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 0.8837974333070235,
"learning_rate": 3.983104351425871e-05,
"loss": 1.1344,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 0.8505877708291516,
"learning_rate": 3.982362866600154e-05,
"loss": 1.1657,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 0.9699129661345991,
"learning_rate": 3.9816055308537806e-05,
"loss": 1.1678,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 1.0529815616985327,
"learning_rate": 3.980832350242388e-05,
"loss": 1.0934,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 0.9215835251815282,
"learning_rate": 3.980043330948306e-05,
"loss": 1.135,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 1.1708926996244031,
"learning_rate": 3.9792384792805124e-05,
"loss": 1.1239,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 0.7731533141555509,
"learning_rate": 3.978417801674579e-05,
"loss": 1.1203,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 1.0723196797155925,
"learning_rate": 3.9775813046926244e-05,
"loss": 1.1409,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 0.8478577489514697,
"learning_rate": 3.976728995023254e-05,
"loss": 1.099,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 0.9458781772237718,
"learning_rate": 3.975860879481514e-05,
"loss": 1.1946,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 0.9189756509143774,
"learning_rate": 3.974976965008833e-05,
"loss": 1.1491,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 0.9932412329383904,
"learning_rate": 3.9740772586729674e-05,
"loss": 1.1067,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 1.4483613647330318,
"learning_rate": 3.973161767667946e-05,
"loss": 1.0867,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 0.6879745391392531,
"learning_rate": 3.972230499314009e-05,
"loss": 1.1443,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 1.4891110353152914,
"learning_rate": 3.9712834610575545e-05,
"loss": 1.129,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 1.1002006055464477,
"learning_rate": 3.9703206604710746e-05,
"loss": 1.1191,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 0.6315816412153679,
"learning_rate": 3.969342105253098e-05,
"loss": 1.1766,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 0.8201167367858538,
"learning_rate": 3.9683478032281246e-05,
"loss": 1.1379,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 1.2794620297198607,
"learning_rate": 3.967337762346568e-05,
"loss": 1.1534,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 1.002781794032513,
"learning_rate": 3.9663119906846885e-05,
"loss": 1.1556,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.7052412310215408,
"learning_rate": 3.965270496444528e-05,
"loss": 1.1859,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 0.8228654494589739,
"learning_rate": 3.9642132879538465e-05,
"loss": 1.1431,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 1.0071354762405516,
"learning_rate": 3.9631403736660566e-05,
"loss": 1.162,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 0.9852445402856153,
"learning_rate": 3.962051762160151e-05,
"loss": 1.085,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 0.8648873337313019,
"learning_rate": 3.960947462140638e-05,
"loss": 1.1247,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 1.3226768447809554,
"learning_rate": 3.959827482437472e-05,
"loss": 1.1654,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 0.686559413179171,
"learning_rate": 3.958691832005981e-05,
"loss": 1.1477,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 1.3709335423435067,
"learning_rate": 3.9575405199267944e-05,
"loss": 1.1132,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 0.9938516283418347,
"learning_rate": 3.956373555405774e-05,
"loss": 1.1218,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 0.8719104243062763,
"learning_rate": 3.955190947773937e-05,
"loss": 1.1444,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 0.7694437461118343,
"learning_rate": 3.9539927064873824e-05,
"loss": 1.1473,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 1.0246339683686938,
"learning_rate": 3.952778841127214e-05,
"loss": 1.1309,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 1.0092004212744536,
"learning_rate": 3.95154936139947e-05,
"loss": 1.1465,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 0.6915468418156808,
"learning_rate": 3.950304277135034e-05,
"loss": 1.1212,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 0.8271581117616249,
"learning_rate": 3.94904359828957e-05,
"loss": 1.1173,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 0.8597748416558273,
"learning_rate": 3.94776733494343e-05,
"loss": 1.183,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 0.6261367813463263,
"learning_rate": 3.9464754973015826e-05,
"loss": 1.1283,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 0.7286050306453891,
"learning_rate": 3.9451680956935275e-05,
"loss": 1.1249,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.9106927921980229,
"learning_rate": 3.943845140573213e-05,
"loss": 1.112,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.5611703289538206,
"learning_rate": 3.942506642518952e-05,
"loss": 1.1076,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 0.9647943764664386,
"learning_rate": 3.9411526122333396e-05,
"loss": 1.1284,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 0.8930819499599111,
"learning_rate": 3.939783060543167e-05,
"loss": 1.1472,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 0.7936634987102481,
"learning_rate": 3.938397998399333e-05,
"loss": 1.1264,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.8637701407428976,
"learning_rate": 3.936997436876756e-05,
"loss": 1.1228,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.6629523422422716,
"learning_rate": 3.935581387174289e-05,
"loss": 1.1631,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 0.6543969262881622,
"learning_rate": 3.9341498606146295e-05,
"loss": 1.1561,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.7421398057930748,
"learning_rate": 3.932702868644224e-05,
"loss": 1.1037,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.8361892586752294,
"learning_rate": 3.931240422833181e-05,
"loss": 1.1506,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.5767040860155221,
"learning_rate": 3.9297625348751806e-05,
"loss": 1.1162,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 0.6726899738802726,
"learning_rate": 3.9282692165873744e-05,
"loss": 1.1146,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 0.6112638551525303,
"learning_rate": 3.9267604799102945e-05,
"loss": 1.1575,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.9250910047274333,
"learning_rate": 3.9252363369077615e-05,
"loss": 1.1322,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 0.779871279079651,
"learning_rate": 3.9236967997667796e-05,
"loss": 1.1183,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 1.001007509315641,
"learning_rate": 3.9221418807974496e-05,
"loss": 1.1456,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 0.8469546793796525,
"learning_rate": 3.92057159243286e-05,
"loss": 1.1749,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 1.0091500345513935,
"learning_rate": 3.9189859472289956e-05,
"loss": 1.1334,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 1.0369533531075905,
"learning_rate": 3.9173849578646324e-05,
"loss": 1.1181,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 1.1456226140096697,
"learning_rate": 3.9157686371412395e-05,
"loss": 1.0735,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.6237881895822553,
"learning_rate": 3.914136997982874e-05,
"loss": 1.1208,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 1.0077091505081075,
"learning_rate": 3.912490053436079e-05,
"loss": 1.1749,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 1.0683988924443741,
"learning_rate": 3.9108278166697805e-05,
"loss": 1.1315,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.8408995258147826,
"learning_rate": 3.909150300975178e-05,
"loss": 1.1422,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 1.0477820580218669,
"learning_rate": 3.907457519765642e-05,
"loss": 1.1449,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.5319717870504315,
"learning_rate": 3.905749486576607e-05,
"loss": 1.1293,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.776321911350439,
"learning_rate": 3.90402621506546e-05,
"loss": 1.1224,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 0.7912847336320562,
"learning_rate": 3.902287719011435e-05,
"loss": 1.1455,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 1.156807304629412,
"learning_rate": 3.900534012315499e-05,
"loss": 1.163,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 0.7190814412676784,
"learning_rate": 3.8987651090002454e-05,
"loss": 1.124,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 0.9095622356082349,
"learning_rate": 3.896981023209777e-05,
"loss": 1.1507,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 0.6747308598634724,
"learning_rate": 3.8951817692095966e-05,
"loss": 1.1652,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 0.7213253116881573,
"learning_rate": 3.893367361386491e-05,
"loss": 1.116,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 0.5558115039242595,
"learning_rate": 3.891537814248417e-05,
"loss": 1.1127,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 0.6638303660093081,
"learning_rate": 3.889693142424384e-05,
"loss": 1.0987,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 0.6921772379219434,
"learning_rate": 3.8878333606643405e-05,
"loss": 1.1286,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 0.6424358943672711,
"learning_rate": 3.885958483839049e-05,
"loss": 1.1288,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 0.7039008382363606,
"learning_rate": 3.8840685269399786e-05,
"loss": 1.1593,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 0.6243856388979439,
"learning_rate": 3.882163505079171e-05,
"loss": 1.0794,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 0.6537349998497267,
"learning_rate": 3.880243433489132e-05,
"loss": 1.0691,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 0.7574086325428867,
"learning_rate": 3.878308327522702e-05,
"loss": 1.0518,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 0.5413360874738604,
"learning_rate": 3.876358202652936e-05,
"loss": 1.0773,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 0.8051001749822325,
"learning_rate": 3.8743930744729835e-05,
"loss": 1.035,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.722273212472283,
"learning_rate": 3.8724129586959535e-05,
"loss": 1.0735,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.7634223737947115,
"learning_rate": 3.870417871154801e-05,
"loss": 1.0396,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 0.640401347712926,
"learning_rate": 3.868407827802192e-05,
"loss": 1.0348,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.8698658592001626,
"learning_rate": 3.8663828447103804e-05,
"loss": 1.052,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.917626378726624,
"learning_rate": 3.8643429380710765e-05,
"loss": 1.0963,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.6714008202666484,
"learning_rate": 3.862288124195319e-05,
"loss": 1.0656,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.7100242409067843,
"learning_rate": 3.8602184195133464e-05,
"loss": 1.049,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 0.742561272310745,
"learning_rate": 3.8581338405744604e-05,
"loss": 1.0659,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 0.6539574271462976,
"learning_rate": 3.8560344040468996e-05,
"loss": 1.0394,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.666434382668697,
"learning_rate": 3.853920126717703e-05,
"loss": 1.0641,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 0.6279867425226482,
"learning_rate": 3.851791025492574e-05,
"loss": 1.0577,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.6283616907537433,
"learning_rate": 3.8496471173957515e-05,
"loss": 1.069,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.7167169844323926,
"learning_rate": 3.847488419569866e-05,
"loss": 1.0324,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.706364829170772,
"learning_rate": 3.8453149492758077e-05,
"loss": 1.0313,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 0.780988822576478,
"learning_rate": 3.8431267238925886e-05,
"loss": 1.0646,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.6329604445428967,
"learning_rate": 3.840923760917199e-05,
"loss": 1.0977,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 0.6573933836324901,
"learning_rate": 3.838706077964473e-05,
"loss": 1.1094,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 0.8368329329603137,
"learning_rate": 3.836473692766944e-05,
"loss": 1.0431,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 1.0194682880476533,
"learning_rate": 3.834226623174705e-05,
"loss": 1.0536,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.7463804257502517,
"learning_rate": 3.831964887155264e-05,
"loss": 1.0513,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.7600055186932484,
"learning_rate": 3.829688502793403e-05,
"loss": 1.0793,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.9118267452849349,
"learning_rate": 3.827397488291029e-05,
"loss": 1.0806,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 1.0481119658856406,
"learning_rate": 3.8250918619670343e-05,
"loss": 1.0625,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.8518776715174656,
"learning_rate": 3.822771642257145e-05,
"loss": 1.0401,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.6600113288294617,
"learning_rate": 3.820436847713776e-05,
"loss": 1.0336,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 0.8975885579339071,
"learning_rate": 3.8180874970058826e-05,
"loss": 1.0762,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 1.2447451970522259,
"learning_rate": 3.8157236089188103e-05,
"loss": 1.0691,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.7101666428450536,
"learning_rate": 3.813345202354145e-05,
"loss": 1.0738,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 0.63450306589063,
"learning_rate": 3.810952296329563e-05,
"loss": 1.0656,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 1.1237055974371688,
"learning_rate": 3.808544909978676e-05,
"loss": 1.0355,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.995404455211516,
"learning_rate": 3.806123062550882e-05,
"loss": 1.0197,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.5829658609621662,
"learning_rate": 3.803686773411208e-05,
"loss": 1.0259,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.8334038198176196,
"learning_rate": 3.801236062040158e-05,
"loss": 1.0428,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 1.1735389494402326,
"learning_rate": 3.798770948033553e-05,
"loss": 1.0661,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.7522866270764116,
"learning_rate": 3.796291451102381e-05,
"loss": 1.0493,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.6489691353630607,
"learning_rate": 3.793797591072631e-05,
"loss": 1.0334,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.9509890145187752,
"learning_rate": 3.791289387885144e-05,
"loss": 1.075,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.9013475543389008,
"learning_rate": 3.788766861595443e-05,
"loss": 1.0497,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 0.6986338787912735,
"learning_rate": 3.7862300323735835e-05,
"loss": 1.0396,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 0.7081500152734426,
"learning_rate": 3.783678920503983e-05,
"loss": 1.033,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 1.0806766828015621,
"learning_rate": 3.781113546385265e-05,
"loss": 1.0702,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.8494375114074504,
"learning_rate": 3.7785339305300916e-05,
"loss": 1.0594,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.651091888964896,
"learning_rate": 3.7759400935650025e-05,
"loss": 1.0709,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.6946341309143321,
"learning_rate": 3.77333205623025e-05,
"loss": 1.0733,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.6109292477607564,
"learning_rate": 3.7707098393796295e-05,
"loss": 1.0856,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.5691513228759845,
"learning_rate": 3.7680734639803185e-05,
"loss": 1.0053,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 0.6696749395105275,
"learning_rate": 3.765422951112704e-05,
"loss": 1.0613,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 1.0578827835487337,
"learning_rate": 3.762758321970216e-05,
"loss": 1.0722,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.8474576647725708,
"learning_rate": 3.7600795978591584e-05,
"loss": 1.0832,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.7319496518368656,
"learning_rate": 3.757386800198538e-05,
"loss": 1.0471,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.5668416992475949,
"learning_rate": 3.7546799505198925e-05,
"loss": 1.038,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.9374730368940797,
"learning_rate": 3.7519590704671197e-05,
"loss": 1.0552,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.8335158748231377,
"learning_rate": 3.749224181796305e-05,
"loss": 1.0624,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.7590101267993233,
"learning_rate": 3.746475306375545e-05,
"loss": 1.0764,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.6921752432731415,
"learning_rate": 3.743712466184774e-05,
"loss": 1.0417,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.6289669158823302,
"learning_rate": 3.7409356833155885e-05,
"loss": 1.0686,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.6320965406535632,
"learning_rate": 3.73814497997107e-05,
"loss": 1.0743,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.6411401584160247,
"learning_rate": 3.7353403784656095e-05,
"loss": 1.0446,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.5668220063159032,
"learning_rate": 3.732521901224724e-05,
"loss": 1.0566,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.6749106600717525,
"learning_rate": 3.7296895707848844e-05,
"loss": 1.0347,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.5863950037733027,
"learning_rate": 3.7268434097933275e-05,
"loss": 1.0431,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.6306804067206873,
"learning_rate": 3.723983441007882e-05,
"loss": 1.085,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.8386386938490404,
"learning_rate": 3.721109687296781e-05,
"loss": 1.091,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.5856140120536691,
"learning_rate": 3.718222171638484e-05,
"loss": 1.1001,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 0.5620760782911751,
"learning_rate": 3.715320917121488e-05,
"loss": 1.0928,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.5683465922756271,
"learning_rate": 3.712405946944149e-05,
"loss": 1.0944,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.4837773988461266,
"learning_rate": 3.709477284414488e-05,
"loss": 1.062,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.5645889661906858,
"learning_rate": 3.7065349529500136e-05,
"loss": 1.0518,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 0.5023966126231083,
"learning_rate": 3.7035789760775305e-05,
"loss": 1.0608,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.5922313303599036,
"learning_rate": 3.70060937743295e-05,
"loss": 1.0894,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.5135131330065896,
"learning_rate": 3.6976261807611045e-05,
"loss": 1.045,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.6351397302697024,
"learning_rate": 3.694629409915555e-05,
"loss": 1.0736,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.49073506039369147,
"learning_rate": 3.691619088858401e-05,
"loss": 1.0417,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.5848209739519431,
"learning_rate": 3.68859524166009e-05,
"loss": 1.0752,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.5501841849931649,
"learning_rate": 3.6855578924992226e-05,
"loss": 1.0451,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.6549789978293852,
"learning_rate": 3.6825070656623626e-05,
"loss": 1.0467,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.5415934230285804,
"learning_rate": 3.67944278554384e-05,
"loss": 1.1,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 0.675625797065519,
"learning_rate": 3.676365076645557e-05,
"loss": 1.045,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.5824468668699212,
"learning_rate": 3.673273963576791e-05,
"loss": 1.0577,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.5597018498179365,
"learning_rate": 3.670169471054001e-05,
"loss": 1.0847,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.6245334345550559,
"learning_rate": 3.6670516239006254e-05,
"loss": 1.0701,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.5286657083195474,
"learning_rate": 3.663920447046888e-05,
"loss": 1.0727,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.5428668546236456,
"learning_rate": 3.6607759655295954e-05,
"loss": 1.0444,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.5508290184145147,
"learning_rate": 3.657618204491937e-05,
"loss": 1.0653,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.6296515488168701,
"learning_rate": 3.6544471891832875e-05,
"loss": 1.0738,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.6473235530346321,
"learning_rate": 3.6512629449589983e-05,
"loss": 0.9932,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.8506262322183419,
"learning_rate": 3.6480654972802045e-05,
"loss": 1.0965,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.7852400137135864,
"learning_rate": 3.644854871713611e-05,
"loss": 1.0691,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.5029227638411485,
"learning_rate": 3.6416310939312924e-05,
"loss": 1.0378,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.584598185923697,
"learning_rate": 3.638394189710493e-05,
"loss": 1.014,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.7414990096929093,
"learning_rate": 3.635144184933412e-05,
"loss": 1.0513,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.5278029049464411,
"learning_rate": 3.6318811055869994e-05,
"loss": 1.0834,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.6264004394619018,
"learning_rate": 3.628604977762752e-05,
"loss": 1.0648,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.7218628427379681,
"learning_rate": 3.6253158276565004e-05,
"loss": 1.0639,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.6321241449677946,
"learning_rate": 3.6220136815682016e-05,
"loss": 1.0453,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.7045279956739332,
"learning_rate": 3.618698565901727e-05,
"loss": 1.0763,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.697243639600045,
"learning_rate": 3.615370507164654e-05,
"loss": 1.023,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.7542340813524014,
"learning_rate": 3.61202953196805e-05,
"loss": 1.0575,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 0.6579117602780626,
"learning_rate": 3.6086756670262654e-05,
"loss": 1.0672,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.7744439829490025,
"learning_rate": 3.6053089391567135e-05,
"loss": 1.0957,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.882534151933419,
"learning_rate": 3.601929375279662e-05,
"loss": 1.0842,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.8060752399959978,
"learning_rate": 3.598537002418012e-05,
"loss": 1.0895,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.6478259306398396,
"learning_rate": 3.5951318476970885e-05,
"loss": 1.0492,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.8485160873391678,
"learning_rate": 3.591713938344416e-05,
"loss": 1.0845,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.7201567394408395,
"learning_rate": 3.588283301689507e-05,
"loss": 1.0515,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 0.9212242250846121,
"learning_rate": 3.5848399651636424e-05,
"loss": 1.075,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.6137618186897199,
"learning_rate": 3.5813839562996473e-05,
"loss": 1.0447,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.6362591778564929,
"learning_rate": 3.5779153027316775e-05,
"loss": 1.0614,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.8241050477087799,
"learning_rate": 3.574434032194994e-05,
"loss": 1.0472,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.6052772489016539,
"learning_rate": 3.5709401725257436e-05,
"loss": 1.0601,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.6224042612653543,
"learning_rate": 3.5674337516607346e-05,
"loss": 1.0323,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.8196464464703571,
"learning_rate": 3.563914797637216e-05,
"loss": 1.0799,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.7265107513632911,
"learning_rate": 3.560383338592649e-05,
"loss": 1.0202,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.5213143005813987,
"learning_rate": 3.556839402764487e-05,
"loss": 1.0508,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.8215265074364294,
"learning_rate": 3.553283018489946e-05,
"loss": 1.0588,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.6542670122736728,
"learning_rate": 3.54971421420578e-05,
"loss": 1.0114,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.5690578032374667,
"learning_rate": 3.5461330184480524e-05,
"loss": 1.0493,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 0.6270987216779925,
"learning_rate": 3.54253945985191e-05,
"loss": 1.0434,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.6285832167905369,
"learning_rate": 3.5389335671513524e-05,
"loss": 1.038,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.5562895883922117,
"learning_rate": 3.5353153691789986e-05,
"loss": 1.0807,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.71957176095894,
"learning_rate": 3.531684894865866e-05,
"loss": 1.06,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.5426608061659263,
"learning_rate": 3.528042173241128e-05,
"loss": 1.0467,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.6102913923908276,
"learning_rate": 3.524387233431892e-05,
"loss": 1.0693,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.4614411248210411,
"learning_rate": 3.520720104662958e-05,
"loss": 0.9726,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.5030265538633998,
"learning_rate": 3.5170408162565904e-05,
"loss": 1.072,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.4766494619795491,
"learning_rate": 3.513349397632281e-05,
"loss": 1.0588,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.4897845589629,
"learning_rate": 3.5096458783065145e-05,
"loss": 1.0883,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.5562380461187335,
"learning_rate": 3.505930287892533e-05,
"loss": 1.0802,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.4964072473009913,
"learning_rate": 3.502202656100099e-05,
"loss": 1.0349,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.4434983265361099,
"learning_rate": 3.498463012735256e-05,
"loss": 1.0875,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.4962961527376089,
"learning_rate": 3.494711387700094e-05,
"loss": 1.066,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.47242217428317174,
"learning_rate": 3.490947810992505e-05,
"loss": 1.0261,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 0.49482753364322457,
"learning_rate": 3.48717231270595e-05,
"loss": 1.0785,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.4643857602953866,
"learning_rate": 3.483384923029211e-05,
"loss": 1.0297,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.524486195671521,
"learning_rate": 3.479585672246156e-05,
"loss": 1.1068,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.4820167625282482,
"learning_rate": 3.4757745907354924e-05,
"loss": 1.0276,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.5985971398793304,
"learning_rate": 3.471951708970527e-05,
"loss": 1.0133,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.46610105829569337,
"learning_rate": 3.468117057518921e-05,
"loss": 1.0659,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.5465217009460402,
"learning_rate": 3.4642706670424426e-05,
"loss": 1.0688,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.46203809549932345,
"learning_rate": 3.460412568296731e-05,
"loss": 1.0286,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.4401474643062085,
"learning_rate": 3.456542792131039e-05,
"loss": 1.0331,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.5255165305830425,
"learning_rate": 3.4526613694879936e-05,
"loss": 1.0888,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.4444365992465688,
"learning_rate": 3.448768331403348e-05,
"loss": 1.0462,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.5596268746253393,
"learning_rate": 3.44486370900573e-05,
"loss": 1.0635,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.510539313608525,
"learning_rate": 3.440947533516396e-05,
"loss": 1.0431,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.5753300895595368,
"learning_rate": 3.4370198362489816e-05,
"loss": 1.041,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 0.6061711478273885,
"learning_rate": 3.43308064860925e-05,
"loss": 1.0691,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.49910104387216864,
"learning_rate": 3.4291300020948393e-05,
"loss": 1.0229,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.6741732127041528,
"learning_rate": 3.425167928295015e-05,
"loss": 1.0602,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.5381949618894597,
"learning_rate": 3.421194458890411e-05,
"loss": 1.0868,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.5529283417868694,
"learning_rate": 3.4172096256527865e-05,
"loss": 1.07,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.507299387380791,
"learning_rate": 3.4132134604447595e-05,
"loss": 1.056,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.5217180326227107,
"learning_rate": 3.409205995219561e-05,
"loss": 1.0927,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.47725065808530187,
"learning_rate": 3.4051872620207765e-05,
"loss": 1.0485,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.46975270665651686,
"learning_rate": 3.40115729298209e-05,
"loss": 1.047,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.6070019702372729,
"learning_rate": 3.397116120327025e-05,
"loss": 1.0507,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.40865192123618144,
"learning_rate": 3.3930637763686916e-05,
"loss": 1.0598,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.5248549320108038,
"learning_rate": 3.389000293509524e-05,
"loss": 1.0299,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.46190316000742787,
"learning_rate": 3.3849257042410225e-05,
"loss": 1.0531,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.4749533259630375,
"learning_rate": 3.380840041143494e-05,
"loss": 0.9987,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.439181351166407,
"learning_rate": 3.376743336885792e-05,
"loss": 1.038,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.5142867850280964,
"learning_rate": 3.372635624225055e-05,
"loss": 1.0613,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.4189891481276028,
"learning_rate": 3.3685169360064445e-05,
"loss": 1.064,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.5080069764795369,
"learning_rate": 3.3643873051628804e-05,
"loss": 1.0604,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.5171315615159465,
"learning_rate": 3.3602467647147814e-05,
"loss": 1.0658,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.4564112845021097,
"learning_rate": 3.356095347769799e-05,
"loss": 1.0757,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.557123370885434,
"learning_rate": 3.351933087522552e-05,
"loss": 1.0281,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.5011538212980309,
"learning_rate": 3.3477600172543626e-05,
"loss": 1.076,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.45820391730128335,
"learning_rate": 3.3435761703329894e-05,
"loss": 1.054,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.4661988062646259,
"learning_rate": 3.3393815802123606e-05,
"loss": 1.0273,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.49497223529744927,
"learning_rate": 3.335176280432308e-05,
"loss": 1.099,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.4874925986260594,
"learning_rate": 3.330960304618295e-05,
"loss": 1.0658,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.49195307924118475,
"learning_rate": 3.326733686481153e-05,
"loss": 1.0936,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.4937569637482176,
"learning_rate": 3.3224964598168084e-05,
"loss": 1.0518,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.453946104314142,
"learning_rate": 3.318248658506012e-05,
"loss": 1.0767,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.7310513320670512,
"learning_rate": 3.313990316514069e-05,
"loss": 1.0787,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.48115363049638055,
"learning_rate": 3.309721467890571e-05,
"loss": 1.0734,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.7589868508813559,
"learning_rate": 3.305442146769115e-05,
"loss": 1.049,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.5737086972272768,
"learning_rate": 3.30115238736704e-05,
"loss": 1.0678,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.564815796981172,
"learning_rate": 3.296852223985148e-05,
"loss": 1.0908,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 0.5146861673861154,
"learning_rate": 3.2925416910074315e-05,
"loss": 1.0611,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.5359818443062468,
"learning_rate": 3.288220822900796e-05,
"loss": 1.0673,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.5330267167133351,
"learning_rate": 3.283889654214788e-05,
"loss": 1.0964,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.4491526780588605,
"learning_rate": 3.279548219581317e-05,
"loss": 1.0538,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.49053696274469377,
"learning_rate": 3.275196553714379e-05,
"loss": 1.0975,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.4501457528050316,
"learning_rate": 3.270834691409778e-05,
"loss": 1.0372,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.48917576936983775,
"learning_rate": 3.266462667544849e-05,
"loss": 1.0192,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.4774165986346771,
"learning_rate": 3.262080517078178e-05,
"loss": 1.0809,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.5049325140500066,
"learning_rate": 3.257688275049323e-05,
"loss": 1.0765,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.4677583056477508,
"learning_rate": 3.253285976578535e-05,
"loss": 1.0747,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.529268054043032,
"learning_rate": 3.2488736568664756e-05,
"loss": 1.0594,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.5107251886933598,
"learning_rate": 3.244451351193935e-05,
"loss": 1.0488,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 1.126662818209734,
"learning_rate": 3.240019094921551e-05,
"loss": 1.1106,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.5463301926273862,
"learning_rate": 3.235576923489527e-05,
"loss": 1.0232,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.6756712038068936,
"learning_rate": 3.2311248724173476e-05,
"loss": 1.0538,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.6625493374513745,
"learning_rate": 3.226662977303494e-05,
"loss": 1.0581,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.6048993518394153,
"learning_rate": 3.2221912738251626e-05,
"loss": 1.1004,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.8781416478125487,
"learning_rate": 3.2177097977379736e-05,
"loss": 1.0977,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.5237142943848294,
"learning_rate": 3.21321858487569e-05,
"loss": 1.0478,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.5517258875086348,
"learning_rate": 3.208717671149931e-05,
"loss": 1.0415,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.542887581507311,
"learning_rate": 3.204207092549882e-05,
"loss": 1.0698,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 0.6134676616234251,
"learning_rate": 3.19968688514201e-05,
"loss": 1.0495,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.43210502348878294,
"learning_rate": 3.195157085069774e-05,
"loss": 1.0886,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.49404198905808444,
"learning_rate": 3.190617728553332e-05,
"loss": 1.0506,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.5796902792708722,
"learning_rate": 3.1860688518892606e-05,
"loss": 1.0917,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.69865423885493,
"learning_rate": 3.181510491450254e-05,
"loss": 1.0684,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.5178197032099471,
"learning_rate": 3.176942683684842e-05,
"loss": 1.0583,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.6388344688732914,
"learning_rate": 3.1723654651170934e-05,
"loss": 1.1121,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.5701341127957145,
"learning_rate": 3.167778872346327e-05,
"loss": 1.0598,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.5135115662772876,
"learning_rate": 3.163182942046815e-05,
"loss": 1.0867,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.6250890888655332,
"learning_rate": 3.158577710967494e-05,
"loss": 1.0358,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.5738361677136236,
"learning_rate": 3.153963215931669e-05,
"loss": 1.0624,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.6036892469305825,
"learning_rate": 3.14933949383672e-05,
"loss": 1.0404,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.515329029805921,
"learning_rate": 3.144706581653805e-05,
"loss": 1.0719,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.6138000264843487,
"learning_rate": 3.140064516427566e-05,
"loss": 1.0647,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.5185824039745996,
"learning_rate": 3.135413335275833e-05,
"loss": 1.0558,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.5189872624292938,
"learning_rate": 3.130753075389327e-05,
"loss": 1.0479,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.44910488662692866,
"learning_rate": 3.1260837740313625e-05,
"loss": 1.0524,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.4234750106171584,
"learning_rate": 3.121405468537547e-05,
"loss": 1.05,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.48006698344898596,
"learning_rate": 3.116718196315488e-05,
"loss": 1.0818,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.5364004263762511,
"learning_rate": 3.1120219948444894e-05,
"loss": 1.0584,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.4601899045028249,
"learning_rate": 3.107316901675252e-05,
"loss": 1.0311,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.42419944087118977,
"learning_rate": 3.1026029544295745e-05,
"loss": 1.0643,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.5437590393520472,
"learning_rate": 3.097880190800057e-05,
"loss": 1.0436,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.4421604474439033,
"learning_rate": 3.093148648549788e-05,
"loss": 1.0481,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.5130506713321273,
"learning_rate": 3.088408365512055e-05,
"loss": 1.1177,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.5440470548962699,
"learning_rate": 3.083659379590034e-05,
"loss": 1.0241,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.4564658766139945,
"learning_rate": 3.0789017287564913e-05,
"loss": 1.0747,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.5572976009093089,
"learning_rate": 3.074135451053475e-05,
"loss": 1.0963,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.6633635088155329,
"learning_rate": 3.069360584592015e-05,
"loss": 1.0446,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.46048459429897187,
"learning_rate": 3.064577167551816e-05,
"loss": 1.0675,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.575292920166596,
"learning_rate": 3.059785238180953e-05,
"loss": 1.038,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.5177018534858858,
"learning_rate": 3.054984834795566e-05,
"loss": 1.0662,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.493537116196442,
"learning_rate": 3.0501759957795537e-05,
"loss": 1.0431,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.49309901893686864,
"learning_rate": 3.0453587595842627e-05,
"loss": 1.0411,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.5737354606285742,
"learning_rate": 3.0405331647281856e-05,
"loss": 1.0598,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.5666988271868785,
"learning_rate": 3.0356992497966508e-05,
"loss": 1.0686,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.6270764970080948,
"learning_rate": 3.0308570534415137e-05,
"loss": 1.0531,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.6227870872904515,
"learning_rate": 3.0260066143808467e-05,
"loss": 1.1028,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.6976635733218007,
"learning_rate": 3.021147971398632e-05,
"loss": 1.0872,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.7078551216411557,
"learning_rate": 3.0162811633444478e-05,
"loss": 1.0698,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.5812429961013541,
"learning_rate": 3.0114062291331635e-05,
"loss": 1.0695,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.6299908150453891,
"learning_rate": 3.0065232077446217e-05,
"loss": 1.0807,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.6262546012341539,
"learning_rate": 3.001632138223332e-05,
"loss": 1.0246,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.6676853796931481,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.9571,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 0.9227734836562714,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.9873,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 0.6746150235458205,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.9883,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.6242513663095817,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.9483,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.5834179716141904,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.9973,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.6671257773114515,
"learning_rate": 2.972118904869376e-05,
"loss": 0.9729,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.5886451711803886,
"learning_rate": 2.9671725970312563e-05,
"loss": 1.0063,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.8004006692973887,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.9409,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.5356848043994012,
"learning_rate": 2.95725682050846e-05,
"loss": 0.8855,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.7796155923342688,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.9559,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.5946243761002199,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.9676,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.740310708213363,
"learning_rate": 2.942325848714505e-05,
"loss": 0.9491,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.572586511741152,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.9727,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.6433386109509598,
"learning_rate": 2.932334127129686e-05,
"loss": 0.9462,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.5082101681535388,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.9797,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.49703338690398924,
"learning_rate": 2.922312585929131e-05,
"loss": 0.9662,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.5253571587342661,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.9483,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.5076251038131386,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.9586,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.5992232335977219,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.9892,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.5110202015429519,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.9463,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.4689094772450536,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.9831,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.5623019764601669,
"learning_rate": 2.8920722546128863e-05,
"loss": 1.0011,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.47424573256675867,
"learning_rate": 2.887006998351372e-05,
"loss": 0.9645,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.6041725918338372,
"learning_rate": 2.881934649605535e-05,
"loss": 0.9858,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.48776645756748893,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.975,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.5071179560283076,
"learning_rate": 2.8717688369507425e-05,
"loss": 1.0144,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.5726868622757821,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.9615,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.4306254330930751,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.9678,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.5251356589932966,
"learning_rate": 2.856467940121025e-05,
"loss": 0.9536,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.46769805077736665,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.9813,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.5098008888749036,
"learning_rate": 2.846233032790149e-05,
"loss": 0.9896,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.4929850521986396,
"learning_rate": 2.84110540896642e-05,
"loss": 0.93,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.4986721776345704,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.9427,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.5570911379283031,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.9559,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.5785696739689463,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.9631,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.517278371571282,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.9581,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.6873859010256615,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.9891,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.4695518391126016,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.9622,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.5989364698326688,
"learning_rate": 2.805026032173773e-05,
"loss": 0.9706,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.5886480376231554,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.9505,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.4769538257329641,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.9593,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.6289105918096688,
"learning_rate": 2.7894660620408e-05,
"loss": 0.9922,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.45461925588786656,
"learning_rate": 2.784266724883173e-05,
"loss": 0.9784,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.5779718942215372,
"learning_rate": 2.7790611167497128e-05,
"loss": 1.0072,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.6194311987851115,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.9669,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 2.9540151617765718,
"learning_rate": 2.768631254100716e-05,
"loss": 0.9351,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.71492594140357,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.9561,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.698142171222291,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.9784,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.4377728332872389,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.9988,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.6404545116305363,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.9916,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.4887263922878127,
"learning_rate": 2.742449775134085e-05,
"loss": 0.9685,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.579574017697458,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.9655,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.5998814775742671,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.9298,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.487106385807781,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.9602,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.6704237753917207,
"learning_rate": 2.7213974824975785e-05,
"loss": 1.0223,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.4870849075956372,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.9659,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.5568461924834262,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.9587,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.5221852475384431,
"learning_rate": 2.705547548999996e-05,
"loss": 1.0065,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.5064423416350499,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.965,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.5745333953898081,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.993,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.45212005940835803,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.9357,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.5125093803050575,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.9634,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.36758547400875446,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.9551,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.6332817936826868,
"learning_rate": 2.67369650552155e-05,
"loss": 0.9644,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.5176228435823944,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.9384,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.4203729393226097,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.9461,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.5127838890867554,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.9503,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.4971740072450579,
"learning_rate": 2.652354173593614e-05,
"loss": 0.9959,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.42759868682967034,
"learning_rate": 2.647005443331317e-05,
"loss": 1.0197,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.4402956985462504,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.995,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.4221032302286683,
"learning_rate": 2.636292505301542e-05,
"loss": 0.9511,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.4991896122560376,
"learning_rate": 2.630928383194431e-05,
"loss": 0.9612,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.4854275354136225,
"learning_rate": 2.625559216200781e-05,
"loss": 0.9705,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.44805537474547363,
"learning_rate": 2.620185047252308e-05,
"loss": 0.9459,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.43283120821118554,
"learning_rate": 2.614805919320723e-05,
"loss": 0.9766,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.4443954711557908,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.9838,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.41727905372023816,
"learning_rate": 2.60403295859298e-05,
"loss": 0.9623,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.43802936346083865,
"learning_rate": 2.5986392119371297e-05,
"loss": 1.0113,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.5452215867942365,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.9725,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.39696162546321845,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.9385,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.4886003852790203,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.9647,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.49857156096394917,
"learning_rate": 2.577016790136453e-05,
"loss": 0.9833,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.44018973958026275,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.9781,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.45855269200293003,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.983,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.5006418302192106,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.9391,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.453724798368577,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.9814,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.5029591450884874,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.9533,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.46470895927813494,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.9222,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.4408188116165881,
"learning_rate": 2.539001591449889e-05,
"loss": 0.9484,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.4648374060483106,
"learning_rate": 2.53355326151989e-05,
"loss": 0.9478,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.42106625864963354,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.9733,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.4075124665627105,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.9973,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.4397006819602384,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.9948,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.4250974255010926,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.9647,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.36613895417264164,
"learning_rate": 2.506248490185627e-05,
"loss": 0.9558,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.45952859336399365,
"learning_rate": 2.500775217241923e-05,
"loss": 0.9639,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.5168092025555308,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.9962,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.4611802649017691,
"learning_rate": 2.4898167025948e-05,
"loss": 0.9569,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.6233117152104348,
"learning_rate": 2.4843315485153703e-05,
"loss": 1.0024,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.4833601701710105,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.9829,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.5436689305494857,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.9662,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.5268072806833968,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.9896,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.37611675367588143,
"learning_rate": 2.46235264422994e-05,
"loss": 0.9382,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.40944898664504914,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.9731,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.5027503050339424,
"learning_rate": 2.451340834511592e-05,
"loss": 0.9877,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.4833471375601258,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.9781,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.493307356115667,
"learning_rate": 2.440314589184958e-05,
"loss": 0.9269,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.5640309175374231,
"learning_rate": 2.434796163357541e-05,
"loss": 0.9589,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.6928655378683046,
"learning_rate": 2.429274260911588e-05,
"loss": 0.9907,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.5176595610899134,
"learning_rate": 2.423748926000084e-05,
"loss": 0.9763,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.6314166370258304,
"learning_rate": 2.418220202803459e-05,
"loss": 1.0056,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.42765993045611816,
"learning_rate": 2.412688135529237e-05,
"loss": 0.944,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.6247498343154921,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.9686,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.4507707231973821,
"learning_rate": 2.4016141457114372e-05,
"loss": 1.0066,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.4342573538280011,
"learning_rate": 2.396072311715187e-05,
"loss": 0.9726,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.45749670098069267,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.9295,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.42113962486957385,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.9812,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.42412107471854654,
"learning_rate": 2.379427985200223e-05,
"loss": 0.9705,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.3978261181213712,
"learning_rate": 2.373873749394972e-05,
"loss": 0.9687,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.4433514456668265,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.9579,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.41523400146330874,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.9475,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.4196136802589086,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.9598,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.4946899840938009,
"learning_rate": 2.351627355800233e-05,
"loss": 0.9892,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.3681686542499818,
"learning_rate": 2.346058617160256e-05,
"loss": 0.9546,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.45472345025541167,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.9336,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.3971140994609525,
"learning_rate": 2.334912883203469e-05,
"loss": 0.9923,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.4507856371903565,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.999,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.47755026005635864,
"learning_rate": 2.323756437450136e-05,
"loss": 0.9288,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.4832538363769859,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.9737,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.46583534523224923,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.9364,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.4530345235503499,
"learning_rate": 2.307002464849052e-05,
"loss": 0.9443,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.48816969338306937,
"learning_rate": 2.30141283818846e-05,
"loss": 0.9546,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.4380267848530856,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.9498,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.5140535632068574,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.9918,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.44609550877619264,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.9648,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.39846386766685743,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.9781,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.5424752440996405,
"learning_rate": 2.273429448107257e-05,
"loss": 0.9905,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.45953105255863286,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.9949,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.5131154493475529,
"learning_rate": 2.262220474347102e-05,
"loss": 0.9768,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.5078568790724532,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.9531,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.4182617414860122,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.9735,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.5631777648235136,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.9204,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.41680555900651595,
"learning_rate": 2.2397777251561302e-05,
"loss": 1.0001,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.4212102645156039,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.9473,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.42658543554456063,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.9686,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.380150163494584,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.9485,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.5048210473686359,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.9853,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.4621815692551444,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.947,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.46242234261766124,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.9442,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.5125358443522363,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.9675,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.39080834804360415,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.9953,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.5422572826363028,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.9786,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.4108300511962947,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.9498,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.4683164810158951,
"learning_rate": 2.17791064266051e-05,
"loss": 0.9503,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.4302177828520039,
"learning_rate": 2.172276926861741e-05,
"loss": 0.9543,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.4748848454619635,
"learning_rate": 2.166641833541255e-05,
"loss": 0.9677,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.40896829273136664,
"learning_rate": 2.161005407757109e-05,
"loss": 0.9436,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.4507773392028119,
"learning_rate": 2.155367694578013e-05,
"loss": 0.9811,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.5183081162963791,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.9766,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.36414382325255146,
"learning_rate": 2.1440885863609208e-05,
"loss": 1.002,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.4487473115010249,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.97,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.3979056717036823,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.9753,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.5369048845555157,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.9341,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.485325006594555,
"learning_rate": 2.1215169053088825e-05,
"loss": 1.0097,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.46295796542643536,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.9726,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.41954698574034244,
"learning_rate": 2.110225054402733e-05,
"loss": 0.9993,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.4753896632821536,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.9751,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.39392631509493664,
"learning_rate": 2.09892967807724e-05,
"loss": 0.973,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.46464112178365186,
"learning_rate": 2.093280780774835e-05,
"loss": 0.987,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.4169954484577574,
"learning_rate": 2.087631137601793e-05,
"loss": 0.9781,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.6664597768145081,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.9995,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.4235735066680114,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.97,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.5471392011948052,
"learning_rate": 2.070678184630458e-05,
"loss": 0.9793,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.4521649479892543,
"learning_rate": 2.06502600977305e-05,
"loss": 0.9481,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.4812805391342667,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.9894,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.4973163520687782,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.9836,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.4619919615006928,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.9928,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.4689310415348912,
"learning_rate": 2.042412562887594e-05,
"loss": 0.9687,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.42369880442271735,
"learning_rate": 2.0367582403231033e-05,
"loss": 1.0047,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.43416963346219645,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.9218,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.4699760889898711,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.9859,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.4352918416717998,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.9664,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.45342136734030297,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.9586,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.5128847106989224,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.9636,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.4197975411023253,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.9514,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.48426151766349124,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.9618,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.3434878818206008,
"learning_rate": 1.99151687695709e-05,
"loss": 0.9998,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.5007154750340169,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.9386,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.4496825211978448,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.9199,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.44906977014301364,
"learning_rate": 1.974551241345447e-05,
"loss": 0.966,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.5014340686780795,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.9613,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.38691664649309243,
"learning_rate": 1.963241759676897e-05,
"loss": 0.9763,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.4235670947391535,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.9603,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.360660257006403,
"learning_rate": 1.951933453677638e-05,
"loss": 0.9768,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.45763949003675397,
"learning_rate": 1.9462798545816932e-05,
"loss": 1.003,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.37933197169590177,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.9426,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.430533590233685,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.9725,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.3991646547306603,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.9656,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.3913598732276072,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.9692,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.3573304931447671,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.979,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.3483354348339445,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.9372,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.33027035894878165,
"learning_rate": 1.906719219225165e-05,
"loss": 0.9424,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.3901190210045387,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.9568,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.34679531927776197,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.9588,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.37676333329026074,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.9597,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.38184357276158903,
"learning_rate": 1.884128556891707e-05,
"loss": 0.9621,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.3739261857490889,
"learning_rate": 1.878483094691119e-05,
"loss": 0.9722,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.34941931094441897,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.9519,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.3685868065262604,
"learning_rate": 1.867195130360947e-05,
"loss": 0.991,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.38171768672155226,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.9795,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.397877505921855,
"learning_rate": 1.8559114136390795e-05,
"loss": 1.0058,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.3841906813852783,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.9902,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.3880462848724178,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.9707,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.4707708731628955,
"learning_rate": 1.838994592242891e-05,
"loss": 0.9879,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.3914270963025876,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.9471,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.40934148158620104,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.9765,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.4647575840446459,
"learning_rate": 1.82208935733949e-05,
"loss": 0.9714,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.37901913282489175,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.9568,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.41036105501375064,
"learning_rate": 1.810826238483892e-05,
"loss": 0.9647,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.4623727640029296,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.9499,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.39965130601508425,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.9913,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.5026982437467358,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.9586,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.4292319093607236,
"learning_rate": 1.788318512320838e-05,
"loss": 1.0023,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.40419487937106074,
"learning_rate": 1.782695699829051e-05,
"loss": 0.9599,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.4362265231495698,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.9601,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.4230187067283155,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.9409,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.3764524733325473,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.9519,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.39875668735643094,
"learning_rate": 1.76022227484387e-05,
"loss": 0.9798,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.42246619525935003,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.9544,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.43601215470225835,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.9873,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.3591255360363613,
"learning_rate": 1.743387180005809e-05,
"loss": 0.9685,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.43617362119007136,
"learning_rate": 1.737779525652899e-05,
"loss": 1.0012,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.37770956312308746,
"learning_rate": 1.732173968007914e-05,
"loss": 0.9923,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.36699186754103297,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.9824,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.43910777343662816,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.9126,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.3869401940316402,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.9766,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.36660372109092826,
"learning_rate": 1.709773600686015e-05,
"loss": 0.9584,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.6124462076439555,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.9937,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.37240532161525647,
"learning_rate": 1.69858716181154e-05,
"loss": 0.9548,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.48008381600454136,
"learning_rate": 1.692997535150948e-05,
"loss": 0.9726,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.38296066031727716,
"learning_rate": 1.687410363273884e-05,
"loss": 0.9262,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.4401938378769744,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.9577,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.4031565577371851,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.9754,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.4596597545211076,
"learning_rate": 1.67066402299235e-05,
"loss": 0.9631,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.36286541761456736,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.9579,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.3977642700102292,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.9718,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.46056873391417563,
"learning_rate": 1.653941382839745e-05,
"loss": 0.9731,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.6160939538433302,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.8983,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.49577774900069976,
"learning_rate": 1.642806717162757e-05,
"loss": 0.9113,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.9358643730951103,
"learning_rate": 1.637243646233718e-05,
"loss": 0.8876,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.8323764623138731,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.8905,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.7041986628474554,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.8862,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.752039340042091,
"learning_rate": 1.620572014799777e-05,
"loss": 0.8533,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.5271708956269997,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.832,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.7416183524064531,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.8639,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.529631090765108,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.8784,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.6339341630765161,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.8967,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.5661162992089498,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.871,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.5869473327082323,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.8586,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.6697910426038437,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.8501,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.6876620150870485,
"learning_rate": 1.576251073999917e-05,
"loss": 0.8922,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.6251447611374782,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.8665,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.6451307103240754,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.8486,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.5867941903281395,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.8345,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.5899320446785735,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.8815,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.48712877425048795,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.8312,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.5689027714765984,
"learning_rate": 1.543151434154755e-05,
"loss": 0.87,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.41708257796467374,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.8554,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.548988161782812,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.8731,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.5485447201884545,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.8689,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.4901185623909021,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.8418,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.5335362882410097,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.8501,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.4076015243315452,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.8151,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.4767688223349791,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.8837,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.4773692976397019,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.8841,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.34679331130823154,
"learning_rate": 1.493751509814374e-05,
"loss": 0.8662,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.4378002806114679,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.8564,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.444032912288178,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.8875,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.4051968136626235,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.8713,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.4747638542499672,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.8728,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.40737003669559635,
"learning_rate": 1.466446738480111e-05,
"loss": 0.8826,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.48570838436109554,
"learning_rate": 1.460998408550112e-05,
"loss": 0.8722,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.3765483555199967,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.853,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.37000196945197733,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.8484,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.3756504451511358,
"learning_rate": 1.444679451900437e-05,
"loss": 0.8909,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.3818648422186509,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.8695,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.36083205837643145,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.8155,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.39382750482228013,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.8816,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.39237503521554123,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.9037,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.36707606679928123,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.8788,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.3920968107493961,
"learning_rate": 1.412162598317602e-05,
"loss": 0.8991,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.48508847531611493,
"learning_rate": 1.406759321421907e-05,
"loss": 0.9096,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.36101748627252195,
"learning_rate": 1.401360788062871e-05,
"loss": 0.8641,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.38190967653025926,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.8755,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.4090523612302533,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.8527,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.38312367113907225,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.9118,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.36889930791336956,
"learning_rate": 1.379814952747693e-05,
"loss": 0.8692,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.36547063266761204,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.8805,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.3540802327641334,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.8721,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.3453960923702268,
"learning_rate": 1.363707494698459e-05,
"loss": 0.928,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.3601591709281372,
"learning_rate": 1.358348460369265e-05,
"loss": 0.8574,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.3654641301121,
"learning_rate": 1.352994556668684e-05,
"loss": 0.8821,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.3402505011430359,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.8924,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.3958411678344889,
"learning_rate": 1.342302312350678e-05,
"loss": 0.8679,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.3878910259068604,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.85,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.38512523473079413,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.9018,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.39371210554303093,
"learning_rate": 1.326303494478451e-05,
"loss": 0.873,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.3824784130828964,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.8549,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.3452670633344668,
"learning_rate": 1.315664479122898e-05,
"loss": 0.8743,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.36377691392009914,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.872,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.35710333507689435,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.8715,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.3778450930749334,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.8921,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.33449449146906945,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.9011,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.33267155579476454,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.8508,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.36857636849825987,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.884,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.33826996687920596,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.8551,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.36435606024357925,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.8609,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.3965297183009045,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.8864,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.3498243399058178,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.8567,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.37797013239619986,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.8803,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.3522201290741201,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.8685,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.3878292177590774,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.8992,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.3437097060975112,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.925,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.4150296960966345,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.8609,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.3372091063531155,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.8739,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.3499657925096952,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.8456,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.35193071306561186,
"learning_rate": 1.220938883250288e-05,
"loss": 0.8387,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.35799927380075,
"learning_rate": 1.215733275116827e-05,
"loss": 0.8757,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.3614299037485971,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.8809,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.34906918019395766,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.872,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.3275202618519123,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.8956,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.3472610308789315,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.8652,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.3478006282534011,
"learning_rate": 1.189800129803139e-05,
"loss": 0.882,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.3447679590549789,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.8957,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.3229770663426597,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.8987,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.33970530580890573,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.8574,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.3401214251255015,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.8993,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.35662340101536966,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.8595,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.3323612122610349,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.8559,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.34375776543721753,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.8367,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.323725753069784,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.8889,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.3309982835599207,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.9305,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.31723797758223843,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.8262,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.34559047594835857,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.854,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.3473193182875884,
"learning_rate": 1.128231163049258e-05,
"loss": 0.9083,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.33678202241468724,
"learning_rate": 1.123144751066259e-05,
"loss": 0.9153,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.32675000367617957,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.8764,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.3394085068072581,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.8908,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.3492751305882892,
"learning_rate": 1.107927745387114e-05,
"loss": 0.8587,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.3872101820052885,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.9153,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.3557178744287592,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.887,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.410733817546045,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.8658,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.3205243696869335,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.8937,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.3886895227235977,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.914,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.4066892916917628,
"learning_rate": 1.077687414070869e-05,
"loss": 0.8772,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.36084802945957795,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.876,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.4006312668864561,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.9068,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.39109700079669846,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.8799,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.4483851301822993,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.867,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.33557745394077537,
"learning_rate": 1.052689572753247e-05,
"loss": 0.8814,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.3440452170209105,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.8869,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.36190951672307625,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.8632,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.35996018036168864,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.8836,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.3511411936486199,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.8611,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.3529534832193384,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.8723,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.3612143797603554,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.8177,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.3891646339204038,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.8658,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.32395034770164216,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.8941,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.3654161605985213,
"learning_rate": 1.008173988718005e-05,
"loss": 0.8687,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.34701234452001806,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.8909,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.3491350569780909,
"learning_rate": 9.983678617766677e-06,
"loss": 0.9101,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.3347817073954634,
"learning_rate": 9.934767922553788e-06,
"loss": 0.8915,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.3413180784978985,
"learning_rate": 9.885937708668371e-06,
"loss": 0.8978,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.3941102460049602,
"learning_rate": 9.837188366555524e-06,
"loss": 0.8501,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.34530758521240157,
"learning_rate": 9.788520286013691e-06,
"loss": 0.856,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.35734441947679835,
"learning_rate": 9.739933856191537e-06,
"loss": 0.8664,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.3840330702354431,
"learning_rate": 9.691429465584866e-06,
"loss": 0.8942,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.3182607054914394,
"learning_rate": 9.643007502033492e-06,
"loss": 0.8781,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.37640507623565644,
"learning_rate": 9.594668352718152e-06,
"loss": 0.8475,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.3731198677352327,
"learning_rate": 9.546412404157383e-06,
"loss": 0.8633,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.3433030958860107,
"learning_rate": 9.498240042204471e-06,
"loss": 0.869,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.35184373526236906,
"learning_rate": 9.450151652044339e-06,
"loss": 0.8624,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.3278963884231831,
"learning_rate": 9.402147618190472e-06,
"loss": 0.9044,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.3309458926250334,
"learning_rate": 9.354228324481841e-06,
"loss": 0.8615,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.31727238946159936,
"learning_rate": 9.306394154079852e-06,
"loss": 0.9125,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.4097820763439508,
"learning_rate": 9.258645489465254e-06,
"loss": 0.8533,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.3635368752211317,
"learning_rate": 9.21098271243509e-06,
"loss": 0.8799,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.34487234630902047,
"learning_rate": 9.163406204099659e-06,
"loss": 0.8708,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.3254292157670719,
"learning_rate": 9.115916344879452e-06,
"loss": 0.8923,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.3231949009242057,
"learning_rate": 9.068513514502126e-06,
"loss": 0.8174,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.39449014845654184,
"learning_rate": 9.021198091999439e-06,
"loss": 0.8575,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.3150815481909019,
"learning_rate": 8.973970455704251e-06,
"loss": 0.8733,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.31020185323010735,
"learning_rate": 8.926830983247495e-06,
"loss": 0.8876,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.3320746834965565,
"learning_rate": 8.879780051555116e-06,
"loss": 0.8854,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.32725755863124073,
"learning_rate": 8.83281803684512e-06,
"loss": 0.8877,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.3155711514130716,
"learning_rate": 8.785945314624532e-06,
"loss": 0.8685,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.3095487279519808,
"learning_rate": 8.739162259686385e-06,
"loss": 0.884,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.34353271095410687,
"learning_rate": 8.692469246106734e-06,
"loss": 0.9027,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.347558765332325,
"learning_rate": 8.645866647241677e-06,
"loss": 0.8758,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.3124022995633425,
"learning_rate": 8.599354835724348e-06,
"loss": 0.8867,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.6884664605756583,
"learning_rate": 8.552934183461958e-06,
"loss": 0.9089,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.3442774488436335,
"learning_rate": 8.506605061632805e-06,
"loss": 0.9038,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.3132723908453103,
"learning_rate": 8.460367840683308e-06,
"loss": 0.8917,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.3134515795322092,
"learning_rate": 8.414222890325065e-06,
"loss": 0.8886,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.3379834549415887,
"learning_rate": 8.368170579531855e-06,
"loss": 0.8905,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.3379082992411319,
"learning_rate": 8.322211276536733e-06,
"loss": 0.8644,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.42230513794295704,
"learning_rate": 8.27634534882907e-06,
"loss": 0.9031,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.3115760553931749,
"learning_rate": 8.230573163151585e-06,
"loss": 0.8932,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.34913765721726475,
"learning_rate": 8.184895085497466e-06,
"loss": 0.8981,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.39846368506075064,
"learning_rate": 8.139311481107401e-06,
"loss": 0.8943,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.3353532712037025,
"learning_rate": 8.093822714466686e-06,
"loss": 0.897,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.3301688665687742,
"learning_rate": 8.048429149302271e-06,
"loss": 0.8687,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.34397671424943954,
"learning_rate": 8.003131148579902e-06,
"loss": 0.8844,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.3677311605804765,
"learning_rate": 7.957929074501182e-06,
"loss": 0.8653,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.3449951525117154,
"learning_rate": 7.912823288500695e-06,
"loss": 0.8688,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.372397321947319,
"learning_rate": 7.867814151243102e-06,
"loss": 0.9103,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.3692640832877265,
"learning_rate": 7.822902022620267e-06,
"loss": 0.9084,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.3309612871015139,
"learning_rate": 7.778087261748377e-06,
"loss": 0.8822,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.3450129883552686,
"learning_rate": 7.733370226965056e-06,
"loss": 0.8792,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.361695848305926,
"learning_rate": 7.688751275826527e-06,
"loss": 0.9132,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.33691735860207095,
"learning_rate": 7.644230765104735e-06,
"loss": 0.8794,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.3539703306478141,
"learning_rate": 7.599809050784502e-06,
"loss": 0.8894,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.30888105455465836,
"learning_rate": 7.555486488060661e-06,
"loss": 0.8898,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.3769043569817299,
"learning_rate": 7.511263431335249e-06,
"loss": 0.8552,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.3583504932576194,
"learning_rate": 7.467140234214656e-06,
"loss": 0.8648,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.30529911530962145,
"learning_rate": 7.423117249506777e-06,
"loss": 0.8667,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.6612999504463676,
"learning_rate": 7.379194829218231e-06,
"loss": 0.9629,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.3337889582367484,
"learning_rate": 7.335373324551518e-06,
"loss": 0.8702,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.329963819841515,
"learning_rate": 7.291653085902224e-06,
"loss": 0.88,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.33937239073002484,
"learning_rate": 7.248034462856213e-06,
"loss": 0.8439,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.33741557663829286,
"learning_rate": 7.20451780418683e-06,
"loss": 0.87,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.3390414817715081,
"learning_rate": 7.161103457852128e-06,
"loss": 0.876,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.3248543407432924,
"learning_rate": 7.117791770992046e-06,
"loss": 0.8986,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.36101397018266285,
"learning_rate": 7.074583089925688e-06,
"loss": 0.8632,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.33386388484965257,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.8626,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.3255879132862254,
"learning_rate": 6.988476126329604e-06,
"loss": 0.8923,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.3116614498759148,
"learning_rate": 6.945578532308857e-06,
"loss": 0.906,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.333762399767386,
"learning_rate": 6.902785321094301e-06,
"loss": 0.8754,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.34114928750448775,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.8894,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.2983533675140726,
"learning_rate": 6.817513414939892e-06,
"loss": 0.8699,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.328577126290748,
"learning_rate": 6.775035401831924e-06,
"loss": 0.8514,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.2950311364055555,
"learning_rate": 6.732663135188473e-06,
"loss": 0.8475,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.3166991007983124,
"learning_rate": 6.690396953817055e-06,
"loss": 0.8716,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.34557686888599054,
"learning_rate": 6.648237195676927e-06,
"loss": 0.8787,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.3104818895266977,
"learning_rate": 6.606184197876395e-06,
"loss": 0.8772,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.3536155099407264,
"learning_rate": 6.564238296670109e-06,
"loss": 0.9037,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.3099036737186759,
"learning_rate": 6.52239982745638e-06,
"loss": 0.8937,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.32461746124513213,
"learning_rate": 6.480669124774484e-06,
"loss": 0.8605,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.3233186455016826,
"learning_rate": 6.439046522302013e-06,
"loss": 0.8988,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.30094068852611044,
"learning_rate": 6.397532352852191e-06,
"loss": 0.8666,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.3086525956646806,
"learning_rate": 6.356126948371202e-06,
"loss": 0.8789,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.3265065640238558,
"learning_rate": 6.314830639935559e-06,
"loss": 0.8546,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.3285098672153688,
"learning_rate": 6.273643757749448e-06,
"loss": 0.8952,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.3369157986159673,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.9056,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.32062254054615325,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.8879,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.311410614625169,
"learning_rate": 6.150742957589784e-06,
"loss": 0.8842,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.32379029141992594,
"learning_rate": 6.109997064904774e-06,
"loss": 0.88,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.3085681802055483,
"learning_rate": 6.069362236313092e-06,
"loss": 0.864,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.31681189041027064,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.8526,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.2951203322139567,
"learning_rate": 5.988427070179108e-06,
"loss": 0.9041,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.3035815173705235,
"learning_rate": 5.948127379792234e-06,
"loss": 0.8613,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.28665427243374103,
"learning_rate": 5.90794004780439e-06,
"loss": 0.9179,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.3293158304736524,
"learning_rate": 5.867865395552405e-06,
"loss": 0.8709,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.30236664697403837,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.8843,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.3237910494345962,
"learning_rate": 5.788055411095892e-06,
"loss": 0.8724,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.3247570080752945,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.8552,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.2963639200101463,
"learning_rate": 5.708699979051613e-06,
"loss": 0.9002,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.37813637471419775,
"learning_rate": 5.66919351390751e-06,
"loss": 0.8847,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 5.026484395567524,
"learning_rate": 5.629801637510188e-06,
"loss": 0.9653,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.3226276133031668,
"learning_rate": 5.590524664836043e-06,
"loss": 0.8703,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.3598803150662204,
"learning_rate": 5.551362909942708e-06,
"loss": 0.8952,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.31864934228560016,
"learning_rate": 5.51231668596653e-06,
"loss": 0.838,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.30623216663567454,
"learning_rate": 5.47338630512007e-06,
"loss": 0.8667,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.32652114408798716,
"learning_rate": 5.434572078689617e-06,
"loss": 0.9045,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.35344100444656096,
"learning_rate": 5.395874317032694e-06,
"loss": 0.8712,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.32338421102077175,
"learning_rate": 5.357293329575573e-06,
"loss": 0.8635,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.3013155404694907,
"learning_rate": 5.318829424810797e-06,
"loss": 0.8963,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.31856872838675404,
"learning_rate": 5.280482910294728e-06,
"loss": 0.857,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.3220626222656101,
"learning_rate": 5.242254092645078e-06,
"loss": 0.872,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.30415450656004833,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.8648,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.34772609078349836,
"learning_rate": 5.166150769707892e-06,
"loss": 0.878,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.3116234115857467,
"learning_rate": 5.128276872940503e-06,
"loss": 0.8898,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.3119759385325859,
"learning_rate": 5.090521890074953e-06,
"loss": 0.8465,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.30015169498246064,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.8748,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.33712919129340413,
"learning_rate": 5.015369872647442e-06,
"loss": 0.8817,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.3154604548684715,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.8638,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.3142871802401897,
"learning_rate": 4.940697121074675e-06,
"loss": 0.8619,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.31973156784337087,
"learning_rate": 4.903541216934864e-06,
"loss": 0.8901,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.2988655074787824,
"learning_rate": 4.8665060236772e-06,
"loss": 0.8638,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.30657002329109534,
"learning_rate": 4.829591837434102e-06,
"loss": 0.8603,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.3232563376191705,
"learning_rate": 4.792798953370424e-06,
"loss": 0.9247,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.3158681427813082,
"learning_rate": 4.756127665681083e-06,
"loss": 0.8926,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.31733456049121817,
"learning_rate": 4.719578267588725e-06,
"loss": 0.8737,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.3146922292388705,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.8558,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.5103851646131561,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.8149,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.4218424619724515,
"learning_rate": 4.610664328486485e-06,
"loss": 0.7997,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.39025493261395594,
"learning_rate": 4.5746054014809e-06,
"loss": 0.8091,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.559703896704067,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.8313,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.4680534318347236,
"learning_rate": 4.502857857942204e-06,
"loss": 0.7959,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.41678603599529124,
"learning_rate": 4.467169815100549e-06,
"loss": 0.8032,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.508990887149863,
"learning_rate": 4.431605972355137e-06,
"loss": 0.8211,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.45705463856767337,
"learning_rate": 4.396166614073516e-06,
"loss": 0.8269,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.40814681417074916,
"learning_rate": 4.360852023627846e-06,
"loss": 0.7983,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.46480173427390503,
"learning_rate": 4.325662483392656e-06,
"loss": 0.8313,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.4209482068637294,
"learning_rate": 4.29059827474257e-06,
"loss": 0.8059,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.40508388856112026,
"learning_rate": 4.255659678050063e-06,
"loss": 0.8146,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.3558210921252431,
"learning_rate": 4.22084697268323e-06,
"loss": 0.8284,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.3680671345389187,
"learning_rate": 4.186160437003535e-06,
"loss": 0.7728,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.3442253726290396,
"learning_rate": 4.151600348363582e-06,
"loss": 0.8194,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.34678442383956365,
"learning_rate": 4.11716698310493e-06,
"loss": 0.8438,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.39547592594820596,
"learning_rate": 4.082860616555848e-06,
"loss": 0.7856,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.3600402687492049,
"learning_rate": 4.048681523029125e-06,
"loss": 0.8233,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.351788849869779,
"learning_rate": 4.014629975819882e-06,
"loss": 0.8294,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.335294643673735,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.8079,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.3345125230000129,
"learning_rate": 3.94691060843287e-06,
"loss": 0.7934,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.3767219310945382,
"learning_rate": 3.913243329737353e-06,
"loss": 0.8137,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.36187461840054147,
"learning_rate": 3.879704680319503e-06,
"loss": 0.7785,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.32477714717235356,
"learning_rate": 3.846294928353471e-06,
"loss": 0.8131,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.35066961059607116,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.8073,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.3268876743143901,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.8471,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.33162953071534573,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.812,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.34943990792696805,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.8262,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.3673694042125052,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.8394,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.3316851689587801,
"learning_rate": 3.648558150665886e-06,
"loss": 0.845,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.3023610357689909,
"learning_rate": 3.616058102895075e-06,
"loss": 0.7925,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.3384847885555847,
"learning_rate": 3.583689060687079e-06,
"loss": 0.7877,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.3193471245760559,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.8268,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.33352262055801896,
"learning_rate": 3.51934502719796e-06,
"loss": 0.7779,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.3273808981030412,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.8128,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.3195413141386823,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.7815,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.3476570835571762,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.7876,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.3312517408639938,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.807,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.28443218589758035,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.8164,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.29465950833562526,
"learning_rate": 3.32948376099375e-06,
"loss": 0.8079,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.3062114392022538,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.8217,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.3297545691631378,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.8232,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.3039720591327221,
"learning_rate": 3.236349233544438e-06,
"loss": 0.8198,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.3058850994996888,
"learning_rate": 3.205572144561604e-06,
"loss": 0.8323,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.3241656782135017,
"learning_rate": 3.174929343376374e-06,
"loss": 0.8213,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.2965667378182129,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.802,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.3061088354546626,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.7939,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.3165094215939869,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.8266,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.2953968767714831,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.8292,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.3003353170589662,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.833,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.32271619013765535,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.7858,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.30727347711200553,
"learning_rate": 2.964210239224696e-06,
"loss": 0.7994,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.3236357348666755,
"learning_rate": 2.934650470499869e-06,
"loss": 0.8334,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.29058106059286626,
"learning_rate": 2.905227155855128e-06,
"loss": 0.7815,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.29962741250237485,
"learning_rate": 2.875940530558519e-06,
"loss": 0.8181,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.3140564728045957,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.8224,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.317448515796565,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.7687,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.28885945039678584,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.8345,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.43918155927218044,
"learning_rate": 2.760165589921184e-06,
"loss": 0.8269,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.30308846717180116,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.8176,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.3000508935633916,
"learning_rate": 2.703104292151164e-06,
"loss": 0.8086,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 2.71494083793442,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.8056,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.29065614747910185,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.8271,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.28990687779127794,
"learning_rate": 2.618550200289303e-06,
"loss": 0.799,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.30271080560619923,
"learning_rate": 2.59064316684412e-06,
"loss": 0.7672,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.2894301682576959,
"learning_rate": 2.562875338152264e-06,
"loss": 0.8388,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.3050138486045336,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.8235,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.3159387507470509,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.8198,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.31553333432619235,
"learning_rate": 2.480409295328803e-06,
"loss": 0.8053,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.29505207521919485,
"learning_rate": 2.45320049480108e-06,
"loss": 0.8204,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.32042595164797016,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.791,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.28854959913015316,
"learning_rate": 2.399204021408421e-06,
"loss": 0.7894,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.29138401836472305,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.7602,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.3000956153421491,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.8078,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.31362115547604213,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.7834,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.3089379304757348,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.8064,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.30720201897777916,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.8146,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.32321911732025327,
"learning_rate": 2.24059906434998e-06,
"loss": 0.785,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.3441643102315247,
"learning_rate": 2.214660694699091e-06,
"loss": 0.8045,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.3399986200472356,
"learning_rate": 2.188864536147355e-06,
"loss": 0.831,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.2967717773579162,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.8065,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.3226601591781755,
"learning_rate": 2.137699676264171e-06,
"loss": 0.8233,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.3353205814897212,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.794,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.3205666683652532,
"learning_rate": 2.08710612114857e-06,
"loss": 0.8037,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.2859985864748409,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.8146,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.30822894288998914,
"learning_rate": 2.0370854889762e-06,
"loss": 0.8339,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.2964922662903834,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.8201,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.3507949708930202,
"learning_rate": 1.987639379598427e-06,
"loss": 0.8101,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.3235859517302435,
"learning_rate": 1.963132265887919e-06,
"loss": 0.7947,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.32269860440403514,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.7747,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.31151437986564207,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.7914,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.3225589070184838,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.8003,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.302520692761897,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.8018,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.29739687407427023,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.8243,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.3055660550566014,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.82,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.3120166316154685,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.7987,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.3077145992168689,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.8218,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.32282807793951956,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.7788,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.33230153380570115,
"learning_rate": 1.726025117089718e-06,
"loss": 0.7837,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.2992428517940992,
"learning_rate": 1.70311497206598e-06,
"loss": 0.8173,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.3169970362010369,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.7686,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.31607250300940043,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.834,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.30419211897121246,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.797,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.30898529659509283,
"learning_rate": 1.612939220355272e-06,
"loss": 0.8039,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.3054267297263573,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.8149,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.3202663802361201,
"learning_rate": 1.568732761074121e-06,
"loss": 0.821,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.32614584943357466,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.8128,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.3044148287850404,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.7974,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.31032921333045305,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.7926,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.303179923875663,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.8293,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.2955019975625422,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.8089,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.2957241098504014,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.8072,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.2951727366239985,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.8208,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.2979967714259443,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.8117,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.3118612849656152,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.7836,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.30894636536475156,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.7724,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.28483987328221055,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.8176,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.315099912119238,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.7522,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.2916150806157721,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.8174,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.3030959299314956,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.8178,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.3074341056102494,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.8179,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.29264787434759765,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.7938,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.30806985997702346,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.8034,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.30981771809153613,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.821,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.2828721369888072,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.7827,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.2862860800713558,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.8124,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.32029613102069254,
"learning_rate": 1.140415161609505e-06,
"loss": 0.7771,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.29294457469806034,
"learning_rate": 1.121666393356604e-06,
"loss": 0.8136,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.29590207665761453,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.832,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.2806364666357281,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.8224,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.2981181306149069,
"learning_rate": 1.066326386135097e-06,
"loss": 0.8005,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.27547960469568833,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.8184,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.28217340799060553,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.8189,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.2840630126167186,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.7898,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.2942589032860491,
"learning_rate": 9.9465987684501e-07,
"loss": 0.8319,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.2905504839401652,
"learning_rate": 9.771228098856534e-07,
"loss": 0.7874,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.29083217223342617,
"learning_rate": 9.597378493454013e-07,
"loss": 0.8108,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.3644634599054222,
"learning_rate": 9.425051342339353e-07,
"loss": 0.8266,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.31631095687099897,
"learning_rate": 9.254248023435864e-07,
"loss": 0.7644,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.29031532865159126,
"learning_rate": 9.084969902482288e-07,
"loss": 0.8202,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.28332845076530694,
"learning_rate": 8.917218333021993e-07,
"loss": 0.7844,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.29295268997940943,
"learning_rate": 8.750994656392109e-07,
"loss": 0.7887,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.3057790751460989,
"learning_rate": 8.58630020171265e-07,
"loss": 0.8251,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.3152978575653759,
"learning_rate": 8.423136285876099e-07,
"loss": 0.7861,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.291294640521585,
"learning_rate": 8.26150421353682e-07,
"loss": 0.8356,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.3061744636451065,
"learning_rate": 8.101405277100549e-07,
"loss": 0.8274,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.2938184274853799,
"learning_rate": 7.942840756714077e-07,
"loss": 0.8478,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.30943322927933026,
"learning_rate": 7.785811920255093e-07,
"loss": 0.8474,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.2983791518455771,
"learning_rate": 7.630320023322024e-07,
"loss": 0.7852,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.30373415132995923,
"learning_rate": 7.476366309223903e-07,
"loss": 0.8449,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.2790057867935526,
"learning_rate": 7.323952008970537e-07,
"loss": 0.8147,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.29452462673869684,
"learning_rate": 7.173078341262618e-07,
"loss": 0.7755,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.2976909345653335,
"learning_rate": 7.023746512481989e-07,
"loss": 0.7911,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.2947222731568572,
"learning_rate": 6.875957716681902e-07,
"loss": 0.7845,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.2979979714753094,
"learning_rate": 6.729713135577687e-07,
"loss": 0.8265,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.3082472356536131,
"learning_rate": 6.585013938537143e-07,
"loss": 0.7946,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.2920075099823839,
"learning_rate": 6.441861282571115e-07,
"loss": 0.7928,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.2759498231141673,
"learning_rate": 6.300256312324471e-07,
"loss": 0.7959,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.2937296176915284,
"learning_rate": 6.160200160066776e-07,
"loss": 0.7822,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.28322194499900494,
"learning_rate": 6.021693945683327e-07,
"loss": 0.8007,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.3083068879607132,
"learning_rate": 5.884738776666044e-07,
"loss": 0.8302,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.2923237567101745,
"learning_rate": 5.749335748104878e-07,
"loss": 0.7964,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.29582331628112873,
"learning_rate": 5.615485942678778e-07,
"loss": 0.7988,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.31241399928125924,
"learning_rate": 5.483190430647289e-07,
"loss": 0.8099,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.31086285474993586,
"learning_rate": 5.352450269841747e-07,
"loss": 0.8005,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.2791908540998049,
"learning_rate": 5.22326650565701e-07,
"loss": 0.813,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.27998217691969074,
"learning_rate": 5.095640171043048e-07,
"loss": 0.8449,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.29442243929669404,
"learning_rate": 4.969572286496594e-07,
"loss": 0.7969,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.2820881895982896,
"learning_rate": 4.845063860053079e-07,
"loss": 0.7977,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.28744355204347616,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.7715,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.3078946614024075,
"learning_rate": 4.600729351261857e-07,
"loss": 0.7949,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.33534407912821623,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.7539,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.3024940188236861,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.7779,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.281328158039516,
"learning_rate": 4.245948007320588e-07,
"loss": 0.8279,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.2838052874363946,
"learning_rate": 4.130816799401971e-07,
"loss": 0.7812,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.2839351717686519,
"learning_rate": 4.017251756252827e-07,
"loss": 0.8226,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.2883447322519683,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.7949,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.29943531872162954,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.7972,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.30551460797147073,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.8155,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.28206647979449306,
"learning_rate": 3.578671204615347e-07,
"loss": 0.799,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.2981738837659964,
"learning_rate": 3.472950355547267e-07,
"loss": 0.8261,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.29450360463298025,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.7968,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.29326553200671096,
"learning_rate": 3.266223765343224e-07,
"loss": 0.7806,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.29746987021274507,
"learning_rate": 3.16521967718757e-07,
"loss": 0.8063,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.30011014388698715,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.8129,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.2911233546720341,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.7962,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.28614536616856506,
"learning_rate": 2.871653894244597e-07,
"loss": 0.8057,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.2958472128699463,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.8223,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.2982750186122814,
"learning_rate": 2.683823233205485e-07,
"loss": 0.7943,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.28956794816501485,
"learning_rate": 2.592274132703287e-07,
"loss": 0.7863,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.2773035369968297,
"learning_rate": 2.502303499116754e-07,
"loss": 0.8039,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.27628988542680194,
"learning_rate": 2.413912051848666e-07,
"loss": 0.8265,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.3289457199856895,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.7728,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.28734136434101254,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.7947,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.30082012976757816,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.8116,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.30134161608105997,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.8289,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.28478162491215603,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.8156,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.274660450693692,
"learning_rate": 1.916764975761254e-07,
"loss": 0.7919,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.3073109230357621,
"learning_rate": 1.839446914621923e-07,
"loss": 0.7844,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.29965243498729216,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.8109,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.2977512859782371,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.7793,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.2842958376885735,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.851,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.2773680763460668,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.8114,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.28416697430730087,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.8362,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.29239034013334425,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.7954,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.29009544067536985,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.7925,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.29120918829646536,
"learning_rate": 1.277993209368744e-07,
"loss": 0.8196,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.2956218028114307,
"learning_rate": 1.214956261808542e-07,
"loss": 0.7999,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.2835991816022542,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.8291,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.27439754391695675,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.7935,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.2888746005385869,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.823,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.29706429637218507,
"learning_rate": 9.787081204466253e-08,
"loss": 0.7874,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.5391692111350939,
"learning_rate": 9.236233586200627e-08,
"loss": 0.824,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.2954145495695477,
"learning_rate": 8.70130406196279e-08,
"loss": 0.8365,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.27824222771029966,
"learning_rate": 8.182296909035047e-08,
"loss": 0.7925,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.28966339007715836,
"learning_rate": 7.679216277384439e-08,
"loss": 0.818,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.26546905213786737,
"learning_rate": 7.19206618963053e-08,
"loss": 0.8354,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.2844482491815184,
"learning_rate": 6.720850541012347e-08,
"loss": 0.8094,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.27444457277430284,
"learning_rate": 6.265573099357714e-08,
"loss": 0.7904,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.3096195695929248,
"learning_rate": 5.826237505053067e-08,
"loss": 0.807,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.2766415896435269,
"learning_rate": 5.4028472710137e-08,
"loss": 0.7936,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.29113003041053337,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.8622,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.29827827093692466,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.8255,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.28689632573873725,
"learning_rate": 4.228381947004812e-08,
"loss": 0.8138,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.2858667982902077,
"learning_rate": 3.868805732811343e-08,
"loss": 0.7799,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.3037455384882988,
"learning_rate": 3.525190530455636e-08,
"loss": 0.8091,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.30707910382346454,
"learning_rate": 3.197539087475887e-08,
"loss": 0.8074,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.29597646924544097,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.8057,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.285699142698408,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.8121,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.2993024849855605,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.8327,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.2817829376812754,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.7954,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.2800409245050442,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.818,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.2906882180116521,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.8006,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.2829109855910991,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.8177,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.2862111044225296,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.8035,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.3262342337911503,
"learning_rate": 9.67435416023843e-09,
"loss": 0.789,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.3023028044865876,
"learning_rate": 7.995445897566178e-09,
"loss": 0.8206,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.293690708032478,
"learning_rate": 6.476393170578021e-09,
"loss": 0.8288,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.29892283353389204,
"learning_rate": 5.11720812557348e-09,
"loss": 0.8331,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.2764959476929851,
"learning_rate": 3.917901630563492e-09,
"loss": 0.8151,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.2904640000629348,
"learning_rate": 2.878483275170485e-09,
"loss": 0.825,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.30672850933570145,
"learning_rate": 1.998961370555108e-09,
"loss": 0.801,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.28810340906681026,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.7771,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.27677613537438034,
"learning_rate": 7.196337656112384e-10,
"loss": 0.8017,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.30054980176138574,
"learning_rate": 3.198382947511292e-10,
"loss": 0.7813,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.2944339213288348,
"learning_rate": 7.995973352770137e-11,
"loss": 0.8107,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.30326108117828826,
"learning_rate": 0.0,
"loss": 0.8365,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 3.4808888073377546e+18,
"train_loss": 0.9800670660459079,
"train_runtime": 31845.5415,
"train_samples_per_second": 4.961,
"train_steps_per_second": 0.039
}
],
"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": 3.4808888073377546e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}