b2_science_fasttext_pos_all / trainer_state.json
neginr's picture
End of training
c8e03c7 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": 7.688179863076495,
"learning_rate": 3.2258064516129035e-07,
"loss": 1.1874,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 7.73724068920001,
"learning_rate": 6.451612903225807e-07,
"loss": 1.1608,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 7.631094744198076,
"learning_rate": 9.67741935483871e-07,
"loss": 1.1354,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 7.751598346618965,
"learning_rate": 1.2903225806451614e-06,
"loss": 1.1622,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 7.166801378868399,
"learning_rate": 1.6129032258064516e-06,
"loss": 1.1403,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 7.246769634222115,
"learning_rate": 1.935483870967742e-06,
"loss": 1.1551,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 5.828531204563078,
"learning_rate": 2.2580645161290324e-06,
"loss": 1.1095,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 5.175792883539393,
"learning_rate": 2.580645161290323e-06,
"loss": 1.0644,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 2.943444652784608,
"learning_rate": 2.903225806451613e-06,
"loss": 1.0061,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 2.682834688438473,
"learning_rate": 3.225806451612903e-06,
"loss": 0.9963,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 2.491571253564692,
"learning_rate": 3.548387096774194e-06,
"loss": 1.0161,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 4.1224876467180716,
"learning_rate": 3.870967741935484e-06,
"loss": 0.9902,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 4.712693446591885,
"learning_rate": 4.193548387096774e-06,
"loss": 1.0044,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 4.531550022601154,
"learning_rate": 4.516129032258065e-06,
"loss": 0.9801,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 4.212834518348826,
"learning_rate": 4.838709677419355e-06,
"loss": 0.9764,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 3.049641643401561,
"learning_rate": 5.161290322580646e-06,
"loss": 0.9042,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 2.9670923624340317,
"learning_rate": 5.483870967741935e-06,
"loss": 0.9328,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 2.754908778187194,
"learning_rate": 5.806451612903226e-06,
"loss": 0.9131,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 2.3596075181278167,
"learning_rate": 6.129032258064517e-06,
"loss": 0.8698,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.9119590840123668,
"learning_rate": 6.451612903225806e-06,
"loss": 0.8945,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 1.621104646293026,
"learning_rate": 6.774193548387097e-06,
"loss": 0.8491,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.6854375091967153,
"learning_rate": 7.096774193548388e-06,
"loss": 0.8263,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.8127847127306422,
"learning_rate": 7.4193548387096784e-06,
"loss": 0.8241,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 1.885994687692818,
"learning_rate": 7.741935483870968e-06,
"loss": 0.8495,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 1.5013626137141376,
"learning_rate": 8.064516129032258e-06,
"loss": 0.8644,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 1.4069553550808411,
"learning_rate": 8.387096774193549e-06,
"loss": 0.8375,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 1.330491367168891,
"learning_rate": 8.70967741935484e-06,
"loss": 0.8124,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 1.169072357326092,
"learning_rate": 9.03225806451613e-06,
"loss": 0.8235,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 1.1239080192946074,
"learning_rate": 9.35483870967742e-06,
"loss": 0.7937,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 1.078022704851394,
"learning_rate": 9.67741935483871e-06,
"loss": 0.809,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 1.0033641593622902,
"learning_rate": 1e-05,
"loss": 0.8072,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 1.0195135102935031,
"learning_rate": 1.0322580645161291e-05,
"loss": 0.7994,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 0.8176714298296772,
"learning_rate": 1.0645161290322582e-05,
"loss": 0.7978,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 0.7556401588652623,
"learning_rate": 1.096774193548387e-05,
"loss": 0.7877,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 0.8569379817670468,
"learning_rate": 1.1290322580645164e-05,
"loss": 0.7789,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.951780605361866,
"learning_rate": 1.1612903225806453e-05,
"loss": 0.8109,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.7959902160815683,
"learning_rate": 1.1935483870967743e-05,
"loss": 0.7739,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.8269900269177051,
"learning_rate": 1.2258064516129034e-05,
"loss": 0.786,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.8190001552815327,
"learning_rate": 1.2580645161290324e-05,
"loss": 0.7829,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 0.7952817571835004,
"learning_rate": 1.2903225806451613e-05,
"loss": 0.7259,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.7693159298006822,
"learning_rate": 1.3225806451612903e-05,
"loss": 0.7826,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.8003013049756755,
"learning_rate": 1.3548387096774194e-05,
"loss": 0.7617,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.8260284058832189,
"learning_rate": 1.3870967741935486e-05,
"loss": 0.7865,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.8373222322784319,
"learning_rate": 1.4193548387096776e-05,
"loss": 0.7595,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.6897787368286888,
"learning_rate": 1.4516129032258066e-05,
"loss": 0.7594,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 0.7609392819336408,
"learning_rate": 1.4838709677419357e-05,
"loss": 0.7738,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.7583183304592515,
"learning_rate": 1.5161290322580646e-05,
"loss": 0.7578,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.9161446965137497,
"learning_rate": 1.5483870967741936e-05,
"loss": 0.7495,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 0.6901895227163013,
"learning_rate": 1.5806451612903226e-05,
"loss": 0.7591,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 0.7824270638190308,
"learning_rate": 1.6129032258064517e-05,
"loss": 0.7389,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 0.7915386090739668,
"learning_rate": 1.6451612903225807e-05,
"loss": 0.7556,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.9881537617097294,
"learning_rate": 1.6774193548387098e-05,
"loss": 0.7409,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 0.783097970233911,
"learning_rate": 1.7096774193548388e-05,
"loss": 0.7224,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.8817574879695529,
"learning_rate": 1.741935483870968e-05,
"loss": 0.7363,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.8582909271145768,
"learning_rate": 1.774193548387097e-05,
"loss": 0.7482,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 0.7852467586780839,
"learning_rate": 1.806451612903226e-05,
"loss": 0.7319,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.8741132494703043,
"learning_rate": 1.838709677419355e-05,
"loss": 0.7421,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 0.8824418187743397,
"learning_rate": 1.870967741935484e-05,
"loss": 0.7638,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 0.8397268081054035,
"learning_rate": 1.903225806451613e-05,
"loss": 0.7538,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 0.8349238916547099,
"learning_rate": 1.935483870967742e-05,
"loss": 0.7153,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 0.9100383390694933,
"learning_rate": 1.967741935483871e-05,
"loss": 0.7688,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 0.9033366271482325,
"learning_rate": 2e-05,
"loss": 0.7214,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 0.7911363241786531,
"learning_rate": 2.0322580645161292e-05,
"loss": 0.7527,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 0.858064630693319,
"learning_rate": 2.0645161290322582e-05,
"loss": 0.7505,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 0.7351239586372001,
"learning_rate": 2.0967741935483873e-05,
"loss": 0.7388,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 0.9837446685101224,
"learning_rate": 2.1290322580645163e-05,
"loss": 0.7524,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 0.738040402632276,
"learning_rate": 2.161290322580645e-05,
"loss": 0.7227,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 0.8365080039488958,
"learning_rate": 2.193548387096774e-05,
"loss": 0.7077,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 0.83567033571223,
"learning_rate": 2.225806451612903e-05,
"loss": 0.7228,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 0.8693446342123202,
"learning_rate": 2.2580645161290328e-05,
"loss": 0.7258,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 0.8944872854994569,
"learning_rate": 2.2903225806451615e-05,
"loss": 0.7132,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 0.9489119149125574,
"learning_rate": 2.3225806451612906e-05,
"loss": 0.7109,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 1.1225573931977324,
"learning_rate": 2.3548387096774196e-05,
"loss": 0.7145,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 0.8368190190065234,
"learning_rate": 2.3870967741935486e-05,
"loss": 0.6949,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 0.8881284672300371,
"learning_rate": 2.4193548387096777e-05,
"loss": 0.7197,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 1.0114219778379718,
"learning_rate": 2.4516129032258067e-05,
"loss": 0.7544,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 0.8791257219740465,
"learning_rate": 2.4838709677419358e-05,
"loss": 0.7268,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 0.9723042631947911,
"learning_rate": 2.5161290322580648e-05,
"loss": 0.7342,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 0.802891182715911,
"learning_rate": 2.548387096774194e-05,
"loss": 0.7188,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 0.8307929581838872,
"learning_rate": 2.5806451612903226e-05,
"loss": 0.7312,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 1.0096059599232061,
"learning_rate": 2.6129032258064516e-05,
"loss": 0.7181,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 0.8042488095524879,
"learning_rate": 2.6451612903225806e-05,
"loss": 0.7478,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 0.9758163352465477,
"learning_rate": 2.6774193548387097e-05,
"loss": 0.7765,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 0.8204704393685959,
"learning_rate": 2.7096774193548387e-05,
"loss": 0.74,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 0.8858402076986247,
"learning_rate": 2.7419354838709678e-05,
"loss": 0.713,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 0.9863138585787664,
"learning_rate": 2.774193548387097e-05,
"loss": 0.7454,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 0.7396799161683018,
"learning_rate": 2.8064516129032262e-05,
"loss": 0.7358,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 0.8312967427223847,
"learning_rate": 2.8387096774193552e-05,
"loss": 0.7108,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 0.7378260753091391,
"learning_rate": 2.8709677419354843e-05,
"loss": 0.7281,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 0.7986144747268478,
"learning_rate": 2.9032258064516133e-05,
"loss": 0.7095,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 0.8219733969099428,
"learning_rate": 2.9354838709677423e-05,
"loss": 0.7143,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 0.8527848135135395,
"learning_rate": 2.9677419354838714e-05,
"loss": 0.7202,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 0.9841667467547359,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.7361,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 0.8286501916970592,
"learning_rate": 3.032258064516129e-05,
"loss": 0.7024,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 1.08610471063365,
"learning_rate": 3.0645161290322585e-05,
"loss": 0.7431,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 0.7576349729629345,
"learning_rate": 3.096774193548387e-05,
"loss": 0.713,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 0.8882395279359305,
"learning_rate": 3.1290322580645166e-05,
"loss": 0.7338,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 1.1193813125628032,
"learning_rate": 3.161290322580645e-05,
"loss": 0.7517,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 0.777239487021528,
"learning_rate": 3.193548387096774e-05,
"loss": 0.728,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 1.2218698209573908,
"learning_rate": 3.2258064516129034e-05,
"loss": 0.7414,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 0.8256226905241524,
"learning_rate": 3.258064516129033e-05,
"loss": 0.7454,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 0.9646673001494978,
"learning_rate": 3.2903225806451614e-05,
"loss": 0.7256,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 0.9563653832674829,
"learning_rate": 3.322580645161291e-05,
"loss": 0.7297,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 0.8775983048693005,
"learning_rate": 3.3548387096774195e-05,
"loss": 0.722,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 0.8705209072424104,
"learning_rate": 3.387096774193549e-05,
"loss": 0.71,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 1.0350320628745142,
"learning_rate": 3.4193548387096776e-05,
"loss": 0.7402,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 0.9570396596798337,
"learning_rate": 3.451612903225807e-05,
"loss": 0.737,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 0.8843198115500295,
"learning_rate": 3.483870967741936e-05,
"loss": 0.7109,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 1.1038521874464942,
"learning_rate": 3.516129032258065e-05,
"loss": 0.7295,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 0.8681744115348454,
"learning_rate": 3.548387096774194e-05,
"loss": 0.7228,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 1.1144088313969713,
"learning_rate": 3.5806451612903225e-05,
"loss": 0.7199,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 0.9540042527378029,
"learning_rate": 3.612903225806452e-05,
"loss": 0.7307,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 0.8982071906920962,
"learning_rate": 3.6451612903225805e-05,
"loss": 0.6954,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 0.9294003807154252,
"learning_rate": 3.67741935483871e-05,
"loss": 0.7309,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 0.9465813210211562,
"learning_rate": 3.7096774193548386e-05,
"loss": 0.7201,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 1.0152187688810679,
"learning_rate": 3.741935483870968e-05,
"loss": 0.7373,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 0.9187206085222418,
"learning_rate": 3.7741935483870974e-05,
"loss": 0.7126,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 1.1120951696716903,
"learning_rate": 3.806451612903226e-05,
"loss": 0.7441,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 0.9334361729490989,
"learning_rate": 3.8387096774193555e-05,
"loss": 0.7214,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 0.9149624403920302,
"learning_rate": 3.870967741935484e-05,
"loss": 0.7466,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 1.031378569803417,
"learning_rate": 3.9032258064516135e-05,
"loss": 0.7507,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 0.9754705898191901,
"learning_rate": 3.935483870967742e-05,
"loss": 0.7186,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 0.9482480248567565,
"learning_rate": 3.9677419354838716e-05,
"loss": 0.7599,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 1.1624051186888371,
"learning_rate": 4e-05,
"loss": 0.738,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 0.9134932439933501,
"learning_rate": 3.9999920040266474e-05,
"loss": 0.723,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 0.9597004427597369,
"learning_rate": 3.999968016170525e-05,
"loss": 0.71,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 1.079475893416654,
"learning_rate": 3.9999280366234394e-05,
"loss": 0.7259,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 0.9523035428593268,
"learning_rate": 3.999872065705065e-05,
"loss": 0.726,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 0.9400129663197848,
"learning_rate": 3.999800103862945e-05,
"loss": 0.7117,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 0.9936725590547946,
"learning_rate": 3.9997121516724834e-05,
"loss": 0.7366,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 0.8883317534580429,
"learning_rate": 3.999608209836944e-05,
"loss": 0.7193,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 1.0808931814607143,
"learning_rate": 3.999488279187443e-05,
"loss": 0.7362,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 0.8499232078066253,
"learning_rate": 3.9993523606829425e-05,
"loss": 0.727,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 1.1322132006022083,
"learning_rate": 3.9992004554102435e-05,
"loss": 0.7146,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 0.7978191624151612,
"learning_rate": 3.999032564583976e-05,
"loss": 0.7135,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 1.0202218839333206,
"learning_rate": 3.998848689546591e-05,
"loss": 0.7239,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 0.841213839398795,
"learning_rate": 3.9986488317683496e-05,
"loss": 0.7429,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 0.953991407812021,
"learning_rate": 3.9984329928473067e-05,
"loss": 0.7318,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 0.8668420820443138,
"learning_rate": 3.998201174509306e-05,
"loss": 0.7338,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 0.8615092110982262,
"learning_rate": 3.997953378607961e-05,
"loss": 0.7154,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 0.880662320660072,
"learning_rate": 3.9976896071246403e-05,
"loss": 0.7259,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 0.9039479022119844,
"learning_rate": 3.997409862168454e-05,
"loss": 0.7359,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 0.8850181689052126,
"learning_rate": 3.997114145976237e-05,
"loss": 0.7345,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 0.8341441970088066,
"learning_rate": 3.9968024609125245e-05,
"loss": 0.7375,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 0.7704387340856799,
"learning_rate": 3.996474809469545e-05,
"loss": 0.7041,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 0.9260726541864716,
"learning_rate": 3.996131194267189e-05,
"loss": 0.7429,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 0.7406766301484381,
"learning_rate": 3.995771618052995e-05,
"loss": 0.741,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 0.8686390846123309,
"learning_rate": 3.9953960837021256e-05,
"loss": 0.6825,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 0.7075391388398525,
"learning_rate": 3.995004594217343e-05,
"loss": 0.7076,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 0.8085123302441101,
"learning_rate": 3.994597152728987e-05,
"loss": 0.7521,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 0.7840845343103943,
"learning_rate": 3.994173762494947e-05,
"loss": 0.7058,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 0.684633599365101,
"learning_rate": 3.9937344269006424e-05,
"loss": 0.7163,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 0.7350487314303473,
"learning_rate": 3.993279149458988e-05,
"loss": 0.7038,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 0.8218125666376468,
"learning_rate": 3.9928079338103695e-05,
"loss": 0.7049,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 0.9304946713872457,
"learning_rate": 3.992320783722616e-05,
"loss": 0.6994,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.759140666751379,
"learning_rate": 3.991817703090965e-05,
"loss": 0.7232,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 0.8353960982712303,
"learning_rate": 3.991298695938038e-05,
"loss": 0.6866,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 0.7193223770805731,
"learning_rate": 3.9907637664138e-05,
"loss": 0.763,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 0.6249958155997818,
"learning_rate": 3.990212918795534e-05,
"loss": 0.6934,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 0.6876824432216262,
"learning_rate": 3.9896461574878036e-05,
"loss": 0.7221,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 0.6449579166597611,
"learning_rate": 3.989063487022416e-05,
"loss": 0.694,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 0.7397860575723113,
"learning_rate": 3.9884649120583903e-05,
"loss": 0.7056,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 0.6436104396212061,
"learning_rate": 3.987850437381915e-05,
"loss": 0.7728,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 0.7172207064444239,
"learning_rate": 3.987220067906313e-05,
"loss": 0.7277,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 0.6840740124279503,
"learning_rate": 3.986573808672002e-05,
"loss": 0.7057,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 0.6888328107359344,
"learning_rate": 3.9859116648464534e-05,
"loss": 0.7077,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 0.5975672579687001,
"learning_rate": 3.985233641724152e-05,
"loss": 0.6964,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 0.6550501157581855,
"learning_rate": 3.984539744726554e-05,
"loss": 0.6883,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 0.6472741439742256,
"learning_rate": 3.983829979402038e-05,
"loss": 0.7047,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 0.6973024761964902,
"learning_rate": 3.983104351425871e-05,
"loss": 0.7175,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 0.6312743272587445,
"learning_rate": 3.982362866600154e-05,
"loss": 0.7119,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 0.6736519518198361,
"learning_rate": 3.9816055308537806e-05,
"loss": 0.7046,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 0.7190680123227876,
"learning_rate": 3.980832350242388e-05,
"loss": 0.7318,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 0.8046436781438411,
"learning_rate": 3.980043330948306e-05,
"loss": 0.7005,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 0.6538247768235116,
"learning_rate": 3.9792384792805124e-05,
"loss": 0.6888,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 0.7667103195416531,
"learning_rate": 3.978417801674579e-05,
"loss": 0.6843,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 0.6336606824456341,
"learning_rate": 3.9775813046926244e-05,
"loss": 0.7279,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 0.6119576857460497,
"learning_rate": 3.976728995023254e-05,
"loss": 0.7198,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 0.727702272418164,
"learning_rate": 3.975860879481514e-05,
"loss": 0.7181,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 0.5913609847398213,
"learning_rate": 3.974976965008833e-05,
"loss": 0.7299,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 0.6799350352323783,
"learning_rate": 3.9740772586729674e-05,
"loss": 0.7252,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.6331378337807121,
"learning_rate": 3.973161767667946e-05,
"loss": 0.6915,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 0.7146394852734002,
"learning_rate": 3.972230499314009e-05,
"loss": 0.6833,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 0.67915566984774,
"learning_rate": 3.9712834610575545e-05,
"loss": 0.7489,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 0.7688523714036912,
"learning_rate": 3.9703206604710746e-05,
"loss": 0.7034,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 0.6612472358161223,
"learning_rate": 3.969342105253098e-05,
"loss": 0.6692,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 0.7712396400318495,
"learning_rate": 3.9683478032281246e-05,
"loss": 0.7121,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 0.7854152913705449,
"learning_rate": 3.967337762346568e-05,
"loss": 0.6978,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 0.7581739539440755,
"learning_rate": 3.9663119906846885e-05,
"loss": 0.7022,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.6544122471268238,
"learning_rate": 3.965270496444528e-05,
"loss": 0.6844,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 0.7996036563118832,
"learning_rate": 3.9642132879538465e-05,
"loss": 0.7149,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 0.7242046979271821,
"learning_rate": 3.9631403736660566e-05,
"loss": 0.767,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 0.7685838203442277,
"learning_rate": 3.962051762160151e-05,
"loss": 0.7085,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 0.6727510218675633,
"learning_rate": 3.960947462140638e-05,
"loss": 0.7157,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 0.7667798505282744,
"learning_rate": 3.959827482437472e-05,
"loss": 0.732,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 0.6037440912174639,
"learning_rate": 3.958691832005981e-05,
"loss": 0.6655,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 0.6917359364509442,
"learning_rate": 3.9575405199267944e-05,
"loss": 0.731,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 0.8023801587783738,
"learning_rate": 3.956373555405774e-05,
"loss": 0.7128,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 0.6657037950753883,
"learning_rate": 3.955190947773937e-05,
"loss": 0.7114,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 0.8259233940732522,
"learning_rate": 3.9539927064873824e-05,
"loss": 0.7228,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 0.6785415824145135,
"learning_rate": 3.952778841127214e-05,
"loss": 0.7449,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 0.7441745045615109,
"learning_rate": 3.95154936139947e-05,
"loss": 0.699,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 0.7446726817250175,
"learning_rate": 3.950304277135034e-05,
"loss": 0.7283,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 0.7151064959051683,
"learning_rate": 3.94904359828957e-05,
"loss": 0.6944,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 0.8158213961381493,
"learning_rate": 3.94776733494343e-05,
"loss": 0.7408,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 0.5790968466613144,
"learning_rate": 3.9464754973015826e-05,
"loss": 0.6713,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 0.6779226589147272,
"learning_rate": 3.9451680956935275e-05,
"loss": 0.7154,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.7005757188902749,
"learning_rate": 3.943845140573213e-05,
"loss": 0.6985,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.7342365247564205,
"learning_rate": 3.942506642518952e-05,
"loss": 0.7163,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 0.707517645696881,
"learning_rate": 3.9411526122333396e-05,
"loss": 0.7103,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 0.6967294941759781,
"learning_rate": 3.939783060543167e-05,
"loss": 0.699,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 0.8991710461856597,
"learning_rate": 3.938397998399333e-05,
"loss": 0.707,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.7209133458561278,
"learning_rate": 3.936997436876756e-05,
"loss": 0.7272,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.7924571468653665,
"learning_rate": 3.935581387174289e-05,
"loss": 0.7094,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 0.6943588154202069,
"learning_rate": 3.9341498606146295e-05,
"loss": 0.683,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.8063030809707221,
"learning_rate": 3.932702868644224e-05,
"loss": 0.7119,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.6786440515780363,
"learning_rate": 3.931240422833181e-05,
"loss": 0.704,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.7058773514457327,
"learning_rate": 3.9297625348751806e-05,
"loss": 0.6869,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 0.642622801511278,
"learning_rate": 3.9282692165873744e-05,
"loss": 0.7154,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 0.7215971466811734,
"learning_rate": 3.9267604799102945e-05,
"loss": 0.7099,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.7460542345944461,
"learning_rate": 3.9252363369077615e-05,
"loss": 0.7267,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 0.5797252778365133,
"learning_rate": 3.9236967997667796e-05,
"loss": 0.6688,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 0.7337040170948218,
"learning_rate": 3.9221418807974496e-05,
"loss": 0.7205,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 0.6936266976021214,
"learning_rate": 3.92057159243286e-05,
"loss": 0.7117,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 0.763975642029759,
"learning_rate": 3.9189859472289956e-05,
"loss": 0.6968,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 0.595732533459729,
"learning_rate": 3.9173849578646324e-05,
"loss": 0.7027,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 0.8394873563019619,
"learning_rate": 3.9157686371412395e-05,
"loss": 0.675,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.659954937311976,
"learning_rate": 3.914136997982874e-05,
"loss": 0.7107,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 0.656350527349446,
"learning_rate": 3.912490053436079e-05,
"loss": 0.703,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 0.6180243756987046,
"learning_rate": 3.9108278166697805e-05,
"loss": 0.6663,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.67803108340052,
"learning_rate": 3.909150300975178e-05,
"loss": 0.6948,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 0.7166217642018008,
"learning_rate": 3.907457519765642e-05,
"loss": 0.6894,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.6430957869082683,
"learning_rate": 3.905749486576607e-05,
"loss": 0.7017,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.6735107028710873,
"learning_rate": 3.90402621506546e-05,
"loss": 0.699,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 0.7134718367818326,
"learning_rate": 3.902287719011435e-05,
"loss": 0.668,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.6247252880697669,
"learning_rate": 3.900534012315499e-05,
"loss": 0.6753,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 0.7990317245357405,
"learning_rate": 3.8987651090002454e-05,
"loss": 0.711,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 0.6152714747253478,
"learning_rate": 3.896981023209777e-05,
"loss": 0.6905,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 0.7516404362473899,
"learning_rate": 3.8951817692095966e-05,
"loss": 0.7125,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 0.6723911153521945,
"learning_rate": 3.893367361386491e-05,
"loss": 0.6846,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 0.6363154797014123,
"learning_rate": 3.891537814248417e-05,
"loss": 0.6803,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 0.679930600776297,
"learning_rate": 3.889693142424384e-05,
"loss": 0.6835,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 0.6509723586572839,
"learning_rate": 3.8878333606643405e-05,
"loss": 0.6951,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 0.6337798189385028,
"learning_rate": 3.885958483839049e-05,
"loss": 0.7053,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 0.7415071007302394,
"learning_rate": 3.8840685269399786e-05,
"loss": 0.7056,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 0.707852684655345,
"learning_rate": 3.882163505079171e-05,
"loss": 0.7167,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 0.7728937853259498,
"learning_rate": 3.880243433489132e-05,
"loss": 0.6946,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 0.8358924984566849,
"learning_rate": 3.878308327522702e-05,
"loss": 0.5458,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 0.7055359085993456,
"learning_rate": 3.876358202652936e-05,
"loss": 0.5434,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 0.7149845668048026,
"learning_rate": 3.8743930744729835e-05,
"loss": 0.5171,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.8743183951057208,
"learning_rate": 3.8724129586959535e-05,
"loss": 0.5545,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.7574497447445379,
"learning_rate": 3.870417871154801e-05,
"loss": 0.5338,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 0.817213112549742,
"learning_rate": 3.868407827802192e-05,
"loss": 0.5096,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.7640967670871153,
"learning_rate": 3.8663828447103804e-05,
"loss": 0.5456,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.764057057837443,
"learning_rate": 3.8643429380710765e-05,
"loss": 0.5098,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.7515459005534402,
"learning_rate": 3.862288124195319e-05,
"loss": 0.5452,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.9358237618854517,
"learning_rate": 3.8602184195133464e-05,
"loss": 0.5641,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 0.7161727147532368,
"learning_rate": 3.8581338405744604e-05,
"loss": 0.5176,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 0.7734990206620908,
"learning_rate": 3.8560344040468996e-05,
"loss": 0.5296,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.7060816553925373,
"learning_rate": 3.853920126717703e-05,
"loss": 0.5429,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 0.6855962822408632,
"learning_rate": 3.851791025492574e-05,
"loss": 0.5491,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.5795684224052543,
"learning_rate": 3.8496471173957515e-05,
"loss": 0.5127,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.7491904144017308,
"learning_rate": 3.847488419569866e-05,
"loss": 0.5179,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.6415881316410517,
"learning_rate": 3.8453149492758077e-05,
"loss": 0.5104,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 0.7443806276974807,
"learning_rate": 3.8431267238925886e-05,
"loss": 0.5272,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.6517176987943015,
"learning_rate": 3.840923760917199e-05,
"loss": 0.5383,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 0.7049897777111443,
"learning_rate": 3.838706077964473e-05,
"loss": 0.4865,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 0.7459073003396374,
"learning_rate": 3.836473692766944e-05,
"loss": 0.5225,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 0.6974454103944512,
"learning_rate": 3.834226623174705e-05,
"loss": 0.5237,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.6735897209397119,
"learning_rate": 3.831964887155264e-05,
"loss": 0.5366,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.7633692254565808,
"learning_rate": 3.829688502793403e-05,
"loss": 0.515,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.7383167138908566,
"learning_rate": 3.827397488291029e-05,
"loss": 0.5307,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.7325345071641233,
"learning_rate": 3.8250918619670343e-05,
"loss": 0.5158,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.7545278814385925,
"learning_rate": 3.822771642257145e-05,
"loss": 0.5546,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.7674735435020933,
"learning_rate": 3.820436847713776e-05,
"loss": 0.5177,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 0.870491924605219,
"learning_rate": 3.8180874970058826e-05,
"loss": 0.4888,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.6915196852036231,
"learning_rate": 3.8157236089188103e-05,
"loss": 0.5119,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.9599526945254842,
"learning_rate": 3.813345202354145e-05,
"loss": 0.5267,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 0.6290444309382922,
"learning_rate": 3.810952296329563e-05,
"loss": 0.5246,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.8583710313409495,
"learning_rate": 3.808544909978676e-05,
"loss": 0.5127,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.6124550399938298,
"learning_rate": 3.806123062550882e-05,
"loss": 0.5327,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.7357111134096048,
"learning_rate": 3.803686773411208e-05,
"loss": 0.527,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.7888159540768396,
"learning_rate": 3.801236062040158e-05,
"loss": 0.5267,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 0.6623319283557203,
"learning_rate": 3.798770948033553e-05,
"loss": 0.5227,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.7554925890085307,
"learning_rate": 3.796291451102381e-05,
"loss": 0.5296,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.6622812232577049,
"learning_rate": 3.793797591072631e-05,
"loss": 0.518,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.6821249182567702,
"learning_rate": 3.791289387885144e-05,
"loss": 0.5542,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.6880109495939181,
"learning_rate": 3.788766861595443e-05,
"loss": 0.545,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 0.6576218278479847,
"learning_rate": 3.7862300323735835e-05,
"loss": 0.5119,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 0.5646667318396298,
"learning_rate": 3.783678920503983e-05,
"loss": 0.5108,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 0.6424645661530645,
"learning_rate": 3.781113546385265e-05,
"loss": 0.516,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.6541945459275476,
"learning_rate": 3.7785339305300916e-05,
"loss": 0.5595,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.6292765687629019,
"learning_rate": 3.7759400935650025e-05,
"loss": 0.5199,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.825843324349711,
"learning_rate": 3.77333205623025e-05,
"loss": 0.5586,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.5930567342630753,
"learning_rate": 3.7707098393796295e-05,
"loss": 0.5187,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.7085598203328148,
"learning_rate": 3.7680734639803185e-05,
"loss": 0.5319,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 0.5705708997778567,
"learning_rate": 3.765422951112704e-05,
"loss": 0.5228,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 0.6450176859544487,
"learning_rate": 3.762758321970216e-05,
"loss": 0.5508,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.602884222871647,
"learning_rate": 3.7600795978591584e-05,
"loss": 0.513,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.5920849708601709,
"learning_rate": 3.757386800198538e-05,
"loss": 0.5497,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.590056306748609,
"learning_rate": 3.7546799505198925e-05,
"loss": 0.5231,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.6782644798300747,
"learning_rate": 3.7519590704671197e-05,
"loss": 0.54,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.609768989718998,
"learning_rate": 3.749224181796305e-05,
"loss": 0.5134,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.5981151868009856,
"learning_rate": 3.746475306375545e-05,
"loss": 0.5735,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.6418487262480571,
"learning_rate": 3.743712466184774e-05,
"loss": 0.5463,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.5856444828233683,
"learning_rate": 3.7409356833155885e-05,
"loss": 0.5087,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.6326159295763438,
"learning_rate": 3.73814497997107e-05,
"loss": 0.5107,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.6206361304883243,
"learning_rate": 3.7353403784656095e-05,
"loss": 0.5091,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.6496795764167369,
"learning_rate": 3.732521901224724e-05,
"loss": 0.5119,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.6481418372222592,
"learning_rate": 3.7296895707848844e-05,
"loss": 0.5387,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.6024949880036563,
"learning_rate": 3.7268434097933275e-05,
"loss": 0.5306,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.5989479740945335,
"learning_rate": 3.723983441007882e-05,
"loss": 0.5074,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.6054831729708476,
"learning_rate": 3.721109687296781e-05,
"loss": 0.5516,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.6675396166411166,
"learning_rate": 3.718222171638484e-05,
"loss": 0.5211,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 0.5677893787900778,
"learning_rate": 3.715320917121488e-05,
"loss": 0.5145,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.6420344311424931,
"learning_rate": 3.712405946944149e-05,
"loss": 0.5724,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.5876526578915247,
"learning_rate": 3.709477284414488e-05,
"loss": 0.5212,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.5949396523678797,
"learning_rate": 3.7065349529500136e-05,
"loss": 0.5146,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 0.6414690627380765,
"learning_rate": 3.7035789760775305e-05,
"loss": 0.5231,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.5717562699542857,
"learning_rate": 3.70060937743295e-05,
"loss": 0.5128,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.6849165952848193,
"learning_rate": 3.6976261807611045e-05,
"loss": 0.5567,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.6076714668197977,
"learning_rate": 3.694629409915555e-05,
"loss": 0.5174,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.6347171120718708,
"learning_rate": 3.691619088858401e-05,
"loss": 0.529,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.6205889673885685,
"learning_rate": 3.68859524166009e-05,
"loss": 0.5217,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.6587231712792305,
"learning_rate": 3.6855578924992226e-05,
"loss": 0.5347,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.5855821013258194,
"learning_rate": 3.6825070656623626e-05,
"loss": 0.5093,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.6471817739886698,
"learning_rate": 3.67944278554384e-05,
"loss": 0.5293,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 0.553523984073177,
"learning_rate": 3.676365076645557e-05,
"loss": 0.5124,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.6521719219585318,
"learning_rate": 3.673273963576791e-05,
"loss": 0.5062,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.6687309993835046,
"learning_rate": 3.670169471054001e-05,
"loss": 0.5389,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.6005161715200066,
"learning_rate": 3.6670516239006254e-05,
"loss": 0.5378,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.5933007235485613,
"learning_rate": 3.663920447046888e-05,
"loss": 0.5536,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.6423651858123454,
"learning_rate": 3.6607759655295954e-05,
"loss": 0.5207,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.6465850789703599,
"learning_rate": 3.657618204491937e-05,
"loss": 0.5292,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.7827370200381917,
"learning_rate": 3.6544471891832875e-05,
"loss": 0.5239,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.6661128145441453,
"learning_rate": 3.6512629449589983e-05,
"loss": 0.5478,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.7448665629230511,
"learning_rate": 3.6480654972802045e-05,
"loss": 0.5062,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.5934303057533601,
"learning_rate": 3.644854871713611e-05,
"loss": 0.5549,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.7910046098505605,
"learning_rate": 3.6416310939312924e-05,
"loss": 0.525,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.5595098714600337,
"learning_rate": 3.638394189710493e-05,
"loss": 0.5078,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.7837802868962804,
"learning_rate": 3.635144184933412e-05,
"loss": 0.5254,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.545657243756117,
"learning_rate": 3.6318811055869994e-05,
"loss": 0.5362,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.6734454506506956,
"learning_rate": 3.628604977762752e-05,
"loss": 0.5353,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.5727910215867763,
"learning_rate": 3.6253158276565004e-05,
"loss": 0.532,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.7111445975690669,
"learning_rate": 3.6220136815682016e-05,
"loss": 0.5623,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.7132597244005114,
"learning_rate": 3.618698565901727e-05,
"loss": 0.5481,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.6419189982037454,
"learning_rate": 3.615370507164654e-05,
"loss": 0.5445,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.7182124657472628,
"learning_rate": 3.61202953196805e-05,
"loss": 0.5484,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 0.6051102582169863,
"learning_rate": 3.6086756670262654e-05,
"loss": 0.5276,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.6412776910234702,
"learning_rate": 3.6053089391567135e-05,
"loss": 0.5492,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.6519133625689907,
"learning_rate": 3.601929375279662e-05,
"loss": 0.5393,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.6469256481808465,
"learning_rate": 3.598537002418012e-05,
"loss": 0.5267,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.6362209932201125,
"learning_rate": 3.5951318476970885e-05,
"loss": 0.5137,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.6704407965370529,
"learning_rate": 3.591713938344416e-05,
"loss": 0.5219,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.5945203415953054,
"learning_rate": 3.588283301689507e-05,
"loss": 0.5293,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 0.6400084069911575,
"learning_rate": 3.5848399651636424e-05,
"loss": 0.5451,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.5609988196044138,
"learning_rate": 3.5813839562996473e-05,
"loss": 0.5279,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.653234655414005,
"learning_rate": 3.5779153027316775e-05,
"loss": 0.5277,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.6215294965015982,
"learning_rate": 3.574434032194994e-05,
"loss": 0.5135,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.6113047187805281,
"learning_rate": 3.5709401725257436e-05,
"loss": 0.5637,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.5962487103282899,
"learning_rate": 3.5674337516607346e-05,
"loss": 0.5354,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.576155910755862,
"learning_rate": 3.563914797637216e-05,
"loss": 0.5282,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.6298291730946849,
"learning_rate": 3.560383338592649e-05,
"loss": 0.5537,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.6257984939543578,
"learning_rate": 3.556839402764487e-05,
"loss": 0.538,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.5972514931491779,
"learning_rate": 3.553283018489946e-05,
"loss": 0.5404,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.5753550994095172,
"learning_rate": 3.54971421420578e-05,
"loss": 0.5368,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.6186322903137738,
"learning_rate": 3.5461330184480524e-05,
"loss": 0.5173,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 0.6313686326322433,
"learning_rate": 3.54253945985191e-05,
"loss": 0.5488,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.6679073078755815,
"learning_rate": 3.5389335671513524e-05,
"loss": 0.5386,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.5890923758119357,
"learning_rate": 3.5353153691789986e-05,
"loss": 0.5369,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.6199314644155113,
"learning_rate": 3.531684894865866e-05,
"loss": 0.5355,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.6169139183614876,
"learning_rate": 3.528042173241128e-05,
"loss": 0.5243,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.6752840617709727,
"learning_rate": 3.524387233431892e-05,
"loss": 0.526,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.6115551792343791,
"learning_rate": 3.520720104662958e-05,
"loss": 0.5444,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.7024745042329117,
"learning_rate": 3.5170408162565904e-05,
"loss": 0.5251,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.6378801115610163,
"learning_rate": 3.513349397632281e-05,
"loss": 0.5188,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.7254253653326832,
"learning_rate": 3.5096458783065145e-05,
"loss": 0.5351,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.6124618087527659,
"learning_rate": 3.505930287892533e-05,
"loss": 0.5416,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.739213282241347,
"learning_rate": 3.502202656100099e-05,
"loss": 0.5401,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.5637725742814717,
"learning_rate": 3.498463012735256e-05,
"loss": 0.5246,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.7021609218467839,
"learning_rate": 3.494711387700094e-05,
"loss": 0.5466,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.6176471076598312,
"learning_rate": 3.490947810992505e-05,
"loss": 0.5598,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 0.6507807586894181,
"learning_rate": 3.48717231270595e-05,
"loss": 0.5093,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.655014225434428,
"learning_rate": 3.483384923029211e-05,
"loss": 0.536,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.6741517952614879,
"learning_rate": 3.479585672246156e-05,
"loss": 0.5301,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.7129786387071837,
"learning_rate": 3.4757745907354924e-05,
"loss": 0.5417,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.5771192086287208,
"learning_rate": 3.471951708970527e-05,
"loss": 0.5068,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.7240633525502648,
"learning_rate": 3.468117057518921e-05,
"loss": 0.53,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.6583059283345285,
"learning_rate": 3.4642706670424426e-05,
"loss": 0.5353,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.590775658301056,
"learning_rate": 3.460412568296731e-05,
"loss": 0.531,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.5741168906663832,
"learning_rate": 3.456542792131039e-05,
"loss": 0.5514,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.614159588556863,
"learning_rate": 3.4526613694879936e-05,
"loss": 0.5296,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.5035445546385336,
"learning_rate": 3.448768331403348e-05,
"loss": 0.5233,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.5967815442591875,
"learning_rate": 3.44486370900573e-05,
"loss": 0.5333,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.5369753287537966,
"learning_rate": 3.440947533516396e-05,
"loss": 0.515,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.5988047189469946,
"learning_rate": 3.4370198362489816e-05,
"loss": 0.5608,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 0.550233961611126,
"learning_rate": 3.43308064860925e-05,
"loss": 0.5287,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.6170735529928311,
"learning_rate": 3.4291300020948393e-05,
"loss": 0.5311,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.5318768701410294,
"learning_rate": 3.425167928295015e-05,
"loss": 0.527,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.6479365672183711,
"learning_rate": 3.421194458890411e-05,
"loss": 0.5303,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.5610796093145791,
"learning_rate": 3.4172096256527865e-05,
"loss": 0.5492,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.5483971030953886,
"learning_rate": 3.4132134604447595e-05,
"loss": 0.513,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.572234289393692,
"learning_rate": 3.409205995219561e-05,
"loss": 0.5342,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.5422145037225615,
"learning_rate": 3.4051872620207765e-05,
"loss": 0.5281,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.5806446018294129,
"learning_rate": 3.40115729298209e-05,
"loss": 0.5217,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.5262825513980086,
"learning_rate": 3.397116120327025e-05,
"loss": 0.531,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.5568670994247226,
"learning_rate": 3.3930637763686916e-05,
"loss": 0.5272,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.5521809476218789,
"learning_rate": 3.389000293509524e-05,
"loss": 0.5247,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.6732930496795585,
"learning_rate": 3.3849257042410225e-05,
"loss": 0.5406,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.5349710330464233,
"learning_rate": 3.380840041143494e-05,
"loss": 0.5443,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.6215121271844373,
"learning_rate": 3.376743336885792e-05,
"loss": 0.5376,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.5910043086684263,
"learning_rate": 3.372635624225055e-05,
"loss": 0.5168,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.6010605406673307,
"learning_rate": 3.3685169360064445e-05,
"loss": 0.5238,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.5983385210421989,
"learning_rate": 3.3643873051628804e-05,
"loss": 0.5483,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.6160911703041266,
"learning_rate": 3.3602467647147814e-05,
"loss": 0.5362,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.6523211884040047,
"learning_rate": 3.356095347769799e-05,
"loss": 0.5257,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.5656589605029323,
"learning_rate": 3.351933087522552e-05,
"loss": 0.5238,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.6172528163300359,
"learning_rate": 3.3477600172543626e-05,
"loss": 0.5314,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.581882548160425,
"learning_rate": 3.3435761703329894e-05,
"loss": 0.5472,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.5776436151428005,
"learning_rate": 3.3393815802123606e-05,
"loss": 0.5371,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.5989261001127604,
"learning_rate": 3.335176280432308e-05,
"loss": 0.5301,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.5169090876325297,
"learning_rate": 3.330960304618295e-05,
"loss": 0.5558,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.5581367987101147,
"learning_rate": 3.326733686481153e-05,
"loss": 0.5361,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.5502217645388633,
"learning_rate": 3.3224964598168084e-05,
"loss": 0.534,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.5483060103442325,
"learning_rate": 3.318248658506012e-05,
"loss": 0.5563,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.5007407897793369,
"learning_rate": 3.313990316514069e-05,
"loss": 0.5298,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.5486931732398301,
"learning_rate": 3.309721467890571e-05,
"loss": 0.5383,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.5055853600409609,
"learning_rate": 3.305442146769115e-05,
"loss": 0.5217,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.5416638843045197,
"learning_rate": 3.30115238736704e-05,
"loss": 0.5157,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.504186603594939,
"learning_rate": 3.296852223985148e-05,
"loss": 0.5265,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 0.649789858024631,
"learning_rate": 3.2925416910074315e-05,
"loss": 0.5298,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.5093238348222194,
"learning_rate": 3.288220822900796e-05,
"loss": 0.5274,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.6278293518530692,
"learning_rate": 3.283889654214788e-05,
"loss": 0.5248,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.5313370448959177,
"learning_rate": 3.279548219581317e-05,
"loss": 0.5336,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.6158050208715182,
"learning_rate": 3.275196553714379e-05,
"loss": 0.536,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.5460357442753421,
"learning_rate": 3.270834691409778e-05,
"loss": 0.5394,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.5477451792724171,
"learning_rate": 3.266462667544849e-05,
"loss": 0.5236,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.5947919306925887,
"learning_rate": 3.262080517078178e-05,
"loss": 0.5415,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.5542926542963483,
"learning_rate": 3.257688275049323e-05,
"loss": 0.5494,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.5260672778933645,
"learning_rate": 3.253285976578535e-05,
"loss": 0.5453,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.5275787118835654,
"learning_rate": 3.2488736568664756e-05,
"loss": 0.5319,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.5535842569522101,
"learning_rate": 3.244451351193935e-05,
"loss": 0.5369,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 0.6553488353721117,
"learning_rate": 3.240019094921551e-05,
"loss": 0.5259,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.570329864007222,
"learning_rate": 3.235576923489527e-05,
"loss": 0.5165,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.5544243848676627,
"learning_rate": 3.2311248724173476e-05,
"loss": 0.5152,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.5939886865768988,
"learning_rate": 3.226662977303494e-05,
"loss": 0.5348,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.5667398930439542,
"learning_rate": 3.2221912738251626e-05,
"loss": 0.5436,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.5889806814602448,
"learning_rate": 3.2177097977379736e-05,
"loss": 0.5329,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.6271068395076822,
"learning_rate": 3.21321858487569e-05,
"loss": 0.5378,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.5596349931978875,
"learning_rate": 3.208717671149931e-05,
"loss": 0.5353,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.5809517934398234,
"learning_rate": 3.204207092549882e-05,
"loss": 0.5303,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 0.599345918359521,
"learning_rate": 3.19968688514201e-05,
"loss": 0.5511,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.549606257129112,
"learning_rate": 3.195157085069774e-05,
"loss": 0.5322,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.5553580294186322,
"learning_rate": 3.190617728553332e-05,
"loss": 0.5168,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.5771815153639682,
"learning_rate": 3.1860688518892606e-05,
"loss": 0.5331,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.618500689663578,
"learning_rate": 3.181510491450254e-05,
"loss": 0.5497,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.6353453300804216,
"learning_rate": 3.176942683684842e-05,
"loss": 0.5403,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.5648258032969621,
"learning_rate": 3.1723654651170934e-05,
"loss": 0.5423,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.716534721670375,
"learning_rate": 3.167778872346327e-05,
"loss": 0.5113,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.5670431177170387,
"learning_rate": 3.163182942046815e-05,
"loss": 0.5249,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.6454471083240536,
"learning_rate": 3.158577710967494e-05,
"loss": 0.5201,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.6195642115703753,
"learning_rate": 3.153963215931669e-05,
"loss": 0.5202,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.5369612151368317,
"learning_rate": 3.14933949383672e-05,
"loss": 0.5063,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.6338297165437473,
"learning_rate": 3.144706581653805e-05,
"loss": 0.5514,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.5571990473077454,
"learning_rate": 3.140064516427566e-05,
"loss": 0.5423,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.5540844797284916,
"learning_rate": 3.135413335275833e-05,
"loss": 0.5251,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.5562369587257164,
"learning_rate": 3.130753075389327e-05,
"loss": 0.5138,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.5313602650574335,
"learning_rate": 3.1260837740313625e-05,
"loss": 0.5332,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.5363721002620643,
"learning_rate": 3.121405468537547e-05,
"loss": 0.519,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.5248106919685209,
"learning_rate": 3.116718196315488e-05,
"loss": 0.5498,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.6062849065147908,
"learning_rate": 3.1120219948444894e-05,
"loss": 0.5374,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.4859289564905834,
"learning_rate": 3.107316901675252e-05,
"loss": 0.518,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.5318451308037363,
"learning_rate": 3.1026029544295745e-05,
"loss": 0.5293,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.5495205394607339,
"learning_rate": 3.097880190800057e-05,
"loss": 0.5236,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.5457904420013693,
"learning_rate": 3.093148648549788e-05,
"loss": 0.5246,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.5215398203920888,
"learning_rate": 3.088408365512055e-05,
"loss": 0.5236,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.517130890113966,
"learning_rate": 3.083659379590034e-05,
"loss": 0.5571,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.5433319869097124,
"learning_rate": 3.0789017287564913e-05,
"loss": 0.521,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.5618878186603158,
"learning_rate": 3.074135451053475e-05,
"loss": 0.5144,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.5747714727840256,
"learning_rate": 3.069360584592015e-05,
"loss": 0.5427,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.5577288940117395,
"learning_rate": 3.064577167551816e-05,
"loss": 0.5331,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.5837185855856337,
"learning_rate": 3.059785238180953e-05,
"loss": 0.513,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.5257146944778895,
"learning_rate": 3.054984834795566e-05,
"loss": 0.5064,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.5170671238170341,
"learning_rate": 3.0501759957795537e-05,
"loss": 0.506,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.5515204195462049,
"learning_rate": 3.0453587595842627e-05,
"loss": 0.5292,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.5180160794261073,
"learning_rate": 3.0405331647281856e-05,
"loss": 0.5731,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.5041110512179401,
"learning_rate": 3.0356992497966508e-05,
"loss": 0.4993,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.5526807865620657,
"learning_rate": 3.0308570534415137e-05,
"loss": 0.5294,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.5921107889933835,
"learning_rate": 3.0260066143808467e-05,
"loss": 0.5388,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.5453905709801151,
"learning_rate": 3.021147971398632e-05,
"loss": 0.5419,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.5857343334245345,
"learning_rate": 3.0162811633444478e-05,
"loss": 0.4981,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.6092309149281664,
"learning_rate": 3.0114062291331635e-05,
"loss": 0.5335,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.5521741007267083,
"learning_rate": 3.0065232077446217e-05,
"loss": 0.4993,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.6334058108569546,
"learning_rate": 3.001632138223332e-05,
"loss": 0.5042,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.9145852995642506,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.323,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 0.670779548617799,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.3743,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 1.128704717482836,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.3429,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.619785976978244,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.3218,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.6646206321165281,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.3424,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.6834659659483117,
"learning_rate": 2.972118904869376e-05,
"loss": 0.3419,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.5911193521792004,
"learning_rate": 2.9671725970312563e-05,
"loss": 0.3219,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.6012826022090074,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.317,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.6283269953396701,
"learning_rate": 2.95725682050846e-05,
"loss": 0.3073,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.5861033019340514,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.3279,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.619650722301193,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.3648,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.5676172248289785,
"learning_rate": 2.942325848714505e-05,
"loss": 0.3193,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.6343788190650698,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.3665,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.5303466513603013,
"learning_rate": 2.932334127129686e-05,
"loss": 0.3241,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.5071481853608703,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.3522,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.5696237595604806,
"learning_rate": 2.922312585929131e-05,
"loss": 0.3338,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.5141585011115314,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.2972,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.6160065710808597,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.3216,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.5782128278010258,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.3372,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.5607511591001173,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.3113,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.5299218231492704,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.3197,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.4989740264461604,
"learning_rate": 2.8920722546128863e-05,
"loss": 0.35,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.5768156027453323,
"learning_rate": 2.887006998351372e-05,
"loss": 0.314,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.49871979327311416,
"learning_rate": 2.881934649605535e-05,
"loss": 0.3618,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.5587486559668555,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.3268,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.582173723618914,
"learning_rate": 2.8717688369507425e-05,
"loss": 0.3228,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.5139557125174097,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.3251,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.5663837974589488,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.3426,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.4975889034610367,
"learning_rate": 2.856467940121025e-05,
"loss": 0.3415,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.5297633030785877,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.3093,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.5314723554420299,
"learning_rate": 2.846233032790149e-05,
"loss": 0.312,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.5065752905844398,
"learning_rate": 2.84110540896642e-05,
"loss": 0.3222,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.5526948628705061,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.3212,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.4918618600642695,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.343,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.4913709607292483,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.3284,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.5351620231984423,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.327,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.4885682214946447,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.3242,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.5182275053876507,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.3357,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.5154517072613337,
"learning_rate": 2.805026032173773e-05,
"loss": 0.3318,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.4757770494764873,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.3228,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.5232660215385057,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.3517,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.4858670191356373,
"learning_rate": 2.7894660620408e-05,
"loss": 0.341,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.5055161582334698,
"learning_rate": 2.784266724883173e-05,
"loss": 0.3224,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.48200985744093006,
"learning_rate": 2.7790611167497128e-05,
"loss": 0.3258,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.4863282836418158,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.3358,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.48801947618207825,
"learning_rate": 2.768631254100716e-05,
"loss": 0.3156,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.483616424018069,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.3247,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.5059596186684188,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.319,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.4799755283852371,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.3232,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.4969311067511534,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.3339,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.4971695758864356,
"learning_rate": 2.742449775134085e-05,
"loss": 0.3317,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.4664029668623959,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.3419,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.47375273659899436,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.3358,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.504344733369845,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.3303,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.4986872719067707,
"learning_rate": 2.7213974824975785e-05,
"loss": 0.3233,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.48283116135170046,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.3314,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.45530746370630165,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.3256,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.4555070927812883,
"learning_rate": 2.705547548999996e-05,
"loss": 0.3481,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.5019915732991275,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.3375,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.4766314928269214,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.3322,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.48800699520379875,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.3563,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.49368293670726277,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.3341,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.551140140419053,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.3412,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.5041327198207888,
"learning_rate": 2.67369650552155e-05,
"loss": 0.3218,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.5380145883184225,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.3453,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.5071943026758169,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.3018,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.5196969856017222,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.3346,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.5428279409297717,
"learning_rate": 2.652354173593614e-05,
"loss": 0.3132,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.5115729091272088,
"learning_rate": 2.647005443331317e-05,
"loss": 0.3565,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.48455505375995295,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.3206,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.5170610367653151,
"learning_rate": 2.636292505301542e-05,
"loss": 0.3349,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.5228410950443815,
"learning_rate": 2.630928383194431e-05,
"loss": 0.3181,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.4777477972496495,
"learning_rate": 2.625559216200781e-05,
"loss": 0.3432,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.4906832013040439,
"learning_rate": 2.620185047252308e-05,
"loss": 0.3512,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.47140728849921976,
"learning_rate": 2.614805919320723e-05,
"loss": 0.3254,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.5015325843570141,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.3482,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.4790253202199389,
"learning_rate": 2.60403295859298e-05,
"loss": 0.3197,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.47457819343541563,
"learning_rate": 2.5986392119371297e-05,
"loss": 0.3466,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.47057627638989424,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.3277,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.5206299122128734,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.3373,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.5007689230167514,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.3364,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.5379850316673305,
"learning_rate": 2.577016790136453e-05,
"loss": 0.33,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.5730839260632903,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.3119,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.4731692312311912,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.3371,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.5726049837139522,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.3338,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.549085008306066,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.3376,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.48873125439449994,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.3494,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.542200122996558,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.3233,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.5071121262998766,
"learning_rate": 2.539001591449889e-05,
"loss": 0.3523,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.5121099031114538,
"learning_rate": 2.53355326151989e-05,
"loss": 0.329,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.5232549499915455,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.3107,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.531878310370974,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.3846,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.45628944285858164,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.3228,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.5111017029385709,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.324,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.5150968157425151,
"learning_rate": 2.506248490185627e-05,
"loss": 0.3286,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.46851011391450226,
"learning_rate": 2.500775217241923e-05,
"loss": 0.3313,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.4902561429982104,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.3471,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.48706083009489465,
"learning_rate": 2.4898167025948e-05,
"loss": 0.3255,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.510003346759521,
"learning_rate": 2.4843315485153703e-05,
"loss": 0.3604,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.49023648766978567,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.3424,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.5033864289874085,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.3255,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.5187367618966164,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.3327,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.4657014956100843,
"learning_rate": 2.46235264422994e-05,
"loss": 0.355,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.5213966883371273,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.3264,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.483707214976313,
"learning_rate": 2.451340834511592e-05,
"loss": 0.3458,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.494610435875465,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.3441,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.5232917883492346,
"learning_rate": 2.440314589184958e-05,
"loss": 0.3263,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.4831142873282886,
"learning_rate": 2.434796163357541e-05,
"loss": 0.3295,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.4622070871613757,
"learning_rate": 2.429274260911588e-05,
"loss": 0.3301,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.5220504152874641,
"learning_rate": 2.423748926000084e-05,
"loss": 0.3382,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.4900600872699815,
"learning_rate": 2.418220202803459e-05,
"loss": 0.3495,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.4770909762186715,
"learning_rate": 2.412688135529237e-05,
"loss": 0.3312,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.4456822386096439,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.3333,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.5207085262700667,
"learning_rate": 2.4016141457114372e-05,
"loss": 0.3315,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.4644166093312412,
"learning_rate": 2.396072311715187e-05,
"loss": 0.3667,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.44403375530788314,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.3387,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.49537261696919255,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.322,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.45244998698532596,
"learning_rate": 2.379427985200223e-05,
"loss": 0.3283,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.5230397021161494,
"learning_rate": 2.373873749394972e-05,
"loss": 0.3218,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.48973824836940966,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.3054,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.5099334429544269,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.3159,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.4874152149354653,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.3476,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.5098917964481762,
"learning_rate": 2.351627355800233e-05,
"loss": 0.3055,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.5182654251304336,
"learning_rate": 2.346058617160256e-05,
"loss": 0.3374,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.49109515027819156,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.3222,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.5230449506958444,
"learning_rate": 2.334912883203469e-05,
"loss": 0.3289,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.49213911740585936,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.3292,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.477063322521347,
"learning_rate": 2.323756437450136e-05,
"loss": 0.3414,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.4803332765895492,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.3249,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.511407329074668,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.3363,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.4777892722838141,
"learning_rate": 2.307002464849052e-05,
"loss": 0.3248,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.4964593071948704,
"learning_rate": 2.30141283818846e-05,
"loss": 0.3501,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.48433577345618206,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.3469,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.481597262233917,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.3312,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.5073099422758671,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.341,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.4898003733900408,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.3568,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.4666290867080843,
"learning_rate": 2.273429448107257e-05,
"loss": 0.3353,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.46679445003323,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.3538,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.5042606893263124,
"learning_rate": 2.262220474347102e-05,
"loss": 0.3438,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.5251900067345789,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.3359,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.4475771802146326,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.3251,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.4871961654196704,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.325,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.5481557523216184,
"learning_rate": 2.2397777251561302e-05,
"loss": 0.3312,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.4552744670188722,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.3428,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.5028324419219297,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.3297,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.5021802600476243,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.3292,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.474008124009915,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.3564,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.5030843684352323,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.3282,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.4970219824850111,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.3227,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.447122575504269,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.3252,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.4969574010439868,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.3417,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.5040465507412758,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.3543,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.49610049872160983,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.3491,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.4895463887857399,
"learning_rate": 2.17791064266051e-05,
"loss": 0.3363,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.4658823240652524,
"learning_rate": 2.172276926861741e-05,
"loss": 0.3449,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.4939434156410938,
"learning_rate": 2.166641833541255e-05,
"loss": 0.3202,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.44756561834162173,
"learning_rate": 2.161005407757109e-05,
"loss": 0.3197,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.4843097613872332,
"learning_rate": 2.155367694578013e-05,
"loss": 0.343,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.4906871486612406,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.3099,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.46390680753244695,
"learning_rate": 2.1440885863609208e-05,
"loss": 0.3292,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.46222418808240456,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.3251,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.47279336076321865,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.3341,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.5164036512526087,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.3554,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.44779805841066594,
"learning_rate": 2.1215169053088825e-05,
"loss": 0.331,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.5266058054833428,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.3131,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.4892153605117747,
"learning_rate": 2.110225054402733e-05,
"loss": 0.3209,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.4548059752549856,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.3408,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.4756127578471704,
"learning_rate": 2.09892967807724e-05,
"loss": 0.3547,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.48513767537418473,
"learning_rate": 2.093280780774835e-05,
"loss": 0.3382,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.45535614105289723,
"learning_rate": 2.087631137601793e-05,
"loss": 0.3504,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.4433201209405159,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.3386,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.47609456261581806,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.3351,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.438336357914269,
"learning_rate": 2.070678184630458e-05,
"loss": 0.3171,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.47580653567834896,
"learning_rate": 2.06502600977305e-05,
"loss": 0.3363,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.4655482829440276,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.3509,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.4667952465489613,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.3213,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.4447071377700525,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.3369,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.4735674926955234,
"learning_rate": 2.042412562887594e-05,
"loss": 0.3338,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.47508938316920724,
"learning_rate": 2.0367582403231033e-05,
"loss": 0.3588,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.48739331192996177,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.3238,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.4876527925147944,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.3165,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.472129957888408,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.3254,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.45034434855900163,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.3449,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.4973078096971464,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.3363,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.46820589100572435,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.3487,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.48217261619325874,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.3372,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.4507304301631925,
"learning_rate": 1.99151687695709e-05,
"loss": 0.3399,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.4662454006012672,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.3252,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.4702786581350403,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.3106,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.48078919448977414,
"learning_rate": 1.974551241345447e-05,
"loss": 0.3328,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.49631664812726234,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.3431,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.4455924238696512,
"learning_rate": 1.963241759676897e-05,
"loss": 0.3205,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.44729638876799993,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.3364,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.4744586425727677,
"learning_rate": 1.951933453677638e-05,
"loss": 0.3377,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.43901562443248177,
"learning_rate": 1.9462798545816932e-05,
"loss": 0.3266,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.45704406244416,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.3303,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.4446413761749309,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.338,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.4599475264064279,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.3149,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.44960761569886787,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.3469,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.49811989827751774,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.3349,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.4777522992614898,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.33,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.46555447520140464,
"learning_rate": 1.906719219225165e-05,
"loss": 0.3295,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.5055695393246648,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.3416,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.45789326119568396,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.3074,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.4941570863027057,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.3376,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.5051067300293707,
"learning_rate": 1.884128556891707e-05,
"loss": 0.314,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.43752092121687386,
"learning_rate": 1.878483094691119e-05,
"loss": 0.3112,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.4647269160584646,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.3406,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.5166452467408496,
"learning_rate": 1.867195130360947e-05,
"loss": 0.3364,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.47249974647376847,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.3258,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.4674606848313089,
"learning_rate": 1.8559114136390795e-05,
"loss": 0.3379,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.5488939778866391,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.3483,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.4699682022278003,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.3288,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.4913874018128893,
"learning_rate": 1.838994592242891e-05,
"loss": 0.3182,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.4798632397815016,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.3273,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.4711816065730948,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.3426,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.4725385556818966,
"learning_rate": 1.82208935733949e-05,
"loss": 0.3471,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.5018910913609181,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.3124,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.48260675057559593,
"learning_rate": 1.810826238483892e-05,
"loss": 0.3541,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.45929200315482777,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.3328,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.5057162002738401,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.3141,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.4765358598830407,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.3407,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.46059496583508885,
"learning_rate": 1.788318512320838e-05,
"loss": 0.3419,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.5053571831378909,
"learning_rate": 1.782695699829051e-05,
"loss": 0.3243,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.4918154606070306,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.3103,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.485631943459722,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.3304,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.4777566821112099,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.3213,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.4469058417009066,
"learning_rate": 1.76022227484387e-05,
"loss": 0.3502,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.45733623473117063,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.3239,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.46583940904808063,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.3423,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.483826993440775,
"learning_rate": 1.743387180005809e-05,
"loss": 0.3312,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.4454865007333446,
"learning_rate": 1.737779525652899e-05,
"loss": 0.3474,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.46188980872204044,
"learning_rate": 1.732173968007914e-05,
"loss": 0.3081,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.4850806121906629,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.3147,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.46909455216309903,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.3286,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.44889347291543147,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.3258,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.4815955121424446,
"learning_rate": 1.709773600686015e-05,
"loss": 0.3263,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.46367565546776707,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.3475,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.46306296988066253,
"learning_rate": 1.69858716181154e-05,
"loss": 0.3242,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.4773111869062809,
"learning_rate": 1.692997535150948e-05,
"loss": 0.3381,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.46134181291924675,
"learning_rate": 1.687410363273884e-05,
"loss": 0.3226,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.46765347690725967,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.3383,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.43299464141939037,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.3428,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.4292992426546697,
"learning_rate": 1.67066402299235e-05,
"loss": 0.3388,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.47159294252370426,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.3405,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.48648679684136786,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.3173,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.44536152311847943,
"learning_rate": 1.653941382839745e-05,
"loss": 0.309,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.5886588231262978,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.2017,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.4629197493088794,
"learning_rate": 1.642806717162757e-05,
"loss": 0.2332,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.3953819470676322,
"learning_rate": 1.637243646233718e-05,
"loss": 0.1989,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.5694734012378638,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.1901,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.6931010800696135,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.2198,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.4839060821409577,
"learning_rate": 1.620572014799777e-05,
"loss": 0.2112,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.5175433279010421,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.2343,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.5178108860676731,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.1896,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.46189824411918184,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.1968,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.4368887520184704,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.1921,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.40853066461669024,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.1974,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.43618252107578603,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.1897,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.46402471994342737,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.1915,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.45607868816088787,
"learning_rate": 1.576251073999917e-05,
"loss": 0.2012,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.4111180327416229,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.204,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.4168678226148701,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.1623,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.3812767005390402,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.1909,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.4269292336973488,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.1906,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.4082802799354187,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.2061,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.3812233657758722,
"learning_rate": 1.543151434154755e-05,
"loss": 0.2075,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.4022121721152854,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.1989,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.3998832297722869,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.1875,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.4336372828944708,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.2061,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.41987628734869653,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.2204,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.38939258588725717,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.2029,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.43256083390705485,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.2032,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.42070031301194283,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.1912,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.39194613179682813,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.2139,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.3989944685120803,
"learning_rate": 1.493751509814374e-05,
"loss": 0.1959,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.42616835060638647,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.1788,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.3978545911903889,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.21,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.4118253179831609,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.1835,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.38928751555759683,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.2077,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.3793588288183995,
"learning_rate": 1.466446738480111e-05,
"loss": 0.195,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.37177664591058807,
"learning_rate": 1.460998408550112e-05,
"loss": 0.1976,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.3774545800518267,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.2122,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.3902824007215372,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.206,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.3527072484819745,
"learning_rate": 1.444679451900437e-05,
"loss": 0.2026,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.3856219557894371,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.2066,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.4096112745952951,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.2082,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.3710715100635714,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.2057,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.3816395902901732,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.1935,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.4060266839676747,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.2044,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.37522507687712137,
"learning_rate": 1.412162598317602e-05,
"loss": 0.1833,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.3786958267738831,
"learning_rate": 1.406759321421907e-05,
"loss": 0.21,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.38639796290415646,
"learning_rate": 1.401360788062871e-05,
"loss": 0.1893,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.3655339616224023,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.1851,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.37656471153621407,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.1804,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.3888736957910231,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.1932,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.3681955877839395,
"learning_rate": 1.379814952747693e-05,
"loss": 0.1929,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.3914835409552191,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.1758,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.3839224216575425,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.181,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.36625069194653104,
"learning_rate": 1.363707494698459e-05,
"loss": 0.2265,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.37902835782228284,
"learning_rate": 1.358348460369265e-05,
"loss": 0.2155,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.3693761438693311,
"learning_rate": 1.352994556668684e-05,
"loss": 0.1894,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.364530932490001,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.2151,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.39277914403162534,
"learning_rate": 1.342302312350678e-05,
"loss": 0.1932,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.3934341249693444,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.1802,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.3687272221815745,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.1811,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.35941738330437284,
"learning_rate": 1.326303494478451e-05,
"loss": 0.2002,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.3980999295533112,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.208,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.3791518705225066,
"learning_rate": 1.315664479122898e-05,
"loss": 0.1946,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.38687512194360063,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.1883,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.38383726221280884,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.196,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.36970102173475694,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.1892,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.3916924357657929,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.1866,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.3846575514589827,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.2047,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.36908490133470484,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.2069,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.4241595233326808,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.2255,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.3810801468184774,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.2119,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.38845922393140353,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.2072,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.38414480469566065,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.1789,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.3858072202361327,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.2088,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.36220151193479083,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.2294,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.3731565960958359,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.205,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.41111947591407544,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.1698,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.36688751684713233,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.1993,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.3614975275205416,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.1854,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.3882633645170946,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.1804,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.3779948010165169,
"learning_rate": 1.220938883250288e-05,
"loss": 0.1962,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.379129748345763,
"learning_rate": 1.215733275116827e-05,
"loss": 0.1871,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.37381603811902425,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.1968,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.3682101750011163,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.1759,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.3664881871068986,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.1927,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.3852581028048795,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.2049,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.36619488741047224,
"learning_rate": 1.189800129803139e-05,
"loss": 0.1906,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.3675984889512967,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.2038,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.36679675090653363,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.2205,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.3756712869480593,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.2086,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.3874712272061355,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.1833,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.3771408070301882,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.2069,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.37295780401463363,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.1801,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.3599363970602588,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.2211,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.3511599323368448,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.1928,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.3702347074126345,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.1798,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.3649151433646086,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.1967,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.3771576671934007,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.2,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.3730589775049065,
"learning_rate": 1.128231163049258e-05,
"loss": 0.2186,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.3733120443770708,
"learning_rate": 1.123144751066259e-05,
"loss": 0.1881,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.3542591488524675,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.2025,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.37729729724294164,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.2149,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.38852993142490305,
"learning_rate": 1.107927745387114e-05,
"loss": 0.2075,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.35985846285631895,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.1895,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.3775564717833132,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.1841,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.3855580980100854,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.1993,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.3623869516506494,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.197,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.3804234992387937,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.1863,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.38828065108255977,
"learning_rate": 1.077687414070869e-05,
"loss": 0.193,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.3671728286277728,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.1856,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.36201077865063624,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.184,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.35329347925792826,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.2091,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.3723380526677439,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.1838,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.35877381615276355,
"learning_rate": 1.052689572753247e-05,
"loss": 0.2109,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.361105887130568,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.198,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.36283612541822996,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.2126,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.3690151361353973,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.2239,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.37834143048814534,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.1707,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.39410945889996324,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.1988,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.3638565011925731,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.1691,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.36947446032603004,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.2101,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.37700286231492963,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.1984,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.3567766678279858,
"learning_rate": 1.008173988718005e-05,
"loss": 0.2166,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.3639255085385596,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.2036,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.3526148073496369,
"learning_rate": 9.983678617766677e-06,
"loss": 0.2051,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.3522666626169635,
"learning_rate": 9.934767922553788e-06,
"loss": 0.1971,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.3815049421013661,
"learning_rate": 9.885937708668371e-06,
"loss": 0.2117,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.3533121980279878,
"learning_rate": 9.837188366555524e-06,
"loss": 0.2152,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.3699060028188981,
"learning_rate": 9.788520286013691e-06,
"loss": 0.2073,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.34890089376721556,
"learning_rate": 9.739933856191537e-06,
"loss": 0.2215,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.35120030357567134,
"learning_rate": 9.691429465584866e-06,
"loss": 0.1838,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.3472681564497665,
"learning_rate": 9.643007502033492e-06,
"loss": 0.1863,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.36953622004301606,
"learning_rate": 9.594668352718152e-06,
"loss": 0.1801,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.38495157906298705,
"learning_rate": 9.546412404157383e-06,
"loss": 0.2093,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.36516226904982074,
"learning_rate": 9.498240042204471e-06,
"loss": 0.2176,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.36850041119964216,
"learning_rate": 9.450151652044339e-06,
"loss": 0.2091,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.37324486711560895,
"learning_rate": 9.402147618190472e-06,
"loss": 0.2043,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.3747883143749412,
"learning_rate": 9.354228324481841e-06,
"loss": 0.2177,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.3930372448516834,
"learning_rate": 9.306394154079852e-06,
"loss": 0.1821,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.3642973962931714,
"learning_rate": 9.258645489465254e-06,
"loss": 0.1948,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.36975120016004975,
"learning_rate": 9.21098271243509e-06,
"loss": 0.1837,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.3859673334296028,
"learning_rate": 9.163406204099659e-06,
"loss": 0.2031,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.3683467755870678,
"learning_rate": 9.115916344879452e-06,
"loss": 0.1885,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.3707252089448323,
"learning_rate": 9.068513514502126e-06,
"loss": 0.1944,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.3560357715399127,
"learning_rate": 9.021198091999439e-06,
"loss": 0.1879,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.36208380722143574,
"learning_rate": 8.973970455704251e-06,
"loss": 0.1841,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.3661056838961244,
"learning_rate": 8.926830983247495e-06,
"loss": 0.1912,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.36621711012556596,
"learning_rate": 8.879780051555116e-06,
"loss": 0.1864,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.3593768525376992,
"learning_rate": 8.83281803684512e-06,
"loss": 0.2063,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.35299123815026917,
"learning_rate": 8.785945314624532e-06,
"loss": 0.2016,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.3511576402080915,
"learning_rate": 8.739162259686385e-06,
"loss": 0.2065,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.3677216924389236,
"learning_rate": 8.692469246106734e-06,
"loss": 0.1942,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.3719514714142007,
"learning_rate": 8.645866647241677e-06,
"loss": 0.2078,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.35923221099054775,
"learning_rate": 8.599354835724348e-06,
"loss": 0.1822,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.36953253682781284,
"learning_rate": 8.552934183461958e-06,
"loss": 0.1735,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.3658611483300556,
"learning_rate": 8.506605061632805e-06,
"loss": 0.1789,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.3772006292764284,
"learning_rate": 8.460367840683308e-06,
"loss": 0.1996,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.3720753294526855,
"learning_rate": 8.414222890325065e-06,
"loss": 0.2041,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.3583991728004014,
"learning_rate": 8.368170579531855e-06,
"loss": 0.1913,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.3709101788411572,
"learning_rate": 8.322211276536733e-06,
"loss": 0.2014,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.3805586473574677,
"learning_rate": 8.27634534882907e-06,
"loss": 0.1833,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.3560038121743905,
"learning_rate": 8.230573163151585e-06,
"loss": 0.1851,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.3615522298667963,
"learning_rate": 8.184895085497466e-06,
"loss": 0.1996,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.35770906593086355,
"learning_rate": 8.139311481107401e-06,
"loss": 0.1917,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.37430470119218506,
"learning_rate": 8.093822714466686e-06,
"loss": 0.2129,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.35120588016070686,
"learning_rate": 8.048429149302271e-06,
"loss": 0.2028,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.35309373454930915,
"learning_rate": 8.003131148579902e-06,
"loss": 0.1896,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.35189021553026395,
"learning_rate": 7.957929074501182e-06,
"loss": 0.2362,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.3453500552980331,
"learning_rate": 7.912823288500695e-06,
"loss": 0.1984,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.3624292199617923,
"learning_rate": 7.867814151243102e-06,
"loss": 0.2022,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.34361679503161807,
"learning_rate": 7.822902022620267e-06,
"loss": 0.212,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.3738168401124924,
"learning_rate": 7.778087261748377e-06,
"loss": 0.1785,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.3532873816446482,
"learning_rate": 7.733370226965056e-06,
"loss": 0.1804,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.34393413138388257,
"learning_rate": 7.688751275826527e-06,
"loss": 0.1837,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.3649898199618412,
"learning_rate": 7.644230765104735e-06,
"loss": 0.1809,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.37451971432742376,
"learning_rate": 7.599809050784502e-06,
"loss": 0.1772,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.362714851985462,
"learning_rate": 7.555486488060661e-06,
"loss": 0.1991,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.3706247847582733,
"learning_rate": 7.511263431335249e-06,
"loss": 0.1964,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.358574719081014,
"learning_rate": 7.467140234214656e-06,
"loss": 0.1761,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.36856155401925034,
"learning_rate": 7.423117249506777e-06,
"loss": 0.1733,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.37860289864786956,
"learning_rate": 7.379194829218231e-06,
"loss": 0.1815,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.353264686011739,
"learning_rate": 7.335373324551518e-06,
"loss": 0.1906,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.35689138385212615,
"learning_rate": 7.291653085902224e-06,
"loss": 0.212,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.3610511494591467,
"learning_rate": 7.248034462856213e-06,
"loss": 0.2092,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.36168515212095675,
"learning_rate": 7.20451780418683e-06,
"loss": 0.18,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.368102292724857,
"learning_rate": 7.161103457852128e-06,
"loss": 0.1938,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.35183137181552737,
"learning_rate": 7.117791770992046e-06,
"loss": 0.2201,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.3426944744597596,
"learning_rate": 7.074583089925688e-06,
"loss": 0.2,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.34915931605094597,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.2122,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.36368612821024093,
"learning_rate": 6.988476126329604e-06,
"loss": 0.1967,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.3682309536450583,
"learning_rate": 6.945578532308857e-06,
"loss": 0.2057,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.3628654775955241,
"learning_rate": 6.902785321094301e-06,
"loss": 0.1881,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.3703097572300821,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.187,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.353035636436125,
"learning_rate": 6.817513414939892e-06,
"loss": 0.1936,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.36773064534392313,
"learning_rate": 6.775035401831924e-06,
"loss": 0.2045,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.36336836380376014,
"learning_rate": 6.732663135188473e-06,
"loss": 0.1992,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.35606285895305734,
"learning_rate": 6.690396953817055e-06,
"loss": 0.2124,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.3631441081787053,
"learning_rate": 6.648237195676927e-06,
"loss": 0.1729,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.36047748609406643,
"learning_rate": 6.606184197876395e-06,
"loss": 0.1942,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.3738777295264479,
"learning_rate": 6.564238296670109e-06,
"loss": 0.1927,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.34783808023970064,
"learning_rate": 6.52239982745638e-06,
"loss": 0.2018,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.34745449702381864,
"learning_rate": 6.480669124774484e-06,
"loss": 0.2122,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.3674032380875255,
"learning_rate": 6.439046522302013e-06,
"loss": 0.1917,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.34739146718519004,
"learning_rate": 6.397532352852191e-06,
"loss": 0.2255,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.3559894391210528,
"learning_rate": 6.356126948371202e-06,
"loss": 0.2102,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.35290456283536553,
"learning_rate": 6.314830639935559e-06,
"loss": 0.2179,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.3622964733046889,
"learning_rate": 6.273643757749448e-06,
"loss": 0.1988,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.3552215078844384,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.1729,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.35838943409633256,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.1868,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.37364423213221365,
"learning_rate": 6.150742957589784e-06,
"loss": 0.2042,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.3543309456207856,
"learning_rate": 6.109997064904774e-06,
"loss": 0.2006,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.3512756248662111,
"learning_rate": 6.069362236313092e-06,
"loss": 0.2127,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.36960302787205757,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.1941,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.34175855806012123,
"learning_rate": 5.988427070179108e-06,
"loss": 0.2118,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.34182147012691877,
"learning_rate": 5.948127379792234e-06,
"loss": 0.1838,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.3556922884530027,
"learning_rate": 5.90794004780439e-06,
"loss": 0.1976,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.35631663963058213,
"learning_rate": 5.867865395552405e-06,
"loss": 0.2054,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.3432825355463661,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.1621,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.3495190455158872,
"learning_rate": 5.788055411095892e-06,
"loss": 0.2053,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.3463993159596917,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.1691,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.34839532227291486,
"learning_rate": 5.708699979051613e-06,
"loss": 0.1831,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.35974150784558523,
"learning_rate": 5.66919351390751e-06,
"loss": 0.1668,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.3674812377731477,
"learning_rate": 5.629801637510188e-06,
"loss": 0.1998,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.3830713984096136,
"learning_rate": 5.590524664836043e-06,
"loss": 0.1908,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.34061598853176955,
"learning_rate": 5.551362909942708e-06,
"loss": 0.1767,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.3540624625770054,
"learning_rate": 5.51231668596653e-06,
"loss": 0.2007,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.34518318700485684,
"learning_rate": 5.47338630512007e-06,
"loss": 0.202,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.36575530841884896,
"learning_rate": 5.434572078689617e-06,
"loss": 0.1701,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.37061165585519384,
"learning_rate": 5.395874317032694e-06,
"loss": 0.1998,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.352650425045049,
"learning_rate": 5.357293329575573e-06,
"loss": 0.1734,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.3538802443141665,
"learning_rate": 5.318829424810797e-06,
"loss": 0.1934,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.3562027659550614,
"learning_rate": 5.280482910294728e-06,
"loss": 0.2051,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.35319414278768935,
"learning_rate": 5.242254092645078e-06,
"loss": 0.1961,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.3445178433988308,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.2053,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.35122079317807064,
"learning_rate": 5.166150769707892e-06,
"loss": 0.2121,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.3470717101715046,
"learning_rate": 5.128276872940503e-06,
"loss": 0.213,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.3513619692684897,
"learning_rate": 5.090521890074953e-06,
"loss": 0.1562,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.35742812704897586,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.1822,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.3548655846476114,
"learning_rate": 5.015369872647442e-06,
"loss": 0.1742,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.3631353107620423,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.1938,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.3607801165449767,
"learning_rate": 4.940697121074675e-06,
"loss": 0.1908,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.3782516104279248,
"learning_rate": 4.903541216934864e-06,
"loss": 0.1789,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.3477490971243548,
"learning_rate": 4.8665060236772e-06,
"loss": 0.1977,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.35852362551098105,
"learning_rate": 4.829591837434102e-06,
"loss": 0.1862,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.35643576665300364,
"learning_rate": 4.792798953370424e-06,
"loss": 0.1802,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.3640180582749226,
"learning_rate": 4.756127665681083e-06,
"loss": 0.1902,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.3521714560630425,
"learning_rate": 4.719578267588725e-06,
"loss": 0.2012,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.3148407004998768,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.2008,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.4239997102327347,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.1339,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.3822680190502947,
"learning_rate": 4.610664328486485e-06,
"loss": 0.1448,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.3251019805941982,
"learning_rate": 4.5746054014809e-06,
"loss": 0.1422,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.26126825198362963,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.1384,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.2683893011822292,
"learning_rate": 4.502857857942204e-06,
"loss": 0.1369,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.3126360402902176,
"learning_rate": 4.467169815100549e-06,
"loss": 0.1336,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.3343041501641993,
"learning_rate": 4.431605972355137e-06,
"loss": 0.1316,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.4108403564820445,
"learning_rate": 4.396166614073516e-06,
"loss": 0.1495,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.3490898046151905,
"learning_rate": 4.360852023627846e-06,
"loss": 0.1246,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.3337299998949254,
"learning_rate": 4.325662483392656e-06,
"loss": 0.1426,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.308578636422647,
"learning_rate": 4.29059827474257e-06,
"loss": 0.1485,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.2896611684881828,
"learning_rate": 4.255659678050063e-06,
"loss": 0.1402,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.2987585006224719,
"learning_rate": 4.22084697268323e-06,
"loss": 0.1301,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.3051339017586801,
"learning_rate": 4.186160437003535e-06,
"loss": 0.1575,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.3005590261253238,
"learning_rate": 4.151600348363582e-06,
"loss": 0.1151,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.30421487943772874,
"learning_rate": 4.11716698310493e-06,
"loss": 0.1408,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.27791696888302664,
"learning_rate": 4.082860616555848e-06,
"loss": 0.1355,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.2627174948563368,
"learning_rate": 4.048681523029125e-06,
"loss": 0.1203,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.2810930165319879,
"learning_rate": 4.014629975819882e-06,
"loss": 0.1335,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.26528132309620567,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.1298,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.3120105432563743,
"learning_rate": 3.94691060843287e-06,
"loss": 0.134,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.2825627833343796,
"learning_rate": 3.913243329737353e-06,
"loss": 0.1361,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.26257471514442404,
"learning_rate": 3.879704680319503e-06,
"loss": 0.1279,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.26817173065717237,
"learning_rate": 3.846294928353471e-06,
"loss": 0.1297,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.2754288685335605,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.1472,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.2796323284341435,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.1197,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.2556714980106988,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.1233,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.26817200905268807,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.1144,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.26941628623652886,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.1175,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.2663532798863268,
"learning_rate": 3.648558150665886e-06,
"loss": 0.1327,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.25732321899839744,
"learning_rate": 3.616058102895075e-06,
"loss": 0.1392,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.2712325227324777,
"learning_rate": 3.583689060687079e-06,
"loss": 0.113,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.27765004472913013,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.1509,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.2613885506667227,
"learning_rate": 3.51934502719796e-06,
"loss": 0.1387,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.24959214248113462,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.1404,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.24707394092491525,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.1124,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.2624694010068742,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.122,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.26377566644646266,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.1414,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.274628372417117,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.1406,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.2671878816307565,
"learning_rate": 3.32948376099375e-06,
"loss": 0.1361,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.26473157747659914,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.1162,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.25870165547967267,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.128,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.2747020709685174,
"learning_rate": 3.236349233544438e-06,
"loss": 0.1488,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.2653331202790173,
"learning_rate": 3.205572144561604e-06,
"loss": 0.145,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.26813998721352966,
"learning_rate": 3.174929343376374e-06,
"loss": 0.1561,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.25667588920992346,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.1484,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.27966166799738645,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.1164,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.2569615615798571,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.136,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.2509829046010024,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.1382,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.2555276001096836,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.1348,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.2569765818739751,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.1223,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.25285902540703653,
"learning_rate": 2.964210239224696e-06,
"loss": 0.144,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.2627819896122741,
"learning_rate": 2.934650470499869e-06,
"loss": 0.1358,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.2578178098189507,
"learning_rate": 2.905227155855128e-06,
"loss": 0.1336,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.2555410091301199,
"learning_rate": 2.875940530558519e-06,
"loss": 0.146,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.24497818480417102,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.1178,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.25511281797221397,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.13,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.25458274188056285,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.1408,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.27612858895184317,
"learning_rate": 2.760165589921184e-06,
"loss": 0.1344,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.26819943075359104,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.1394,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.2576167807556612,
"learning_rate": 2.703104292151164e-06,
"loss": 0.0993,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.2555690098771205,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.1412,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.260102184874047,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.1359,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.2703404495443971,
"learning_rate": 2.618550200289303e-06,
"loss": 0.136,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.2640392599609205,
"learning_rate": 2.59064316684412e-06,
"loss": 0.1411,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.27068444926797686,
"learning_rate": 2.562875338152264e-06,
"loss": 0.1108,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.26483649509615326,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.1423,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.27591866684904215,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.1434,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.26647190405109306,
"learning_rate": 2.480409295328803e-06,
"loss": 0.1496,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.26181547296949376,
"learning_rate": 2.45320049480108e-06,
"loss": 0.1453,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.2709397929578278,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.1433,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.26688336634638143,
"learning_rate": 2.399204021408421e-06,
"loss": 0.1326,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.2743109333373425,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.1233,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.25886314317998194,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.1359,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.258948121649209,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.1371,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.26527858497624723,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.1216,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.25952161996239775,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.1542,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.2511429120846962,
"learning_rate": 2.24059906434998e-06,
"loss": 0.1242,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.2668443342869992,
"learning_rate": 2.214660694699091e-06,
"loss": 0.1412,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.261456388614094,
"learning_rate": 2.188864536147355e-06,
"loss": 0.1306,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.251984230942002,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.1278,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.2690718548580669,
"learning_rate": 2.137699676264171e-06,
"loss": 0.1175,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.2529084314342833,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.135,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.2705201527540329,
"learning_rate": 2.08710612114857e-06,
"loss": 0.149,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.25126275428282857,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.1505,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.26602268296882575,
"learning_rate": 2.0370854889762e-06,
"loss": 0.1456,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.269515388260461,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.1357,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.2591622367020616,
"learning_rate": 1.987639379598427e-06,
"loss": 0.1224,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.25795394275160183,
"learning_rate": 1.963132265887919e-06,
"loss": 0.1385,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.25625736611066946,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.126,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.2648105171582911,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.1493,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.2546358821382365,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.1398,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.2590353661000284,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.1354,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.27554382605835154,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.1357,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.26864429517986455,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.1413,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.24804135549137207,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.134,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.25854321116420037,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.1247,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.2492960702378381,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.1415,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.2703770822855009,
"learning_rate": 1.726025117089718e-06,
"loss": 0.1402,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.24693733485330607,
"learning_rate": 1.70311497206598e-06,
"loss": 0.1207,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.2583949482502939,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.1169,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.2482304476444032,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.1102,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.2675135763989509,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.1514,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.25715566488297475,
"learning_rate": 1.612939220355272e-06,
"loss": 0.1487,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.25266006755386367,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.1194,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.25244698783185876,
"learning_rate": 1.568732761074121e-06,
"loss": 0.138,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.2706572314799759,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.1272,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.26229245777434984,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.1302,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.27852481421823916,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.1352,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.26735458606337875,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.1503,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.26757659566319486,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.1363,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.26826183912026264,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.139,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.24731233927584245,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.1386,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.2600561534865456,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.1208,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.2639452414739435,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.1214,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.2385548118122125,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.1206,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.2619851898446747,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.1453,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.25390303879044224,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.1266,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.2545096207108936,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.1369,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.2633578252152403,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.1369,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.26356301068189936,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.1395,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.260137591810319,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.1359,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.25589838170249335,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.1109,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.24836554022189777,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.1382,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.2583587076809429,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.1637,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.2588908596363637,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.1595,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.254471262625897,
"learning_rate": 1.140415161609505e-06,
"loss": 0.1276,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.25746934564462903,
"learning_rate": 1.121666393356604e-06,
"loss": 0.1225,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.25443951599173703,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.1501,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.26357147045632784,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.1347,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.25403197400587213,
"learning_rate": 1.066326386135097e-06,
"loss": 0.1268,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.2746368300390469,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.1093,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.28462480268306334,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.1367,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.26957407425120294,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.0996,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.24427019571621197,
"learning_rate": 9.9465987684501e-07,
"loss": 0.1217,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.2687295600792783,
"learning_rate": 9.771228098856534e-07,
"loss": 0.1322,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.2598893612358071,
"learning_rate": 9.597378493454013e-07,
"loss": 0.1498,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.2562416406569612,
"learning_rate": 9.425051342339353e-07,
"loss": 0.1203,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.253197100935636,
"learning_rate": 9.254248023435864e-07,
"loss": 0.1591,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.2629729543306587,
"learning_rate": 9.084969902482288e-07,
"loss": 0.1403,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.25855262271230844,
"learning_rate": 8.917218333021993e-07,
"loss": 0.1264,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.26691617635371573,
"learning_rate": 8.750994656392109e-07,
"loss": 0.119,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.26723750770780474,
"learning_rate": 8.58630020171265e-07,
"loss": 0.135,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.2672875761446365,
"learning_rate": 8.423136285876099e-07,
"loss": 0.1298,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.2747386648717876,
"learning_rate": 8.26150421353682e-07,
"loss": 0.1344,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.244559283487711,
"learning_rate": 8.101405277100549e-07,
"loss": 0.1304,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.26186863781530534,
"learning_rate": 7.942840756714077e-07,
"loss": 0.1408,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.2519411785803035,
"learning_rate": 7.785811920255093e-07,
"loss": 0.1314,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.2717751127957952,
"learning_rate": 7.630320023322024e-07,
"loss": 0.1415,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.2577859667076002,
"learning_rate": 7.476366309223903e-07,
"loss": 0.1261,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.25757945338559696,
"learning_rate": 7.323952008970537e-07,
"loss": 0.1242,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.2605405824858115,
"learning_rate": 7.173078341262618e-07,
"loss": 0.1315,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.25668992324350937,
"learning_rate": 7.023746512481989e-07,
"loss": 0.1483,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.2601209034804113,
"learning_rate": 6.875957716681902e-07,
"loss": 0.1323,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.2674591348542859,
"learning_rate": 6.729713135577687e-07,
"loss": 0.161,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.25797381008360465,
"learning_rate": 6.585013938537143e-07,
"loss": 0.1271,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.24469186537342696,
"learning_rate": 6.441861282571115e-07,
"loss": 0.1323,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.2648666040259314,
"learning_rate": 6.300256312324471e-07,
"loss": 0.1506,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.25469027192886673,
"learning_rate": 6.160200160066776e-07,
"loss": 0.123,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.2626651681392501,
"learning_rate": 6.021693945683327e-07,
"loss": 0.1385,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.2525857088954639,
"learning_rate": 5.884738776666044e-07,
"loss": 0.1352,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.2465504262939218,
"learning_rate": 5.749335748104878e-07,
"loss": 0.1302,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.25839849460784997,
"learning_rate": 5.615485942678778e-07,
"loss": 0.1025,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.2697155846746826,
"learning_rate": 5.483190430647289e-07,
"loss": 0.1295,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.25808628842613407,
"learning_rate": 5.352450269841747e-07,
"loss": 0.1286,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.2518687155104697,
"learning_rate": 5.22326650565701e-07,
"loss": 0.1494,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.2605762108182109,
"learning_rate": 5.095640171043048e-07,
"loss": 0.1417,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.2552637685098596,
"learning_rate": 4.969572286496594e-07,
"loss": 0.1146,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.247450086327206,
"learning_rate": 4.845063860053079e-07,
"loss": 0.1323,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.24370633526257293,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.1034,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.2730916119779298,
"learning_rate": 4.600729351261857e-07,
"loss": 0.1526,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.2583319355130422,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.135,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.2694844331233726,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.1363,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.264286855416339,
"learning_rate": 4.245948007320588e-07,
"loss": 0.152,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.2548128048910112,
"learning_rate": 4.130816799401971e-07,
"loss": 0.1172,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.26265495295273367,
"learning_rate": 4.017251756252827e-07,
"loss": 0.1512,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.2588550611936135,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.1201,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.2576742375280421,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.1461,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.25366607596928437,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.1349,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.28769386056678464,
"learning_rate": 3.578671204615347e-07,
"loss": 0.1426,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.25349447910672984,
"learning_rate": 3.472950355547267e-07,
"loss": 0.1326,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.24835883738087094,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.1168,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.2488234021776992,
"learning_rate": 3.266223765343224e-07,
"loss": 0.1379,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.26122178288687536,
"learning_rate": 3.16521967718757e-07,
"loss": 0.1273,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.25478641103767213,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.1123,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.26572800837608446,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.1195,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.2574242660795796,
"learning_rate": 2.871653894244597e-07,
"loss": 0.1378,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.24351664115205504,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.1271,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.25704827509650024,
"learning_rate": 2.683823233205485e-07,
"loss": 0.1522,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.2666942698056165,
"learning_rate": 2.592274132703287e-07,
"loss": 0.1385,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.26979255707745053,
"learning_rate": 2.502303499116754e-07,
"loss": 0.1378,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.256173943551224,
"learning_rate": 2.413912051848666e-07,
"loss": 0.1258,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.257606730258103,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.152,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.2616227005466666,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.1528,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.2586035646665131,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.1405,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.257032678920735,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.1162,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.2720666496096381,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.1437,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.27331727332838607,
"learning_rate": 1.916764975761254e-07,
"loss": 0.1257,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.25106464199128864,
"learning_rate": 1.839446914621923e-07,
"loss": 0.1343,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.2617300972831417,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.1624,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.255132742582193,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.1346,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.2510759738012655,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.1157,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.24774931743271755,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.1436,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.2716470957568605,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.1416,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.25292686406719816,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.1455,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.24752786740936514,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.1363,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.2574830783716652,
"learning_rate": 1.277993209368744e-07,
"loss": 0.1382,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.24085768595163284,
"learning_rate": 1.214956261808542e-07,
"loss": 0.1294,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.25874665907959987,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.1481,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.2655974215673269,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.1301,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.26081783932754415,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.1383,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.262428992717117,
"learning_rate": 9.787081204466253e-08,
"loss": 0.1554,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.25468840456440733,
"learning_rate": 9.236233586200627e-08,
"loss": 0.1204,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.2549496903682309,
"learning_rate": 8.70130406196279e-08,
"loss": 0.1374,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.25328191221612434,
"learning_rate": 8.182296909035047e-08,
"loss": 0.0991,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.24230716875870095,
"learning_rate": 7.679216277384439e-08,
"loss": 0.1588,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.26197600797239945,
"learning_rate": 7.19206618963053e-08,
"loss": 0.1398,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.2574915299655057,
"learning_rate": 6.720850541012347e-08,
"loss": 0.1257,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.2578244027098705,
"learning_rate": 6.265573099357714e-08,
"loss": 0.1332,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.2550751850866068,
"learning_rate": 5.826237505053067e-08,
"loss": 0.1184,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.29920404270593143,
"learning_rate": 5.4028472710137e-08,
"loss": 0.1413,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.2592169877779951,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.1568,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.26109317521198894,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.1203,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.2626205563123868,
"learning_rate": 4.228381947004812e-08,
"loss": 0.1396,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.2579106959547022,
"learning_rate": 3.868805732811343e-08,
"loss": 0.1499,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.2576449224696159,
"learning_rate": 3.525190530455636e-08,
"loss": 0.1309,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.26287618133784174,
"learning_rate": 3.197539087475887e-08,
"loss": 0.144,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.26436487737088743,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.1351,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.2547712614787337,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.1185,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.2582624831258652,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.1708,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.2570570439873706,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.135,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.2570601267478999,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.1359,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.2634677390977056,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.1428,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.26918923521379823,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.1416,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.26116766858971446,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.1207,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.2629437259989721,
"learning_rate": 9.67435416023843e-09,
"loss": 0.1354,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.2541540373958804,
"learning_rate": 7.995445897566178e-09,
"loss": 0.1392,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.27223613733446245,
"learning_rate": 6.476393170578021e-09,
"loss": 0.1532,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.25900877786323234,
"learning_rate": 5.11720812557348e-09,
"loss": 0.1465,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.2589726414087966,
"learning_rate": 3.917901630563492e-09,
"loss": 0.1453,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.2510455701771674,
"learning_rate": 2.878483275170485e-09,
"loss": 0.1614,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.25056410124278705,
"learning_rate": 1.998961370555108e-09,
"loss": 0.13,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.2524064621383995,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.1481,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.2577195757623055,
"learning_rate": 7.196337656112384e-10,
"loss": 0.1458,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.2611351646140211,
"learning_rate": 3.198382947511292e-10,
"loss": 0.1315,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.25667726105510424,
"learning_rate": 7.995973352770137e-11,
"loss": 0.1188,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.2490225575692216,
"learning_rate": 0.0,
"loss": 0.1316,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 1.30279386186462e+18,
"train_loss": 0.3894373575383835,
"train_runtime": 17437.6756,
"train_samples_per_second": 9.061,
"train_steps_per_second": 0.071
}
],
"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": 1.30279386186462e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}