neginr's picture
End of training
8f42adf verified
raw
history blame
215 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 500,
"global_step": 1235,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004048582995951417,
"grad_norm": 6.881321048999711,
"learning_rate": 3.2258064516129035e-07,
"loss": 0.7985,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 7.345358648272824,
"learning_rate": 6.451612903225807e-07,
"loss": 0.8224,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 6.972151086709958,
"learning_rate": 9.67741935483871e-07,
"loss": 0.8036,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 6.850953807234061,
"learning_rate": 1.2903225806451614e-06,
"loss": 0.7915,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 6.667408550791196,
"learning_rate": 1.6129032258064516e-06,
"loss": 0.7865,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 6.332179938169585,
"learning_rate": 1.935483870967742e-06,
"loss": 0.7861,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 5.2375860154803835,
"learning_rate": 2.2580645161290324e-06,
"loss": 0.7372,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 5.009398225651841,
"learning_rate": 2.580645161290323e-06,
"loss": 0.7596,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 3.0214547398340286,
"learning_rate": 2.903225806451613e-06,
"loss": 0.695,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 2.7599255072162374,
"learning_rate": 3.225806451612903e-06,
"loss": 0.6957,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 2.5565893257482055,
"learning_rate": 3.548387096774194e-06,
"loss": 0.6668,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 3.7616337916812443,
"learning_rate": 3.870967741935484e-06,
"loss": 0.6358,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 3.924570680969341,
"learning_rate": 4.193548387096774e-06,
"loss": 0.6547,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 3.690129710508207,
"learning_rate": 4.516129032258065e-06,
"loss": 0.6374,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 3.3783934808740064,
"learning_rate": 4.838709677419355e-06,
"loss": 0.6035,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 2.7428461400340045,
"learning_rate": 5.161290322580646e-06,
"loss": 0.5563,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 3.029230013940835,
"learning_rate": 5.483870967741935e-06,
"loss": 0.5808,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 2.5886370959686142,
"learning_rate": 5.806451612903226e-06,
"loss": 0.5579,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 2.0937971696952364,
"learning_rate": 6.129032258064517e-06,
"loss": 0.5465,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.572433430405842,
"learning_rate": 6.451612903225806e-06,
"loss": 0.5338,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 1.3177282504248022,
"learning_rate": 6.774193548387097e-06,
"loss": 0.5141,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.8529404350779977,
"learning_rate": 7.096774193548388e-06,
"loss": 0.5214,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.9556497088243536,
"learning_rate": 7.4193548387096784e-06,
"loss": 0.4988,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 1.476075842991829,
"learning_rate": 7.741935483870968e-06,
"loss": 0.4962,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 0.9200560348392686,
"learning_rate": 8.064516129032258e-06,
"loss": 0.4828,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 0.9359579190275144,
"learning_rate": 8.387096774193549e-06,
"loss": 0.4888,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 0.943071076216648,
"learning_rate": 8.70967741935484e-06,
"loss": 0.4708,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 1.0371079447128948,
"learning_rate": 9.03225806451613e-06,
"loss": 0.4721,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 0.8436729179235692,
"learning_rate": 9.35483870967742e-06,
"loss": 0.4643,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 0.7326085731641712,
"learning_rate": 9.67741935483871e-06,
"loss": 0.4826,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 0.8627157007941662,
"learning_rate": 1e-05,
"loss": 0.4581,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 0.9263581921898835,
"learning_rate": 1.0322580645161291e-05,
"loss": 0.4601,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 0.7553282062946043,
"learning_rate": 1.0645161290322582e-05,
"loss": 0.4504,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 0.7581804106731883,
"learning_rate": 1.096774193548387e-05,
"loss": 0.463,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 0.7380989998640367,
"learning_rate": 1.1290322580645164e-05,
"loss": 0.4526,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.7735296607078446,
"learning_rate": 1.1612903225806453e-05,
"loss": 0.4404,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.7070014647979432,
"learning_rate": 1.1935483870967743e-05,
"loss": 0.4162,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.6065207673453902,
"learning_rate": 1.2258064516129034e-05,
"loss": 0.4431,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.7366518209903369,
"learning_rate": 1.2580645161290324e-05,
"loss": 0.4334,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 0.5673623173727267,
"learning_rate": 1.2903225806451613e-05,
"loss": 0.4257,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.5417783644006262,
"learning_rate": 1.3225806451612903e-05,
"loss": 0.4175,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.6375900188857361,
"learning_rate": 1.3548387096774194e-05,
"loss": 0.4348,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.6256295211020415,
"learning_rate": 1.3870967741935486e-05,
"loss": 0.421,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.5627701527191482,
"learning_rate": 1.4193548387096776e-05,
"loss": 0.4089,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.5540926859411169,
"learning_rate": 1.4516129032258066e-05,
"loss": 0.4332,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 0.5948773856308074,
"learning_rate": 1.4838709677419357e-05,
"loss": 0.4126,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.49798888358345106,
"learning_rate": 1.5161290322580646e-05,
"loss": 0.384,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.6079992427422003,
"learning_rate": 1.5483870967741936e-05,
"loss": 0.4034,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 0.49921687358427214,
"learning_rate": 1.5806451612903226e-05,
"loss": 0.4156,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 0.5712502502840897,
"learning_rate": 1.6129032258064517e-05,
"loss": 0.4086,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 0.5089764501390577,
"learning_rate": 1.6451612903225807e-05,
"loss": 0.406,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.577558157507632,
"learning_rate": 1.6774193548387098e-05,
"loss": 0.4178,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 0.58306259961128,
"learning_rate": 1.7096774193548388e-05,
"loss": 0.4258,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.5538590504408656,
"learning_rate": 1.741935483870968e-05,
"loss": 0.395,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.5783014766967746,
"learning_rate": 1.774193548387097e-05,
"loss": 0.3938,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 0.5897624614974293,
"learning_rate": 1.806451612903226e-05,
"loss": 0.3935,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.6056190598284228,
"learning_rate": 1.838709677419355e-05,
"loss": 0.4154,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 0.6778922365763834,
"learning_rate": 1.870967741935484e-05,
"loss": 0.4024,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 0.5496445019224415,
"learning_rate": 1.903225806451613e-05,
"loss": 0.3974,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 0.682176959760106,
"learning_rate": 1.935483870967742e-05,
"loss": 0.4053,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 0.668277754114471,
"learning_rate": 1.967741935483871e-05,
"loss": 0.3875,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 0.6409602929962404,
"learning_rate": 2e-05,
"loss": 0.3918,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 0.7948305284292113,
"learning_rate": 2.0322580645161292e-05,
"loss": 0.3915,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 0.5466240423759453,
"learning_rate": 2.0645161290322582e-05,
"loss": 0.4105,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 0.6267391789155579,
"learning_rate": 2.0967741935483873e-05,
"loss": 0.393,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 0.5245419906248381,
"learning_rate": 2.1290322580645163e-05,
"loss": 0.3797,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 0.6079392675242722,
"learning_rate": 2.161290322580645e-05,
"loss": 0.395,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 0.6542519429504802,
"learning_rate": 2.193548387096774e-05,
"loss": 0.3879,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 0.67127837663462,
"learning_rate": 2.225806451612903e-05,
"loss": 0.3728,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 0.6519014269992068,
"learning_rate": 2.2580645161290328e-05,
"loss": 0.3991,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 0.6413654383141485,
"learning_rate": 2.2903225806451615e-05,
"loss": 0.384,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 0.7761395340311835,
"learning_rate": 2.3225806451612906e-05,
"loss": 0.3812,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 0.6392360251414426,
"learning_rate": 2.3548387096774196e-05,
"loss": 0.3799,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 0.6344126804945156,
"learning_rate": 2.3870967741935486e-05,
"loss": 0.3869,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 0.48952732569006074,
"learning_rate": 2.4193548387096777e-05,
"loss": 0.3903,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 0.5974593572852642,
"learning_rate": 2.4516129032258067e-05,
"loss": 0.3871,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 0.6119391220123684,
"learning_rate": 2.4838709677419358e-05,
"loss": 0.3797,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 0.6505506039192074,
"learning_rate": 2.5161290322580648e-05,
"loss": 0.364,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 0.8456607059587417,
"learning_rate": 2.548387096774194e-05,
"loss": 0.3841,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 0.7711706864790735,
"learning_rate": 2.5806451612903226e-05,
"loss": 0.3853,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 0.6269103271646045,
"learning_rate": 2.6129032258064516e-05,
"loss": 0.3931,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 0.7198112035118428,
"learning_rate": 2.6451612903225806e-05,
"loss": 0.3948,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 0.7752634379246502,
"learning_rate": 2.6774193548387097e-05,
"loss": 0.3733,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 0.623149413928178,
"learning_rate": 2.7096774193548387e-05,
"loss": 0.3857,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 0.6427853487918697,
"learning_rate": 2.7419354838709678e-05,
"loss": 0.3643,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 0.7266940695862811,
"learning_rate": 2.774193548387097e-05,
"loss": 0.3803,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 0.7071244677841849,
"learning_rate": 2.8064516129032262e-05,
"loss": 0.3799,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 0.6328596180888287,
"learning_rate": 2.8387096774193552e-05,
"loss": 0.3796,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 0.6865531842349164,
"learning_rate": 2.8709677419354843e-05,
"loss": 0.3866,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 0.6933938283516059,
"learning_rate": 2.9032258064516133e-05,
"loss": 0.3738,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 0.5852363513185772,
"learning_rate": 2.9354838709677423e-05,
"loss": 0.3773,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 0.6336100828589462,
"learning_rate": 2.9677419354838714e-05,
"loss": 0.38,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 0.6847854172752209,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.3746,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 0.6765196942359786,
"learning_rate": 3.032258064516129e-05,
"loss": 0.3692,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 0.7618485705399742,
"learning_rate": 3.0645161290322585e-05,
"loss": 0.3697,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 1.0091201811047514,
"learning_rate": 3.096774193548387e-05,
"loss": 0.3785,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 0.7545502681722757,
"learning_rate": 3.1290322580645166e-05,
"loss": 0.374,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 0.7401380162579622,
"learning_rate": 3.161290322580645e-05,
"loss": 0.3777,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 0.5829933874761833,
"learning_rate": 3.193548387096774e-05,
"loss": 0.3804,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 0.8801448048118413,
"learning_rate": 3.2258064516129034e-05,
"loss": 0.3734,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 0.6758108001148754,
"learning_rate": 3.258064516129033e-05,
"loss": 0.3653,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 0.7106607932519297,
"learning_rate": 3.2903225806451614e-05,
"loss": 0.3776,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 0.6826624984507279,
"learning_rate": 3.322580645161291e-05,
"loss": 0.3877,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 0.8030132658598782,
"learning_rate": 3.3548387096774195e-05,
"loss": 0.373,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 0.5411799411339533,
"learning_rate": 3.387096774193549e-05,
"loss": 0.3809,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 0.6734573552459837,
"learning_rate": 3.4193548387096776e-05,
"loss": 0.3859,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 0.5683555263239974,
"learning_rate": 3.451612903225807e-05,
"loss": 0.3696,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 0.5815672551083655,
"learning_rate": 3.483870967741936e-05,
"loss": 0.3816,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 0.7049240596973112,
"learning_rate": 3.516129032258065e-05,
"loss": 0.378,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 0.5331900168194205,
"learning_rate": 3.548387096774194e-05,
"loss": 0.3727,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 0.5522496180917865,
"learning_rate": 3.5806451612903225e-05,
"loss": 0.3695,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 0.6620617359908778,
"learning_rate": 3.612903225806452e-05,
"loss": 0.3747,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 0.7412721595951489,
"learning_rate": 3.6451612903225805e-05,
"loss": 0.3845,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 0.6782348254050736,
"learning_rate": 3.67741935483871e-05,
"loss": 0.3796,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 0.6652637557897376,
"learning_rate": 3.7096774193548386e-05,
"loss": 0.3661,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 0.6866558039799981,
"learning_rate": 3.741935483870968e-05,
"loss": 0.3867,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 0.8159544677175238,
"learning_rate": 3.7741935483870974e-05,
"loss": 0.3721,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 1.0448831242502137,
"learning_rate": 3.806451612903226e-05,
"loss": 0.3789,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 1.1527203616466875,
"learning_rate": 3.8387096774193555e-05,
"loss": 0.3767,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 0.6159042149194461,
"learning_rate": 3.870967741935484e-05,
"loss": 0.3681,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 1.0301724101722216,
"learning_rate": 3.9032258064516135e-05,
"loss": 0.3715,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 1.2369000317913417,
"learning_rate": 3.935483870967742e-05,
"loss": 0.3722,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 0.6524241000691695,
"learning_rate": 3.9677419354838716e-05,
"loss": 0.3694,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 1.0997535325711614,
"learning_rate": 4e-05,
"loss": 0.3747,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 0.9131194496936479,
"learning_rate": 3.9999920040266474e-05,
"loss": 0.3847,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 0.9156629443937606,
"learning_rate": 3.999968016170525e-05,
"loss": 0.3647,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 0.8989473793465564,
"learning_rate": 3.9999280366234394e-05,
"loss": 0.3695,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 0.854678419105073,
"learning_rate": 3.999872065705065e-05,
"loss": 0.3717,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 0.7875829950433242,
"learning_rate": 3.999800103862945e-05,
"loss": 0.3701,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 0.511024678734643,
"learning_rate": 3.9997121516724834e-05,
"loss": 0.3645,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 0.7606853050163473,
"learning_rate": 3.999608209836944e-05,
"loss": 0.3639,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 0.7264287015311619,
"learning_rate": 3.999488279187443e-05,
"loss": 0.3942,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 0.616997440038911,
"learning_rate": 3.9993523606829425e-05,
"loss": 0.3828,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 0.6134968745193577,
"learning_rate": 3.9992004554102435e-05,
"loss": 0.3551,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 0.600309591156177,
"learning_rate": 3.999032564583976e-05,
"loss": 0.387,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 0.5831655780248199,
"learning_rate": 3.998848689546591e-05,
"loss": 0.3775,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 0.5258547326281815,
"learning_rate": 3.9986488317683496e-05,
"loss": 0.387,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 0.6444261344892638,
"learning_rate": 3.9984329928473067e-05,
"loss": 0.3817,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 0.5559726885266529,
"learning_rate": 3.998201174509306e-05,
"loss": 0.3757,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 0.5941422934691541,
"learning_rate": 3.997953378607961e-05,
"loss": 0.3656,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 0.5850134304305487,
"learning_rate": 3.9976896071246403e-05,
"loss": 0.383,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 0.5179239800565875,
"learning_rate": 3.997409862168454e-05,
"loss": 0.3835,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 0.4688242769853943,
"learning_rate": 3.997114145976237e-05,
"loss": 0.361,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 0.47733313166288716,
"learning_rate": 3.9968024609125245e-05,
"loss": 0.3682,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 0.42369664238068755,
"learning_rate": 3.996474809469545e-05,
"loss": 0.3648,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 0.4785114821202125,
"learning_rate": 3.996131194267189e-05,
"loss": 0.3637,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 0.40213178634825053,
"learning_rate": 3.995771618052995e-05,
"loss": 0.3403,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 0.5048599747763787,
"learning_rate": 3.9953960837021256e-05,
"loss": 0.3656,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 0.4863659457879787,
"learning_rate": 3.995004594217343e-05,
"loss": 0.3726,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 0.46828657706493093,
"learning_rate": 3.994597152728987e-05,
"loss": 0.3723,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 0.48833857039214695,
"learning_rate": 3.994173762494947e-05,
"loss": 0.3632,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 0.43888554133553087,
"learning_rate": 3.9937344269006424e-05,
"loss": 0.3643,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 0.4510871807340462,
"learning_rate": 3.993279149458988e-05,
"loss": 0.3678,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 0.4644398903866322,
"learning_rate": 3.9928079338103695e-05,
"loss": 0.3744,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 0.6816465943360329,
"learning_rate": 3.992320783722616e-05,
"loss": 0.3674,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.4698077744145719,
"learning_rate": 3.991817703090965e-05,
"loss": 0.3588,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 0.5746807727847477,
"learning_rate": 3.991298695938038e-05,
"loss": 0.3868,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 0.6262294561918806,
"learning_rate": 3.9907637664138e-05,
"loss": 0.3549,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 0.5355298141961358,
"learning_rate": 3.990212918795534e-05,
"loss": 0.3551,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 0.5669412872360924,
"learning_rate": 3.9896461574878036e-05,
"loss": 0.3654,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 0.5220804771232758,
"learning_rate": 3.989063487022416e-05,
"loss": 0.3588,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 0.5520595986949121,
"learning_rate": 3.9884649120583903e-05,
"loss": 0.3777,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 0.5001008584893774,
"learning_rate": 3.987850437381915e-05,
"loss": 0.3489,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 0.594715504325838,
"learning_rate": 3.987220067906313e-05,
"loss": 0.368,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 0.43784239183318224,
"learning_rate": 3.986573808672002e-05,
"loss": 0.364,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 0.4986575313049066,
"learning_rate": 3.9859116648464534e-05,
"loss": 0.3544,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 0.48786641201043845,
"learning_rate": 3.985233641724152e-05,
"loss": 0.3535,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 0.39282128895282487,
"learning_rate": 3.984539744726554e-05,
"loss": 0.3602,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 0.6211091344803816,
"learning_rate": 3.983829979402038e-05,
"loss": 0.3655,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 0.46645326592591846,
"learning_rate": 3.983104351425871e-05,
"loss": 0.3586,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 0.5983034576071964,
"learning_rate": 3.982362866600154e-05,
"loss": 0.3782,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 0.5108937292910668,
"learning_rate": 3.9816055308537806e-05,
"loss": 0.344,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 0.5197307632061816,
"learning_rate": 3.980832350242388e-05,
"loss": 0.3553,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 0.518034086819224,
"learning_rate": 3.980043330948306e-05,
"loss": 0.346,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 0.6159021924059812,
"learning_rate": 3.9792384792805124e-05,
"loss": 0.3462,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 0.41266972944341523,
"learning_rate": 3.978417801674579e-05,
"loss": 0.3482,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 0.5309718621723455,
"learning_rate": 3.9775813046926244e-05,
"loss": 0.3514,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 0.487602821319539,
"learning_rate": 3.976728995023254e-05,
"loss": 0.3624,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 0.5442413724641324,
"learning_rate": 3.975860879481514e-05,
"loss": 0.3705,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 0.5103616464292019,
"learning_rate": 3.974976965008833e-05,
"loss": 0.3604,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 0.4518570193725807,
"learning_rate": 3.9740772586729674e-05,
"loss": 0.3464,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.5358679034132297,
"learning_rate": 3.973161767667946e-05,
"loss": 0.3737,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 0.42244962565851824,
"learning_rate": 3.972230499314009e-05,
"loss": 0.3364,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 0.5543845574131623,
"learning_rate": 3.9712834610575545e-05,
"loss": 0.3657,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 0.5637322503627609,
"learning_rate": 3.9703206604710746e-05,
"loss": 0.3505,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 0.4992135900927081,
"learning_rate": 3.969342105253098e-05,
"loss": 0.351,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 0.41320800226457055,
"learning_rate": 3.9683478032281246e-05,
"loss": 0.3595,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 0.4516877982270651,
"learning_rate": 3.967337762346568e-05,
"loss": 0.3518,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 0.4326574061642712,
"learning_rate": 3.9663119906846885e-05,
"loss": 0.3711,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.4238287434293606,
"learning_rate": 3.965270496444528e-05,
"loss": 0.3514,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 0.47443014919338206,
"learning_rate": 3.9642132879538465e-05,
"loss": 0.3446,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 0.4616793121991076,
"learning_rate": 3.9631403736660566e-05,
"loss": 0.3382,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 0.46842976514395507,
"learning_rate": 3.962051762160151e-05,
"loss": 0.3659,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 0.46359344621052884,
"learning_rate": 3.960947462140638e-05,
"loss": 0.3439,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 0.46231320227964867,
"learning_rate": 3.959827482437472e-05,
"loss": 0.3438,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 0.6112562408857575,
"learning_rate": 3.958691832005981e-05,
"loss": 0.3708,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 0.4536649782978675,
"learning_rate": 3.9575405199267944e-05,
"loss": 0.3657,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 0.5388664031962368,
"learning_rate": 3.956373555405774e-05,
"loss": 0.3617,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 0.5165645513690353,
"learning_rate": 3.955190947773937e-05,
"loss": 0.3601,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 0.4572220485274068,
"learning_rate": 3.9539927064873824e-05,
"loss": 0.3364,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 0.4943433629823981,
"learning_rate": 3.952778841127214e-05,
"loss": 0.3413,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 0.4954408611236279,
"learning_rate": 3.95154936139947e-05,
"loss": 0.3439,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 0.49151237971035405,
"learning_rate": 3.950304277135034e-05,
"loss": 0.3433,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 0.6064696214083403,
"learning_rate": 3.94904359828957e-05,
"loss": 0.3696,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 0.4760414575549844,
"learning_rate": 3.94776733494343e-05,
"loss": 0.3452,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 0.4892639364264894,
"learning_rate": 3.9464754973015826e-05,
"loss": 0.3517,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 0.5044957358211062,
"learning_rate": 3.9451680956935275e-05,
"loss": 0.3602,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.5425240316536258,
"learning_rate": 3.943845140573213e-05,
"loss": 0.3493,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.41452788337360763,
"learning_rate": 3.942506642518952e-05,
"loss": 0.3665,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 0.5495586919362497,
"learning_rate": 3.9411526122333396e-05,
"loss": 0.3441,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 0.426295597122564,
"learning_rate": 3.939783060543167e-05,
"loss": 0.3619,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 0.4458770026428559,
"learning_rate": 3.938397998399333e-05,
"loss": 0.3485,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.46914545313742584,
"learning_rate": 3.936997436876756e-05,
"loss": 0.3581,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.4716473702224292,
"learning_rate": 3.935581387174289e-05,
"loss": 0.3433,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 0.465398827050473,
"learning_rate": 3.9341498606146295e-05,
"loss": 0.3546,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.49490191007514334,
"learning_rate": 3.932702868644224e-05,
"loss": 0.363,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.5009516462747736,
"learning_rate": 3.931240422833181e-05,
"loss": 0.3615,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.4365892204461152,
"learning_rate": 3.9297625348751806e-05,
"loss": 0.342,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 0.5477277704005414,
"learning_rate": 3.9282692165873744e-05,
"loss": 0.3356,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 0.46122154091683776,
"learning_rate": 3.9267604799102945e-05,
"loss": 0.3314,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.5044628628258826,
"learning_rate": 3.9252363369077615e-05,
"loss": 0.357,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 0.6524630107671445,
"learning_rate": 3.9236967997667796e-05,
"loss": 0.3591,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 0.4543360464875463,
"learning_rate": 3.9221418807974496e-05,
"loss": 0.3612,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 0.5923567683411362,
"learning_rate": 3.92057159243286e-05,
"loss": 0.3485,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 0.43625245620174646,
"learning_rate": 3.9189859472289956e-05,
"loss": 0.3335,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 0.4399635364218596,
"learning_rate": 3.9173849578646324e-05,
"loss": 0.3567,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 0.4977936457787769,
"learning_rate": 3.9157686371412395e-05,
"loss": 0.3458,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.4302299054465774,
"learning_rate": 3.914136997982874e-05,
"loss": 0.3383,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 0.49021842164385554,
"learning_rate": 3.912490053436079e-05,
"loss": 0.3667,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 0.48327740261635704,
"learning_rate": 3.9108278166697805e-05,
"loss": 0.3351,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.5240681372472347,
"learning_rate": 3.909150300975178e-05,
"loss": 0.3341,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 0.4361889983387936,
"learning_rate": 3.907457519765642e-05,
"loss": 0.3362,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.4536707870474062,
"learning_rate": 3.905749486576607e-05,
"loss": 0.3415,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.47390631316232223,
"learning_rate": 3.90402621506546e-05,
"loss": 0.3479,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 0.4323544206509106,
"learning_rate": 3.902287719011435e-05,
"loss": 0.35,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.5046885837670478,
"learning_rate": 3.900534012315499e-05,
"loss": 0.3508,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 0.4252196013730918,
"learning_rate": 3.8987651090002454e-05,
"loss": 0.3544,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 0.44805281729805224,
"learning_rate": 3.896981023209777e-05,
"loss": 0.3632,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 0.35837213863829076,
"learning_rate": 3.8951817692095966e-05,
"loss": 0.3361,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 0.4481776951580174,
"learning_rate": 3.893367361386491e-05,
"loss": 0.3545,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 0.400170805134326,
"learning_rate": 3.891537814248417e-05,
"loss": 0.3479,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 0.4560839650326389,
"learning_rate": 3.889693142424384e-05,
"loss": 0.3604,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 0.42555472569480196,
"learning_rate": 3.8878333606643405e-05,
"loss": 0.3632,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 0.49216508724642805,
"learning_rate": 3.885958483839049e-05,
"loss": 0.3487,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 0.4101504859146444,
"learning_rate": 3.8840685269399786e-05,
"loss": 0.3502,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 0.4386924043756538,
"learning_rate": 3.882163505079171e-05,
"loss": 0.3524,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 0.4135252417847887,
"learning_rate": 3.880243433489132e-05,
"loss": 0.3419,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 0.43255912409413405,
"learning_rate": 3.878308327522702e-05,
"loss": 0.2946,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 0.38386101106202564,
"learning_rate": 3.876358202652936e-05,
"loss": 0.2884,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 0.45135365128566346,
"learning_rate": 3.8743930744729835e-05,
"loss": 0.2941,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.4644476646488923,
"learning_rate": 3.8724129586959535e-05,
"loss": 0.302,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.4232585392537929,
"learning_rate": 3.870417871154801e-05,
"loss": 0.3013,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 0.4737367618552959,
"learning_rate": 3.868407827802192e-05,
"loss": 0.2973,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.4441969994054439,
"learning_rate": 3.8663828447103804e-05,
"loss": 0.291,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.44371882330972173,
"learning_rate": 3.8643429380710765e-05,
"loss": 0.2966,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.514475111620334,
"learning_rate": 3.862288124195319e-05,
"loss": 0.2938,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.4302671735593257,
"learning_rate": 3.8602184195133464e-05,
"loss": 0.3001,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 0.44823465202122353,
"learning_rate": 3.8581338405744604e-05,
"loss": 0.3091,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 0.412866009661847,
"learning_rate": 3.8560344040468996e-05,
"loss": 0.2934,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.4365410839322923,
"learning_rate": 3.853920126717703e-05,
"loss": 0.3019,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 0.39718480875683515,
"learning_rate": 3.851791025492574e-05,
"loss": 0.3031,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.4482689345274002,
"learning_rate": 3.8496471173957515e-05,
"loss": 0.2844,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.45915165968951616,
"learning_rate": 3.847488419569866e-05,
"loss": 0.3132,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.46798016890166605,
"learning_rate": 3.8453149492758077e-05,
"loss": 0.2773,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 0.3905572205934363,
"learning_rate": 3.8431267238925886e-05,
"loss": 0.2784,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.45719330850806317,
"learning_rate": 3.840923760917199e-05,
"loss": 0.2981,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 0.407602775938911,
"learning_rate": 3.838706077964473e-05,
"loss": 0.3038,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 0.44118144655152647,
"learning_rate": 3.836473692766944e-05,
"loss": 0.289,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 0.3877962467338273,
"learning_rate": 3.834226623174705e-05,
"loss": 0.2957,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.4622184891070544,
"learning_rate": 3.831964887155264e-05,
"loss": 0.2968,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.408783418579291,
"learning_rate": 3.829688502793403e-05,
"loss": 0.2827,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.3571372360136095,
"learning_rate": 3.827397488291029e-05,
"loss": 0.2897,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.3957124968523676,
"learning_rate": 3.8250918619670343e-05,
"loss": 0.303,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.42226034903857,
"learning_rate": 3.822771642257145e-05,
"loss": 0.2921,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.41813574610613896,
"learning_rate": 3.820436847713776e-05,
"loss": 0.296,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 0.4566537519958979,
"learning_rate": 3.8180874970058826e-05,
"loss": 0.29,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.4155229445397747,
"learning_rate": 3.8157236089188103e-05,
"loss": 0.2898,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.3605001888290419,
"learning_rate": 3.813345202354145e-05,
"loss": 0.2936,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 0.39259652102710485,
"learning_rate": 3.810952296329563e-05,
"loss": 0.3091,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.3451958525658247,
"learning_rate": 3.808544909978676e-05,
"loss": 0.2857,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.4157673647962679,
"learning_rate": 3.806123062550882e-05,
"loss": 0.3071,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.3765693192057601,
"learning_rate": 3.803686773411208e-05,
"loss": 0.2975,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.4595731553784716,
"learning_rate": 3.801236062040158e-05,
"loss": 0.2902,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 0.4259964560051856,
"learning_rate": 3.798770948033553e-05,
"loss": 0.2875,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.42780638721641195,
"learning_rate": 3.796291451102381e-05,
"loss": 0.275,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.4884778476630945,
"learning_rate": 3.793797591072631e-05,
"loss": 0.29,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.42101711637088907,
"learning_rate": 3.791289387885144e-05,
"loss": 0.3038,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.5056215682881441,
"learning_rate": 3.788766861595443e-05,
"loss": 0.2864,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 0.44784911133563376,
"learning_rate": 3.7862300323735835e-05,
"loss": 0.2984,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 0.5179228150472691,
"learning_rate": 3.783678920503983e-05,
"loss": 0.3117,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 0.38547914489465845,
"learning_rate": 3.781113546385265e-05,
"loss": 0.2986,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.43239907137919803,
"learning_rate": 3.7785339305300916e-05,
"loss": 0.2988,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.3712429814641452,
"learning_rate": 3.7759400935650025e-05,
"loss": 0.2868,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.4679869232042905,
"learning_rate": 3.77333205623025e-05,
"loss": 0.3043,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.4337645114257828,
"learning_rate": 3.7707098393796295e-05,
"loss": 0.2967,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.4210296286024566,
"learning_rate": 3.7680734639803185e-05,
"loss": 0.2995,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 0.4435404623323763,
"learning_rate": 3.765422951112704e-05,
"loss": 0.2934,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 0.4051188804513351,
"learning_rate": 3.762758321970216e-05,
"loss": 0.2883,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.4169113582957879,
"learning_rate": 3.7600795978591584e-05,
"loss": 0.305,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.3626999982624648,
"learning_rate": 3.757386800198538e-05,
"loss": 0.2872,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.3968563817425368,
"learning_rate": 3.7546799505198925e-05,
"loss": 0.2892,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.38823446417042623,
"learning_rate": 3.7519590704671197e-05,
"loss": 0.3083,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.38799008956516484,
"learning_rate": 3.749224181796305e-05,
"loss": 0.3029,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.39834386144138045,
"learning_rate": 3.746475306375545e-05,
"loss": 0.2901,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.35261948732540266,
"learning_rate": 3.743712466184774e-05,
"loss": 0.2985,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.4365713231315396,
"learning_rate": 3.7409356833155885e-05,
"loss": 0.3001,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.4592999811520726,
"learning_rate": 3.73814497997107e-05,
"loss": 0.2988,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.3937916978991155,
"learning_rate": 3.7353403784656095e-05,
"loss": 0.2882,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.44932918410132805,
"learning_rate": 3.732521901224724e-05,
"loss": 0.2919,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.39822329238733567,
"learning_rate": 3.7296895707848844e-05,
"loss": 0.284,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.36006291261618395,
"learning_rate": 3.7268434097933275e-05,
"loss": 0.2886,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.4141406298369155,
"learning_rate": 3.723983441007882e-05,
"loss": 0.2931,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.44540723730102877,
"learning_rate": 3.721109687296781e-05,
"loss": 0.3034,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.42120188882331705,
"learning_rate": 3.718222171638484e-05,
"loss": 0.2905,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 0.4281318589223922,
"learning_rate": 3.715320917121488e-05,
"loss": 0.2825,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.4503656755509057,
"learning_rate": 3.712405946944149e-05,
"loss": 0.3061,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.38114678729583606,
"learning_rate": 3.709477284414488e-05,
"loss": 0.3057,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.42800703175559457,
"learning_rate": 3.7065349529500136e-05,
"loss": 0.294,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 0.37692566303492975,
"learning_rate": 3.7035789760775305e-05,
"loss": 0.3069,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.4142271401592637,
"learning_rate": 3.70060937743295e-05,
"loss": 0.2821,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.3738437647467631,
"learning_rate": 3.6976261807611045e-05,
"loss": 0.3006,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.42243146596318853,
"learning_rate": 3.694629409915555e-05,
"loss": 0.2806,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.45425686723809916,
"learning_rate": 3.691619088858401e-05,
"loss": 0.2981,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.4062028197344348,
"learning_rate": 3.68859524166009e-05,
"loss": 0.2969,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.3894381321971832,
"learning_rate": 3.6855578924992226e-05,
"loss": 0.2923,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.44823814452407795,
"learning_rate": 3.6825070656623626e-05,
"loss": 0.2931,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.470430895867103,
"learning_rate": 3.67944278554384e-05,
"loss": 0.2942,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 0.438204056871131,
"learning_rate": 3.676365076645557e-05,
"loss": 0.2886,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.47983807682810803,
"learning_rate": 3.673273963576791e-05,
"loss": 0.2917,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.3849806329400015,
"learning_rate": 3.670169471054001e-05,
"loss": 0.2894,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.42050444700135786,
"learning_rate": 3.6670516239006254e-05,
"loss": 0.3045,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.3905493792787207,
"learning_rate": 3.663920447046888e-05,
"loss": 0.2984,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.4121993287801604,
"learning_rate": 3.6607759655295954e-05,
"loss": 0.2928,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.4459818907704544,
"learning_rate": 3.657618204491937e-05,
"loss": 0.2952,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.39717305009753207,
"learning_rate": 3.6544471891832875e-05,
"loss": 0.2959,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.4799115778940018,
"learning_rate": 3.6512629449589983e-05,
"loss": 0.2969,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.3958758946524484,
"learning_rate": 3.6480654972802045e-05,
"loss": 0.2881,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.4600438630892322,
"learning_rate": 3.644854871713611e-05,
"loss": 0.2903,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.4155116777327393,
"learning_rate": 3.6416310939312924e-05,
"loss": 0.3029,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.404935940015595,
"learning_rate": 3.638394189710493e-05,
"loss": 0.2933,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.4510851850905217,
"learning_rate": 3.635144184933412e-05,
"loss": 0.2918,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.3689634425773272,
"learning_rate": 3.6318811055869994e-05,
"loss": 0.3104,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.41306360224280453,
"learning_rate": 3.628604977762752e-05,
"loss": 0.3032,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.37172589814952844,
"learning_rate": 3.6253158276565004e-05,
"loss": 0.2888,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.365012267895697,
"learning_rate": 3.6220136815682016e-05,
"loss": 0.297,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.34606768739122795,
"learning_rate": 3.618698565901727e-05,
"loss": 0.2998,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.3626189944246459,
"learning_rate": 3.615370507164654e-05,
"loss": 0.3051,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.33918544473260676,
"learning_rate": 3.61202953196805e-05,
"loss": 0.2951,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 0.34104818561126127,
"learning_rate": 3.6086756670262654e-05,
"loss": 0.3046,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.3507993202514925,
"learning_rate": 3.6053089391567135e-05,
"loss": 0.2985,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.39280197245842585,
"learning_rate": 3.601929375279662e-05,
"loss": 0.3121,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.3679274550621327,
"learning_rate": 3.598537002418012e-05,
"loss": 0.2956,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.3652256228164707,
"learning_rate": 3.5951318476970885e-05,
"loss": 0.2943,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.4188958454452056,
"learning_rate": 3.591713938344416e-05,
"loss": 0.2931,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.36166470387834354,
"learning_rate": 3.588283301689507e-05,
"loss": 0.2906,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 0.395631888409455,
"learning_rate": 3.5848399651636424e-05,
"loss": 0.2988,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.37947683290085654,
"learning_rate": 3.5813839562996473e-05,
"loss": 0.2773,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.3731474082514876,
"learning_rate": 3.5779153027316775e-05,
"loss": 0.297,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.3898434271484996,
"learning_rate": 3.574434032194994e-05,
"loss": 0.3057,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.3927065683005924,
"learning_rate": 3.5709401725257436e-05,
"loss": 0.3,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.4244584127780789,
"learning_rate": 3.5674337516607346e-05,
"loss": 0.2816,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.4137441129587813,
"learning_rate": 3.563914797637216e-05,
"loss": 0.3084,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.4069718527331933,
"learning_rate": 3.560383338592649e-05,
"loss": 0.3084,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.36261587208712315,
"learning_rate": 3.556839402764487e-05,
"loss": 0.2867,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.4154119306760094,
"learning_rate": 3.553283018489946e-05,
"loss": 0.3016,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.4102246098811583,
"learning_rate": 3.54971421420578e-05,
"loss": 0.2929,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.4031732820124687,
"learning_rate": 3.5461330184480524e-05,
"loss": 0.2839,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 0.3521395654163643,
"learning_rate": 3.54253945985191e-05,
"loss": 0.2962,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.36465064084128923,
"learning_rate": 3.5389335671513524e-05,
"loss": 0.3042,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.3695343653054988,
"learning_rate": 3.5353153691789986e-05,
"loss": 0.3046,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.4279151357140074,
"learning_rate": 3.531684894865866e-05,
"loss": 0.2963,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.39212305850656887,
"learning_rate": 3.528042173241128e-05,
"loss": 0.3095,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.36564713638117485,
"learning_rate": 3.524387233431892e-05,
"loss": 0.2829,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.38320038320965666,
"learning_rate": 3.520720104662958e-05,
"loss": 0.3018,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.3525069943055677,
"learning_rate": 3.5170408162565904e-05,
"loss": 0.2998,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.4013522833351578,
"learning_rate": 3.513349397632281e-05,
"loss": 0.2806,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.40635498673607173,
"learning_rate": 3.5096458783065145e-05,
"loss": 0.3188,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.38206741290739565,
"learning_rate": 3.505930287892533e-05,
"loss": 0.2897,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.3891078266322259,
"learning_rate": 3.502202656100099e-05,
"loss": 0.2726,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.36417296718578124,
"learning_rate": 3.498463012735256e-05,
"loss": 0.2854,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.38547538066174797,
"learning_rate": 3.494711387700094e-05,
"loss": 0.2876,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.36687321527650774,
"learning_rate": 3.490947810992505e-05,
"loss": 0.2968,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 0.3615778100563679,
"learning_rate": 3.48717231270595e-05,
"loss": 0.2969,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.36953136459928065,
"learning_rate": 3.483384923029211e-05,
"loss": 0.3251,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.3508004374950771,
"learning_rate": 3.479585672246156e-05,
"loss": 0.2907,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.36251939006604317,
"learning_rate": 3.4757745907354924e-05,
"loss": 0.2937,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.3244845795056114,
"learning_rate": 3.471951708970527e-05,
"loss": 0.2953,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.3629084985168257,
"learning_rate": 3.468117057518921e-05,
"loss": 0.2955,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.379260000704293,
"learning_rate": 3.4642706670424426e-05,
"loss": 0.3047,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.3226584697885049,
"learning_rate": 3.460412568296731e-05,
"loss": 0.3111,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.38597354581693216,
"learning_rate": 3.456542792131039e-05,
"loss": 0.3073,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.3075883176471981,
"learning_rate": 3.4526613694879936e-05,
"loss": 0.2951,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.3526844289210075,
"learning_rate": 3.448768331403348e-05,
"loss": 0.2801,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.3256536904615551,
"learning_rate": 3.44486370900573e-05,
"loss": 0.2856,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.3470963577994215,
"learning_rate": 3.440947533516396e-05,
"loss": 0.2973,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.3756657368646931,
"learning_rate": 3.4370198362489816e-05,
"loss": 0.2987,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 0.35186962163387947,
"learning_rate": 3.43308064860925e-05,
"loss": 0.2833,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.37937953434542576,
"learning_rate": 3.4291300020948393e-05,
"loss": 0.3052,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.3569464855372478,
"learning_rate": 3.425167928295015e-05,
"loss": 0.2971,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.37638660322816686,
"learning_rate": 3.421194458890411e-05,
"loss": 0.2845,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.361616296345157,
"learning_rate": 3.4172096256527865e-05,
"loss": 0.2995,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.3881825888498009,
"learning_rate": 3.4132134604447595e-05,
"loss": 0.2855,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.38379409012932847,
"learning_rate": 3.409205995219561e-05,
"loss": 0.2997,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.3874940951520242,
"learning_rate": 3.4051872620207765e-05,
"loss": 0.2902,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.3846769067500781,
"learning_rate": 3.40115729298209e-05,
"loss": 0.2923,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.36536261131027215,
"learning_rate": 3.397116120327025e-05,
"loss": 0.3012,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.40248348401115996,
"learning_rate": 3.3930637763686916e-05,
"loss": 0.296,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.34370714789162954,
"learning_rate": 3.389000293509524e-05,
"loss": 0.2832,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.43174189294125837,
"learning_rate": 3.3849257042410225e-05,
"loss": 0.3016,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.34105555020774103,
"learning_rate": 3.380840041143494e-05,
"loss": 0.2836,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.3445220433987209,
"learning_rate": 3.376743336885792e-05,
"loss": 0.3069,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.32329203602313966,
"learning_rate": 3.372635624225055e-05,
"loss": 0.2773,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.34867261193927873,
"learning_rate": 3.3685169360064445e-05,
"loss": 0.305,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.3474455770549151,
"learning_rate": 3.3643873051628804e-05,
"loss": 0.2887,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.3117490587585434,
"learning_rate": 3.3602467647147814e-05,
"loss": 0.2898,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.36780708309803783,
"learning_rate": 3.356095347769799e-05,
"loss": 0.2964,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.3205837560164388,
"learning_rate": 3.351933087522552e-05,
"loss": 0.283,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.36966467267696423,
"learning_rate": 3.3477600172543626e-05,
"loss": 0.2946,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.34846683036184967,
"learning_rate": 3.3435761703329894e-05,
"loss": 0.3014,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.3486792925339909,
"learning_rate": 3.3393815802123606e-05,
"loss": 0.2925,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.391962270856151,
"learning_rate": 3.335176280432308e-05,
"loss": 0.3097,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.3424418649022183,
"learning_rate": 3.330960304618295e-05,
"loss": 0.2855,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.3711239614186727,
"learning_rate": 3.326733686481153e-05,
"loss": 0.3027,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.35644633431413764,
"learning_rate": 3.3224964598168084e-05,
"loss": 0.281,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.3874171356324364,
"learning_rate": 3.318248658506012e-05,
"loss": 0.3038,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.3754877608870576,
"learning_rate": 3.313990316514069e-05,
"loss": 0.3017,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.3500151132800928,
"learning_rate": 3.309721467890571e-05,
"loss": 0.2839,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.42072075906351464,
"learning_rate": 3.305442146769115e-05,
"loss": 0.2922,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.38004790767712243,
"learning_rate": 3.30115238736704e-05,
"loss": 0.2986,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.3964953910724282,
"learning_rate": 3.296852223985148e-05,
"loss": 0.2987,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 0.3929098895244896,
"learning_rate": 3.2925416910074315e-05,
"loss": 0.2935,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.37547697139105823,
"learning_rate": 3.288220822900796e-05,
"loss": 0.2852,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.3945055168933806,
"learning_rate": 3.283889654214788e-05,
"loss": 0.2835,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.3789239771554095,
"learning_rate": 3.279548219581317e-05,
"loss": 0.3012,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.35120314617553533,
"learning_rate": 3.275196553714379e-05,
"loss": 0.2984,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.36749471897258823,
"learning_rate": 3.270834691409778e-05,
"loss": 0.289,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.39921956956691923,
"learning_rate": 3.266462667544849e-05,
"loss": 0.2876,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.34247581697414803,
"learning_rate": 3.262080517078178e-05,
"loss": 0.2997,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.41051612937015813,
"learning_rate": 3.257688275049323e-05,
"loss": 0.2765,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.315374740903345,
"learning_rate": 3.253285976578535e-05,
"loss": 0.2933,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.38805667135644134,
"learning_rate": 3.2488736568664756e-05,
"loss": 0.2905,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.29936458928480114,
"learning_rate": 3.244451351193935e-05,
"loss": 0.2913,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 0.3947631888064866,
"learning_rate": 3.240019094921551e-05,
"loss": 0.2997,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.32309942016059673,
"learning_rate": 3.235576923489527e-05,
"loss": 0.2907,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.347996701213138,
"learning_rate": 3.2311248724173476e-05,
"loss": 0.2857,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.34224662402520684,
"learning_rate": 3.226662977303494e-05,
"loss": 0.2921,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.3610825307723578,
"learning_rate": 3.2221912738251626e-05,
"loss": 0.3115,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.3762830280379083,
"learning_rate": 3.2177097977379736e-05,
"loss": 0.3061,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.33732993057218397,
"learning_rate": 3.21321858487569e-05,
"loss": 0.2918,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.3597278983784664,
"learning_rate": 3.208717671149931e-05,
"loss": 0.2946,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.3373124456624458,
"learning_rate": 3.204207092549882e-05,
"loss": 0.3002,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 0.3535866739425677,
"learning_rate": 3.19968688514201e-05,
"loss": 0.2843,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.32546311968620967,
"learning_rate": 3.195157085069774e-05,
"loss": 0.3018,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.3129029799335298,
"learning_rate": 3.190617728553332e-05,
"loss": 0.289,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.3616429822805402,
"learning_rate": 3.1860688518892606e-05,
"loss": 0.2978,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.3513245394939147,
"learning_rate": 3.181510491450254e-05,
"loss": 0.2961,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.3142502773385173,
"learning_rate": 3.176942683684842e-05,
"loss": 0.2998,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.32690989092478556,
"learning_rate": 3.1723654651170934e-05,
"loss": 0.2928,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.291417688428564,
"learning_rate": 3.167778872346327e-05,
"loss": 0.2841,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.30841894529185265,
"learning_rate": 3.163182942046815e-05,
"loss": 0.282,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.30771157387257475,
"learning_rate": 3.158577710967494e-05,
"loss": 0.2873,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.3359161777110256,
"learning_rate": 3.153963215931669e-05,
"loss": 0.2874,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.3054385833594419,
"learning_rate": 3.14933949383672e-05,
"loss": 0.2957,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.35677445859172496,
"learning_rate": 3.144706581653805e-05,
"loss": 0.2951,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.3244941331304331,
"learning_rate": 3.140064516427566e-05,
"loss": 0.3008,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.3919012806882817,
"learning_rate": 3.135413335275833e-05,
"loss": 0.3076,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.3228393206921307,
"learning_rate": 3.130753075389327e-05,
"loss": 0.2857,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.29062525489554625,
"learning_rate": 3.1260837740313625e-05,
"loss": 0.2928,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.3693475632146356,
"learning_rate": 3.121405468537547e-05,
"loss": 0.3019,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.31650009868568335,
"learning_rate": 3.116718196315488e-05,
"loss": 0.2893,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.31240654632867015,
"learning_rate": 3.1120219948444894e-05,
"loss": 0.2944,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.35426202586361943,
"learning_rate": 3.107316901675252e-05,
"loss": 0.2958,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.3303012140144008,
"learning_rate": 3.1026029544295745e-05,
"loss": 0.2964,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.3325072596476312,
"learning_rate": 3.097880190800057e-05,
"loss": 0.3082,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.36245036159060456,
"learning_rate": 3.093148648549788e-05,
"loss": 0.2972,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.3706691065693111,
"learning_rate": 3.088408365512055e-05,
"loss": 0.3016,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.31093457174647243,
"learning_rate": 3.083659379590034e-05,
"loss": 0.2941,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.3510724882864077,
"learning_rate": 3.0789017287564913e-05,
"loss": 0.3043,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.3318534996283806,
"learning_rate": 3.074135451053475e-05,
"loss": 0.2989,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.3445172942596027,
"learning_rate": 3.069360584592015e-05,
"loss": 0.3036,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.37350625246948393,
"learning_rate": 3.064577167551816e-05,
"loss": 0.2875,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.33716961234119613,
"learning_rate": 3.059785238180953e-05,
"loss": 0.2902,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.38118201662060974,
"learning_rate": 3.054984834795566e-05,
"loss": 0.2883,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.3700429308079092,
"learning_rate": 3.0501759957795537e-05,
"loss": 0.2896,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.36101990123992245,
"learning_rate": 3.0453587595842627e-05,
"loss": 0.2862,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.4193001503444059,
"learning_rate": 3.0405331647281856e-05,
"loss": 0.3107,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.33928188717104546,
"learning_rate": 3.0356992497966508e-05,
"loss": 0.2794,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.3548502929466175,
"learning_rate": 3.0308570534415137e-05,
"loss": 0.2824,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.3727595532807279,
"learning_rate": 3.0260066143808467e-05,
"loss": 0.2823,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.3394021631379387,
"learning_rate": 3.021147971398632e-05,
"loss": 0.294,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.3903063121376847,
"learning_rate": 3.0162811633444478e-05,
"loss": 0.2859,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.35111309561729265,
"learning_rate": 3.0114062291331635e-05,
"loss": 0.2994,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.3478974839362705,
"learning_rate": 3.0065232077446217e-05,
"loss": 0.271,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.37583675638111574,
"learning_rate": 3.001632138223332e-05,
"loss": 0.2839,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.4309775008581306,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.227,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 0.36579028705528316,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.2062,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 0.9128804282264695,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.2377,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.45829418572318664,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.2383,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.3917906578057329,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.2074,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.38275095451526703,
"learning_rate": 2.972118904869376e-05,
"loss": 0.2282,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.3258187047161393,
"learning_rate": 2.9671725970312563e-05,
"loss": 0.2097,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.3859669595724809,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.2266,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.36081101184809017,
"learning_rate": 2.95725682050846e-05,
"loss": 0.2181,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.3589601201583789,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.2281,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.3786415729215701,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.2136,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.34525939812540085,
"learning_rate": 2.942325848714505e-05,
"loss": 0.208,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.3740619143370787,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.2075,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.3312481477418027,
"learning_rate": 2.932334127129686e-05,
"loss": 0.2234,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.35913549630681174,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.2178,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.3359973310728297,
"learning_rate": 2.922312585929131e-05,
"loss": 0.2135,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.35065520762573826,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.2324,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.35310765834749996,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.2252,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.3641137452553996,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.2269,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.3415819452427269,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.2106,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.3483714384207297,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.2115,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.3799214259119207,
"learning_rate": 2.8920722546128863e-05,
"loss": 0.2327,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.3750442285057045,
"learning_rate": 2.887006998351372e-05,
"loss": 0.2197,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.3154432940337058,
"learning_rate": 2.881934649605535e-05,
"loss": 0.2214,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.35487476622549396,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.2091,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.34076724229493616,
"learning_rate": 2.8717688369507425e-05,
"loss": 0.2128,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.34869555801020935,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.2215,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.3245209398153679,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.2265,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.31879151541249123,
"learning_rate": 2.856467940121025e-05,
"loss": 0.2133,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.3466902240960824,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.2192,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.3271410983983892,
"learning_rate": 2.846233032790149e-05,
"loss": 0.2131,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.3211828750229001,
"learning_rate": 2.84110540896642e-05,
"loss": 0.2151,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.32281645407320453,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.2106,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.3165193215530264,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.2191,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.33502207710661824,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.2213,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.32950826313776,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.221,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.32829848321177835,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.2202,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.2985919969907737,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.2138,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.32435395696509917,
"learning_rate": 2.805026032173773e-05,
"loss": 0.2195,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.3202160353929606,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.2161,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.340472316965653,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.2144,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.33600559245119005,
"learning_rate": 2.7894660620408e-05,
"loss": 0.2041,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.3285838964413386,
"learning_rate": 2.784266724883173e-05,
"loss": 0.2109,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.3142693321057148,
"learning_rate": 2.7790611167497128e-05,
"loss": 0.2308,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.32001913520077857,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.2345,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.3150863987050536,
"learning_rate": 2.768631254100716e-05,
"loss": 0.2339,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.2892626720040568,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.2205,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.30136487308193294,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.2167,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.3244463236598499,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.2083,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.36768875753426616,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.2221,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.29446454290701857,
"learning_rate": 2.742449775134085e-05,
"loss": 0.2257,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.3356711798683238,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.226,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.36324863602792695,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.2227,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.31596777304691864,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.203,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.33246397713539305,
"learning_rate": 2.7213974824975785e-05,
"loss": 0.2161,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.3261899271227943,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.2417,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.31742938708530094,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.2344,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.3029770773890946,
"learning_rate": 2.705547548999996e-05,
"loss": 0.2279,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.33227761874974615,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.2074,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.29510501263412,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.218,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.3150584222774071,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.2069,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.32590472687878913,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.2216,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.30584871667163144,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.2057,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.3412934785020945,
"learning_rate": 2.67369650552155e-05,
"loss": 0.2281,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.3179766070868756,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.2201,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.31139253652338367,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.22,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.3183812170954294,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.2184,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.36094149545991255,
"learning_rate": 2.652354173593614e-05,
"loss": 0.2228,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.28849801321184565,
"learning_rate": 2.647005443331317e-05,
"loss": 0.2184,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.3832118781396914,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.2248,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.31956535939346914,
"learning_rate": 2.636292505301542e-05,
"loss": 0.2275,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.326181253514031,
"learning_rate": 2.630928383194431e-05,
"loss": 0.2205,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.32874161519396417,
"learning_rate": 2.625559216200781e-05,
"loss": 0.2251,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.34586602529260635,
"learning_rate": 2.620185047252308e-05,
"loss": 0.2135,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.32940093156655464,
"learning_rate": 2.614805919320723e-05,
"loss": 0.2395,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.3444742374069792,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.2172,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.3000769350239719,
"learning_rate": 2.60403295859298e-05,
"loss": 0.2161,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.34535977316520194,
"learning_rate": 2.5986392119371297e-05,
"loss": 0.2269,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.31807527373353456,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.2225,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.3089355342757059,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.2347,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.34688726811869647,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.215,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.3177971706345458,
"learning_rate": 2.577016790136453e-05,
"loss": 0.2331,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.3098495883929342,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.2081,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.3084070294006283,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.2214,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.3378973628286093,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.2312,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.3305989092705123,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.2439,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.3431136649012861,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.2318,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.32928059090551665,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.2164,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.3149874494267551,
"learning_rate": 2.539001591449889e-05,
"loss": 0.2256,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.3425461958699717,
"learning_rate": 2.53355326151989e-05,
"loss": 0.2212,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.308115356951747,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.2354,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.35055807145423257,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.2251,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.30228252662278116,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.2196,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.3187146817024279,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.2102,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.3124544559335303,
"learning_rate": 2.506248490185627e-05,
"loss": 0.2059,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.35789272792267535,
"learning_rate": 2.500775217241923e-05,
"loss": 0.2215,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.33867844092110666,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.2176,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.33279462579600305,
"learning_rate": 2.4898167025948e-05,
"loss": 0.2096,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.3500082172831406,
"learning_rate": 2.4843315485153703e-05,
"loss": 0.2114,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.29102682696658705,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.2152,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.34060196437034956,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.2231,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.3290791019899497,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.223,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.32305954458190067,
"learning_rate": 2.46235264422994e-05,
"loss": 0.2329,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.350695317075137,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.2355,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.354137107733857,
"learning_rate": 2.451340834511592e-05,
"loss": 0.2368,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.3558023509108764,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.2153,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.34116897038227795,
"learning_rate": 2.440314589184958e-05,
"loss": 0.2186,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.3385405834568051,
"learning_rate": 2.434796163357541e-05,
"loss": 0.2515,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.3626474508561237,
"learning_rate": 2.429274260911588e-05,
"loss": 0.2178,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.3139560482755046,
"learning_rate": 2.423748926000084e-05,
"loss": 0.2243,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.3614322806540988,
"learning_rate": 2.418220202803459e-05,
"loss": 0.2139,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.3207013219182415,
"learning_rate": 2.412688135529237e-05,
"loss": 0.2459,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.38792909440303036,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.2294,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.3078833679549232,
"learning_rate": 2.4016141457114372e-05,
"loss": 0.2231,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.3575084833011102,
"learning_rate": 2.396072311715187e-05,
"loss": 0.2247,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.36483192628666156,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.2164,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.31888828538346503,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.2329,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.36750159242198815,
"learning_rate": 2.379427985200223e-05,
"loss": 0.222,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.33473626515225885,
"learning_rate": 2.373873749394972e-05,
"loss": 0.2258,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.3469148869626964,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.2131,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.3444350763528098,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.2103,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.34804658480515477,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.2227,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.305756901120702,
"learning_rate": 2.351627355800233e-05,
"loss": 0.2292,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.3280442404937715,
"learning_rate": 2.346058617160256e-05,
"loss": 0.228,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.31227351692779587,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.2285,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.3027205633355427,
"learning_rate": 2.334912883203469e-05,
"loss": 0.2206,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.28834263693722606,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.2212,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.3088479838540525,
"learning_rate": 2.323756437450136e-05,
"loss": 0.2267,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.2824572920558381,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.2172,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.3074978740296837,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.2272,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.27364239268241713,
"learning_rate": 2.307002464849052e-05,
"loss": 0.2201,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.3132563096528576,
"learning_rate": 2.30141283818846e-05,
"loss": 0.2117,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.30115377066777327,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.2205,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.2850617235827318,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.2247,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.3624409884171958,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.2314,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.2885471971602888,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.2059,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.3371057263308071,
"learning_rate": 2.273429448107257e-05,
"loss": 0.2178,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.32285746036451346,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.2217,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.32562222671055474,
"learning_rate": 2.262220474347102e-05,
"loss": 0.2298,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.3142399923904699,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.2103,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.2941438793239499,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.2237,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.32389511645921526,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.2137,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.2908177767221533,
"learning_rate": 2.2397777251561302e-05,
"loss": 0.2301,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.335110328059866,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.2278,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.31428178636986787,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.2181,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.31447656725742124,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.2184,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.35430596610101744,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.2225,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.31776788138969486,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.1966,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.32448379203562155,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.2193,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.31393640717752586,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.2045,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.2860381916928262,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.2426,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.3400283137889665,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.2181,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.2933954030643628,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.2245,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.3350770847423531,
"learning_rate": 2.17791064266051e-05,
"loss": 0.2124,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.30646416752357064,
"learning_rate": 2.172276926861741e-05,
"loss": 0.2375,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.32574869550394053,
"learning_rate": 2.166641833541255e-05,
"loss": 0.2116,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.3371556499394173,
"learning_rate": 2.161005407757109e-05,
"loss": 0.2399,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.30887606424334424,
"learning_rate": 2.155367694578013e-05,
"loss": 0.222,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.3538420454967925,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.2199,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.2929786987631004,
"learning_rate": 2.1440885863609208e-05,
"loss": 0.2233,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.33112150720980277,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.2196,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.3001313800700288,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.2243,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.2955369398855487,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.2355,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.30747354089895723,
"learning_rate": 2.1215169053088825e-05,
"loss": 0.2137,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.30588607211179775,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.2062,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.28779759783739206,
"learning_rate": 2.110225054402733e-05,
"loss": 0.2153,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.3437009601937799,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.2365,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.29194220580538505,
"learning_rate": 2.09892967807724e-05,
"loss": 0.2306,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.3179119058075917,
"learning_rate": 2.093280780774835e-05,
"loss": 0.219,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.31707619018100625,
"learning_rate": 2.087631137601793e-05,
"loss": 0.226,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.29623859814697634,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.222,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.2907333287034914,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.2298,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.2762875905252602,
"learning_rate": 2.070678184630458e-05,
"loss": 0.2164,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.3239250472209428,
"learning_rate": 2.06502600977305e-05,
"loss": 0.233,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.2744884724272244,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.2388,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.2862732891542417,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.2151,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.29311491051376687,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.2325,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.2830206905041068,
"learning_rate": 2.042412562887594e-05,
"loss": 0.2199,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.30659784208821633,
"learning_rate": 2.0367582403231033e-05,
"loss": 0.2211,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.2905383837446636,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.2074,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.30131331499472097,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.2126,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.3084380780997318,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.2348,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.28843956088824224,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.2157,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.2982201349965727,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.2244,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.2904553917115968,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.2302,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.3059974418412637,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.2203,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.2858087267631303,
"learning_rate": 1.99151687695709e-05,
"loss": 0.2363,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.3287867875186797,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.2313,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.27937294639659566,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.2155,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.2769490924845332,
"learning_rate": 1.974551241345447e-05,
"loss": 0.2237,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.33818824474176945,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.2286,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.2853262876023708,
"learning_rate": 1.963241759676897e-05,
"loss": 0.2363,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.3125145355225224,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.2313,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.29493600948430987,
"learning_rate": 1.951933453677638e-05,
"loss": 0.2104,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.31073208829310156,
"learning_rate": 1.9462798545816932e-05,
"loss": 0.2124,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.28659567077906184,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.2136,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.3185910814600777,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.2045,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.2826071551990233,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.2168,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.30023441728776096,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.2188,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.306322586099109,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.2187,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.3092502339524204,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.2161,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.3262189437887334,
"learning_rate": 1.906719219225165e-05,
"loss": 0.2282,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.27951774230740606,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.2203,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.3025641791258419,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.2191,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.3177827684196955,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.2351,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.27801201494661576,
"learning_rate": 1.884128556891707e-05,
"loss": 0.2215,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.336177317517899,
"learning_rate": 1.878483094691119e-05,
"loss": 0.2197,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.3190410301930468,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.2337,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.2915190876768107,
"learning_rate": 1.867195130360947e-05,
"loss": 0.2144,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.31925143092116004,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.228,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.2881148169420616,
"learning_rate": 1.8559114136390795e-05,
"loss": 0.2262,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.29867868857382107,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.2073,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.3063890591648297,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.2287,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.303550593282036,
"learning_rate": 1.838994592242891e-05,
"loss": 0.2056,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.29282951105912053,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.2199,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.3366768741168272,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.2212,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.30816064016657835,
"learning_rate": 1.82208935733949e-05,
"loss": 0.2139,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.2999265885539367,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.2112,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.3595598713590131,
"learning_rate": 1.810826238483892e-05,
"loss": 0.2175,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.30529561548691186,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.2217,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.3002849336157821,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.2169,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.3644843336200708,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.2137,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.29248573939177736,
"learning_rate": 1.788318512320838e-05,
"loss": 0.2148,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.3187818773378321,
"learning_rate": 1.782695699829051e-05,
"loss": 0.2133,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.30484068506709433,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.2201,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.3099015550725969,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.2265,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.32112203606598233,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.225,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.30342580300962047,
"learning_rate": 1.76022227484387e-05,
"loss": 0.2115,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.30155648475437674,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.2079,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.3167489382364859,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.2307,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.29741523633903,
"learning_rate": 1.743387180005809e-05,
"loss": 0.2208,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.284355839338277,
"learning_rate": 1.737779525652899e-05,
"loss": 0.2289,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.3083483782414399,
"learning_rate": 1.732173968007914e-05,
"loss": 0.2238,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.2863580557007244,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.2144,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.2832742009425353,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.2079,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.3050259881763377,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.2221,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.2875181843078019,
"learning_rate": 1.709773600686015e-05,
"loss": 0.2127,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.28856644358755734,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.2313,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.2796792491127687,
"learning_rate": 1.69858716181154e-05,
"loss": 0.1997,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.29656297192928727,
"learning_rate": 1.692997535150948e-05,
"loss": 0.2176,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.2786166821928314,
"learning_rate": 1.687410363273884e-05,
"loss": 0.2188,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.28243888240686177,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.2238,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.2868244662416192,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.2148,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.28448354389082814,
"learning_rate": 1.67066402299235e-05,
"loss": 0.2113,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.27145660509400277,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.2234,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.29486383250274945,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.2218,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.3033468221750191,
"learning_rate": 1.653941382839745e-05,
"loss": 0.2082,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.397154765990684,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.154,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.31724090813668876,
"learning_rate": 1.642806717162757e-05,
"loss": 0.1621,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.3309493935868425,
"learning_rate": 1.637243646233718e-05,
"loss": 0.1581,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.4561764231645493,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.1551,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.40742918070955675,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.1557,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.3451877214136376,
"learning_rate": 1.620572014799777e-05,
"loss": 0.158,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.3879841598746224,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.1625,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.3333264989102448,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.1483,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.35282540014817515,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.1645,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.3258716875545596,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.1412,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.3500770667323516,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.153,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.35578422416159156,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.1528,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.327428289428976,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.1561,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.33309502111342526,
"learning_rate": 1.576251073999917e-05,
"loss": 0.1554,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.3062273897524306,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.1404,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.362718785857218,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.1453,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.33583223732445733,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.1526,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.2935567022659122,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.1591,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.34500761954667325,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.1531,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.3322578078679699,
"learning_rate": 1.543151434154755e-05,
"loss": 0.1576,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.2974749239498471,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.1552,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.310211654882383,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.154,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.2870962692545311,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.1412,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.2873221042772823,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.1489,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.31240733092553125,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.1485,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.2705337681983615,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.1642,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.29111302013953444,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.1607,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.30581344782050396,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.1513,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.29965042243403034,
"learning_rate": 1.493751509814374e-05,
"loss": 0.151,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.3142239971049906,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.1539,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.29658094452962847,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.1657,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.2996520891656327,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.1463,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.3053996658022313,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.1463,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.2615142391163506,
"learning_rate": 1.466446738480111e-05,
"loss": 0.156,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.3107428060839664,
"learning_rate": 1.460998408550112e-05,
"loss": 0.1408,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.285230296815445,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.1367,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.28322115386860847,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.1585,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.27620321053932395,
"learning_rate": 1.444679451900437e-05,
"loss": 0.1621,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.2722857169593467,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.1492,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.27289156741592324,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.1573,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.2756380904974883,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.1462,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.26864160945293925,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.1467,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.2772810555363997,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.1356,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.2968552566771805,
"learning_rate": 1.412162598317602e-05,
"loss": 0.1479,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.2699856746407695,
"learning_rate": 1.406759321421907e-05,
"loss": 0.1551,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.2828065525955027,
"learning_rate": 1.401360788062871e-05,
"loss": 0.1588,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.2703209034034535,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.1442,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.2765292975310289,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.1301,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.2605317963615092,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.1434,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.2777966011222208,
"learning_rate": 1.379814952747693e-05,
"loss": 0.1501,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.2931715876312652,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.1636,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.2970989164273828,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.1481,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.2720706035166885,
"learning_rate": 1.363707494698459e-05,
"loss": 0.1553,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.2845482591423535,
"learning_rate": 1.358348460369265e-05,
"loss": 0.1486,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.27296654928527103,
"learning_rate": 1.352994556668684e-05,
"loss": 0.1514,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.2786499156990954,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.1626,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.28053197783381384,
"learning_rate": 1.342302312350678e-05,
"loss": 0.1534,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.2803503823015207,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.1474,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.28015280593629205,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.1529,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.3116614258729012,
"learning_rate": 1.326303494478451e-05,
"loss": 0.1535,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.28761386997833255,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.1611,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.2771075200561231,
"learning_rate": 1.315664479122898e-05,
"loss": 0.1606,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.2773124667075005,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.1626,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.2752343290055939,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.1455,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.30041624572423,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.1516,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.28464815790944403,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.1569,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.28646694264897826,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.1477,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.28218937309294073,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.1541,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.2766769708813757,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.1565,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.2777285585306405,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.1331,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.280667187073768,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.147,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.27594290103487773,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.1308,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.27282032215386176,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.1518,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.2752358987616325,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.1817,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.269660375985427,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.143,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.27066584074085553,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.1487,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.2876389891863877,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.1446,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.2738781919235518,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.1429,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.2649577797139933,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.1404,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.27677091121011294,
"learning_rate": 1.220938883250288e-05,
"loss": 0.1755,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.26833462555675436,
"learning_rate": 1.215733275116827e-05,
"loss": 0.1485,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.2807099500703982,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.1465,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.2827646416391426,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.1544,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.28925620627085374,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.1405,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.26627576885224136,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.1535,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.30701001064698724,
"learning_rate": 1.189800129803139e-05,
"loss": 0.1549,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.2761489609709447,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.1719,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.28822249176147313,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.154,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.2903286602312146,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.141,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.2702422609280513,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.1567,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.2859224614853225,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.1612,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.27562782818898446,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.1612,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.26462207640099084,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.1438,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.2847800886994583,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.1535,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.27034294684101673,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.1417,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.2695829435174694,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.1621,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.27764708548421635,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.1553,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.2788737666645289,
"learning_rate": 1.128231163049258e-05,
"loss": 0.1417,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.28092863282346253,
"learning_rate": 1.123144751066259e-05,
"loss": 0.1558,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.2659867270338007,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.1338,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.2714107122801702,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.1565,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.26319233865989317,
"learning_rate": 1.107927745387114e-05,
"loss": 0.1571,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.2779891755355749,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.158,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.2737744928276358,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.1546,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.27558115789575033,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.1329,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.2656819695365569,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.1669,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.2840196145081312,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.1469,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.2705179347229032,
"learning_rate": 1.077687414070869e-05,
"loss": 0.1642,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.2895289339998133,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.1426,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.28825164048576984,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.1511,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.2874326379490155,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.1434,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.27943834199556433,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.14,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.27482126190709016,
"learning_rate": 1.052689572753247e-05,
"loss": 0.1451,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.2705501733765905,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.1573,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.2953494895821463,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.1313,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.2868315239828167,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.1529,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.2652934624395068,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.1467,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.2747461384029013,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.1568,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.2856705570961946,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.14,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.2665687658103036,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.1504,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.2607073062179364,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.1629,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.26573247463368915,
"learning_rate": 1.008173988718005e-05,
"loss": 0.1469,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.2881251903437462,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.1549,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.29093635700063825,
"learning_rate": 9.983678617766677e-06,
"loss": 0.1752,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.2515194862450111,
"learning_rate": 9.934767922553788e-06,
"loss": 0.1512,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.2847418281646088,
"learning_rate": 9.885937708668371e-06,
"loss": 0.1508,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.29245375344174906,
"learning_rate": 9.837188366555524e-06,
"loss": 0.1463,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.27651672436222824,
"learning_rate": 9.788520286013691e-06,
"loss": 0.1451,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.2804916067409061,
"learning_rate": 9.739933856191537e-06,
"loss": 0.1419,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.2773065068642153,
"learning_rate": 9.691429465584866e-06,
"loss": 0.1433,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.27375559518205517,
"learning_rate": 9.643007502033492e-06,
"loss": 0.1532,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.2830465935957425,
"learning_rate": 9.594668352718152e-06,
"loss": 0.1732,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.2749654364196413,
"learning_rate": 9.546412404157383e-06,
"loss": 0.1526,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.27630492990212674,
"learning_rate": 9.498240042204471e-06,
"loss": 0.1564,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.2600408060611999,
"learning_rate": 9.450151652044339e-06,
"loss": 0.1533,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.2788724139346997,
"learning_rate": 9.402147618190472e-06,
"loss": 0.1526,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.2695136567684646,
"learning_rate": 9.354228324481841e-06,
"loss": 0.1626,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.27292129128144854,
"learning_rate": 9.306394154079852e-06,
"loss": 0.1537,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.2685242603784752,
"learning_rate": 9.258645489465254e-06,
"loss": 0.161,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.2695940770743695,
"learning_rate": 9.21098271243509e-06,
"loss": 0.1522,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.2630083976060877,
"learning_rate": 9.163406204099659e-06,
"loss": 0.1545,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.27308650953308256,
"learning_rate": 9.115916344879452e-06,
"loss": 0.134,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.2766994155736057,
"learning_rate": 9.068513514502126e-06,
"loss": 0.1493,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.2699509838435309,
"learning_rate": 9.021198091999439e-06,
"loss": 0.1421,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.276837898504343,
"learning_rate": 8.973970455704251e-06,
"loss": 0.1464,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.26247278274297686,
"learning_rate": 8.926830983247495e-06,
"loss": 0.1418,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.2783528751270847,
"learning_rate": 8.879780051555116e-06,
"loss": 0.1597,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.2598621685207637,
"learning_rate": 8.83281803684512e-06,
"loss": 0.1523,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.276002826698117,
"learning_rate": 8.785945314624532e-06,
"loss": 0.1351,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.2693729333630508,
"learning_rate": 8.739162259686385e-06,
"loss": 0.1545,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.26779021704402417,
"learning_rate": 8.692469246106734e-06,
"loss": 0.1542,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.261457233210652,
"learning_rate": 8.645866647241677e-06,
"loss": 0.143,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.26777302244317513,
"learning_rate": 8.599354835724348e-06,
"loss": 0.1441,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.2752994186012688,
"learning_rate": 8.552934183461958e-06,
"loss": 0.1461,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.25308233064112295,
"learning_rate": 8.506605061632805e-06,
"loss": 0.1773,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.25897884480768957,
"learning_rate": 8.460367840683308e-06,
"loss": 0.152,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.2794732715451756,
"learning_rate": 8.414222890325065e-06,
"loss": 0.1323,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.26225257453671713,
"learning_rate": 8.368170579531855e-06,
"loss": 0.1462,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.2641475555083238,
"learning_rate": 8.322211276536733e-06,
"loss": 0.1398,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.273114011220243,
"learning_rate": 8.27634534882907e-06,
"loss": 0.1308,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.274028462632121,
"learning_rate": 8.230573163151585e-06,
"loss": 0.1354,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.26043767917517735,
"learning_rate": 8.184895085497466e-06,
"loss": 0.1546,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.2713702991415884,
"learning_rate": 8.139311481107401e-06,
"loss": 0.1515,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.26647846271607534,
"learning_rate": 8.093822714466686e-06,
"loss": 0.1517,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.2637649211856946,
"learning_rate": 8.048429149302271e-06,
"loss": 0.1693,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.26160995707596774,
"learning_rate": 8.003131148579902e-06,
"loss": 0.1464,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.2682883437488621,
"learning_rate": 7.957929074501182e-06,
"loss": 0.1468,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.2572694845969744,
"learning_rate": 7.912823288500695e-06,
"loss": 0.1475,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.26058922191820194,
"learning_rate": 7.867814151243102e-06,
"loss": 0.1632,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.2737163891529833,
"learning_rate": 7.822902022620267e-06,
"loss": 0.1431,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.26821023293949703,
"learning_rate": 7.778087261748377e-06,
"loss": 0.1528,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.2641992861301323,
"learning_rate": 7.733370226965056e-06,
"loss": 0.1357,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.24617006605069922,
"learning_rate": 7.688751275826527e-06,
"loss": 0.1598,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.2640995830766692,
"learning_rate": 7.644230765104735e-06,
"loss": 0.1514,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.26003337482906347,
"learning_rate": 7.599809050784502e-06,
"loss": 0.133,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.25756416893652456,
"learning_rate": 7.555486488060661e-06,
"loss": 0.1428,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.2645128631938312,
"learning_rate": 7.511263431335249e-06,
"loss": 0.1473,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.25613433550127446,
"learning_rate": 7.467140234214656e-06,
"loss": 0.1671,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.2564548200050186,
"learning_rate": 7.423117249506777e-06,
"loss": 0.1353,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.2578989271807504,
"learning_rate": 7.379194829218231e-06,
"loss": 0.1449,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.26260908004683664,
"learning_rate": 7.335373324551518e-06,
"loss": 0.136,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.258751623833181,
"learning_rate": 7.291653085902224e-06,
"loss": 0.1651,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.2750016260961553,
"learning_rate": 7.248034462856213e-06,
"loss": 0.1526,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.26817602646397765,
"learning_rate": 7.20451780418683e-06,
"loss": 0.156,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.26629593939058444,
"learning_rate": 7.161103457852128e-06,
"loss": 0.1587,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.250604037051657,
"learning_rate": 7.117791770992046e-06,
"loss": 0.1567,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.26263039834519103,
"learning_rate": 7.074583089925688e-06,
"loss": 0.1506,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.2612823497485445,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.1455,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.24420583441662602,
"learning_rate": 6.988476126329604e-06,
"loss": 0.1432,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.262707499182776,
"learning_rate": 6.945578532308857e-06,
"loss": 0.1439,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.26424851376296193,
"learning_rate": 6.902785321094301e-06,
"loss": 0.1575,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.266477945260017,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.1358,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.27065259284908255,
"learning_rate": 6.817513414939892e-06,
"loss": 0.1671,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.2567945428869787,
"learning_rate": 6.775035401831924e-06,
"loss": 0.157,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.24895945642456452,
"learning_rate": 6.732663135188473e-06,
"loss": 0.144,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.25720648824503595,
"learning_rate": 6.690396953817055e-06,
"loss": 0.1499,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.26289643963252957,
"learning_rate": 6.648237195676927e-06,
"loss": 0.1557,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.2730819534520769,
"learning_rate": 6.606184197876395e-06,
"loss": 0.1395,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.2595178748391058,
"learning_rate": 6.564238296670109e-06,
"loss": 0.1468,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.2642087084071547,
"learning_rate": 6.52239982745638e-06,
"loss": 0.1452,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.2665287706592239,
"learning_rate": 6.480669124774484e-06,
"loss": 0.1529,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.26304838627082067,
"learning_rate": 6.439046522302013e-06,
"loss": 0.1619,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.2644911347946117,
"learning_rate": 6.397532352852191e-06,
"loss": 0.1412,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.270954512851971,
"learning_rate": 6.356126948371202e-06,
"loss": 0.1492,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.25365365232608467,
"learning_rate": 6.314830639935559e-06,
"loss": 0.1596,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.27410865492691355,
"learning_rate": 6.273643757749448e-06,
"loss": 0.141,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.2655241726727382,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.1561,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.2580080672419028,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.1505,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.2564656344033108,
"learning_rate": 6.150742957589784e-06,
"loss": 0.1496,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.2552708273827649,
"learning_rate": 6.109997064904774e-06,
"loss": 0.1625,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.25667249208566706,
"learning_rate": 6.069362236313092e-06,
"loss": 0.1299,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.26216845621048446,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.1651,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.26363622622663857,
"learning_rate": 5.988427070179108e-06,
"loss": 0.1567,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.2629341242313908,
"learning_rate": 5.948127379792234e-06,
"loss": 0.157,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.25251858995233945,
"learning_rate": 5.90794004780439e-06,
"loss": 0.1514,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.25867453870111473,
"learning_rate": 5.867865395552405e-06,
"loss": 0.1609,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.2698227386779464,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.161,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.2626768402477025,
"learning_rate": 5.788055411095892e-06,
"loss": 0.1376,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.25525916652094827,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.1463,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.25196534679766697,
"learning_rate": 5.708699979051613e-06,
"loss": 0.1403,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.25946207585470615,
"learning_rate": 5.66919351390751e-06,
"loss": 0.1387,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.262691017871041,
"learning_rate": 5.629801637510188e-06,
"loss": 0.1515,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.267062202039777,
"learning_rate": 5.590524664836043e-06,
"loss": 0.1369,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.26469013789970497,
"learning_rate": 5.551362909942708e-06,
"loss": 0.1359,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.2615429349959115,
"learning_rate": 5.51231668596653e-06,
"loss": 0.1391,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.25572609526830975,
"learning_rate": 5.47338630512007e-06,
"loss": 0.1454,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.2680200286018108,
"learning_rate": 5.434572078689617e-06,
"loss": 0.1301,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.2565476109650748,
"learning_rate": 5.395874317032694e-06,
"loss": 0.1513,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.2590516710581462,
"learning_rate": 5.357293329575573e-06,
"loss": 0.1472,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.26097877871573305,
"learning_rate": 5.318829424810797e-06,
"loss": 0.1667,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.25894915842545013,
"learning_rate": 5.280482910294728e-06,
"loss": 0.1451,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.26186051981052993,
"learning_rate": 5.242254092645078e-06,
"loss": 0.1395,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.26093131628354055,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.1564,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.2566547101467263,
"learning_rate": 5.166150769707892e-06,
"loss": 0.1546,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.24984394537915675,
"learning_rate": 5.128276872940503e-06,
"loss": 0.1351,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.2592946865641581,
"learning_rate": 5.090521890074953e-06,
"loss": 0.1447,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.26217624266005723,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.1373,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.259897227281541,
"learning_rate": 5.015369872647442e-06,
"loss": 0.1389,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.2581422941705377,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.137,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.261182948189605,
"learning_rate": 4.940697121074675e-06,
"loss": 0.1408,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.26075994958634224,
"learning_rate": 4.903541216934864e-06,
"loss": 0.1448,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.2613205372528744,
"learning_rate": 4.8665060236772e-06,
"loss": 0.1346,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.2525017895426983,
"learning_rate": 4.829591837434102e-06,
"loss": 0.1507,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.25580221017391963,
"learning_rate": 4.792798953370424e-06,
"loss": 0.1497,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.25218331207749517,
"learning_rate": 4.756127665681083e-06,
"loss": 0.1461,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.2634174818794725,
"learning_rate": 4.719578267588725e-06,
"loss": 0.1393,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.24419918003200825,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.135,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.33789773913686305,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.1272,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.2957350127486132,
"learning_rate": 4.610664328486485e-06,
"loss": 0.1092,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.24776779838530963,
"learning_rate": 4.5746054014809e-06,
"loss": 0.1005,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.23384780956622314,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.1089,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.23456097656258745,
"learning_rate": 4.502857857942204e-06,
"loss": 0.0926,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.3239181604816918,
"learning_rate": 4.467169815100549e-06,
"loss": 0.1118,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.32961319508498604,
"learning_rate": 4.431605972355137e-06,
"loss": 0.0991,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.30576486613258447,
"learning_rate": 4.396166614073516e-06,
"loss": 0.1138,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.2805597657697385,
"learning_rate": 4.360852023627846e-06,
"loss": 0.0986,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.2612318500826581,
"learning_rate": 4.325662483392656e-06,
"loss": 0.1078,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.24556777813301536,
"learning_rate": 4.29059827474257e-06,
"loss": 0.1026,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.24052406327443757,
"learning_rate": 4.255659678050063e-06,
"loss": 0.0956,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.26300782836235265,
"learning_rate": 4.22084697268323e-06,
"loss": 0.1047,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.2566254772883887,
"learning_rate": 4.186160437003535e-06,
"loss": 0.1197,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.25259269651448957,
"learning_rate": 4.151600348363582e-06,
"loss": 0.0976,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.24281598886716832,
"learning_rate": 4.11716698310493e-06,
"loss": 0.1172,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.24202065932170075,
"learning_rate": 4.082860616555848e-06,
"loss": 0.108,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.24487781916264215,
"learning_rate": 4.048681523029125e-06,
"loss": 0.0881,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.24917650213929096,
"learning_rate": 4.014629975819882e-06,
"loss": 0.0987,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.2553879219438997,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.1122,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.26027984314343966,
"learning_rate": 3.94691060843287e-06,
"loss": 0.1105,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.23928142070017153,
"learning_rate": 3.913243329737353e-06,
"loss": 0.1056,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.24581640464970217,
"learning_rate": 3.879704680319503e-06,
"loss": 0.107,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.24444405084182458,
"learning_rate": 3.846294928353471e-06,
"loss": 0.1082,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.2509751443908099,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.0925,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.2491477440714071,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.1062,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.2458158932158229,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.117,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.24704851814610795,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.1008,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.22954704543648605,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.114,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.22647010862847253,
"learning_rate": 3.648558150665886e-06,
"loss": 0.0902,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.23621546740807103,
"learning_rate": 3.616058102895075e-06,
"loss": 0.0999,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.24404341403905255,
"learning_rate": 3.583689060687079e-06,
"loss": 0.0932,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.2347840224303744,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.0914,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.24661711535736056,
"learning_rate": 3.51934502719796e-06,
"loss": 0.0837,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.2231659432538797,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.1092,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.24339325733653985,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.1039,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.22401126505990984,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.0991,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.22184732001273355,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.129,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.24251448974072753,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.1092,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.2342287761384104,
"learning_rate": 3.32948376099375e-06,
"loss": 0.0943,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.22526576001795623,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.1216,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.23668462548684166,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.1065,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.2340769046870369,
"learning_rate": 3.236349233544438e-06,
"loss": 0.085,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.24485731128231403,
"learning_rate": 3.205572144561604e-06,
"loss": 0.0991,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.2380530018314139,
"learning_rate": 3.174929343376374e-06,
"loss": 0.0971,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.2349491939237908,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.116,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.2424467775121758,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.0979,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.245556603941716,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.1084,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.22204352028966487,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.0967,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.2234123476929975,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.1188,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.222977827761059,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.1037,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.23296063987737273,
"learning_rate": 2.964210239224696e-06,
"loss": 0.0911,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.2372087567605575,
"learning_rate": 2.934650470499869e-06,
"loss": 0.1049,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.22897639750794235,
"learning_rate": 2.905227155855128e-06,
"loss": 0.1079,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.23577054200606293,
"learning_rate": 2.875940530558519e-06,
"loss": 0.0986,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.2324951574717562,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.1026,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.24092825538141685,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.0955,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.24333499983716972,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.1133,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.23630114019260876,
"learning_rate": 2.760165589921184e-06,
"loss": 0.099,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.2358240235965679,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.0956,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.2331439944629347,
"learning_rate": 2.703104292151164e-06,
"loss": 0.099,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.22674874036789794,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.1006,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.22124371593839703,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.0986,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.22558292621393475,
"learning_rate": 2.618550200289303e-06,
"loss": 0.1012,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.23290144653397318,
"learning_rate": 2.59064316684412e-06,
"loss": 0.1007,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.2322996346173084,
"learning_rate": 2.562875338152264e-06,
"loss": 0.1045,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.23015304670036893,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.1154,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.22942959921334324,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.096,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.2193613417611723,
"learning_rate": 2.480409295328803e-06,
"loss": 0.1092,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.23144139510682243,
"learning_rate": 2.45320049480108e-06,
"loss": 0.0877,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.22249436207867992,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.1015,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.23314443849145458,
"learning_rate": 2.399204021408421e-06,
"loss": 0.1084,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.23095915839254844,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.0924,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.22970477740187228,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.1011,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.24048669020830696,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.1001,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.23084497773246698,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.1032,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.22349851229578155,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.1054,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.23096047909814085,
"learning_rate": 2.24059906434998e-06,
"loss": 0.0992,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.22769899418894932,
"learning_rate": 2.214660694699091e-06,
"loss": 0.1081,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.23295193461225647,
"learning_rate": 2.188864536147355e-06,
"loss": 0.1092,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.22208253008980294,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.0937,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.2269625160773298,
"learning_rate": 2.137699676264171e-06,
"loss": 0.0905,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.22607975924651225,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.0954,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.22546500427913985,
"learning_rate": 2.08710612114857e-06,
"loss": 0.1007,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.22651874148408946,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.1094,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.2249833341376805,
"learning_rate": 2.0370854889762e-06,
"loss": 0.09,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.2360425398315627,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.0985,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.2380743272152553,
"learning_rate": 1.987639379598427e-06,
"loss": 0.1071,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.23117146013016154,
"learning_rate": 1.963132265887919e-06,
"loss": 0.1003,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.23762640271276392,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.117,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.235756777939297,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.1026,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.23125916061399776,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.0927,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.238409367324413,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.0908,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.2262661976997714,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.1019,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.2303932767794935,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.1216,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.22738171818066427,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.0986,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.238848245969983,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.1119,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.22926926813479895,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.1001,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.22516443814578077,
"learning_rate": 1.726025117089718e-06,
"loss": 0.0933,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.2168050207496254,
"learning_rate": 1.70311497206598e-06,
"loss": 0.1089,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.23126417436297905,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.1163,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.22299395192062102,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.1148,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.23298011758644444,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.0962,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.22795017865658862,
"learning_rate": 1.612939220355272e-06,
"loss": 0.0972,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.23271806190838082,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.0917,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.21667300960188707,
"learning_rate": 1.568732761074121e-06,
"loss": 0.0934,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.22772567707809827,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.1037,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.23759662743465496,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.0959,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.22732710462682074,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.094,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.23213579091107917,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.0946,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.22279122766217904,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.115,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.23068801211168646,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.1011,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.23472746133899192,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.1087,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.22866617789656332,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.0907,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.23635967218434703,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.1044,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.22958295579318871,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.1118,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.22098488516028056,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.1013,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.22985132788823026,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.0957,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.22460504861826397,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.103,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.2253142723559058,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.0999,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.22863803648540212,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.1089,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.22435250742007445,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.1046,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.22471875726497356,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.1036,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.22421823983721534,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.1053,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.22123803477656132,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.1057,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.22402085411264133,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.0976,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.23834841179779914,
"learning_rate": 1.140415161609505e-06,
"loss": 0.1052,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.22559589281837372,
"learning_rate": 1.121666393356604e-06,
"loss": 0.1144,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.24664201416955786,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.0936,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.23023923952974068,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.1147,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.23306791272153263,
"learning_rate": 1.066326386135097e-06,
"loss": 0.091,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.2262773837581165,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.1012,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.22646613245246505,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.1051,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.2252440603934367,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.1095,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.22039521641145626,
"learning_rate": 9.9465987684501e-07,
"loss": 0.0958,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.22157416728451465,
"learning_rate": 9.771228098856534e-07,
"loss": 0.111,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.22917176471035586,
"learning_rate": 9.597378493454013e-07,
"loss": 0.1082,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.21980843967051458,
"learning_rate": 9.425051342339353e-07,
"loss": 0.1079,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.2292806507513386,
"learning_rate": 9.254248023435864e-07,
"loss": 0.1003,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.23344570012101404,
"learning_rate": 9.084969902482288e-07,
"loss": 0.0961,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.2312019560743656,
"learning_rate": 8.917218333021993e-07,
"loss": 0.1017,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.23262522292552454,
"learning_rate": 8.750994656392109e-07,
"loss": 0.086,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.22749569863095784,
"learning_rate": 8.58630020171265e-07,
"loss": 0.1048,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.23603565461606835,
"learning_rate": 8.423136285876099e-07,
"loss": 0.0817,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.22287997172355753,
"learning_rate": 8.26150421353682e-07,
"loss": 0.1057,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.2247362380661727,
"learning_rate": 8.101405277100549e-07,
"loss": 0.103,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.22608344401415317,
"learning_rate": 7.942840756714077e-07,
"loss": 0.1065,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.22339231148109834,
"learning_rate": 7.785811920255093e-07,
"loss": 0.105,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.2429004809815547,
"learning_rate": 7.630320023322024e-07,
"loss": 0.0941,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.22611168026582534,
"learning_rate": 7.476366309223903e-07,
"loss": 0.1115,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.23837003157882386,
"learning_rate": 7.323952008970537e-07,
"loss": 0.1022,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.2293350039849728,
"learning_rate": 7.173078341262618e-07,
"loss": 0.1074,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.2350995616973901,
"learning_rate": 7.023746512481989e-07,
"loss": 0.1146,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.21493492941891973,
"learning_rate": 6.875957716681902e-07,
"loss": 0.1232,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.23404647515534627,
"learning_rate": 6.729713135577687e-07,
"loss": 0.1033,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.2271798280242011,
"learning_rate": 6.585013938537143e-07,
"loss": 0.1038,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.22636986943357,
"learning_rate": 6.441861282571115e-07,
"loss": 0.1095,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.22396402338269794,
"learning_rate": 6.300256312324471e-07,
"loss": 0.1088,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.22858017832673141,
"learning_rate": 6.160200160066776e-07,
"loss": 0.0938,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.2356997943030238,
"learning_rate": 6.021693945683327e-07,
"loss": 0.1033,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.22484181068377615,
"learning_rate": 5.884738776666044e-07,
"loss": 0.0931,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.22597213346033274,
"learning_rate": 5.749335748104878e-07,
"loss": 0.1052,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.22983274716275262,
"learning_rate": 5.615485942678778e-07,
"loss": 0.0978,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.22187110235598362,
"learning_rate": 5.483190430647289e-07,
"loss": 0.1011,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.2358867101610163,
"learning_rate": 5.352450269841747e-07,
"loss": 0.1023,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.22132384196282548,
"learning_rate": 5.22326650565701e-07,
"loss": 0.105,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.22241329004569751,
"learning_rate": 5.095640171043048e-07,
"loss": 0.1082,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.2275513177984334,
"learning_rate": 4.969572286496594e-07,
"loss": 0.1002,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.22619291114352763,
"learning_rate": 4.845063860053079e-07,
"loss": 0.1037,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.21698002275210243,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.1051,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.2252160614776298,
"learning_rate": 4.600729351261857e-07,
"loss": 0.1188,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.2317029642420103,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.0938,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.22344276777621558,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.1026,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.2255890938446797,
"learning_rate": 4.245948007320588e-07,
"loss": 0.1121,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.22638363621838312,
"learning_rate": 4.130816799401971e-07,
"loss": 0.1009,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.21866192074762125,
"learning_rate": 4.017251756252827e-07,
"loss": 0.1276,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.2271659408960713,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.1079,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.22803553813283048,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.102,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.21791281245202854,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.1035,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.23129342124036184,
"learning_rate": 3.578671204615347e-07,
"loss": 0.1083,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.23054059479869465,
"learning_rate": 3.472950355547267e-07,
"loss": 0.1045,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.218647022317662,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.0957,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.2271554733703652,
"learning_rate": 3.266223765343224e-07,
"loss": 0.0964,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.2147556067587474,
"learning_rate": 3.16521967718757e-07,
"loss": 0.1073,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.23634738553976922,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.1002,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.22137561538547257,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.1065,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.22651465795869857,
"learning_rate": 2.871653894244597e-07,
"loss": 0.0903,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.2376678933430918,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.0908,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.2239207756813193,
"learning_rate": 2.683823233205485e-07,
"loss": 0.1196,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.23560164515193027,
"learning_rate": 2.592274132703287e-07,
"loss": 0.1107,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.2320876607569991,
"learning_rate": 2.502303499116754e-07,
"loss": 0.108,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.22541527961251612,
"learning_rate": 2.413912051848666e-07,
"loss": 0.1042,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.2128147171256533,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.1137,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.22009212668103206,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.0984,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.21618540242477013,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.0996,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.22087285189842126,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.1114,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.21876521217224926,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.1174,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.2249919659223271,
"learning_rate": 1.916764975761254e-07,
"loss": 0.1037,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.2196275943459484,
"learning_rate": 1.839446914621923e-07,
"loss": 0.1138,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.22216671359902881,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.1048,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.2199327196186011,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.0941,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.22458340677024724,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.093,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.2205027800497888,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.1049,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.2193899828913749,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.0859,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.22416955440257558,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.1032,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.21998105392389797,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.1152,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.22601783986496707,
"learning_rate": 1.277993209368744e-07,
"loss": 0.1074,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.22722722070783605,
"learning_rate": 1.214956261808542e-07,
"loss": 0.1019,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.21892812990402458,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.1209,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.22353487255931725,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.1071,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.22208213787428646,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.1047,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.2298536221504355,
"learning_rate": 9.787081204466253e-08,
"loss": 0.1002,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.23178356072622433,
"learning_rate": 9.236233586200627e-08,
"loss": 0.0904,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.21832483609887907,
"learning_rate": 8.70130406196279e-08,
"loss": 0.1126,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.21993240215064788,
"learning_rate": 8.182296909035047e-08,
"loss": 0.105,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.2185176006594575,
"learning_rate": 7.679216277384439e-08,
"loss": 0.0973,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.22237874429888582,
"learning_rate": 7.19206618963053e-08,
"loss": 0.0837,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.23003571627256583,
"learning_rate": 6.720850541012347e-08,
"loss": 0.1058,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.2127723617905437,
"learning_rate": 6.265573099357714e-08,
"loss": 0.1053,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.224829697603223,
"learning_rate": 5.826237505053067e-08,
"loss": 0.0945,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.2293297274678159,
"learning_rate": 5.4028472710137e-08,
"loss": 0.108,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.21172493282131324,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.1089,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.23022236270867638,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.1093,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.21750537393043187,
"learning_rate": 4.228381947004812e-08,
"loss": 0.1088,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.2294055775619191,
"learning_rate": 3.868805732811343e-08,
"loss": 0.0977,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.2280804748221771,
"learning_rate": 3.525190530455636e-08,
"loss": 0.1116,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.22512200292957613,
"learning_rate": 3.197539087475887e-08,
"loss": 0.113,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.22277238338752897,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.1005,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.23207848224904018,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.0916,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.21786265597758886,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.1033,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.22270643857453643,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.1043,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.22716870460518837,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.0953,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.2334875858204294,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.111,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.22856465371297024,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.0889,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.22436031768975057,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.1079,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.23336017210411938,
"learning_rate": 9.67435416023843e-09,
"loss": 0.1014,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.22527100534578548,
"learning_rate": 7.995445897566178e-09,
"loss": 0.088,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.2337654194085413,
"learning_rate": 6.476393170578021e-09,
"loss": 0.1061,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.23992012220797498,
"learning_rate": 5.11720812557348e-09,
"loss": 0.0903,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.23043349089631976,
"learning_rate": 3.917901630563492e-09,
"loss": 0.0942,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.21555564511294645,
"learning_rate": 2.878483275170485e-09,
"loss": 0.0859,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.22515576680071656,
"learning_rate": 1.998961370555108e-09,
"loss": 0.1119,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.21946784356059817,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.1028,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.22629153300242977,
"learning_rate": 7.196337656112384e-10,
"loss": 0.1199,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.22094496476065184,
"learning_rate": 3.198382947511292e-10,
"loss": 0.0998,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.2287977823692179,
"learning_rate": 7.995973352770137e-11,
"loss": 0.1111,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.21243235990304002,
"learning_rate": 0.0,
"loss": 0.1176,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 1.5403539686878085e+18,
"train_loss": 0.2344522412247986,
"train_runtime": 20680.2069,
"train_samples_per_second": 7.64,
"train_steps_per_second": 0.06
}
],
"logging_steps": 1,
"max_steps": 1235,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.5403539686878085e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}