b2_code_fasttext / trainer_state.json
neginr's picture
End of training
2557836 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 500,
"global_step": 1235,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004048582995951417,
"grad_norm": 6.567379829729217,
"learning_rate": 3.2258064516129035e-07,
"loss": 1.0637,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 6.616549389493247,
"learning_rate": 6.451612903225807e-07,
"loss": 1.0656,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 6.748973240331154,
"learning_rate": 9.67741935483871e-07,
"loss": 1.0751,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 6.366179016425402,
"learning_rate": 1.2903225806451614e-06,
"loss": 1.0401,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 6.128970857561569,
"learning_rate": 1.6129032258064516e-06,
"loss": 1.0519,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 6.0068096518213885,
"learning_rate": 1.935483870967742e-06,
"loss": 1.0386,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 4.7664201706077725,
"learning_rate": 2.2580645161290324e-06,
"loss": 1.0149,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 4.351372971040162,
"learning_rate": 2.580645161290323e-06,
"loss": 1.0103,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 2.9404103715103074,
"learning_rate": 2.903225806451613e-06,
"loss": 0.9405,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 2.649344217647464,
"learning_rate": 3.225806451612903e-06,
"loss": 0.9312,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 2.6007843885680084,
"learning_rate": 3.548387096774194e-06,
"loss": 0.9652,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 3.7892562952917777,
"learning_rate": 3.870967741935484e-06,
"loss": 0.9154,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 4.126829455639829,
"learning_rate": 4.193548387096774e-06,
"loss": 0.9443,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 3.9878912675133975,
"learning_rate": 4.516129032258065e-06,
"loss": 0.9418,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 3.565735518097172,
"learning_rate": 4.838709677419355e-06,
"loss": 0.893,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 2.6718620628400913,
"learning_rate": 5.161290322580646e-06,
"loss": 0.8732,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 2.4014655125421274,
"learning_rate": 5.483870967741935e-06,
"loss": 0.8602,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 1.973550247938392,
"learning_rate": 5.806451612903226e-06,
"loss": 0.8288,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 1.4868639809561965,
"learning_rate": 6.129032258064517e-06,
"loss": 0.8226,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.2766734736581322,
"learning_rate": 6.451612903225806e-06,
"loss": 0.839,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 1.1541663065098036,
"learning_rate": 6.774193548387097e-06,
"loss": 0.8101,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.3633265387728988,
"learning_rate": 7.096774193548388e-06,
"loss": 0.7905,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.1124527806302156,
"learning_rate": 7.4193548387096784e-06,
"loss": 0.7753,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 1.0912658918169746,
"learning_rate": 7.741935483870968e-06,
"loss": 0.7969,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 0.8965958963240429,
"learning_rate": 8.064516129032258e-06,
"loss": 0.7556,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 0.926623802323802,
"learning_rate": 8.387096774193549e-06,
"loss": 0.7632,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 0.87362709067711,
"learning_rate": 8.70967741935484e-06,
"loss": 0.7354,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 0.8056752506482027,
"learning_rate": 9.03225806451613e-06,
"loss": 0.7422,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 0.7720751404145114,
"learning_rate": 9.35483870967742e-06,
"loss": 0.7166,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 0.8868900904391396,
"learning_rate": 9.67741935483871e-06,
"loss": 0.7199,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 0.8685401155806047,
"learning_rate": 1e-05,
"loss": 0.7481,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 0.7891052145696525,
"learning_rate": 1.0322580645161291e-05,
"loss": 0.7103,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 0.665625934325174,
"learning_rate": 1.0645161290322582e-05,
"loss": 0.7403,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 0.8016936671178536,
"learning_rate": 1.096774193548387e-05,
"loss": 0.7109,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 0.7169015556949915,
"learning_rate": 1.1290322580645164e-05,
"loss": 0.6952,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.6739511904656749,
"learning_rate": 1.1612903225806453e-05,
"loss": 0.6813,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.7874325425487126,
"learning_rate": 1.1935483870967743e-05,
"loss": 0.7203,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.6976652245496162,
"learning_rate": 1.2258064516129034e-05,
"loss": 0.6945,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.8694797386571121,
"learning_rate": 1.2580645161290324e-05,
"loss": 0.7096,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 0.69069606901595,
"learning_rate": 1.2903225806451613e-05,
"loss": 0.6952,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.7923509747759255,
"learning_rate": 1.3225806451612903e-05,
"loss": 0.7096,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.6757545534737325,
"learning_rate": 1.3548387096774194e-05,
"loss": 0.6857,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.5214426576231466,
"learning_rate": 1.3870967741935486e-05,
"loss": 0.6948,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.5500514441285806,
"learning_rate": 1.4193548387096776e-05,
"loss": 0.6565,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.7129980374589238,
"learning_rate": 1.4516129032258066e-05,
"loss": 0.6714,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 0.4827921205478151,
"learning_rate": 1.4838709677419357e-05,
"loss": 0.678,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.5859080729349087,
"learning_rate": 1.5161290322580646e-05,
"loss": 0.6852,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.6356447170900633,
"learning_rate": 1.5483870967741936e-05,
"loss": 0.6754,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 0.5386668156065767,
"learning_rate": 1.5806451612903226e-05,
"loss": 0.6676,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 0.6538665847365316,
"learning_rate": 1.6129032258064517e-05,
"loss": 0.6718,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 0.5913652955635377,
"learning_rate": 1.6451612903225807e-05,
"loss": 0.6645,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.5755146307461849,
"learning_rate": 1.6774193548387098e-05,
"loss": 0.6548,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 0.5706481632629784,
"learning_rate": 1.7096774193548388e-05,
"loss": 0.6588,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.6115747340152574,
"learning_rate": 1.741935483870968e-05,
"loss": 0.6491,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.6342402793351102,
"learning_rate": 1.774193548387097e-05,
"loss": 0.6694,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 0.6417745057708247,
"learning_rate": 1.806451612903226e-05,
"loss": 0.6586,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.6234077196004995,
"learning_rate": 1.838709677419355e-05,
"loss": 0.647,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 0.773602013030957,
"learning_rate": 1.870967741935484e-05,
"loss": 0.6628,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 0.6718919840828889,
"learning_rate": 1.903225806451613e-05,
"loss": 0.6595,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 0.7218217981228736,
"learning_rate": 1.935483870967742e-05,
"loss": 0.6565,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 0.6638791467034282,
"learning_rate": 1.967741935483871e-05,
"loss": 0.6486,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 0.6477677347215384,
"learning_rate": 2e-05,
"loss": 0.6437,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 0.6339433915615369,
"learning_rate": 2.0322580645161292e-05,
"loss": 0.6685,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 0.9725383653900108,
"learning_rate": 2.0645161290322582e-05,
"loss": 0.6628,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 0.9277116877242957,
"learning_rate": 2.0967741935483873e-05,
"loss": 0.6407,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 0.6213109381859517,
"learning_rate": 2.1290322580645163e-05,
"loss": 0.6544,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 0.6297738385276114,
"learning_rate": 2.161290322580645e-05,
"loss": 0.6499,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 0.7858972533056193,
"learning_rate": 2.193548387096774e-05,
"loss": 0.642,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 1.13766949974504,
"learning_rate": 2.225806451612903e-05,
"loss": 0.639,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 0.7328646052869894,
"learning_rate": 2.2580645161290328e-05,
"loss": 0.6557,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 0.7118782153124682,
"learning_rate": 2.2903225806451615e-05,
"loss": 0.6432,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 1.1945318516208183,
"learning_rate": 2.3225806451612906e-05,
"loss": 0.6486,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 1.1069670802932936,
"learning_rate": 2.3548387096774196e-05,
"loss": 0.6415,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 0.6723300828585277,
"learning_rate": 2.3870967741935486e-05,
"loss": 0.6493,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 1.1154602831916363,
"learning_rate": 2.4193548387096777e-05,
"loss": 0.6299,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 1.2234174211324371,
"learning_rate": 2.4516129032258067e-05,
"loss": 0.6583,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 0.8569995482802254,
"learning_rate": 2.4838709677419358e-05,
"loss": 0.6315,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 1.2789501252413824,
"learning_rate": 2.5161290322580648e-05,
"loss": 0.6313,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 0.9341232446864609,
"learning_rate": 2.548387096774194e-05,
"loss": 0.6259,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 0.9884125192767299,
"learning_rate": 2.5806451612903226e-05,
"loss": 0.6385,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 1.1732815162776824,
"learning_rate": 2.6129032258064516e-05,
"loss": 0.6327,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 1.0979430280396432,
"learning_rate": 2.6451612903225806e-05,
"loss": 0.6353,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 1.0246956730086025,
"learning_rate": 2.6774193548387097e-05,
"loss": 0.6249,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 1.0306363167508155,
"learning_rate": 2.7096774193548387e-05,
"loss": 0.6301,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 1.5273496566684133,
"learning_rate": 2.7419354838709678e-05,
"loss": 0.6441,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 1.269676920279862,
"learning_rate": 2.774193548387097e-05,
"loss": 0.6333,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 1.8114215860393885,
"learning_rate": 2.8064516129032262e-05,
"loss": 0.6355,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 1.492562976096259,
"learning_rate": 2.8387096774193552e-05,
"loss": 0.6348,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 1.8665776964045384,
"learning_rate": 2.8709677419354843e-05,
"loss": 0.6528,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 1.5213362827796335,
"learning_rate": 2.9032258064516133e-05,
"loss": 0.6266,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 1.9207257072664337,
"learning_rate": 2.9354838709677423e-05,
"loss": 0.6307,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 1.7997879524972542,
"learning_rate": 2.9677419354838714e-05,
"loss": 0.6453,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 1.6086393116459399,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.6375,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 1.3962741621101615,
"learning_rate": 3.032258064516129e-05,
"loss": 0.6261,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 1.2494785276847427,
"learning_rate": 3.0645161290322585e-05,
"loss": 0.6186,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 1.0281021934084318,
"learning_rate": 3.096774193548387e-05,
"loss": 0.613,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 1.2279031043746442,
"learning_rate": 3.1290322580645166e-05,
"loss": 0.6304,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 0.8255932036640569,
"learning_rate": 3.161290322580645e-05,
"loss": 0.6272,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 0.9403012641858256,
"learning_rate": 3.193548387096774e-05,
"loss": 0.6154,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 1.0969184454462626,
"learning_rate": 3.2258064516129034e-05,
"loss": 0.6457,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 1.194447420412311,
"learning_rate": 3.258064516129033e-05,
"loss": 0.632,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 0.8796748596080799,
"learning_rate": 3.2903225806451614e-05,
"loss": 0.6194,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 0.893265362361624,
"learning_rate": 3.322580645161291e-05,
"loss": 0.6178,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 1.1785611873485426,
"learning_rate": 3.3548387096774195e-05,
"loss": 0.623,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 0.8448112314265869,
"learning_rate": 3.387096774193549e-05,
"loss": 0.6192,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 0.7932945913256584,
"learning_rate": 3.4193548387096776e-05,
"loss": 0.6114,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 0.7284363976096005,
"learning_rate": 3.451612903225807e-05,
"loss": 0.6262,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 0.9783825649141459,
"learning_rate": 3.483870967741936e-05,
"loss": 0.6111,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 1.1751609546113029,
"learning_rate": 3.516129032258065e-05,
"loss": 0.6353,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 0.8063172486716059,
"learning_rate": 3.548387096774194e-05,
"loss": 0.6136,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 1.0712968296385812,
"learning_rate": 3.5806451612903225e-05,
"loss": 0.619,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 1.2661117452712727,
"learning_rate": 3.612903225806452e-05,
"loss": 0.6177,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 0.8335145643810444,
"learning_rate": 3.6451612903225805e-05,
"loss": 0.622,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 1.046161239576529,
"learning_rate": 3.67741935483871e-05,
"loss": 0.6242,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 1.0132750432883362,
"learning_rate": 3.7096774193548386e-05,
"loss": 0.6089,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 0.9103296297760488,
"learning_rate": 3.741935483870968e-05,
"loss": 0.625,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 0.9364513017950385,
"learning_rate": 3.7741935483870974e-05,
"loss": 0.6076,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 1.4232135752763118,
"learning_rate": 3.806451612903226e-05,
"loss": 0.619,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 0.7679031859598143,
"learning_rate": 3.8387096774193555e-05,
"loss": 0.6242,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 1.2912533960613235,
"learning_rate": 3.870967741935484e-05,
"loss": 0.6263,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 0.8970342606959881,
"learning_rate": 3.9032258064516135e-05,
"loss": 0.6199,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 1.3853584733432007,
"learning_rate": 3.935483870967742e-05,
"loss": 0.6103,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 1.010748204402799,
"learning_rate": 3.9677419354838716e-05,
"loss": 0.6186,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 1.3756086067717637,
"learning_rate": 4e-05,
"loss": 0.6053,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 0.8085796282947588,
"learning_rate": 3.9999920040266474e-05,
"loss": 0.6192,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 1.1301182681652477,
"learning_rate": 3.999968016170525e-05,
"loss": 0.6286,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 1.187758455575281,
"learning_rate": 3.9999280366234394e-05,
"loss": 0.6116,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 0.9176869156555297,
"learning_rate": 3.999872065705065e-05,
"loss": 0.6074,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 1.0196250561736895,
"learning_rate": 3.999800103862945e-05,
"loss": 0.6048,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 1.081023414044582,
"learning_rate": 3.9997121516724834e-05,
"loss": 0.6112,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 0.8175128841513086,
"learning_rate": 3.999608209836944e-05,
"loss": 0.604,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 0.9080557096039521,
"learning_rate": 3.999488279187443e-05,
"loss": 0.6145,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 0.896099630159905,
"learning_rate": 3.9993523606829425e-05,
"loss": 0.611,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 0.654759711983695,
"learning_rate": 3.9992004554102435e-05,
"loss": 0.6159,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 0.6340958953855346,
"learning_rate": 3.999032564583976e-05,
"loss": 0.6216,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 0.5372248725374853,
"learning_rate": 3.998848689546591e-05,
"loss": 0.6025,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 0.6754946892804272,
"learning_rate": 3.9986488317683496e-05,
"loss": 0.6162,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 0.7772108041944729,
"learning_rate": 3.9984329928473067e-05,
"loss": 0.6149,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 0.7114602438407199,
"learning_rate": 3.998201174509306e-05,
"loss": 0.6135,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 0.6355621498174238,
"learning_rate": 3.997953378607961e-05,
"loss": 0.6083,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 0.6582160326896682,
"learning_rate": 3.9976896071246403e-05,
"loss": 0.5955,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 0.6179185327667744,
"learning_rate": 3.997409862168454e-05,
"loss": 0.6191,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 0.7724742764333579,
"learning_rate": 3.997114145976237e-05,
"loss": 0.5961,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 0.6646362582943439,
"learning_rate": 3.9968024609125245e-05,
"loss": 0.5982,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 0.6249393272741437,
"learning_rate": 3.996474809469545e-05,
"loss": 0.5886,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 0.5850674340556162,
"learning_rate": 3.996131194267189e-05,
"loss": 0.6134,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 0.7063267926213389,
"learning_rate": 3.995771618052995e-05,
"loss": 0.5936,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 0.7073461476734361,
"learning_rate": 3.9953960837021256e-05,
"loss": 0.6214,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 0.6347227218066128,
"learning_rate": 3.995004594217343e-05,
"loss": 0.5927,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 0.616861564053352,
"learning_rate": 3.994597152728987e-05,
"loss": 0.6002,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 0.662555002011711,
"learning_rate": 3.994173762494947e-05,
"loss": 0.6039,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 0.5823914296715781,
"learning_rate": 3.9937344269006424e-05,
"loss": 0.6091,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 0.6946957471286906,
"learning_rate": 3.993279149458988e-05,
"loss": 0.6005,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 0.6900822770586325,
"learning_rate": 3.9928079338103695e-05,
"loss": 0.6015,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 0.5484848687514599,
"learning_rate": 3.992320783722616e-05,
"loss": 0.6,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.8165291107313926,
"learning_rate": 3.991817703090965e-05,
"loss": 0.5968,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 0.8481835309107851,
"learning_rate": 3.991298695938038e-05,
"loss": 0.6064,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 0.7244895287056498,
"learning_rate": 3.9907637664138e-05,
"loss": 0.6249,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 0.8225751250410662,
"learning_rate": 3.990212918795534e-05,
"loss": 0.6292,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 0.6006560208357979,
"learning_rate": 3.9896461574878036e-05,
"loss": 0.5938,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 0.7940529465718681,
"learning_rate": 3.989063487022416e-05,
"loss": 0.5903,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 0.8489188975230876,
"learning_rate": 3.9884649120583903e-05,
"loss": 0.5919,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 0.8678557286067451,
"learning_rate": 3.987850437381915e-05,
"loss": 0.6081,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 0.6565860664464264,
"learning_rate": 3.987220067906313e-05,
"loss": 0.6159,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 0.7161317978957609,
"learning_rate": 3.986573808672002e-05,
"loss": 0.6214,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 0.7657817029016185,
"learning_rate": 3.9859116648464534e-05,
"loss": 0.6052,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 0.7449046437477294,
"learning_rate": 3.985233641724152e-05,
"loss": 0.6009,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 0.9139395435482843,
"learning_rate": 3.984539744726554e-05,
"loss": 0.6123,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 0.8200347245510937,
"learning_rate": 3.983829979402038e-05,
"loss": 0.5964,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 0.7011506910009733,
"learning_rate": 3.983104351425871e-05,
"loss": 0.5894,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 0.827490181429491,
"learning_rate": 3.982362866600154e-05,
"loss": 0.5935,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 0.7124258682820735,
"learning_rate": 3.9816055308537806e-05,
"loss": 0.5833,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 0.5377976394412157,
"learning_rate": 3.980832350242388e-05,
"loss": 0.5944,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 0.5000236481976327,
"learning_rate": 3.980043330948306e-05,
"loss": 0.5952,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 0.5873558795440958,
"learning_rate": 3.9792384792805124e-05,
"loss": 0.6052,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 0.6145517365931847,
"learning_rate": 3.978417801674579e-05,
"loss": 0.5825,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 0.5357182269501898,
"learning_rate": 3.9775813046926244e-05,
"loss": 0.5846,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 0.6501210044303178,
"learning_rate": 3.976728995023254e-05,
"loss": 0.5825,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 0.6490347604920063,
"learning_rate": 3.975860879481514e-05,
"loss": 0.5972,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 0.607705646478506,
"learning_rate": 3.974976965008833e-05,
"loss": 0.6067,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 0.7007425217541702,
"learning_rate": 3.9740772586729674e-05,
"loss": 0.6191,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.7117392611954662,
"learning_rate": 3.973161767667946e-05,
"loss": 0.5897,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 0.5736878766938066,
"learning_rate": 3.972230499314009e-05,
"loss": 0.5935,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 0.6097697660144247,
"learning_rate": 3.9712834610575545e-05,
"loss": 0.6005,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 0.6212813545409719,
"learning_rate": 3.9703206604710746e-05,
"loss": 0.5907,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 0.5796266007542467,
"learning_rate": 3.969342105253098e-05,
"loss": 0.5976,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 0.6356278524753456,
"learning_rate": 3.9683478032281246e-05,
"loss": 0.595,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 0.667896386685234,
"learning_rate": 3.967337762346568e-05,
"loss": 0.5927,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 0.4880173702891493,
"learning_rate": 3.9663119906846885e-05,
"loss": 0.6021,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.631988662946068,
"learning_rate": 3.965270496444528e-05,
"loss": 0.597,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 0.5664813115831782,
"learning_rate": 3.9642132879538465e-05,
"loss": 0.6002,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 0.46933194148891366,
"learning_rate": 3.9631403736660566e-05,
"loss": 0.588,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 0.5448142723088519,
"learning_rate": 3.962051762160151e-05,
"loss": 0.5945,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 0.6574535894294661,
"learning_rate": 3.960947462140638e-05,
"loss": 0.587,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 0.565375706366302,
"learning_rate": 3.959827482437472e-05,
"loss": 0.5974,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 0.41922874039808833,
"learning_rate": 3.958691832005981e-05,
"loss": 0.5912,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 0.5141715084467168,
"learning_rate": 3.9575405199267944e-05,
"loss": 0.5989,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 0.4663952548199157,
"learning_rate": 3.956373555405774e-05,
"loss": 0.5842,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 0.40416580728625784,
"learning_rate": 3.955190947773937e-05,
"loss": 0.5828,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 0.6124633301305779,
"learning_rate": 3.9539927064873824e-05,
"loss": 0.5788,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 0.550449549747645,
"learning_rate": 3.952778841127214e-05,
"loss": 0.5806,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 0.4953018559244568,
"learning_rate": 3.95154936139947e-05,
"loss": 0.5688,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 0.4679025038751024,
"learning_rate": 3.950304277135034e-05,
"loss": 0.59,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 0.575998979937518,
"learning_rate": 3.94904359828957e-05,
"loss": 0.5848,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 0.5023068502474253,
"learning_rate": 3.94776733494343e-05,
"loss": 0.5828,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 0.43698568332170074,
"learning_rate": 3.9464754973015826e-05,
"loss": 0.5838,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 0.4653721183690986,
"learning_rate": 3.9451680956935275e-05,
"loss": 0.5774,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.49951845594834116,
"learning_rate": 3.943845140573213e-05,
"loss": 0.5907,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.5414999948095397,
"learning_rate": 3.942506642518952e-05,
"loss": 0.5948,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 0.595516265615837,
"learning_rate": 3.9411526122333396e-05,
"loss": 0.603,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 0.46649690974851377,
"learning_rate": 3.939783060543167e-05,
"loss": 0.5846,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 0.5172470480808606,
"learning_rate": 3.938397998399333e-05,
"loss": 0.5784,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.5118902364756727,
"learning_rate": 3.936997436876756e-05,
"loss": 0.572,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.5558647757076058,
"learning_rate": 3.935581387174289e-05,
"loss": 0.5766,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 0.5559404536676781,
"learning_rate": 3.9341498606146295e-05,
"loss": 0.5774,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.50649302112337,
"learning_rate": 3.932702868644224e-05,
"loss": 0.5705,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.5355479766437198,
"learning_rate": 3.931240422833181e-05,
"loss": 0.5626,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.48353021502822474,
"learning_rate": 3.9297625348751806e-05,
"loss": 0.5707,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 0.7550944886426096,
"learning_rate": 3.9282692165873744e-05,
"loss": 0.5793,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 0.6516975727547778,
"learning_rate": 3.9267604799102945e-05,
"loss": 0.5889,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.5915037733503842,
"learning_rate": 3.9252363369077615e-05,
"loss": 0.5743,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 0.526598413329693,
"learning_rate": 3.9236967997667796e-05,
"loss": 0.5661,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 0.669437927519776,
"learning_rate": 3.9221418807974496e-05,
"loss": 0.5833,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 0.6843541850394781,
"learning_rate": 3.92057159243286e-05,
"loss": 0.5806,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 0.6475919333714825,
"learning_rate": 3.9189859472289956e-05,
"loss": 0.6011,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 0.5240280782568577,
"learning_rate": 3.9173849578646324e-05,
"loss": 0.5818,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 0.4468271893875613,
"learning_rate": 3.9157686371412395e-05,
"loss": 0.5879,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.5435953553278234,
"learning_rate": 3.914136997982874e-05,
"loss": 0.5695,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 0.5222236107288969,
"learning_rate": 3.912490053436079e-05,
"loss": 0.583,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 0.48247545359366334,
"learning_rate": 3.9108278166697805e-05,
"loss": 0.5873,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.5159039768220861,
"learning_rate": 3.909150300975178e-05,
"loss": 0.5702,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 0.563397833491016,
"learning_rate": 3.907457519765642e-05,
"loss": 0.5672,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.47152572137261134,
"learning_rate": 3.905749486576607e-05,
"loss": 0.5589,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.4960766302257666,
"learning_rate": 3.90402621506546e-05,
"loss": 0.5768,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 0.4555399227218288,
"learning_rate": 3.902287719011435e-05,
"loss": 0.5782,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.5715665320590803,
"learning_rate": 3.900534012315499e-05,
"loss": 0.5807,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 0.5407873842927828,
"learning_rate": 3.8987651090002454e-05,
"loss": 0.5702,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 0.6176617186407979,
"learning_rate": 3.896981023209777e-05,
"loss": 0.5706,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 0.5845018642940558,
"learning_rate": 3.8951817692095966e-05,
"loss": 0.5737,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 0.4620516926014223,
"learning_rate": 3.893367361386491e-05,
"loss": 0.5847,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 0.6078843188127038,
"learning_rate": 3.891537814248417e-05,
"loss": 0.5672,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 0.7669890845928758,
"learning_rate": 3.889693142424384e-05,
"loss": 0.5834,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 0.7799267999637822,
"learning_rate": 3.8878333606643405e-05,
"loss": 0.5863,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 0.6890670129078139,
"learning_rate": 3.885958483839049e-05,
"loss": 0.5756,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 0.6123935315623268,
"learning_rate": 3.8840685269399786e-05,
"loss": 0.5822,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 0.8170319232021415,
"learning_rate": 3.882163505079171e-05,
"loss": 0.5946,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 0.6388662886258882,
"learning_rate": 3.880243433489132e-05,
"loss": 0.5793,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 0.5677685518976022,
"learning_rate": 3.878308327522702e-05,
"loss": 0.5402,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 0.6147289628057021,
"learning_rate": 3.876358202652936e-05,
"loss": 0.5398,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 0.6783790513385738,
"learning_rate": 3.8743930744729835e-05,
"loss": 0.5389,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.5990803873404211,
"learning_rate": 3.8724129586959535e-05,
"loss": 0.5265,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.5218317312476443,
"learning_rate": 3.870417871154801e-05,
"loss": 0.5325,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 0.5316975343949741,
"learning_rate": 3.868407827802192e-05,
"loss": 0.515,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.6840070067765623,
"learning_rate": 3.8663828447103804e-05,
"loss": 0.5349,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.5472721809945392,
"learning_rate": 3.8643429380710765e-05,
"loss": 0.5161,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.505001993244571,
"learning_rate": 3.862288124195319e-05,
"loss": 0.5296,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.6547316331668308,
"learning_rate": 3.8602184195133464e-05,
"loss": 0.5265,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 0.5495571207819933,
"learning_rate": 3.8581338405744604e-05,
"loss": 0.5106,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 0.4433077699184542,
"learning_rate": 3.8560344040468996e-05,
"loss": 0.5247,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.6120733654552979,
"learning_rate": 3.853920126717703e-05,
"loss": 0.5171,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 0.5583029571922499,
"learning_rate": 3.851791025492574e-05,
"loss": 0.5127,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.4881528138825233,
"learning_rate": 3.8496471173957515e-05,
"loss": 0.5269,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.5624657061458714,
"learning_rate": 3.847488419569866e-05,
"loss": 0.5374,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.5519526526095401,
"learning_rate": 3.8453149492758077e-05,
"loss": 0.5187,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 0.5181240630535938,
"learning_rate": 3.8431267238925886e-05,
"loss": 0.524,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.5282895537232197,
"learning_rate": 3.840923760917199e-05,
"loss": 0.5265,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 0.5624941513758152,
"learning_rate": 3.838706077964473e-05,
"loss": 0.5302,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 0.5442562438184323,
"learning_rate": 3.836473692766944e-05,
"loss": 0.5314,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 0.506694345495052,
"learning_rate": 3.834226623174705e-05,
"loss": 0.5313,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.4559004590725417,
"learning_rate": 3.831964887155264e-05,
"loss": 0.5184,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.505361507643699,
"learning_rate": 3.829688502793403e-05,
"loss": 0.5421,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.5201041132786677,
"learning_rate": 3.827397488291029e-05,
"loss": 0.5273,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.4348975817366091,
"learning_rate": 3.8250918619670343e-05,
"loss": 0.5165,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.5249593759888069,
"learning_rate": 3.822771642257145e-05,
"loss": 0.5165,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.538567134901824,
"learning_rate": 3.820436847713776e-05,
"loss": 0.5142,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 0.460908577006086,
"learning_rate": 3.8180874970058826e-05,
"loss": 0.5215,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.5767321687844561,
"learning_rate": 3.8157236089188103e-05,
"loss": 0.5165,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.5444689368353063,
"learning_rate": 3.813345202354145e-05,
"loss": 0.5416,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 0.48934305104364745,
"learning_rate": 3.810952296329563e-05,
"loss": 0.537,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.5773925995615876,
"learning_rate": 3.808544909978676e-05,
"loss": 0.5205,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.5933270973685919,
"learning_rate": 3.806123062550882e-05,
"loss": 0.5412,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.5060828719312787,
"learning_rate": 3.803686773411208e-05,
"loss": 0.5366,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.5703965818511507,
"learning_rate": 3.801236062040158e-05,
"loss": 0.5162,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 0.4788412975625315,
"learning_rate": 3.798770948033553e-05,
"loss": 0.5204,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.5932296211936338,
"learning_rate": 3.796291451102381e-05,
"loss": 0.5205,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.5644380544849608,
"learning_rate": 3.793797591072631e-05,
"loss": 0.52,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.539106476858554,
"learning_rate": 3.791289387885144e-05,
"loss": 0.5158,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.6159726793246251,
"learning_rate": 3.788766861595443e-05,
"loss": 0.5068,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 0.4790524353140686,
"learning_rate": 3.7862300323735835e-05,
"loss": 0.5296,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 0.5203005976876327,
"learning_rate": 3.783678920503983e-05,
"loss": 0.5218,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 0.58146411260701,
"learning_rate": 3.781113546385265e-05,
"loss": 0.5182,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.4770199604878711,
"learning_rate": 3.7785339305300916e-05,
"loss": 0.5356,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.43152573136918515,
"learning_rate": 3.7759400935650025e-05,
"loss": 0.54,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.6368259736646043,
"learning_rate": 3.77333205623025e-05,
"loss": 0.5375,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.5351333097430191,
"learning_rate": 3.7707098393796295e-05,
"loss": 0.5307,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.476427662784345,
"learning_rate": 3.7680734639803185e-05,
"loss": 0.5256,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 0.5153062730969878,
"learning_rate": 3.765422951112704e-05,
"loss": 0.5164,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 0.501219367614121,
"learning_rate": 3.762758321970216e-05,
"loss": 0.52,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.8149003758026235,
"learning_rate": 3.7600795978591584e-05,
"loss": 0.5017,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.5712667631922967,
"learning_rate": 3.757386800198538e-05,
"loss": 0.5391,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.669020213928777,
"learning_rate": 3.7546799505198925e-05,
"loss": 0.5404,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.7889070900387266,
"learning_rate": 3.7519590704671197e-05,
"loss": 0.5241,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.6092674973213665,
"learning_rate": 3.749224181796305e-05,
"loss": 0.5318,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.7857949485377873,
"learning_rate": 3.746475306375545e-05,
"loss": 0.5271,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.6437029505769302,
"learning_rate": 3.743712466184774e-05,
"loss": 0.5202,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.8116921025651157,
"learning_rate": 3.7409356833155885e-05,
"loss": 0.5202,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.488508276198876,
"learning_rate": 3.73814497997107e-05,
"loss": 0.5267,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.5650093869957042,
"learning_rate": 3.7353403784656095e-05,
"loss": 0.519,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.6030734017232058,
"learning_rate": 3.732521901224724e-05,
"loss": 0.5341,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.456503932793933,
"learning_rate": 3.7296895707848844e-05,
"loss": 0.5002,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.5242673822348783,
"learning_rate": 3.7268434097933275e-05,
"loss": 0.5312,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.6043849182055482,
"learning_rate": 3.723983441007882e-05,
"loss": 0.5047,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.40905190032901345,
"learning_rate": 3.721109687296781e-05,
"loss": 0.5174,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.41797810534437,
"learning_rate": 3.718222171638484e-05,
"loss": 0.5212,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 0.5583686854159029,
"learning_rate": 3.715320917121488e-05,
"loss": 0.5261,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.442706675370137,
"learning_rate": 3.712405946944149e-05,
"loss": 0.5198,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.4519531824979171,
"learning_rate": 3.709477284414488e-05,
"loss": 0.5123,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.49375384913140924,
"learning_rate": 3.7065349529500136e-05,
"loss": 0.5326,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 0.5311026605973748,
"learning_rate": 3.7035789760775305e-05,
"loss": 0.5197,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.3819999441785956,
"learning_rate": 3.70060937743295e-05,
"loss": 0.5095,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.5475897494651518,
"learning_rate": 3.6976261807611045e-05,
"loss": 0.5313,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.5133319181831394,
"learning_rate": 3.694629409915555e-05,
"loss": 0.5489,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.461259848023181,
"learning_rate": 3.691619088858401e-05,
"loss": 0.5085,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.4532647587647387,
"learning_rate": 3.68859524166009e-05,
"loss": 0.5127,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.558335805142642,
"learning_rate": 3.6855578924992226e-05,
"loss": 0.509,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.4866703652366528,
"learning_rate": 3.6825070656623626e-05,
"loss": 0.5037,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.4498702305468001,
"learning_rate": 3.67944278554384e-05,
"loss": 0.5369,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 0.39466364866146353,
"learning_rate": 3.676365076645557e-05,
"loss": 0.5374,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.40446530697347355,
"learning_rate": 3.673273963576791e-05,
"loss": 0.5199,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.4132200761612057,
"learning_rate": 3.670169471054001e-05,
"loss": 0.5228,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.4414386199779288,
"learning_rate": 3.6670516239006254e-05,
"loss": 0.525,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.4898051279703619,
"learning_rate": 3.663920447046888e-05,
"loss": 0.5319,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.43614320078653745,
"learning_rate": 3.6607759655295954e-05,
"loss": 0.5199,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.47368661690306585,
"learning_rate": 3.657618204491937e-05,
"loss": 0.517,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.5035318544728339,
"learning_rate": 3.6544471891832875e-05,
"loss": 0.5152,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.37798493271177164,
"learning_rate": 3.6512629449589983e-05,
"loss": 0.5383,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.5160626888204923,
"learning_rate": 3.6480654972802045e-05,
"loss": 0.5444,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.42022799695247676,
"learning_rate": 3.644854871713611e-05,
"loss": 0.5251,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.5016151695966443,
"learning_rate": 3.6416310939312924e-05,
"loss": 0.5296,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.4206362554043026,
"learning_rate": 3.638394189710493e-05,
"loss": 0.5087,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.4755739659433779,
"learning_rate": 3.635144184933412e-05,
"loss": 0.5154,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.39956234282221886,
"learning_rate": 3.6318811055869994e-05,
"loss": 0.5065,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.47718565460964824,
"learning_rate": 3.628604977762752e-05,
"loss": 0.5169,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.44234043730852934,
"learning_rate": 3.6253158276565004e-05,
"loss": 0.5138,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.4363847244730975,
"learning_rate": 3.6220136815682016e-05,
"loss": 0.5193,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.4046680480926559,
"learning_rate": 3.618698565901727e-05,
"loss": 0.5344,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.3939295327306027,
"learning_rate": 3.615370507164654e-05,
"loss": 0.5317,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.38456831046474665,
"learning_rate": 3.61202953196805e-05,
"loss": 0.5239,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 0.43407017715832774,
"learning_rate": 3.6086756670262654e-05,
"loss": 0.5453,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.39904816694136686,
"learning_rate": 3.6053089391567135e-05,
"loss": 0.5412,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.41487523566999807,
"learning_rate": 3.601929375279662e-05,
"loss": 0.5247,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.4365796333135234,
"learning_rate": 3.598537002418012e-05,
"loss": 0.5237,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.5542918589324212,
"learning_rate": 3.5951318476970885e-05,
"loss": 0.5283,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.40449871776695,
"learning_rate": 3.591713938344416e-05,
"loss": 0.5052,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.4653498498322578,
"learning_rate": 3.588283301689507e-05,
"loss": 0.5324,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 0.49562317961711855,
"learning_rate": 3.5848399651636424e-05,
"loss": 0.5073,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.40186532307630907,
"learning_rate": 3.5813839562996473e-05,
"loss": 0.5181,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.4728880981566691,
"learning_rate": 3.5779153027316775e-05,
"loss": 0.5332,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.3646837591708386,
"learning_rate": 3.574434032194994e-05,
"loss": 0.513,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.4003534060844194,
"learning_rate": 3.5709401725257436e-05,
"loss": 0.5224,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.3952508008687929,
"learning_rate": 3.5674337516607346e-05,
"loss": 0.5156,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.3661320181435376,
"learning_rate": 3.563914797637216e-05,
"loss": 0.5003,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.4256258490164142,
"learning_rate": 3.560383338592649e-05,
"loss": 0.5272,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.3662878311676565,
"learning_rate": 3.556839402764487e-05,
"loss": 0.5298,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.3854078816846968,
"learning_rate": 3.553283018489946e-05,
"loss": 0.5327,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.4702546413293467,
"learning_rate": 3.54971421420578e-05,
"loss": 0.5299,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.38201013929222594,
"learning_rate": 3.5461330184480524e-05,
"loss": 0.5125,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 0.39324739033745554,
"learning_rate": 3.54253945985191e-05,
"loss": 0.5311,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.33559176972390903,
"learning_rate": 3.5389335671513524e-05,
"loss": 0.5196,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.3781299058578177,
"learning_rate": 3.5353153691789986e-05,
"loss": 0.5265,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.37373910402217897,
"learning_rate": 3.531684894865866e-05,
"loss": 0.512,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.3894264781613061,
"learning_rate": 3.528042173241128e-05,
"loss": 0.5175,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.37012187761282855,
"learning_rate": 3.524387233431892e-05,
"loss": 0.4979,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.4001476476543467,
"learning_rate": 3.520720104662958e-05,
"loss": 0.5198,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.40162095343474785,
"learning_rate": 3.5170408162565904e-05,
"loss": 0.5245,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.38689779451333056,
"learning_rate": 3.513349397632281e-05,
"loss": 0.5373,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.33728469429076136,
"learning_rate": 3.5096458783065145e-05,
"loss": 0.5229,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.4608948870658328,
"learning_rate": 3.505930287892533e-05,
"loss": 0.5116,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.420324311730209,
"learning_rate": 3.502202656100099e-05,
"loss": 0.5253,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.46951546916789105,
"learning_rate": 3.498463012735256e-05,
"loss": 0.5283,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.504912973613517,
"learning_rate": 3.494711387700094e-05,
"loss": 0.5231,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.49898906886619787,
"learning_rate": 3.490947810992505e-05,
"loss": 0.5206,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 0.48953100429478535,
"learning_rate": 3.48717231270595e-05,
"loss": 0.5296,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.408670388038461,
"learning_rate": 3.483384923029211e-05,
"loss": 0.5272,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.3881080089803089,
"learning_rate": 3.479585672246156e-05,
"loss": 0.5002,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.4057460309372445,
"learning_rate": 3.4757745907354924e-05,
"loss": 0.522,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.4456066961526123,
"learning_rate": 3.471951708970527e-05,
"loss": 0.5106,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.3872550835375557,
"learning_rate": 3.468117057518921e-05,
"loss": 0.5244,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.440779344072978,
"learning_rate": 3.4642706670424426e-05,
"loss": 0.5066,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.41241343342797643,
"learning_rate": 3.460412568296731e-05,
"loss": 0.5201,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.41087340745132256,
"learning_rate": 3.456542792131039e-05,
"loss": 0.5047,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.47496890846944606,
"learning_rate": 3.4526613694879936e-05,
"loss": 0.5321,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.46161509031536113,
"learning_rate": 3.448768331403348e-05,
"loss": 0.5283,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.40741314938635287,
"learning_rate": 3.44486370900573e-05,
"loss": 0.5378,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.4765811070455373,
"learning_rate": 3.440947533516396e-05,
"loss": 0.5265,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.559260155952037,
"learning_rate": 3.4370198362489816e-05,
"loss": 0.5334,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 0.4331793671344221,
"learning_rate": 3.43308064860925e-05,
"loss": 0.5038,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.3669418136842725,
"learning_rate": 3.4291300020948393e-05,
"loss": 0.523,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.44832461681265606,
"learning_rate": 3.425167928295015e-05,
"loss": 0.5069,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.3575876027013466,
"learning_rate": 3.421194458890411e-05,
"loss": 0.5239,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.38985420886795996,
"learning_rate": 3.4172096256527865e-05,
"loss": 0.5278,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.40616325887276017,
"learning_rate": 3.4132134604447595e-05,
"loss": 0.5286,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.42512788341570645,
"learning_rate": 3.409205995219561e-05,
"loss": 0.5174,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.42653438921699827,
"learning_rate": 3.4051872620207765e-05,
"loss": 0.5145,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.41379222721244324,
"learning_rate": 3.40115729298209e-05,
"loss": 0.5049,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.43904282879939216,
"learning_rate": 3.397116120327025e-05,
"loss": 0.504,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.37780827423058627,
"learning_rate": 3.3930637763686916e-05,
"loss": 0.5236,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.4275407043972215,
"learning_rate": 3.389000293509524e-05,
"loss": 0.515,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.3658601035096516,
"learning_rate": 3.3849257042410225e-05,
"loss": 0.524,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.40348041250456645,
"learning_rate": 3.380840041143494e-05,
"loss": 0.5147,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.4395918798343256,
"learning_rate": 3.376743336885792e-05,
"loss": 0.5284,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.3938208443443045,
"learning_rate": 3.372635624225055e-05,
"loss": 0.5286,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.41179079230478566,
"learning_rate": 3.3685169360064445e-05,
"loss": 0.506,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.3996932704057503,
"learning_rate": 3.3643873051628804e-05,
"loss": 0.5037,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.4164843360173594,
"learning_rate": 3.3602467647147814e-05,
"loss": 0.5283,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.3958531701112546,
"learning_rate": 3.356095347769799e-05,
"loss": 0.5315,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.4146899125066927,
"learning_rate": 3.351933087522552e-05,
"loss": 0.5258,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.4939193198933418,
"learning_rate": 3.3477600172543626e-05,
"loss": 0.5155,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.40343113807183006,
"learning_rate": 3.3435761703329894e-05,
"loss": 0.5312,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.3826272167230999,
"learning_rate": 3.3393815802123606e-05,
"loss": 0.5242,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.45866196437925344,
"learning_rate": 3.335176280432308e-05,
"loss": 0.4992,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.37615421242515773,
"learning_rate": 3.330960304618295e-05,
"loss": 0.5028,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.392925680586739,
"learning_rate": 3.326733686481153e-05,
"loss": 0.5093,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.41564062738966606,
"learning_rate": 3.3224964598168084e-05,
"loss": 0.5248,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.3571253763054853,
"learning_rate": 3.318248658506012e-05,
"loss": 0.5223,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.39121146781199767,
"learning_rate": 3.313990316514069e-05,
"loss": 0.5365,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.3291193544780531,
"learning_rate": 3.309721467890571e-05,
"loss": 0.521,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.35230869835370127,
"learning_rate": 3.305442146769115e-05,
"loss": 0.5188,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.38478183193349763,
"learning_rate": 3.30115238736704e-05,
"loss": 0.5227,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.3443390589994284,
"learning_rate": 3.296852223985148e-05,
"loss": 0.5213,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 0.3443221448918253,
"learning_rate": 3.2925416910074315e-05,
"loss": 0.5102,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.3665464119375583,
"learning_rate": 3.288220822900796e-05,
"loss": 0.5282,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.37118641080432396,
"learning_rate": 3.283889654214788e-05,
"loss": 0.5156,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.41681712984494995,
"learning_rate": 3.279548219581317e-05,
"loss": 0.5289,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.39101287139908025,
"learning_rate": 3.275196553714379e-05,
"loss": 0.5256,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.4060068208022471,
"learning_rate": 3.270834691409778e-05,
"loss": 0.5251,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.4125579175785946,
"learning_rate": 3.266462667544849e-05,
"loss": 0.4962,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.4106429705260203,
"learning_rate": 3.262080517078178e-05,
"loss": 0.5277,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.4010230763530757,
"learning_rate": 3.257688275049323e-05,
"loss": 0.5179,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.44896828178634696,
"learning_rate": 3.253285976578535e-05,
"loss": 0.5203,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.46984640034051434,
"learning_rate": 3.2488736568664756e-05,
"loss": 0.5184,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.4040128948658912,
"learning_rate": 3.244451351193935e-05,
"loss": 0.5241,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 0.6148313120046134,
"learning_rate": 3.240019094921551e-05,
"loss": 0.5268,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.4632880571416557,
"learning_rate": 3.235576923489527e-05,
"loss": 0.4992,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.4762802917629513,
"learning_rate": 3.2311248724173476e-05,
"loss": 0.5156,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.5973231617952175,
"learning_rate": 3.226662977303494e-05,
"loss": 0.5232,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.3545949927850837,
"learning_rate": 3.2221912738251626e-05,
"loss": 0.515,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.5571308541829948,
"learning_rate": 3.2177097977379736e-05,
"loss": 0.5205,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.40122664883841613,
"learning_rate": 3.21321858487569e-05,
"loss": 0.5041,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.5103772249648183,
"learning_rate": 3.208717671149931e-05,
"loss": 0.5115,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.39230967221851504,
"learning_rate": 3.204207092549882e-05,
"loss": 0.5083,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 0.48633782340036813,
"learning_rate": 3.19968688514201e-05,
"loss": 0.5228,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.47222418894770546,
"learning_rate": 3.195157085069774e-05,
"loss": 0.5268,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.4033361406609526,
"learning_rate": 3.190617728553332e-05,
"loss": 0.5106,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.5341017198235001,
"learning_rate": 3.1860688518892606e-05,
"loss": 0.5172,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.36757098526871357,
"learning_rate": 3.181510491450254e-05,
"loss": 0.5361,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.5055475822632897,
"learning_rate": 3.176942683684842e-05,
"loss": 0.5324,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.3317167315405086,
"learning_rate": 3.1723654651170934e-05,
"loss": 0.5175,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.5063281005438391,
"learning_rate": 3.167778872346327e-05,
"loss": 0.5019,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.4492745102689018,
"learning_rate": 3.163182942046815e-05,
"loss": 0.5371,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.3937967880921271,
"learning_rate": 3.158577710967494e-05,
"loss": 0.5127,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.45534822852438994,
"learning_rate": 3.153963215931669e-05,
"loss": 0.5208,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.4985800141466128,
"learning_rate": 3.14933949383672e-05,
"loss": 0.5291,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.42957708046475473,
"learning_rate": 3.144706581653805e-05,
"loss": 0.5202,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.5420933193340414,
"learning_rate": 3.140064516427566e-05,
"loss": 0.5185,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.416687869859613,
"learning_rate": 3.135413335275833e-05,
"loss": 0.5138,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.541562983752296,
"learning_rate": 3.130753075389327e-05,
"loss": 0.5346,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.5263663780833188,
"learning_rate": 3.1260837740313625e-05,
"loss": 0.5257,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.3899246977462424,
"learning_rate": 3.121405468537547e-05,
"loss": 0.5132,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.3956462874460845,
"learning_rate": 3.116718196315488e-05,
"loss": 0.5141,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.4018400129910295,
"learning_rate": 3.1120219948444894e-05,
"loss": 0.5374,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.40431628556752186,
"learning_rate": 3.107316901675252e-05,
"loss": 0.5316,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.39001674739142095,
"learning_rate": 3.1026029544295745e-05,
"loss": 0.5284,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.3962189772241596,
"learning_rate": 3.097880190800057e-05,
"loss": 0.5243,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.36478027186531253,
"learning_rate": 3.093148648549788e-05,
"loss": 0.5226,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.5180449959409986,
"learning_rate": 3.088408365512055e-05,
"loss": 0.5274,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.43066653642961816,
"learning_rate": 3.083659379590034e-05,
"loss": 0.5324,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.4948928832848403,
"learning_rate": 3.0789017287564913e-05,
"loss": 0.5126,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.44553992392913416,
"learning_rate": 3.074135451053475e-05,
"loss": 0.4944,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.3817042637074389,
"learning_rate": 3.069360584592015e-05,
"loss": 0.5201,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.40987677475263945,
"learning_rate": 3.064577167551816e-05,
"loss": 0.5061,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.38258245744672703,
"learning_rate": 3.059785238180953e-05,
"loss": 0.5291,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.4005170964371923,
"learning_rate": 3.054984834795566e-05,
"loss": 0.5249,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.38311002872922556,
"learning_rate": 3.0501759957795537e-05,
"loss": 0.5203,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.3982618931616072,
"learning_rate": 3.0453587595842627e-05,
"loss": 0.5118,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.4306399795885287,
"learning_rate": 3.0405331647281856e-05,
"loss": 0.5174,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.3172722552733599,
"learning_rate": 3.0356992497966508e-05,
"loss": 0.5172,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.37327794294218064,
"learning_rate": 3.0308570534415137e-05,
"loss": 0.5218,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.39002590169564255,
"learning_rate": 3.0260066143808467e-05,
"loss": 0.5283,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.3853660132264016,
"learning_rate": 3.021147971398632e-05,
"loss": 0.5097,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.48899644115774726,
"learning_rate": 3.0162811633444478e-05,
"loss": 0.5169,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.3865787831376619,
"learning_rate": 3.0114062291331635e-05,
"loss": 0.5137,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.4404596461129848,
"learning_rate": 3.0065232077446217e-05,
"loss": 0.5314,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.41857698115810316,
"learning_rate": 3.001632138223332e-05,
"loss": 0.5154,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.5366713961201395,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.4606,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 0.4626429339585698,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.4402,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 0.721389525541343,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.4475,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.4803612599107758,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.4437,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.5054208571721194,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.4316,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.46712562019444287,
"learning_rate": 2.972118904869376e-05,
"loss": 0.4548,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.6149955895612947,
"learning_rate": 2.9671725970312563e-05,
"loss": 0.4568,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.4233646200332559,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.4385,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.6193999724555395,
"learning_rate": 2.95725682050846e-05,
"loss": 0.4514,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.5390746520594105,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.4357,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.5125628000225935,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.4634,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.5636389964237921,
"learning_rate": 2.942325848714505e-05,
"loss": 0.4461,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.49501790915936345,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.4589,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.47084169871934345,
"learning_rate": 2.932334127129686e-05,
"loss": 0.4363,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.44247562232273785,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.4333,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.47630337375190057,
"learning_rate": 2.922312585929131e-05,
"loss": 0.4393,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.3946374018255513,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.4459,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.4988173196429385,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.4396,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.4379302526725672,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.4374,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.48886244425387093,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.4489,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.4805781964319383,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.4448,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.4338018028173234,
"learning_rate": 2.8920722546128863e-05,
"loss": 0.4235,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.4926659280663074,
"learning_rate": 2.887006998351372e-05,
"loss": 0.4598,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.4393192492876481,
"learning_rate": 2.881934649605535e-05,
"loss": 0.4582,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.4984452150172177,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.4407,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.512948556170671,
"learning_rate": 2.8717688369507425e-05,
"loss": 0.4326,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.4622278127456779,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.4394,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.8059913923518713,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.4573,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.4324378299726228,
"learning_rate": 2.856467940121025e-05,
"loss": 0.4476,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.4576047276577876,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.4239,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.37144509196027026,
"learning_rate": 2.846233032790149e-05,
"loss": 0.4252,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.4650328272715631,
"learning_rate": 2.84110540896642e-05,
"loss": 0.4635,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.36614277714314697,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.4704,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.42378161481515597,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.4451,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.33933281731969933,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.4376,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.39864924111058003,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.4481,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.3637487335270627,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.4248,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.40218129408206965,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.4516,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.3949603726763716,
"learning_rate": 2.805026032173773e-05,
"loss": 0.4518,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.37011929529833526,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.4491,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.3931641401199811,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.4597,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.36173028137615987,
"learning_rate": 2.7894660620408e-05,
"loss": 0.452,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.36089854543357164,
"learning_rate": 2.784266724883173e-05,
"loss": 0.4435,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.8808328720854052,
"learning_rate": 2.7790611167497128e-05,
"loss": 0.463,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.39994622602181273,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.4538,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.422277991641746,
"learning_rate": 2.768631254100716e-05,
"loss": 0.4745,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.363353602829565,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.424,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.4353441666578217,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.4425,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.8150663238740433,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.4709,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.5044034738788512,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.449,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.33936839538328684,
"learning_rate": 2.742449775134085e-05,
"loss": 0.4507,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.4050849531467314,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.4578,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.42462508079987693,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.4435,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.3385252628953948,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.4614,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.352088769385564,
"learning_rate": 2.7213974824975785e-05,
"loss": 0.4515,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.37012966421717836,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.4309,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.3400918504269219,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.4465,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.32805101908224243,
"learning_rate": 2.705547548999996e-05,
"loss": 0.4509,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.8574679588389204,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.4638,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.7049186435157787,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.4379,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.38922857347679213,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.4652,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.3751344395226707,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.4591,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.3780578609262804,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.4502,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.4257266950510736,
"learning_rate": 2.67369650552155e-05,
"loss": 0.4601,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.3606911957309391,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.4585,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.3877523612277333,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.4509,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.41666626064728257,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.4619,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.31690390175419564,
"learning_rate": 2.652354173593614e-05,
"loss": 0.4569,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.7356626854694025,
"learning_rate": 2.647005443331317e-05,
"loss": 0.4695,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.3258902989122571,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.4647,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.43778731566570317,
"learning_rate": 2.636292505301542e-05,
"loss": 0.4376,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.3686067722825784,
"learning_rate": 2.630928383194431e-05,
"loss": 0.4598,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.4623925784936642,
"learning_rate": 2.625559216200781e-05,
"loss": 0.4664,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.35804723782729336,
"learning_rate": 2.620185047252308e-05,
"loss": 0.4578,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.3847900350190687,
"learning_rate": 2.614805919320723e-05,
"loss": 0.4507,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.36431559595324087,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.4362,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.3228734397646898,
"learning_rate": 2.60403295859298e-05,
"loss": 0.443,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.33864471568972904,
"learning_rate": 2.5986392119371297e-05,
"loss": 0.4681,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.32917868585324916,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.4464,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.32732445338832294,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.4379,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.29820227332080207,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.4604,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.3432700184982519,
"learning_rate": 2.577016790136453e-05,
"loss": 0.4672,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.3003053632744563,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.4612,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.34051424622729953,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.4376,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.3820767427045452,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.448,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.30243162886383046,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.4657,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.32619662515920966,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.4317,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.34257653341705135,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.447,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.3103642347933657,
"learning_rate": 2.539001591449889e-05,
"loss": 0.4624,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.332078889125599,
"learning_rate": 2.53355326151989e-05,
"loss": 0.438,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.31226659312182886,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.4457,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.3844622645707507,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.4573,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.32009837090507776,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.466,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.36781284916324364,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.4585,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.31975719583607826,
"learning_rate": 2.506248490185627e-05,
"loss": 0.4552,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.34617405132142265,
"learning_rate": 2.500775217241923e-05,
"loss": 0.4598,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.3409764834581472,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.4496,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.39224233523954616,
"learning_rate": 2.4898167025948e-05,
"loss": 0.4674,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.3078244563216794,
"learning_rate": 2.4843315485153703e-05,
"loss": 0.4575,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.43231244665095053,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.4669,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.29183937436750956,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.4725,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.3520312183924523,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.4473,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.34053175598557567,
"learning_rate": 2.46235264422994e-05,
"loss": 0.4474,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.34947849098093414,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.4393,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.3677211780526643,
"learning_rate": 2.451340834511592e-05,
"loss": 0.449,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.3206296992084034,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.4677,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.3015526421778376,
"learning_rate": 2.440314589184958e-05,
"loss": 0.4429,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.34847092606385904,
"learning_rate": 2.434796163357541e-05,
"loss": 0.4409,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.3243041735483128,
"learning_rate": 2.429274260911588e-05,
"loss": 0.4457,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.3244335717040414,
"learning_rate": 2.423748926000084e-05,
"loss": 0.4548,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.31066900145517334,
"learning_rate": 2.418220202803459e-05,
"loss": 0.435,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.3343628764609248,
"learning_rate": 2.412688135529237e-05,
"loss": 0.4513,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.2848348604444999,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.4635,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.4048222567805918,
"learning_rate": 2.4016141457114372e-05,
"loss": 0.4426,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.31742740410981013,
"learning_rate": 2.396072311715187e-05,
"loss": 0.4682,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.39247689722467005,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.4403,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.32824924548894036,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.4472,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.36617181565105417,
"learning_rate": 2.379427985200223e-05,
"loss": 0.459,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.31524510042205545,
"learning_rate": 2.373873749394972e-05,
"loss": 0.4584,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.37588189852426845,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.4441,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.32526769412395623,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.4449,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.5040989990873995,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.4559,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.34483852512390817,
"learning_rate": 2.351627355800233e-05,
"loss": 0.4414,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.4450233672339188,
"learning_rate": 2.346058617160256e-05,
"loss": 0.4391,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.34897653753779634,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.4485,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.5302806512647216,
"learning_rate": 2.334912883203469e-05,
"loss": 0.4477,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.3442459122612514,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.4337,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.3593313641359571,
"learning_rate": 2.323756437450136e-05,
"loss": 0.4431,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.34106484235412715,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.4434,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.38048816352224946,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.4463,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.39585293859519266,
"learning_rate": 2.307002464849052e-05,
"loss": 0.448,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.3810218901987287,
"learning_rate": 2.30141283818846e-05,
"loss": 0.4322,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.3382583989637951,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.4359,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.361583537693359,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.4601,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.3430271918652335,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.4258,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.3541048173784491,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.4526,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.30525237629981916,
"learning_rate": 2.273429448107257e-05,
"loss": 0.4424,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.296756611345705,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.4528,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.36150114945771006,
"learning_rate": 2.262220474347102e-05,
"loss": 0.4564,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.3069990467063392,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.4442,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.330650518624278,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.4639,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.36081924572358576,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.4477,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.3174124951808157,
"learning_rate": 2.2397777251561302e-05,
"loss": 0.4638,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.3619065156995915,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.4318,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.32060891110526246,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.4481,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.320376971918371,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.4626,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.37775858397014583,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.442,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.3233199544931668,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.4514,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.37832203072107007,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.4424,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.3669721427301362,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.4513,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.33421136345754876,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.4691,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.35498463211344583,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.4255,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.3493427295921005,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.4304,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.35277297247648465,
"learning_rate": 2.17791064266051e-05,
"loss": 0.4369,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.3458420833405429,
"learning_rate": 2.172276926861741e-05,
"loss": 0.4529,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.3963356394492606,
"learning_rate": 2.166641833541255e-05,
"loss": 0.455,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.3280284395840259,
"learning_rate": 2.161005407757109e-05,
"loss": 0.4441,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.354555089585402,
"learning_rate": 2.155367694578013e-05,
"loss": 0.4585,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.36505505126953736,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.4445,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.3610422926659515,
"learning_rate": 2.1440885863609208e-05,
"loss": 0.4468,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.3262257066742579,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.4571,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.38876906151347945,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.4475,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.3543604198168496,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.4512,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.370103850544902,
"learning_rate": 2.1215169053088825e-05,
"loss": 0.4407,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.37215471258771077,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.4585,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.2973326308525486,
"learning_rate": 2.110225054402733e-05,
"loss": 0.4556,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.37519483165966583,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.4422,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.36158648361650125,
"learning_rate": 2.09892967807724e-05,
"loss": 0.4438,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.3452039956708305,
"learning_rate": 2.093280780774835e-05,
"loss": 0.4512,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.3975968232081312,
"learning_rate": 2.087631137601793e-05,
"loss": 0.4715,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.29615868531576917,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.4707,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.391143773833073,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.4573,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.30078244289229794,
"learning_rate": 2.070678184630458e-05,
"loss": 0.4519,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.3404433257091126,
"learning_rate": 2.06502600977305e-05,
"loss": 0.4523,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.3581839313792603,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.4501,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.32091792766107224,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.4421,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.3743164979695902,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.4609,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.3165536088419245,
"learning_rate": 2.042412562887594e-05,
"loss": 0.455,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.33979078975903654,
"learning_rate": 2.0367582403231033e-05,
"loss": 0.4529,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.3383291763713177,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.4415,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.3289078265785243,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.4448,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.3254454754107151,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.4569,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.3458189392025663,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.4385,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.3349986436167749,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.4576,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.3638574018292932,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.4233,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.33395021051245516,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.4655,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.350068545071552,
"learning_rate": 1.99151687695709e-05,
"loss": 0.4365,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.30888498695013106,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.4543,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.3990849699590566,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.4487,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.31139250009661074,
"learning_rate": 1.974551241345447e-05,
"loss": 0.4567,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.3345412661136026,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.4437,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.4059950355665284,
"learning_rate": 1.963241759676897e-05,
"loss": 0.4285,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.3690449084610853,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.4541,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.33619982684492106,
"learning_rate": 1.951933453677638e-05,
"loss": 0.4396,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.33492326288647495,
"learning_rate": 1.9462798545816932e-05,
"loss": 0.4377,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.3067654608929252,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.4478,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.32925016803512336,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.4513,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.35811598373467285,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.4405,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.3383114839148709,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.4505,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.3858421029931187,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.4429,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.3361451162722987,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.4541,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.345277214675037,
"learning_rate": 1.906719219225165e-05,
"loss": 0.4487,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.3735683606516725,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.4559,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.34843309666699046,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.4405,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.37763657004329726,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.4394,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.3052688028101917,
"learning_rate": 1.884128556891707e-05,
"loss": 0.4428,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.3797419062813991,
"learning_rate": 1.878483094691119e-05,
"loss": 0.4372,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.3410282941485193,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.4343,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.3309099823195336,
"learning_rate": 1.867195130360947e-05,
"loss": 0.4376,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.3368057394904786,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.4599,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.3290339252777393,
"learning_rate": 1.8559114136390795e-05,
"loss": 0.4505,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.2818897854095426,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.4494,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.63669542483964,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.457,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.29994355853502386,
"learning_rate": 1.838994592242891e-05,
"loss": 0.4315,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.34296914847431964,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.4574,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.3158015508325669,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.455,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.326055819130679,
"learning_rate": 1.82208935733949e-05,
"loss": 0.4399,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.3114948231418578,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.4651,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.3163094575526374,
"learning_rate": 1.810826238483892e-05,
"loss": 0.4535,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.35134830732947714,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.4326,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.32656082547002563,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.4559,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.3729482147707082,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.4399,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.3360070783659262,
"learning_rate": 1.788318512320838e-05,
"loss": 0.4517,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.32419021602893194,
"learning_rate": 1.782695699829051e-05,
"loss": 0.4495,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.3210911020487191,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.4472,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.35948655798548373,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.4424,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.2929137571982241,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.453,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.36285989461512913,
"learning_rate": 1.76022227484387e-05,
"loss": 0.4527,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.31127293255550653,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.4401,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.31480754157448826,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.4454,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.3389044177247185,
"learning_rate": 1.743387180005809e-05,
"loss": 0.46,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.30295197285363057,
"learning_rate": 1.737779525652899e-05,
"loss": 0.4541,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.307505228670382,
"learning_rate": 1.732173968007914e-05,
"loss": 0.4457,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.36899891151455627,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.4579,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.29662485849877873,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.4427,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.3465727345073777,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.4609,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.32394287809758765,
"learning_rate": 1.709773600686015e-05,
"loss": 0.4507,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.30707878060821453,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.4579,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.32219297819224113,
"learning_rate": 1.69858716181154e-05,
"loss": 0.4449,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.29205121216468233,
"learning_rate": 1.692997535150948e-05,
"loss": 0.4398,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.29462453504374997,
"learning_rate": 1.687410363273884e-05,
"loss": 0.4442,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.3231360787662342,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.4687,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.6576880589495006,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.4517,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.3131480425737225,
"learning_rate": 1.67066402299235e-05,
"loss": 0.4363,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.3348991338724005,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.4487,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.291868665530291,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.4371,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.3195449832441071,
"learning_rate": 1.653941382839745e-05,
"loss": 0.4502,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 1.1424298936965598,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.402,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.3383824805327573,
"learning_rate": 1.642806717162757e-05,
"loss": 0.3888,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.589920619544513,
"learning_rate": 1.637243646233718e-05,
"loss": 0.3818,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.43741491879160965,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.3916,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.47539062289925166,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.3754,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.42137119017741514,
"learning_rate": 1.620572014799777e-05,
"loss": 0.374,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.3852418538377848,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.3632,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.4086312059908428,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.3793,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.46850269756461593,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.3937,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.33863901150384634,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.3666,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.4073123772113444,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.3928,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.4044972364823636,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.3811,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.3014085728502069,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.3783,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.3830807583361465,
"learning_rate": 1.576251073999917e-05,
"loss": 0.375,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.35000459463642375,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.3987,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.2846922083811925,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.3858,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.3256658674257771,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.3882,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.33984872386113946,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.3632,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.3133714506279018,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.3793,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.3273987117111861,
"learning_rate": 1.543151434154755e-05,
"loss": 0.3797,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.34719421432907194,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.3756,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.2678924658455305,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.3779,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.3126598551894825,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.3707,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.3109917201304041,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.3999,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.2865734077954445,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.369,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.3425365947447981,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.3844,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.3141734943794524,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.3897,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.3124219088125161,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.3539,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.3323646333748089,
"learning_rate": 1.493751509814374e-05,
"loss": 0.3846,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.3236751567778616,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.3826,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.28040636733878826,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.3795,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.3324043186259119,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.3834,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.3180727148512536,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.3817,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.3040819951386089,
"learning_rate": 1.466446738480111e-05,
"loss": 0.3812,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.5714736641028277,
"learning_rate": 1.460998408550112e-05,
"loss": 0.3696,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.30192934483627293,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.3754,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.2845223466098713,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.3781,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.3188765455140033,
"learning_rate": 1.444679451900437e-05,
"loss": 0.3684,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.308520900415538,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.3847,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.29772895421268525,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.3854,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.3027481325268989,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.3963,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.28751642750698336,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.3865,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.290593888125531,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.4029,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.2790078918898438,
"learning_rate": 1.412162598317602e-05,
"loss": 0.3846,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.2899632574676317,
"learning_rate": 1.406759321421907e-05,
"loss": 0.3996,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.28365658356150847,
"learning_rate": 1.401360788062871e-05,
"loss": 0.3721,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.26479454997786744,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.3882,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.33641549892563993,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.3807,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.2823715362897143,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.3829,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.30718417409709564,
"learning_rate": 1.379814952747693e-05,
"loss": 0.3635,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.3087308624733637,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.404,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.4110029186056012,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.3613,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.3111524415039772,
"learning_rate": 1.363707494698459e-05,
"loss": 0.3597,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.2866397959448614,
"learning_rate": 1.358348460369265e-05,
"loss": 0.3884,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.3075654705395617,
"learning_rate": 1.352994556668684e-05,
"loss": 0.3728,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.29012663399765865,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.3672,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.29950192538009096,
"learning_rate": 1.342302312350678e-05,
"loss": 0.3942,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.29796257907715895,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.3748,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.2732791047338982,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.3829,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.3079733414602006,
"learning_rate": 1.326303494478451e-05,
"loss": 0.3876,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.28832250555641686,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.3847,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.3155110984465533,
"learning_rate": 1.315664479122898e-05,
"loss": 0.3717,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.28467285891519284,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.4077,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.3000151477441149,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.3773,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.26023573546455525,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.3848,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.29190555315364386,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.3747,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.292625781307788,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.3868,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.2584305472375794,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.3864,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.29146100506384814,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.3761,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.33724649432021425,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.366,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.26804010285877977,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.4101,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.28261641455021663,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.3953,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.3432003124327908,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.3728,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.2840540005445252,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.3623,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.27908308245868907,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.3808,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.3032002947539616,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.3661,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.302107123102299,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.3871,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.2583360511427107,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.3944,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.2959185303508997,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.3716,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.27985559102433294,
"learning_rate": 1.220938883250288e-05,
"loss": 0.3814,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.47450492996507365,
"learning_rate": 1.215733275116827e-05,
"loss": 0.4023,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.3240086825205765,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.3916,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.267404515365025,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.3522,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.28250666482405884,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.394,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.2964970033616171,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.3657,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.27671239109960233,
"learning_rate": 1.189800129803139e-05,
"loss": 0.3838,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.2627171439905052,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.4016,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.282783682346681,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.383,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.27349069396115194,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.373,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.2692144764529452,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.3803,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.2847573787807442,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.3901,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.2697762143001218,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.41,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.28033159514515316,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.3808,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.29869596407293575,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.3849,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.2902805331741144,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.3887,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.26048474830328905,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.3872,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.2855251063833262,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.3817,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.3392740423878247,
"learning_rate": 1.128231163049258e-05,
"loss": 0.3741,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.2846209032666356,
"learning_rate": 1.123144751066259e-05,
"loss": 0.3914,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.27522224493439523,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.3712,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.330423352819347,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.3804,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.31189170244292475,
"learning_rate": 1.107927745387114e-05,
"loss": 0.3617,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.29093023567392584,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.3865,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.29730752908301106,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.373,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.28093592399446804,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.3775,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.2706866261950757,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.3847,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.27148639227820087,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.3791,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.2818181801504894,
"learning_rate": 1.077687414070869e-05,
"loss": 0.385,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.28882768162964034,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.3884,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.2704037351881482,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.3882,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.27905479599273075,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.3913,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.26337547306944625,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.3872,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.28711327262671554,
"learning_rate": 1.052689572753247e-05,
"loss": 0.3767,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.26476845662103726,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.3941,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.27259913231501803,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.3905,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.2928240465557256,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.3872,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.25937165208831053,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.3888,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.2776835877309511,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.3833,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.2482394374329225,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.3767,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.3003994875002175,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.3658,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.26135812126806884,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.3767,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.25675310525770106,
"learning_rate": 1.008173988718005e-05,
"loss": 0.3702,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.2553766970658665,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.3835,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.2759747171248721,
"learning_rate": 9.983678617766677e-06,
"loss": 0.3831,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.2818622614158607,
"learning_rate": 9.934767922553788e-06,
"loss": 0.386,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.2668320060603505,
"learning_rate": 9.885937708668371e-06,
"loss": 0.3987,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.26444621949246055,
"learning_rate": 9.837188366555524e-06,
"loss": 0.3844,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.26246826774777116,
"learning_rate": 9.788520286013691e-06,
"loss": 0.3795,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.3083408754982388,
"learning_rate": 9.739933856191537e-06,
"loss": 0.3789,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.39855634296426945,
"learning_rate": 9.691429465584866e-06,
"loss": 0.3877,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.2560805862964639,
"learning_rate": 9.643007502033492e-06,
"loss": 0.3804,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.2571850704456232,
"learning_rate": 9.594668352718152e-06,
"loss": 0.4061,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.2757590991852375,
"learning_rate": 9.546412404157383e-06,
"loss": 0.3718,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.2629713420583283,
"learning_rate": 9.498240042204471e-06,
"loss": 0.3927,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.2625353294272889,
"learning_rate": 9.450151652044339e-06,
"loss": 0.3651,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.27649245965405705,
"learning_rate": 9.402147618190472e-06,
"loss": 0.383,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.25857410118290675,
"learning_rate": 9.354228324481841e-06,
"loss": 0.3975,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.26555768736747143,
"learning_rate": 9.306394154079852e-06,
"loss": 0.3726,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.2787365462254729,
"learning_rate": 9.258645489465254e-06,
"loss": 0.3866,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.26843271751601655,
"learning_rate": 9.21098271243509e-06,
"loss": 0.3779,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.26228326995683143,
"learning_rate": 9.163406204099659e-06,
"loss": 0.3822,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.2607873705300564,
"learning_rate": 9.115916344879452e-06,
"loss": 0.3879,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.2749002490221343,
"learning_rate": 9.068513514502126e-06,
"loss": 0.3761,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.26219332640435317,
"learning_rate": 9.021198091999439e-06,
"loss": 0.3798,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.2537090891539065,
"learning_rate": 8.973970455704251e-06,
"loss": 0.3722,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.24819312042430586,
"learning_rate": 8.926830983247495e-06,
"loss": 0.3996,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.27426598239037553,
"learning_rate": 8.879780051555116e-06,
"loss": 0.3946,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.2855418499726935,
"learning_rate": 8.83281803684512e-06,
"loss": 0.3646,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.25358343100676806,
"learning_rate": 8.785945314624532e-06,
"loss": 0.3853,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.27807019923270687,
"learning_rate": 8.739162259686385e-06,
"loss": 0.3883,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.27228074277568426,
"learning_rate": 8.692469246106734e-06,
"loss": 0.3658,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.27898821439840044,
"learning_rate": 8.645866647241677e-06,
"loss": 0.3962,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.2775722728529585,
"learning_rate": 8.599354835724348e-06,
"loss": 0.3729,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.2536967587533296,
"learning_rate": 8.552934183461958e-06,
"loss": 0.3642,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.2616595935286664,
"learning_rate": 8.506605061632805e-06,
"loss": 0.3906,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.27783817088844737,
"learning_rate": 8.460367840683308e-06,
"loss": 0.3708,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.2683581254102091,
"learning_rate": 8.414222890325065e-06,
"loss": 0.3924,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.256156656338443,
"learning_rate": 8.368170579531855e-06,
"loss": 0.3809,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.25337578241974207,
"learning_rate": 8.322211276536733e-06,
"loss": 0.3858,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.25178799260267204,
"learning_rate": 8.27634534882907e-06,
"loss": 0.3835,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.27429515345277605,
"learning_rate": 8.230573163151585e-06,
"loss": 0.3621,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.2534786660792886,
"learning_rate": 8.184895085497466e-06,
"loss": 0.3845,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.2803934447012294,
"learning_rate": 8.139311481107401e-06,
"loss": 0.3571,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.24405727276095823,
"learning_rate": 8.093822714466686e-06,
"loss": 0.3797,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.26862695881361254,
"learning_rate": 8.048429149302271e-06,
"loss": 0.3817,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.26571237986477875,
"learning_rate": 8.003131148579902e-06,
"loss": 0.3981,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.25619381329336827,
"learning_rate": 7.957929074501182e-06,
"loss": 0.3747,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.245728443470987,
"learning_rate": 7.912823288500695e-06,
"loss": 0.3907,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.24842931272046406,
"learning_rate": 7.867814151243102e-06,
"loss": 0.3697,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.25512822381780115,
"learning_rate": 7.822902022620267e-06,
"loss": 0.382,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.26415316816826884,
"learning_rate": 7.778087261748377e-06,
"loss": 0.3818,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.2740399389221058,
"learning_rate": 7.733370226965056e-06,
"loss": 0.3648,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.2557960041919949,
"learning_rate": 7.688751275826527e-06,
"loss": 0.3807,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.2727879476339929,
"learning_rate": 7.644230765104735e-06,
"loss": 0.3433,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.2811567771928,
"learning_rate": 7.599809050784502e-06,
"loss": 0.3776,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.26368948236531004,
"learning_rate": 7.555486488060661e-06,
"loss": 0.3771,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.26506958081264104,
"learning_rate": 7.511263431335249e-06,
"loss": 0.3806,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.30024155592748836,
"learning_rate": 7.467140234214656e-06,
"loss": 0.3892,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.2815317503574504,
"learning_rate": 7.423117249506777e-06,
"loss": 0.3835,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.276774416516275,
"learning_rate": 7.379194829218231e-06,
"loss": 0.3833,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.2409207337367531,
"learning_rate": 7.335373324551518e-06,
"loss": 0.3846,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.26035117441574057,
"learning_rate": 7.291653085902224e-06,
"loss": 0.3982,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.26780580693903083,
"learning_rate": 7.248034462856213e-06,
"loss": 0.3775,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.2691973332288732,
"learning_rate": 7.20451780418683e-06,
"loss": 0.3702,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.26384710610395873,
"learning_rate": 7.161103457852128e-06,
"loss": 0.3747,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.27554568330794244,
"learning_rate": 7.117791770992046e-06,
"loss": 0.3809,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.2723832990538713,
"learning_rate": 7.074583089925688e-06,
"loss": 0.366,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.26546450811845096,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.373,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.2592703443693665,
"learning_rate": 6.988476126329604e-06,
"loss": 0.374,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.3016132235354125,
"learning_rate": 6.945578532308857e-06,
"loss": 0.3903,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.27935488085631227,
"learning_rate": 6.902785321094301e-06,
"loss": 0.3916,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.24700437215499121,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.3826,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.25652797317412884,
"learning_rate": 6.817513414939892e-06,
"loss": 0.3858,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.26765907243149106,
"learning_rate": 6.775035401831924e-06,
"loss": 0.3896,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.26103066279135057,
"learning_rate": 6.732663135188473e-06,
"loss": 0.3953,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.24449485797491613,
"learning_rate": 6.690396953817055e-06,
"loss": 0.3657,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.2586930912186053,
"learning_rate": 6.648237195676927e-06,
"loss": 0.3634,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.25315773012400383,
"learning_rate": 6.606184197876395e-06,
"loss": 0.3876,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.2655804908122219,
"learning_rate": 6.564238296670109e-06,
"loss": 0.3572,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.24021357431723755,
"learning_rate": 6.52239982745638e-06,
"loss": 0.3815,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.24578091780763267,
"learning_rate": 6.480669124774484e-06,
"loss": 0.3786,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.23891934514236285,
"learning_rate": 6.439046522302013e-06,
"loss": 0.3913,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.24653546980773147,
"learning_rate": 6.397532352852191e-06,
"loss": 0.3593,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.23896455915205866,
"learning_rate": 6.356126948371202e-06,
"loss": 0.3742,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.2372816978890914,
"learning_rate": 6.314830639935559e-06,
"loss": 0.3738,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.24480888592357009,
"learning_rate": 6.273643757749448e-06,
"loss": 0.3861,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.4241550481707884,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.3664,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.240912342808302,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.3798,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.2420248321756086,
"learning_rate": 6.150742957589784e-06,
"loss": 0.3876,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.25166702104323246,
"learning_rate": 6.109997064904774e-06,
"loss": 0.3787,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.24770130729902023,
"learning_rate": 6.069362236313092e-06,
"loss": 0.3744,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.23901553192844796,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.3837,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.2701203073707638,
"learning_rate": 5.988427070179108e-06,
"loss": 0.369,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.2320214223204591,
"learning_rate": 5.948127379792234e-06,
"loss": 0.4063,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.25903679615714664,
"learning_rate": 5.90794004780439e-06,
"loss": 0.3865,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.2568833195296441,
"learning_rate": 5.867865395552405e-06,
"loss": 0.3689,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.24835563313763623,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.3837,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.239485476140789,
"learning_rate": 5.788055411095892e-06,
"loss": 0.3904,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.2474014534572385,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.3757,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.2439262810814573,
"learning_rate": 5.708699979051613e-06,
"loss": 0.3827,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.2372401871150628,
"learning_rate": 5.66919351390751e-06,
"loss": 0.3935,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.24011208377636564,
"learning_rate": 5.629801637510188e-06,
"loss": 0.3905,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 1.2359783512466322,
"learning_rate": 5.590524664836043e-06,
"loss": 0.3648,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.24646571761961944,
"learning_rate": 5.551362909942708e-06,
"loss": 0.3848,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.24954192843029466,
"learning_rate": 5.51231668596653e-06,
"loss": 0.3672,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.24799876810245827,
"learning_rate": 5.47338630512007e-06,
"loss": 0.3624,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.24430089846152486,
"learning_rate": 5.434572078689617e-06,
"loss": 0.3893,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.24455653096684413,
"learning_rate": 5.395874317032694e-06,
"loss": 0.3911,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.23566420961951792,
"learning_rate": 5.357293329575573e-06,
"loss": 0.3898,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.23764848424348714,
"learning_rate": 5.318829424810797e-06,
"loss": 0.379,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.24676578682905181,
"learning_rate": 5.280482910294728e-06,
"loss": 0.3832,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.2400866354679534,
"learning_rate": 5.242254092645078e-06,
"loss": 0.38,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.24054304112246006,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.3809,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.22925509881472866,
"learning_rate": 5.166150769707892e-06,
"loss": 0.4029,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.24614921300701226,
"learning_rate": 5.128276872940503e-06,
"loss": 0.3726,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.23777999035181593,
"learning_rate": 5.090521890074953e-06,
"loss": 0.394,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.25669753579717863,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.3685,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.2524679174095785,
"learning_rate": 5.015369872647442e-06,
"loss": 0.3653,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.2523613421786001,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.3673,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.27389889002189943,
"learning_rate": 4.940697121074675e-06,
"loss": 0.3986,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.2503982346056133,
"learning_rate": 4.903541216934864e-06,
"loss": 0.3882,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.23762213855181474,
"learning_rate": 4.8665060236772e-06,
"loss": 0.3795,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.23490535990824501,
"learning_rate": 4.829591837434102e-06,
"loss": 0.3965,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.2650966692953968,
"learning_rate": 4.792798953370424e-06,
"loss": 0.3807,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.23426101376977013,
"learning_rate": 4.756127665681083e-06,
"loss": 0.3847,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.25505773024856576,
"learning_rate": 4.719578267588725e-06,
"loss": 0.3563,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.6286034835518819,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.3965,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.3717197222554223,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.3527,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.3123242777495726,
"learning_rate": 4.610664328486485e-06,
"loss": 0.3456,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.25043055818002113,
"learning_rate": 4.5746054014809e-06,
"loss": 0.3293,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.36972635397397025,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.3447,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.43136748918527484,
"learning_rate": 4.502857857942204e-06,
"loss": 0.3299,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.3087538568083091,
"learning_rate": 4.467169815100549e-06,
"loss": 0.3453,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.28822240973588065,
"learning_rate": 4.431605972355137e-06,
"loss": 0.3455,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.3490723336176667,
"learning_rate": 4.396166614073516e-06,
"loss": 0.3535,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.31779261816375504,
"learning_rate": 4.360852023627846e-06,
"loss": 0.3301,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.26445138674747515,
"learning_rate": 4.325662483392656e-06,
"loss": 0.3402,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.2570646511660508,
"learning_rate": 4.29059827474257e-06,
"loss": 0.3492,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.299567620227929,
"learning_rate": 4.255659678050063e-06,
"loss": 0.3382,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.3363479733287015,
"learning_rate": 4.22084697268323e-06,
"loss": 0.3333,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.30357063696890807,
"learning_rate": 4.186160437003535e-06,
"loss": 0.3347,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.26266911574125373,
"learning_rate": 4.151600348363582e-06,
"loss": 0.3303,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.2711844496777647,
"learning_rate": 4.11716698310493e-06,
"loss": 0.3477,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.27748424652131876,
"learning_rate": 4.082860616555848e-06,
"loss": 0.3172,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.2787654544210855,
"learning_rate": 4.048681523029125e-06,
"loss": 0.346,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.27938557591467633,
"learning_rate": 4.014629975819882e-06,
"loss": 0.3259,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.2567370030017558,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.3554,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.2649528353676226,
"learning_rate": 3.94691060843287e-06,
"loss": 0.336,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.29081720111139825,
"learning_rate": 3.913243329737353e-06,
"loss": 0.3176,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.273271900460297,
"learning_rate": 3.879704680319503e-06,
"loss": 0.3209,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.23257323806946706,
"learning_rate": 3.846294928353471e-06,
"loss": 0.3673,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.24378237462144078,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.3332,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.2545857580990604,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.3258,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.24894452115750007,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.3468,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.25011215995716807,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.3444,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.24447509508347304,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.3291,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.24592024817960503,
"learning_rate": 3.648558150665886e-06,
"loss": 0.3302,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.23741075550502652,
"learning_rate": 3.616058102895075e-06,
"loss": 0.3348,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.25629881818023453,
"learning_rate": 3.583689060687079e-06,
"loss": 0.3365,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.3431539496066457,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.3433,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.25083808316692857,
"learning_rate": 3.51934502719796e-06,
"loss": 0.3119,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.24351962302301056,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.3342,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.2513377616894056,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.338,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.2409043076326343,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.316,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.2267127864382647,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.3686,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.2283315471592896,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.3508,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.24563226414337266,
"learning_rate": 3.32948376099375e-06,
"loss": 0.3161,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.24142447834842973,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.3235,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.2473495857222565,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.3416,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.23280030302349408,
"learning_rate": 3.236349233544438e-06,
"loss": 0.3208,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.24243365171571807,
"learning_rate": 3.205572144561604e-06,
"loss": 0.3246,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.23931730606542917,
"learning_rate": 3.174929343376374e-06,
"loss": 0.3474,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.2379737884435119,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.3413,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.2339933457293264,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.3472,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.2326073798933234,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.3304,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.22478390296070422,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.3378,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.23469382388742022,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.3494,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.2363051483836901,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.3392,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.25154079498821574,
"learning_rate": 2.964210239224696e-06,
"loss": 0.3317,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.23182086598067025,
"learning_rate": 2.934650470499869e-06,
"loss": 0.3364,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.22321425293809527,
"learning_rate": 2.905227155855128e-06,
"loss": 0.355,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.2377019008419826,
"learning_rate": 2.875940530558519e-06,
"loss": 0.3441,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.24496677831909705,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.3378,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.2262905641100641,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.3534,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.24039058411559228,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.3309,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.6916783368933088,
"learning_rate": 2.760165589921184e-06,
"loss": 0.36,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.6031056430727239,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.353,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.23728952755406468,
"learning_rate": 2.703104292151164e-06,
"loss": 0.3393,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.22698180955966504,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.3328,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.2540974862927897,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.3318,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.23156254242028762,
"learning_rate": 2.618550200289303e-06,
"loss": 0.351,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.24002751541076764,
"learning_rate": 2.59064316684412e-06,
"loss": 0.3372,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.6387043265920096,
"learning_rate": 2.562875338152264e-06,
"loss": 0.3099,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.2337117955836568,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.3197,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.22761237149999683,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.356,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.23623824010563968,
"learning_rate": 2.480409295328803e-06,
"loss": 0.3316,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.21618579834266008,
"learning_rate": 2.45320049480108e-06,
"loss": 0.3373,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.23274844093827804,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.3251,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.2293722050712042,
"learning_rate": 2.399204021408421e-06,
"loss": 0.3189,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.23630630413711312,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.3093,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.23619517369231818,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.3532,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.23784473904765763,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.3332,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.2366076849719523,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.3122,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.23513891974269752,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.3351,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.23696320179313132,
"learning_rate": 2.24059906434998e-06,
"loss": 0.3223,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.22060755357472392,
"learning_rate": 2.214660694699091e-06,
"loss": 0.3341,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.21968587425379732,
"learning_rate": 2.188864536147355e-06,
"loss": 0.3404,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.2222442587628407,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.3178,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.23709265581529906,
"learning_rate": 2.137699676264171e-06,
"loss": 0.3195,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.223075919034264,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.334,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.2405281631580415,
"learning_rate": 2.08710612114857e-06,
"loss": 0.3234,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.2347706865166277,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.3067,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.23930006283815752,
"learning_rate": 2.0370854889762e-06,
"loss": 0.3274,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.24647326528815924,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.3155,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.22426187891626015,
"learning_rate": 1.987639379598427e-06,
"loss": 0.3513,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.24446697119832558,
"learning_rate": 1.963132265887919e-06,
"loss": 0.2955,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.23399015582005503,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.3241,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.228550023319382,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.3304,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.22852930720597925,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.339,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.22218506101829977,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.3405,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.2228226651733688,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.3579,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.227758548268771,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.3339,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.22388923383829049,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.3304,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.22793532622111176,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.3191,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.21504461345234227,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.3526,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.22506675050444927,
"learning_rate": 1.726025117089718e-06,
"loss": 0.3313,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.22706918115560093,
"learning_rate": 1.70311497206598e-06,
"loss": 0.347,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.2228166672852491,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.3357,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.23170083422869456,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.3369,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.2254905152257692,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.3351,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.23405718974331155,
"learning_rate": 1.612939220355272e-06,
"loss": 0.3172,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.2265632169789599,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.3381,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.23796756478782286,
"learning_rate": 1.568732761074121e-06,
"loss": 0.3189,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.23986637203294403,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.3087,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.22011581612348016,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.3376,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.22454098934004948,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.3505,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.20683704102281147,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.341,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.23019904799781657,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.3398,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.22911475720603164,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.3315,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.2265880387098166,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.3413,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.21951813095864053,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.3418,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.22092885681800487,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.344,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.23655832124652088,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.3456,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.2182822109538011,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.3188,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.228209584173349,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.3408,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.22343649454576003,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.3234,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.22571535267452814,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.3514,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.22615729450908992,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.3403,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.23122288294262516,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.3451,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.2245902772426389,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.3414,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.22745421627103718,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.3262,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.2173750260067848,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.342,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.21446943971119936,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.3535,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.21804765225170616,
"learning_rate": 1.140415161609505e-06,
"loss": 0.3467,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.2341390918139744,
"learning_rate": 1.121666393356604e-06,
"loss": 0.3374,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.2449113552281814,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.312,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.2183633402398427,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.3302,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.2208779549048004,
"learning_rate": 1.066326386135097e-06,
"loss": 0.3396,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.2225969007466801,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.336,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.22778862772416847,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.3262,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.2243950712434488,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.3324,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.22367055762177226,
"learning_rate": 9.9465987684501e-07,
"loss": 0.3364,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.2273310703163356,
"learning_rate": 9.771228098856534e-07,
"loss": 0.3414,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.21066369969804036,
"learning_rate": 9.597378493454013e-07,
"loss": 0.3531,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.23771525550722664,
"learning_rate": 9.425051342339353e-07,
"loss": 0.3128,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.23547328197208217,
"learning_rate": 9.254248023435864e-07,
"loss": 0.3503,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.22651136652850862,
"learning_rate": 9.084969902482288e-07,
"loss": 0.3214,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.22358968191038434,
"learning_rate": 8.917218333021993e-07,
"loss": 0.3268,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.2079952763911423,
"learning_rate": 8.750994656392109e-07,
"loss": 0.35,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.218585108900732,
"learning_rate": 8.58630020171265e-07,
"loss": 0.3395,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.23305677853962334,
"learning_rate": 8.423136285876099e-07,
"loss": 0.3278,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.2220443186289935,
"learning_rate": 8.26150421353682e-07,
"loss": 0.339,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.259806352261638,
"learning_rate": 8.101405277100549e-07,
"loss": 0.3212,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.2456985212744622,
"learning_rate": 7.942840756714077e-07,
"loss": 0.3195,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.22175843709128137,
"learning_rate": 7.785811920255093e-07,
"loss": 0.3387,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.2224397633716813,
"learning_rate": 7.630320023322024e-07,
"loss": 0.3307,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.22137123432028086,
"learning_rate": 7.476366309223903e-07,
"loss": 0.333,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.2958178557083139,
"learning_rate": 7.323952008970537e-07,
"loss": 0.3151,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.2263424047898576,
"learning_rate": 7.173078341262618e-07,
"loss": 0.33,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.22454656635845324,
"learning_rate": 7.023746512481989e-07,
"loss": 0.3444,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.22151185712362592,
"learning_rate": 6.875957716681902e-07,
"loss": 0.3336,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.21439742780246213,
"learning_rate": 6.729713135577687e-07,
"loss": 0.3558,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.23849355222397384,
"learning_rate": 6.585013938537143e-07,
"loss": 0.3204,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.2185073406574934,
"learning_rate": 6.441861282571115e-07,
"loss": 0.3313,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.22690502976812432,
"learning_rate": 6.300256312324471e-07,
"loss": 0.322,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.2236132627119189,
"learning_rate": 6.160200160066776e-07,
"loss": 0.3279,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.2136365962067497,
"learning_rate": 6.021693945683327e-07,
"loss": 0.3339,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.2278391680869195,
"learning_rate": 5.884738776666044e-07,
"loss": 0.3301,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.21790617505921447,
"learning_rate": 5.749335748104878e-07,
"loss": 0.3449,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.22018834724308373,
"learning_rate": 5.615485942678778e-07,
"loss": 0.3233,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.2151714040271756,
"learning_rate": 5.483190430647289e-07,
"loss": 0.3367,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.21696231271818636,
"learning_rate": 5.352450269841747e-07,
"loss": 0.3316,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.22122592009295264,
"learning_rate": 5.22326650565701e-07,
"loss": 0.3414,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.2052917708681775,
"learning_rate": 5.095640171043048e-07,
"loss": 0.338,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.2332106060682708,
"learning_rate": 4.969572286496594e-07,
"loss": 0.31,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.2225175657226836,
"learning_rate": 4.845063860053079e-07,
"loss": 0.3398,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.22429752019412985,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.3432,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.22262528627110936,
"learning_rate": 4.600729351261857e-07,
"loss": 0.3321,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.22106752605897048,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.3477,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.2207807117325936,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.3364,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 1.3789159273384628,
"learning_rate": 4.245948007320588e-07,
"loss": 0.3407,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.22128238989920185,
"learning_rate": 4.130816799401971e-07,
"loss": 0.3509,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.22622643586506072,
"learning_rate": 4.017251756252827e-07,
"loss": 0.3319,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.24748077719273956,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.3017,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.21217073230945824,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.3484,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.21728751128243054,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.3366,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.23398748127389982,
"learning_rate": 3.578671204615347e-07,
"loss": 0.3212,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.2178948186771216,
"learning_rate": 3.472950355547267e-07,
"loss": 0.3529,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.22816047624743535,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.3366,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 1.0883920723877127,
"learning_rate": 3.266223765343224e-07,
"loss": 0.3424,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.22008574251316024,
"learning_rate": 3.16521967718757e-07,
"loss": 0.3363,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.23408399677933736,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.3175,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.22238076190475872,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.3388,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.23918161532619805,
"learning_rate": 2.871653894244597e-07,
"loss": 0.3067,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.22195486351745577,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.342,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.2215669921585236,
"learning_rate": 2.683823233205485e-07,
"loss": 0.3431,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.21752690624223067,
"learning_rate": 2.592274132703287e-07,
"loss": 0.343,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.22936075604527434,
"learning_rate": 2.502303499116754e-07,
"loss": 0.3146,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.21650889766284384,
"learning_rate": 2.413912051848666e-07,
"loss": 0.3398,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.2191200623787321,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.3226,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.21334282000236113,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.3448,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.22369019281737357,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.3353,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.2183194334126865,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.3229,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.21967753210728055,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.333,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.22527237109144752,
"learning_rate": 1.916764975761254e-07,
"loss": 0.3307,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.2332352904459758,
"learning_rate": 1.839446914621923e-07,
"loss": 0.3096,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.2214027406308652,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.338,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.21465661654805746,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.3443,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.22589285395387032,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.3423,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.2214270701735002,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.3378,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.2219729283860785,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.3344,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.22222482976153624,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.3328,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.21636297209627392,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.3409,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.22615058513976602,
"learning_rate": 1.277993209368744e-07,
"loss": 0.3292,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.22463156874719023,
"learning_rate": 1.214956261808542e-07,
"loss": 0.3332,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.21137041253410274,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.3331,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.21801566166953248,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.3192,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.22437991672205262,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.3394,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.2208295272761821,
"learning_rate": 9.787081204466253e-08,
"loss": 0.3448,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.21776737183806022,
"learning_rate": 9.236233586200627e-08,
"loss": 0.3454,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.2143635881195194,
"learning_rate": 8.70130406196279e-08,
"loss": 0.3207,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.22067081359862212,
"learning_rate": 8.182296909035047e-08,
"loss": 0.326,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.20701783791010908,
"learning_rate": 7.679216277384439e-08,
"loss": 0.3422,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.22714594128899607,
"learning_rate": 7.19206618963053e-08,
"loss": 0.3205,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.22341884355016248,
"learning_rate": 6.720850541012347e-08,
"loss": 0.3203,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.2215386604100352,
"learning_rate": 6.265573099357714e-08,
"loss": 0.3071,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.2206890924925404,
"learning_rate": 5.826237505053067e-08,
"loss": 0.341,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.22718975077590237,
"learning_rate": 5.4028472710137e-08,
"loss": 0.3293,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.226373726100088,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.3339,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.2054998581653382,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.3386,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.22542698888092735,
"learning_rate": 4.228381947004812e-08,
"loss": 0.3298,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.21038959793358317,
"learning_rate": 3.868805732811343e-08,
"loss": 0.3407,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.21349774995168586,
"learning_rate": 3.525190530455636e-08,
"loss": 0.3542,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.22309529813939522,
"learning_rate": 3.197539087475887e-08,
"loss": 0.3268,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.2138359496874014,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.3364,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.22966750351861115,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.3094,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.22002398313415397,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.3253,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.21842820681058223,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.3509,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.2254209029899217,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.3169,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.23176609483393243,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.3309,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.22424271402043172,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.333,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.2163436191868354,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.343,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.22186797657714857,
"learning_rate": 9.67435416023843e-09,
"loss": 0.3251,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.2194564591905599,
"learning_rate": 7.995445897566178e-09,
"loss": 0.3269,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.23774912606277648,
"learning_rate": 6.476393170578021e-09,
"loss": 0.3007,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.2119392690661949,
"learning_rate": 5.11720812557348e-09,
"loss": 0.3495,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.2331155801457071,
"learning_rate": 3.917901630563492e-09,
"loss": 0.3012,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.21703139834885107,
"learning_rate": 2.878483275170485e-09,
"loss": 0.3437,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.20570175073958702,
"learning_rate": 1.998961370555108e-09,
"loss": 0.355,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.21786897728952997,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.3315,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.2313693167566134,
"learning_rate": 7.196337656112384e-10,
"loss": 0.3058,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.21578037884091492,
"learning_rate": 3.198382947511292e-10,
"loss": 0.3494,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.3493582055201771,
"learning_rate": 7.995973352770137e-11,
"loss": 0.3539,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.2187729075596966,
"learning_rate": 0.0,
"loss": 0.3435,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 4.023057813021393e+18,
"train_loss": 0.46734252733740245,
"train_runtime": 28976.3848,
"train_samples_per_second": 5.453,
"train_steps_per_second": 0.043
}
],
"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": 4.023057813021393e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}