qwen3-32b-mo-posttrained / trainer_state.json
obalcells's picture
Add files using upload-large-folder tool
4c19c17 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 0,
"global_step": 2544,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00039308176100628933,
"grad_norm": 0.6179488301277161,
"learning_rate": 5e-06,
"loss": 0.6711,
"step": 1
},
{
"epoch": 0.0007861635220125787,
"grad_norm": 0.608638346195221,
"learning_rate": 4.998034591194969e-06,
"loss": 0.6346,
"step": 2
},
{
"epoch": 0.0011792452830188679,
"grad_norm": 0.8736597299575806,
"learning_rate": 4.996069182389937e-06,
"loss": 0.7717,
"step": 3
},
{
"epoch": 0.0015723270440251573,
"grad_norm": 0.7820959687232971,
"learning_rate": 4.994103773584906e-06,
"loss": 0.6385,
"step": 4
},
{
"epoch": 0.0019654088050314465,
"grad_norm": 0.8728240132331848,
"learning_rate": 4.992138364779874e-06,
"loss": 0.7611,
"step": 5
},
{
"epoch": 0.0023584905660377358,
"grad_norm": 0.8467546105384827,
"learning_rate": 4.990172955974843e-06,
"loss": 0.6658,
"step": 6
},
{
"epoch": 0.002751572327044025,
"grad_norm": 0.6698756814002991,
"learning_rate": 4.988207547169812e-06,
"loss": 0.6047,
"step": 7
},
{
"epoch": 0.0031446540880503146,
"grad_norm": 0.813163161277771,
"learning_rate": 4.98624213836478e-06,
"loss": 0.7491,
"step": 8
},
{
"epoch": 0.003537735849056604,
"grad_norm": 0.6491667628288269,
"learning_rate": 4.984276729559749e-06,
"loss": 0.6431,
"step": 9
},
{
"epoch": 0.003930817610062893,
"grad_norm": 0.5689978003501892,
"learning_rate": 4.9823113207547175e-06,
"loss": 0.5927,
"step": 10
},
{
"epoch": 0.004323899371069182,
"grad_norm": 0.62506103515625,
"learning_rate": 4.980345911949686e-06,
"loss": 0.6352,
"step": 11
},
{
"epoch": 0.0047169811320754715,
"grad_norm": 0.8910884857177734,
"learning_rate": 4.978380503144654e-06,
"loss": 0.7081,
"step": 12
},
{
"epoch": 0.005110062893081761,
"grad_norm": 0.8217619061470032,
"learning_rate": 4.976415094339623e-06,
"loss": 0.766,
"step": 13
},
{
"epoch": 0.00550314465408805,
"grad_norm": 2.929856300354004,
"learning_rate": 4.974449685534591e-06,
"loss": 0.8389,
"step": 14
},
{
"epoch": 0.00589622641509434,
"grad_norm": 0.7581937313079834,
"learning_rate": 4.97248427672956e-06,
"loss": 0.7413,
"step": 15
},
{
"epoch": 0.006289308176100629,
"grad_norm": 0.8592962622642517,
"learning_rate": 4.970518867924529e-06,
"loss": 0.7364,
"step": 16
},
{
"epoch": 0.0066823899371069185,
"grad_norm": 0.9449328780174255,
"learning_rate": 4.968553459119497e-06,
"loss": 0.7131,
"step": 17
},
{
"epoch": 0.007075471698113208,
"grad_norm": 0.754513680934906,
"learning_rate": 4.966588050314466e-06,
"loss": 0.6652,
"step": 18
},
{
"epoch": 0.007468553459119497,
"grad_norm": 0.8711069226264954,
"learning_rate": 4.9646226415094346e-06,
"loss": 0.7313,
"step": 19
},
{
"epoch": 0.007861635220125786,
"grad_norm": 0.5097134709358215,
"learning_rate": 4.962657232704403e-06,
"loss": 0.59,
"step": 20
},
{
"epoch": 0.008254716981132075,
"grad_norm": 0.8065551519393921,
"learning_rate": 4.960691823899371e-06,
"loss": 0.6255,
"step": 21
},
{
"epoch": 0.008647798742138365,
"grad_norm": 0.8206204771995544,
"learning_rate": 4.958726415094339e-06,
"loss": 0.7028,
"step": 22
},
{
"epoch": 0.009040880503144654,
"grad_norm": 0.5936712026596069,
"learning_rate": 4.956761006289308e-06,
"loss": 0.6245,
"step": 23
},
{
"epoch": 0.009433962264150943,
"grad_norm": 0.9390531778335571,
"learning_rate": 4.954795597484277e-06,
"loss": 0.7964,
"step": 24
},
{
"epoch": 0.009827044025157232,
"grad_norm": 0.943632960319519,
"learning_rate": 4.952830188679246e-06,
"loss": 0.8413,
"step": 25
},
{
"epoch": 0.010220125786163521,
"grad_norm": 0.943383514881134,
"learning_rate": 4.950864779874214e-06,
"loss": 0.7278,
"step": 26
},
{
"epoch": 0.01061320754716981,
"grad_norm": 0.8025944828987122,
"learning_rate": 4.948899371069183e-06,
"loss": 0.6554,
"step": 27
},
{
"epoch": 0.0110062893081761,
"grad_norm": 0.8679471015930176,
"learning_rate": 4.946933962264152e-06,
"loss": 0.7201,
"step": 28
},
{
"epoch": 0.01139937106918239,
"grad_norm": 0.6243231892585754,
"learning_rate": 4.94496855345912e-06,
"loss": 0.6312,
"step": 29
},
{
"epoch": 0.01179245283018868,
"grad_norm": 0.7542524337768555,
"learning_rate": 4.9430031446540885e-06,
"loss": 0.7101,
"step": 30
},
{
"epoch": 0.01218553459119497,
"grad_norm": 0.5379254221916199,
"learning_rate": 4.9410377358490565e-06,
"loss": 0.6923,
"step": 31
},
{
"epoch": 0.012578616352201259,
"grad_norm": 0.7621724009513855,
"learning_rate": 4.939072327044025e-06,
"loss": 0.5914,
"step": 32
},
{
"epoch": 0.012971698113207548,
"grad_norm": 0.6130362153053284,
"learning_rate": 4.937106918238994e-06,
"loss": 0.6196,
"step": 33
},
{
"epoch": 0.013364779874213837,
"grad_norm": 0.9599727392196655,
"learning_rate": 4.935141509433963e-06,
"loss": 0.7632,
"step": 34
},
{
"epoch": 0.013757861635220126,
"grad_norm": 0.7335894703865051,
"learning_rate": 4.933176100628931e-06,
"loss": 0.6306,
"step": 35
},
{
"epoch": 0.014150943396226415,
"grad_norm": 0.826854944229126,
"learning_rate": 4.9312106918239e-06,
"loss": 0.7473,
"step": 36
},
{
"epoch": 0.014544025157232705,
"grad_norm": 0.824235200881958,
"learning_rate": 4.929245283018868e-06,
"loss": 0.7596,
"step": 37
},
{
"epoch": 0.014937106918238994,
"grad_norm": 0.8440677523612976,
"learning_rate": 4.927279874213837e-06,
"loss": 0.7103,
"step": 38
},
{
"epoch": 0.015330188679245283,
"grad_norm": 0.5619874596595764,
"learning_rate": 4.9253144654088056e-06,
"loss": 0.5681,
"step": 39
},
{
"epoch": 0.015723270440251572,
"grad_norm": 0.5799058675765991,
"learning_rate": 4.9233490566037736e-06,
"loss": 0.5679,
"step": 40
},
{
"epoch": 0.01611635220125786,
"grad_norm": 0.6457942724227905,
"learning_rate": 4.921383647798742e-06,
"loss": 0.6299,
"step": 41
},
{
"epoch": 0.01650943396226415,
"grad_norm": 0.8218834400177002,
"learning_rate": 4.919418238993711e-06,
"loss": 0.6707,
"step": 42
},
{
"epoch": 0.01690251572327044,
"grad_norm": 0.9053443074226379,
"learning_rate": 4.91745283018868e-06,
"loss": 0.7555,
"step": 43
},
{
"epoch": 0.01729559748427673,
"grad_norm": 0.994422972202301,
"learning_rate": 4.915487421383648e-06,
"loss": 0.842,
"step": 44
},
{
"epoch": 0.01768867924528302,
"grad_norm": 0.7779326438903809,
"learning_rate": 4.913522012578616e-06,
"loss": 0.7549,
"step": 45
},
{
"epoch": 0.018081761006289308,
"grad_norm": 0.7225157022476196,
"learning_rate": 4.911556603773585e-06,
"loss": 0.7462,
"step": 46
},
{
"epoch": 0.018474842767295597,
"grad_norm": 0.7550550103187561,
"learning_rate": 4.909591194968554e-06,
"loss": 0.649,
"step": 47
},
{
"epoch": 0.018867924528301886,
"grad_norm": 0.9289947748184204,
"learning_rate": 4.907625786163523e-06,
"loss": 0.6272,
"step": 48
},
{
"epoch": 0.019261006289308175,
"grad_norm": 1.0065691471099854,
"learning_rate": 4.905660377358491e-06,
"loss": 0.7957,
"step": 49
},
{
"epoch": 0.019654088050314465,
"grad_norm": 0.8711187839508057,
"learning_rate": 4.9036949685534595e-06,
"loss": 0.7571,
"step": 50
},
{
"epoch": 0.020047169811320754,
"grad_norm": 0.6377099752426147,
"learning_rate": 4.901729559748428e-06,
"loss": 0.5242,
"step": 51
},
{
"epoch": 0.020440251572327043,
"grad_norm": 0.7619996070861816,
"learning_rate": 4.899764150943397e-06,
"loss": 0.7809,
"step": 52
},
{
"epoch": 0.020833333333333332,
"grad_norm": 0.5642449259757996,
"learning_rate": 4.897798742138365e-06,
"loss": 0.5511,
"step": 53
},
{
"epoch": 0.02122641509433962,
"grad_norm": 0.8692660927772522,
"learning_rate": 4.895833333333333e-06,
"loss": 0.8009,
"step": 54
},
{
"epoch": 0.02161949685534591,
"grad_norm": 0.810824990272522,
"learning_rate": 4.893867924528302e-06,
"loss": 0.6917,
"step": 55
},
{
"epoch": 0.0220125786163522,
"grad_norm": 0.8436955213546753,
"learning_rate": 4.891902515723271e-06,
"loss": 0.7338,
"step": 56
},
{
"epoch": 0.02240566037735849,
"grad_norm": 0.8777972459793091,
"learning_rate": 4.88993710691824e-06,
"loss": 0.8016,
"step": 57
},
{
"epoch": 0.02279874213836478,
"grad_norm": 0.728876531124115,
"learning_rate": 4.887971698113208e-06,
"loss": 0.7344,
"step": 58
},
{
"epoch": 0.023191823899371068,
"grad_norm": 0.6115991473197937,
"learning_rate": 4.8860062893081766e-06,
"loss": 0.6687,
"step": 59
},
{
"epoch": 0.02358490566037736,
"grad_norm": 0.7880122661590576,
"learning_rate": 4.884040880503145e-06,
"loss": 0.6835,
"step": 60
},
{
"epoch": 0.02397798742138365,
"grad_norm": 0.8009531497955322,
"learning_rate": 4.882075471698113e-06,
"loss": 0.7227,
"step": 61
},
{
"epoch": 0.02437106918238994,
"grad_norm": 0.8650614023208618,
"learning_rate": 4.880110062893082e-06,
"loss": 0.7894,
"step": 62
},
{
"epoch": 0.024764150943396228,
"grad_norm": 0.8107253909111023,
"learning_rate": 4.87814465408805e-06,
"loss": 0.7877,
"step": 63
},
{
"epoch": 0.025157232704402517,
"grad_norm": 0.54973965883255,
"learning_rate": 4.876179245283019e-06,
"loss": 0.6527,
"step": 64
},
{
"epoch": 0.025550314465408806,
"grad_norm": 0.7287852168083191,
"learning_rate": 4.874213836477988e-06,
"loss": 0.7399,
"step": 65
},
{
"epoch": 0.025943396226415096,
"grad_norm": 0.6669225692749023,
"learning_rate": 4.872248427672957e-06,
"loss": 0.7235,
"step": 66
},
{
"epoch": 0.026336477987421385,
"grad_norm": 0.8761881589889526,
"learning_rate": 4.870283018867925e-06,
"loss": 0.7705,
"step": 67
},
{
"epoch": 0.026729559748427674,
"grad_norm": 0.7960617542266846,
"learning_rate": 4.868317610062894e-06,
"loss": 0.7063,
"step": 68
},
{
"epoch": 0.027122641509433963,
"grad_norm": 0.853482723236084,
"learning_rate": 4.866352201257862e-06,
"loss": 0.7587,
"step": 69
},
{
"epoch": 0.027515723270440252,
"grad_norm": 0.6334282755851746,
"learning_rate": 4.8643867924528305e-06,
"loss": 0.6111,
"step": 70
},
{
"epoch": 0.02790880503144654,
"grad_norm": 0.7821253538131714,
"learning_rate": 4.862421383647799e-06,
"loss": 0.7218,
"step": 71
},
{
"epoch": 0.02830188679245283,
"grad_norm": 0.8890393376350403,
"learning_rate": 4.860455974842767e-06,
"loss": 0.6888,
"step": 72
},
{
"epoch": 0.02869496855345912,
"grad_norm": 0.7255067229270935,
"learning_rate": 4.858490566037736e-06,
"loss": 0.7819,
"step": 73
},
{
"epoch": 0.02908805031446541,
"grad_norm": 0.8044862747192383,
"learning_rate": 4.856525157232705e-06,
"loss": 0.6712,
"step": 74
},
{
"epoch": 0.0294811320754717,
"grad_norm": 0.8763513565063477,
"learning_rate": 4.854559748427674e-06,
"loss": 0.7362,
"step": 75
},
{
"epoch": 0.029874213836477988,
"grad_norm": 0.8919411301612854,
"learning_rate": 4.852594339622642e-06,
"loss": 0.7703,
"step": 76
},
{
"epoch": 0.030267295597484277,
"grad_norm": 0.5968775153160095,
"learning_rate": 4.85062893081761e-06,
"loss": 0.5671,
"step": 77
},
{
"epoch": 0.030660377358490566,
"grad_norm": 0.9242513179779053,
"learning_rate": 4.848663522012579e-06,
"loss": 0.7903,
"step": 78
},
{
"epoch": 0.031053459119496855,
"grad_norm": 0.7358946204185486,
"learning_rate": 4.8466981132075476e-06,
"loss": 0.6479,
"step": 79
},
{
"epoch": 0.031446540880503145,
"grad_norm": 0.7624842524528503,
"learning_rate": 4.844732704402516e-06,
"loss": 0.6906,
"step": 80
},
{
"epoch": 0.031839622641509434,
"grad_norm": 0.8793920278549194,
"learning_rate": 4.842767295597484e-06,
"loss": 0.7352,
"step": 81
},
{
"epoch": 0.03223270440251572,
"grad_norm": 0.8826913833618164,
"learning_rate": 4.840801886792453e-06,
"loss": 0.6802,
"step": 82
},
{
"epoch": 0.03262578616352201,
"grad_norm": 0.7771472930908203,
"learning_rate": 4.838836477987422e-06,
"loss": 0.7369,
"step": 83
},
{
"epoch": 0.0330188679245283,
"grad_norm": 0.7909108400344849,
"learning_rate": 4.836871069182391e-06,
"loss": 0.6792,
"step": 84
},
{
"epoch": 0.03341194968553459,
"grad_norm": 0.6148988008499146,
"learning_rate": 4.834905660377359e-06,
"loss": 0.6902,
"step": 85
},
{
"epoch": 0.03380503144654088,
"grad_norm": 0.5967656970024109,
"learning_rate": 4.832940251572327e-06,
"loss": 0.604,
"step": 86
},
{
"epoch": 0.03419811320754717,
"grad_norm": 0.7446857690811157,
"learning_rate": 4.830974842767296e-06,
"loss": 0.6957,
"step": 87
},
{
"epoch": 0.03459119496855346,
"grad_norm": 0.9904138445854187,
"learning_rate": 4.829009433962265e-06,
"loss": 0.7777,
"step": 88
},
{
"epoch": 0.03498427672955975,
"grad_norm": 0.7725400924682617,
"learning_rate": 4.8270440251572335e-06,
"loss": 0.7172,
"step": 89
},
{
"epoch": 0.03537735849056604,
"grad_norm": 0.5844003558158875,
"learning_rate": 4.8250786163522015e-06,
"loss": 0.5907,
"step": 90
},
{
"epoch": 0.035770440251572326,
"grad_norm": 0.5471771955490112,
"learning_rate": 4.82311320754717e-06,
"loss": 0.6292,
"step": 91
},
{
"epoch": 0.036163522012578615,
"grad_norm": 0.8446871638298035,
"learning_rate": 4.821147798742139e-06,
"loss": 0.751,
"step": 92
},
{
"epoch": 0.036556603773584904,
"grad_norm": 0.8345248103141785,
"learning_rate": 4.819182389937107e-06,
"loss": 0.7633,
"step": 93
},
{
"epoch": 0.036949685534591194,
"grad_norm": 0.6620730757713318,
"learning_rate": 4.817216981132076e-06,
"loss": 0.6758,
"step": 94
},
{
"epoch": 0.03734276729559748,
"grad_norm": 0.9627453684806824,
"learning_rate": 4.815251572327044e-06,
"loss": 0.7996,
"step": 95
},
{
"epoch": 0.03773584905660377,
"grad_norm": 0.7127353549003601,
"learning_rate": 4.813286163522013e-06,
"loss": 0.6164,
"step": 96
},
{
"epoch": 0.03812893081761006,
"grad_norm": 0.7763054966926575,
"learning_rate": 4.811320754716982e-06,
"loss": 0.7451,
"step": 97
},
{
"epoch": 0.03852201257861635,
"grad_norm": 0.7480806708335876,
"learning_rate": 4.809355345911951e-06,
"loss": 0.7245,
"step": 98
},
{
"epoch": 0.03891509433962264,
"grad_norm": 0.643818736076355,
"learning_rate": 4.807389937106919e-06,
"loss": 0.6136,
"step": 99
},
{
"epoch": 0.03930817610062893,
"grad_norm": 0.8345184922218323,
"learning_rate": 4.8054245283018874e-06,
"loss": 0.8518,
"step": 100
},
{
"epoch": 0.03970125786163522,
"grad_norm": 0.6038028001785278,
"learning_rate": 4.8034591194968554e-06,
"loss": 0.6947,
"step": 101
},
{
"epoch": 0.04009433962264151,
"grad_norm": 0.725440263748169,
"learning_rate": 4.801493710691824e-06,
"loss": 0.7538,
"step": 102
},
{
"epoch": 0.0404874213836478,
"grad_norm": 0.780044436454773,
"learning_rate": 4.799528301886793e-06,
"loss": 0.7612,
"step": 103
},
{
"epoch": 0.040880503144654086,
"grad_norm": 0.7466725707054138,
"learning_rate": 4.797562893081761e-06,
"loss": 0.7378,
"step": 104
},
{
"epoch": 0.041273584905660375,
"grad_norm": 0.6727305054664612,
"learning_rate": 4.79559748427673e-06,
"loss": 0.6899,
"step": 105
},
{
"epoch": 0.041666666666666664,
"grad_norm": 0.5716968774795532,
"learning_rate": 4.793632075471699e-06,
"loss": 0.6941,
"step": 106
},
{
"epoch": 0.042059748427672954,
"grad_norm": 0.7646844983100891,
"learning_rate": 4.791666666666668e-06,
"loss": 0.7435,
"step": 107
},
{
"epoch": 0.04245283018867924,
"grad_norm": 0.8681058287620544,
"learning_rate": 4.789701257861636e-06,
"loss": 0.6819,
"step": 108
},
{
"epoch": 0.04284591194968553,
"grad_norm": 0.7446102499961853,
"learning_rate": 4.787735849056604e-06,
"loss": 0.8048,
"step": 109
},
{
"epoch": 0.04323899371069182,
"grad_norm": 0.6049733757972717,
"learning_rate": 4.7857704402515725e-06,
"loss": 0.6671,
"step": 110
},
{
"epoch": 0.04363207547169811,
"grad_norm": 0.7207951545715332,
"learning_rate": 4.783805031446541e-06,
"loss": 0.7046,
"step": 111
},
{
"epoch": 0.0440251572327044,
"grad_norm": 1.1246774196624756,
"learning_rate": 4.78183962264151e-06,
"loss": 0.7162,
"step": 112
},
{
"epoch": 0.04441823899371069,
"grad_norm": 0.8429887890815735,
"learning_rate": 4.779874213836478e-06,
"loss": 0.7881,
"step": 113
},
{
"epoch": 0.04481132075471698,
"grad_norm": 0.6717969179153442,
"learning_rate": 4.777908805031447e-06,
"loss": 0.6929,
"step": 114
},
{
"epoch": 0.04520440251572327,
"grad_norm": 0.6223211884498596,
"learning_rate": 4.775943396226416e-06,
"loss": 0.7129,
"step": 115
},
{
"epoch": 0.04559748427672956,
"grad_norm": 0.7531170845031738,
"learning_rate": 4.773977987421384e-06,
"loss": 0.6495,
"step": 116
},
{
"epoch": 0.045990566037735846,
"grad_norm": 0.811154305934906,
"learning_rate": 4.772012578616353e-06,
"loss": 0.6907,
"step": 117
},
{
"epoch": 0.046383647798742135,
"grad_norm": 0.617703914642334,
"learning_rate": 4.770047169811321e-06,
"loss": 0.6547,
"step": 118
},
{
"epoch": 0.046776729559748424,
"grad_norm": 0.704272985458374,
"learning_rate": 4.76808176100629e-06,
"loss": 0.6847,
"step": 119
},
{
"epoch": 0.04716981132075472,
"grad_norm": 0.5939571261405945,
"learning_rate": 4.7661163522012584e-06,
"loss": 0.5115,
"step": 120
},
{
"epoch": 0.04756289308176101,
"grad_norm": 0.8179329037666321,
"learning_rate": 4.764150943396227e-06,
"loss": 0.6877,
"step": 121
},
{
"epoch": 0.0479559748427673,
"grad_norm": 0.6399481892585754,
"learning_rate": 4.762185534591195e-06,
"loss": 0.6536,
"step": 122
},
{
"epoch": 0.04834905660377359,
"grad_norm": 0.7095329165458679,
"learning_rate": 4.760220125786164e-06,
"loss": 0.702,
"step": 123
},
{
"epoch": 0.04874213836477988,
"grad_norm": 0.5246897339820862,
"learning_rate": 4.758254716981133e-06,
"loss": 0.5748,
"step": 124
},
{
"epoch": 0.049135220125786166,
"grad_norm": 0.7650740742683411,
"learning_rate": 4.756289308176101e-06,
"loss": 0.66,
"step": 125
},
{
"epoch": 0.049528301886792456,
"grad_norm": 0.6977266669273376,
"learning_rate": 4.75432389937107e-06,
"loss": 0.6854,
"step": 126
},
{
"epoch": 0.049921383647798745,
"grad_norm": 0.7853943109512329,
"learning_rate": 4.752358490566038e-06,
"loss": 0.6482,
"step": 127
},
{
"epoch": 0.050314465408805034,
"grad_norm": 0.6174890398979187,
"learning_rate": 4.750393081761007e-06,
"loss": 0.6224,
"step": 128
},
{
"epoch": 0.05070754716981132,
"grad_norm": 0.7074907422065735,
"learning_rate": 4.7484276729559755e-06,
"loss": 0.7791,
"step": 129
},
{
"epoch": 0.05110062893081761,
"grad_norm": 0.582131564617157,
"learning_rate": 4.7464622641509435e-06,
"loss": 0.6559,
"step": 130
},
{
"epoch": 0.0514937106918239,
"grad_norm": 0.6367083787918091,
"learning_rate": 4.744496855345912e-06,
"loss": 0.6907,
"step": 131
},
{
"epoch": 0.05188679245283019,
"grad_norm": 0.7958084940910339,
"learning_rate": 4.742531446540881e-06,
"loss": 0.6969,
"step": 132
},
{
"epoch": 0.05227987421383648,
"grad_norm": 0.7190726399421692,
"learning_rate": 4.740566037735849e-06,
"loss": 0.6497,
"step": 133
},
{
"epoch": 0.05267295597484277,
"grad_norm": 0.7292737364768982,
"learning_rate": 4.738600628930818e-06,
"loss": 0.6986,
"step": 134
},
{
"epoch": 0.05306603773584906,
"grad_norm": 0.7295096516609192,
"learning_rate": 4.736635220125787e-06,
"loss": 0.6405,
"step": 135
},
{
"epoch": 0.05345911949685535,
"grad_norm": 0.6752004623413086,
"learning_rate": 4.734669811320755e-06,
"loss": 0.6729,
"step": 136
},
{
"epoch": 0.05385220125786164,
"grad_norm": 0.7118735909461975,
"learning_rate": 4.732704402515724e-06,
"loss": 0.7154,
"step": 137
},
{
"epoch": 0.054245283018867926,
"grad_norm": 0.6835665106773376,
"learning_rate": 4.730738993710693e-06,
"loss": 0.7332,
"step": 138
},
{
"epoch": 0.054638364779874216,
"grad_norm": 0.5906879305839539,
"learning_rate": 4.728773584905661e-06,
"loss": 0.6506,
"step": 139
},
{
"epoch": 0.055031446540880505,
"grad_norm": 0.6482188701629639,
"learning_rate": 4.7268081761006294e-06,
"loss": 0.6625,
"step": 140
},
{
"epoch": 0.055424528301886794,
"grad_norm": 0.6386060118675232,
"learning_rate": 4.7248427672955974e-06,
"loss": 0.671,
"step": 141
},
{
"epoch": 0.05581761006289308,
"grad_norm": 0.8564355373382568,
"learning_rate": 4.722877358490566e-06,
"loss": 0.7994,
"step": 142
},
{
"epoch": 0.05621069182389937,
"grad_norm": 0.9191057085990906,
"learning_rate": 4.720911949685535e-06,
"loss": 0.7847,
"step": 143
},
{
"epoch": 0.05660377358490566,
"grad_norm": 0.9450004696846008,
"learning_rate": 4.718946540880503e-06,
"loss": 0.8885,
"step": 144
},
{
"epoch": 0.05699685534591195,
"grad_norm": 0.7979057431221008,
"learning_rate": 4.716981132075472e-06,
"loss": 0.728,
"step": 145
},
{
"epoch": 0.05738993710691824,
"grad_norm": 0.6612402200698853,
"learning_rate": 4.715015723270441e-06,
"loss": 0.7153,
"step": 146
},
{
"epoch": 0.05778301886792453,
"grad_norm": 0.7722911238670349,
"learning_rate": 4.71305031446541e-06,
"loss": 0.7354,
"step": 147
},
{
"epoch": 0.05817610062893082,
"grad_norm": 0.6579909324645996,
"learning_rate": 4.711084905660378e-06,
"loss": 0.6978,
"step": 148
},
{
"epoch": 0.05856918238993711,
"grad_norm": 0.5801395773887634,
"learning_rate": 4.709119496855346e-06,
"loss": 0.6306,
"step": 149
},
{
"epoch": 0.0589622641509434,
"grad_norm": 0.636152446269989,
"learning_rate": 4.7071540880503145e-06,
"loss": 0.6385,
"step": 150
},
{
"epoch": 0.059355345911949686,
"grad_norm": 0.6418371796607971,
"learning_rate": 4.705188679245283e-06,
"loss": 0.6696,
"step": 151
},
{
"epoch": 0.059748427672955975,
"grad_norm": 0.7373504042625427,
"learning_rate": 4.703223270440252e-06,
"loss": 0.759,
"step": 152
},
{
"epoch": 0.060141509433962265,
"grad_norm": 0.8479204773902893,
"learning_rate": 4.70125786163522e-06,
"loss": 0.7603,
"step": 153
},
{
"epoch": 0.060534591194968554,
"grad_norm": 0.8076416254043579,
"learning_rate": 4.699292452830189e-06,
"loss": 0.6841,
"step": 154
},
{
"epoch": 0.06092767295597484,
"grad_norm": 0.850490152835846,
"learning_rate": 4.697327044025158e-06,
"loss": 0.8684,
"step": 155
},
{
"epoch": 0.06132075471698113,
"grad_norm": 0.9423735737800598,
"learning_rate": 4.695361635220127e-06,
"loss": 0.7566,
"step": 156
},
{
"epoch": 0.06171383647798742,
"grad_norm": 0.8819036483764648,
"learning_rate": 4.693396226415095e-06,
"loss": 0.7559,
"step": 157
},
{
"epoch": 0.06210691823899371,
"grad_norm": 0.7582936882972717,
"learning_rate": 4.691430817610063e-06,
"loss": 0.6666,
"step": 158
},
{
"epoch": 0.0625,
"grad_norm": 0.890673816204071,
"learning_rate": 4.689465408805032e-06,
"loss": 0.8589,
"step": 159
},
{
"epoch": 0.06289308176100629,
"grad_norm": 0.739288866519928,
"learning_rate": 4.6875000000000004e-06,
"loss": 0.6166,
"step": 160
},
{
"epoch": 0.06328616352201258,
"grad_norm": 0.8375418186187744,
"learning_rate": 4.685534591194969e-06,
"loss": 0.7561,
"step": 161
},
{
"epoch": 0.06367924528301887,
"grad_norm": 0.7377603054046631,
"learning_rate": 4.683569182389937e-06,
"loss": 0.679,
"step": 162
},
{
"epoch": 0.06407232704402516,
"grad_norm": 0.47440454363822937,
"learning_rate": 4.681603773584906e-06,
"loss": 0.588,
"step": 163
},
{
"epoch": 0.06446540880503145,
"grad_norm": 0.6441667675971985,
"learning_rate": 4.679638364779874e-06,
"loss": 0.6479,
"step": 164
},
{
"epoch": 0.06485849056603774,
"grad_norm": 0.6721105575561523,
"learning_rate": 4.677672955974843e-06,
"loss": 0.698,
"step": 165
},
{
"epoch": 0.06525157232704402,
"grad_norm": 0.6793725490570068,
"learning_rate": 4.675707547169812e-06,
"loss": 0.7317,
"step": 166
},
{
"epoch": 0.06564465408805031,
"grad_norm": 0.6439410448074341,
"learning_rate": 4.67374213836478e-06,
"loss": 0.6335,
"step": 167
},
{
"epoch": 0.0660377358490566,
"grad_norm": 0.9626488089561462,
"learning_rate": 4.671776729559749e-06,
"loss": 0.6908,
"step": 168
},
{
"epoch": 0.06643081761006289,
"grad_norm": 0.8366047739982605,
"learning_rate": 4.6698113207547175e-06,
"loss": 0.7689,
"step": 169
},
{
"epoch": 0.06682389937106918,
"grad_norm": 0.8190625309944153,
"learning_rate": 4.667845911949686e-06,
"loss": 0.8159,
"step": 170
},
{
"epoch": 0.06721698113207547,
"grad_norm": 0.6961092352867126,
"learning_rate": 4.665880503144654e-06,
"loss": 0.6823,
"step": 171
},
{
"epoch": 0.06761006289308176,
"grad_norm": 0.6719902157783508,
"learning_rate": 4.663915094339622e-06,
"loss": 0.6265,
"step": 172
},
{
"epoch": 0.06800314465408805,
"grad_norm": 0.7299097776412964,
"learning_rate": 4.661949685534591e-06,
"loss": 0.6841,
"step": 173
},
{
"epoch": 0.06839622641509434,
"grad_norm": 0.7970944046974182,
"learning_rate": 4.65998427672956e-06,
"loss": 0.7457,
"step": 174
},
{
"epoch": 0.06878930817610063,
"grad_norm": 0.7069805264472961,
"learning_rate": 4.658018867924529e-06,
"loss": 0.7327,
"step": 175
},
{
"epoch": 0.06918238993710692,
"grad_norm": 0.7414055466651917,
"learning_rate": 4.656053459119497e-06,
"loss": 0.7568,
"step": 176
},
{
"epoch": 0.0695754716981132,
"grad_norm": 0.49029892683029175,
"learning_rate": 4.654088050314466e-06,
"loss": 0.5729,
"step": 177
},
{
"epoch": 0.0699685534591195,
"grad_norm": 0.6802594661712646,
"learning_rate": 4.652122641509435e-06,
"loss": 0.675,
"step": 178
},
{
"epoch": 0.07036163522012578,
"grad_norm": 0.7009592652320862,
"learning_rate": 4.6501572327044034e-06,
"loss": 0.6595,
"step": 179
},
{
"epoch": 0.07075471698113207,
"grad_norm": 0.641412615776062,
"learning_rate": 4.6481918238993714e-06,
"loss": 0.736,
"step": 180
},
{
"epoch": 0.07114779874213836,
"grad_norm": 0.6692111492156982,
"learning_rate": 4.6462264150943394e-06,
"loss": 0.6133,
"step": 181
},
{
"epoch": 0.07154088050314465,
"grad_norm": 0.7276618480682373,
"learning_rate": 4.644261006289308e-06,
"loss": 0.7415,
"step": 182
},
{
"epoch": 0.07193396226415094,
"grad_norm": 0.7225254774093628,
"learning_rate": 4.642295597484277e-06,
"loss": 0.6454,
"step": 183
},
{
"epoch": 0.07232704402515723,
"grad_norm": 0.7878135442733765,
"learning_rate": 4.640330188679246e-06,
"loss": 0.7739,
"step": 184
},
{
"epoch": 0.07272012578616352,
"grad_norm": 0.6281365156173706,
"learning_rate": 4.638364779874214e-06,
"loss": 0.6871,
"step": 185
},
{
"epoch": 0.07311320754716981,
"grad_norm": 0.5062946081161499,
"learning_rate": 4.636399371069183e-06,
"loss": 0.6082,
"step": 186
},
{
"epoch": 0.0735062893081761,
"grad_norm": 0.8338656425476074,
"learning_rate": 4.634433962264152e-06,
"loss": 0.7692,
"step": 187
},
{
"epoch": 0.07389937106918239,
"grad_norm": 0.8063732385635376,
"learning_rate": 4.63246855345912e-06,
"loss": 0.7289,
"step": 188
},
{
"epoch": 0.07429245283018868,
"grad_norm": 0.7640451788902283,
"learning_rate": 4.6305031446540885e-06,
"loss": 0.6945,
"step": 189
},
{
"epoch": 0.07468553459119497,
"grad_norm": 0.6079548597335815,
"learning_rate": 4.6285377358490565e-06,
"loss": 0.7091,
"step": 190
},
{
"epoch": 0.07507861635220126,
"grad_norm": 0.6782593131065369,
"learning_rate": 4.626572327044025e-06,
"loss": 0.7146,
"step": 191
},
{
"epoch": 0.07547169811320754,
"grad_norm": 0.7054331302642822,
"learning_rate": 4.624606918238994e-06,
"loss": 0.6718,
"step": 192
},
{
"epoch": 0.07586477987421383,
"grad_norm": 0.8395931720733643,
"learning_rate": 4.622641509433963e-06,
"loss": 0.788,
"step": 193
},
{
"epoch": 0.07625786163522012,
"grad_norm": 0.5841469168663025,
"learning_rate": 4.620676100628931e-06,
"loss": 0.5985,
"step": 194
},
{
"epoch": 0.07665094339622641,
"grad_norm": 0.7057890295982361,
"learning_rate": 4.6187106918239e-06,
"loss": 0.7399,
"step": 195
},
{
"epoch": 0.0770440251572327,
"grad_norm": 0.8152762055397034,
"learning_rate": 4.616745283018868e-06,
"loss": 0.7198,
"step": 196
},
{
"epoch": 0.07743710691823899,
"grad_norm": 0.7154718637466431,
"learning_rate": 4.614779874213837e-06,
"loss": 0.6707,
"step": 197
},
{
"epoch": 0.07783018867924528,
"grad_norm": 0.6757956743240356,
"learning_rate": 4.612814465408806e-06,
"loss": 0.6888,
"step": 198
},
{
"epoch": 0.07822327044025157,
"grad_norm": 0.5570179224014282,
"learning_rate": 4.610849056603774e-06,
"loss": 0.7266,
"step": 199
},
{
"epoch": 0.07861635220125786,
"grad_norm": 0.7227308750152588,
"learning_rate": 4.6088836477987424e-06,
"loss": 0.6412,
"step": 200
},
{
"epoch": 0.07900943396226415,
"grad_norm": 0.779802143573761,
"learning_rate": 4.606918238993711e-06,
"loss": 0.6805,
"step": 201
},
{
"epoch": 0.07940251572327044,
"grad_norm": 0.7713460922241211,
"learning_rate": 4.60495283018868e-06,
"loss": 0.7828,
"step": 202
},
{
"epoch": 0.07979559748427673,
"grad_norm": 0.8046134114265442,
"learning_rate": 4.602987421383648e-06,
"loss": 0.746,
"step": 203
},
{
"epoch": 0.08018867924528301,
"grad_norm": 0.8605005145072937,
"learning_rate": 4.601022012578616e-06,
"loss": 0.7178,
"step": 204
},
{
"epoch": 0.0805817610062893,
"grad_norm": 0.8474591374397278,
"learning_rate": 4.599056603773585e-06,
"loss": 0.7684,
"step": 205
},
{
"epoch": 0.0809748427672956,
"grad_norm": 0.6036484241485596,
"learning_rate": 4.597091194968554e-06,
"loss": 0.6697,
"step": 206
},
{
"epoch": 0.08136792452830188,
"grad_norm": 0.6357397437095642,
"learning_rate": 4.595125786163523e-06,
"loss": 0.7092,
"step": 207
},
{
"epoch": 0.08176100628930817,
"grad_norm": 0.7884801030158997,
"learning_rate": 4.593160377358491e-06,
"loss": 0.7166,
"step": 208
},
{
"epoch": 0.08215408805031446,
"grad_norm": 0.8826138973236084,
"learning_rate": 4.5911949685534595e-06,
"loss": 0.7295,
"step": 209
},
{
"epoch": 0.08254716981132075,
"grad_norm": 0.8796142935752869,
"learning_rate": 4.589229559748428e-06,
"loss": 0.7467,
"step": 210
},
{
"epoch": 0.08294025157232704,
"grad_norm": 0.7213162183761597,
"learning_rate": 4.587264150943397e-06,
"loss": 0.7449,
"step": 211
},
{
"epoch": 0.08333333333333333,
"grad_norm": 0.726154625415802,
"learning_rate": 4.585298742138365e-06,
"loss": 0.7174,
"step": 212
},
{
"epoch": 0.08372641509433962,
"grad_norm": 0.7808687686920166,
"learning_rate": 4.583333333333333e-06,
"loss": 0.7424,
"step": 213
},
{
"epoch": 0.08411949685534591,
"grad_norm": 0.7380768656730652,
"learning_rate": 4.581367924528302e-06,
"loss": 0.6517,
"step": 214
},
{
"epoch": 0.0845125786163522,
"grad_norm": 0.7482620477676392,
"learning_rate": 4.579402515723271e-06,
"loss": 0.7987,
"step": 215
},
{
"epoch": 0.08490566037735849,
"grad_norm": 0.6737194061279297,
"learning_rate": 4.57743710691824e-06,
"loss": 0.7814,
"step": 216
},
{
"epoch": 0.08529874213836477,
"grad_norm": 0.5724939703941345,
"learning_rate": 4.575471698113208e-06,
"loss": 0.6881,
"step": 217
},
{
"epoch": 0.08569182389937106,
"grad_norm": 0.7268404960632324,
"learning_rate": 4.573506289308177e-06,
"loss": 0.635,
"step": 218
},
{
"epoch": 0.08608490566037735,
"grad_norm": 0.80220627784729,
"learning_rate": 4.5715408805031454e-06,
"loss": 0.6562,
"step": 219
},
{
"epoch": 0.08647798742138364,
"grad_norm": 0.6595987677574158,
"learning_rate": 4.5695754716981134e-06,
"loss": 0.7667,
"step": 220
},
{
"epoch": 0.08687106918238993,
"grad_norm": 0.6321592330932617,
"learning_rate": 4.567610062893082e-06,
"loss": 0.6549,
"step": 221
},
{
"epoch": 0.08726415094339622,
"grad_norm": 0.6521388292312622,
"learning_rate": 4.56564465408805e-06,
"loss": 0.6896,
"step": 222
},
{
"epoch": 0.08765723270440251,
"grad_norm": 0.7109624147415161,
"learning_rate": 4.563679245283019e-06,
"loss": 0.6206,
"step": 223
},
{
"epoch": 0.0880503144654088,
"grad_norm": 0.763735830783844,
"learning_rate": 4.561713836477988e-06,
"loss": 0.7069,
"step": 224
},
{
"epoch": 0.08844339622641509,
"grad_norm": 0.6934118866920471,
"learning_rate": 4.559748427672957e-06,
"loss": 0.6804,
"step": 225
},
{
"epoch": 0.08883647798742138,
"grad_norm": 0.8565337061882019,
"learning_rate": 4.557783018867925e-06,
"loss": 0.8271,
"step": 226
},
{
"epoch": 0.08922955974842767,
"grad_norm": 0.7634615898132324,
"learning_rate": 4.555817610062894e-06,
"loss": 0.743,
"step": 227
},
{
"epoch": 0.08962264150943396,
"grad_norm": 0.7572429776191711,
"learning_rate": 4.553852201257862e-06,
"loss": 0.7626,
"step": 228
},
{
"epoch": 0.09001572327044025,
"grad_norm": 0.6485528945922852,
"learning_rate": 4.5518867924528305e-06,
"loss": 0.5809,
"step": 229
},
{
"epoch": 0.09040880503144653,
"grad_norm": 0.886944591999054,
"learning_rate": 4.549921383647799e-06,
"loss": 0.6669,
"step": 230
},
{
"epoch": 0.09080188679245282,
"grad_norm": 0.6645367741584778,
"learning_rate": 4.547955974842767e-06,
"loss": 0.6842,
"step": 231
},
{
"epoch": 0.09119496855345911,
"grad_norm": 0.5729372501373291,
"learning_rate": 4.545990566037736e-06,
"loss": 0.6792,
"step": 232
},
{
"epoch": 0.0915880503144654,
"grad_norm": 0.6556676030158997,
"learning_rate": 4.544025157232705e-06,
"loss": 0.6453,
"step": 233
},
{
"epoch": 0.09198113207547169,
"grad_norm": 0.5750808715820312,
"learning_rate": 4.542059748427674e-06,
"loss": 0.6589,
"step": 234
},
{
"epoch": 0.09237421383647798,
"grad_norm": 0.9425899982452393,
"learning_rate": 4.540094339622642e-06,
"loss": 0.8726,
"step": 235
},
{
"epoch": 0.09276729559748427,
"grad_norm": 0.8481728434562683,
"learning_rate": 4.53812893081761e-06,
"loss": 0.8292,
"step": 236
},
{
"epoch": 0.09316037735849056,
"grad_norm": 0.8018584251403809,
"learning_rate": 4.536163522012579e-06,
"loss": 0.7986,
"step": 237
},
{
"epoch": 0.09355345911949685,
"grad_norm": 0.8461182117462158,
"learning_rate": 4.534198113207548e-06,
"loss": 0.7869,
"step": 238
},
{
"epoch": 0.09394654088050315,
"grad_norm": 0.7935497164726257,
"learning_rate": 4.5322327044025165e-06,
"loss": 0.7203,
"step": 239
},
{
"epoch": 0.09433962264150944,
"grad_norm": 0.6823428273200989,
"learning_rate": 4.5302672955974844e-06,
"loss": 0.6635,
"step": 240
},
{
"epoch": 0.09473270440251573,
"grad_norm": 0.5465632081031799,
"learning_rate": 4.528301886792453e-06,
"loss": 0.6232,
"step": 241
},
{
"epoch": 0.09512578616352202,
"grad_norm": 0.6571717858314514,
"learning_rate": 4.526336477987422e-06,
"loss": 0.7142,
"step": 242
},
{
"epoch": 0.09551886792452831,
"grad_norm": 0.7379636764526367,
"learning_rate": 4.52437106918239e-06,
"loss": 0.7079,
"step": 243
},
{
"epoch": 0.0959119496855346,
"grad_norm": 0.6985920667648315,
"learning_rate": 4.522405660377359e-06,
"loss": 0.7008,
"step": 244
},
{
"epoch": 0.09630503144654089,
"grad_norm": 0.6371440291404724,
"learning_rate": 4.520440251572327e-06,
"loss": 0.6373,
"step": 245
},
{
"epoch": 0.09669811320754718,
"grad_norm": 0.7277976274490356,
"learning_rate": 4.518474842767296e-06,
"loss": 0.757,
"step": 246
},
{
"epoch": 0.09709119496855347,
"grad_norm": 0.798163652420044,
"learning_rate": 4.516509433962265e-06,
"loss": 0.7949,
"step": 247
},
{
"epoch": 0.09748427672955975,
"grad_norm": 0.8415144085884094,
"learning_rate": 4.5145440251572335e-06,
"loss": 0.7264,
"step": 248
},
{
"epoch": 0.09787735849056604,
"grad_norm": 0.6439629793167114,
"learning_rate": 4.5125786163522015e-06,
"loss": 0.6566,
"step": 249
},
{
"epoch": 0.09827044025157233,
"grad_norm": 0.6127240657806396,
"learning_rate": 4.51061320754717e-06,
"loss": 0.6251,
"step": 250
},
{
"epoch": 0.09866352201257862,
"grad_norm": 0.6476905941963196,
"learning_rate": 4.508647798742139e-06,
"loss": 0.693,
"step": 251
},
{
"epoch": 0.09905660377358491,
"grad_norm": 0.67047518491745,
"learning_rate": 4.506682389937107e-06,
"loss": 0.6418,
"step": 252
},
{
"epoch": 0.0994496855345912,
"grad_norm": 0.6989288926124573,
"learning_rate": 4.504716981132076e-06,
"loss": 0.7372,
"step": 253
},
{
"epoch": 0.09984276729559749,
"grad_norm": 0.6157780289649963,
"learning_rate": 4.502751572327044e-06,
"loss": 0.6491,
"step": 254
},
{
"epoch": 0.10023584905660378,
"grad_norm": 0.5550264120101929,
"learning_rate": 4.500786163522013e-06,
"loss": 0.7074,
"step": 255
},
{
"epoch": 0.10062893081761007,
"grad_norm": 0.7177658677101135,
"learning_rate": 4.498820754716982e-06,
"loss": 0.7136,
"step": 256
},
{
"epoch": 0.10102201257861636,
"grad_norm": 0.7320812940597534,
"learning_rate": 4.49685534591195e-06,
"loss": 0.64,
"step": 257
},
{
"epoch": 0.10141509433962265,
"grad_norm": 0.7332838773727417,
"learning_rate": 4.494889937106919e-06,
"loss": 0.6539,
"step": 258
},
{
"epoch": 0.10180817610062894,
"grad_norm": 0.6120893955230713,
"learning_rate": 4.4929245283018875e-06,
"loss": 0.6177,
"step": 259
},
{
"epoch": 0.10220125786163523,
"grad_norm": 0.6673860549926758,
"learning_rate": 4.4909591194968555e-06,
"loss": 0.6163,
"step": 260
},
{
"epoch": 0.10259433962264151,
"grad_norm": 0.8789912462234497,
"learning_rate": 4.488993710691824e-06,
"loss": 0.777,
"step": 261
},
{
"epoch": 0.1029874213836478,
"grad_norm": 0.6589669585227966,
"learning_rate": 4.487028301886793e-06,
"loss": 0.6881,
"step": 262
},
{
"epoch": 0.10338050314465409,
"grad_norm": 0.6306723952293396,
"learning_rate": 4.485062893081761e-06,
"loss": 0.5902,
"step": 263
},
{
"epoch": 0.10377358490566038,
"grad_norm": 0.8177103400230408,
"learning_rate": 4.48309748427673e-06,
"loss": 0.672,
"step": 264
},
{
"epoch": 0.10416666666666667,
"grad_norm": 0.6623080372810364,
"learning_rate": 4.481132075471699e-06,
"loss": 0.6139,
"step": 265
},
{
"epoch": 0.10455974842767296,
"grad_norm": 0.8297303915023804,
"learning_rate": 4.479166666666667e-06,
"loss": 0.6836,
"step": 266
},
{
"epoch": 0.10495283018867925,
"grad_norm": 0.6303706765174866,
"learning_rate": 4.477201257861636e-06,
"loss": 0.5841,
"step": 267
},
{
"epoch": 0.10534591194968554,
"grad_norm": 0.9960982203483582,
"learning_rate": 4.475235849056604e-06,
"loss": 0.7599,
"step": 268
},
{
"epoch": 0.10573899371069183,
"grad_norm": 0.8094085454940796,
"learning_rate": 4.4732704402515725e-06,
"loss": 0.726,
"step": 269
},
{
"epoch": 0.10613207547169812,
"grad_norm": 0.6393566131591797,
"learning_rate": 4.471305031446541e-06,
"loss": 0.6011,
"step": 270
},
{
"epoch": 0.1065251572327044,
"grad_norm": 0.8219441175460815,
"learning_rate": 4.469339622641509e-06,
"loss": 0.7919,
"step": 271
},
{
"epoch": 0.1069182389937107,
"grad_norm": 0.8910709619522095,
"learning_rate": 4.467374213836478e-06,
"loss": 0.8094,
"step": 272
},
{
"epoch": 0.10731132075471699,
"grad_norm": 0.7936025261878967,
"learning_rate": 4.465408805031447e-06,
"loss": 0.764,
"step": 273
},
{
"epoch": 0.10770440251572327,
"grad_norm": 0.7737945914268494,
"learning_rate": 4.463443396226416e-06,
"loss": 0.703,
"step": 274
},
{
"epoch": 0.10809748427672956,
"grad_norm": 0.7185357213020325,
"learning_rate": 4.461477987421384e-06,
"loss": 0.6663,
"step": 275
},
{
"epoch": 0.10849056603773585,
"grad_norm": 0.7443729043006897,
"learning_rate": 4.459512578616352e-06,
"loss": 0.7273,
"step": 276
},
{
"epoch": 0.10888364779874214,
"grad_norm": 0.5857657790184021,
"learning_rate": 4.457547169811321e-06,
"loss": 0.5461,
"step": 277
},
{
"epoch": 0.10927672955974843,
"grad_norm": 0.8231582641601562,
"learning_rate": 4.45558176100629e-06,
"loss": 0.8081,
"step": 278
},
{
"epoch": 0.10966981132075472,
"grad_norm": 0.6273958086967468,
"learning_rate": 4.4536163522012585e-06,
"loss": 0.6891,
"step": 279
},
{
"epoch": 0.11006289308176101,
"grad_norm": 0.6534463167190552,
"learning_rate": 4.4516509433962265e-06,
"loss": 0.6882,
"step": 280
},
{
"epoch": 0.1104559748427673,
"grad_norm": 0.8038327097892761,
"learning_rate": 4.449685534591195e-06,
"loss": 0.7665,
"step": 281
},
{
"epoch": 0.11084905660377359,
"grad_norm": 0.627996027469635,
"learning_rate": 4.447720125786164e-06,
"loss": 0.6937,
"step": 282
},
{
"epoch": 0.11124213836477988,
"grad_norm": 0.6092222332954407,
"learning_rate": 4.445754716981133e-06,
"loss": 0.6068,
"step": 283
},
{
"epoch": 0.11163522012578617,
"grad_norm": 0.8283336162567139,
"learning_rate": 4.443789308176101e-06,
"loss": 0.709,
"step": 284
},
{
"epoch": 0.11202830188679246,
"grad_norm": 0.6859408020973206,
"learning_rate": 4.441823899371069e-06,
"loss": 0.6376,
"step": 285
},
{
"epoch": 0.11242138364779874,
"grad_norm": 0.5870697498321533,
"learning_rate": 4.439858490566038e-06,
"loss": 0.6381,
"step": 286
},
{
"epoch": 0.11281446540880503,
"grad_norm": 0.889226496219635,
"learning_rate": 4.437893081761007e-06,
"loss": 0.7246,
"step": 287
},
{
"epoch": 0.11320754716981132,
"grad_norm": 0.59443598985672,
"learning_rate": 4.4359276729559755e-06,
"loss": 0.7014,
"step": 288
},
{
"epoch": 0.11360062893081761,
"grad_norm": 0.6627135276794434,
"learning_rate": 4.4339622641509435e-06,
"loss": 0.7241,
"step": 289
},
{
"epoch": 0.1139937106918239,
"grad_norm": 0.6983840465545654,
"learning_rate": 4.431996855345912e-06,
"loss": 0.5844,
"step": 290
},
{
"epoch": 0.11438679245283019,
"grad_norm": 0.6393117904663086,
"learning_rate": 4.430031446540881e-06,
"loss": 0.6425,
"step": 291
},
{
"epoch": 0.11477987421383648,
"grad_norm": 0.9656357169151306,
"learning_rate": 4.428066037735849e-06,
"loss": 0.8342,
"step": 292
},
{
"epoch": 0.11517295597484277,
"grad_norm": 0.9343078136444092,
"learning_rate": 4.426100628930818e-06,
"loss": 0.7548,
"step": 293
},
{
"epoch": 0.11556603773584906,
"grad_norm": 0.6172775030136108,
"learning_rate": 4.424135220125786e-06,
"loss": 0.658,
"step": 294
},
{
"epoch": 0.11595911949685535,
"grad_norm": 0.9789079427719116,
"learning_rate": 4.422169811320755e-06,
"loss": 0.7684,
"step": 295
},
{
"epoch": 0.11635220125786164,
"grad_norm": 0.669203519821167,
"learning_rate": 4.420204402515724e-06,
"loss": 0.7957,
"step": 296
},
{
"epoch": 0.11674528301886793,
"grad_norm": 0.6653300523757935,
"learning_rate": 4.418238993710693e-06,
"loss": 0.7306,
"step": 297
},
{
"epoch": 0.11713836477987422,
"grad_norm": 0.5106072425842285,
"learning_rate": 4.416273584905661e-06,
"loss": 0.5035,
"step": 298
},
{
"epoch": 0.1175314465408805,
"grad_norm": 0.6548547148704529,
"learning_rate": 4.4143081761006295e-06,
"loss": 0.756,
"step": 299
},
{
"epoch": 0.1179245283018868,
"grad_norm": 0.7807608842849731,
"learning_rate": 4.4123427672955975e-06,
"loss": 0.7485,
"step": 300
},
{
"epoch": 0.11831761006289308,
"grad_norm": 0.7343317270278931,
"learning_rate": 4.410377358490566e-06,
"loss": 0.7478,
"step": 301
},
{
"epoch": 0.11871069182389937,
"grad_norm": 0.740220308303833,
"learning_rate": 4.408411949685535e-06,
"loss": 0.7745,
"step": 302
},
{
"epoch": 0.11910377358490566,
"grad_norm": 0.90440833568573,
"learning_rate": 4.406446540880503e-06,
"loss": 0.7141,
"step": 303
},
{
"epoch": 0.11949685534591195,
"grad_norm": 0.6514386534690857,
"learning_rate": 4.404481132075472e-06,
"loss": 0.7274,
"step": 304
},
{
"epoch": 0.11988993710691824,
"grad_norm": 0.6821758151054382,
"learning_rate": 4.402515723270441e-06,
"loss": 0.6617,
"step": 305
},
{
"epoch": 0.12028301886792453,
"grad_norm": 0.6932318210601807,
"learning_rate": 4.40055031446541e-06,
"loss": 0.6407,
"step": 306
},
{
"epoch": 0.12067610062893082,
"grad_norm": 0.5298123955726624,
"learning_rate": 4.398584905660378e-06,
"loss": 0.6049,
"step": 307
},
{
"epoch": 0.12106918238993711,
"grad_norm": 0.9546761512756348,
"learning_rate": 4.396619496855346e-06,
"loss": 0.645,
"step": 308
},
{
"epoch": 0.1214622641509434,
"grad_norm": 0.8384695053100586,
"learning_rate": 4.3946540880503145e-06,
"loss": 0.7069,
"step": 309
},
{
"epoch": 0.12185534591194969,
"grad_norm": 0.8737980127334595,
"learning_rate": 4.392688679245283e-06,
"loss": 0.8138,
"step": 310
},
{
"epoch": 0.12224842767295598,
"grad_norm": 0.7366362810134888,
"learning_rate": 4.390723270440252e-06,
"loss": 0.7766,
"step": 311
},
{
"epoch": 0.12264150943396226,
"grad_norm": 0.8029090166091919,
"learning_rate": 4.38875786163522e-06,
"loss": 0.8323,
"step": 312
},
{
"epoch": 0.12303459119496855,
"grad_norm": 0.6308997869491577,
"learning_rate": 4.386792452830189e-06,
"loss": 0.6881,
"step": 313
},
{
"epoch": 0.12342767295597484,
"grad_norm": 0.6481279134750366,
"learning_rate": 4.384827044025158e-06,
"loss": 0.6119,
"step": 314
},
{
"epoch": 0.12382075471698113,
"grad_norm": 0.9859546422958374,
"learning_rate": 4.382861635220127e-06,
"loss": 0.8092,
"step": 315
},
{
"epoch": 0.12421383647798742,
"grad_norm": 0.7419601082801819,
"learning_rate": 4.380896226415095e-06,
"loss": 0.8057,
"step": 316
},
{
"epoch": 0.12460691823899371,
"grad_norm": 0.9320324659347534,
"learning_rate": 4.378930817610063e-06,
"loss": 0.8796,
"step": 317
},
{
"epoch": 0.125,
"grad_norm": 0.766592800617218,
"learning_rate": 4.376965408805032e-06,
"loss": 0.7031,
"step": 318
},
{
"epoch": 0.1253930817610063,
"grad_norm": 0.7485993504524231,
"learning_rate": 4.3750000000000005e-06,
"loss": 0.8188,
"step": 319
},
{
"epoch": 0.12578616352201258,
"grad_norm": 0.8741160035133362,
"learning_rate": 4.373034591194969e-06,
"loss": 0.7379,
"step": 320
},
{
"epoch": 0.12617924528301888,
"grad_norm": 0.6361350417137146,
"learning_rate": 4.371069182389937e-06,
"loss": 0.5549,
"step": 321
},
{
"epoch": 0.12657232704402516,
"grad_norm": 0.6133933067321777,
"learning_rate": 4.369103773584906e-06,
"loss": 0.7498,
"step": 322
},
{
"epoch": 0.12696540880503146,
"grad_norm": 0.7134450674057007,
"learning_rate": 4.367138364779874e-06,
"loss": 0.7332,
"step": 323
},
{
"epoch": 0.12735849056603774,
"grad_norm": 0.7126860022544861,
"learning_rate": 4.365172955974843e-06,
"loss": 0.7213,
"step": 324
},
{
"epoch": 0.12775157232704404,
"grad_norm": 0.593816876411438,
"learning_rate": 4.363207547169812e-06,
"loss": 0.6368,
"step": 325
},
{
"epoch": 0.1281446540880503,
"grad_norm": 0.9041202068328857,
"learning_rate": 4.36124213836478e-06,
"loss": 0.8875,
"step": 326
},
{
"epoch": 0.12853773584905662,
"grad_norm": 0.7038397789001465,
"learning_rate": 4.359276729559749e-06,
"loss": 0.7488,
"step": 327
},
{
"epoch": 0.1289308176100629,
"grad_norm": 0.7919421195983887,
"learning_rate": 4.3573113207547175e-06,
"loss": 0.8105,
"step": 328
},
{
"epoch": 0.1293238993710692,
"grad_norm": 0.5377156138420105,
"learning_rate": 4.355345911949686e-06,
"loss": 0.6061,
"step": 329
},
{
"epoch": 0.12971698113207547,
"grad_norm": 0.8786885142326355,
"learning_rate": 4.353380503144654e-06,
"loss": 0.7679,
"step": 330
},
{
"epoch": 0.13011006289308177,
"grad_norm": 0.9694565534591675,
"learning_rate": 4.351415094339622e-06,
"loss": 0.6846,
"step": 331
},
{
"epoch": 0.13050314465408805,
"grad_norm": 0.9756545424461365,
"learning_rate": 4.349449685534591e-06,
"loss": 0.774,
"step": 332
},
{
"epoch": 0.13089622641509435,
"grad_norm": 0.5200220942497253,
"learning_rate": 4.34748427672956e-06,
"loss": 0.7123,
"step": 333
},
{
"epoch": 0.13128930817610063,
"grad_norm": 0.744434654712677,
"learning_rate": 4.345518867924529e-06,
"loss": 0.8015,
"step": 334
},
{
"epoch": 0.13168238993710693,
"grad_norm": 0.7287289500236511,
"learning_rate": 4.343553459119497e-06,
"loss": 0.6902,
"step": 335
},
{
"epoch": 0.1320754716981132,
"grad_norm": 0.5760185122489929,
"learning_rate": 4.341588050314466e-06,
"loss": 0.6176,
"step": 336
},
{
"epoch": 0.1324685534591195,
"grad_norm": 0.6589622497558594,
"learning_rate": 4.339622641509435e-06,
"loss": 0.6738,
"step": 337
},
{
"epoch": 0.13286163522012578,
"grad_norm": 0.5201132297515869,
"learning_rate": 4.3376572327044035e-06,
"loss": 0.6948,
"step": 338
},
{
"epoch": 0.1332547169811321,
"grad_norm": 0.7743415832519531,
"learning_rate": 4.3356918238993715e-06,
"loss": 0.7061,
"step": 339
},
{
"epoch": 0.13364779874213836,
"grad_norm": 0.8509595990180969,
"learning_rate": 4.3337264150943395e-06,
"loss": 0.818,
"step": 340
},
{
"epoch": 0.13404088050314467,
"grad_norm": 0.6129952073097229,
"learning_rate": 4.331761006289308e-06,
"loss": 0.67,
"step": 341
},
{
"epoch": 0.13443396226415094,
"grad_norm": 0.7439064383506775,
"learning_rate": 4.329795597484277e-06,
"loss": 0.694,
"step": 342
},
{
"epoch": 0.13482704402515724,
"grad_norm": 0.6204108595848083,
"learning_rate": 4.327830188679246e-06,
"loss": 0.6495,
"step": 343
},
{
"epoch": 0.13522012578616352,
"grad_norm": 0.6858063340187073,
"learning_rate": 4.325864779874214e-06,
"loss": 0.7362,
"step": 344
},
{
"epoch": 0.13561320754716982,
"grad_norm": 0.7320376038551331,
"learning_rate": 4.323899371069183e-06,
"loss": 0.7172,
"step": 345
},
{
"epoch": 0.1360062893081761,
"grad_norm": 0.8180674910545349,
"learning_rate": 4.321933962264152e-06,
"loss": 0.7797,
"step": 346
},
{
"epoch": 0.1363993710691824,
"grad_norm": 0.5736013650894165,
"learning_rate": 4.31996855345912e-06,
"loss": 0.616,
"step": 347
},
{
"epoch": 0.13679245283018868,
"grad_norm": 0.8550167679786682,
"learning_rate": 4.3180031446540885e-06,
"loss": 0.7434,
"step": 348
},
{
"epoch": 0.13718553459119498,
"grad_norm": 0.45813804864883423,
"learning_rate": 4.3160377358490565e-06,
"loss": 0.6526,
"step": 349
},
{
"epoch": 0.13757861635220126,
"grad_norm": 0.5863099694252014,
"learning_rate": 4.314072327044025e-06,
"loss": 0.6058,
"step": 350
},
{
"epoch": 0.13797169811320756,
"grad_norm": 0.6945333480834961,
"learning_rate": 4.312106918238994e-06,
"loss": 0.7083,
"step": 351
},
{
"epoch": 0.13836477987421383,
"grad_norm": 0.6828585863113403,
"learning_rate": 4.310141509433963e-06,
"loss": 0.7042,
"step": 352
},
{
"epoch": 0.13875786163522014,
"grad_norm": 0.8388162851333618,
"learning_rate": 4.308176100628931e-06,
"loss": 0.7886,
"step": 353
},
{
"epoch": 0.1391509433962264,
"grad_norm": 0.7719033360481262,
"learning_rate": 4.3062106918239e-06,
"loss": 0.7356,
"step": 354
},
{
"epoch": 0.13954402515723272,
"grad_norm": 0.6767169833183289,
"learning_rate": 4.304245283018868e-06,
"loss": 0.6896,
"step": 355
},
{
"epoch": 0.139937106918239,
"grad_norm": 0.8583604097366333,
"learning_rate": 4.302279874213837e-06,
"loss": 0.7292,
"step": 356
},
{
"epoch": 0.1403301886792453,
"grad_norm": 0.9142056107521057,
"learning_rate": 4.300314465408806e-06,
"loss": 0.7471,
"step": 357
},
{
"epoch": 0.14072327044025157,
"grad_norm": 0.8602889180183411,
"learning_rate": 4.298349056603774e-06,
"loss": 0.7981,
"step": 358
},
{
"epoch": 0.14111635220125787,
"grad_norm": 0.6194784641265869,
"learning_rate": 4.2963836477987425e-06,
"loss": 0.6096,
"step": 359
},
{
"epoch": 0.14150943396226415,
"grad_norm": 0.7596156597137451,
"learning_rate": 4.294418238993711e-06,
"loss": 0.7749,
"step": 360
},
{
"epoch": 0.14190251572327045,
"grad_norm": 0.7424924969673157,
"learning_rate": 4.29245283018868e-06,
"loss": 0.7054,
"step": 361
},
{
"epoch": 0.14229559748427673,
"grad_norm": 0.6850123405456543,
"learning_rate": 4.290487421383648e-06,
"loss": 0.7081,
"step": 362
},
{
"epoch": 0.14268867924528303,
"grad_norm": 0.6266891360282898,
"learning_rate": 4.288522012578616e-06,
"loss": 0.6598,
"step": 363
},
{
"epoch": 0.1430817610062893,
"grad_norm": 0.9214758276939392,
"learning_rate": 4.286556603773585e-06,
"loss": 0.803,
"step": 364
},
{
"epoch": 0.1434748427672956,
"grad_norm": 0.625169038772583,
"learning_rate": 4.284591194968554e-06,
"loss": 0.7067,
"step": 365
},
{
"epoch": 0.14386792452830188,
"grad_norm": 0.7211024761199951,
"learning_rate": 4.282625786163523e-06,
"loss": 0.6772,
"step": 366
},
{
"epoch": 0.14426100628930819,
"grad_norm": 0.6380763649940491,
"learning_rate": 4.280660377358491e-06,
"loss": 0.5542,
"step": 367
},
{
"epoch": 0.14465408805031446,
"grad_norm": 0.7800401449203491,
"learning_rate": 4.2786949685534595e-06,
"loss": 0.6208,
"step": 368
},
{
"epoch": 0.14504716981132076,
"grad_norm": 0.7703373432159424,
"learning_rate": 4.276729559748428e-06,
"loss": 0.7827,
"step": 369
},
{
"epoch": 0.14544025157232704,
"grad_norm": 0.8166595101356506,
"learning_rate": 4.274764150943396e-06,
"loss": 0.7365,
"step": 370
},
{
"epoch": 0.14583333333333334,
"grad_norm": 0.8058414459228516,
"learning_rate": 4.272798742138365e-06,
"loss": 0.7371,
"step": 371
},
{
"epoch": 0.14622641509433962,
"grad_norm": 0.6813973784446716,
"learning_rate": 4.270833333333333e-06,
"loss": 0.6075,
"step": 372
},
{
"epoch": 0.14661949685534592,
"grad_norm": 0.820884108543396,
"learning_rate": 4.268867924528302e-06,
"loss": 0.6709,
"step": 373
},
{
"epoch": 0.1470125786163522,
"grad_norm": 0.6750672459602356,
"learning_rate": 4.266902515723271e-06,
"loss": 0.6993,
"step": 374
},
{
"epoch": 0.1474056603773585,
"grad_norm": 0.8457156419754028,
"learning_rate": 4.26493710691824e-06,
"loss": 0.624,
"step": 375
},
{
"epoch": 0.14779874213836477,
"grad_norm": 0.731155514717102,
"learning_rate": 4.262971698113208e-06,
"loss": 0.6787,
"step": 376
},
{
"epoch": 0.14819182389937108,
"grad_norm": 0.8950203061103821,
"learning_rate": 4.261006289308177e-06,
"loss": 0.8821,
"step": 377
},
{
"epoch": 0.14858490566037735,
"grad_norm": 0.7730862498283386,
"learning_rate": 4.2590408805031455e-06,
"loss": 0.7336,
"step": 378
},
{
"epoch": 0.14897798742138366,
"grad_norm": 0.7011619806289673,
"learning_rate": 4.2570754716981135e-06,
"loss": 0.7048,
"step": 379
},
{
"epoch": 0.14937106918238993,
"grad_norm": 0.7763076424598694,
"learning_rate": 4.255110062893082e-06,
"loss": 0.7306,
"step": 380
},
{
"epoch": 0.14976415094339623,
"grad_norm": 0.5881472826004028,
"learning_rate": 4.25314465408805e-06,
"loss": 0.6327,
"step": 381
},
{
"epoch": 0.1501572327044025,
"grad_norm": 0.8390628099441528,
"learning_rate": 4.251179245283019e-06,
"loss": 0.7809,
"step": 382
},
{
"epoch": 0.1505503144654088,
"grad_norm": 0.7930539846420288,
"learning_rate": 4.249213836477988e-06,
"loss": 0.7712,
"step": 383
},
{
"epoch": 0.1509433962264151,
"grad_norm": 0.790570855140686,
"learning_rate": 4.247248427672956e-06,
"loss": 0.8046,
"step": 384
},
{
"epoch": 0.1513364779874214,
"grad_norm": 0.8940498232841492,
"learning_rate": 4.245283018867925e-06,
"loss": 0.8344,
"step": 385
},
{
"epoch": 0.15172955974842767,
"grad_norm": 0.7356125712394714,
"learning_rate": 4.243317610062894e-06,
"loss": 0.6898,
"step": 386
},
{
"epoch": 0.15212264150943397,
"grad_norm": 0.8964545726776123,
"learning_rate": 4.241352201257862e-06,
"loss": 0.7632,
"step": 387
},
{
"epoch": 0.15251572327044025,
"grad_norm": 0.49109452962875366,
"learning_rate": 4.2393867924528306e-06,
"loss": 0.5786,
"step": 388
},
{
"epoch": 0.15290880503144655,
"grad_norm": 0.7633875608444214,
"learning_rate": 4.237421383647799e-06,
"loss": 0.7852,
"step": 389
},
{
"epoch": 0.15330188679245282,
"grad_norm": 0.6968268752098083,
"learning_rate": 4.235455974842767e-06,
"loss": 0.7091,
"step": 390
},
{
"epoch": 0.15369496855345913,
"grad_norm": 0.7840529680252075,
"learning_rate": 4.233490566037736e-06,
"loss": 0.7062,
"step": 391
},
{
"epoch": 0.1540880503144654,
"grad_norm": 0.7724788188934326,
"learning_rate": 4.231525157232705e-06,
"loss": 0.765,
"step": 392
},
{
"epoch": 0.1544811320754717,
"grad_norm": 0.5707865953445435,
"learning_rate": 4.229559748427673e-06,
"loss": 0.5554,
"step": 393
},
{
"epoch": 0.15487421383647798,
"grad_norm": 0.8885861039161682,
"learning_rate": 4.227594339622642e-06,
"loss": 0.8018,
"step": 394
},
{
"epoch": 0.15526729559748428,
"grad_norm": 0.6703863739967346,
"learning_rate": 4.22562893081761e-06,
"loss": 0.6943,
"step": 395
},
{
"epoch": 0.15566037735849056,
"grad_norm": 0.7018299102783203,
"learning_rate": 4.223663522012579e-06,
"loss": 0.6833,
"step": 396
},
{
"epoch": 0.15605345911949686,
"grad_norm": 0.7630490660667419,
"learning_rate": 4.221698113207548e-06,
"loss": 0.6415,
"step": 397
},
{
"epoch": 0.15644654088050314,
"grad_norm": 0.5800950527191162,
"learning_rate": 4.219732704402516e-06,
"loss": 0.688,
"step": 398
},
{
"epoch": 0.15683962264150944,
"grad_norm": 0.5614280104637146,
"learning_rate": 4.2177672955974845e-06,
"loss": 0.6229,
"step": 399
},
{
"epoch": 0.15723270440251572,
"grad_norm": 0.5274897217750549,
"learning_rate": 4.215801886792453e-06,
"loss": 0.6089,
"step": 400
},
{
"epoch": 0.15762578616352202,
"grad_norm": 0.8656290173530579,
"learning_rate": 4.213836477987422e-06,
"loss": 0.7518,
"step": 401
},
{
"epoch": 0.1580188679245283,
"grad_norm": 0.6752126216888428,
"learning_rate": 4.21187106918239e-06,
"loss": 0.6422,
"step": 402
},
{
"epoch": 0.1584119496855346,
"grad_norm": 0.8268985748291016,
"learning_rate": 4.209905660377359e-06,
"loss": 0.7097,
"step": 403
},
{
"epoch": 0.15880503144654087,
"grad_norm": 0.6740929484367371,
"learning_rate": 4.207940251572327e-06,
"loss": 0.6664,
"step": 404
},
{
"epoch": 0.15919811320754718,
"grad_norm": 0.5292866826057434,
"learning_rate": 4.205974842767296e-06,
"loss": 0.6727,
"step": 405
},
{
"epoch": 0.15959119496855345,
"grad_norm": 0.8358549475669861,
"learning_rate": 4.204009433962265e-06,
"loss": 0.761,
"step": 406
},
{
"epoch": 0.15998427672955975,
"grad_norm": 0.7481523752212524,
"learning_rate": 4.202044025157233e-06,
"loss": 0.7075,
"step": 407
},
{
"epoch": 0.16037735849056603,
"grad_norm": 0.8299019932746887,
"learning_rate": 4.2000786163522016e-06,
"loss": 0.7715,
"step": 408
},
{
"epoch": 0.16077044025157233,
"grad_norm": 0.8758783340454102,
"learning_rate": 4.19811320754717e-06,
"loss": 0.7267,
"step": 409
},
{
"epoch": 0.1611635220125786,
"grad_norm": 0.7504462003707886,
"learning_rate": 4.196147798742139e-06,
"loss": 0.7733,
"step": 410
},
{
"epoch": 0.1615566037735849,
"grad_norm": 0.739465594291687,
"learning_rate": 4.194182389937107e-06,
"loss": 0.7599,
"step": 411
},
{
"epoch": 0.1619496855345912,
"grad_norm": 0.764616847038269,
"learning_rate": 4.192216981132075e-06,
"loss": 0.796,
"step": 412
},
{
"epoch": 0.1623427672955975,
"grad_norm": 0.9454963803291321,
"learning_rate": 4.190251572327044e-06,
"loss": 0.6471,
"step": 413
},
{
"epoch": 0.16273584905660377,
"grad_norm": 0.8122548460960388,
"learning_rate": 4.188286163522013e-06,
"loss": 0.7363,
"step": 414
},
{
"epoch": 0.16312893081761007,
"grad_norm": 0.5691152811050415,
"learning_rate": 4.186320754716982e-06,
"loss": 0.66,
"step": 415
},
{
"epoch": 0.16352201257861634,
"grad_norm": 0.850400447845459,
"learning_rate": 4.18435534591195e-06,
"loss": 0.7467,
"step": 416
},
{
"epoch": 0.16391509433962265,
"grad_norm": 0.596206784248352,
"learning_rate": 4.182389937106919e-06,
"loss": 0.6127,
"step": 417
},
{
"epoch": 0.16430817610062892,
"grad_norm": 0.6859114170074463,
"learning_rate": 4.1804245283018875e-06,
"loss": 0.7056,
"step": 418
},
{
"epoch": 0.16470125786163523,
"grad_norm": 0.6605796813964844,
"learning_rate": 4.1784591194968555e-06,
"loss": 0.731,
"step": 419
},
{
"epoch": 0.1650943396226415,
"grad_norm": 1.04123055934906,
"learning_rate": 4.176493710691824e-06,
"loss": 0.8215,
"step": 420
},
{
"epoch": 0.1654874213836478,
"grad_norm": 0.8572551012039185,
"learning_rate": 4.174528301886792e-06,
"loss": 0.6851,
"step": 421
},
{
"epoch": 0.16588050314465408,
"grad_norm": 0.7791615128517151,
"learning_rate": 4.172562893081761e-06,
"loss": 0.6942,
"step": 422
},
{
"epoch": 0.16627358490566038,
"grad_norm": 0.7698145508766174,
"learning_rate": 4.17059748427673e-06,
"loss": 0.7898,
"step": 423
},
{
"epoch": 0.16666666666666666,
"grad_norm": 0.5505489110946655,
"learning_rate": 4.168632075471699e-06,
"loss": 0.6072,
"step": 424
},
{
"epoch": 0.16705974842767296,
"grad_norm": 0.5675012469291687,
"learning_rate": 4.166666666666667e-06,
"loss": 0.5704,
"step": 425
},
{
"epoch": 0.16745283018867924,
"grad_norm": 0.6114552021026611,
"learning_rate": 4.164701257861636e-06,
"loss": 0.7437,
"step": 426
},
{
"epoch": 0.16784591194968554,
"grad_norm": 0.6155975461006165,
"learning_rate": 4.162735849056604e-06,
"loss": 0.6329,
"step": 427
},
{
"epoch": 0.16823899371069181,
"grad_norm": 0.8240042328834534,
"learning_rate": 4.1607704402515726e-06,
"loss": 0.7863,
"step": 428
},
{
"epoch": 0.16863207547169812,
"grad_norm": 0.6015107035636902,
"learning_rate": 4.158805031446541e-06,
"loss": 0.635,
"step": 429
},
{
"epoch": 0.1690251572327044,
"grad_norm": 0.7438596487045288,
"learning_rate": 4.156839622641509e-06,
"loss": 0.6761,
"step": 430
},
{
"epoch": 0.1694182389937107,
"grad_norm": 0.918503999710083,
"learning_rate": 4.154874213836478e-06,
"loss": 0.9299,
"step": 431
},
{
"epoch": 0.16981132075471697,
"grad_norm": 0.617569625377655,
"learning_rate": 4.152908805031447e-06,
"loss": 0.6299,
"step": 432
},
{
"epoch": 0.17020440251572327,
"grad_norm": 0.4728482663631439,
"learning_rate": 4.150943396226416e-06,
"loss": 0.513,
"step": 433
},
{
"epoch": 0.17059748427672955,
"grad_norm": 0.4552503526210785,
"learning_rate": 4.148977987421384e-06,
"loss": 0.5973,
"step": 434
},
{
"epoch": 0.17099056603773585,
"grad_norm": 0.6574108600616455,
"learning_rate": 4.147012578616352e-06,
"loss": 0.621,
"step": 435
},
{
"epoch": 0.17138364779874213,
"grad_norm": 0.7741186022758484,
"learning_rate": 4.145047169811321e-06,
"loss": 0.6835,
"step": 436
},
{
"epoch": 0.17177672955974843,
"grad_norm": 0.7707587480545044,
"learning_rate": 4.14308176100629e-06,
"loss": 0.7238,
"step": 437
},
{
"epoch": 0.1721698113207547,
"grad_norm": 0.7448788285255432,
"learning_rate": 4.1411163522012585e-06,
"loss": 0.6611,
"step": 438
},
{
"epoch": 0.172562893081761,
"grad_norm": 0.6879774928092957,
"learning_rate": 4.1391509433962265e-06,
"loss": 0.7054,
"step": 439
},
{
"epoch": 0.17295597484276728,
"grad_norm": 0.6766829490661621,
"learning_rate": 4.137185534591195e-06,
"loss": 0.6771,
"step": 440
},
{
"epoch": 0.1733490566037736,
"grad_norm": 0.6811165809631348,
"learning_rate": 4.135220125786164e-06,
"loss": 0.6199,
"step": 441
},
{
"epoch": 0.17374213836477986,
"grad_norm": 0.440508097410202,
"learning_rate": 4.133254716981133e-06,
"loss": 0.5699,
"step": 442
},
{
"epoch": 0.17413522012578617,
"grad_norm": 0.8049497604370117,
"learning_rate": 4.131289308176101e-06,
"loss": 0.7223,
"step": 443
},
{
"epoch": 0.17452830188679244,
"grad_norm": 0.8328004479408264,
"learning_rate": 4.129323899371069e-06,
"loss": 0.7609,
"step": 444
},
{
"epoch": 0.17492138364779874,
"grad_norm": 0.6496450304985046,
"learning_rate": 4.127358490566038e-06,
"loss": 0.6415,
"step": 445
},
{
"epoch": 0.17531446540880502,
"grad_norm": 0.6597641706466675,
"learning_rate": 4.125393081761007e-06,
"loss": 0.6432,
"step": 446
},
{
"epoch": 0.17570754716981132,
"grad_norm": 0.5789406299591064,
"learning_rate": 4.1234276729559756e-06,
"loss": 0.6719,
"step": 447
},
{
"epoch": 0.1761006289308176,
"grad_norm": 0.7399960160255432,
"learning_rate": 4.1214622641509436e-06,
"loss": 0.6227,
"step": 448
},
{
"epoch": 0.1764937106918239,
"grad_norm": 0.8376094698905945,
"learning_rate": 4.119496855345912e-06,
"loss": 0.7648,
"step": 449
},
{
"epoch": 0.17688679245283018,
"grad_norm": 0.6618998646736145,
"learning_rate": 4.117531446540881e-06,
"loss": 0.6317,
"step": 450
},
{
"epoch": 0.17727987421383648,
"grad_norm": 0.7996476292610168,
"learning_rate": 4.115566037735849e-06,
"loss": 0.769,
"step": 451
},
{
"epoch": 0.17767295597484276,
"grad_norm": 0.9092757105827332,
"learning_rate": 4.113600628930818e-06,
"loss": 0.7786,
"step": 452
},
{
"epoch": 0.17806603773584906,
"grad_norm": 0.9256893396377563,
"learning_rate": 4.111635220125786e-06,
"loss": 0.7543,
"step": 453
},
{
"epoch": 0.17845911949685533,
"grad_norm": 0.9227033853530884,
"learning_rate": 4.109669811320755e-06,
"loss": 0.8396,
"step": 454
},
{
"epoch": 0.17885220125786164,
"grad_norm": 0.8288350701332092,
"learning_rate": 4.107704402515724e-06,
"loss": 0.8043,
"step": 455
},
{
"epoch": 0.1792452830188679,
"grad_norm": 0.6936625838279724,
"learning_rate": 4.105738993710693e-06,
"loss": 0.6979,
"step": 456
},
{
"epoch": 0.17963836477987422,
"grad_norm": 0.7924924492835999,
"learning_rate": 4.103773584905661e-06,
"loss": 0.7173,
"step": 457
},
{
"epoch": 0.1800314465408805,
"grad_norm": 0.8403724431991577,
"learning_rate": 4.1018081761006295e-06,
"loss": 0.7789,
"step": 458
},
{
"epoch": 0.1804245283018868,
"grad_norm": 0.8265434503555298,
"learning_rate": 4.0998427672955975e-06,
"loss": 0.8309,
"step": 459
},
{
"epoch": 0.18081761006289307,
"grad_norm": 0.6624492406845093,
"learning_rate": 4.097877358490566e-06,
"loss": 0.6211,
"step": 460
},
{
"epoch": 0.18121069182389937,
"grad_norm": 0.658221423625946,
"learning_rate": 4.095911949685535e-06,
"loss": 0.7571,
"step": 461
},
{
"epoch": 0.18160377358490565,
"grad_norm": 0.7930863499641418,
"learning_rate": 4.093946540880503e-06,
"loss": 0.7986,
"step": 462
},
{
"epoch": 0.18199685534591195,
"grad_norm": 0.6550056338310242,
"learning_rate": 4.091981132075472e-06,
"loss": 0.7128,
"step": 463
},
{
"epoch": 0.18238993710691823,
"grad_norm": 0.7230806946754456,
"learning_rate": 4.090015723270441e-06,
"loss": 0.7227,
"step": 464
},
{
"epoch": 0.18278301886792453,
"grad_norm": 0.7128236293792725,
"learning_rate": 4.08805031446541e-06,
"loss": 0.6747,
"step": 465
},
{
"epoch": 0.1831761006289308,
"grad_norm": 0.9017447829246521,
"learning_rate": 4.086084905660378e-06,
"loss": 0.8768,
"step": 466
},
{
"epoch": 0.1835691823899371,
"grad_norm": 0.4649348855018616,
"learning_rate": 4.084119496855346e-06,
"loss": 0.6119,
"step": 467
},
{
"epoch": 0.18396226415094338,
"grad_norm": 0.5495778322219849,
"learning_rate": 4.0821540880503146e-06,
"loss": 0.6199,
"step": 468
},
{
"epoch": 0.1843553459119497,
"grad_norm": 0.7424394488334656,
"learning_rate": 4.080188679245283e-06,
"loss": 0.7827,
"step": 469
},
{
"epoch": 0.18474842767295596,
"grad_norm": 0.725179135799408,
"learning_rate": 4.078223270440252e-06,
"loss": 0.745,
"step": 470
},
{
"epoch": 0.18514150943396226,
"grad_norm": 0.9258477687835693,
"learning_rate": 4.07625786163522e-06,
"loss": 0.7315,
"step": 471
},
{
"epoch": 0.18553459119496854,
"grad_norm": 0.6924487352371216,
"learning_rate": 4.074292452830189e-06,
"loss": 0.7518,
"step": 472
},
{
"epoch": 0.18592767295597484,
"grad_norm": 0.8073963522911072,
"learning_rate": 4.072327044025158e-06,
"loss": 0.7905,
"step": 473
},
{
"epoch": 0.18632075471698112,
"grad_norm": 0.8046379685401917,
"learning_rate": 4.070361635220127e-06,
"loss": 0.6722,
"step": 474
},
{
"epoch": 0.18671383647798742,
"grad_norm": 0.6345374584197998,
"learning_rate": 4.068396226415095e-06,
"loss": 0.6682,
"step": 475
},
{
"epoch": 0.1871069182389937,
"grad_norm": 0.5131593346595764,
"learning_rate": 4.066430817610063e-06,
"loss": 0.5422,
"step": 476
},
{
"epoch": 0.1875,
"grad_norm": 0.7456933856010437,
"learning_rate": 4.064465408805032e-06,
"loss": 0.7014,
"step": 477
},
{
"epoch": 0.1878930817610063,
"grad_norm": 0.49726560711860657,
"learning_rate": 4.0625000000000005e-06,
"loss": 0.5569,
"step": 478
},
{
"epoch": 0.18828616352201258,
"grad_norm": 0.9552895426750183,
"learning_rate": 4.060534591194969e-06,
"loss": 0.8429,
"step": 479
},
{
"epoch": 0.18867924528301888,
"grad_norm": 0.7442042827606201,
"learning_rate": 4.058569182389937e-06,
"loss": 0.6934,
"step": 480
},
{
"epoch": 0.18907232704402516,
"grad_norm": 0.6966462731361389,
"learning_rate": 4.056603773584906e-06,
"loss": 0.6674,
"step": 481
},
{
"epoch": 0.18946540880503146,
"grad_norm": 0.7582611441612244,
"learning_rate": 4.054638364779874e-06,
"loss": 0.7893,
"step": 482
},
{
"epoch": 0.18985849056603774,
"grad_norm": 0.6868255734443665,
"learning_rate": 4.052672955974843e-06,
"loss": 0.729,
"step": 483
},
{
"epoch": 0.19025157232704404,
"grad_norm": 0.6857454776763916,
"learning_rate": 4.050707547169812e-06,
"loss": 0.7615,
"step": 484
},
{
"epoch": 0.1906446540880503,
"grad_norm": 0.5415035486221313,
"learning_rate": 4.04874213836478e-06,
"loss": 0.5995,
"step": 485
},
{
"epoch": 0.19103773584905662,
"grad_norm": 0.869529128074646,
"learning_rate": 4.046776729559749e-06,
"loss": 0.736,
"step": 486
},
{
"epoch": 0.1914308176100629,
"grad_norm": 0.5622003078460693,
"learning_rate": 4.0448113207547176e-06,
"loss": 0.6055,
"step": 487
},
{
"epoch": 0.1918238993710692,
"grad_norm": 0.7869160771369934,
"learning_rate": 4.042845911949686e-06,
"loss": 0.7368,
"step": 488
},
{
"epoch": 0.19221698113207547,
"grad_norm": 0.6204540729522705,
"learning_rate": 4.040880503144654e-06,
"loss": 0.6903,
"step": 489
},
{
"epoch": 0.19261006289308177,
"grad_norm": 0.7842543721199036,
"learning_rate": 4.038915094339622e-06,
"loss": 0.7921,
"step": 490
},
{
"epoch": 0.19300314465408805,
"grad_norm": 0.9983119964599609,
"learning_rate": 4.036949685534591e-06,
"loss": 0.8429,
"step": 491
},
{
"epoch": 0.19339622641509435,
"grad_norm": 0.7088050842285156,
"learning_rate": 4.03498427672956e-06,
"loss": 0.7067,
"step": 492
},
{
"epoch": 0.19378930817610063,
"grad_norm": 0.6921375393867493,
"learning_rate": 4.033018867924529e-06,
"loss": 0.6958,
"step": 493
},
{
"epoch": 0.19418238993710693,
"grad_norm": 0.6930464506149292,
"learning_rate": 4.031053459119497e-06,
"loss": 0.7457,
"step": 494
},
{
"epoch": 0.1945754716981132,
"grad_norm": 0.5345468521118164,
"learning_rate": 4.029088050314466e-06,
"loss": 0.5737,
"step": 495
},
{
"epoch": 0.1949685534591195,
"grad_norm": 0.6762393712997437,
"learning_rate": 4.027122641509435e-06,
"loss": 0.6965,
"step": 496
},
{
"epoch": 0.19536163522012578,
"grad_norm": 0.8481380343437195,
"learning_rate": 4.025157232704403e-06,
"loss": 0.7292,
"step": 497
},
{
"epoch": 0.1957547169811321,
"grad_norm": 0.6349161267280579,
"learning_rate": 4.0231918238993715e-06,
"loss": 0.6951,
"step": 498
},
{
"epoch": 0.19614779874213836,
"grad_norm": 0.7153770923614502,
"learning_rate": 4.0212264150943395e-06,
"loss": 0.7123,
"step": 499
},
{
"epoch": 0.19654088050314467,
"grad_norm": 0.8139598369598389,
"learning_rate": 4.019261006289308e-06,
"loss": 0.6608,
"step": 500
},
{
"epoch": 0.19693396226415094,
"grad_norm": 0.8879462480545044,
"learning_rate": 4.017295597484277e-06,
"loss": 0.7747,
"step": 501
},
{
"epoch": 0.19732704402515724,
"grad_norm": 0.6974735856056213,
"learning_rate": 4.015330188679246e-06,
"loss": 0.6547,
"step": 502
},
{
"epoch": 0.19772012578616352,
"grad_norm": 0.6531393527984619,
"learning_rate": 4.013364779874214e-06,
"loss": 0.7181,
"step": 503
},
{
"epoch": 0.19811320754716982,
"grad_norm": 0.816124439239502,
"learning_rate": 4.011399371069183e-06,
"loss": 0.7493,
"step": 504
},
{
"epoch": 0.1985062893081761,
"grad_norm": 0.6029940247535706,
"learning_rate": 4.009433962264152e-06,
"loss": 0.6877,
"step": 505
},
{
"epoch": 0.1988993710691824,
"grad_norm": 0.6231976747512817,
"learning_rate": 4.00746855345912e-06,
"loss": 0.6116,
"step": 506
},
{
"epoch": 0.19929245283018868,
"grad_norm": 0.9103032350540161,
"learning_rate": 4.0055031446540886e-06,
"loss": 0.7972,
"step": 507
},
{
"epoch": 0.19968553459119498,
"grad_norm": 0.8262636661529541,
"learning_rate": 4.0035377358490566e-06,
"loss": 0.7012,
"step": 508
},
{
"epoch": 0.20007861635220126,
"grad_norm": 0.6213559508323669,
"learning_rate": 4.001572327044025e-06,
"loss": 0.6343,
"step": 509
},
{
"epoch": 0.20047169811320756,
"grad_norm": 0.7573500275611877,
"learning_rate": 3.999606918238994e-06,
"loss": 0.7495,
"step": 510
},
{
"epoch": 0.20086477987421383,
"grad_norm": 0.673606276512146,
"learning_rate": 3.997641509433962e-06,
"loss": 0.7025,
"step": 511
},
{
"epoch": 0.20125786163522014,
"grad_norm": 0.6274670958518982,
"learning_rate": 3.995676100628931e-06,
"loss": 0.6696,
"step": 512
},
{
"epoch": 0.2016509433962264,
"grad_norm": 0.8562602996826172,
"learning_rate": 3.9937106918239e-06,
"loss": 0.7578,
"step": 513
},
{
"epoch": 0.20204402515723272,
"grad_norm": 0.7410380840301514,
"learning_rate": 3.991745283018868e-06,
"loss": 0.6718,
"step": 514
},
{
"epoch": 0.202437106918239,
"grad_norm": 0.7216569185256958,
"learning_rate": 3.989779874213837e-06,
"loss": 0.7281,
"step": 515
},
{
"epoch": 0.2028301886792453,
"grad_norm": 0.8041703701019287,
"learning_rate": 3.987814465408806e-06,
"loss": 0.7772,
"step": 516
},
{
"epoch": 0.20322327044025157,
"grad_norm": 0.8699491620063782,
"learning_rate": 3.985849056603774e-06,
"loss": 0.712,
"step": 517
},
{
"epoch": 0.20361635220125787,
"grad_norm": 0.6603866219520569,
"learning_rate": 3.9838836477987425e-06,
"loss": 0.6642,
"step": 518
},
{
"epoch": 0.20400943396226415,
"grad_norm": 0.8222591280937195,
"learning_rate": 3.981918238993711e-06,
"loss": 0.6949,
"step": 519
},
{
"epoch": 0.20440251572327045,
"grad_norm": 1.0058287382125854,
"learning_rate": 3.979952830188679e-06,
"loss": 0.8021,
"step": 520
},
{
"epoch": 0.20479559748427673,
"grad_norm": 0.49957284331321716,
"learning_rate": 3.977987421383648e-06,
"loss": 0.6174,
"step": 521
},
{
"epoch": 0.20518867924528303,
"grad_norm": 0.6434182524681091,
"learning_rate": 3.976022012578616e-06,
"loss": 0.6434,
"step": 522
},
{
"epoch": 0.2055817610062893,
"grad_norm": 0.7129496932029724,
"learning_rate": 3.974056603773585e-06,
"loss": 0.7648,
"step": 523
},
{
"epoch": 0.2059748427672956,
"grad_norm": 0.740412712097168,
"learning_rate": 3.972091194968554e-06,
"loss": 0.8125,
"step": 524
},
{
"epoch": 0.20636792452830188,
"grad_norm": 0.6600255966186523,
"learning_rate": 3.970125786163522e-06,
"loss": 0.7483,
"step": 525
},
{
"epoch": 0.20676100628930819,
"grad_norm": 0.791252851486206,
"learning_rate": 3.968160377358491e-06,
"loss": 0.7463,
"step": 526
},
{
"epoch": 0.20715408805031446,
"grad_norm": 0.9195122122764587,
"learning_rate": 3.9661949685534596e-06,
"loss": 0.7814,
"step": 527
},
{
"epoch": 0.20754716981132076,
"grad_norm": 0.5734022259712219,
"learning_rate": 3.964229559748428e-06,
"loss": 0.6191,
"step": 528
},
{
"epoch": 0.20794025157232704,
"grad_norm": 0.6958749890327454,
"learning_rate": 3.962264150943396e-06,
"loss": 0.7291,
"step": 529
},
{
"epoch": 0.20833333333333334,
"grad_norm": 0.7866143584251404,
"learning_rate": 3.960298742138365e-06,
"loss": 0.7372,
"step": 530
},
{
"epoch": 0.20872641509433962,
"grad_norm": 0.867087185382843,
"learning_rate": 3.958333333333333e-06,
"loss": 0.7039,
"step": 531
},
{
"epoch": 0.20911949685534592,
"grad_norm": 0.5408624410629272,
"learning_rate": 3.956367924528302e-06,
"loss": 0.5969,
"step": 532
},
{
"epoch": 0.2095125786163522,
"grad_norm": 0.7537875175476074,
"learning_rate": 3.954402515723271e-06,
"loss": 0.6292,
"step": 533
},
{
"epoch": 0.2099056603773585,
"grad_norm": 0.409322053194046,
"learning_rate": 3.952437106918239e-06,
"loss": 0.5501,
"step": 534
},
{
"epoch": 0.21029874213836477,
"grad_norm": 0.5843974351882935,
"learning_rate": 3.950471698113208e-06,
"loss": 0.5957,
"step": 535
},
{
"epoch": 0.21069182389937108,
"grad_norm": 0.6247936487197876,
"learning_rate": 3.948506289308177e-06,
"loss": 0.5623,
"step": 536
},
{
"epoch": 0.21108490566037735,
"grad_norm": 0.8161054253578186,
"learning_rate": 3.9465408805031455e-06,
"loss": 0.7704,
"step": 537
},
{
"epoch": 0.21147798742138366,
"grad_norm": 0.9021421074867249,
"learning_rate": 3.9445754716981135e-06,
"loss": 0.8109,
"step": 538
},
{
"epoch": 0.21187106918238993,
"grad_norm": 0.6959788203239441,
"learning_rate": 3.9426100628930815e-06,
"loss": 0.777,
"step": 539
},
{
"epoch": 0.21226415094339623,
"grad_norm": 0.7252652049064636,
"learning_rate": 3.94064465408805e-06,
"loss": 0.749,
"step": 540
},
{
"epoch": 0.2126572327044025,
"grad_norm": 0.8188115954399109,
"learning_rate": 3.938679245283019e-06,
"loss": 0.7095,
"step": 541
},
{
"epoch": 0.2130503144654088,
"grad_norm": 0.795558512210846,
"learning_rate": 3.936713836477988e-06,
"loss": 0.7168,
"step": 542
},
{
"epoch": 0.2134433962264151,
"grad_norm": 0.7858245968818665,
"learning_rate": 3.934748427672956e-06,
"loss": 0.7351,
"step": 543
},
{
"epoch": 0.2138364779874214,
"grad_norm": 0.6433058977127075,
"learning_rate": 3.932783018867925e-06,
"loss": 0.7065,
"step": 544
},
{
"epoch": 0.21422955974842767,
"grad_norm": 0.7799449563026428,
"learning_rate": 3.930817610062894e-06,
"loss": 0.832,
"step": 545
},
{
"epoch": 0.21462264150943397,
"grad_norm": 0.8454453945159912,
"learning_rate": 3.928852201257862e-06,
"loss": 0.7682,
"step": 546
},
{
"epoch": 0.21501572327044025,
"grad_norm": 0.667979896068573,
"learning_rate": 3.926886792452831e-06,
"loss": 0.714,
"step": 547
},
{
"epoch": 0.21540880503144655,
"grad_norm": 0.5971770286560059,
"learning_rate": 3.924921383647799e-06,
"loss": 0.6487,
"step": 548
},
{
"epoch": 0.21580188679245282,
"grad_norm": 0.8958584070205688,
"learning_rate": 3.9229559748427674e-06,
"loss": 0.8031,
"step": 549
},
{
"epoch": 0.21619496855345913,
"grad_norm": 0.6960045695304871,
"learning_rate": 3.920990566037736e-06,
"loss": 0.6723,
"step": 550
},
{
"epoch": 0.2165880503144654,
"grad_norm": 0.611761212348938,
"learning_rate": 3.919025157232705e-06,
"loss": 0.6592,
"step": 551
},
{
"epoch": 0.2169811320754717,
"grad_norm": 0.5321255922317505,
"learning_rate": 3.917059748427673e-06,
"loss": 0.6484,
"step": 552
},
{
"epoch": 0.21737421383647798,
"grad_norm": 0.6533987522125244,
"learning_rate": 3.915094339622642e-06,
"loss": 0.66,
"step": 553
},
{
"epoch": 0.21776729559748428,
"grad_norm": 0.7539795637130737,
"learning_rate": 3.91312893081761e-06,
"loss": 0.7443,
"step": 554
},
{
"epoch": 0.21816037735849056,
"grad_norm": 0.8132233619689941,
"learning_rate": 3.911163522012579e-06,
"loss": 0.7057,
"step": 555
},
{
"epoch": 0.21855345911949686,
"grad_norm": 0.6597614288330078,
"learning_rate": 3.909198113207548e-06,
"loss": 0.6665,
"step": 556
},
{
"epoch": 0.21894654088050314,
"grad_norm": 0.6477934122085571,
"learning_rate": 3.907232704402516e-06,
"loss": 0.6939,
"step": 557
},
{
"epoch": 0.21933962264150944,
"grad_norm": 0.625823438167572,
"learning_rate": 3.9052672955974845e-06,
"loss": 0.7008,
"step": 558
},
{
"epoch": 0.21973270440251572,
"grad_norm": 0.5783600211143494,
"learning_rate": 3.903301886792453e-06,
"loss": 0.6639,
"step": 559
},
{
"epoch": 0.22012578616352202,
"grad_norm": 0.8364136815071106,
"learning_rate": 3.901336477987422e-06,
"loss": 0.752,
"step": 560
},
{
"epoch": 0.2205188679245283,
"grad_norm": 0.8422330617904663,
"learning_rate": 3.89937106918239e-06,
"loss": 0.7395,
"step": 561
},
{
"epoch": 0.2209119496855346,
"grad_norm": 0.8582128882408142,
"learning_rate": 3.897405660377358e-06,
"loss": 0.7526,
"step": 562
},
{
"epoch": 0.22130503144654087,
"grad_norm": 0.7453737258911133,
"learning_rate": 3.895440251572327e-06,
"loss": 0.7141,
"step": 563
},
{
"epoch": 0.22169811320754718,
"grad_norm": 0.718410313129425,
"learning_rate": 3.893474842767296e-06,
"loss": 0.6268,
"step": 564
},
{
"epoch": 0.22209119496855345,
"grad_norm": 0.8502873778343201,
"learning_rate": 3.891509433962265e-06,
"loss": 0.7307,
"step": 565
},
{
"epoch": 0.22248427672955975,
"grad_norm": 0.9926005005836487,
"learning_rate": 3.889544025157233e-06,
"loss": 0.8091,
"step": 566
},
{
"epoch": 0.22287735849056603,
"grad_norm": 0.728136420249939,
"learning_rate": 3.887578616352202e-06,
"loss": 0.7117,
"step": 567
},
{
"epoch": 0.22327044025157233,
"grad_norm": 0.7908682823181152,
"learning_rate": 3.8856132075471704e-06,
"loss": 0.6548,
"step": 568
},
{
"epoch": 0.2236635220125786,
"grad_norm": 0.7978211641311646,
"learning_rate": 3.883647798742139e-06,
"loss": 0.7527,
"step": 569
},
{
"epoch": 0.2240566037735849,
"grad_norm": 0.7200964093208313,
"learning_rate": 3.881682389937107e-06,
"loss": 0.7185,
"step": 570
},
{
"epoch": 0.2244496855345912,
"grad_norm": 0.6247283816337585,
"learning_rate": 3.879716981132075e-06,
"loss": 0.6826,
"step": 571
},
{
"epoch": 0.2248427672955975,
"grad_norm": 0.7651311755180359,
"learning_rate": 3.877751572327044e-06,
"loss": 0.8396,
"step": 572
},
{
"epoch": 0.22523584905660377,
"grad_norm": 0.7035654783248901,
"learning_rate": 3.875786163522013e-06,
"loss": 0.7125,
"step": 573
},
{
"epoch": 0.22562893081761007,
"grad_norm": 0.7168806195259094,
"learning_rate": 3.873820754716982e-06,
"loss": 0.6672,
"step": 574
},
{
"epoch": 0.22602201257861634,
"grad_norm": 0.6484406590461731,
"learning_rate": 3.87185534591195e-06,
"loss": 0.6316,
"step": 575
},
{
"epoch": 0.22641509433962265,
"grad_norm": 0.48049476742744446,
"learning_rate": 3.869889937106919e-06,
"loss": 0.6057,
"step": 576
},
{
"epoch": 0.22680817610062892,
"grad_norm": 0.7538033723831177,
"learning_rate": 3.8679245283018875e-06,
"loss": 0.8158,
"step": 577
},
{
"epoch": 0.22720125786163523,
"grad_norm": 0.6726552248001099,
"learning_rate": 3.8659591194968555e-06,
"loss": 0.7271,
"step": 578
},
{
"epoch": 0.2275943396226415,
"grad_norm": 0.6002660393714905,
"learning_rate": 3.863993710691824e-06,
"loss": 0.5521,
"step": 579
},
{
"epoch": 0.2279874213836478,
"grad_norm": 0.7190521359443665,
"learning_rate": 3.862028301886792e-06,
"loss": 0.7687,
"step": 580
},
{
"epoch": 0.22838050314465408,
"grad_norm": 0.7439704537391663,
"learning_rate": 3.860062893081761e-06,
"loss": 0.7556,
"step": 581
},
{
"epoch": 0.22877358490566038,
"grad_norm": 0.7495930790901184,
"learning_rate": 3.85809748427673e-06,
"loss": 0.8008,
"step": 582
},
{
"epoch": 0.22916666666666666,
"grad_norm": 0.6474454402923584,
"learning_rate": 3.856132075471699e-06,
"loss": 0.6187,
"step": 583
},
{
"epoch": 0.22955974842767296,
"grad_norm": 0.8108791708946228,
"learning_rate": 3.854166666666667e-06,
"loss": 0.7511,
"step": 584
},
{
"epoch": 0.22995283018867924,
"grad_norm": 0.7088963985443115,
"learning_rate": 3.852201257861636e-06,
"loss": 0.6901,
"step": 585
},
{
"epoch": 0.23034591194968554,
"grad_norm": 0.4884248673915863,
"learning_rate": 3.850235849056604e-06,
"loss": 0.6201,
"step": 586
},
{
"epoch": 0.23073899371069181,
"grad_norm": 0.6399824023246765,
"learning_rate": 3.848270440251573e-06,
"loss": 0.7565,
"step": 587
},
{
"epoch": 0.23113207547169812,
"grad_norm": 0.5865756869316101,
"learning_rate": 3.8463050314465414e-06,
"loss": 0.5812,
"step": 588
},
{
"epoch": 0.2315251572327044,
"grad_norm": 0.878975510597229,
"learning_rate": 3.8443396226415094e-06,
"loss": 0.8311,
"step": 589
},
{
"epoch": 0.2319182389937107,
"grad_norm": 0.9516463279724121,
"learning_rate": 3.842374213836478e-06,
"loss": 0.8216,
"step": 590
},
{
"epoch": 0.23231132075471697,
"grad_norm": 0.7245076894760132,
"learning_rate": 3.840408805031447e-06,
"loss": 0.7174,
"step": 591
},
{
"epoch": 0.23270440251572327,
"grad_norm": 0.8299921751022339,
"learning_rate": 3.838443396226416e-06,
"loss": 0.7678,
"step": 592
},
{
"epoch": 0.23309748427672955,
"grad_norm": 0.8144885301589966,
"learning_rate": 3.836477987421384e-06,
"loss": 0.699,
"step": 593
},
{
"epoch": 0.23349056603773585,
"grad_norm": 0.5703035593032837,
"learning_rate": 3.834512578616352e-06,
"loss": 0.6675,
"step": 594
},
{
"epoch": 0.23388364779874213,
"grad_norm": 1.0637414455413818,
"learning_rate": 3.832547169811321e-06,
"loss": 0.7964,
"step": 595
},
{
"epoch": 0.23427672955974843,
"grad_norm": 0.7096083164215088,
"learning_rate": 3.83058176100629e-06,
"loss": 0.7745,
"step": 596
},
{
"epoch": 0.2346698113207547,
"grad_norm": 0.656769335269928,
"learning_rate": 3.8286163522012585e-06,
"loss": 0.738,
"step": 597
},
{
"epoch": 0.235062893081761,
"grad_norm": 0.9162077307701111,
"learning_rate": 3.8266509433962265e-06,
"loss": 0.8483,
"step": 598
},
{
"epoch": 0.23545597484276728,
"grad_norm": 0.7402646541595459,
"learning_rate": 3.824685534591195e-06,
"loss": 0.7006,
"step": 599
},
{
"epoch": 0.2358490566037736,
"grad_norm": 0.7398844361305237,
"learning_rate": 3.822720125786164e-06,
"loss": 0.7937,
"step": 600
},
{
"epoch": 0.23624213836477986,
"grad_norm": 0.6222953200340271,
"learning_rate": 3.820754716981133e-06,
"loss": 0.7122,
"step": 601
},
{
"epoch": 0.23663522012578617,
"grad_norm": 0.4656656086444855,
"learning_rate": 3.818789308176101e-06,
"loss": 0.5313,
"step": 602
},
{
"epoch": 0.23702830188679244,
"grad_norm": 0.8139412999153137,
"learning_rate": 3.816823899371069e-06,
"loss": 0.6777,
"step": 603
},
{
"epoch": 0.23742138364779874,
"grad_norm": 0.6474642753601074,
"learning_rate": 3.814858490566038e-06,
"loss": 0.6521,
"step": 604
},
{
"epoch": 0.23781446540880502,
"grad_norm": 0.7259648442268372,
"learning_rate": 3.8128930817610067e-06,
"loss": 0.7183,
"step": 605
},
{
"epoch": 0.23820754716981132,
"grad_norm": 0.5460947751998901,
"learning_rate": 3.810927672955975e-06,
"loss": 0.6726,
"step": 606
},
{
"epoch": 0.2386006289308176,
"grad_norm": 0.7965362668037415,
"learning_rate": 3.808962264150944e-06,
"loss": 0.7325,
"step": 607
},
{
"epoch": 0.2389937106918239,
"grad_norm": 0.8698051571846008,
"learning_rate": 3.8069968553459124e-06,
"loss": 0.8198,
"step": 608
},
{
"epoch": 0.23938679245283018,
"grad_norm": 0.576199471950531,
"learning_rate": 3.8050314465408813e-06,
"loss": 0.694,
"step": 609
},
{
"epoch": 0.23977987421383648,
"grad_norm": 0.7640446424484253,
"learning_rate": 3.8030660377358493e-06,
"loss": 0.7292,
"step": 610
},
{
"epoch": 0.24017295597484276,
"grad_norm": 0.7617279291152954,
"learning_rate": 3.8011006289308177e-06,
"loss": 0.7423,
"step": 611
},
{
"epoch": 0.24056603773584906,
"grad_norm": 0.8192011713981628,
"learning_rate": 3.7991352201257866e-06,
"loss": 0.7338,
"step": 612
},
{
"epoch": 0.24095911949685533,
"grad_norm": 0.7665988206863403,
"learning_rate": 3.797169811320755e-06,
"loss": 0.7161,
"step": 613
},
{
"epoch": 0.24135220125786164,
"grad_norm": 1.0656249523162842,
"learning_rate": 3.795204402515724e-06,
"loss": 0.7322,
"step": 614
},
{
"epoch": 0.2417452830188679,
"grad_norm": 0.7555919885635376,
"learning_rate": 3.7932389937106922e-06,
"loss": 0.7954,
"step": 615
},
{
"epoch": 0.24213836477987422,
"grad_norm": 0.853146493434906,
"learning_rate": 3.791273584905661e-06,
"loss": 0.8367,
"step": 616
},
{
"epoch": 0.2425314465408805,
"grad_norm": 0.5899045467376709,
"learning_rate": 3.7893081761006295e-06,
"loss": 0.7174,
"step": 617
},
{
"epoch": 0.2429245283018868,
"grad_norm": 0.6919112205505371,
"learning_rate": 3.7873427672955975e-06,
"loss": 0.6688,
"step": 618
},
{
"epoch": 0.24331761006289307,
"grad_norm": 0.6048345565795898,
"learning_rate": 3.7853773584905664e-06,
"loss": 0.676,
"step": 619
},
{
"epoch": 0.24371069182389937,
"grad_norm": 0.8114615082740784,
"learning_rate": 3.7834119496855348e-06,
"loss": 0.8424,
"step": 620
},
{
"epoch": 0.24410377358490565,
"grad_norm": 0.8147010803222656,
"learning_rate": 3.7814465408805036e-06,
"loss": 0.7546,
"step": 621
},
{
"epoch": 0.24449685534591195,
"grad_norm": 0.7359874844551086,
"learning_rate": 3.779481132075472e-06,
"loss": 0.8396,
"step": 622
},
{
"epoch": 0.24488993710691823,
"grad_norm": 0.8015702366828918,
"learning_rate": 3.777515723270441e-06,
"loss": 0.7562,
"step": 623
},
{
"epoch": 0.24528301886792453,
"grad_norm": 0.616278350353241,
"learning_rate": 3.7755503144654093e-06,
"loss": 0.6665,
"step": 624
},
{
"epoch": 0.2456761006289308,
"grad_norm": 0.6622241139411926,
"learning_rate": 3.7735849056603777e-06,
"loss": 0.6277,
"step": 625
},
{
"epoch": 0.2460691823899371,
"grad_norm": 0.7007507681846619,
"learning_rate": 3.771619496855346e-06,
"loss": 0.7347,
"step": 626
},
{
"epoch": 0.24646226415094338,
"grad_norm": 0.7834786176681519,
"learning_rate": 3.7696540880503146e-06,
"loss": 0.7346,
"step": 627
},
{
"epoch": 0.2468553459119497,
"grad_norm": 0.7515172958374023,
"learning_rate": 3.7676886792452834e-06,
"loss": 0.6626,
"step": 628
},
{
"epoch": 0.24724842767295596,
"grad_norm": 0.737511157989502,
"learning_rate": 3.765723270440252e-06,
"loss": 0.6504,
"step": 629
},
{
"epoch": 0.24764150943396226,
"grad_norm": 0.6552973985671997,
"learning_rate": 3.7637578616352207e-06,
"loss": 0.6347,
"step": 630
},
{
"epoch": 0.24803459119496854,
"grad_norm": 0.6942483186721802,
"learning_rate": 3.761792452830189e-06,
"loss": 0.7206,
"step": 631
},
{
"epoch": 0.24842767295597484,
"grad_norm": 0.627299427986145,
"learning_rate": 3.7598270440251576e-06,
"loss": 0.6119,
"step": 632
},
{
"epoch": 0.24882075471698112,
"grad_norm": 0.7206055521965027,
"learning_rate": 3.7578616352201264e-06,
"loss": 0.8129,
"step": 633
},
{
"epoch": 0.24921383647798742,
"grad_norm": 0.6862805485725403,
"learning_rate": 3.7558962264150944e-06,
"loss": 0.656,
"step": 634
},
{
"epoch": 0.2496069182389937,
"grad_norm": 0.7048293948173523,
"learning_rate": 3.7539308176100632e-06,
"loss": 0.6874,
"step": 635
},
{
"epoch": 0.25,
"grad_norm": 0.6825442314147949,
"learning_rate": 3.7519654088050317e-06,
"loss": 0.6697,
"step": 636
},
{
"epoch": 0.2503930817610063,
"grad_norm": 0.6846087574958801,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.7021,
"step": 637
},
{
"epoch": 0.2507861635220126,
"grad_norm": 0.7152469754219055,
"learning_rate": 3.748034591194969e-06,
"loss": 0.78,
"step": 638
},
{
"epoch": 0.2511792452830189,
"grad_norm": 0.8945066332817078,
"learning_rate": 3.7460691823899374e-06,
"loss": 0.7828,
"step": 639
},
{
"epoch": 0.25157232704402516,
"grad_norm": 0.489244282245636,
"learning_rate": 3.744103773584906e-06,
"loss": 0.5905,
"step": 640
},
{
"epoch": 0.25196540880503143,
"grad_norm": 0.751223623752594,
"learning_rate": 3.742138364779874e-06,
"loss": 0.7299,
"step": 641
},
{
"epoch": 0.25235849056603776,
"grad_norm": 0.546923816204071,
"learning_rate": 3.740172955974843e-06,
"loss": 0.6196,
"step": 642
},
{
"epoch": 0.25275157232704404,
"grad_norm": 0.6353660821914673,
"learning_rate": 3.7382075471698115e-06,
"loss": 0.6547,
"step": 643
},
{
"epoch": 0.2531446540880503,
"grad_norm": 0.8038619160652161,
"learning_rate": 3.73624213836478e-06,
"loss": 0.7852,
"step": 644
},
{
"epoch": 0.2535377358490566,
"grad_norm": 0.7789469361305237,
"learning_rate": 3.7342767295597488e-06,
"loss": 0.746,
"step": 645
},
{
"epoch": 0.2539308176100629,
"grad_norm": 0.5475597977638245,
"learning_rate": 3.732311320754717e-06,
"loss": 0.6774,
"step": 646
},
{
"epoch": 0.2543238993710692,
"grad_norm": 0.7746227979660034,
"learning_rate": 3.730345911949686e-06,
"loss": 0.7477,
"step": 647
},
{
"epoch": 0.25471698113207547,
"grad_norm": 0.7397127747535706,
"learning_rate": 3.7283805031446544e-06,
"loss": 0.697,
"step": 648
},
{
"epoch": 0.25511006289308175,
"grad_norm": 0.8050301671028137,
"learning_rate": 3.726415094339623e-06,
"loss": 0.7692,
"step": 649
},
{
"epoch": 0.2555031446540881,
"grad_norm": 0.6419538855552673,
"learning_rate": 3.7244496855345913e-06,
"loss": 0.5424,
"step": 650
},
{
"epoch": 0.25589622641509435,
"grad_norm": 0.6572560667991638,
"learning_rate": 3.7224842767295597e-06,
"loss": 0.6349,
"step": 651
},
{
"epoch": 0.2562893081761006,
"grad_norm": 0.5560314655303955,
"learning_rate": 3.7205188679245286e-06,
"loss": 0.6318,
"step": 652
},
{
"epoch": 0.2566823899371069,
"grad_norm": 0.6834450364112854,
"learning_rate": 3.718553459119497e-06,
"loss": 0.7261,
"step": 653
},
{
"epoch": 0.25707547169811323,
"grad_norm": 0.7266345024108887,
"learning_rate": 3.716588050314466e-06,
"loss": 0.7828,
"step": 654
},
{
"epoch": 0.2574685534591195,
"grad_norm": 0.7601588368415833,
"learning_rate": 3.7146226415094343e-06,
"loss": 0.735,
"step": 655
},
{
"epoch": 0.2578616352201258,
"grad_norm": 0.694202721118927,
"learning_rate": 3.712657232704403e-06,
"loss": 0.7692,
"step": 656
},
{
"epoch": 0.25825471698113206,
"grad_norm": 0.7295196652412415,
"learning_rate": 3.710691823899371e-06,
"loss": 0.7672,
"step": 657
},
{
"epoch": 0.2586477987421384,
"grad_norm": 0.6914157271385193,
"learning_rate": 3.7087264150943395e-06,
"loss": 0.6838,
"step": 658
},
{
"epoch": 0.25904088050314467,
"grad_norm": 0.8847957253456116,
"learning_rate": 3.7067610062893084e-06,
"loss": 0.629,
"step": 659
},
{
"epoch": 0.25943396226415094,
"grad_norm": 0.9601537585258484,
"learning_rate": 3.704795597484277e-06,
"loss": 0.8344,
"step": 660
},
{
"epoch": 0.2598270440251572,
"grad_norm": 0.9131399989128113,
"learning_rate": 3.7028301886792456e-06,
"loss": 0.7924,
"step": 661
},
{
"epoch": 0.26022012578616355,
"grad_norm": 0.6381687521934509,
"learning_rate": 3.700864779874214e-06,
"loss": 0.6754,
"step": 662
},
{
"epoch": 0.2606132075471698,
"grad_norm": 0.638401985168457,
"learning_rate": 3.698899371069183e-06,
"loss": 0.7594,
"step": 663
},
{
"epoch": 0.2610062893081761,
"grad_norm": 0.6575311422348022,
"learning_rate": 3.6969339622641513e-06,
"loss": 0.6448,
"step": 664
},
{
"epoch": 0.2613993710691824,
"grad_norm": 0.6457099914550781,
"learning_rate": 3.6949685534591193e-06,
"loss": 0.7179,
"step": 665
},
{
"epoch": 0.2617924528301887,
"grad_norm": 0.6970727443695068,
"learning_rate": 3.693003144654088e-06,
"loss": 0.7073,
"step": 666
},
{
"epoch": 0.262185534591195,
"grad_norm": 0.7326680421829224,
"learning_rate": 3.6910377358490566e-06,
"loss": 0.7152,
"step": 667
},
{
"epoch": 0.26257861635220126,
"grad_norm": 0.8637464642524719,
"learning_rate": 3.6890723270440254e-06,
"loss": 0.806,
"step": 668
},
{
"epoch": 0.26297169811320753,
"grad_norm": 0.6680986881256104,
"learning_rate": 3.687106918238994e-06,
"loss": 0.6169,
"step": 669
},
{
"epoch": 0.26336477987421386,
"grad_norm": 0.7912803888320923,
"learning_rate": 3.6851415094339627e-06,
"loss": 0.6921,
"step": 670
},
{
"epoch": 0.26375786163522014,
"grad_norm": 0.8088789582252502,
"learning_rate": 3.683176100628931e-06,
"loss": 0.7172,
"step": 671
},
{
"epoch": 0.2641509433962264,
"grad_norm": 0.6131149530410767,
"learning_rate": 3.6812106918239e-06,
"loss": 0.6853,
"step": 672
},
{
"epoch": 0.2645440251572327,
"grad_norm": 0.7404634952545166,
"learning_rate": 3.679245283018868e-06,
"loss": 0.7556,
"step": 673
},
{
"epoch": 0.264937106918239,
"grad_norm": 0.7626749277114868,
"learning_rate": 3.6772798742138364e-06,
"loss": 0.7562,
"step": 674
},
{
"epoch": 0.2653301886792453,
"grad_norm": 0.7413991093635559,
"learning_rate": 3.6753144654088053e-06,
"loss": 0.7693,
"step": 675
},
{
"epoch": 0.26572327044025157,
"grad_norm": 0.6210454106330872,
"learning_rate": 3.6733490566037737e-06,
"loss": 0.5854,
"step": 676
},
{
"epoch": 0.26611635220125784,
"grad_norm": 0.6471617817878723,
"learning_rate": 3.6713836477987425e-06,
"loss": 0.6052,
"step": 677
},
{
"epoch": 0.2665094339622642,
"grad_norm": 0.837889552116394,
"learning_rate": 3.669418238993711e-06,
"loss": 0.8444,
"step": 678
},
{
"epoch": 0.26690251572327045,
"grad_norm": 0.738037645816803,
"learning_rate": 3.66745283018868e-06,
"loss": 0.735,
"step": 679
},
{
"epoch": 0.2672955974842767,
"grad_norm": 0.7115139365196228,
"learning_rate": 3.6654874213836482e-06,
"loss": 0.7668,
"step": 680
},
{
"epoch": 0.267688679245283,
"grad_norm": 0.7722954750061035,
"learning_rate": 3.6635220125786162e-06,
"loss": 0.7568,
"step": 681
},
{
"epoch": 0.26808176100628933,
"grad_norm": 0.7241666913032532,
"learning_rate": 3.661556603773585e-06,
"loss": 0.7394,
"step": 682
},
{
"epoch": 0.2684748427672956,
"grad_norm": 0.6548068523406982,
"learning_rate": 3.6595911949685535e-06,
"loss": 0.7113,
"step": 683
},
{
"epoch": 0.2688679245283019,
"grad_norm": 0.7873329520225525,
"learning_rate": 3.6576257861635223e-06,
"loss": 0.7501,
"step": 684
},
{
"epoch": 0.26926100628930816,
"grad_norm": 0.859569787979126,
"learning_rate": 3.6556603773584908e-06,
"loss": 0.7252,
"step": 685
},
{
"epoch": 0.2696540880503145,
"grad_norm": 0.7971783876419067,
"learning_rate": 3.6536949685534596e-06,
"loss": 0.7615,
"step": 686
},
{
"epoch": 0.27004716981132076,
"grad_norm": 0.8030787110328674,
"learning_rate": 3.651729559748428e-06,
"loss": 0.6718,
"step": 687
},
{
"epoch": 0.27044025157232704,
"grad_norm": 0.7667816877365112,
"learning_rate": 3.649764150943397e-06,
"loss": 0.733,
"step": 688
},
{
"epoch": 0.2708333333333333,
"grad_norm": 0.4544336795806885,
"learning_rate": 3.647798742138365e-06,
"loss": 0.5857,
"step": 689
},
{
"epoch": 0.27122641509433965,
"grad_norm": 0.6503133177757263,
"learning_rate": 3.6458333333333333e-06,
"loss": 0.6427,
"step": 690
},
{
"epoch": 0.2716194968553459,
"grad_norm": 0.9716889262199402,
"learning_rate": 3.643867924528302e-06,
"loss": 0.802,
"step": 691
},
{
"epoch": 0.2720125786163522,
"grad_norm": 0.7119978666305542,
"learning_rate": 3.6419025157232706e-06,
"loss": 0.6406,
"step": 692
},
{
"epoch": 0.27240566037735847,
"grad_norm": 1.1628193855285645,
"learning_rate": 3.6399371069182394e-06,
"loss": 0.7811,
"step": 693
},
{
"epoch": 0.2727987421383648,
"grad_norm": 0.7122523784637451,
"learning_rate": 3.637971698113208e-06,
"loss": 0.7296,
"step": 694
},
{
"epoch": 0.2731918238993711,
"grad_norm": 0.8217347860336304,
"learning_rate": 3.6360062893081767e-06,
"loss": 0.816,
"step": 695
},
{
"epoch": 0.27358490566037735,
"grad_norm": 0.8290286064147949,
"learning_rate": 3.634040880503145e-06,
"loss": 0.7885,
"step": 696
},
{
"epoch": 0.27397798742138363,
"grad_norm": 0.9015904664993286,
"learning_rate": 3.632075471698113e-06,
"loss": 0.801,
"step": 697
},
{
"epoch": 0.27437106918238996,
"grad_norm": 0.6886477470397949,
"learning_rate": 3.630110062893082e-06,
"loss": 0.7511,
"step": 698
},
{
"epoch": 0.27476415094339623,
"grad_norm": 0.7540025115013123,
"learning_rate": 3.6281446540880504e-06,
"loss": 0.7289,
"step": 699
},
{
"epoch": 0.2751572327044025,
"grad_norm": 0.6688190698623657,
"learning_rate": 3.6261792452830192e-06,
"loss": 0.671,
"step": 700
},
{
"epoch": 0.2755503144654088,
"grad_norm": 0.43286392092704773,
"learning_rate": 3.6242138364779876e-06,
"loss": 0.5497,
"step": 701
},
{
"epoch": 0.2759433962264151,
"grad_norm": 0.6663103103637695,
"learning_rate": 3.6222484276729565e-06,
"loss": 0.7239,
"step": 702
},
{
"epoch": 0.2763364779874214,
"grad_norm": 0.7031357884407043,
"learning_rate": 3.620283018867925e-06,
"loss": 0.7108,
"step": 703
},
{
"epoch": 0.27672955974842767,
"grad_norm": 0.7224757075309753,
"learning_rate": 3.6183176100628938e-06,
"loss": 0.7429,
"step": 704
},
{
"epoch": 0.27712264150943394,
"grad_norm": 0.693199634552002,
"learning_rate": 3.6163522012578618e-06,
"loss": 0.7307,
"step": 705
},
{
"epoch": 0.2775157232704403,
"grad_norm": 0.8898524641990662,
"learning_rate": 3.61438679245283e-06,
"loss": 0.7021,
"step": 706
},
{
"epoch": 0.27790880503144655,
"grad_norm": 0.6072456240653992,
"learning_rate": 3.612421383647799e-06,
"loss": 0.6372,
"step": 707
},
{
"epoch": 0.2783018867924528,
"grad_norm": 0.5581188201904297,
"learning_rate": 3.6104559748427675e-06,
"loss": 0.5854,
"step": 708
},
{
"epoch": 0.2786949685534591,
"grad_norm": 0.7197433114051819,
"learning_rate": 3.6084905660377363e-06,
"loss": 0.7336,
"step": 709
},
{
"epoch": 0.27908805031446543,
"grad_norm": 0.7470580339431763,
"learning_rate": 3.6065251572327047e-06,
"loss": 0.7419,
"step": 710
},
{
"epoch": 0.2794811320754717,
"grad_norm": 0.7079648375511169,
"learning_rate": 3.6045597484276736e-06,
"loss": 0.7136,
"step": 711
},
{
"epoch": 0.279874213836478,
"grad_norm": 0.6582914590835571,
"learning_rate": 3.602594339622642e-06,
"loss": 0.6716,
"step": 712
},
{
"epoch": 0.28026729559748426,
"grad_norm": 0.5141453742980957,
"learning_rate": 3.60062893081761e-06,
"loss": 0.6083,
"step": 713
},
{
"epoch": 0.2806603773584906,
"grad_norm": 0.6707072257995605,
"learning_rate": 3.598663522012579e-06,
"loss": 0.647,
"step": 714
},
{
"epoch": 0.28105345911949686,
"grad_norm": 0.9114739894866943,
"learning_rate": 3.5966981132075473e-06,
"loss": 0.7742,
"step": 715
},
{
"epoch": 0.28144654088050314,
"grad_norm": 0.7468785047531128,
"learning_rate": 3.594732704402516e-06,
"loss": 0.8011,
"step": 716
},
{
"epoch": 0.2818396226415094,
"grad_norm": 0.8181868195533752,
"learning_rate": 3.5927672955974845e-06,
"loss": 0.7077,
"step": 717
},
{
"epoch": 0.28223270440251574,
"grad_norm": 0.77386873960495,
"learning_rate": 3.5908018867924534e-06,
"loss": 0.7388,
"step": 718
},
{
"epoch": 0.282625786163522,
"grad_norm": 0.5534411072731018,
"learning_rate": 3.588836477987422e-06,
"loss": 0.5424,
"step": 719
},
{
"epoch": 0.2830188679245283,
"grad_norm": 0.727638840675354,
"learning_rate": 3.5868710691823906e-06,
"loss": 0.7311,
"step": 720
},
{
"epoch": 0.28341194968553457,
"grad_norm": 0.8172037601470947,
"learning_rate": 3.5849056603773586e-06,
"loss": 0.764,
"step": 721
},
{
"epoch": 0.2838050314465409,
"grad_norm": 0.671344518661499,
"learning_rate": 3.582940251572327e-06,
"loss": 0.6785,
"step": 722
},
{
"epoch": 0.2841981132075472,
"grad_norm": 0.6699550747871399,
"learning_rate": 3.580974842767296e-06,
"loss": 0.718,
"step": 723
},
{
"epoch": 0.28459119496855345,
"grad_norm": 0.6324903964996338,
"learning_rate": 3.5790094339622643e-06,
"loss": 0.6398,
"step": 724
},
{
"epoch": 0.2849842767295597,
"grad_norm": 0.9452508091926575,
"learning_rate": 3.577044025157233e-06,
"loss": 0.8274,
"step": 725
},
{
"epoch": 0.28537735849056606,
"grad_norm": 0.6974048614501953,
"learning_rate": 3.5750786163522016e-06,
"loss": 0.6019,
"step": 726
},
{
"epoch": 0.28577044025157233,
"grad_norm": 0.7029326558113098,
"learning_rate": 3.5731132075471705e-06,
"loss": 0.6428,
"step": 727
},
{
"epoch": 0.2861635220125786,
"grad_norm": 0.6052168011665344,
"learning_rate": 3.571147798742139e-06,
"loss": 0.6981,
"step": 728
},
{
"epoch": 0.2865566037735849,
"grad_norm": 0.655536413192749,
"learning_rate": 3.569182389937107e-06,
"loss": 0.6056,
"step": 729
},
{
"epoch": 0.2869496855345912,
"grad_norm": 0.679928719997406,
"learning_rate": 3.5672169811320757e-06,
"loss": 0.7104,
"step": 730
},
{
"epoch": 0.2873427672955975,
"grad_norm": 0.8509970307350159,
"learning_rate": 3.565251572327044e-06,
"loss": 0.7975,
"step": 731
},
{
"epoch": 0.28773584905660377,
"grad_norm": 0.5402770042419434,
"learning_rate": 3.563286163522013e-06,
"loss": 0.6533,
"step": 732
},
{
"epoch": 0.28812893081761004,
"grad_norm": 0.7498342990875244,
"learning_rate": 3.5613207547169814e-06,
"loss": 0.7319,
"step": 733
},
{
"epoch": 0.28852201257861637,
"grad_norm": 0.7832929491996765,
"learning_rate": 3.5593553459119503e-06,
"loss": 0.7626,
"step": 734
},
{
"epoch": 0.28891509433962265,
"grad_norm": 0.7070589065551758,
"learning_rate": 3.5573899371069187e-06,
"loss": 0.6936,
"step": 735
},
{
"epoch": 0.2893081761006289,
"grad_norm": 0.6468908190727234,
"learning_rate": 3.5554245283018875e-06,
"loss": 0.7404,
"step": 736
},
{
"epoch": 0.2897012578616352,
"grad_norm": 0.6880595684051514,
"learning_rate": 3.5534591194968555e-06,
"loss": 0.7445,
"step": 737
},
{
"epoch": 0.29009433962264153,
"grad_norm": 0.5934287309646606,
"learning_rate": 3.551493710691824e-06,
"loss": 0.6459,
"step": 738
},
{
"epoch": 0.2904874213836478,
"grad_norm": 0.6091917753219604,
"learning_rate": 3.549528301886793e-06,
"loss": 0.6284,
"step": 739
},
{
"epoch": 0.2908805031446541,
"grad_norm": 0.771747887134552,
"learning_rate": 3.5475628930817612e-06,
"loss": 0.7136,
"step": 740
},
{
"epoch": 0.29127358490566035,
"grad_norm": 0.7308130860328674,
"learning_rate": 3.54559748427673e-06,
"loss": 0.6943,
"step": 741
},
{
"epoch": 0.2916666666666667,
"grad_norm": 0.7461852431297302,
"learning_rate": 3.5436320754716985e-06,
"loss": 0.6675,
"step": 742
},
{
"epoch": 0.29205974842767296,
"grad_norm": 0.7564466595649719,
"learning_rate": 3.5416666666666673e-06,
"loss": 0.7069,
"step": 743
},
{
"epoch": 0.29245283018867924,
"grad_norm": 0.5232489705085754,
"learning_rate": 3.5397012578616358e-06,
"loss": 0.6144,
"step": 744
},
{
"epoch": 0.2928459119496855,
"grad_norm": 0.7904266715049744,
"learning_rate": 3.5377358490566038e-06,
"loss": 0.6261,
"step": 745
},
{
"epoch": 0.29323899371069184,
"grad_norm": 0.5785681009292603,
"learning_rate": 3.5357704402515726e-06,
"loss": 0.6222,
"step": 746
},
{
"epoch": 0.2936320754716981,
"grad_norm": 0.7388083934783936,
"learning_rate": 3.533805031446541e-06,
"loss": 0.7888,
"step": 747
},
{
"epoch": 0.2940251572327044,
"grad_norm": 0.8282852172851562,
"learning_rate": 3.53183962264151e-06,
"loss": 0.802,
"step": 748
},
{
"epoch": 0.29441823899371067,
"grad_norm": 0.49576103687286377,
"learning_rate": 3.5298742138364783e-06,
"loss": 0.6417,
"step": 749
},
{
"epoch": 0.294811320754717,
"grad_norm": 0.6288707256317139,
"learning_rate": 3.527908805031447e-06,
"loss": 0.6481,
"step": 750
},
{
"epoch": 0.2952044025157233,
"grad_norm": 0.7539206147193909,
"learning_rate": 3.5259433962264156e-06,
"loss": 0.6968,
"step": 751
},
{
"epoch": 0.29559748427672955,
"grad_norm": 0.8830596208572388,
"learning_rate": 3.523977987421384e-06,
"loss": 0.6507,
"step": 752
},
{
"epoch": 0.2959905660377358,
"grad_norm": 0.6779255270957947,
"learning_rate": 3.5220125786163524e-06,
"loss": 0.6633,
"step": 753
},
{
"epoch": 0.29638364779874216,
"grad_norm": 0.6763765215873718,
"learning_rate": 3.520047169811321e-06,
"loss": 0.6736,
"step": 754
},
{
"epoch": 0.29677672955974843,
"grad_norm": 0.5210625529289246,
"learning_rate": 3.5180817610062897e-06,
"loss": 0.6692,
"step": 755
},
{
"epoch": 0.2971698113207547,
"grad_norm": 0.6614780426025391,
"learning_rate": 3.516116352201258e-06,
"loss": 0.621,
"step": 756
},
{
"epoch": 0.297562893081761,
"grad_norm": 0.6689832210540771,
"learning_rate": 3.514150943396227e-06,
"loss": 0.7056,
"step": 757
},
{
"epoch": 0.2979559748427673,
"grad_norm": 0.7676858305931091,
"learning_rate": 3.5121855345911954e-06,
"loss": 0.7341,
"step": 758
},
{
"epoch": 0.2983490566037736,
"grad_norm": 0.8308287262916565,
"learning_rate": 3.510220125786164e-06,
"loss": 0.7485,
"step": 759
},
{
"epoch": 0.29874213836477986,
"grad_norm": 0.8294158577919006,
"learning_rate": 3.5082547169811327e-06,
"loss": 0.7848,
"step": 760
},
{
"epoch": 0.29913522012578614,
"grad_norm": 0.6138144135475159,
"learning_rate": 3.5062893081761007e-06,
"loss": 0.738,
"step": 761
},
{
"epoch": 0.29952830188679247,
"grad_norm": 0.7953334450721741,
"learning_rate": 3.5043238993710695e-06,
"loss": 0.7642,
"step": 762
},
{
"epoch": 0.29992138364779874,
"grad_norm": 0.780980110168457,
"learning_rate": 3.502358490566038e-06,
"loss": 0.6746,
"step": 763
},
{
"epoch": 0.300314465408805,
"grad_norm": 0.8279568552970886,
"learning_rate": 3.5003930817610068e-06,
"loss": 0.7613,
"step": 764
},
{
"epoch": 0.3007075471698113,
"grad_norm": 0.7951825857162476,
"learning_rate": 3.498427672955975e-06,
"loss": 0.6971,
"step": 765
},
{
"epoch": 0.3011006289308176,
"grad_norm": 0.5353666543960571,
"learning_rate": 3.4964622641509436e-06,
"loss": 0.6,
"step": 766
},
{
"epoch": 0.3014937106918239,
"grad_norm": 0.7189162969589233,
"learning_rate": 3.4944968553459125e-06,
"loss": 0.7565,
"step": 767
},
{
"epoch": 0.3018867924528302,
"grad_norm": 0.8963586091995239,
"learning_rate": 3.492531446540881e-06,
"loss": 0.8469,
"step": 768
},
{
"epoch": 0.30227987421383645,
"grad_norm": 0.8720613718032837,
"learning_rate": 3.4905660377358493e-06,
"loss": 0.5281,
"step": 769
},
{
"epoch": 0.3026729559748428,
"grad_norm": 0.6833882331848145,
"learning_rate": 3.4886006289308177e-06,
"loss": 0.7026,
"step": 770
},
{
"epoch": 0.30306603773584906,
"grad_norm": 0.7453868985176086,
"learning_rate": 3.4866352201257866e-06,
"loss": 0.7832,
"step": 771
},
{
"epoch": 0.30345911949685533,
"grad_norm": 0.5843695998191833,
"learning_rate": 3.484669811320755e-06,
"loss": 0.6635,
"step": 772
},
{
"epoch": 0.3038522012578616,
"grad_norm": 0.8642531633377075,
"learning_rate": 3.4827044025157234e-06,
"loss": 0.7573,
"step": 773
},
{
"epoch": 0.30424528301886794,
"grad_norm": 0.94108647108078,
"learning_rate": 3.4807389937106923e-06,
"loss": 0.7097,
"step": 774
},
{
"epoch": 0.3046383647798742,
"grad_norm": 0.5342159271240234,
"learning_rate": 3.4787735849056607e-06,
"loss": 0.548,
"step": 775
},
{
"epoch": 0.3050314465408805,
"grad_norm": 0.6646297574043274,
"learning_rate": 3.4768081761006295e-06,
"loss": 0.6086,
"step": 776
},
{
"epoch": 0.30542452830188677,
"grad_norm": 0.7132428884506226,
"learning_rate": 3.4748427672955975e-06,
"loss": 0.7515,
"step": 777
},
{
"epoch": 0.3058176100628931,
"grad_norm": 0.7163389921188354,
"learning_rate": 3.472877358490566e-06,
"loss": 0.7391,
"step": 778
},
{
"epoch": 0.3062106918238994,
"grad_norm": 0.5722758173942566,
"learning_rate": 3.470911949685535e-06,
"loss": 0.6492,
"step": 779
},
{
"epoch": 0.30660377358490565,
"grad_norm": 0.7335605025291443,
"learning_rate": 3.4689465408805032e-06,
"loss": 0.6941,
"step": 780
},
{
"epoch": 0.3069968553459119,
"grad_norm": 0.6022026538848877,
"learning_rate": 3.466981132075472e-06,
"loss": 0.6586,
"step": 781
},
{
"epoch": 0.30738993710691825,
"grad_norm": 0.778712809085846,
"learning_rate": 3.4650157232704405e-06,
"loss": 0.7883,
"step": 782
},
{
"epoch": 0.30778301886792453,
"grad_norm": 0.7263333201408386,
"learning_rate": 3.4630503144654094e-06,
"loss": 0.7667,
"step": 783
},
{
"epoch": 0.3081761006289308,
"grad_norm": 0.5462168455123901,
"learning_rate": 3.4610849056603778e-06,
"loss": 0.5282,
"step": 784
},
{
"epoch": 0.3085691823899371,
"grad_norm": 0.6476829648017883,
"learning_rate": 3.4591194968553458e-06,
"loss": 0.7465,
"step": 785
},
{
"epoch": 0.3089622641509434,
"grad_norm": 0.7814940810203552,
"learning_rate": 3.4571540880503146e-06,
"loss": 0.8167,
"step": 786
},
{
"epoch": 0.3093553459119497,
"grad_norm": 0.5674806833267212,
"learning_rate": 3.455188679245283e-06,
"loss": 0.6162,
"step": 787
},
{
"epoch": 0.30974842767295596,
"grad_norm": 0.6166163086891174,
"learning_rate": 3.453223270440252e-06,
"loss": 0.6851,
"step": 788
},
{
"epoch": 0.31014150943396224,
"grad_norm": 0.7583447098731995,
"learning_rate": 3.4512578616352203e-06,
"loss": 0.7175,
"step": 789
},
{
"epoch": 0.31053459119496857,
"grad_norm": 0.7047259211540222,
"learning_rate": 3.449292452830189e-06,
"loss": 0.7308,
"step": 790
},
{
"epoch": 0.31092767295597484,
"grad_norm": 0.6118384003639221,
"learning_rate": 3.4473270440251576e-06,
"loss": 0.7096,
"step": 791
},
{
"epoch": 0.3113207547169811,
"grad_norm": 0.5853203535079956,
"learning_rate": 3.4453616352201264e-06,
"loss": 0.6245,
"step": 792
},
{
"epoch": 0.3117138364779874,
"grad_norm": 0.8413107395172119,
"learning_rate": 3.4433962264150944e-06,
"loss": 0.6582,
"step": 793
},
{
"epoch": 0.3121069182389937,
"grad_norm": 0.6036621928215027,
"learning_rate": 3.441430817610063e-06,
"loss": 0.7031,
"step": 794
},
{
"epoch": 0.3125,
"grad_norm": 0.5434512495994568,
"learning_rate": 3.4394654088050317e-06,
"loss": 0.575,
"step": 795
},
{
"epoch": 0.3128930817610063,
"grad_norm": 0.7164487242698669,
"learning_rate": 3.4375e-06,
"loss": 0.648,
"step": 796
},
{
"epoch": 0.3132861635220126,
"grad_norm": 0.8002411723136902,
"learning_rate": 3.435534591194969e-06,
"loss": 0.7368,
"step": 797
},
{
"epoch": 0.3136792452830189,
"grad_norm": 0.7365742325782776,
"learning_rate": 3.4335691823899374e-06,
"loss": 0.6677,
"step": 798
},
{
"epoch": 0.31407232704402516,
"grad_norm": 0.6155741810798645,
"learning_rate": 3.4316037735849062e-06,
"loss": 0.6554,
"step": 799
},
{
"epoch": 0.31446540880503143,
"grad_norm": 0.9236664772033691,
"learning_rate": 3.4296383647798742e-06,
"loss": 0.7217,
"step": 800
},
{
"epoch": 0.31485849056603776,
"grad_norm": 0.6514955759048462,
"learning_rate": 3.4276729559748427e-06,
"loss": 0.7505,
"step": 801
},
{
"epoch": 0.31525157232704404,
"grad_norm": 0.7860538959503174,
"learning_rate": 3.4257075471698115e-06,
"loss": 0.735,
"step": 802
},
{
"epoch": 0.3156446540880503,
"grad_norm": 0.6035215258598328,
"learning_rate": 3.42374213836478e-06,
"loss": 0.6345,
"step": 803
},
{
"epoch": 0.3160377358490566,
"grad_norm": 0.5767151117324829,
"learning_rate": 3.4217767295597488e-06,
"loss": 0.587,
"step": 804
},
{
"epoch": 0.3164308176100629,
"grad_norm": 0.7364680171012878,
"learning_rate": 3.419811320754717e-06,
"loss": 0.6299,
"step": 805
},
{
"epoch": 0.3168238993710692,
"grad_norm": 0.7540648579597473,
"learning_rate": 3.417845911949686e-06,
"loss": 0.6572,
"step": 806
},
{
"epoch": 0.31721698113207547,
"grad_norm": 0.6114500164985657,
"learning_rate": 3.4158805031446545e-06,
"loss": 0.6718,
"step": 807
},
{
"epoch": 0.31761006289308175,
"grad_norm": 0.6366127729415894,
"learning_rate": 3.4139150943396225e-06,
"loss": 0.6923,
"step": 808
},
{
"epoch": 0.3180031446540881,
"grad_norm": 0.660406231880188,
"learning_rate": 3.4119496855345913e-06,
"loss": 0.7035,
"step": 809
},
{
"epoch": 0.31839622641509435,
"grad_norm": 0.8871625065803528,
"learning_rate": 3.4099842767295597e-06,
"loss": 0.7382,
"step": 810
},
{
"epoch": 0.3187893081761006,
"grad_norm": 0.8379966616630554,
"learning_rate": 3.4080188679245286e-06,
"loss": 0.6823,
"step": 811
},
{
"epoch": 0.3191823899371069,
"grad_norm": 0.6468477249145508,
"learning_rate": 3.406053459119497e-06,
"loss": 0.6711,
"step": 812
},
{
"epoch": 0.31957547169811323,
"grad_norm": 0.8465133905410767,
"learning_rate": 3.404088050314466e-06,
"loss": 0.7453,
"step": 813
},
{
"epoch": 0.3199685534591195,
"grad_norm": 0.7196464538574219,
"learning_rate": 3.4021226415094343e-06,
"loss": 0.7479,
"step": 814
},
{
"epoch": 0.3203616352201258,
"grad_norm": 0.624432384967804,
"learning_rate": 3.400157232704403e-06,
"loss": 0.6132,
"step": 815
},
{
"epoch": 0.32075471698113206,
"grad_norm": 0.5828042030334473,
"learning_rate": 3.398191823899371e-06,
"loss": 0.6637,
"step": 816
},
{
"epoch": 0.3211477987421384,
"grad_norm": 0.834619402885437,
"learning_rate": 3.3962264150943395e-06,
"loss": 0.6848,
"step": 817
},
{
"epoch": 0.32154088050314467,
"grad_norm": 0.5867133736610413,
"learning_rate": 3.3942610062893084e-06,
"loss": 0.6045,
"step": 818
},
{
"epoch": 0.32193396226415094,
"grad_norm": 0.6242752075195312,
"learning_rate": 3.392295597484277e-06,
"loss": 0.7146,
"step": 819
},
{
"epoch": 0.3223270440251572,
"grad_norm": 0.6345254778862,
"learning_rate": 3.3903301886792457e-06,
"loss": 0.6952,
"step": 820
},
{
"epoch": 0.32272012578616355,
"grad_norm": 0.6565260291099548,
"learning_rate": 3.388364779874214e-06,
"loss": 0.7156,
"step": 821
},
{
"epoch": 0.3231132075471698,
"grad_norm": 0.5501343607902527,
"learning_rate": 3.386399371069183e-06,
"loss": 0.6694,
"step": 822
},
{
"epoch": 0.3235062893081761,
"grad_norm": 0.5736883878707886,
"learning_rate": 3.3844339622641514e-06,
"loss": 0.6241,
"step": 823
},
{
"epoch": 0.3238993710691824,
"grad_norm": 0.6627534031867981,
"learning_rate": 3.3824685534591194e-06,
"loss": 0.6586,
"step": 824
},
{
"epoch": 0.3242924528301887,
"grad_norm": 0.8529284596443176,
"learning_rate": 3.380503144654088e-06,
"loss": 0.8123,
"step": 825
},
{
"epoch": 0.324685534591195,
"grad_norm": 0.6788104176521301,
"learning_rate": 3.3785377358490566e-06,
"loss": 0.6853,
"step": 826
},
{
"epoch": 0.32507861635220126,
"grad_norm": 0.7295207381248474,
"learning_rate": 3.3765723270440255e-06,
"loss": 0.7803,
"step": 827
},
{
"epoch": 0.32547169811320753,
"grad_norm": 0.6535108089447021,
"learning_rate": 3.374606918238994e-06,
"loss": 0.6285,
"step": 828
},
{
"epoch": 0.32586477987421386,
"grad_norm": 0.6590744256973267,
"learning_rate": 3.3726415094339627e-06,
"loss": 0.6025,
"step": 829
},
{
"epoch": 0.32625786163522014,
"grad_norm": 0.6202201247215271,
"learning_rate": 3.370676100628931e-06,
"loss": 0.6706,
"step": 830
},
{
"epoch": 0.3266509433962264,
"grad_norm": 0.8043216466903687,
"learning_rate": 3.3687106918239e-06,
"loss": 0.8022,
"step": 831
},
{
"epoch": 0.3270440251572327,
"grad_norm": 0.6206783652305603,
"learning_rate": 3.366745283018868e-06,
"loss": 0.6426,
"step": 832
},
{
"epoch": 0.327437106918239,
"grad_norm": 0.5996221899986267,
"learning_rate": 3.3647798742138364e-06,
"loss": 0.6917,
"step": 833
},
{
"epoch": 0.3278301886792453,
"grad_norm": 0.9232528805732727,
"learning_rate": 3.3628144654088053e-06,
"loss": 0.7662,
"step": 834
},
{
"epoch": 0.32822327044025157,
"grad_norm": 0.614966094493866,
"learning_rate": 3.3608490566037737e-06,
"loss": 0.6517,
"step": 835
},
{
"epoch": 0.32861635220125784,
"grad_norm": 0.7594910860061646,
"learning_rate": 3.3588836477987426e-06,
"loss": 0.7745,
"step": 836
},
{
"epoch": 0.3290094339622642,
"grad_norm": 0.88567715883255,
"learning_rate": 3.356918238993711e-06,
"loss": 0.9014,
"step": 837
},
{
"epoch": 0.32940251572327045,
"grad_norm": 0.6599681973457336,
"learning_rate": 3.35495283018868e-06,
"loss": 0.7314,
"step": 838
},
{
"epoch": 0.3297955974842767,
"grad_norm": 0.8103616237640381,
"learning_rate": 3.3529874213836482e-06,
"loss": 0.6539,
"step": 839
},
{
"epoch": 0.330188679245283,
"grad_norm": 0.6297152042388916,
"learning_rate": 3.3510220125786162e-06,
"loss": 0.7186,
"step": 840
},
{
"epoch": 0.33058176100628933,
"grad_norm": 0.4803082048892975,
"learning_rate": 3.349056603773585e-06,
"loss": 0.5624,
"step": 841
},
{
"epoch": 0.3309748427672956,
"grad_norm": 0.6797827482223511,
"learning_rate": 3.3470911949685535e-06,
"loss": 0.6068,
"step": 842
},
{
"epoch": 0.3313679245283019,
"grad_norm": 0.7400256395339966,
"learning_rate": 3.3451257861635224e-06,
"loss": 0.7665,
"step": 843
},
{
"epoch": 0.33176100628930816,
"grad_norm": 0.7882441282272339,
"learning_rate": 3.3431603773584908e-06,
"loss": 0.8042,
"step": 844
},
{
"epoch": 0.3321540880503145,
"grad_norm": 0.6411010026931763,
"learning_rate": 3.3411949685534596e-06,
"loss": 0.6757,
"step": 845
},
{
"epoch": 0.33254716981132076,
"grad_norm": 0.5777365565299988,
"learning_rate": 3.339229559748428e-06,
"loss": 0.6486,
"step": 846
},
{
"epoch": 0.33294025157232704,
"grad_norm": 0.6347094178199768,
"learning_rate": 3.337264150943397e-06,
"loss": 0.6444,
"step": 847
},
{
"epoch": 0.3333333333333333,
"grad_norm": 0.7500883936882019,
"learning_rate": 3.335298742138365e-06,
"loss": 0.7722,
"step": 848
},
{
"epoch": 0.33372641509433965,
"grad_norm": 0.7402048110961914,
"learning_rate": 3.3333333333333333e-06,
"loss": 0.6775,
"step": 849
},
{
"epoch": 0.3341194968553459,
"grad_norm": 0.7049366235733032,
"learning_rate": 3.331367924528302e-06,
"loss": 0.7145,
"step": 850
},
{
"epoch": 0.3345125786163522,
"grad_norm": 0.6983188986778259,
"learning_rate": 3.3294025157232706e-06,
"loss": 0.6404,
"step": 851
},
{
"epoch": 0.33490566037735847,
"grad_norm": 0.7494722008705139,
"learning_rate": 3.3274371069182394e-06,
"loss": 0.7189,
"step": 852
},
{
"epoch": 0.3352987421383648,
"grad_norm": 0.6518175005912781,
"learning_rate": 3.325471698113208e-06,
"loss": 0.6393,
"step": 853
},
{
"epoch": 0.3356918238993711,
"grad_norm": 0.7594310641288757,
"learning_rate": 3.3235062893081767e-06,
"loss": 0.7145,
"step": 854
},
{
"epoch": 0.33608490566037735,
"grad_norm": 0.7364822626113892,
"learning_rate": 3.321540880503145e-06,
"loss": 0.7964,
"step": 855
},
{
"epoch": 0.33647798742138363,
"grad_norm": 0.7396246790885925,
"learning_rate": 3.319575471698113e-06,
"loss": 0.7024,
"step": 856
},
{
"epoch": 0.33687106918238996,
"grad_norm": 0.6513952612876892,
"learning_rate": 3.317610062893082e-06,
"loss": 0.7112,
"step": 857
},
{
"epoch": 0.33726415094339623,
"grad_norm": 0.773344874382019,
"learning_rate": 3.3156446540880504e-06,
"loss": 0.674,
"step": 858
},
{
"epoch": 0.3376572327044025,
"grad_norm": 0.7500815987586975,
"learning_rate": 3.3136792452830192e-06,
"loss": 0.7577,
"step": 859
},
{
"epoch": 0.3380503144654088,
"grad_norm": 0.9222676753997803,
"learning_rate": 3.3117138364779877e-06,
"loss": 0.6666,
"step": 860
},
{
"epoch": 0.3384433962264151,
"grad_norm": 0.6305007934570312,
"learning_rate": 3.3097484276729565e-06,
"loss": 0.6433,
"step": 861
},
{
"epoch": 0.3388364779874214,
"grad_norm": 0.8446972966194153,
"learning_rate": 3.307783018867925e-06,
"loss": 0.7822,
"step": 862
},
{
"epoch": 0.33922955974842767,
"grad_norm": 0.603020429611206,
"learning_rate": 3.305817610062894e-06,
"loss": 0.6409,
"step": 863
},
{
"epoch": 0.33962264150943394,
"grad_norm": 0.7890493869781494,
"learning_rate": 3.3038522012578618e-06,
"loss": 0.7357,
"step": 864
},
{
"epoch": 0.3400157232704403,
"grad_norm": 0.8309876918792725,
"learning_rate": 3.30188679245283e-06,
"loss": 0.7366,
"step": 865
},
{
"epoch": 0.34040880503144655,
"grad_norm": 0.627488911151886,
"learning_rate": 3.299921383647799e-06,
"loss": 0.6664,
"step": 866
},
{
"epoch": 0.3408018867924528,
"grad_norm": 0.4105185866355896,
"learning_rate": 3.2979559748427675e-06,
"loss": 0.5654,
"step": 867
},
{
"epoch": 0.3411949685534591,
"grad_norm": 0.6174553632736206,
"learning_rate": 3.2959905660377363e-06,
"loss": 0.6844,
"step": 868
},
{
"epoch": 0.34158805031446543,
"grad_norm": 0.79210364818573,
"learning_rate": 3.2940251572327048e-06,
"loss": 0.6967,
"step": 869
},
{
"epoch": 0.3419811320754717,
"grad_norm": 0.741703450679779,
"learning_rate": 3.2920597484276736e-06,
"loss": 0.7365,
"step": 870
},
{
"epoch": 0.342374213836478,
"grad_norm": 0.7099801301956177,
"learning_rate": 3.290094339622642e-06,
"loss": 0.7366,
"step": 871
},
{
"epoch": 0.34276729559748426,
"grad_norm": 0.6283672451972961,
"learning_rate": 3.28812893081761e-06,
"loss": 0.6858,
"step": 872
},
{
"epoch": 0.3431603773584906,
"grad_norm": 0.7430014610290527,
"learning_rate": 3.286163522012579e-06,
"loss": 0.7411,
"step": 873
},
{
"epoch": 0.34355345911949686,
"grad_norm": 0.7987573146820068,
"learning_rate": 3.2841981132075473e-06,
"loss": 0.7208,
"step": 874
},
{
"epoch": 0.34394654088050314,
"grad_norm": 0.9213460087776184,
"learning_rate": 3.282232704402516e-06,
"loss": 0.7469,
"step": 875
},
{
"epoch": 0.3443396226415094,
"grad_norm": 0.6389803886413574,
"learning_rate": 3.2802672955974846e-06,
"loss": 0.5962,
"step": 876
},
{
"epoch": 0.34473270440251574,
"grad_norm": 0.79592365026474,
"learning_rate": 3.2783018867924534e-06,
"loss": 0.6545,
"step": 877
},
{
"epoch": 0.345125786163522,
"grad_norm": 0.7427257299423218,
"learning_rate": 3.276336477987422e-06,
"loss": 0.6595,
"step": 878
},
{
"epoch": 0.3455188679245283,
"grad_norm": 0.7178816199302673,
"learning_rate": 3.2743710691823903e-06,
"loss": 0.6458,
"step": 879
},
{
"epoch": 0.34591194968553457,
"grad_norm": 0.623972475528717,
"learning_rate": 3.2724056603773587e-06,
"loss": 0.6473,
"step": 880
},
{
"epoch": 0.3463050314465409,
"grad_norm": 1.9782747030258179,
"learning_rate": 3.270440251572327e-06,
"loss": 0.7788,
"step": 881
},
{
"epoch": 0.3466981132075472,
"grad_norm": 0.5571057796478271,
"learning_rate": 3.268474842767296e-06,
"loss": 0.6573,
"step": 882
},
{
"epoch": 0.34709119496855345,
"grad_norm": 0.7872087359428406,
"learning_rate": 3.2665094339622644e-06,
"loss": 0.8557,
"step": 883
},
{
"epoch": 0.3474842767295597,
"grad_norm": 0.7490341663360596,
"learning_rate": 3.2645440251572332e-06,
"loss": 0.6929,
"step": 884
},
{
"epoch": 0.34787735849056606,
"grad_norm": 1.0249555110931396,
"learning_rate": 3.2625786163522016e-06,
"loss": 0.8578,
"step": 885
},
{
"epoch": 0.34827044025157233,
"grad_norm": 0.8417578339576721,
"learning_rate": 3.26061320754717e-06,
"loss": 0.6486,
"step": 886
},
{
"epoch": 0.3486635220125786,
"grad_norm": 0.7490862011909485,
"learning_rate": 3.258647798742139e-06,
"loss": 0.6985,
"step": 887
},
{
"epoch": 0.3490566037735849,
"grad_norm": 0.7455279231071472,
"learning_rate": 3.256682389937107e-06,
"loss": 0.7236,
"step": 888
},
{
"epoch": 0.3494496855345912,
"grad_norm": 0.7189223170280457,
"learning_rate": 3.2547169811320758e-06,
"loss": 0.7213,
"step": 889
},
{
"epoch": 0.3498427672955975,
"grad_norm": 0.8472079634666443,
"learning_rate": 3.252751572327044e-06,
"loss": 0.7119,
"step": 890
},
{
"epoch": 0.35023584905660377,
"grad_norm": 0.7449663281440735,
"learning_rate": 3.250786163522013e-06,
"loss": 0.6698,
"step": 891
},
{
"epoch": 0.35062893081761004,
"grad_norm": 0.7291968464851379,
"learning_rate": 3.2488207547169814e-06,
"loss": 0.6671,
"step": 892
},
{
"epoch": 0.35102201257861637,
"grad_norm": 0.7697893381118774,
"learning_rate": 3.24685534591195e-06,
"loss": 0.655,
"step": 893
},
{
"epoch": 0.35141509433962265,
"grad_norm": 0.8186383843421936,
"learning_rate": 3.2448899371069187e-06,
"loss": 0.802,
"step": 894
},
{
"epoch": 0.3518081761006289,
"grad_norm": 0.8444569110870361,
"learning_rate": 3.242924528301887e-06,
"loss": 0.7839,
"step": 895
},
{
"epoch": 0.3522012578616352,
"grad_norm": 0.6955016851425171,
"learning_rate": 3.2409591194968556e-06,
"loss": 0.7148,
"step": 896
},
{
"epoch": 0.35259433962264153,
"grad_norm": 0.9391415119171143,
"learning_rate": 3.238993710691824e-06,
"loss": 0.7665,
"step": 897
},
{
"epoch": 0.3529874213836478,
"grad_norm": 0.6215878129005432,
"learning_rate": 3.237028301886793e-06,
"loss": 0.6199,
"step": 898
},
{
"epoch": 0.3533805031446541,
"grad_norm": 0.7451786398887634,
"learning_rate": 3.2350628930817613e-06,
"loss": 0.7502,
"step": 899
},
{
"epoch": 0.35377358490566035,
"grad_norm": 0.6543328166007996,
"learning_rate": 3.2330974842767297e-06,
"loss": 0.7764,
"step": 900
},
{
"epoch": 0.3541666666666667,
"grad_norm": 0.5691241025924683,
"learning_rate": 3.2311320754716985e-06,
"loss": 0.651,
"step": 901
},
{
"epoch": 0.35455974842767296,
"grad_norm": 0.6460458636283875,
"learning_rate": 3.229166666666667e-06,
"loss": 0.6138,
"step": 902
},
{
"epoch": 0.35495283018867924,
"grad_norm": 0.5493612885475159,
"learning_rate": 3.227201257861636e-06,
"loss": 0.6271,
"step": 903
},
{
"epoch": 0.3553459119496855,
"grad_norm": 0.4768497943878174,
"learning_rate": 3.225235849056604e-06,
"loss": 0.604,
"step": 904
},
{
"epoch": 0.35573899371069184,
"grad_norm": 0.7839314937591553,
"learning_rate": 3.2232704402515726e-06,
"loss": 0.7275,
"step": 905
},
{
"epoch": 0.3561320754716981,
"grad_norm": 1.228136420249939,
"learning_rate": 3.221305031446541e-06,
"loss": 0.7606,
"step": 906
},
{
"epoch": 0.3565251572327044,
"grad_norm": 0.9644804000854492,
"learning_rate": 3.2193396226415095e-06,
"loss": 0.7522,
"step": 907
},
{
"epoch": 0.35691823899371067,
"grad_norm": 0.6227933764457703,
"learning_rate": 3.2173742138364783e-06,
"loss": 0.6799,
"step": 908
},
{
"epoch": 0.357311320754717,
"grad_norm": 0.669668972492218,
"learning_rate": 3.2154088050314468e-06,
"loss": 0.7111,
"step": 909
},
{
"epoch": 0.3577044025157233,
"grad_norm": 0.8379939198493958,
"learning_rate": 3.2134433962264156e-06,
"loss": 0.7601,
"step": 910
},
{
"epoch": 0.35809748427672955,
"grad_norm": 0.6675577759742737,
"learning_rate": 3.211477987421384e-06,
"loss": 0.7216,
"step": 911
},
{
"epoch": 0.3584905660377358,
"grad_norm": 0.5988037586212158,
"learning_rate": 3.209512578616352e-06,
"loss": 0.599,
"step": 912
},
{
"epoch": 0.35888364779874216,
"grad_norm": 0.90444415807724,
"learning_rate": 3.207547169811321e-06,
"loss": 0.775,
"step": 913
},
{
"epoch": 0.35927672955974843,
"grad_norm": 0.7017625570297241,
"learning_rate": 3.2055817610062893e-06,
"loss": 0.6722,
"step": 914
},
{
"epoch": 0.3596698113207547,
"grad_norm": 0.6909486651420593,
"learning_rate": 3.203616352201258e-06,
"loss": 0.696,
"step": 915
},
{
"epoch": 0.360062893081761,
"grad_norm": 0.6290726065635681,
"learning_rate": 3.2016509433962266e-06,
"loss": 0.6263,
"step": 916
},
{
"epoch": 0.3604559748427673,
"grad_norm": 0.725310206413269,
"learning_rate": 3.1996855345911954e-06,
"loss": 0.7277,
"step": 917
},
{
"epoch": 0.3608490566037736,
"grad_norm": 0.5694523453712463,
"learning_rate": 3.197720125786164e-06,
"loss": 0.6312,
"step": 918
},
{
"epoch": 0.36124213836477986,
"grad_norm": 1.0193957090377808,
"learning_rate": 3.1957547169811327e-06,
"loss": 0.7906,
"step": 919
},
{
"epoch": 0.36163522012578614,
"grad_norm": 0.7084755301475525,
"learning_rate": 3.1937893081761007e-06,
"loss": 0.6921,
"step": 920
},
{
"epoch": 0.36202830188679247,
"grad_norm": 0.8145508170127869,
"learning_rate": 3.191823899371069e-06,
"loss": 0.8052,
"step": 921
},
{
"epoch": 0.36242138364779874,
"grad_norm": 0.47967100143432617,
"learning_rate": 3.189858490566038e-06,
"loss": 0.7053,
"step": 922
},
{
"epoch": 0.362814465408805,
"grad_norm": 0.7048394680023193,
"learning_rate": 3.1878930817610064e-06,
"loss": 0.7151,
"step": 923
},
{
"epoch": 0.3632075471698113,
"grad_norm": 0.8402290344238281,
"learning_rate": 3.1859276729559752e-06,
"loss": 0.8252,
"step": 924
},
{
"epoch": 0.3636006289308176,
"grad_norm": 0.6275447010993958,
"learning_rate": 3.1839622641509436e-06,
"loss": 0.6283,
"step": 925
},
{
"epoch": 0.3639937106918239,
"grad_norm": 0.7651583552360535,
"learning_rate": 3.1819968553459125e-06,
"loss": 0.7499,
"step": 926
},
{
"epoch": 0.3643867924528302,
"grad_norm": 0.6212397813796997,
"learning_rate": 3.180031446540881e-06,
"loss": 0.7296,
"step": 927
},
{
"epoch": 0.36477987421383645,
"grad_norm": 0.5684332251548767,
"learning_rate": 3.178066037735849e-06,
"loss": 0.7057,
"step": 928
},
{
"epoch": 0.3651729559748428,
"grad_norm": 0.8502115607261658,
"learning_rate": 3.1761006289308178e-06,
"loss": 0.7432,
"step": 929
},
{
"epoch": 0.36556603773584906,
"grad_norm": 0.8275956511497498,
"learning_rate": 3.174135220125786e-06,
"loss": 0.7317,
"step": 930
},
{
"epoch": 0.36595911949685533,
"grad_norm": 0.5729609727859497,
"learning_rate": 3.172169811320755e-06,
"loss": 0.6759,
"step": 931
},
{
"epoch": 0.3663522012578616,
"grad_norm": 0.7062389254570007,
"learning_rate": 3.1702044025157235e-06,
"loss": 0.641,
"step": 932
},
{
"epoch": 0.36674528301886794,
"grad_norm": 0.6936527490615845,
"learning_rate": 3.1682389937106923e-06,
"loss": 0.7001,
"step": 933
},
{
"epoch": 0.3671383647798742,
"grad_norm": 0.7760637998580933,
"learning_rate": 3.1662735849056607e-06,
"loss": 0.704,
"step": 934
},
{
"epoch": 0.3675314465408805,
"grad_norm": 0.7277079224586487,
"learning_rate": 3.1643081761006296e-06,
"loss": 0.7574,
"step": 935
},
{
"epoch": 0.36792452830188677,
"grad_norm": 0.6786287426948547,
"learning_rate": 3.1623427672955976e-06,
"loss": 0.7064,
"step": 936
},
{
"epoch": 0.3683176100628931,
"grad_norm": 0.692562997341156,
"learning_rate": 3.160377358490566e-06,
"loss": 0.689,
"step": 937
},
{
"epoch": 0.3687106918238994,
"grad_norm": 0.6396228671073914,
"learning_rate": 3.158411949685535e-06,
"loss": 0.6896,
"step": 938
},
{
"epoch": 0.36910377358490565,
"grad_norm": 0.5706999897956848,
"learning_rate": 3.1564465408805033e-06,
"loss": 0.6359,
"step": 939
},
{
"epoch": 0.3694968553459119,
"grad_norm": 0.5492956638336182,
"learning_rate": 3.154481132075472e-06,
"loss": 0.7069,
"step": 940
},
{
"epoch": 0.36988993710691825,
"grad_norm": 0.6056336760520935,
"learning_rate": 3.1525157232704405e-06,
"loss": 0.624,
"step": 941
},
{
"epoch": 0.37028301886792453,
"grad_norm": 0.623134195804596,
"learning_rate": 3.1505503144654094e-06,
"loss": 0.5751,
"step": 942
},
{
"epoch": 0.3706761006289308,
"grad_norm": 0.6053591966629028,
"learning_rate": 3.148584905660378e-06,
"loss": 0.646,
"step": 943
},
{
"epoch": 0.3710691823899371,
"grad_norm": 0.8523058891296387,
"learning_rate": 3.146619496855346e-06,
"loss": 0.6387,
"step": 944
},
{
"epoch": 0.3714622641509434,
"grad_norm": 0.8016600012779236,
"learning_rate": 3.1446540880503146e-06,
"loss": 0.7592,
"step": 945
},
{
"epoch": 0.3718553459119497,
"grad_norm": 0.8801593780517578,
"learning_rate": 3.142688679245283e-06,
"loss": 0.7922,
"step": 946
},
{
"epoch": 0.37224842767295596,
"grad_norm": 0.7623356580734253,
"learning_rate": 3.140723270440252e-06,
"loss": 0.7018,
"step": 947
},
{
"epoch": 0.37264150943396224,
"grad_norm": 0.5109173059463501,
"learning_rate": 3.1387578616352203e-06,
"loss": 0.5964,
"step": 948
},
{
"epoch": 0.37303459119496857,
"grad_norm": 0.7063978314399719,
"learning_rate": 3.136792452830189e-06,
"loss": 0.7222,
"step": 949
},
{
"epoch": 0.37342767295597484,
"grad_norm": 0.6266635060310364,
"learning_rate": 3.1348270440251576e-06,
"loss": 0.61,
"step": 950
},
{
"epoch": 0.3738207547169811,
"grad_norm": 0.905984103679657,
"learning_rate": 3.1328616352201265e-06,
"loss": 0.8007,
"step": 951
},
{
"epoch": 0.3742138364779874,
"grad_norm": 0.790346622467041,
"learning_rate": 3.1308962264150945e-06,
"loss": 0.6798,
"step": 952
},
{
"epoch": 0.3746069182389937,
"grad_norm": 0.6115738153457642,
"learning_rate": 3.128930817610063e-06,
"loss": 0.6357,
"step": 953
},
{
"epoch": 0.375,
"grad_norm": 0.48611846566200256,
"learning_rate": 3.1269654088050317e-06,
"loss": 0.5873,
"step": 954
},
{
"epoch": 0.3753930817610063,
"grad_norm": 0.7266239523887634,
"learning_rate": 3.125e-06,
"loss": 0.7688,
"step": 955
},
{
"epoch": 0.3757861635220126,
"grad_norm": 0.7782076001167297,
"learning_rate": 3.123034591194969e-06,
"loss": 0.8019,
"step": 956
},
{
"epoch": 0.3761792452830189,
"grad_norm": 0.9216904640197754,
"learning_rate": 3.1210691823899374e-06,
"loss": 0.8132,
"step": 957
},
{
"epoch": 0.37657232704402516,
"grad_norm": 0.7297998070716858,
"learning_rate": 3.1191037735849063e-06,
"loss": 0.6923,
"step": 958
},
{
"epoch": 0.37696540880503143,
"grad_norm": 0.5805630683898926,
"learning_rate": 3.1171383647798743e-06,
"loss": 0.695,
"step": 959
},
{
"epoch": 0.37735849056603776,
"grad_norm": 0.7553980350494385,
"learning_rate": 3.1151729559748427e-06,
"loss": 0.7686,
"step": 960
},
{
"epoch": 0.37775157232704404,
"grad_norm": 0.6018033027648926,
"learning_rate": 3.1132075471698115e-06,
"loss": 0.6671,
"step": 961
},
{
"epoch": 0.3781446540880503,
"grad_norm": 0.6250959038734436,
"learning_rate": 3.11124213836478e-06,
"loss": 0.573,
"step": 962
},
{
"epoch": 0.3785377358490566,
"grad_norm": 0.6415625214576721,
"learning_rate": 3.109276729559749e-06,
"loss": 0.725,
"step": 963
},
{
"epoch": 0.3789308176100629,
"grad_norm": 0.603950023651123,
"learning_rate": 3.1073113207547172e-06,
"loss": 0.6799,
"step": 964
},
{
"epoch": 0.3793238993710692,
"grad_norm": 0.7063471078872681,
"learning_rate": 3.105345911949686e-06,
"loss": 0.7013,
"step": 965
},
{
"epoch": 0.37971698113207547,
"grad_norm": 0.4992704689502716,
"learning_rate": 3.1033805031446545e-06,
"loss": 0.5954,
"step": 966
},
{
"epoch": 0.38011006289308175,
"grad_norm": 0.8308043479919434,
"learning_rate": 3.1014150943396225e-06,
"loss": 0.7526,
"step": 967
},
{
"epoch": 0.3805031446540881,
"grad_norm": 0.939951479434967,
"learning_rate": 3.0994496855345913e-06,
"loss": 0.8788,
"step": 968
},
{
"epoch": 0.38089622641509435,
"grad_norm": 0.7524932026863098,
"learning_rate": 3.0974842767295598e-06,
"loss": 0.7605,
"step": 969
},
{
"epoch": 0.3812893081761006,
"grad_norm": 0.574846625328064,
"learning_rate": 3.0955188679245286e-06,
"loss": 0.5673,
"step": 970
},
{
"epoch": 0.3816823899371069,
"grad_norm": 0.748582124710083,
"learning_rate": 3.093553459119497e-06,
"loss": 0.6908,
"step": 971
},
{
"epoch": 0.38207547169811323,
"grad_norm": 0.7016021013259888,
"learning_rate": 3.091588050314466e-06,
"loss": 0.7152,
"step": 972
},
{
"epoch": 0.3824685534591195,
"grad_norm": 0.5162691473960876,
"learning_rate": 3.0896226415094343e-06,
"loss": 0.6187,
"step": 973
},
{
"epoch": 0.3828616352201258,
"grad_norm": 0.6713387966156006,
"learning_rate": 3.087657232704403e-06,
"loss": 0.6808,
"step": 974
},
{
"epoch": 0.38325471698113206,
"grad_norm": 0.7065016031265259,
"learning_rate": 3.085691823899371e-06,
"loss": 0.7291,
"step": 975
},
{
"epoch": 0.3836477987421384,
"grad_norm": 0.9689213633537292,
"learning_rate": 3.0837264150943396e-06,
"loss": 0.7438,
"step": 976
},
{
"epoch": 0.38404088050314467,
"grad_norm": 0.8745124936103821,
"learning_rate": 3.0817610062893084e-06,
"loss": 0.738,
"step": 977
},
{
"epoch": 0.38443396226415094,
"grad_norm": 0.6272825002670288,
"learning_rate": 3.079795597484277e-06,
"loss": 0.6102,
"step": 978
},
{
"epoch": 0.3848270440251572,
"grad_norm": 0.7032546401023865,
"learning_rate": 3.0778301886792457e-06,
"loss": 0.663,
"step": 979
},
{
"epoch": 0.38522012578616355,
"grad_norm": 0.5791395902633667,
"learning_rate": 3.075864779874214e-06,
"loss": 0.6876,
"step": 980
},
{
"epoch": 0.3856132075471698,
"grad_norm": 0.7973341345787048,
"learning_rate": 3.073899371069183e-06,
"loss": 0.7888,
"step": 981
},
{
"epoch": 0.3860062893081761,
"grad_norm": 0.7559678554534912,
"learning_rate": 3.0719339622641514e-06,
"loss": 0.7037,
"step": 982
},
{
"epoch": 0.3863993710691824,
"grad_norm": 0.5916761755943298,
"learning_rate": 3.0699685534591194e-06,
"loss": 0.5459,
"step": 983
},
{
"epoch": 0.3867924528301887,
"grad_norm": 0.7382115721702576,
"learning_rate": 3.0680031446540882e-06,
"loss": 0.7399,
"step": 984
},
{
"epoch": 0.387185534591195,
"grad_norm": 0.4721450209617615,
"learning_rate": 3.0660377358490567e-06,
"loss": 0.5853,
"step": 985
},
{
"epoch": 0.38757861635220126,
"grad_norm": 0.6330047845840454,
"learning_rate": 3.0640723270440255e-06,
"loss": 0.6551,
"step": 986
},
{
"epoch": 0.38797169811320753,
"grad_norm": 0.669251561164856,
"learning_rate": 3.062106918238994e-06,
"loss": 0.7543,
"step": 987
},
{
"epoch": 0.38836477987421386,
"grad_norm": 0.7437297105789185,
"learning_rate": 3.0601415094339628e-06,
"loss": 0.7081,
"step": 988
},
{
"epoch": 0.38875786163522014,
"grad_norm": 0.6762712001800537,
"learning_rate": 3.058176100628931e-06,
"loss": 0.7043,
"step": 989
},
{
"epoch": 0.3891509433962264,
"grad_norm": 0.7705761790275574,
"learning_rate": 3.0562106918239e-06,
"loss": 0.7358,
"step": 990
},
{
"epoch": 0.3895440251572327,
"grad_norm": 0.6579947471618652,
"learning_rate": 3.054245283018868e-06,
"loss": 0.6816,
"step": 991
},
{
"epoch": 0.389937106918239,
"grad_norm": 0.7796435952186584,
"learning_rate": 3.0522798742138365e-06,
"loss": 0.7559,
"step": 992
},
{
"epoch": 0.3903301886792453,
"grad_norm": 0.8162177801132202,
"learning_rate": 3.0503144654088053e-06,
"loss": 0.6995,
"step": 993
},
{
"epoch": 0.39072327044025157,
"grad_norm": 0.8338039517402649,
"learning_rate": 3.0483490566037737e-06,
"loss": 0.7201,
"step": 994
},
{
"epoch": 0.39111635220125784,
"grad_norm": 0.5165339112281799,
"learning_rate": 3.0463836477987426e-06,
"loss": 0.6729,
"step": 995
},
{
"epoch": 0.3915094339622642,
"grad_norm": 0.7111666202545166,
"learning_rate": 3.044418238993711e-06,
"loss": 0.7035,
"step": 996
},
{
"epoch": 0.39190251572327045,
"grad_norm": 0.7852262258529663,
"learning_rate": 3.04245283018868e-06,
"loss": 0.7045,
"step": 997
},
{
"epoch": 0.3922955974842767,
"grad_norm": 0.8401679396629333,
"learning_rate": 3.0404874213836483e-06,
"loss": 0.7991,
"step": 998
},
{
"epoch": 0.392688679245283,
"grad_norm": 0.6312411427497864,
"learning_rate": 3.0385220125786163e-06,
"loss": 0.6865,
"step": 999
},
{
"epoch": 0.39308176100628933,
"grad_norm": 0.6451079249382019,
"learning_rate": 3.036556603773585e-06,
"loss": 0.7092,
"step": 1000
},
{
"epoch": 0.3934748427672956,
"grad_norm": 0.9083591103553772,
"learning_rate": 3.0345911949685535e-06,
"loss": 0.7286,
"step": 1001
},
{
"epoch": 0.3938679245283019,
"grad_norm": 0.9039638042449951,
"learning_rate": 3.0326257861635224e-06,
"loss": 0.7167,
"step": 1002
},
{
"epoch": 0.39426100628930816,
"grad_norm": 0.7467072010040283,
"learning_rate": 3.030660377358491e-06,
"loss": 0.7391,
"step": 1003
},
{
"epoch": 0.3946540880503145,
"grad_norm": 0.5601504445075989,
"learning_rate": 3.0286949685534597e-06,
"loss": 0.5742,
"step": 1004
},
{
"epoch": 0.39504716981132076,
"grad_norm": 0.6848729252815247,
"learning_rate": 3.026729559748428e-06,
"loss": 0.658,
"step": 1005
},
{
"epoch": 0.39544025157232704,
"grad_norm": 0.7195609211921692,
"learning_rate": 3.0247641509433965e-06,
"loss": 0.7212,
"step": 1006
},
{
"epoch": 0.3958333333333333,
"grad_norm": 0.8385602235794067,
"learning_rate": 3.022798742138365e-06,
"loss": 0.6466,
"step": 1007
},
{
"epoch": 0.39622641509433965,
"grad_norm": 0.7911286354064941,
"learning_rate": 3.0208333333333334e-06,
"loss": 0.7952,
"step": 1008
},
{
"epoch": 0.3966194968553459,
"grad_norm": 0.6798897981643677,
"learning_rate": 3.018867924528302e-06,
"loss": 0.6843,
"step": 1009
},
{
"epoch": 0.3970125786163522,
"grad_norm": 0.7566647529602051,
"learning_rate": 3.0169025157232706e-06,
"loss": 0.6833,
"step": 1010
},
{
"epoch": 0.39740566037735847,
"grad_norm": 0.6413459181785583,
"learning_rate": 3.0149371069182395e-06,
"loss": 0.6541,
"step": 1011
},
{
"epoch": 0.3977987421383648,
"grad_norm": 0.7588937282562256,
"learning_rate": 3.012971698113208e-06,
"loss": 0.6587,
"step": 1012
},
{
"epoch": 0.3981918238993711,
"grad_norm": 0.5974403023719788,
"learning_rate": 3.0110062893081763e-06,
"loss": 0.6554,
"step": 1013
},
{
"epoch": 0.39858490566037735,
"grad_norm": 0.7417561411857605,
"learning_rate": 3.009040880503145e-06,
"loss": 0.774,
"step": 1014
},
{
"epoch": 0.39897798742138363,
"grad_norm": 0.5492700934410095,
"learning_rate": 3.007075471698113e-06,
"loss": 0.5991,
"step": 1015
},
{
"epoch": 0.39937106918238996,
"grad_norm": 0.6512425541877747,
"learning_rate": 3.005110062893082e-06,
"loss": 0.6667,
"step": 1016
},
{
"epoch": 0.39976415094339623,
"grad_norm": 0.86830073595047,
"learning_rate": 3.0031446540880504e-06,
"loss": 0.749,
"step": 1017
},
{
"epoch": 0.4001572327044025,
"grad_norm": 0.5500980615615845,
"learning_rate": 3.0011792452830193e-06,
"loss": 0.6733,
"step": 1018
},
{
"epoch": 0.4005503144654088,
"grad_norm": 0.5683943033218384,
"learning_rate": 2.9992138364779877e-06,
"loss": 0.702,
"step": 1019
},
{
"epoch": 0.4009433962264151,
"grad_norm": 0.6889972686767578,
"learning_rate": 2.997248427672956e-06,
"loss": 0.6315,
"step": 1020
},
{
"epoch": 0.4013364779874214,
"grad_norm": 0.5596610307693481,
"learning_rate": 2.995283018867925e-06,
"loss": 0.6797,
"step": 1021
},
{
"epoch": 0.40172955974842767,
"grad_norm": 0.7305598258972168,
"learning_rate": 2.9933176100628934e-06,
"loss": 0.6031,
"step": 1022
},
{
"epoch": 0.40212264150943394,
"grad_norm": 0.5764161348342896,
"learning_rate": 2.991352201257862e-06,
"loss": 0.662,
"step": 1023
},
{
"epoch": 0.4025157232704403,
"grad_norm": 0.6990935206413269,
"learning_rate": 2.9893867924528302e-06,
"loss": 0.7198,
"step": 1024
},
{
"epoch": 0.40290880503144655,
"grad_norm": 0.932759165763855,
"learning_rate": 2.987421383647799e-06,
"loss": 0.7283,
"step": 1025
},
{
"epoch": 0.4033018867924528,
"grad_norm": 0.8571686148643494,
"learning_rate": 2.9854559748427675e-06,
"loss": 0.7241,
"step": 1026
},
{
"epoch": 0.4036949685534591,
"grad_norm": 0.4383394718170166,
"learning_rate": 2.983490566037736e-06,
"loss": 0.5464,
"step": 1027
},
{
"epoch": 0.40408805031446543,
"grad_norm": 0.8871681094169617,
"learning_rate": 2.9815251572327048e-06,
"loss": 0.7994,
"step": 1028
},
{
"epoch": 0.4044811320754717,
"grad_norm": 0.791783332824707,
"learning_rate": 2.979559748427673e-06,
"loss": 0.7788,
"step": 1029
},
{
"epoch": 0.404874213836478,
"grad_norm": 0.7186205387115479,
"learning_rate": 2.977594339622642e-06,
"loss": 0.8328,
"step": 1030
},
{
"epoch": 0.40526729559748426,
"grad_norm": 0.5765904188156128,
"learning_rate": 2.97562893081761e-06,
"loss": 0.5772,
"step": 1031
},
{
"epoch": 0.4056603773584906,
"grad_norm": 0.7256649732589722,
"learning_rate": 2.973663522012579e-06,
"loss": 0.7274,
"step": 1032
},
{
"epoch": 0.40605345911949686,
"grad_norm": 0.6426676511764526,
"learning_rate": 2.9716981132075473e-06,
"loss": 0.6697,
"step": 1033
},
{
"epoch": 0.40644654088050314,
"grad_norm": 0.6621936559677124,
"learning_rate": 2.9697327044025157e-06,
"loss": 0.7231,
"step": 1034
},
{
"epoch": 0.4068396226415094,
"grad_norm": 0.6613886952400208,
"learning_rate": 2.9677672955974846e-06,
"loss": 0.6572,
"step": 1035
},
{
"epoch": 0.40723270440251574,
"grad_norm": 0.7529765367507935,
"learning_rate": 2.965801886792453e-06,
"loss": 0.6945,
"step": 1036
},
{
"epoch": 0.407625786163522,
"grad_norm": 0.7473201751708984,
"learning_rate": 2.963836477987422e-06,
"loss": 0.6465,
"step": 1037
},
{
"epoch": 0.4080188679245283,
"grad_norm": 0.7480108737945557,
"learning_rate": 2.9618710691823903e-06,
"loss": 0.6609,
"step": 1038
},
{
"epoch": 0.40841194968553457,
"grad_norm": 0.7269521355628967,
"learning_rate": 2.9599056603773587e-06,
"loss": 0.6201,
"step": 1039
},
{
"epoch": 0.4088050314465409,
"grad_norm": 0.7366546988487244,
"learning_rate": 2.957940251572327e-06,
"loss": 0.7434,
"step": 1040
},
{
"epoch": 0.4091981132075472,
"grad_norm": 0.6154018044471741,
"learning_rate": 2.9559748427672955e-06,
"loss": 0.7025,
"step": 1041
},
{
"epoch": 0.40959119496855345,
"grad_norm": 0.4963739514350891,
"learning_rate": 2.9540094339622644e-06,
"loss": 0.5953,
"step": 1042
},
{
"epoch": 0.4099842767295597,
"grad_norm": 0.667646586894989,
"learning_rate": 2.952044025157233e-06,
"loss": 0.6923,
"step": 1043
},
{
"epoch": 0.41037735849056606,
"grad_norm": 0.9173179864883423,
"learning_rate": 2.9500786163522017e-06,
"loss": 0.8013,
"step": 1044
},
{
"epoch": 0.41077044025157233,
"grad_norm": 0.911105751991272,
"learning_rate": 2.94811320754717e-06,
"loss": 0.6601,
"step": 1045
},
{
"epoch": 0.4111635220125786,
"grad_norm": 0.7773126363754272,
"learning_rate": 2.946147798742139e-06,
"loss": 0.6257,
"step": 1046
},
{
"epoch": 0.4115566037735849,
"grad_norm": 0.7495285272598267,
"learning_rate": 2.944182389937107e-06,
"loss": 0.7145,
"step": 1047
},
{
"epoch": 0.4119496855345912,
"grad_norm": 0.6634531617164612,
"learning_rate": 2.9422169811320754e-06,
"loss": 0.6084,
"step": 1048
},
{
"epoch": 0.4123427672955975,
"grad_norm": 0.6873542666435242,
"learning_rate": 2.940251572327044e-06,
"loss": 0.6509,
"step": 1049
},
{
"epoch": 0.41273584905660377,
"grad_norm": 0.7291033267974854,
"learning_rate": 2.9382861635220126e-06,
"loss": 0.7902,
"step": 1050
},
{
"epoch": 0.41312893081761004,
"grad_norm": 0.6920795440673828,
"learning_rate": 2.9363207547169815e-06,
"loss": 0.7051,
"step": 1051
},
{
"epoch": 0.41352201257861637,
"grad_norm": 0.791043221950531,
"learning_rate": 2.93435534591195e-06,
"loss": 0.743,
"step": 1052
},
{
"epoch": 0.41391509433962265,
"grad_norm": 0.4831397235393524,
"learning_rate": 2.9323899371069187e-06,
"loss": 0.6538,
"step": 1053
},
{
"epoch": 0.4143081761006289,
"grad_norm": 0.6534374952316284,
"learning_rate": 2.930424528301887e-06,
"loss": 0.5843,
"step": 1054
},
{
"epoch": 0.4147012578616352,
"grad_norm": 0.6760728359222412,
"learning_rate": 2.928459119496855e-06,
"loss": 0.5952,
"step": 1055
},
{
"epoch": 0.41509433962264153,
"grad_norm": 0.6800026297569275,
"learning_rate": 2.926493710691824e-06,
"loss": 0.6641,
"step": 1056
},
{
"epoch": 0.4154874213836478,
"grad_norm": 0.9393219947814941,
"learning_rate": 2.9245283018867924e-06,
"loss": 0.9093,
"step": 1057
},
{
"epoch": 0.4158805031446541,
"grad_norm": 0.6511555910110474,
"learning_rate": 2.9225628930817613e-06,
"loss": 0.5853,
"step": 1058
},
{
"epoch": 0.41627358490566035,
"grad_norm": 0.6415976285934448,
"learning_rate": 2.9205974842767297e-06,
"loss": 0.6553,
"step": 1059
},
{
"epoch": 0.4166666666666667,
"grad_norm": 0.8780087232589722,
"learning_rate": 2.9186320754716986e-06,
"loss": 0.7701,
"step": 1060
},
{
"epoch": 0.41705974842767296,
"grad_norm": 0.8306125998497009,
"learning_rate": 2.916666666666667e-06,
"loss": 0.7697,
"step": 1061
},
{
"epoch": 0.41745283018867924,
"grad_norm": 0.8024614453315735,
"learning_rate": 2.914701257861636e-06,
"loss": 0.7135,
"step": 1062
},
{
"epoch": 0.4178459119496855,
"grad_norm": 0.7001165151596069,
"learning_rate": 2.912735849056604e-06,
"loss": 0.808,
"step": 1063
},
{
"epoch": 0.41823899371069184,
"grad_norm": 0.7673312425613403,
"learning_rate": 2.9107704402515722e-06,
"loss": 0.6674,
"step": 1064
},
{
"epoch": 0.4186320754716981,
"grad_norm": 0.7684844732284546,
"learning_rate": 2.908805031446541e-06,
"loss": 0.7711,
"step": 1065
},
{
"epoch": 0.4190251572327044,
"grad_norm": 0.6615145206451416,
"learning_rate": 2.9068396226415095e-06,
"loss": 0.6721,
"step": 1066
},
{
"epoch": 0.41941823899371067,
"grad_norm": 0.6366955041885376,
"learning_rate": 2.9048742138364784e-06,
"loss": 0.7147,
"step": 1067
},
{
"epoch": 0.419811320754717,
"grad_norm": 0.6137346625328064,
"learning_rate": 2.9029088050314468e-06,
"loss": 0.6816,
"step": 1068
},
{
"epoch": 0.4202044025157233,
"grad_norm": 0.8133223652839661,
"learning_rate": 2.9009433962264156e-06,
"loss": 0.7522,
"step": 1069
},
{
"epoch": 0.42059748427672955,
"grad_norm": 0.8280315399169922,
"learning_rate": 2.898977987421384e-06,
"loss": 0.7723,
"step": 1070
},
{
"epoch": 0.4209905660377358,
"grad_norm": 0.7544548511505127,
"learning_rate": 2.897012578616352e-06,
"loss": 0.7582,
"step": 1071
},
{
"epoch": 0.42138364779874216,
"grad_norm": 0.7089150547981262,
"learning_rate": 2.895047169811321e-06,
"loss": 0.7005,
"step": 1072
},
{
"epoch": 0.42177672955974843,
"grad_norm": 0.6980677843093872,
"learning_rate": 2.8930817610062893e-06,
"loss": 0.7053,
"step": 1073
},
{
"epoch": 0.4221698113207547,
"grad_norm": 0.7463325262069702,
"learning_rate": 2.891116352201258e-06,
"loss": 0.7242,
"step": 1074
},
{
"epoch": 0.422562893081761,
"grad_norm": 0.7267622947692871,
"learning_rate": 2.8891509433962266e-06,
"loss": 0.7427,
"step": 1075
},
{
"epoch": 0.4229559748427673,
"grad_norm": 0.7070834636688232,
"learning_rate": 2.8871855345911954e-06,
"loss": 0.7286,
"step": 1076
},
{
"epoch": 0.4233490566037736,
"grad_norm": 0.6570694446563721,
"learning_rate": 2.885220125786164e-06,
"loss": 0.7529,
"step": 1077
},
{
"epoch": 0.42374213836477986,
"grad_norm": 0.712973415851593,
"learning_rate": 2.8832547169811327e-06,
"loss": 0.651,
"step": 1078
},
{
"epoch": 0.42413522012578614,
"grad_norm": 0.5979359745979309,
"learning_rate": 2.8812893081761007e-06,
"loss": 0.6453,
"step": 1079
},
{
"epoch": 0.42452830188679247,
"grad_norm": 0.5918779373168945,
"learning_rate": 2.879323899371069e-06,
"loss": 0.6151,
"step": 1080
},
{
"epoch": 0.42492138364779874,
"grad_norm": 0.7464630007743835,
"learning_rate": 2.877358490566038e-06,
"loss": 0.6879,
"step": 1081
},
{
"epoch": 0.425314465408805,
"grad_norm": 0.9332670569419861,
"learning_rate": 2.8753930817610064e-06,
"loss": 0.7885,
"step": 1082
},
{
"epoch": 0.4257075471698113,
"grad_norm": 0.6473411321640015,
"learning_rate": 2.8734276729559752e-06,
"loss": 0.7375,
"step": 1083
},
{
"epoch": 0.4261006289308176,
"grad_norm": 0.47155243158340454,
"learning_rate": 2.8714622641509437e-06,
"loss": 0.6112,
"step": 1084
},
{
"epoch": 0.4264937106918239,
"grad_norm": 0.7482328414916992,
"learning_rate": 2.8694968553459125e-06,
"loss": 0.6727,
"step": 1085
},
{
"epoch": 0.4268867924528302,
"grad_norm": 0.7974027395248413,
"learning_rate": 2.867531446540881e-06,
"loss": 0.704,
"step": 1086
},
{
"epoch": 0.42727987421383645,
"grad_norm": 1.0152580738067627,
"learning_rate": 2.865566037735849e-06,
"loss": 0.8248,
"step": 1087
},
{
"epoch": 0.4276729559748428,
"grad_norm": 0.6010071039199829,
"learning_rate": 2.8636006289308178e-06,
"loss": 0.7007,
"step": 1088
},
{
"epoch": 0.42806603773584906,
"grad_norm": 0.7331111431121826,
"learning_rate": 2.861635220125786e-06,
"loss": 0.724,
"step": 1089
},
{
"epoch": 0.42845911949685533,
"grad_norm": 0.5621688961982727,
"learning_rate": 2.859669811320755e-06,
"loss": 0.6619,
"step": 1090
},
{
"epoch": 0.4288522012578616,
"grad_norm": 0.7338464856147766,
"learning_rate": 2.8577044025157235e-06,
"loss": 0.6799,
"step": 1091
},
{
"epoch": 0.42924528301886794,
"grad_norm": 0.6197801828384399,
"learning_rate": 2.8557389937106923e-06,
"loss": 0.6716,
"step": 1092
},
{
"epoch": 0.4296383647798742,
"grad_norm": 0.6634385585784912,
"learning_rate": 2.8537735849056608e-06,
"loss": 0.7379,
"step": 1093
},
{
"epoch": 0.4300314465408805,
"grad_norm": 0.4419562816619873,
"learning_rate": 2.8518081761006296e-06,
"loss": 0.5685,
"step": 1094
},
{
"epoch": 0.43042452830188677,
"grad_norm": 0.6517072319984436,
"learning_rate": 2.8498427672955976e-06,
"loss": 0.6169,
"step": 1095
},
{
"epoch": 0.4308176100628931,
"grad_norm": 0.6441124677658081,
"learning_rate": 2.847877358490566e-06,
"loss": 0.7095,
"step": 1096
},
{
"epoch": 0.4312106918238994,
"grad_norm": 0.7386425137519836,
"learning_rate": 2.845911949685535e-06,
"loss": 0.683,
"step": 1097
},
{
"epoch": 0.43160377358490565,
"grad_norm": 0.6729860305786133,
"learning_rate": 2.8439465408805033e-06,
"loss": 0.7076,
"step": 1098
},
{
"epoch": 0.4319968553459119,
"grad_norm": 0.7944410443305969,
"learning_rate": 2.841981132075472e-06,
"loss": 0.7319,
"step": 1099
},
{
"epoch": 0.43238993710691825,
"grad_norm": 0.6964000463485718,
"learning_rate": 2.8400157232704406e-06,
"loss": 0.6645,
"step": 1100
},
{
"epoch": 0.43278301886792453,
"grad_norm": 0.579157292842865,
"learning_rate": 2.8380503144654094e-06,
"loss": 0.6367,
"step": 1101
},
{
"epoch": 0.4331761006289308,
"grad_norm": 1.0114585161209106,
"learning_rate": 2.836084905660378e-06,
"loss": 0.6603,
"step": 1102
},
{
"epoch": 0.4335691823899371,
"grad_norm": 0.7365342974662781,
"learning_rate": 2.834119496855346e-06,
"loss": 0.7374,
"step": 1103
},
{
"epoch": 0.4339622641509434,
"grad_norm": 0.883794903755188,
"learning_rate": 2.8321540880503147e-06,
"loss": 0.7307,
"step": 1104
},
{
"epoch": 0.4343553459119497,
"grad_norm": 0.962576687335968,
"learning_rate": 2.830188679245283e-06,
"loss": 0.7837,
"step": 1105
},
{
"epoch": 0.43474842767295596,
"grad_norm": 0.6893895268440247,
"learning_rate": 2.828223270440252e-06,
"loss": 0.7266,
"step": 1106
},
{
"epoch": 0.43514150943396224,
"grad_norm": 0.7192584276199341,
"learning_rate": 2.8262578616352204e-06,
"loss": 0.7782,
"step": 1107
},
{
"epoch": 0.43553459119496857,
"grad_norm": 0.6722439527511597,
"learning_rate": 2.8242924528301892e-06,
"loss": 0.6523,
"step": 1108
},
{
"epoch": 0.43592767295597484,
"grad_norm": 0.5174304246902466,
"learning_rate": 2.8223270440251576e-06,
"loss": 0.5865,
"step": 1109
},
{
"epoch": 0.4363207547169811,
"grad_norm": 0.8323651552200317,
"learning_rate": 2.8203616352201265e-06,
"loss": 0.628,
"step": 1110
},
{
"epoch": 0.4367138364779874,
"grad_norm": 0.6904851198196411,
"learning_rate": 2.8183962264150945e-06,
"loss": 0.7357,
"step": 1111
},
{
"epoch": 0.4371069182389937,
"grad_norm": 0.8741118907928467,
"learning_rate": 2.816430817610063e-06,
"loss": 0.6698,
"step": 1112
},
{
"epoch": 0.4375,
"grad_norm": 0.8097689747810364,
"learning_rate": 2.8144654088050318e-06,
"loss": 0.6432,
"step": 1113
},
{
"epoch": 0.4378930817610063,
"grad_norm": 0.695667564868927,
"learning_rate": 2.8125e-06,
"loss": 0.6357,
"step": 1114
},
{
"epoch": 0.4382861635220126,
"grad_norm": 0.6892862319946289,
"learning_rate": 2.810534591194969e-06,
"loss": 0.5995,
"step": 1115
},
{
"epoch": 0.4386792452830189,
"grad_norm": 0.6002050638198853,
"learning_rate": 2.8085691823899374e-06,
"loss": 0.645,
"step": 1116
},
{
"epoch": 0.43907232704402516,
"grad_norm": 0.781235933303833,
"learning_rate": 2.8066037735849063e-06,
"loss": 0.7493,
"step": 1117
},
{
"epoch": 0.43946540880503143,
"grad_norm": 0.6305332779884338,
"learning_rate": 2.8046383647798743e-06,
"loss": 0.6129,
"step": 1118
},
{
"epoch": 0.43985849056603776,
"grad_norm": 0.7028183341026306,
"learning_rate": 2.8026729559748427e-06,
"loss": 0.6023,
"step": 1119
},
{
"epoch": 0.44025157232704404,
"grad_norm": 0.643160343170166,
"learning_rate": 2.8007075471698116e-06,
"loss": 0.6627,
"step": 1120
},
{
"epoch": 0.4406446540880503,
"grad_norm": 0.8067569732666016,
"learning_rate": 2.79874213836478e-06,
"loss": 0.7184,
"step": 1121
},
{
"epoch": 0.4410377358490566,
"grad_norm": 0.8040790557861328,
"learning_rate": 2.796776729559749e-06,
"loss": 0.827,
"step": 1122
},
{
"epoch": 0.4414308176100629,
"grad_norm": 0.7451358437538147,
"learning_rate": 2.7948113207547173e-06,
"loss": 0.689,
"step": 1123
},
{
"epoch": 0.4418238993710692,
"grad_norm": 0.7173066735267639,
"learning_rate": 2.792845911949686e-06,
"loss": 0.659,
"step": 1124
},
{
"epoch": 0.44221698113207547,
"grad_norm": 0.6626467108726501,
"learning_rate": 2.7908805031446545e-06,
"loss": 0.6617,
"step": 1125
},
{
"epoch": 0.44261006289308175,
"grad_norm": 0.7028668522834778,
"learning_rate": 2.7889150943396225e-06,
"loss": 0.5878,
"step": 1126
},
{
"epoch": 0.4430031446540881,
"grad_norm": 1.0061613321304321,
"learning_rate": 2.7869496855345914e-06,
"loss": 0.6455,
"step": 1127
},
{
"epoch": 0.44339622641509435,
"grad_norm": 0.8330791592597961,
"learning_rate": 2.78498427672956e-06,
"loss": 0.814,
"step": 1128
},
{
"epoch": 0.4437893081761006,
"grad_norm": 0.4720902740955353,
"learning_rate": 2.7830188679245286e-06,
"loss": 0.614,
"step": 1129
},
{
"epoch": 0.4441823899371069,
"grad_norm": 0.8637996315956116,
"learning_rate": 2.781053459119497e-06,
"loss": 0.7339,
"step": 1130
},
{
"epoch": 0.44457547169811323,
"grad_norm": 0.6245742440223694,
"learning_rate": 2.779088050314466e-06,
"loss": 0.6456,
"step": 1131
},
{
"epoch": 0.4449685534591195,
"grad_norm": 0.8598979711532593,
"learning_rate": 2.7771226415094343e-06,
"loss": 0.7526,
"step": 1132
},
{
"epoch": 0.4453616352201258,
"grad_norm": 0.5663571953773499,
"learning_rate": 2.7751572327044028e-06,
"loss": 0.6222,
"step": 1133
},
{
"epoch": 0.44575471698113206,
"grad_norm": 0.623297929763794,
"learning_rate": 2.773191823899371e-06,
"loss": 0.6565,
"step": 1134
},
{
"epoch": 0.4461477987421384,
"grad_norm": 0.5693382620811462,
"learning_rate": 2.7712264150943396e-06,
"loss": 0.6879,
"step": 1135
},
{
"epoch": 0.44654088050314467,
"grad_norm": 0.6104090213775635,
"learning_rate": 2.7692610062893085e-06,
"loss": 0.7053,
"step": 1136
},
{
"epoch": 0.44693396226415094,
"grad_norm": 0.6537973284721375,
"learning_rate": 2.767295597484277e-06,
"loss": 0.7444,
"step": 1137
},
{
"epoch": 0.4473270440251572,
"grad_norm": 0.6328637599945068,
"learning_rate": 2.7653301886792457e-06,
"loss": 0.6699,
"step": 1138
},
{
"epoch": 0.44772012578616355,
"grad_norm": 0.6024653315544128,
"learning_rate": 2.763364779874214e-06,
"loss": 0.6727,
"step": 1139
},
{
"epoch": 0.4481132075471698,
"grad_norm": 0.6828110218048096,
"learning_rate": 2.7613993710691826e-06,
"loss": 0.749,
"step": 1140
},
{
"epoch": 0.4485062893081761,
"grad_norm": 0.7768702507019043,
"learning_rate": 2.7594339622641514e-06,
"loss": 0.6888,
"step": 1141
},
{
"epoch": 0.4488993710691824,
"grad_norm": 0.6281961798667908,
"learning_rate": 2.7574685534591194e-06,
"loss": 0.6153,
"step": 1142
},
{
"epoch": 0.4492924528301887,
"grad_norm": 0.8371246457099915,
"learning_rate": 2.7555031446540883e-06,
"loss": 0.6991,
"step": 1143
},
{
"epoch": 0.449685534591195,
"grad_norm": 0.6776599287986755,
"learning_rate": 2.7535377358490567e-06,
"loss": 0.7433,
"step": 1144
},
{
"epoch": 0.45007861635220126,
"grad_norm": 0.6529794335365295,
"learning_rate": 2.7515723270440255e-06,
"loss": 0.6997,
"step": 1145
},
{
"epoch": 0.45047169811320753,
"grad_norm": 0.7979133725166321,
"learning_rate": 2.749606918238994e-06,
"loss": 0.6323,
"step": 1146
},
{
"epoch": 0.45086477987421386,
"grad_norm": 0.582793653011322,
"learning_rate": 2.7476415094339624e-06,
"loss": 0.5929,
"step": 1147
},
{
"epoch": 0.45125786163522014,
"grad_norm": 0.6265951991081238,
"learning_rate": 2.7456761006289312e-06,
"loss": 0.7778,
"step": 1148
},
{
"epoch": 0.4516509433962264,
"grad_norm": 0.7754130363464355,
"learning_rate": 2.7437106918238996e-06,
"loss": 0.7124,
"step": 1149
},
{
"epoch": 0.4520440251572327,
"grad_norm": 0.7138745188713074,
"learning_rate": 2.741745283018868e-06,
"loss": 0.7295,
"step": 1150
},
{
"epoch": 0.452437106918239,
"grad_norm": 0.797276496887207,
"learning_rate": 2.7397798742138365e-06,
"loss": 0.6609,
"step": 1151
},
{
"epoch": 0.4528301886792453,
"grad_norm": 0.5241825580596924,
"learning_rate": 2.7378144654088053e-06,
"loss": 0.6283,
"step": 1152
},
{
"epoch": 0.45322327044025157,
"grad_norm": 0.652489960193634,
"learning_rate": 2.7358490566037738e-06,
"loss": 0.653,
"step": 1153
},
{
"epoch": 0.45361635220125784,
"grad_norm": 0.6647413969039917,
"learning_rate": 2.733883647798742e-06,
"loss": 0.6794,
"step": 1154
},
{
"epoch": 0.4540094339622642,
"grad_norm": 0.7657634615898132,
"learning_rate": 2.731918238993711e-06,
"loss": 0.7784,
"step": 1155
},
{
"epoch": 0.45440251572327045,
"grad_norm": 0.516899585723877,
"learning_rate": 2.7299528301886795e-06,
"loss": 0.6434,
"step": 1156
},
{
"epoch": 0.4547955974842767,
"grad_norm": 0.6802290678024292,
"learning_rate": 2.7279874213836483e-06,
"loss": 0.6896,
"step": 1157
},
{
"epoch": 0.455188679245283,
"grad_norm": 0.8072131276130676,
"learning_rate": 2.7260220125786163e-06,
"loss": 0.7599,
"step": 1158
},
{
"epoch": 0.45558176100628933,
"grad_norm": 0.6927570104598999,
"learning_rate": 2.724056603773585e-06,
"loss": 0.6437,
"step": 1159
},
{
"epoch": 0.4559748427672956,
"grad_norm": 0.8016869425773621,
"learning_rate": 2.7220911949685536e-06,
"loss": 0.7327,
"step": 1160
},
{
"epoch": 0.4563679245283019,
"grad_norm": 0.7104259133338928,
"learning_rate": 2.720125786163522e-06,
"loss": 0.7478,
"step": 1161
},
{
"epoch": 0.45676100628930816,
"grad_norm": 0.7174919843673706,
"learning_rate": 2.718160377358491e-06,
"loss": 0.7053,
"step": 1162
},
{
"epoch": 0.4571540880503145,
"grad_norm": 0.6440824270248413,
"learning_rate": 2.7161949685534593e-06,
"loss": 0.6774,
"step": 1163
},
{
"epoch": 0.45754716981132076,
"grad_norm": 0.7074658274650574,
"learning_rate": 2.714229559748428e-06,
"loss": 0.8195,
"step": 1164
},
{
"epoch": 0.45794025157232704,
"grad_norm": 0.6479617357254028,
"learning_rate": 2.7122641509433965e-06,
"loss": 0.5404,
"step": 1165
},
{
"epoch": 0.4583333333333333,
"grad_norm": 0.5966088771820068,
"learning_rate": 2.710298742138365e-06,
"loss": 0.5773,
"step": 1166
},
{
"epoch": 0.45872641509433965,
"grad_norm": 0.4643804132938385,
"learning_rate": 2.7083333333333334e-06,
"loss": 0.5974,
"step": 1167
},
{
"epoch": 0.4591194968553459,
"grad_norm": 0.9964118599891663,
"learning_rate": 2.706367924528302e-06,
"loss": 0.7279,
"step": 1168
},
{
"epoch": 0.4595125786163522,
"grad_norm": 0.7080931663513184,
"learning_rate": 2.7044025157232706e-06,
"loss": 0.632,
"step": 1169
},
{
"epoch": 0.45990566037735847,
"grad_norm": 0.852836012840271,
"learning_rate": 2.702437106918239e-06,
"loss": 0.7415,
"step": 1170
},
{
"epoch": 0.4602987421383648,
"grad_norm": 1.1155188083648682,
"learning_rate": 2.700471698113208e-06,
"loss": 0.7661,
"step": 1171
},
{
"epoch": 0.4606918238993711,
"grad_norm": 0.7656535506248474,
"learning_rate": 2.6985062893081763e-06,
"loss": 0.6953,
"step": 1172
},
{
"epoch": 0.46108490566037735,
"grad_norm": 0.7052552700042725,
"learning_rate": 2.696540880503145e-06,
"loss": 0.6607,
"step": 1173
},
{
"epoch": 0.46147798742138363,
"grad_norm": 0.7101826667785645,
"learning_rate": 2.694575471698113e-06,
"loss": 0.7377,
"step": 1174
},
{
"epoch": 0.46187106918238996,
"grad_norm": 0.7010018229484558,
"learning_rate": 2.6926100628930816e-06,
"loss": 0.6243,
"step": 1175
},
{
"epoch": 0.46226415094339623,
"grad_norm": 0.7229256629943848,
"learning_rate": 2.6906446540880505e-06,
"loss": 0.8002,
"step": 1176
},
{
"epoch": 0.4626572327044025,
"grad_norm": 1.0095576047897339,
"learning_rate": 2.688679245283019e-06,
"loss": 0.7517,
"step": 1177
},
{
"epoch": 0.4630503144654088,
"grad_norm": 0.6959314942359924,
"learning_rate": 2.6867138364779877e-06,
"loss": 0.7039,
"step": 1178
},
{
"epoch": 0.4634433962264151,
"grad_norm": 0.7328331470489502,
"learning_rate": 2.684748427672956e-06,
"loss": 0.7125,
"step": 1179
},
{
"epoch": 0.4638364779874214,
"grad_norm": 0.6622053980827332,
"learning_rate": 2.682783018867925e-06,
"loss": 0.7208,
"step": 1180
},
{
"epoch": 0.46422955974842767,
"grad_norm": 0.5427559018135071,
"learning_rate": 2.6808176100628934e-06,
"loss": 0.6375,
"step": 1181
},
{
"epoch": 0.46462264150943394,
"grad_norm": 0.5504891276359558,
"learning_rate": 2.6788522012578614e-06,
"loss": 0.6036,
"step": 1182
},
{
"epoch": 0.4650157232704403,
"grad_norm": 0.550934374332428,
"learning_rate": 2.6768867924528303e-06,
"loss": 0.6007,
"step": 1183
},
{
"epoch": 0.46540880503144655,
"grad_norm": 1.369096279144287,
"learning_rate": 2.6749213836477987e-06,
"loss": 0.6921,
"step": 1184
},
{
"epoch": 0.4658018867924528,
"grad_norm": 0.7848041653633118,
"learning_rate": 2.6729559748427675e-06,
"loss": 0.726,
"step": 1185
},
{
"epoch": 0.4661949685534591,
"grad_norm": 0.6354577541351318,
"learning_rate": 2.670990566037736e-06,
"loss": 0.6748,
"step": 1186
},
{
"epoch": 0.46658805031446543,
"grad_norm": 0.6736332774162292,
"learning_rate": 2.669025157232705e-06,
"loss": 0.6592,
"step": 1187
},
{
"epoch": 0.4669811320754717,
"grad_norm": 0.730667233467102,
"learning_rate": 2.6670597484276732e-06,
"loss": 0.7694,
"step": 1188
},
{
"epoch": 0.467374213836478,
"grad_norm": 0.6425909996032715,
"learning_rate": 2.665094339622642e-06,
"loss": 0.688,
"step": 1189
},
{
"epoch": 0.46776729559748426,
"grad_norm": 0.8513888120651245,
"learning_rate": 2.66312893081761e-06,
"loss": 0.6896,
"step": 1190
},
{
"epoch": 0.4681603773584906,
"grad_norm": 0.49088338017463684,
"learning_rate": 2.6611635220125785e-06,
"loss": 0.6787,
"step": 1191
},
{
"epoch": 0.46855345911949686,
"grad_norm": 0.6328742504119873,
"learning_rate": 2.6591981132075473e-06,
"loss": 0.5571,
"step": 1192
},
{
"epoch": 0.46894654088050314,
"grad_norm": 0.7458985447883606,
"learning_rate": 2.6572327044025158e-06,
"loss": 0.7479,
"step": 1193
},
{
"epoch": 0.4693396226415094,
"grad_norm": 0.8140206933021545,
"learning_rate": 2.6552672955974846e-06,
"loss": 0.6879,
"step": 1194
},
{
"epoch": 0.46973270440251574,
"grad_norm": 0.7769689559936523,
"learning_rate": 2.653301886792453e-06,
"loss": 0.8116,
"step": 1195
},
{
"epoch": 0.470125786163522,
"grad_norm": 0.6339801549911499,
"learning_rate": 2.651336477987422e-06,
"loss": 0.6731,
"step": 1196
},
{
"epoch": 0.4705188679245283,
"grad_norm": 0.8083868026733398,
"learning_rate": 2.6493710691823903e-06,
"loss": 0.7727,
"step": 1197
},
{
"epoch": 0.47091194968553457,
"grad_norm": 0.5975171327590942,
"learning_rate": 2.6474056603773583e-06,
"loss": 0.5837,
"step": 1198
},
{
"epoch": 0.4713050314465409,
"grad_norm": 0.653993546962738,
"learning_rate": 2.645440251572327e-06,
"loss": 0.7466,
"step": 1199
},
{
"epoch": 0.4716981132075472,
"grad_norm": 0.8181992769241333,
"learning_rate": 2.6434748427672956e-06,
"loss": 0.7268,
"step": 1200
},
{
"epoch": 0.47209119496855345,
"grad_norm": 0.7353447079658508,
"learning_rate": 2.6415094339622644e-06,
"loss": 0.7286,
"step": 1201
},
{
"epoch": 0.4724842767295597,
"grad_norm": 0.7426215410232544,
"learning_rate": 2.639544025157233e-06,
"loss": 0.615,
"step": 1202
},
{
"epoch": 0.47287735849056606,
"grad_norm": 0.6470339298248291,
"learning_rate": 2.6375786163522017e-06,
"loss": 0.6689,
"step": 1203
},
{
"epoch": 0.47327044025157233,
"grad_norm": 0.5557639598846436,
"learning_rate": 2.63561320754717e-06,
"loss": 0.6682,
"step": 1204
},
{
"epoch": 0.4736635220125786,
"grad_norm": 0.6796771287918091,
"learning_rate": 2.633647798742139e-06,
"loss": 0.5612,
"step": 1205
},
{
"epoch": 0.4740566037735849,
"grad_norm": 0.7698836922645569,
"learning_rate": 2.631682389937107e-06,
"loss": 0.7072,
"step": 1206
},
{
"epoch": 0.4744496855345912,
"grad_norm": 0.5960604548454285,
"learning_rate": 2.6297169811320754e-06,
"loss": 0.6947,
"step": 1207
},
{
"epoch": 0.4748427672955975,
"grad_norm": 0.8375264406204224,
"learning_rate": 2.6277515723270442e-06,
"loss": 0.6709,
"step": 1208
},
{
"epoch": 0.47523584905660377,
"grad_norm": 0.8032341599464417,
"learning_rate": 2.6257861635220127e-06,
"loss": 0.7742,
"step": 1209
},
{
"epoch": 0.47562893081761004,
"grad_norm": 0.651500940322876,
"learning_rate": 2.6238207547169815e-06,
"loss": 0.6779,
"step": 1210
},
{
"epoch": 0.47602201257861637,
"grad_norm": 0.5570948123931885,
"learning_rate": 2.62185534591195e-06,
"loss": 0.5874,
"step": 1211
},
{
"epoch": 0.47641509433962265,
"grad_norm": 0.6360973715782166,
"learning_rate": 2.6198899371069188e-06,
"loss": 0.6548,
"step": 1212
},
{
"epoch": 0.4768081761006289,
"grad_norm": 0.6472859978675842,
"learning_rate": 2.617924528301887e-06,
"loss": 0.6233,
"step": 1213
},
{
"epoch": 0.4772012578616352,
"grad_norm": 0.7749592065811157,
"learning_rate": 2.615959119496855e-06,
"loss": 0.6699,
"step": 1214
},
{
"epoch": 0.47759433962264153,
"grad_norm": 0.7135629653930664,
"learning_rate": 2.613993710691824e-06,
"loss": 0.6771,
"step": 1215
},
{
"epoch": 0.4779874213836478,
"grad_norm": 0.9305992722511292,
"learning_rate": 2.6120283018867925e-06,
"loss": 0.7834,
"step": 1216
},
{
"epoch": 0.4783805031446541,
"grad_norm": 0.7440532445907593,
"learning_rate": 2.6100628930817613e-06,
"loss": 0.7296,
"step": 1217
},
{
"epoch": 0.47877358490566035,
"grad_norm": 0.5722053050994873,
"learning_rate": 2.6080974842767297e-06,
"loss": 0.6113,
"step": 1218
},
{
"epoch": 0.4791666666666667,
"grad_norm": 0.7441682815551758,
"learning_rate": 2.6061320754716986e-06,
"loss": 0.7816,
"step": 1219
},
{
"epoch": 0.47955974842767296,
"grad_norm": 0.6952428221702576,
"learning_rate": 2.604166666666667e-06,
"loss": 0.7452,
"step": 1220
},
{
"epoch": 0.47995283018867924,
"grad_norm": 0.5781884789466858,
"learning_rate": 2.602201257861636e-06,
"loss": 0.6234,
"step": 1221
},
{
"epoch": 0.4803459119496855,
"grad_norm": 0.5742769241333008,
"learning_rate": 2.600235849056604e-06,
"loss": 0.6458,
"step": 1222
},
{
"epoch": 0.48073899371069184,
"grad_norm": 0.6358216404914856,
"learning_rate": 2.5982704402515723e-06,
"loss": 0.7413,
"step": 1223
},
{
"epoch": 0.4811320754716981,
"grad_norm": 0.737565815448761,
"learning_rate": 2.596305031446541e-06,
"loss": 0.7362,
"step": 1224
},
{
"epoch": 0.4815251572327044,
"grad_norm": 0.7103287577629089,
"learning_rate": 2.5943396226415095e-06,
"loss": 0.7034,
"step": 1225
},
{
"epoch": 0.48191823899371067,
"grad_norm": 0.7302315831184387,
"learning_rate": 2.5923742138364784e-06,
"loss": 0.6661,
"step": 1226
},
{
"epoch": 0.482311320754717,
"grad_norm": 0.7760244607925415,
"learning_rate": 2.590408805031447e-06,
"loss": 0.71,
"step": 1227
},
{
"epoch": 0.4827044025157233,
"grad_norm": 0.681980311870575,
"learning_rate": 2.5884433962264157e-06,
"loss": 0.6689,
"step": 1228
},
{
"epoch": 0.48309748427672955,
"grad_norm": 0.5686315298080444,
"learning_rate": 2.586477987421384e-06,
"loss": 0.5814,
"step": 1229
},
{
"epoch": 0.4834905660377358,
"grad_norm": 0.7008070945739746,
"learning_rate": 2.584512578616352e-06,
"loss": 0.7371,
"step": 1230
},
{
"epoch": 0.48388364779874216,
"grad_norm": 0.5374528765678406,
"learning_rate": 2.582547169811321e-06,
"loss": 0.6654,
"step": 1231
},
{
"epoch": 0.48427672955974843,
"grad_norm": 0.7389048337936401,
"learning_rate": 2.5805817610062894e-06,
"loss": 0.6388,
"step": 1232
},
{
"epoch": 0.4846698113207547,
"grad_norm": 0.7383620738983154,
"learning_rate": 2.578616352201258e-06,
"loss": 0.6442,
"step": 1233
},
{
"epoch": 0.485062893081761,
"grad_norm": 0.7014018893241882,
"learning_rate": 2.5766509433962266e-06,
"loss": 0.7557,
"step": 1234
},
{
"epoch": 0.4854559748427673,
"grad_norm": 0.572113573551178,
"learning_rate": 2.5746855345911955e-06,
"loss": 0.6471,
"step": 1235
},
{
"epoch": 0.4858490566037736,
"grad_norm": 0.817406952381134,
"learning_rate": 2.572720125786164e-06,
"loss": 0.7496,
"step": 1236
},
{
"epoch": 0.48624213836477986,
"grad_norm": 0.7988145351409912,
"learning_rate": 2.5707547169811327e-06,
"loss": 0.6985,
"step": 1237
},
{
"epoch": 0.48663522012578614,
"grad_norm": 1.037604808807373,
"learning_rate": 2.5687893081761007e-06,
"loss": 0.658,
"step": 1238
},
{
"epoch": 0.48702830188679247,
"grad_norm": 0.7785852551460266,
"learning_rate": 2.566823899371069e-06,
"loss": 0.6326,
"step": 1239
},
{
"epoch": 0.48742138364779874,
"grad_norm": 0.5972993969917297,
"learning_rate": 2.564858490566038e-06,
"loss": 0.7045,
"step": 1240
},
{
"epoch": 0.487814465408805,
"grad_norm": 0.6306884288787842,
"learning_rate": 2.5628930817610064e-06,
"loss": 0.6497,
"step": 1241
},
{
"epoch": 0.4882075471698113,
"grad_norm": 0.6400014162063599,
"learning_rate": 2.5609276729559753e-06,
"loss": 0.6644,
"step": 1242
},
{
"epoch": 0.4886006289308176,
"grad_norm": 0.6530547142028809,
"learning_rate": 2.5589622641509437e-06,
"loss": 0.6021,
"step": 1243
},
{
"epoch": 0.4889937106918239,
"grad_norm": 0.6254391074180603,
"learning_rate": 2.5569968553459125e-06,
"loss": 0.566,
"step": 1244
},
{
"epoch": 0.4893867924528302,
"grad_norm": 0.8972374200820923,
"learning_rate": 2.555031446540881e-06,
"loss": 0.7834,
"step": 1245
},
{
"epoch": 0.48977987421383645,
"grad_norm": 0.6880585551261902,
"learning_rate": 2.553066037735849e-06,
"loss": 0.6778,
"step": 1246
},
{
"epoch": 0.4901729559748428,
"grad_norm": 0.6857219338417053,
"learning_rate": 2.551100628930818e-06,
"loss": 0.6539,
"step": 1247
},
{
"epoch": 0.49056603773584906,
"grad_norm": 0.8150197863578796,
"learning_rate": 2.5491352201257862e-06,
"loss": 0.6957,
"step": 1248
},
{
"epoch": 0.49095911949685533,
"grad_norm": 0.7534979581832886,
"learning_rate": 2.547169811320755e-06,
"loss": 0.7511,
"step": 1249
},
{
"epoch": 0.4913522012578616,
"grad_norm": 0.8706692457199097,
"learning_rate": 2.5452044025157235e-06,
"loss": 0.731,
"step": 1250
},
{
"epoch": 0.49174528301886794,
"grad_norm": 0.7131131291389465,
"learning_rate": 2.5432389937106924e-06,
"loss": 0.7211,
"step": 1251
},
{
"epoch": 0.4921383647798742,
"grad_norm": 0.5392166972160339,
"learning_rate": 2.5412735849056608e-06,
"loss": 0.6422,
"step": 1252
},
{
"epoch": 0.4925314465408805,
"grad_norm": 0.742170512676239,
"learning_rate": 2.539308176100629e-06,
"loss": 0.6779,
"step": 1253
},
{
"epoch": 0.49292452830188677,
"grad_norm": 0.6728838682174683,
"learning_rate": 2.5373427672955976e-06,
"loss": 0.6447,
"step": 1254
},
{
"epoch": 0.4933176100628931,
"grad_norm": 0.8849697113037109,
"learning_rate": 2.535377358490566e-06,
"loss": 0.7495,
"step": 1255
},
{
"epoch": 0.4937106918238994,
"grad_norm": 0.6279146671295166,
"learning_rate": 2.533411949685535e-06,
"loss": 0.6699,
"step": 1256
},
{
"epoch": 0.49410377358490565,
"grad_norm": 0.6752800941467285,
"learning_rate": 2.5314465408805033e-06,
"loss": 0.7338,
"step": 1257
},
{
"epoch": 0.4944968553459119,
"grad_norm": 0.4384029507637024,
"learning_rate": 2.529481132075472e-06,
"loss": 0.5315,
"step": 1258
},
{
"epoch": 0.49488993710691825,
"grad_norm": 0.6677165031433105,
"learning_rate": 2.5275157232704406e-06,
"loss": 0.6005,
"step": 1259
},
{
"epoch": 0.49528301886792453,
"grad_norm": 0.6500537395477295,
"learning_rate": 2.525550314465409e-06,
"loss": 0.6748,
"step": 1260
},
{
"epoch": 0.4956761006289308,
"grad_norm": 0.6206688284873962,
"learning_rate": 2.523584905660378e-06,
"loss": 0.5696,
"step": 1261
},
{
"epoch": 0.4960691823899371,
"grad_norm": 0.6357614994049072,
"learning_rate": 2.521619496855346e-06,
"loss": 0.6926,
"step": 1262
},
{
"epoch": 0.4964622641509434,
"grad_norm": 0.7353872060775757,
"learning_rate": 2.5196540880503147e-06,
"loss": 0.7206,
"step": 1263
},
{
"epoch": 0.4968553459119497,
"grad_norm": 0.5979838967323303,
"learning_rate": 2.517688679245283e-06,
"loss": 0.6479,
"step": 1264
},
{
"epoch": 0.49724842767295596,
"grad_norm": 0.6727257370948792,
"learning_rate": 2.515723270440252e-06,
"loss": 0.8071,
"step": 1265
},
{
"epoch": 0.49764150943396224,
"grad_norm": 0.5943746566772461,
"learning_rate": 2.5137578616352204e-06,
"loss": 0.6149,
"step": 1266
},
{
"epoch": 0.49803459119496857,
"grad_norm": 0.7032964825630188,
"learning_rate": 2.511792452830189e-06,
"loss": 0.741,
"step": 1267
},
{
"epoch": 0.49842767295597484,
"grad_norm": 0.9545643925666809,
"learning_rate": 2.5098270440251577e-06,
"loss": 0.7462,
"step": 1268
},
{
"epoch": 0.4988207547169811,
"grad_norm": 0.6547558903694153,
"learning_rate": 2.507861635220126e-06,
"loss": 0.6382,
"step": 1269
},
{
"epoch": 0.4992138364779874,
"grad_norm": 0.7001852989196777,
"learning_rate": 2.5058962264150945e-06,
"loss": 0.7457,
"step": 1270
},
{
"epoch": 0.4996069182389937,
"grad_norm": 0.7509366273880005,
"learning_rate": 2.503930817610063e-06,
"loss": 0.6197,
"step": 1271
},
{
"epoch": 0.5,
"grad_norm": 0.7888211607933044,
"learning_rate": 2.5019654088050318e-06,
"loss": 0.5822,
"step": 1272
},
{
"epoch": 0.5003930817610063,
"grad_norm": 0.9375798106193542,
"learning_rate": 2.5e-06,
"loss": 0.8303,
"step": 1273
},
{
"epoch": 0.5007861635220126,
"grad_norm": 0.7427085638046265,
"learning_rate": 2.4980345911949686e-06,
"loss": 0.7339,
"step": 1274
},
{
"epoch": 0.5011792452830188,
"grad_norm": 0.9575438499450684,
"learning_rate": 2.496069182389937e-06,
"loss": 0.8694,
"step": 1275
},
{
"epoch": 0.5015723270440252,
"grad_norm": 0.6350022554397583,
"learning_rate": 2.494103773584906e-06,
"loss": 0.663,
"step": 1276
},
{
"epoch": 0.5019654088050315,
"grad_norm": 0.7169939279556274,
"learning_rate": 2.4921383647798743e-06,
"loss": 0.6301,
"step": 1277
},
{
"epoch": 0.5023584905660378,
"grad_norm": 0.9066876173019409,
"learning_rate": 2.490172955974843e-06,
"loss": 0.7394,
"step": 1278
},
{
"epoch": 0.502751572327044,
"grad_norm": 0.6524776220321655,
"learning_rate": 2.4882075471698116e-06,
"loss": 0.7454,
"step": 1279
},
{
"epoch": 0.5031446540880503,
"grad_norm": 0.5753188133239746,
"learning_rate": 2.48624213836478e-06,
"loss": 0.588,
"step": 1280
},
{
"epoch": 0.5035377358490566,
"grad_norm": 0.5630155205726624,
"learning_rate": 2.4842767295597484e-06,
"loss": 0.6663,
"step": 1281
},
{
"epoch": 0.5039308176100629,
"grad_norm": 0.7138348817825317,
"learning_rate": 2.4823113207547173e-06,
"loss": 0.7751,
"step": 1282
},
{
"epoch": 0.5043238993710691,
"grad_norm": 0.7923138737678528,
"learning_rate": 2.4803459119496857e-06,
"loss": 0.8231,
"step": 1283
},
{
"epoch": 0.5047169811320755,
"grad_norm": 0.7163272500038147,
"learning_rate": 2.478380503144654e-06,
"loss": 0.7935,
"step": 1284
},
{
"epoch": 0.5051100628930818,
"grad_norm": 0.7271803617477417,
"learning_rate": 2.476415094339623e-06,
"loss": 0.6356,
"step": 1285
},
{
"epoch": 0.5055031446540881,
"grad_norm": 0.6828988790512085,
"learning_rate": 2.4744496855345914e-06,
"loss": 0.6833,
"step": 1286
},
{
"epoch": 0.5058962264150944,
"grad_norm": 0.7776426672935486,
"learning_rate": 2.47248427672956e-06,
"loss": 0.7423,
"step": 1287
},
{
"epoch": 0.5062893081761006,
"grad_norm": 0.8237703442573547,
"learning_rate": 2.4705188679245282e-06,
"loss": 0.7369,
"step": 1288
},
{
"epoch": 0.5066823899371069,
"grad_norm": 0.7904167175292969,
"learning_rate": 2.468553459119497e-06,
"loss": 0.6816,
"step": 1289
},
{
"epoch": 0.5070754716981132,
"grad_norm": 0.7876915335655212,
"learning_rate": 2.4665880503144655e-06,
"loss": 0.7291,
"step": 1290
},
{
"epoch": 0.5074685534591195,
"grad_norm": 0.8483654260635376,
"learning_rate": 2.464622641509434e-06,
"loss": 0.7603,
"step": 1291
},
{
"epoch": 0.5078616352201258,
"grad_norm": 0.6100538372993469,
"learning_rate": 2.4626572327044028e-06,
"loss": 0.6801,
"step": 1292
},
{
"epoch": 0.5082547169811321,
"grad_norm": 0.8378139138221741,
"learning_rate": 2.460691823899371e-06,
"loss": 0.7773,
"step": 1293
},
{
"epoch": 0.5086477987421384,
"grad_norm": 0.7030474543571472,
"learning_rate": 2.45872641509434e-06,
"loss": 0.7336,
"step": 1294
},
{
"epoch": 0.5090408805031447,
"grad_norm": 0.7941507697105408,
"learning_rate": 2.456761006289308e-06,
"loss": 0.6876,
"step": 1295
},
{
"epoch": 0.5094339622641509,
"grad_norm": 0.7438374757766724,
"learning_rate": 2.454795597484277e-06,
"loss": 0.7219,
"step": 1296
},
{
"epoch": 0.5098270440251572,
"grad_norm": 0.7501423358917236,
"learning_rate": 2.4528301886792453e-06,
"loss": 0.8068,
"step": 1297
},
{
"epoch": 0.5102201257861635,
"grad_norm": 0.7288534045219421,
"learning_rate": 2.450864779874214e-06,
"loss": 0.7695,
"step": 1298
},
{
"epoch": 0.5106132075471698,
"grad_norm": 0.6144891977310181,
"learning_rate": 2.4488993710691826e-06,
"loss": 0.7316,
"step": 1299
},
{
"epoch": 0.5110062893081762,
"grad_norm": 0.5642289519309998,
"learning_rate": 2.446933962264151e-06,
"loss": 0.6772,
"step": 1300
},
{
"epoch": 0.5113993710691824,
"grad_norm": 0.5538402795791626,
"learning_rate": 2.44496855345912e-06,
"loss": 0.6071,
"step": 1301
},
{
"epoch": 0.5117924528301887,
"grad_norm": 0.6629433631896973,
"learning_rate": 2.4430031446540883e-06,
"loss": 0.7155,
"step": 1302
},
{
"epoch": 0.512185534591195,
"grad_norm": 0.6620303988456726,
"learning_rate": 2.4410377358490567e-06,
"loss": 0.656,
"step": 1303
},
{
"epoch": 0.5125786163522013,
"grad_norm": 0.8463244438171387,
"learning_rate": 2.439072327044025e-06,
"loss": 0.7947,
"step": 1304
},
{
"epoch": 0.5129716981132075,
"grad_norm": 0.8870863914489746,
"learning_rate": 2.437106918238994e-06,
"loss": 0.7701,
"step": 1305
},
{
"epoch": 0.5133647798742138,
"grad_norm": 0.5720570683479309,
"learning_rate": 2.4351415094339624e-06,
"loss": 0.611,
"step": 1306
},
{
"epoch": 0.5137578616352201,
"grad_norm": 0.6084403991699219,
"learning_rate": 2.433176100628931e-06,
"loss": 0.6312,
"step": 1307
},
{
"epoch": 0.5141509433962265,
"grad_norm": 0.6791754364967346,
"learning_rate": 2.4312106918238997e-06,
"loss": 0.733,
"step": 1308
},
{
"epoch": 0.5145440251572327,
"grad_norm": 0.521419107913971,
"learning_rate": 2.429245283018868e-06,
"loss": 0.6313,
"step": 1309
},
{
"epoch": 0.514937106918239,
"grad_norm": 1.3247402906417847,
"learning_rate": 2.427279874213837e-06,
"loss": 0.736,
"step": 1310
},
{
"epoch": 0.5153301886792453,
"grad_norm": 0.7860870957374573,
"learning_rate": 2.425314465408805e-06,
"loss": 0.7577,
"step": 1311
},
{
"epoch": 0.5157232704402516,
"grad_norm": 0.7706316709518433,
"learning_rate": 2.4233490566037738e-06,
"loss": 0.7456,
"step": 1312
},
{
"epoch": 0.5161163522012578,
"grad_norm": 0.7867957949638367,
"learning_rate": 2.421383647798742e-06,
"loss": 0.7224,
"step": 1313
},
{
"epoch": 0.5165094339622641,
"grad_norm": 0.42827004194259644,
"learning_rate": 2.419418238993711e-06,
"loss": 0.4283,
"step": 1314
},
{
"epoch": 0.5169025157232704,
"grad_norm": 0.6773838400840759,
"learning_rate": 2.4174528301886795e-06,
"loss": 0.7222,
"step": 1315
},
{
"epoch": 0.5172955974842768,
"grad_norm": 0.5185518860816956,
"learning_rate": 2.415487421383648e-06,
"loss": 0.6589,
"step": 1316
},
{
"epoch": 0.5176886792452831,
"grad_norm": 0.601759135723114,
"learning_rate": 2.4135220125786168e-06,
"loss": 0.7045,
"step": 1317
},
{
"epoch": 0.5180817610062893,
"grad_norm": 0.7216753363609314,
"learning_rate": 2.411556603773585e-06,
"loss": 0.7208,
"step": 1318
},
{
"epoch": 0.5184748427672956,
"grad_norm": 0.7023240923881531,
"learning_rate": 2.4095911949685536e-06,
"loss": 0.6846,
"step": 1319
},
{
"epoch": 0.5188679245283019,
"grad_norm": 0.7611101269721985,
"learning_rate": 2.407625786163522e-06,
"loss": 0.7225,
"step": 1320
},
{
"epoch": 0.5192610062893082,
"grad_norm": 0.7871542572975159,
"learning_rate": 2.405660377358491e-06,
"loss": 0.7974,
"step": 1321
},
{
"epoch": 0.5196540880503144,
"grad_norm": 1.0317476987838745,
"learning_rate": 2.4036949685534593e-06,
"loss": 0.7682,
"step": 1322
},
{
"epoch": 0.5200471698113207,
"grad_norm": 0.6276971101760864,
"learning_rate": 2.4017295597484277e-06,
"loss": 0.6334,
"step": 1323
},
{
"epoch": 0.5204402515723271,
"grad_norm": 0.8281777501106262,
"learning_rate": 2.3997641509433966e-06,
"loss": 0.7411,
"step": 1324
},
{
"epoch": 0.5208333333333334,
"grad_norm": 0.6224610209465027,
"learning_rate": 2.397798742138365e-06,
"loss": 0.6522,
"step": 1325
},
{
"epoch": 0.5212264150943396,
"grad_norm": 0.6447500586509705,
"learning_rate": 2.395833333333334e-06,
"loss": 0.6169,
"step": 1326
},
{
"epoch": 0.5216194968553459,
"grad_norm": 0.7196338176727295,
"learning_rate": 2.393867924528302e-06,
"loss": 0.6606,
"step": 1327
},
{
"epoch": 0.5220125786163522,
"grad_norm": 0.8055816292762756,
"learning_rate": 2.3919025157232707e-06,
"loss": 0.765,
"step": 1328
},
{
"epoch": 0.5224056603773585,
"grad_norm": 0.5690597891807556,
"learning_rate": 2.389937106918239e-06,
"loss": 0.5839,
"step": 1329
},
{
"epoch": 0.5227987421383647,
"grad_norm": 0.858963668346405,
"learning_rate": 2.387971698113208e-06,
"loss": 0.8007,
"step": 1330
},
{
"epoch": 0.523191823899371,
"grad_norm": 0.7087966799736023,
"learning_rate": 2.3860062893081764e-06,
"loss": 0.6576,
"step": 1331
},
{
"epoch": 0.5235849056603774,
"grad_norm": 0.7247787117958069,
"learning_rate": 2.384040880503145e-06,
"loss": 0.6832,
"step": 1332
},
{
"epoch": 0.5239779874213837,
"grad_norm": 0.6216225624084473,
"learning_rate": 2.3820754716981136e-06,
"loss": 0.6929,
"step": 1333
},
{
"epoch": 0.52437106918239,
"grad_norm": 0.8185009360313416,
"learning_rate": 2.380110062893082e-06,
"loss": 0.6426,
"step": 1334
},
{
"epoch": 0.5247641509433962,
"grad_norm": 0.7024775147438049,
"learning_rate": 2.3781446540880505e-06,
"loss": 0.6552,
"step": 1335
},
{
"epoch": 0.5251572327044025,
"grad_norm": 0.6155581474304199,
"learning_rate": 2.376179245283019e-06,
"loss": 0.6737,
"step": 1336
},
{
"epoch": 0.5255503144654088,
"grad_norm": 0.8044111728668213,
"learning_rate": 2.3742138364779878e-06,
"loss": 0.7836,
"step": 1337
},
{
"epoch": 0.5259433962264151,
"grad_norm": 0.6128347516059875,
"learning_rate": 2.372248427672956e-06,
"loss": 0.6327,
"step": 1338
},
{
"epoch": 0.5263364779874213,
"grad_norm": 0.7312962412834167,
"learning_rate": 2.3702830188679246e-06,
"loss": 0.7689,
"step": 1339
},
{
"epoch": 0.5267295597484277,
"grad_norm": 0.8646583557128906,
"learning_rate": 2.3683176100628934e-06,
"loss": 0.7858,
"step": 1340
},
{
"epoch": 0.527122641509434,
"grad_norm": 0.8398555517196655,
"learning_rate": 2.366352201257862e-06,
"loss": 0.7816,
"step": 1341
},
{
"epoch": 0.5275157232704403,
"grad_norm": 0.9116210341453552,
"learning_rate": 2.3643867924528303e-06,
"loss": 0.7743,
"step": 1342
},
{
"epoch": 0.5279088050314465,
"grad_norm": 0.7479056715965271,
"learning_rate": 2.3624213836477987e-06,
"loss": 0.7227,
"step": 1343
},
{
"epoch": 0.5283018867924528,
"grad_norm": 0.7436212301254272,
"learning_rate": 2.3604559748427676e-06,
"loss": 0.648,
"step": 1344
},
{
"epoch": 0.5286949685534591,
"grad_norm": 0.8771992325782776,
"learning_rate": 2.358490566037736e-06,
"loss": 0.7361,
"step": 1345
},
{
"epoch": 0.5290880503144654,
"grad_norm": 0.5958253145217896,
"learning_rate": 2.356525157232705e-06,
"loss": 0.6558,
"step": 1346
},
{
"epoch": 0.5294811320754716,
"grad_norm": 0.6558207869529724,
"learning_rate": 2.354559748427673e-06,
"loss": 0.6139,
"step": 1347
},
{
"epoch": 0.529874213836478,
"grad_norm": 0.5549044609069824,
"learning_rate": 2.3525943396226417e-06,
"loss": 0.5447,
"step": 1348
},
{
"epoch": 0.5302672955974843,
"grad_norm": 0.8148266077041626,
"learning_rate": 2.35062893081761e-06,
"loss": 0.6361,
"step": 1349
},
{
"epoch": 0.5306603773584906,
"grad_norm": 0.642961323261261,
"learning_rate": 2.348663522012579e-06,
"loss": 0.6517,
"step": 1350
},
{
"epoch": 0.5310534591194969,
"grad_norm": 0.648349940776825,
"learning_rate": 2.3466981132075474e-06,
"loss": 0.686,
"step": 1351
},
{
"epoch": 0.5314465408805031,
"grad_norm": 0.8965466618537903,
"learning_rate": 2.344732704402516e-06,
"loss": 0.8136,
"step": 1352
},
{
"epoch": 0.5318396226415094,
"grad_norm": 0.7063581943511963,
"learning_rate": 2.3427672955974846e-06,
"loss": 0.7191,
"step": 1353
},
{
"epoch": 0.5322327044025157,
"grad_norm": 0.5758686661720276,
"learning_rate": 2.340801886792453e-06,
"loss": 0.5933,
"step": 1354
},
{
"epoch": 0.532625786163522,
"grad_norm": 0.6889188289642334,
"learning_rate": 2.3388364779874215e-06,
"loss": 0.6888,
"step": 1355
},
{
"epoch": 0.5330188679245284,
"grad_norm": 0.6893822550773621,
"learning_rate": 2.33687106918239e-06,
"loss": 0.6958,
"step": 1356
},
{
"epoch": 0.5334119496855346,
"grad_norm": 0.7186471223831177,
"learning_rate": 2.3349056603773588e-06,
"loss": 0.6508,
"step": 1357
},
{
"epoch": 0.5338050314465409,
"grad_norm": 0.7326532006263733,
"learning_rate": 2.332940251572327e-06,
"loss": 0.6334,
"step": 1358
},
{
"epoch": 0.5341981132075472,
"grad_norm": 0.7797999978065491,
"learning_rate": 2.3309748427672956e-06,
"loss": 0.7347,
"step": 1359
},
{
"epoch": 0.5345911949685535,
"grad_norm": 0.6638525724411011,
"learning_rate": 2.3290094339622645e-06,
"loss": 0.7007,
"step": 1360
},
{
"epoch": 0.5349842767295597,
"grad_norm": 0.6540643572807312,
"learning_rate": 2.327044025157233e-06,
"loss": 0.6356,
"step": 1361
},
{
"epoch": 0.535377358490566,
"grad_norm": 0.5673895478248596,
"learning_rate": 2.3250786163522017e-06,
"loss": 0.6409,
"step": 1362
},
{
"epoch": 0.5357704402515723,
"grad_norm": 0.6130591630935669,
"learning_rate": 2.3231132075471697e-06,
"loss": 0.6848,
"step": 1363
},
{
"epoch": 0.5361635220125787,
"grad_norm": 0.6859002113342285,
"learning_rate": 2.3211477987421386e-06,
"loss": 0.6807,
"step": 1364
},
{
"epoch": 0.5365566037735849,
"grad_norm": 0.5453311800956726,
"learning_rate": 2.319182389937107e-06,
"loss": 0.6775,
"step": 1365
},
{
"epoch": 0.5369496855345912,
"grad_norm": 0.8470795750617981,
"learning_rate": 2.317216981132076e-06,
"loss": 0.7482,
"step": 1366
},
{
"epoch": 0.5373427672955975,
"grad_norm": 0.770240306854248,
"learning_rate": 2.3152515723270443e-06,
"loss": 0.7106,
"step": 1367
},
{
"epoch": 0.5377358490566038,
"grad_norm": 0.7521366477012634,
"learning_rate": 2.3132861635220127e-06,
"loss": 0.7469,
"step": 1368
},
{
"epoch": 0.53812893081761,
"grad_norm": 0.378184974193573,
"learning_rate": 2.3113207547169815e-06,
"loss": 0.4702,
"step": 1369
},
{
"epoch": 0.5385220125786163,
"grad_norm": 0.7602344751358032,
"learning_rate": 2.30935534591195e-06,
"loss": 0.7285,
"step": 1370
},
{
"epoch": 0.5389150943396226,
"grad_norm": 0.5888705849647522,
"learning_rate": 2.3073899371069184e-06,
"loss": 0.6707,
"step": 1371
},
{
"epoch": 0.539308176100629,
"grad_norm": 0.7087979912757874,
"learning_rate": 2.305424528301887e-06,
"loss": 0.7017,
"step": 1372
},
{
"epoch": 0.5397012578616353,
"grad_norm": 0.5701378583908081,
"learning_rate": 2.3034591194968556e-06,
"loss": 0.6899,
"step": 1373
},
{
"epoch": 0.5400943396226415,
"grad_norm": 0.7287231087684631,
"learning_rate": 2.301493710691824e-06,
"loss": 0.6849,
"step": 1374
},
{
"epoch": 0.5404874213836478,
"grad_norm": 0.5905083417892456,
"learning_rate": 2.2995283018867925e-06,
"loss": 0.6571,
"step": 1375
},
{
"epoch": 0.5408805031446541,
"grad_norm": 0.7938220500946045,
"learning_rate": 2.2975628930817613e-06,
"loss": 0.7733,
"step": 1376
},
{
"epoch": 0.5412735849056604,
"grad_norm": 0.839634358882904,
"learning_rate": 2.2955974842767298e-06,
"loss": 0.7902,
"step": 1377
},
{
"epoch": 0.5416666666666666,
"grad_norm": 0.8788559436798096,
"learning_rate": 2.2936320754716986e-06,
"loss": 0.7145,
"step": 1378
},
{
"epoch": 0.5420597484276729,
"grad_norm": 0.6192091107368469,
"learning_rate": 2.2916666666666666e-06,
"loss": 0.6212,
"step": 1379
},
{
"epoch": 0.5424528301886793,
"grad_norm": 0.7988916039466858,
"learning_rate": 2.2897012578616355e-06,
"loss": 0.7482,
"step": 1380
},
{
"epoch": 0.5428459119496856,
"grad_norm": 0.6450135111808777,
"learning_rate": 2.287735849056604e-06,
"loss": 0.7235,
"step": 1381
},
{
"epoch": 0.5432389937106918,
"grad_norm": 0.5630086660385132,
"learning_rate": 2.2857704402515727e-06,
"loss": 0.6807,
"step": 1382
},
{
"epoch": 0.5436320754716981,
"grad_norm": 0.78071528673172,
"learning_rate": 2.283805031446541e-06,
"loss": 0.6887,
"step": 1383
},
{
"epoch": 0.5440251572327044,
"grad_norm": 0.6387832760810852,
"learning_rate": 2.2818396226415096e-06,
"loss": 0.7236,
"step": 1384
},
{
"epoch": 0.5444182389937107,
"grad_norm": 0.7810574173927307,
"learning_rate": 2.2798742138364784e-06,
"loss": 0.7566,
"step": 1385
},
{
"epoch": 0.5448113207547169,
"grad_norm": 0.6254619359970093,
"learning_rate": 2.277908805031447e-06,
"loss": 0.6707,
"step": 1386
},
{
"epoch": 0.5452044025157232,
"grad_norm": 0.9031320214271545,
"learning_rate": 2.2759433962264153e-06,
"loss": 0.8147,
"step": 1387
},
{
"epoch": 0.5455974842767296,
"grad_norm": 0.8739271759986877,
"learning_rate": 2.2739779874213837e-06,
"loss": 0.8782,
"step": 1388
},
{
"epoch": 0.5459905660377359,
"grad_norm": 0.8503692150115967,
"learning_rate": 2.2720125786163525e-06,
"loss": 0.7712,
"step": 1389
},
{
"epoch": 0.5463836477987422,
"grad_norm": 0.6300114989280701,
"learning_rate": 2.270047169811321e-06,
"loss": 0.664,
"step": 1390
},
{
"epoch": 0.5467767295597484,
"grad_norm": 0.7298681735992432,
"learning_rate": 2.2680817610062894e-06,
"loss": 0.7041,
"step": 1391
},
{
"epoch": 0.5471698113207547,
"grad_norm": 0.7003951668739319,
"learning_rate": 2.2661163522012582e-06,
"loss": 0.6233,
"step": 1392
},
{
"epoch": 0.547562893081761,
"grad_norm": 0.7125574350357056,
"learning_rate": 2.2641509433962266e-06,
"loss": 0.7581,
"step": 1393
},
{
"epoch": 0.5479559748427673,
"grad_norm": 0.637005090713501,
"learning_rate": 2.262185534591195e-06,
"loss": 0.6583,
"step": 1394
},
{
"epoch": 0.5483490566037735,
"grad_norm": 0.6715347170829773,
"learning_rate": 2.2602201257861635e-06,
"loss": 0.6761,
"step": 1395
},
{
"epoch": 0.5487421383647799,
"grad_norm": 0.6979790329933167,
"learning_rate": 2.2582547169811323e-06,
"loss": 0.6444,
"step": 1396
},
{
"epoch": 0.5491352201257862,
"grad_norm": 0.6810550689697266,
"learning_rate": 2.2562893081761008e-06,
"loss": 0.6823,
"step": 1397
},
{
"epoch": 0.5495283018867925,
"grad_norm": 0.5503625273704529,
"learning_rate": 2.2543238993710696e-06,
"loss": 0.6193,
"step": 1398
},
{
"epoch": 0.5499213836477987,
"grad_norm": 0.7882553339004517,
"learning_rate": 2.252358490566038e-06,
"loss": 0.7267,
"step": 1399
},
{
"epoch": 0.550314465408805,
"grad_norm": 0.7033482193946838,
"learning_rate": 2.2503930817610065e-06,
"loss": 0.6412,
"step": 1400
},
{
"epoch": 0.5507075471698113,
"grad_norm": 0.710369348526001,
"learning_rate": 2.248427672955975e-06,
"loss": 0.7067,
"step": 1401
},
{
"epoch": 0.5511006289308176,
"grad_norm": 0.8353728652000427,
"learning_rate": 2.2464622641509437e-06,
"loss": 0.7347,
"step": 1402
},
{
"epoch": 0.5514937106918238,
"grad_norm": 0.6869653463363647,
"learning_rate": 2.244496855345912e-06,
"loss": 0.7301,
"step": 1403
},
{
"epoch": 0.5518867924528302,
"grad_norm": 0.776421070098877,
"learning_rate": 2.2425314465408806e-06,
"loss": 0.7137,
"step": 1404
},
{
"epoch": 0.5522798742138365,
"grad_norm": 0.705815851688385,
"learning_rate": 2.2405660377358494e-06,
"loss": 0.776,
"step": 1405
},
{
"epoch": 0.5526729559748428,
"grad_norm": 0.903978705406189,
"learning_rate": 2.238600628930818e-06,
"loss": 0.7144,
"step": 1406
},
{
"epoch": 0.5530660377358491,
"grad_norm": 0.5884318947792053,
"learning_rate": 2.2366352201257863e-06,
"loss": 0.6173,
"step": 1407
},
{
"epoch": 0.5534591194968553,
"grad_norm": 0.6925857067108154,
"learning_rate": 2.2346698113207547e-06,
"loss": 0.66,
"step": 1408
},
{
"epoch": 0.5538522012578616,
"grad_norm": 0.7048287987709045,
"learning_rate": 2.2327044025157235e-06,
"loss": 0.6931,
"step": 1409
},
{
"epoch": 0.5542452830188679,
"grad_norm": 0.7373592257499695,
"learning_rate": 2.230738993710692e-06,
"loss": 0.62,
"step": 1410
},
{
"epoch": 0.5546383647798742,
"grad_norm": 0.842210590839386,
"learning_rate": 2.2287735849056604e-06,
"loss": 0.7606,
"step": 1411
},
{
"epoch": 0.5550314465408805,
"grad_norm": 0.5966304540634155,
"learning_rate": 2.2268081761006292e-06,
"loss": 0.7155,
"step": 1412
},
{
"epoch": 0.5554245283018868,
"grad_norm": 0.6816209554672241,
"learning_rate": 2.2248427672955977e-06,
"loss": 0.7365,
"step": 1413
},
{
"epoch": 0.5558176100628931,
"grad_norm": 0.873467206954956,
"learning_rate": 2.2228773584905665e-06,
"loss": 0.786,
"step": 1414
},
{
"epoch": 0.5562106918238994,
"grad_norm": 0.7664565443992615,
"learning_rate": 2.2209119496855345e-06,
"loss": 0.6794,
"step": 1415
},
{
"epoch": 0.5566037735849056,
"grad_norm": 0.6069918274879456,
"learning_rate": 2.2189465408805033e-06,
"loss": 0.6237,
"step": 1416
},
{
"epoch": 0.5569968553459119,
"grad_norm": 0.634852409362793,
"learning_rate": 2.2169811320754718e-06,
"loss": 0.5996,
"step": 1417
},
{
"epoch": 0.5573899371069182,
"grad_norm": 0.7907192707061768,
"learning_rate": 2.2150157232704406e-06,
"loss": 0.6921,
"step": 1418
},
{
"epoch": 0.5577830188679245,
"grad_norm": 0.9277790188789368,
"learning_rate": 2.213050314465409e-06,
"loss": 0.8145,
"step": 1419
},
{
"epoch": 0.5581761006289309,
"grad_norm": 0.7333400845527649,
"learning_rate": 2.2110849056603775e-06,
"loss": 0.6949,
"step": 1420
},
{
"epoch": 0.5585691823899371,
"grad_norm": 0.6489284634590149,
"learning_rate": 2.2091194968553463e-06,
"loss": 0.6504,
"step": 1421
},
{
"epoch": 0.5589622641509434,
"grad_norm": 0.7281147837638855,
"learning_rate": 2.2071540880503147e-06,
"loss": 0.7114,
"step": 1422
},
{
"epoch": 0.5593553459119497,
"grad_norm": 0.6547876000404358,
"learning_rate": 2.205188679245283e-06,
"loss": 0.6444,
"step": 1423
},
{
"epoch": 0.559748427672956,
"grad_norm": 0.6834612488746643,
"learning_rate": 2.2032232704402516e-06,
"loss": 0.6035,
"step": 1424
},
{
"epoch": 0.5601415094339622,
"grad_norm": 0.5559141039848328,
"learning_rate": 2.2012578616352204e-06,
"loss": 0.7124,
"step": 1425
},
{
"epoch": 0.5605345911949685,
"grad_norm": 0.5917544364929199,
"learning_rate": 2.199292452830189e-06,
"loss": 0.5973,
"step": 1426
},
{
"epoch": 0.5609276729559748,
"grad_norm": 0.6828616261482239,
"learning_rate": 2.1973270440251573e-06,
"loss": 0.728,
"step": 1427
},
{
"epoch": 0.5613207547169812,
"grad_norm": 0.8474870920181274,
"learning_rate": 2.195361635220126e-06,
"loss": 0.7267,
"step": 1428
},
{
"epoch": 0.5617138364779874,
"grad_norm": 0.7436255216598511,
"learning_rate": 2.1933962264150945e-06,
"loss": 0.8014,
"step": 1429
},
{
"epoch": 0.5621069182389937,
"grad_norm": 0.6264328956604004,
"learning_rate": 2.1914308176100634e-06,
"loss": 0.5736,
"step": 1430
},
{
"epoch": 0.5625,
"grad_norm": 0.5994024276733398,
"learning_rate": 2.1894654088050314e-06,
"loss": 0.6813,
"step": 1431
},
{
"epoch": 0.5628930817610063,
"grad_norm": 0.6246083974838257,
"learning_rate": 2.1875000000000002e-06,
"loss": 0.6226,
"step": 1432
},
{
"epoch": 0.5632861635220126,
"grad_norm": 0.5922362804412842,
"learning_rate": 2.1855345911949687e-06,
"loss": 0.6954,
"step": 1433
},
{
"epoch": 0.5636792452830188,
"grad_norm": 0.8046216368675232,
"learning_rate": 2.183569182389937e-06,
"loss": 0.7067,
"step": 1434
},
{
"epoch": 0.5640723270440252,
"grad_norm": 0.6886887550354004,
"learning_rate": 2.181603773584906e-06,
"loss": 0.6492,
"step": 1435
},
{
"epoch": 0.5644654088050315,
"grad_norm": 0.8090947270393372,
"learning_rate": 2.1796383647798743e-06,
"loss": 0.6863,
"step": 1436
},
{
"epoch": 0.5648584905660378,
"grad_norm": 0.7496331930160522,
"learning_rate": 2.177672955974843e-06,
"loss": 0.8188,
"step": 1437
},
{
"epoch": 0.565251572327044,
"grad_norm": 0.5893301963806152,
"learning_rate": 2.175707547169811e-06,
"loss": 0.689,
"step": 1438
},
{
"epoch": 0.5656446540880503,
"grad_norm": 0.8355748057365417,
"learning_rate": 2.17374213836478e-06,
"loss": 0.754,
"step": 1439
},
{
"epoch": 0.5660377358490566,
"grad_norm": 0.6710575222969055,
"learning_rate": 2.1717767295597485e-06,
"loss": 0.6412,
"step": 1440
},
{
"epoch": 0.5664308176100629,
"grad_norm": 0.716201901435852,
"learning_rate": 2.1698113207547173e-06,
"loss": 0.7264,
"step": 1441
},
{
"epoch": 0.5668238993710691,
"grad_norm": 0.7533861994743347,
"learning_rate": 2.1678459119496857e-06,
"loss": 0.6538,
"step": 1442
},
{
"epoch": 0.5672169811320755,
"grad_norm": 0.5604265928268433,
"learning_rate": 2.165880503144654e-06,
"loss": 0.6714,
"step": 1443
},
{
"epoch": 0.5676100628930818,
"grad_norm": 1.0851994752883911,
"learning_rate": 2.163915094339623e-06,
"loss": 0.7477,
"step": 1444
},
{
"epoch": 0.5680031446540881,
"grad_norm": 0.7837111949920654,
"learning_rate": 2.1619496855345914e-06,
"loss": 0.7832,
"step": 1445
},
{
"epoch": 0.5683962264150944,
"grad_norm": 0.9391263723373413,
"learning_rate": 2.15998427672956e-06,
"loss": 0.7333,
"step": 1446
},
{
"epoch": 0.5687893081761006,
"grad_norm": 0.5665994882583618,
"learning_rate": 2.1580188679245283e-06,
"loss": 0.6742,
"step": 1447
},
{
"epoch": 0.5691823899371069,
"grad_norm": 0.6831005215644836,
"learning_rate": 2.156053459119497e-06,
"loss": 0.6595,
"step": 1448
},
{
"epoch": 0.5695754716981132,
"grad_norm": 0.5789435505867004,
"learning_rate": 2.1540880503144655e-06,
"loss": 0.6597,
"step": 1449
},
{
"epoch": 0.5699685534591195,
"grad_norm": 0.8528465032577515,
"learning_rate": 2.152122641509434e-06,
"loss": 0.6558,
"step": 1450
},
{
"epoch": 0.5703616352201258,
"grad_norm": 0.4888685643672943,
"learning_rate": 2.150157232704403e-06,
"loss": 0.649,
"step": 1451
},
{
"epoch": 0.5707547169811321,
"grad_norm": 0.5105802416801453,
"learning_rate": 2.1481918238993712e-06,
"loss": 0.6522,
"step": 1452
},
{
"epoch": 0.5711477987421384,
"grad_norm": 0.5972757935523987,
"learning_rate": 2.14622641509434e-06,
"loss": 0.6633,
"step": 1453
},
{
"epoch": 0.5715408805031447,
"grad_norm": 0.6490737795829773,
"learning_rate": 2.144261006289308e-06,
"loss": 0.6493,
"step": 1454
},
{
"epoch": 0.5719339622641509,
"grad_norm": 0.8007569313049316,
"learning_rate": 2.142295597484277e-06,
"loss": 0.6786,
"step": 1455
},
{
"epoch": 0.5723270440251572,
"grad_norm": 1.0423263311386108,
"learning_rate": 2.1403301886792454e-06,
"loss": 0.8254,
"step": 1456
},
{
"epoch": 0.5727201257861635,
"grad_norm": 0.8007357716560364,
"learning_rate": 2.138364779874214e-06,
"loss": 0.8081,
"step": 1457
},
{
"epoch": 0.5731132075471698,
"grad_norm": 0.821307361125946,
"learning_rate": 2.1363993710691826e-06,
"loss": 0.8097,
"step": 1458
},
{
"epoch": 0.5735062893081762,
"grad_norm": 0.7179580926895142,
"learning_rate": 2.134433962264151e-06,
"loss": 0.6476,
"step": 1459
},
{
"epoch": 0.5738993710691824,
"grad_norm": 0.8438088893890381,
"learning_rate": 2.13246855345912e-06,
"loss": 0.7626,
"step": 1460
},
{
"epoch": 0.5742924528301887,
"grad_norm": 0.7857232689857483,
"learning_rate": 2.1305031446540883e-06,
"loss": 0.7174,
"step": 1461
},
{
"epoch": 0.574685534591195,
"grad_norm": 0.6653187870979309,
"learning_rate": 2.1285377358490567e-06,
"loss": 0.6888,
"step": 1462
},
{
"epoch": 0.5750786163522013,
"grad_norm": 0.8163049221038818,
"learning_rate": 2.126572327044025e-06,
"loss": 0.7406,
"step": 1463
},
{
"epoch": 0.5754716981132075,
"grad_norm": 0.9705492854118347,
"learning_rate": 2.124606918238994e-06,
"loss": 0.7727,
"step": 1464
},
{
"epoch": 0.5758647798742138,
"grad_norm": 0.7452424764633179,
"learning_rate": 2.1226415094339624e-06,
"loss": 0.6851,
"step": 1465
},
{
"epoch": 0.5762578616352201,
"grad_norm": 0.7768338322639465,
"learning_rate": 2.120676100628931e-06,
"loss": 0.6394,
"step": 1466
},
{
"epoch": 0.5766509433962265,
"grad_norm": 0.9185128211975098,
"learning_rate": 2.1187106918238997e-06,
"loss": 0.7532,
"step": 1467
},
{
"epoch": 0.5770440251572327,
"grad_norm": 0.7840537428855896,
"learning_rate": 2.116745283018868e-06,
"loss": 0.7562,
"step": 1468
},
{
"epoch": 0.577437106918239,
"grad_norm": 0.8269118666648865,
"learning_rate": 2.1147798742138365e-06,
"loss": 0.7581,
"step": 1469
},
{
"epoch": 0.5778301886792453,
"grad_norm": 0.6769137382507324,
"learning_rate": 2.112814465408805e-06,
"loss": 0.6631,
"step": 1470
},
{
"epoch": 0.5782232704402516,
"grad_norm": 0.8486701250076294,
"learning_rate": 2.110849056603774e-06,
"loss": 0.8218,
"step": 1471
},
{
"epoch": 0.5786163522012578,
"grad_norm": 0.8664657473564148,
"learning_rate": 2.1088836477987422e-06,
"loss": 0.6504,
"step": 1472
},
{
"epoch": 0.5790094339622641,
"grad_norm": 0.7925930619239807,
"learning_rate": 2.106918238993711e-06,
"loss": 0.7452,
"step": 1473
},
{
"epoch": 0.5794025157232704,
"grad_norm": 1.082800030708313,
"learning_rate": 2.1049528301886795e-06,
"loss": 0.7551,
"step": 1474
},
{
"epoch": 0.5797955974842768,
"grad_norm": 0.8494970202445984,
"learning_rate": 2.102987421383648e-06,
"loss": 0.7789,
"step": 1475
},
{
"epoch": 0.5801886792452831,
"grad_norm": 0.8614088296890259,
"learning_rate": 2.1010220125786164e-06,
"loss": 0.7727,
"step": 1476
},
{
"epoch": 0.5805817610062893,
"grad_norm": 0.9731083512306213,
"learning_rate": 2.099056603773585e-06,
"loss": 0.9001,
"step": 1477
},
{
"epoch": 0.5809748427672956,
"grad_norm": 0.7833426594734192,
"learning_rate": 2.0970911949685536e-06,
"loss": 0.651,
"step": 1478
},
{
"epoch": 0.5813679245283019,
"grad_norm": 0.8220633864402771,
"learning_rate": 2.095125786163522e-06,
"loss": 0.7029,
"step": 1479
},
{
"epoch": 0.5817610062893082,
"grad_norm": 0.7603692412376404,
"learning_rate": 2.093160377358491e-06,
"loss": 0.75,
"step": 1480
},
{
"epoch": 0.5821540880503144,
"grad_norm": 0.6338460445404053,
"learning_rate": 2.0911949685534593e-06,
"loss": 0.6525,
"step": 1481
},
{
"epoch": 0.5825471698113207,
"grad_norm": 0.6147676706314087,
"learning_rate": 2.0892295597484277e-06,
"loss": 0.7159,
"step": 1482
},
{
"epoch": 0.5829402515723271,
"grad_norm": 0.6025592088699341,
"learning_rate": 2.087264150943396e-06,
"loss": 0.6693,
"step": 1483
},
{
"epoch": 0.5833333333333334,
"grad_norm": 0.7551356554031372,
"learning_rate": 2.085298742138365e-06,
"loss": 0.7025,
"step": 1484
},
{
"epoch": 0.5837264150943396,
"grad_norm": 0.704410195350647,
"learning_rate": 2.0833333333333334e-06,
"loss": 0.6545,
"step": 1485
},
{
"epoch": 0.5841194968553459,
"grad_norm": 0.5775546431541443,
"learning_rate": 2.081367924528302e-06,
"loss": 0.6713,
"step": 1486
},
{
"epoch": 0.5845125786163522,
"grad_norm": 0.6387528777122498,
"learning_rate": 2.0794025157232707e-06,
"loss": 0.7125,
"step": 1487
},
{
"epoch": 0.5849056603773585,
"grad_norm": 0.7859731316566467,
"learning_rate": 2.077437106918239e-06,
"loss": 0.713,
"step": 1488
},
{
"epoch": 0.5852987421383647,
"grad_norm": 0.75186687707901,
"learning_rate": 2.075471698113208e-06,
"loss": 0.6788,
"step": 1489
},
{
"epoch": 0.585691823899371,
"grad_norm": 0.6188842058181763,
"learning_rate": 2.073506289308176e-06,
"loss": 0.631,
"step": 1490
},
{
"epoch": 0.5860849056603774,
"grad_norm": 0.6439576148986816,
"learning_rate": 2.071540880503145e-06,
"loss": 0.6488,
"step": 1491
},
{
"epoch": 0.5864779874213837,
"grad_norm": 0.695747435092926,
"learning_rate": 2.0695754716981132e-06,
"loss": 0.6803,
"step": 1492
},
{
"epoch": 0.58687106918239,
"grad_norm": 0.7508324384689331,
"learning_rate": 2.067610062893082e-06,
"loss": 0.5656,
"step": 1493
},
{
"epoch": 0.5872641509433962,
"grad_norm": 0.957109272480011,
"learning_rate": 2.0656446540880505e-06,
"loss": 0.8134,
"step": 1494
},
{
"epoch": 0.5876572327044025,
"grad_norm": 0.5503519773483276,
"learning_rate": 2.063679245283019e-06,
"loss": 0.6449,
"step": 1495
},
{
"epoch": 0.5880503144654088,
"grad_norm": 0.6186985373497009,
"learning_rate": 2.0617138364779878e-06,
"loss": 0.6102,
"step": 1496
},
{
"epoch": 0.5884433962264151,
"grad_norm": 0.6537104249000549,
"learning_rate": 2.059748427672956e-06,
"loss": 0.6503,
"step": 1497
},
{
"epoch": 0.5888364779874213,
"grad_norm": 0.5538856983184814,
"learning_rate": 2.0577830188679246e-06,
"loss": 0.6243,
"step": 1498
},
{
"epoch": 0.5892295597484277,
"grad_norm": 0.8563919067382812,
"learning_rate": 2.055817610062893e-06,
"loss": 0.7528,
"step": 1499
},
{
"epoch": 0.589622641509434,
"grad_norm": 0.6991892457008362,
"learning_rate": 2.053852201257862e-06,
"loss": 0.6813,
"step": 1500
},
{
"epoch": 0.5900157232704403,
"grad_norm": 0.7889415621757507,
"learning_rate": 2.0518867924528303e-06,
"loss": 0.7772,
"step": 1501
},
{
"epoch": 0.5904088050314465,
"grad_norm": 0.8779224157333374,
"learning_rate": 2.0499213836477987e-06,
"loss": 0.7415,
"step": 1502
},
{
"epoch": 0.5908018867924528,
"grad_norm": 0.6833891868591309,
"learning_rate": 2.0479559748427676e-06,
"loss": 0.6829,
"step": 1503
},
{
"epoch": 0.5911949685534591,
"grad_norm": 0.8189573287963867,
"learning_rate": 2.045990566037736e-06,
"loss": 0.7422,
"step": 1504
},
{
"epoch": 0.5915880503144654,
"grad_norm": 0.6660454869270325,
"learning_rate": 2.044025157232705e-06,
"loss": 0.6417,
"step": 1505
},
{
"epoch": 0.5919811320754716,
"grad_norm": 0.7052726745605469,
"learning_rate": 2.042059748427673e-06,
"loss": 0.7036,
"step": 1506
},
{
"epoch": 0.592374213836478,
"grad_norm": 0.7753020524978638,
"learning_rate": 2.0400943396226417e-06,
"loss": 0.7767,
"step": 1507
},
{
"epoch": 0.5927672955974843,
"grad_norm": 0.7334070801734924,
"learning_rate": 2.03812893081761e-06,
"loss": 0.6805,
"step": 1508
},
{
"epoch": 0.5931603773584906,
"grad_norm": 0.6439114212989807,
"learning_rate": 2.036163522012579e-06,
"loss": 0.6393,
"step": 1509
},
{
"epoch": 0.5935534591194969,
"grad_norm": 0.5840775966644287,
"learning_rate": 2.0341981132075474e-06,
"loss": 0.5927,
"step": 1510
},
{
"epoch": 0.5939465408805031,
"grad_norm": 0.9419865012168884,
"learning_rate": 2.032232704402516e-06,
"loss": 0.6955,
"step": 1511
},
{
"epoch": 0.5943396226415094,
"grad_norm": 0.751986026763916,
"learning_rate": 2.0302672955974847e-06,
"loss": 0.7368,
"step": 1512
},
{
"epoch": 0.5947327044025157,
"grad_norm": 0.6326794028282166,
"learning_rate": 2.028301886792453e-06,
"loss": 0.6224,
"step": 1513
},
{
"epoch": 0.595125786163522,
"grad_norm": 0.7244898676872253,
"learning_rate": 2.0263364779874215e-06,
"loss": 0.5973,
"step": 1514
},
{
"epoch": 0.5955188679245284,
"grad_norm": 0.8055747747421265,
"learning_rate": 2.02437106918239e-06,
"loss": 0.7173,
"step": 1515
},
{
"epoch": 0.5959119496855346,
"grad_norm": 0.868969202041626,
"learning_rate": 2.0224056603773588e-06,
"loss": 0.7006,
"step": 1516
},
{
"epoch": 0.5963050314465409,
"grad_norm": 0.8277462720870972,
"learning_rate": 2.020440251572327e-06,
"loss": 0.7562,
"step": 1517
},
{
"epoch": 0.5966981132075472,
"grad_norm": 0.592400312423706,
"learning_rate": 2.0184748427672956e-06,
"loss": 0.691,
"step": 1518
},
{
"epoch": 0.5970911949685535,
"grad_norm": 0.6293091773986816,
"learning_rate": 2.0165094339622645e-06,
"loss": 0.5934,
"step": 1519
},
{
"epoch": 0.5974842767295597,
"grad_norm": 0.6954889893531799,
"learning_rate": 2.014544025157233e-06,
"loss": 0.6127,
"step": 1520
},
{
"epoch": 0.597877358490566,
"grad_norm": 0.8755274415016174,
"learning_rate": 2.0125786163522013e-06,
"loss": 0.7639,
"step": 1521
},
{
"epoch": 0.5982704402515723,
"grad_norm": 0.6167333126068115,
"learning_rate": 2.0106132075471697e-06,
"loss": 0.6421,
"step": 1522
},
{
"epoch": 0.5986635220125787,
"grad_norm": 0.6424795389175415,
"learning_rate": 2.0086477987421386e-06,
"loss": 0.6903,
"step": 1523
},
{
"epoch": 0.5990566037735849,
"grad_norm": 0.6280001401901245,
"learning_rate": 2.006682389937107e-06,
"loss": 0.6295,
"step": 1524
},
{
"epoch": 0.5994496855345912,
"grad_norm": 0.6509950757026672,
"learning_rate": 2.004716981132076e-06,
"loss": 0.6864,
"step": 1525
},
{
"epoch": 0.5998427672955975,
"grad_norm": 0.5580452680587769,
"learning_rate": 2.0027515723270443e-06,
"loss": 0.5824,
"step": 1526
},
{
"epoch": 0.6002358490566038,
"grad_norm": 0.7185975313186646,
"learning_rate": 2.0007861635220127e-06,
"loss": 0.6588,
"step": 1527
},
{
"epoch": 0.60062893081761,
"grad_norm": 0.5868175029754639,
"learning_rate": 1.998820754716981e-06,
"loss": 0.6359,
"step": 1528
},
{
"epoch": 0.6010220125786163,
"grad_norm": 0.7772722244262695,
"learning_rate": 1.99685534591195e-06,
"loss": 0.7399,
"step": 1529
},
{
"epoch": 0.6014150943396226,
"grad_norm": 0.5818542838096619,
"learning_rate": 1.9948899371069184e-06,
"loss": 0.5962,
"step": 1530
},
{
"epoch": 0.601808176100629,
"grad_norm": 0.8383485078811646,
"learning_rate": 1.992924528301887e-06,
"loss": 0.7435,
"step": 1531
},
{
"epoch": 0.6022012578616353,
"grad_norm": 0.9404725432395935,
"learning_rate": 1.9909591194968557e-06,
"loss": 0.7328,
"step": 1532
},
{
"epoch": 0.6025943396226415,
"grad_norm": 0.4395799934864044,
"learning_rate": 1.988993710691824e-06,
"loss": 0.6516,
"step": 1533
},
{
"epoch": 0.6029874213836478,
"grad_norm": 0.7806381583213806,
"learning_rate": 1.9870283018867925e-06,
"loss": 0.7347,
"step": 1534
},
{
"epoch": 0.6033805031446541,
"grad_norm": 0.6776301860809326,
"learning_rate": 1.985062893081761e-06,
"loss": 0.7339,
"step": 1535
},
{
"epoch": 0.6037735849056604,
"grad_norm": 0.7533168792724609,
"learning_rate": 1.9830974842767298e-06,
"loss": 0.7792,
"step": 1536
},
{
"epoch": 0.6041666666666666,
"grad_norm": 0.8690527081489563,
"learning_rate": 1.981132075471698e-06,
"loss": 0.8342,
"step": 1537
},
{
"epoch": 0.6045597484276729,
"grad_norm": 0.9228425621986389,
"learning_rate": 1.9791666666666666e-06,
"loss": 0.7288,
"step": 1538
},
{
"epoch": 0.6049528301886793,
"grad_norm": 0.8432055115699768,
"learning_rate": 1.9772012578616355e-06,
"loss": 0.7749,
"step": 1539
},
{
"epoch": 0.6053459119496856,
"grad_norm": 0.7854190468788147,
"learning_rate": 1.975235849056604e-06,
"loss": 0.7202,
"step": 1540
},
{
"epoch": 0.6057389937106918,
"grad_norm": 0.6895219087600708,
"learning_rate": 1.9732704402515728e-06,
"loss": 0.6995,
"step": 1541
},
{
"epoch": 0.6061320754716981,
"grad_norm": 0.8248894810676575,
"learning_rate": 1.9713050314465408e-06,
"loss": 0.861,
"step": 1542
},
{
"epoch": 0.6065251572327044,
"grad_norm": 0.6397697925567627,
"learning_rate": 1.9693396226415096e-06,
"loss": 0.6816,
"step": 1543
},
{
"epoch": 0.6069182389937107,
"grad_norm": 0.6951172947883606,
"learning_rate": 1.967374213836478e-06,
"loss": 0.7247,
"step": 1544
},
{
"epoch": 0.6073113207547169,
"grad_norm": 0.6109116673469543,
"learning_rate": 1.965408805031447e-06,
"loss": 0.5535,
"step": 1545
},
{
"epoch": 0.6077044025157232,
"grad_norm": 0.614949107170105,
"learning_rate": 1.9634433962264153e-06,
"loss": 0.6682,
"step": 1546
},
{
"epoch": 0.6080974842767296,
"grad_norm": 0.8841148018836975,
"learning_rate": 1.9614779874213837e-06,
"loss": 0.6929,
"step": 1547
},
{
"epoch": 0.6084905660377359,
"grad_norm": 0.6391963362693787,
"learning_rate": 1.9595125786163526e-06,
"loss": 0.7081,
"step": 1548
},
{
"epoch": 0.6088836477987422,
"grad_norm": 1.2931239604949951,
"learning_rate": 1.957547169811321e-06,
"loss": 0.7122,
"step": 1549
},
{
"epoch": 0.6092767295597484,
"grad_norm": 0.6303018927574158,
"learning_rate": 1.9555817610062894e-06,
"loss": 0.6681,
"step": 1550
},
{
"epoch": 0.6096698113207547,
"grad_norm": 0.652459442615509,
"learning_rate": 1.953616352201258e-06,
"loss": 0.6688,
"step": 1551
},
{
"epoch": 0.610062893081761,
"grad_norm": 0.6887359023094177,
"learning_rate": 1.9516509433962267e-06,
"loss": 0.6489,
"step": 1552
},
{
"epoch": 0.6104559748427673,
"grad_norm": 0.6565138697624207,
"learning_rate": 1.949685534591195e-06,
"loss": 0.7555,
"step": 1553
},
{
"epoch": 0.6108490566037735,
"grad_norm": 0.7731680870056152,
"learning_rate": 1.9477201257861635e-06,
"loss": 0.7771,
"step": 1554
},
{
"epoch": 0.6112421383647799,
"grad_norm": 0.776526927947998,
"learning_rate": 1.9457547169811324e-06,
"loss": 0.7239,
"step": 1555
},
{
"epoch": 0.6116352201257862,
"grad_norm": 0.7610769271850586,
"learning_rate": 1.943789308176101e-06,
"loss": 0.6484,
"step": 1556
},
{
"epoch": 0.6120283018867925,
"grad_norm": 0.8110302090644836,
"learning_rate": 1.9418238993710696e-06,
"loss": 0.7609,
"step": 1557
},
{
"epoch": 0.6124213836477987,
"grad_norm": 0.7409825325012207,
"learning_rate": 1.9398584905660376e-06,
"loss": 0.7298,
"step": 1558
},
{
"epoch": 0.612814465408805,
"grad_norm": 0.8083226084709167,
"learning_rate": 1.9378930817610065e-06,
"loss": 0.8445,
"step": 1559
},
{
"epoch": 0.6132075471698113,
"grad_norm": 0.7841634750366211,
"learning_rate": 1.935927672955975e-06,
"loss": 0.7306,
"step": 1560
},
{
"epoch": 0.6136006289308176,
"grad_norm": 0.5882968902587891,
"learning_rate": 1.9339622641509438e-06,
"loss": 0.7096,
"step": 1561
},
{
"epoch": 0.6139937106918238,
"grad_norm": 0.8343744277954102,
"learning_rate": 1.931996855345912e-06,
"loss": 0.707,
"step": 1562
},
{
"epoch": 0.6143867924528302,
"grad_norm": 0.6858739852905273,
"learning_rate": 1.9300314465408806e-06,
"loss": 0.7243,
"step": 1563
},
{
"epoch": 0.6147798742138365,
"grad_norm": 0.7866342663764954,
"learning_rate": 1.9280660377358494e-06,
"loss": 0.712,
"step": 1564
},
{
"epoch": 0.6151729559748428,
"grad_norm": 0.5476449131965637,
"learning_rate": 1.926100628930818e-06,
"loss": 0.6786,
"step": 1565
},
{
"epoch": 0.6155660377358491,
"grad_norm": 0.7673195600509644,
"learning_rate": 1.9241352201257863e-06,
"loss": 0.8331,
"step": 1566
},
{
"epoch": 0.6159591194968553,
"grad_norm": 0.7172686457633972,
"learning_rate": 1.9221698113207547e-06,
"loss": 0.6608,
"step": 1567
},
{
"epoch": 0.6163522012578616,
"grad_norm": 0.7932978272438049,
"learning_rate": 1.9202044025157236e-06,
"loss": 0.8113,
"step": 1568
},
{
"epoch": 0.6167452830188679,
"grad_norm": 0.6243324279785156,
"learning_rate": 1.918238993710692e-06,
"loss": 0.5791,
"step": 1569
},
{
"epoch": 0.6171383647798742,
"grad_norm": 0.6123634576797485,
"learning_rate": 1.9162735849056604e-06,
"loss": 0.6413,
"step": 1570
},
{
"epoch": 0.6175314465408805,
"grad_norm": 0.6223311424255371,
"learning_rate": 1.9143081761006293e-06,
"loss": 0.7057,
"step": 1571
},
{
"epoch": 0.6179245283018868,
"grad_norm": 0.7178637981414795,
"learning_rate": 1.9123427672955977e-06,
"loss": 0.6733,
"step": 1572
},
{
"epoch": 0.6183176100628931,
"grad_norm": 0.6080794930458069,
"learning_rate": 1.9103773584905665e-06,
"loss": 0.6296,
"step": 1573
},
{
"epoch": 0.6187106918238994,
"grad_norm": 0.8301292657852173,
"learning_rate": 1.9084119496855345e-06,
"loss": 0.6362,
"step": 1574
},
{
"epoch": 0.6191037735849056,
"grad_norm": 0.7345229983329773,
"learning_rate": 1.9064465408805034e-06,
"loss": 0.6395,
"step": 1575
},
{
"epoch": 0.6194968553459119,
"grad_norm": 0.6985357403755188,
"learning_rate": 1.904481132075472e-06,
"loss": 0.588,
"step": 1576
},
{
"epoch": 0.6198899371069182,
"grad_norm": 0.870455265045166,
"learning_rate": 1.9025157232704406e-06,
"loss": 0.7942,
"step": 1577
},
{
"epoch": 0.6202830188679245,
"grad_norm": 0.7607539892196655,
"learning_rate": 1.9005503144654089e-06,
"loss": 0.7429,
"step": 1578
},
{
"epoch": 0.6206761006289309,
"grad_norm": 0.7141609191894531,
"learning_rate": 1.8985849056603775e-06,
"loss": 0.7468,
"step": 1579
},
{
"epoch": 0.6210691823899371,
"grad_norm": 0.7514468431472778,
"learning_rate": 1.8966194968553461e-06,
"loss": 0.717,
"step": 1580
},
{
"epoch": 0.6214622641509434,
"grad_norm": 0.6254903674125671,
"learning_rate": 1.8946540880503148e-06,
"loss": 0.7057,
"step": 1581
},
{
"epoch": 0.6218553459119497,
"grad_norm": 0.5091788172721863,
"learning_rate": 1.8926886792452832e-06,
"loss": 0.6443,
"step": 1582
},
{
"epoch": 0.622248427672956,
"grad_norm": 0.7453218698501587,
"learning_rate": 1.8907232704402518e-06,
"loss": 0.7223,
"step": 1583
},
{
"epoch": 0.6226415094339622,
"grad_norm": 0.6915518045425415,
"learning_rate": 1.8887578616352205e-06,
"loss": 0.7333,
"step": 1584
},
{
"epoch": 0.6230345911949685,
"grad_norm": 0.7891326546669006,
"learning_rate": 1.8867924528301889e-06,
"loss": 0.6799,
"step": 1585
},
{
"epoch": 0.6234276729559748,
"grad_norm": 0.7398204207420349,
"learning_rate": 1.8848270440251573e-06,
"loss": 0.8137,
"step": 1586
},
{
"epoch": 0.6238207547169812,
"grad_norm": 0.8431394100189209,
"learning_rate": 1.882861635220126e-06,
"loss": 0.7965,
"step": 1587
},
{
"epoch": 0.6242138364779874,
"grad_norm": 0.4613338112831116,
"learning_rate": 1.8808962264150946e-06,
"loss": 0.5831,
"step": 1588
},
{
"epoch": 0.6246069182389937,
"grad_norm": 0.852630078792572,
"learning_rate": 1.8789308176100632e-06,
"loss": 0.7612,
"step": 1589
},
{
"epoch": 0.625,
"grad_norm": 0.6552901268005371,
"learning_rate": 1.8769654088050316e-06,
"loss": 0.692,
"step": 1590
},
{
"epoch": 0.6253930817610063,
"grad_norm": 0.7663810849189758,
"learning_rate": 1.8750000000000003e-06,
"loss": 0.6616,
"step": 1591
},
{
"epoch": 0.6257861635220126,
"grad_norm": 0.8007197380065918,
"learning_rate": 1.8730345911949687e-06,
"loss": 0.665,
"step": 1592
},
{
"epoch": 0.6261792452830188,
"grad_norm": 0.46744146943092346,
"learning_rate": 1.871069182389937e-06,
"loss": 0.5951,
"step": 1593
},
{
"epoch": 0.6265723270440252,
"grad_norm": 0.6524441242218018,
"learning_rate": 1.8691037735849057e-06,
"loss": 0.6097,
"step": 1594
},
{
"epoch": 0.6269654088050315,
"grad_norm": 0.885798990726471,
"learning_rate": 1.8671383647798744e-06,
"loss": 0.8324,
"step": 1595
},
{
"epoch": 0.6273584905660378,
"grad_norm": 0.7618914842605591,
"learning_rate": 1.865172955974843e-06,
"loss": 0.7295,
"step": 1596
},
{
"epoch": 0.627751572327044,
"grad_norm": 0.6547210216522217,
"learning_rate": 1.8632075471698114e-06,
"loss": 0.7311,
"step": 1597
},
{
"epoch": 0.6281446540880503,
"grad_norm": 0.5837027430534363,
"learning_rate": 1.8612421383647799e-06,
"loss": 0.6344,
"step": 1598
},
{
"epoch": 0.6285377358490566,
"grad_norm": 0.7335178852081299,
"learning_rate": 1.8592767295597485e-06,
"loss": 0.7168,
"step": 1599
},
{
"epoch": 0.6289308176100629,
"grad_norm": 0.7043851017951965,
"learning_rate": 1.8573113207547171e-06,
"loss": 0.7474,
"step": 1600
},
{
"epoch": 0.6293238993710691,
"grad_norm": 0.7553254961967468,
"learning_rate": 1.8553459119496855e-06,
"loss": 0.6551,
"step": 1601
},
{
"epoch": 0.6297169811320755,
"grad_norm": 0.5837968587875366,
"learning_rate": 1.8533805031446542e-06,
"loss": 0.6226,
"step": 1602
},
{
"epoch": 0.6301100628930818,
"grad_norm": 0.8485018014907837,
"learning_rate": 1.8514150943396228e-06,
"loss": 0.8352,
"step": 1603
},
{
"epoch": 0.6305031446540881,
"grad_norm": 0.5657935738563538,
"learning_rate": 1.8494496855345915e-06,
"loss": 0.6117,
"step": 1604
},
{
"epoch": 0.6308962264150944,
"grad_norm": 0.721409797668457,
"learning_rate": 1.8474842767295597e-06,
"loss": 0.7319,
"step": 1605
},
{
"epoch": 0.6312893081761006,
"grad_norm": 0.7759498953819275,
"learning_rate": 1.8455188679245283e-06,
"loss": 0.7062,
"step": 1606
},
{
"epoch": 0.6316823899371069,
"grad_norm": 0.8070909976959229,
"learning_rate": 1.843553459119497e-06,
"loss": 0.746,
"step": 1607
},
{
"epoch": 0.6320754716981132,
"grad_norm": 0.7674651145935059,
"learning_rate": 1.8415880503144656e-06,
"loss": 0.7354,
"step": 1608
},
{
"epoch": 0.6324685534591195,
"grad_norm": 0.617154061794281,
"learning_rate": 1.839622641509434e-06,
"loss": 0.6823,
"step": 1609
},
{
"epoch": 0.6328616352201258,
"grad_norm": 0.8871747255325317,
"learning_rate": 1.8376572327044026e-06,
"loss": 0.7781,
"step": 1610
},
{
"epoch": 0.6332547169811321,
"grad_norm": 0.6469852924346924,
"learning_rate": 1.8356918238993713e-06,
"loss": 0.709,
"step": 1611
},
{
"epoch": 0.6336477987421384,
"grad_norm": 0.5585039854049683,
"learning_rate": 1.83372641509434e-06,
"loss": 0.6595,
"step": 1612
},
{
"epoch": 0.6340408805031447,
"grad_norm": 0.5893180966377258,
"learning_rate": 1.8317610062893081e-06,
"loss": 0.73,
"step": 1613
},
{
"epoch": 0.6344339622641509,
"grad_norm": 0.8004491329193115,
"learning_rate": 1.8297955974842767e-06,
"loss": 0.7444,
"step": 1614
},
{
"epoch": 0.6348270440251572,
"grad_norm": 0.7132411003112793,
"learning_rate": 1.8278301886792454e-06,
"loss": 0.729,
"step": 1615
},
{
"epoch": 0.6352201257861635,
"grad_norm": 0.7293667793273926,
"learning_rate": 1.825864779874214e-06,
"loss": 0.7333,
"step": 1616
},
{
"epoch": 0.6356132075471698,
"grad_norm": 0.5375982522964478,
"learning_rate": 1.8238993710691824e-06,
"loss": 0.6362,
"step": 1617
},
{
"epoch": 0.6360062893081762,
"grad_norm": 0.8626253008842468,
"learning_rate": 1.821933962264151e-06,
"loss": 0.7975,
"step": 1618
},
{
"epoch": 0.6363993710691824,
"grad_norm": 0.5614542365074158,
"learning_rate": 1.8199685534591197e-06,
"loss": 0.6324,
"step": 1619
},
{
"epoch": 0.6367924528301887,
"grad_norm": 0.8241792917251587,
"learning_rate": 1.8180031446540883e-06,
"loss": 0.6727,
"step": 1620
},
{
"epoch": 0.637185534591195,
"grad_norm": 0.7126257419586182,
"learning_rate": 1.8160377358490566e-06,
"loss": 0.6605,
"step": 1621
},
{
"epoch": 0.6375786163522013,
"grad_norm": 0.5685837268829346,
"learning_rate": 1.8140723270440252e-06,
"loss": 0.6986,
"step": 1622
},
{
"epoch": 0.6379716981132075,
"grad_norm": 0.5807716250419617,
"learning_rate": 1.8121069182389938e-06,
"loss": 0.6713,
"step": 1623
},
{
"epoch": 0.6383647798742138,
"grad_norm": 0.6329025030136108,
"learning_rate": 1.8101415094339625e-06,
"loss": 0.717,
"step": 1624
},
{
"epoch": 0.6387578616352201,
"grad_norm": 0.7511675953865051,
"learning_rate": 1.8081761006289309e-06,
"loss": 0.679,
"step": 1625
},
{
"epoch": 0.6391509433962265,
"grad_norm": 0.6215159296989441,
"learning_rate": 1.8062106918238995e-06,
"loss": 0.6537,
"step": 1626
},
{
"epoch": 0.6395440251572327,
"grad_norm": 0.6716428995132446,
"learning_rate": 1.8042452830188682e-06,
"loss": 0.6787,
"step": 1627
},
{
"epoch": 0.639937106918239,
"grad_norm": 0.6762564778327942,
"learning_rate": 1.8022798742138368e-06,
"loss": 0.5875,
"step": 1628
},
{
"epoch": 0.6403301886792453,
"grad_norm": 0.8625909686088562,
"learning_rate": 1.800314465408805e-06,
"loss": 0.6626,
"step": 1629
},
{
"epoch": 0.6407232704402516,
"grad_norm": 0.5872889161109924,
"learning_rate": 1.7983490566037736e-06,
"loss": 0.5813,
"step": 1630
},
{
"epoch": 0.6411163522012578,
"grad_norm": 0.5575698614120483,
"learning_rate": 1.7963836477987423e-06,
"loss": 0.5954,
"step": 1631
},
{
"epoch": 0.6415094339622641,
"grad_norm": 0.6191297173500061,
"learning_rate": 1.794418238993711e-06,
"loss": 0.6503,
"step": 1632
},
{
"epoch": 0.6419025157232704,
"grad_norm": 0.5577523112297058,
"learning_rate": 1.7924528301886793e-06,
"loss": 0.6044,
"step": 1633
},
{
"epoch": 0.6422955974842768,
"grad_norm": 0.7050474882125854,
"learning_rate": 1.790487421383648e-06,
"loss": 0.6392,
"step": 1634
},
{
"epoch": 0.6426886792452831,
"grad_norm": 0.6038858294487,
"learning_rate": 1.7885220125786166e-06,
"loss": 0.6733,
"step": 1635
},
{
"epoch": 0.6430817610062893,
"grad_norm": 0.7716225981712341,
"learning_rate": 1.7865566037735852e-06,
"loss": 0.7461,
"step": 1636
},
{
"epoch": 0.6434748427672956,
"grad_norm": 0.7664049863815308,
"learning_rate": 1.7845911949685534e-06,
"loss": 0.7101,
"step": 1637
},
{
"epoch": 0.6438679245283019,
"grad_norm": 1.2803711891174316,
"learning_rate": 1.782625786163522e-06,
"loss": 0.7732,
"step": 1638
},
{
"epoch": 0.6442610062893082,
"grad_norm": 0.734843373298645,
"learning_rate": 1.7806603773584907e-06,
"loss": 0.5905,
"step": 1639
},
{
"epoch": 0.6446540880503144,
"grad_norm": 0.5314722657203674,
"learning_rate": 1.7786949685534593e-06,
"loss": 0.6339,
"step": 1640
},
{
"epoch": 0.6450471698113207,
"grad_norm": 0.6672811508178711,
"learning_rate": 1.7767295597484278e-06,
"loss": 0.6613,
"step": 1641
},
{
"epoch": 0.6454402515723271,
"grad_norm": 0.6203894019126892,
"learning_rate": 1.7747641509433964e-06,
"loss": 0.6006,
"step": 1642
},
{
"epoch": 0.6458333333333334,
"grad_norm": 0.4984264373779297,
"learning_rate": 1.772798742138365e-06,
"loss": 0.5751,
"step": 1643
},
{
"epoch": 0.6462264150943396,
"grad_norm": 0.641929566860199,
"learning_rate": 1.7708333333333337e-06,
"loss": 0.6167,
"step": 1644
},
{
"epoch": 0.6466194968553459,
"grad_norm": 0.8082166910171509,
"learning_rate": 1.7688679245283019e-06,
"loss": 0.7732,
"step": 1645
},
{
"epoch": 0.6470125786163522,
"grad_norm": 0.5086861848831177,
"learning_rate": 1.7669025157232705e-06,
"loss": 0.6686,
"step": 1646
},
{
"epoch": 0.6474056603773585,
"grad_norm": 0.6730546951293945,
"learning_rate": 1.7649371069182392e-06,
"loss": 0.6727,
"step": 1647
},
{
"epoch": 0.6477987421383647,
"grad_norm": 0.6173148155212402,
"learning_rate": 1.7629716981132078e-06,
"loss": 0.6548,
"step": 1648
},
{
"epoch": 0.648191823899371,
"grad_norm": 0.6544507145881653,
"learning_rate": 1.7610062893081762e-06,
"loss": 0.6062,
"step": 1649
},
{
"epoch": 0.6485849056603774,
"grad_norm": 0.7197887301445007,
"learning_rate": 1.7590408805031448e-06,
"loss": 0.7311,
"step": 1650
},
{
"epoch": 0.6489779874213837,
"grad_norm": 0.7869980931282043,
"learning_rate": 1.7570754716981135e-06,
"loss": 0.641,
"step": 1651
},
{
"epoch": 0.64937106918239,
"grad_norm": 0.8897876143455505,
"learning_rate": 1.755110062893082e-06,
"loss": 0.7382,
"step": 1652
},
{
"epoch": 0.6497641509433962,
"grad_norm": 0.6859142184257507,
"learning_rate": 1.7531446540880503e-06,
"loss": 0.6589,
"step": 1653
},
{
"epoch": 0.6501572327044025,
"grad_norm": 0.8492444753646851,
"learning_rate": 1.751179245283019e-06,
"loss": 0.6852,
"step": 1654
},
{
"epoch": 0.6505503144654088,
"grad_norm": 0.6911584734916687,
"learning_rate": 1.7492138364779876e-06,
"loss": 0.764,
"step": 1655
},
{
"epoch": 0.6509433962264151,
"grad_norm": 0.6768414974212646,
"learning_rate": 1.7472484276729562e-06,
"loss": 0.6881,
"step": 1656
},
{
"epoch": 0.6513364779874213,
"grad_norm": 0.5681583285331726,
"learning_rate": 1.7452830188679247e-06,
"loss": 0.6085,
"step": 1657
},
{
"epoch": 0.6517295597484277,
"grad_norm": 0.6894484758377075,
"learning_rate": 1.7433176100628933e-06,
"loss": 0.599,
"step": 1658
},
{
"epoch": 0.652122641509434,
"grad_norm": 0.5970176458358765,
"learning_rate": 1.7413522012578617e-06,
"loss": 0.6941,
"step": 1659
},
{
"epoch": 0.6525157232704403,
"grad_norm": 0.8096800446510315,
"learning_rate": 1.7393867924528303e-06,
"loss": 0.8707,
"step": 1660
},
{
"epoch": 0.6529088050314465,
"grad_norm": 0.7440771460533142,
"learning_rate": 1.7374213836477988e-06,
"loss": 0.6355,
"step": 1661
},
{
"epoch": 0.6533018867924528,
"grad_norm": 0.7624602913856506,
"learning_rate": 1.7354559748427674e-06,
"loss": 0.7155,
"step": 1662
},
{
"epoch": 0.6536949685534591,
"grad_norm": 0.6406753659248352,
"learning_rate": 1.733490566037736e-06,
"loss": 0.5961,
"step": 1663
},
{
"epoch": 0.6540880503144654,
"grad_norm": 0.7344948649406433,
"learning_rate": 1.7315251572327047e-06,
"loss": 0.7979,
"step": 1664
},
{
"epoch": 0.6544811320754716,
"grad_norm": 0.8572521209716797,
"learning_rate": 1.7295597484276729e-06,
"loss": 0.8135,
"step": 1665
},
{
"epoch": 0.654874213836478,
"grad_norm": 0.8575150966644287,
"learning_rate": 1.7275943396226415e-06,
"loss": 0.7719,
"step": 1666
},
{
"epoch": 0.6552672955974843,
"grad_norm": 0.9233154654502869,
"learning_rate": 1.7256289308176102e-06,
"loss": 0.7775,
"step": 1667
},
{
"epoch": 0.6556603773584906,
"grad_norm": 0.5930840373039246,
"learning_rate": 1.7236635220125788e-06,
"loss": 0.6564,
"step": 1668
},
{
"epoch": 0.6560534591194969,
"grad_norm": 0.6798102855682373,
"learning_rate": 1.7216981132075472e-06,
"loss": 0.6864,
"step": 1669
},
{
"epoch": 0.6564465408805031,
"grad_norm": 0.6273459196090698,
"learning_rate": 1.7197327044025158e-06,
"loss": 0.6616,
"step": 1670
},
{
"epoch": 0.6568396226415094,
"grad_norm": 0.6842119693756104,
"learning_rate": 1.7177672955974845e-06,
"loss": 0.7283,
"step": 1671
},
{
"epoch": 0.6572327044025157,
"grad_norm": 0.8604435324668884,
"learning_rate": 1.7158018867924531e-06,
"loss": 0.6537,
"step": 1672
},
{
"epoch": 0.657625786163522,
"grad_norm": 0.689431369304657,
"learning_rate": 1.7138364779874213e-06,
"loss": 0.6609,
"step": 1673
},
{
"epoch": 0.6580188679245284,
"grad_norm": 0.6235973834991455,
"learning_rate": 1.71187106918239e-06,
"loss": 0.6639,
"step": 1674
},
{
"epoch": 0.6584119496855346,
"grad_norm": 0.9754005670547485,
"learning_rate": 1.7099056603773586e-06,
"loss": 0.6614,
"step": 1675
},
{
"epoch": 0.6588050314465409,
"grad_norm": 0.6192585825920105,
"learning_rate": 1.7079402515723272e-06,
"loss": 0.6758,
"step": 1676
},
{
"epoch": 0.6591981132075472,
"grad_norm": 0.6636350750923157,
"learning_rate": 1.7059748427672957e-06,
"loss": 0.623,
"step": 1677
},
{
"epoch": 0.6595911949685535,
"grad_norm": 0.9329872727394104,
"learning_rate": 1.7040094339622643e-06,
"loss": 0.7694,
"step": 1678
},
{
"epoch": 0.6599842767295597,
"grad_norm": 0.7850617170333862,
"learning_rate": 1.702044025157233e-06,
"loss": 0.7228,
"step": 1679
},
{
"epoch": 0.660377358490566,
"grad_norm": 0.4603402316570282,
"learning_rate": 1.7000786163522016e-06,
"loss": 0.6357,
"step": 1680
},
{
"epoch": 0.6607704402515723,
"grad_norm": 0.6385326981544495,
"learning_rate": 1.6981132075471698e-06,
"loss": 0.6641,
"step": 1681
},
{
"epoch": 0.6611635220125787,
"grad_norm": 0.9375196695327759,
"learning_rate": 1.6961477987421384e-06,
"loss": 0.7057,
"step": 1682
},
{
"epoch": 0.6615566037735849,
"grad_norm": 0.6431612968444824,
"learning_rate": 1.694182389937107e-06,
"loss": 0.7269,
"step": 1683
},
{
"epoch": 0.6619496855345912,
"grad_norm": 0.6182874441146851,
"learning_rate": 1.6922169811320757e-06,
"loss": 0.6025,
"step": 1684
},
{
"epoch": 0.6623427672955975,
"grad_norm": 0.7947145700454712,
"learning_rate": 1.690251572327044e-06,
"loss": 0.741,
"step": 1685
},
{
"epoch": 0.6627358490566038,
"grad_norm": 0.7595776319503784,
"learning_rate": 1.6882861635220127e-06,
"loss": 0.675,
"step": 1686
},
{
"epoch": 0.66312893081761,
"grad_norm": 0.8239551186561584,
"learning_rate": 1.6863207547169814e-06,
"loss": 0.7747,
"step": 1687
},
{
"epoch": 0.6635220125786163,
"grad_norm": 0.6555206775665283,
"learning_rate": 1.68435534591195e-06,
"loss": 0.715,
"step": 1688
},
{
"epoch": 0.6639150943396226,
"grad_norm": 0.8214390873908997,
"learning_rate": 1.6823899371069182e-06,
"loss": 0.7865,
"step": 1689
},
{
"epoch": 0.664308176100629,
"grad_norm": 0.7942861318588257,
"learning_rate": 1.6804245283018869e-06,
"loss": 0.7018,
"step": 1690
},
{
"epoch": 0.6647012578616353,
"grad_norm": 0.7687889337539673,
"learning_rate": 1.6784591194968555e-06,
"loss": 0.6691,
"step": 1691
},
{
"epoch": 0.6650943396226415,
"grad_norm": 0.6641785502433777,
"learning_rate": 1.6764937106918241e-06,
"loss": 0.5706,
"step": 1692
},
{
"epoch": 0.6654874213836478,
"grad_norm": 0.7239528894424438,
"learning_rate": 1.6745283018867925e-06,
"loss": 0.712,
"step": 1693
},
{
"epoch": 0.6658805031446541,
"grad_norm": 0.5714917778968811,
"learning_rate": 1.6725628930817612e-06,
"loss": 0.6542,
"step": 1694
},
{
"epoch": 0.6662735849056604,
"grad_norm": 0.6324212551116943,
"learning_rate": 1.6705974842767298e-06,
"loss": 0.6022,
"step": 1695
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.5100755095481873,
"learning_rate": 1.6686320754716985e-06,
"loss": 0.5698,
"step": 1696
},
{
"epoch": 0.6670597484276729,
"grad_norm": 0.6506298780441284,
"learning_rate": 1.6666666666666667e-06,
"loss": 0.6565,
"step": 1697
},
{
"epoch": 0.6674528301886793,
"grad_norm": 0.6662166118621826,
"learning_rate": 1.6647012578616353e-06,
"loss": 0.6974,
"step": 1698
},
{
"epoch": 0.6678459119496856,
"grad_norm": 0.7481600642204285,
"learning_rate": 1.662735849056604e-06,
"loss": 0.6661,
"step": 1699
},
{
"epoch": 0.6682389937106918,
"grad_norm": 0.8677677512168884,
"learning_rate": 1.6607704402515726e-06,
"loss": 0.744,
"step": 1700
},
{
"epoch": 0.6686320754716981,
"grad_norm": 0.6632601022720337,
"learning_rate": 1.658805031446541e-06,
"loss": 0.7023,
"step": 1701
},
{
"epoch": 0.6690251572327044,
"grad_norm": 0.7465344667434692,
"learning_rate": 1.6568396226415096e-06,
"loss": 0.6483,
"step": 1702
},
{
"epoch": 0.6694182389937107,
"grad_norm": 0.7378121614456177,
"learning_rate": 1.6548742138364783e-06,
"loss": 0.7611,
"step": 1703
},
{
"epoch": 0.6698113207547169,
"grad_norm": 0.9084758162498474,
"learning_rate": 1.652908805031447e-06,
"loss": 0.7727,
"step": 1704
},
{
"epoch": 0.6702044025157232,
"grad_norm": 0.5396172404289246,
"learning_rate": 1.650943396226415e-06,
"loss": 0.5927,
"step": 1705
},
{
"epoch": 0.6705974842767296,
"grad_norm": 0.7573910355567932,
"learning_rate": 1.6489779874213837e-06,
"loss": 0.6607,
"step": 1706
},
{
"epoch": 0.6709905660377359,
"grad_norm": 0.6944689154624939,
"learning_rate": 1.6470125786163524e-06,
"loss": 0.6987,
"step": 1707
},
{
"epoch": 0.6713836477987422,
"grad_norm": 0.725572407245636,
"learning_rate": 1.645047169811321e-06,
"loss": 0.6117,
"step": 1708
},
{
"epoch": 0.6717767295597484,
"grad_norm": 0.7425258755683899,
"learning_rate": 1.6430817610062894e-06,
"loss": 0.7283,
"step": 1709
},
{
"epoch": 0.6721698113207547,
"grad_norm": 0.5473070740699768,
"learning_rate": 1.641116352201258e-06,
"loss": 0.6611,
"step": 1710
},
{
"epoch": 0.672562893081761,
"grad_norm": 0.9759601354598999,
"learning_rate": 1.6391509433962267e-06,
"loss": 0.7078,
"step": 1711
},
{
"epoch": 0.6729559748427673,
"grad_norm": 0.8745030164718628,
"learning_rate": 1.6371855345911951e-06,
"loss": 0.7875,
"step": 1712
},
{
"epoch": 0.6733490566037735,
"grad_norm": 0.6132333874702454,
"learning_rate": 1.6352201257861635e-06,
"loss": 0.6193,
"step": 1713
},
{
"epoch": 0.6737421383647799,
"grad_norm": 0.828861653804779,
"learning_rate": 1.6332547169811322e-06,
"loss": 0.6427,
"step": 1714
},
{
"epoch": 0.6741352201257862,
"grad_norm": 0.7484422326087952,
"learning_rate": 1.6312893081761008e-06,
"loss": 0.738,
"step": 1715
},
{
"epoch": 0.6745283018867925,
"grad_norm": 0.7503340840339661,
"learning_rate": 1.6293238993710695e-06,
"loss": 0.7286,
"step": 1716
},
{
"epoch": 0.6749213836477987,
"grad_norm": 0.5788977742195129,
"learning_rate": 1.6273584905660379e-06,
"loss": 0.6227,
"step": 1717
},
{
"epoch": 0.675314465408805,
"grad_norm": 0.7129281163215637,
"learning_rate": 1.6253930817610065e-06,
"loss": 0.6765,
"step": 1718
},
{
"epoch": 0.6757075471698113,
"grad_norm": 1.2111996412277222,
"learning_rate": 1.623427672955975e-06,
"loss": 0.6887,
"step": 1719
},
{
"epoch": 0.6761006289308176,
"grad_norm": 0.7377118468284607,
"learning_rate": 1.6214622641509436e-06,
"loss": 0.6769,
"step": 1720
},
{
"epoch": 0.6764937106918238,
"grad_norm": 0.4261360466480255,
"learning_rate": 1.619496855345912e-06,
"loss": 0.5887,
"step": 1721
},
{
"epoch": 0.6768867924528302,
"grad_norm": 0.713239312171936,
"learning_rate": 1.6175314465408806e-06,
"loss": 0.663,
"step": 1722
},
{
"epoch": 0.6772798742138365,
"grad_norm": 0.8083055019378662,
"learning_rate": 1.6155660377358493e-06,
"loss": 0.814,
"step": 1723
},
{
"epoch": 0.6776729559748428,
"grad_norm": 0.5922679901123047,
"learning_rate": 1.613600628930818e-06,
"loss": 0.6478,
"step": 1724
},
{
"epoch": 0.6780660377358491,
"grad_norm": 0.5734134316444397,
"learning_rate": 1.6116352201257863e-06,
"loss": 0.6626,
"step": 1725
},
{
"epoch": 0.6784591194968553,
"grad_norm": 0.6487038135528564,
"learning_rate": 1.6096698113207547e-06,
"loss": 0.5848,
"step": 1726
},
{
"epoch": 0.6788522012578616,
"grad_norm": 0.871709406375885,
"learning_rate": 1.6077044025157234e-06,
"loss": 0.6948,
"step": 1727
},
{
"epoch": 0.6792452830188679,
"grad_norm": 0.8098554015159607,
"learning_rate": 1.605738993710692e-06,
"loss": 0.6694,
"step": 1728
},
{
"epoch": 0.6796383647798742,
"grad_norm": 0.6058477163314819,
"learning_rate": 1.6037735849056604e-06,
"loss": 0.6978,
"step": 1729
},
{
"epoch": 0.6800314465408805,
"grad_norm": 0.9011598825454712,
"learning_rate": 1.601808176100629e-06,
"loss": 0.7648,
"step": 1730
},
{
"epoch": 0.6804245283018868,
"grad_norm": 0.7559778690338135,
"learning_rate": 1.5998427672955977e-06,
"loss": 0.7104,
"step": 1731
},
{
"epoch": 0.6808176100628931,
"grad_norm": 0.598142683506012,
"learning_rate": 1.5978773584905663e-06,
"loss": 0.5718,
"step": 1732
},
{
"epoch": 0.6812106918238994,
"grad_norm": 0.6998093128204346,
"learning_rate": 1.5959119496855346e-06,
"loss": 0.7466,
"step": 1733
},
{
"epoch": 0.6816037735849056,
"grad_norm": 0.5081683397293091,
"learning_rate": 1.5939465408805032e-06,
"loss": 0.5868,
"step": 1734
},
{
"epoch": 0.6819968553459119,
"grad_norm": 0.6666524410247803,
"learning_rate": 1.5919811320754718e-06,
"loss": 0.6538,
"step": 1735
},
{
"epoch": 0.6823899371069182,
"grad_norm": 0.7543526887893677,
"learning_rate": 1.5900157232704405e-06,
"loss": 0.6761,
"step": 1736
},
{
"epoch": 0.6827830188679245,
"grad_norm": 0.9586549997329712,
"learning_rate": 1.5880503144654089e-06,
"loss": 0.8252,
"step": 1737
},
{
"epoch": 0.6831761006289309,
"grad_norm": 0.7068762183189392,
"learning_rate": 1.5860849056603775e-06,
"loss": 0.7317,
"step": 1738
},
{
"epoch": 0.6835691823899371,
"grad_norm": 0.7546913623809814,
"learning_rate": 1.5841194968553462e-06,
"loss": 0.7007,
"step": 1739
},
{
"epoch": 0.6839622641509434,
"grad_norm": 0.7934712767601013,
"learning_rate": 1.5821540880503148e-06,
"loss": 0.7111,
"step": 1740
},
{
"epoch": 0.6843553459119497,
"grad_norm": 0.6482877135276794,
"learning_rate": 1.580188679245283e-06,
"loss": 0.6329,
"step": 1741
},
{
"epoch": 0.684748427672956,
"grad_norm": 0.6041259169578552,
"learning_rate": 1.5782232704402516e-06,
"loss": 0.6381,
"step": 1742
},
{
"epoch": 0.6851415094339622,
"grad_norm": 0.7193888425827026,
"learning_rate": 1.5762578616352203e-06,
"loss": 0.6796,
"step": 1743
},
{
"epoch": 0.6855345911949685,
"grad_norm": 0.8158414363861084,
"learning_rate": 1.574292452830189e-06,
"loss": 0.6726,
"step": 1744
},
{
"epoch": 0.6859276729559748,
"grad_norm": 0.7104851007461548,
"learning_rate": 1.5723270440251573e-06,
"loss": 0.7336,
"step": 1745
},
{
"epoch": 0.6863207547169812,
"grad_norm": 0.6518145799636841,
"learning_rate": 1.570361635220126e-06,
"loss": 0.671,
"step": 1746
},
{
"epoch": 0.6867138364779874,
"grad_norm": 0.8247858881950378,
"learning_rate": 1.5683962264150946e-06,
"loss": 0.7484,
"step": 1747
},
{
"epoch": 0.6871069182389937,
"grad_norm": 0.9963441491127014,
"learning_rate": 1.5664308176100632e-06,
"loss": 0.8379,
"step": 1748
},
{
"epoch": 0.6875,
"grad_norm": 0.671656608581543,
"learning_rate": 1.5644654088050314e-06,
"loss": 0.6399,
"step": 1749
},
{
"epoch": 0.6878930817610063,
"grad_norm": 0.6859222650527954,
"learning_rate": 1.5625e-06,
"loss": 0.7087,
"step": 1750
},
{
"epoch": 0.6882861635220126,
"grad_norm": 0.8261825442314148,
"learning_rate": 1.5605345911949687e-06,
"loss": 0.7408,
"step": 1751
},
{
"epoch": 0.6886792452830188,
"grad_norm": 0.7971336841583252,
"learning_rate": 1.5585691823899371e-06,
"loss": 0.7961,
"step": 1752
},
{
"epoch": 0.6890723270440252,
"grad_norm": 0.5762659907341003,
"learning_rate": 1.5566037735849058e-06,
"loss": 0.6201,
"step": 1753
},
{
"epoch": 0.6894654088050315,
"grad_norm": 0.6793798208236694,
"learning_rate": 1.5546383647798744e-06,
"loss": 0.6935,
"step": 1754
},
{
"epoch": 0.6898584905660378,
"grad_norm": 0.69764643907547,
"learning_rate": 1.552672955974843e-06,
"loss": 0.6383,
"step": 1755
},
{
"epoch": 0.690251572327044,
"grad_norm": 0.7773534655570984,
"learning_rate": 1.5507075471698112e-06,
"loss": 0.7312,
"step": 1756
},
{
"epoch": 0.6906446540880503,
"grad_norm": 0.7948174476623535,
"learning_rate": 1.5487421383647799e-06,
"loss": 0.7316,
"step": 1757
},
{
"epoch": 0.6910377358490566,
"grad_norm": 0.8953741788864136,
"learning_rate": 1.5467767295597485e-06,
"loss": 0.7494,
"step": 1758
},
{
"epoch": 0.6914308176100629,
"grad_norm": 0.8447637557983398,
"learning_rate": 1.5448113207547172e-06,
"loss": 0.8018,
"step": 1759
},
{
"epoch": 0.6918238993710691,
"grad_norm": 0.6026454567909241,
"learning_rate": 1.5428459119496856e-06,
"loss": 0.6111,
"step": 1760
},
{
"epoch": 0.6922169811320755,
"grad_norm": 0.6515069007873535,
"learning_rate": 1.5408805031446542e-06,
"loss": 0.6167,
"step": 1761
},
{
"epoch": 0.6926100628930818,
"grad_norm": 0.6022630929946899,
"learning_rate": 1.5389150943396228e-06,
"loss": 0.6743,
"step": 1762
},
{
"epoch": 0.6930031446540881,
"grad_norm": 0.6207025051116943,
"learning_rate": 1.5369496855345915e-06,
"loss": 0.6423,
"step": 1763
},
{
"epoch": 0.6933962264150944,
"grad_norm": 0.7039181590080261,
"learning_rate": 1.5349842767295597e-06,
"loss": 0.6925,
"step": 1764
},
{
"epoch": 0.6937893081761006,
"grad_norm": 0.8616054654121399,
"learning_rate": 1.5330188679245283e-06,
"loss": 0.7946,
"step": 1765
},
{
"epoch": 0.6941823899371069,
"grad_norm": 0.611029863357544,
"learning_rate": 1.531053459119497e-06,
"loss": 0.6359,
"step": 1766
},
{
"epoch": 0.6945754716981132,
"grad_norm": 0.6635613441467285,
"learning_rate": 1.5290880503144656e-06,
"loss": 0.7223,
"step": 1767
},
{
"epoch": 0.6949685534591195,
"grad_norm": 0.5227457284927368,
"learning_rate": 1.527122641509434e-06,
"loss": 0.637,
"step": 1768
},
{
"epoch": 0.6953616352201258,
"grad_norm": 0.6783047318458557,
"learning_rate": 1.5251572327044027e-06,
"loss": 0.7236,
"step": 1769
},
{
"epoch": 0.6957547169811321,
"grad_norm": 0.6703030467033386,
"learning_rate": 1.5231918238993713e-06,
"loss": 0.5561,
"step": 1770
},
{
"epoch": 0.6961477987421384,
"grad_norm": 0.726677656173706,
"learning_rate": 1.52122641509434e-06,
"loss": 0.6362,
"step": 1771
},
{
"epoch": 0.6965408805031447,
"grad_norm": 0.5254794359207153,
"learning_rate": 1.5192610062893081e-06,
"loss": 0.5763,
"step": 1772
},
{
"epoch": 0.6969339622641509,
"grad_norm": 0.6154314279556274,
"learning_rate": 1.5172955974842768e-06,
"loss": 0.6396,
"step": 1773
},
{
"epoch": 0.6973270440251572,
"grad_norm": 0.6501309275627136,
"learning_rate": 1.5153301886792454e-06,
"loss": 0.662,
"step": 1774
},
{
"epoch": 0.6977201257861635,
"grad_norm": 0.7189278602600098,
"learning_rate": 1.513364779874214e-06,
"loss": 0.6595,
"step": 1775
},
{
"epoch": 0.6981132075471698,
"grad_norm": 0.509173572063446,
"learning_rate": 1.5113993710691825e-06,
"loss": 0.6501,
"step": 1776
},
{
"epoch": 0.6985062893081762,
"grad_norm": 0.7481945157051086,
"learning_rate": 1.509433962264151e-06,
"loss": 0.6827,
"step": 1777
},
{
"epoch": 0.6988993710691824,
"grad_norm": 0.7772241830825806,
"learning_rate": 1.5074685534591197e-06,
"loss": 0.646,
"step": 1778
},
{
"epoch": 0.6992924528301887,
"grad_norm": 0.6089989542961121,
"learning_rate": 1.5055031446540882e-06,
"loss": 0.6741,
"step": 1779
},
{
"epoch": 0.699685534591195,
"grad_norm": 0.5924184322357178,
"learning_rate": 1.5035377358490566e-06,
"loss": 0.6832,
"step": 1780
},
{
"epoch": 0.7000786163522013,
"grad_norm": 0.7790172100067139,
"learning_rate": 1.5015723270440252e-06,
"loss": 0.6581,
"step": 1781
},
{
"epoch": 0.7004716981132075,
"grad_norm": 0.8168909549713135,
"learning_rate": 1.4996069182389938e-06,
"loss": 0.6469,
"step": 1782
},
{
"epoch": 0.7008647798742138,
"grad_norm": 0.914035439491272,
"learning_rate": 1.4976415094339625e-06,
"loss": 0.781,
"step": 1783
},
{
"epoch": 0.7012578616352201,
"grad_norm": 0.6542779207229614,
"learning_rate": 1.495676100628931e-06,
"loss": 0.6917,
"step": 1784
},
{
"epoch": 0.7016509433962265,
"grad_norm": 0.7413952946662903,
"learning_rate": 1.4937106918238995e-06,
"loss": 0.7024,
"step": 1785
},
{
"epoch": 0.7020440251572327,
"grad_norm": 1.1687352657318115,
"learning_rate": 1.491745283018868e-06,
"loss": 0.8041,
"step": 1786
},
{
"epoch": 0.702437106918239,
"grad_norm": 0.9856542944908142,
"learning_rate": 1.4897798742138366e-06,
"loss": 0.7874,
"step": 1787
},
{
"epoch": 0.7028301886792453,
"grad_norm": 0.6561897993087769,
"learning_rate": 1.487814465408805e-06,
"loss": 0.6766,
"step": 1788
},
{
"epoch": 0.7032232704402516,
"grad_norm": 0.591860830783844,
"learning_rate": 1.4858490566037737e-06,
"loss": 0.6431,
"step": 1789
},
{
"epoch": 0.7036163522012578,
"grad_norm": 0.626357913017273,
"learning_rate": 1.4838836477987423e-06,
"loss": 0.6461,
"step": 1790
},
{
"epoch": 0.7040094339622641,
"grad_norm": 0.678950309753418,
"learning_rate": 1.481918238993711e-06,
"loss": 0.5908,
"step": 1791
},
{
"epoch": 0.7044025157232704,
"grad_norm": 0.6452150344848633,
"learning_rate": 1.4799528301886794e-06,
"loss": 0.6182,
"step": 1792
},
{
"epoch": 0.7047955974842768,
"grad_norm": 0.9632670283317566,
"learning_rate": 1.4779874213836478e-06,
"loss": 0.7323,
"step": 1793
},
{
"epoch": 0.7051886792452831,
"grad_norm": 0.6055132150650024,
"learning_rate": 1.4760220125786164e-06,
"loss": 0.6887,
"step": 1794
},
{
"epoch": 0.7055817610062893,
"grad_norm": 0.8240327835083008,
"learning_rate": 1.474056603773585e-06,
"loss": 0.7844,
"step": 1795
},
{
"epoch": 0.7059748427672956,
"grad_norm": 0.8026304244995117,
"learning_rate": 1.4720911949685535e-06,
"loss": 0.7681,
"step": 1796
},
{
"epoch": 0.7063679245283019,
"grad_norm": 0.6698837876319885,
"learning_rate": 1.470125786163522e-06,
"loss": 0.6862,
"step": 1797
},
{
"epoch": 0.7067610062893082,
"grad_norm": 0.5456218123435974,
"learning_rate": 1.4681603773584907e-06,
"loss": 0.5955,
"step": 1798
},
{
"epoch": 0.7071540880503144,
"grad_norm": 0.43243128061294556,
"learning_rate": 1.4661949685534594e-06,
"loss": 0.6067,
"step": 1799
},
{
"epoch": 0.7075471698113207,
"grad_norm": 0.7796711325645447,
"learning_rate": 1.4642295597484276e-06,
"loss": 0.7892,
"step": 1800
},
{
"epoch": 0.7079402515723271,
"grad_norm": 0.6946038007736206,
"learning_rate": 1.4622641509433962e-06,
"loss": 0.7192,
"step": 1801
},
{
"epoch": 0.7083333333333334,
"grad_norm": 0.7358932495117188,
"learning_rate": 1.4602987421383649e-06,
"loss": 0.7007,
"step": 1802
},
{
"epoch": 0.7087264150943396,
"grad_norm": 0.608268141746521,
"learning_rate": 1.4583333333333335e-06,
"loss": 0.7635,
"step": 1803
},
{
"epoch": 0.7091194968553459,
"grad_norm": 0.836534321308136,
"learning_rate": 1.456367924528302e-06,
"loss": 0.7502,
"step": 1804
},
{
"epoch": 0.7095125786163522,
"grad_norm": 0.4456639289855957,
"learning_rate": 1.4544025157232705e-06,
"loss": 0.6165,
"step": 1805
},
{
"epoch": 0.7099056603773585,
"grad_norm": 0.7184557318687439,
"learning_rate": 1.4524371069182392e-06,
"loss": 0.6708,
"step": 1806
},
{
"epoch": 0.7102987421383647,
"grad_norm": 0.938624918460846,
"learning_rate": 1.4504716981132078e-06,
"loss": 0.7235,
"step": 1807
},
{
"epoch": 0.710691823899371,
"grad_norm": 0.6916636228561401,
"learning_rate": 1.448506289308176e-06,
"loss": 0.7478,
"step": 1808
},
{
"epoch": 0.7110849056603774,
"grad_norm": 0.6670815944671631,
"learning_rate": 1.4465408805031447e-06,
"loss": 0.7053,
"step": 1809
},
{
"epoch": 0.7114779874213837,
"grad_norm": 0.5720268487930298,
"learning_rate": 1.4445754716981133e-06,
"loss": 0.6503,
"step": 1810
},
{
"epoch": 0.71187106918239,
"grad_norm": 0.7414019107818604,
"learning_rate": 1.442610062893082e-06,
"loss": 0.7367,
"step": 1811
},
{
"epoch": 0.7122641509433962,
"grad_norm": 0.7042202353477478,
"learning_rate": 1.4406446540880504e-06,
"loss": 0.738,
"step": 1812
},
{
"epoch": 0.7126572327044025,
"grad_norm": 0.770480751991272,
"learning_rate": 1.438679245283019e-06,
"loss": 0.7515,
"step": 1813
},
{
"epoch": 0.7130503144654088,
"grad_norm": 0.630929172039032,
"learning_rate": 1.4367138364779876e-06,
"loss": 0.712,
"step": 1814
},
{
"epoch": 0.7134433962264151,
"grad_norm": 0.6110004782676697,
"learning_rate": 1.4347484276729563e-06,
"loss": 0.5542,
"step": 1815
},
{
"epoch": 0.7138364779874213,
"grad_norm": 0.7706511616706848,
"learning_rate": 1.4327830188679245e-06,
"loss": 0.744,
"step": 1816
},
{
"epoch": 0.7142295597484277,
"grad_norm": 0.720784068107605,
"learning_rate": 1.430817610062893e-06,
"loss": 0.7125,
"step": 1817
},
{
"epoch": 0.714622641509434,
"grad_norm": 0.6888326406478882,
"learning_rate": 1.4288522012578617e-06,
"loss": 0.6112,
"step": 1818
},
{
"epoch": 0.7150157232704403,
"grad_norm": 0.6630930304527283,
"learning_rate": 1.4268867924528304e-06,
"loss": 0.6897,
"step": 1819
},
{
"epoch": 0.7154088050314465,
"grad_norm": 0.648844301700592,
"learning_rate": 1.4249213836477988e-06,
"loss": 0.7124,
"step": 1820
},
{
"epoch": 0.7158018867924528,
"grad_norm": 0.6940709948539734,
"learning_rate": 1.4229559748427674e-06,
"loss": 0.7092,
"step": 1821
},
{
"epoch": 0.7161949685534591,
"grad_norm": 0.5596283078193665,
"learning_rate": 1.420990566037736e-06,
"loss": 0.6028,
"step": 1822
},
{
"epoch": 0.7165880503144654,
"grad_norm": 0.8014578819274902,
"learning_rate": 1.4190251572327047e-06,
"loss": 0.8148,
"step": 1823
},
{
"epoch": 0.7169811320754716,
"grad_norm": 0.850028932094574,
"learning_rate": 1.417059748427673e-06,
"loss": 0.7615,
"step": 1824
},
{
"epoch": 0.717374213836478,
"grad_norm": 0.6816880106925964,
"learning_rate": 1.4150943396226415e-06,
"loss": 0.6584,
"step": 1825
},
{
"epoch": 0.7177672955974843,
"grad_norm": 0.8313127160072327,
"learning_rate": 1.4131289308176102e-06,
"loss": 0.8342,
"step": 1826
},
{
"epoch": 0.7181603773584906,
"grad_norm": 0.6028608679771423,
"learning_rate": 1.4111635220125788e-06,
"loss": 0.577,
"step": 1827
},
{
"epoch": 0.7185534591194969,
"grad_norm": 0.6804369688034058,
"learning_rate": 1.4091981132075472e-06,
"loss": 0.6547,
"step": 1828
},
{
"epoch": 0.7189465408805031,
"grad_norm": 0.6613967418670654,
"learning_rate": 1.4072327044025159e-06,
"loss": 0.7059,
"step": 1829
},
{
"epoch": 0.7193396226415094,
"grad_norm": 0.6942279934883118,
"learning_rate": 1.4052672955974845e-06,
"loss": 0.6829,
"step": 1830
},
{
"epoch": 0.7197327044025157,
"grad_norm": 0.74433833360672,
"learning_rate": 1.4033018867924531e-06,
"loss": 0.7944,
"step": 1831
},
{
"epoch": 0.720125786163522,
"grad_norm": 0.6830068230628967,
"learning_rate": 1.4013364779874214e-06,
"loss": 0.6505,
"step": 1832
},
{
"epoch": 0.7205188679245284,
"grad_norm": 0.7548493146896362,
"learning_rate": 1.39937106918239e-06,
"loss": 0.6656,
"step": 1833
},
{
"epoch": 0.7209119496855346,
"grad_norm": 0.7680490612983704,
"learning_rate": 1.3974056603773586e-06,
"loss": 0.6458,
"step": 1834
},
{
"epoch": 0.7213050314465409,
"grad_norm": 0.7552528977394104,
"learning_rate": 1.3954402515723273e-06,
"loss": 0.7192,
"step": 1835
},
{
"epoch": 0.7216981132075472,
"grad_norm": 0.8983709216117859,
"learning_rate": 1.3934748427672957e-06,
"loss": 0.7223,
"step": 1836
},
{
"epoch": 0.7220911949685535,
"grad_norm": 0.7910611629486084,
"learning_rate": 1.3915094339622643e-06,
"loss": 0.7153,
"step": 1837
},
{
"epoch": 0.7224842767295597,
"grad_norm": 0.7908303141593933,
"learning_rate": 1.389544025157233e-06,
"loss": 0.7349,
"step": 1838
},
{
"epoch": 0.722877358490566,
"grad_norm": 0.6163021326065063,
"learning_rate": 1.3875786163522014e-06,
"loss": 0.5848,
"step": 1839
},
{
"epoch": 0.7232704402515723,
"grad_norm": 0.5950599312782288,
"learning_rate": 1.3856132075471698e-06,
"loss": 0.6601,
"step": 1840
},
{
"epoch": 0.7236635220125787,
"grad_norm": 0.8659715056419373,
"learning_rate": 1.3836477987421384e-06,
"loss": 0.6924,
"step": 1841
},
{
"epoch": 0.7240566037735849,
"grad_norm": 0.8773351311683655,
"learning_rate": 1.381682389937107e-06,
"loss": 0.7169,
"step": 1842
},
{
"epoch": 0.7244496855345912,
"grad_norm": 0.9375039339065552,
"learning_rate": 1.3797169811320757e-06,
"loss": 0.7898,
"step": 1843
},
{
"epoch": 0.7248427672955975,
"grad_norm": 1.2229951620101929,
"learning_rate": 1.3777515723270441e-06,
"loss": 0.7186,
"step": 1844
},
{
"epoch": 0.7252358490566038,
"grad_norm": 0.6375990509986877,
"learning_rate": 1.3757861635220128e-06,
"loss": 0.5321,
"step": 1845
},
{
"epoch": 0.72562893081761,
"grad_norm": 0.6560855507850647,
"learning_rate": 1.3738207547169812e-06,
"loss": 0.6099,
"step": 1846
},
{
"epoch": 0.7260220125786163,
"grad_norm": 0.7304661273956299,
"learning_rate": 1.3718553459119498e-06,
"loss": 0.7128,
"step": 1847
},
{
"epoch": 0.7264150943396226,
"grad_norm": 0.6928592920303345,
"learning_rate": 1.3698899371069182e-06,
"loss": 0.6547,
"step": 1848
},
{
"epoch": 0.726808176100629,
"grad_norm": 0.7003422975540161,
"learning_rate": 1.3679245283018869e-06,
"loss": 0.7362,
"step": 1849
},
{
"epoch": 0.7272012578616353,
"grad_norm": 0.7021567225456238,
"learning_rate": 1.3659591194968555e-06,
"loss": 0.6283,
"step": 1850
},
{
"epoch": 0.7275943396226415,
"grad_norm": 0.6775747537612915,
"learning_rate": 1.3639937106918242e-06,
"loss": 0.6058,
"step": 1851
},
{
"epoch": 0.7279874213836478,
"grad_norm": 0.6340211033821106,
"learning_rate": 1.3620283018867926e-06,
"loss": 0.6149,
"step": 1852
},
{
"epoch": 0.7283805031446541,
"grad_norm": 0.7404446601867676,
"learning_rate": 1.360062893081761e-06,
"loss": 0.7413,
"step": 1853
},
{
"epoch": 0.7287735849056604,
"grad_norm": 0.5350498557090759,
"learning_rate": 1.3580974842767296e-06,
"loss": 0.562,
"step": 1854
},
{
"epoch": 0.7291666666666666,
"grad_norm": 0.812415361404419,
"learning_rate": 1.3561320754716983e-06,
"loss": 0.6958,
"step": 1855
},
{
"epoch": 0.7295597484276729,
"grad_norm": 0.6135343313217163,
"learning_rate": 1.3541666666666667e-06,
"loss": 0.6025,
"step": 1856
},
{
"epoch": 0.7299528301886793,
"grad_norm": 0.9421678185462952,
"learning_rate": 1.3522012578616353e-06,
"loss": 0.8228,
"step": 1857
},
{
"epoch": 0.7303459119496856,
"grad_norm": 0.6760720014572144,
"learning_rate": 1.350235849056604e-06,
"loss": 0.7111,
"step": 1858
},
{
"epoch": 0.7307389937106918,
"grad_norm": 0.7153463363647461,
"learning_rate": 1.3482704402515726e-06,
"loss": 0.7172,
"step": 1859
},
{
"epoch": 0.7311320754716981,
"grad_norm": 0.6221895217895508,
"learning_rate": 1.3463050314465408e-06,
"loss": 0.5794,
"step": 1860
},
{
"epoch": 0.7315251572327044,
"grad_norm": 0.79213547706604,
"learning_rate": 1.3443396226415094e-06,
"loss": 0.6885,
"step": 1861
},
{
"epoch": 0.7319182389937107,
"grad_norm": 0.9507853984832764,
"learning_rate": 1.342374213836478e-06,
"loss": 0.7891,
"step": 1862
},
{
"epoch": 0.7323113207547169,
"grad_norm": 0.857855498790741,
"learning_rate": 1.3404088050314467e-06,
"loss": 0.7084,
"step": 1863
},
{
"epoch": 0.7327044025157232,
"grad_norm": 0.6300313472747803,
"learning_rate": 1.3384433962264151e-06,
"loss": 0.6774,
"step": 1864
},
{
"epoch": 0.7330974842767296,
"grad_norm": 0.640471339225769,
"learning_rate": 1.3364779874213838e-06,
"loss": 0.6357,
"step": 1865
},
{
"epoch": 0.7334905660377359,
"grad_norm": 0.602178692817688,
"learning_rate": 1.3345125786163524e-06,
"loss": 0.6094,
"step": 1866
},
{
"epoch": 0.7338836477987422,
"grad_norm": 0.6933202743530273,
"learning_rate": 1.332547169811321e-06,
"loss": 0.7747,
"step": 1867
},
{
"epoch": 0.7342767295597484,
"grad_norm": 0.6867457628250122,
"learning_rate": 1.3305817610062892e-06,
"loss": 0.6436,
"step": 1868
},
{
"epoch": 0.7346698113207547,
"grad_norm": 0.8128904700279236,
"learning_rate": 1.3286163522012579e-06,
"loss": 0.719,
"step": 1869
},
{
"epoch": 0.735062893081761,
"grad_norm": 0.6408858299255371,
"learning_rate": 1.3266509433962265e-06,
"loss": 0.5786,
"step": 1870
},
{
"epoch": 0.7354559748427673,
"grad_norm": 0.7060768604278564,
"learning_rate": 1.3246855345911952e-06,
"loss": 0.6594,
"step": 1871
},
{
"epoch": 0.7358490566037735,
"grad_norm": 0.698584794998169,
"learning_rate": 1.3227201257861636e-06,
"loss": 0.5967,
"step": 1872
},
{
"epoch": 0.7362421383647799,
"grad_norm": 0.9708681106567383,
"learning_rate": 1.3207547169811322e-06,
"loss": 0.707,
"step": 1873
},
{
"epoch": 0.7366352201257862,
"grad_norm": 0.6745409369468689,
"learning_rate": 1.3187893081761008e-06,
"loss": 0.6593,
"step": 1874
},
{
"epoch": 0.7370283018867925,
"grad_norm": 0.7829700708389282,
"learning_rate": 1.3168238993710695e-06,
"loss": 0.6497,
"step": 1875
},
{
"epoch": 0.7374213836477987,
"grad_norm": 0.7159058451652527,
"learning_rate": 1.3148584905660377e-06,
"loss": 0.6358,
"step": 1876
},
{
"epoch": 0.737814465408805,
"grad_norm": 0.7407160401344299,
"learning_rate": 1.3128930817610063e-06,
"loss": 0.6467,
"step": 1877
},
{
"epoch": 0.7382075471698113,
"grad_norm": 0.8269376158714294,
"learning_rate": 1.310927672955975e-06,
"loss": 0.8651,
"step": 1878
},
{
"epoch": 0.7386006289308176,
"grad_norm": 0.8148836493492126,
"learning_rate": 1.3089622641509436e-06,
"loss": 0.7434,
"step": 1879
},
{
"epoch": 0.7389937106918238,
"grad_norm": 0.9783030152320862,
"learning_rate": 1.306996855345912e-06,
"loss": 0.778,
"step": 1880
},
{
"epoch": 0.7393867924528302,
"grad_norm": 0.6194635033607483,
"learning_rate": 1.3050314465408807e-06,
"loss": 0.6069,
"step": 1881
},
{
"epoch": 0.7397798742138365,
"grad_norm": 0.6858715415000916,
"learning_rate": 1.3030660377358493e-06,
"loss": 0.6997,
"step": 1882
},
{
"epoch": 0.7401729559748428,
"grad_norm": 0.6410948038101196,
"learning_rate": 1.301100628930818e-06,
"loss": 0.6923,
"step": 1883
},
{
"epoch": 0.7405660377358491,
"grad_norm": 0.7050250172615051,
"learning_rate": 1.2991352201257861e-06,
"loss": 0.6758,
"step": 1884
},
{
"epoch": 0.7409591194968553,
"grad_norm": 0.9802526235580444,
"learning_rate": 1.2971698113207548e-06,
"loss": 0.8313,
"step": 1885
},
{
"epoch": 0.7413522012578616,
"grad_norm": 0.6277905106544495,
"learning_rate": 1.2952044025157234e-06,
"loss": 0.6329,
"step": 1886
},
{
"epoch": 0.7417452830188679,
"grad_norm": 0.8869853019714355,
"learning_rate": 1.293238993710692e-06,
"loss": 0.8237,
"step": 1887
},
{
"epoch": 0.7421383647798742,
"grad_norm": 0.5675129294395447,
"learning_rate": 1.2912735849056605e-06,
"loss": 0.5823,
"step": 1888
},
{
"epoch": 0.7425314465408805,
"grad_norm": 0.7815062403678894,
"learning_rate": 1.289308176100629e-06,
"loss": 0.799,
"step": 1889
},
{
"epoch": 0.7429245283018868,
"grad_norm": 0.6720836758613586,
"learning_rate": 1.2873427672955977e-06,
"loss": 0.6487,
"step": 1890
},
{
"epoch": 0.7433176100628931,
"grad_norm": 0.7089105844497681,
"learning_rate": 1.2853773584905664e-06,
"loss": 0.7195,
"step": 1891
},
{
"epoch": 0.7437106918238994,
"grad_norm": 0.8437160849571228,
"learning_rate": 1.2834119496855346e-06,
"loss": 0.7736,
"step": 1892
},
{
"epoch": 0.7441037735849056,
"grad_norm": 0.6545000672340393,
"learning_rate": 1.2814465408805032e-06,
"loss": 0.7206,
"step": 1893
},
{
"epoch": 0.7444968553459119,
"grad_norm": 0.8676877021789551,
"learning_rate": 1.2794811320754718e-06,
"loss": 0.7003,
"step": 1894
},
{
"epoch": 0.7448899371069182,
"grad_norm": 0.6981176137924194,
"learning_rate": 1.2775157232704405e-06,
"loss": 0.5798,
"step": 1895
},
{
"epoch": 0.7452830188679245,
"grad_norm": 0.8167281150817871,
"learning_rate": 1.275550314465409e-06,
"loss": 0.6586,
"step": 1896
},
{
"epoch": 0.7456761006289309,
"grad_norm": 0.5205354690551758,
"learning_rate": 1.2735849056603775e-06,
"loss": 0.6101,
"step": 1897
},
{
"epoch": 0.7460691823899371,
"grad_norm": 0.560174286365509,
"learning_rate": 1.2716194968553462e-06,
"loss": 0.6127,
"step": 1898
},
{
"epoch": 0.7464622641509434,
"grad_norm": 0.7096153497695923,
"learning_rate": 1.2696540880503146e-06,
"loss": 0.7173,
"step": 1899
},
{
"epoch": 0.7468553459119497,
"grad_norm": 0.8807818293571472,
"learning_rate": 1.267688679245283e-06,
"loss": 0.8134,
"step": 1900
},
{
"epoch": 0.747248427672956,
"grad_norm": 0.6683316230773926,
"learning_rate": 1.2657232704402517e-06,
"loss": 0.6746,
"step": 1901
},
{
"epoch": 0.7476415094339622,
"grad_norm": 0.7726778984069824,
"learning_rate": 1.2637578616352203e-06,
"loss": 0.7173,
"step": 1902
},
{
"epoch": 0.7480345911949685,
"grad_norm": 0.6927328705787659,
"learning_rate": 1.261792452830189e-06,
"loss": 0.7656,
"step": 1903
},
{
"epoch": 0.7484276729559748,
"grad_norm": 0.8003778457641602,
"learning_rate": 1.2598270440251574e-06,
"loss": 0.7128,
"step": 1904
},
{
"epoch": 0.7488207547169812,
"grad_norm": 0.6433470845222473,
"learning_rate": 1.257861635220126e-06,
"loss": 0.6326,
"step": 1905
},
{
"epoch": 0.7492138364779874,
"grad_norm": 0.6671479344367981,
"learning_rate": 1.2558962264150944e-06,
"loss": 0.7391,
"step": 1906
},
{
"epoch": 0.7496069182389937,
"grad_norm": 0.6235984563827515,
"learning_rate": 1.253930817610063e-06,
"loss": 0.6649,
"step": 1907
},
{
"epoch": 0.75,
"grad_norm": 0.6084699034690857,
"learning_rate": 1.2519654088050315e-06,
"loss": 0.5842,
"step": 1908
},
{
"epoch": 0.7503930817610063,
"grad_norm": 0.8485831022262573,
"learning_rate": 1.25e-06,
"loss": 0.7916,
"step": 1909
},
{
"epoch": 0.7507861635220126,
"grad_norm": 0.8693320751190186,
"learning_rate": 1.2480345911949685e-06,
"loss": 0.7803,
"step": 1910
},
{
"epoch": 0.7511792452830188,
"grad_norm": 0.6483492255210876,
"learning_rate": 1.2460691823899372e-06,
"loss": 0.6133,
"step": 1911
},
{
"epoch": 0.7515723270440252,
"grad_norm": 0.6764208078384399,
"learning_rate": 1.2441037735849058e-06,
"loss": 0.6215,
"step": 1912
},
{
"epoch": 0.7519654088050315,
"grad_norm": 0.593447744846344,
"learning_rate": 1.2421383647798742e-06,
"loss": 0.64,
"step": 1913
},
{
"epoch": 0.7523584905660378,
"grad_norm": 0.6669500470161438,
"learning_rate": 1.2401729559748429e-06,
"loss": 0.6661,
"step": 1914
},
{
"epoch": 0.752751572327044,
"grad_norm": 0.7398116588592529,
"learning_rate": 1.2382075471698115e-06,
"loss": 0.7543,
"step": 1915
},
{
"epoch": 0.7531446540880503,
"grad_norm": 0.7229875326156616,
"learning_rate": 1.23624213836478e-06,
"loss": 0.7341,
"step": 1916
},
{
"epoch": 0.7535377358490566,
"grad_norm": 0.8301966786384583,
"learning_rate": 1.2342767295597485e-06,
"loss": 0.7175,
"step": 1917
},
{
"epoch": 0.7539308176100629,
"grad_norm": 0.7797653079032898,
"learning_rate": 1.232311320754717e-06,
"loss": 0.7472,
"step": 1918
},
{
"epoch": 0.7543238993710691,
"grad_norm": 0.7501053214073181,
"learning_rate": 1.2303459119496856e-06,
"loss": 0.6782,
"step": 1919
},
{
"epoch": 0.7547169811320755,
"grad_norm": 0.5579621195793152,
"learning_rate": 1.228380503144654e-06,
"loss": 0.6215,
"step": 1920
},
{
"epoch": 0.7551100628930818,
"grad_norm": 0.742179274559021,
"learning_rate": 1.2264150943396227e-06,
"loss": 0.6982,
"step": 1921
},
{
"epoch": 0.7555031446540881,
"grad_norm": 0.6128301620483398,
"learning_rate": 1.2244496855345913e-06,
"loss": 0.62,
"step": 1922
},
{
"epoch": 0.7558962264150944,
"grad_norm": 0.7548709511756897,
"learning_rate": 1.22248427672956e-06,
"loss": 0.7231,
"step": 1923
},
{
"epoch": 0.7562893081761006,
"grad_norm": 0.7941972017288208,
"learning_rate": 1.2205188679245284e-06,
"loss": 0.7727,
"step": 1924
},
{
"epoch": 0.7566823899371069,
"grad_norm": 0.926997721195221,
"learning_rate": 1.218553459119497e-06,
"loss": 0.8615,
"step": 1925
},
{
"epoch": 0.7570754716981132,
"grad_norm": 0.7232053279876709,
"learning_rate": 1.2165880503144654e-06,
"loss": 0.6814,
"step": 1926
},
{
"epoch": 0.7574685534591195,
"grad_norm": 0.8009117841720581,
"learning_rate": 1.214622641509434e-06,
"loss": 0.7476,
"step": 1927
},
{
"epoch": 0.7578616352201258,
"grad_norm": 0.9014382362365723,
"learning_rate": 1.2126572327044025e-06,
"loss": 0.7098,
"step": 1928
},
{
"epoch": 0.7582547169811321,
"grad_norm": 0.933290421962738,
"learning_rate": 1.210691823899371e-06,
"loss": 0.83,
"step": 1929
},
{
"epoch": 0.7586477987421384,
"grad_norm": 0.967338502407074,
"learning_rate": 1.2087264150943397e-06,
"loss": 0.8245,
"step": 1930
},
{
"epoch": 0.7590408805031447,
"grad_norm": 0.8377425074577332,
"learning_rate": 1.2067610062893084e-06,
"loss": 0.8229,
"step": 1931
},
{
"epoch": 0.7594339622641509,
"grad_norm": 0.6549360752105713,
"learning_rate": 1.2047955974842768e-06,
"loss": 0.6613,
"step": 1932
},
{
"epoch": 0.7598270440251572,
"grad_norm": 0.5508908033370972,
"learning_rate": 1.2028301886792454e-06,
"loss": 0.577,
"step": 1933
},
{
"epoch": 0.7602201257861635,
"grad_norm": 0.6499086022377014,
"learning_rate": 1.2008647798742139e-06,
"loss": 0.7998,
"step": 1934
},
{
"epoch": 0.7606132075471698,
"grad_norm": 0.5735564827919006,
"learning_rate": 1.1988993710691825e-06,
"loss": 0.6178,
"step": 1935
},
{
"epoch": 0.7610062893081762,
"grad_norm": 0.6505428552627563,
"learning_rate": 1.196933962264151e-06,
"loss": 0.6472,
"step": 1936
},
{
"epoch": 0.7613993710691824,
"grad_norm": 0.7224058508872986,
"learning_rate": 1.1949685534591195e-06,
"loss": 0.6907,
"step": 1937
},
{
"epoch": 0.7617924528301887,
"grad_norm": 0.7389016151428223,
"learning_rate": 1.1930031446540882e-06,
"loss": 0.7499,
"step": 1938
},
{
"epoch": 0.762185534591195,
"grad_norm": 0.6133378148078918,
"learning_rate": 1.1910377358490568e-06,
"loss": 0.6049,
"step": 1939
},
{
"epoch": 0.7625786163522013,
"grad_norm": 0.6055359244346619,
"learning_rate": 1.1890723270440252e-06,
"loss": 0.6544,
"step": 1940
},
{
"epoch": 0.7629716981132075,
"grad_norm": 0.9314672946929932,
"learning_rate": 1.1871069182389939e-06,
"loss": 0.7281,
"step": 1941
},
{
"epoch": 0.7633647798742138,
"grad_norm": 0.7143706679344177,
"learning_rate": 1.1851415094339623e-06,
"loss": 0.6223,
"step": 1942
},
{
"epoch": 0.7637578616352201,
"grad_norm": 0.6476017236709595,
"learning_rate": 1.183176100628931e-06,
"loss": 0.7031,
"step": 1943
},
{
"epoch": 0.7641509433962265,
"grad_norm": 0.6435614228248596,
"learning_rate": 1.1812106918238994e-06,
"loss": 0.5795,
"step": 1944
},
{
"epoch": 0.7645440251572327,
"grad_norm": 0.6348211169242859,
"learning_rate": 1.179245283018868e-06,
"loss": 0.7131,
"step": 1945
},
{
"epoch": 0.764937106918239,
"grad_norm": 0.6219361424446106,
"learning_rate": 1.1772798742138364e-06,
"loss": 0.6381,
"step": 1946
},
{
"epoch": 0.7653301886792453,
"grad_norm": 0.525641143321991,
"learning_rate": 1.175314465408805e-06,
"loss": 0.6006,
"step": 1947
},
{
"epoch": 0.7657232704402516,
"grad_norm": 0.7307358384132385,
"learning_rate": 1.1733490566037737e-06,
"loss": 0.7086,
"step": 1948
},
{
"epoch": 0.7661163522012578,
"grad_norm": 0.7370688915252686,
"learning_rate": 1.1713836477987423e-06,
"loss": 0.6243,
"step": 1949
},
{
"epoch": 0.7665094339622641,
"grad_norm": 0.741445779800415,
"learning_rate": 1.1694182389937107e-06,
"loss": 0.7548,
"step": 1950
},
{
"epoch": 0.7669025157232704,
"grad_norm": 0.6911572813987732,
"learning_rate": 1.1674528301886794e-06,
"loss": 0.7175,
"step": 1951
},
{
"epoch": 0.7672955974842768,
"grad_norm": 0.5443742871284485,
"learning_rate": 1.1654874213836478e-06,
"loss": 0.5393,
"step": 1952
},
{
"epoch": 0.7676886792452831,
"grad_norm": 0.9272948503494263,
"learning_rate": 1.1635220125786164e-06,
"loss": 0.7477,
"step": 1953
},
{
"epoch": 0.7680817610062893,
"grad_norm": 0.7260961532592773,
"learning_rate": 1.1615566037735849e-06,
"loss": 0.714,
"step": 1954
},
{
"epoch": 0.7684748427672956,
"grad_norm": 0.5977525115013123,
"learning_rate": 1.1595911949685535e-06,
"loss": 0.6364,
"step": 1955
},
{
"epoch": 0.7688679245283019,
"grad_norm": 0.5826175212860107,
"learning_rate": 1.1576257861635221e-06,
"loss": 0.5284,
"step": 1956
},
{
"epoch": 0.7692610062893082,
"grad_norm": 0.6195921301841736,
"learning_rate": 1.1556603773584908e-06,
"loss": 0.6581,
"step": 1957
},
{
"epoch": 0.7696540880503144,
"grad_norm": 0.8652958273887634,
"learning_rate": 1.1536949685534592e-06,
"loss": 0.7479,
"step": 1958
},
{
"epoch": 0.7700471698113207,
"grad_norm": 0.5708823204040527,
"learning_rate": 1.1517295597484278e-06,
"loss": 0.5934,
"step": 1959
},
{
"epoch": 0.7704402515723271,
"grad_norm": 0.8159661889076233,
"learning_rate": 1.1497641509433962e-06,
"loss": 0.7493,
"step": 1960
},
{
"epoch": 0.7708333333333334,
"grad_norm": 0.7122158408164978,
"learning_rate": 1.1477987421383649e-06,
"loss": 0.6682,
"step": 1961
},
{
"epoch": 0.7712264150943396,
"grad_norm": 0.8063465356826782,
"learning_rate": 1.1458333333333333e-06,
"loss": 0.7512,
"step": 1962
},
{
"epoch": 0.7716194968553459,
"grad_norm": 0.692222535610199,
"learning_rate": 1.143867924528302e-06,
"loss": 0.6552,
"step": 1963
},
{
"epoch": 0.7720125786163522,
"grad_norm": 0.594934344291687,
"learning_rate": 1.1419025157232706e-06,
"loss": 0.625,
"step": 1964
},
{
"epoch": 0.7724056603773585,
"grad_norm": 0.6492074131965637,
"learning_rate": 1.1399371069182392e-06,
"loss": 0.7297,
"step": 1965
},
{
"epoch": 0.7727987421383647,
"grad_norm": 0.7118626832962036,
"learning_rate": 1.1379716981132076e-06,
"loss": 0.6986,
"step": 1966
},
{
"epoch": 0.773191823899371,
"grad_norm": 0.7138659358024597,
"learning_rate": 1.1360062893081763e-06,
"loss": 0.7432,
"step": 1967
},
{
"epoch": 0.7735849056603774,
"grad_norm": 0.5773758292198181,
"learning_rate": 1.1340408805031447e-06,
"loss": 0.6438,
"step": 1968
},
{
"epoch": 0.7739779874213837,
"grad_norm": 0.6772209405899048,
"learning_rate": 1.1320754716981133e-06,
"loss": 0.6926,
"step": 1969
},
{
"epoch": 0.77437106918239,
"grad_norm": 0.7962756156921387,
"learning_rate": 1.1301100628930817e-06,
"loss": 0.7957,
"step": 1970
},
{
"epoch": 0.7747641509433962,
"grad_norm": 0.7118302583694458,
"learning_rate": 1.1281446540880504e-06,
"loss": 0.6655,
"step": 1971
},
{
"epoch": 0.7751572327044025,
"grad_norm": 0.6722339987754822,
"learning_rate": 1.126179245283019e-06,
"loss": 0.637,
"step": 1972
},
{
"epoch": 0.7755503144654088,
"grad_norm": 0.6867762804031372,
"learning_rate": 1.1242138364779874e-06,
"loss": 0.7632,
"step": 1973
},
{
"epoch": 0.7759433962264151,
"grad_norm": 0.6814719438552856,
"learning_rate": 1.122248427672956e-06,
"loss": 0.7192,
"step": 1974
},
{
"epoch": 0.7763364779874213,
"grad_norm": 0.651362955570221,
"learning_rate": 1.1202830188679247e-06,
"loss": 0.7034,
"step": 1975
},
{
"epoch": 0.7767295597484277,
"grad_norm": 0.6886539459228516,
"learning_rate": 1.1183176100628931e-06,
"loss": 0.7341,
"step": 1976
},
{
"epoch": 0.777122641509434,
"grad_norm": 0.559715986251831,
"learning_rate": 1.1163522012578618e-06,
"loss": 0.6237,
"step": 1977
},
{
"epoch": 0.7775157232704403,
"grad_norm": 0.6480963230133057,
"learning_rate": 1.1143867924528302e-06,
"loss": 0.7159,
"step": 1978
},
{
"epoch": 0.7779088050314465,
"grad_norm": 0.8303337693214417,
"learning_rate": 1.1124213836477988e-06,
"loss": 0.6714,
"step": 1979
},
{
"epoch": 0.7783018867924528,
"grad_norm": 0.4986167848110199,
"learning_rate": 1.1104559748427672e-06,
"loss": 0.4941,
"step": 1980
},
{
"epoch": 0.7786949685534591,
"grad_norm": 0.6292533874511719,
"learning_rate": 1.1084905660377359e-06,
"loss": 0.6057,
"step": 1981
},
{
"epoch": 0.7790880503144654,
"grad_norm": 0.621329665184021,
"learning_rate": 1.1065251572327045e-06,
"loss": 0.5671,
"step": 1982
},
{
"epoch": 0.7794811320754716,
"grad_norm": 0.7734379768371582,
"learning_rate": 1.1045597484276732e-06,
"loss": 0.6787,
"step": 1983
},
{
"epoch": 0.779874213836478,
"grad_norm": 0.8058542013168335,
"learning_rate": 1.1025943396226416e-06,
"loss": 0.8156,
"step": 1984
},
{
"epoch": 0.7802672955974843,
"grad_norm": 0.8468074798583984,
"learning_rate": 1.1006289308176102e-06,
"loss": 0.7736,
"step": 1985
},
{
"epoch": 0.7806603773584906,
"grad_norm": 0.8372936844825745,
"learning_rate": 1.0986635220125786e-06,
"loss": 0.7,
"step": 1986
},
{
"epoch": 0.7810534591194969,
"grad_norm": 0.6944576501846313,
"learning_rate": 1.0966981132075473e-06,
"loss": 0.7471,
"step": 1987
},
{
"epoch": 0.7814465408805031,
"grad_norm": 0.6543607711791992,
"learning_rate": 1.0947327044025157e-06,
"loss": 0.6853,
"step": 1988
},
{
"epoch": 0.7818396226415094,
"grad_norm": 0.7486172318458557,
"learning_rate": 1.0927672955974843e-06,
"loss": 0.6493,
"step": 1989
},
{
"epoch": 0.7822327044025157,
"grad_norm": 0.7002162933349609,
"learning_rate": 1.090801886792453e-06,
"loss": 0.6645,
"step": 1990
},
{
"epoch": 0.782625786163522,
"grad_norm": 0.6126017570495605,
"learning_rate": 1.0888364779874216e-06,
"loss": 0.6208,
"step": 1991
},
{
"epoch": 0.7830188679245284,
"grad_norm": 0.7982561588287354,
"learning_rate": 1.08687106918239e-06,
"loss": 0.7351,
"step": 1992
},
{
"epoch": 0.7834119496855346,
"grad_norm": 0.649348258972168,
"learning_rate": 1.0849056603773587e-06,
"loss": 0.6953,
"step": 1993
},
{
"epoch": 0.7838050314465409,
"grad_norm": 1.1700596809387207,
"learning_rate": 1.082940251572327e-06,
"loss": 0.9023,
"step": 1994
},
{
"epoch": 0.7841981132075472,
"grad_norm": 0.9160242676734924,
"learning_rate": 1.0809748427672957e-06,
"loss": 0.755,
"step": 1995
},
{
"epoch": 0.7845911949685535,
"grad_norm": 0.6683558821678162,
"learning_rate": 1.0790094339622641e-06,
"loss": 0.7367,
"step": 1996
},
{
"epoch": 0.7849842767295597,
"grad_norm": 0.7258625030517578,
"learning_rate": 1.0770440251572328e-06,
"loss": 0.7214,
"step": 1997
},
{
"epoch": 0.785377358490566,
"grad_norm": 0.6731094717979431,
"learning_rate": 1.0750786163522014e-06,
"loss": 0.7589,
"step": 1998
},
{
"epoch": 0.7857704402515723,
"grad_norm": 0.8246821761131287,
"learning_rate": 1.07311320754717e-06,
"loss": 0.7292,
"step": 1999
},
{
"epoch": 0.7861635220125787,
"grad_norm": 0.6704521775245667,
"learning_rate": 1.0711477987421385e-06,
"loss": 0.6973,
"step": 2000
},
{
"epoch": 0.7865566037735849,
"grad_norm": 0.6282249689102173,
"learning_rate": 1.069182389937107e-06,
"loss": 0.6284,
"step": 2001
},
{
"epoch": 0.7869496855345912,
"grad_norm": 0.7296137809753418,
"learning_rate": 1.0672169811320755e-06,
"loss": 0.6992,
"step": 2002
},
{
"epoch": 0.7873427672955975,
"grad_norm": 0.8105663657188416,
"learning_rate": 1.0652515723270442e-06,
"loss": 0.7838,
"step": 2003
},
{
"epoch": 0.7877358490566038,
"grad_norm": 0.7701147794723511,
"learning_rate": 1.0632861635220126e-06,
"loss": 0.7166,
"step": 2004
},
{
"epoch": 0.78812893081761,
"grad_norm": 0.9194126129150391,
"learning_rate": 1.0613207547169812e-06,
"loss": 0.7458,
"step": 2005
},
{
"epoch": 0.7885220125786163,
"grad_norm": 0.5097021460533142,
"learning_rate": 1.0593553459119498e-06,
"loss": 0.6764,
"step": 2006
},
{
"epoch": 0.7889150943396226,
"grad_norm": 0.6882070899009705,
"learning_rate": 1.0573899371069183e-06,
"loss": 0.7084,
"step": 2007
},
{
"epoch": 0.789308176100629,
"grad_norm": 0.7543272972106934,
"learning_rate": 1.055424528301887e-06,
"loss": 0.67,
"step": 2008
},
{
"epoch": 0.7897012578616353,
"grad_norm": 0.7908801436424255,
"learning_rate": 1.0534591194968555e-06,
"loss": 0.7186,
"step": 2009
},
{
"epoch": 0.7900943396226415,
"grad_norm": 0.5779961943626404,
"learning_rate": 1.051493710691824e-06,
"loss": 0.6437,
"step": 2010
},
{
"epoch": 0.7904874213836478,
"grad_norm": 0.7615296244621277,
"learning_rate": 1.0495283018867926e-06,
"loss": 0.6948,
"step": 2011
},
{
"epoch": 0.7908805031446541,
"grad_norm": 0.6817577481269836,
"learning_rate": 1.047562893081761e-06,
"loss": 0.6058,
"step": 2012
},
{
"epoch": 0.7912735849056604,
"grad_norm": 0.5761832594871521,
"learning_rate": 1.0455974842767297e-06,
"loss": 0.6027,
"step": 2013
},
{
"epoch": 0.7916666666666666,
"grad_norm": 0.6301024556159973,
"learning_rate": 1.043632075471698e-06,
"loss": 0.6906,
"step": 2014
},
{
"epoch": 0.7920597484276729,
"grad_norm": 0.8220508694648743,
"learning_rate": 1.0416666666666667e-06,
"loss": 0.7467,
"step": 2015
},
{
"epoch": 0.7924528301886793,
"grad_norm": 0.8649424314498901,
"learning_rate": 1.0397012578616354e-06,
"loss": 0.7823,
"step": 2016
},
{
"epoch": 0.7928459119496856,
"grad_norm": 0.5401425361633301,
"learning_rate": 1.037735849056604e-06,
"loss": 0.5589,
"step": 2017
},
{
"epoch": 0.7932389937106918,
"grad_norm": 0.6610985994338989,
"learning_rate": 1.0357704402515724e-06,
"loss": 0.6608,
"step": 2018
},
{
"epoch": 0.7936320754716981,
"grad_norm": 0.7545170783996582,
"learning_rate": 1.033805031446541e-06,
"loss": 0.7073,
"step": 2019
},
{
"epoch": 0.7940251572327044,
"grad_norm": 0.9143254160881042,
"learning_rate": 1.0318396226415095e-06,
"loss": 0.6852,
"step": 2020
},
{
"epoch": 0.7944182389937107,
"grad_norm": 0.6908055543899536,
"learning_rate": 1.029874213836478e-06,
"loss": 0.7526,
"step": 2021
},
{
"epoch": 0.7948113207547169,
"grad_norm": 0.7849035263061523,
"learning_rate": 1.0279088050314465e-06,
"loss": 0.7286,
"step": 2022
},
{
"epoch": 0.7952044025157232,
"grad_norm": 0.6233143210411072,
"learning_rate": 1.0259433962264152e-06,
"loss": 0.6578,
"step": 2023
},
{
"epoch": 0.7955974842767296,
"grad_norm": 0.6351747512817383,
"learning_rate": 1.0239779874213838e-06,
"loss": 0.6602,
"step": 2024
},
{
"epoch": 0.7959905660377359,
"grad_norm": 0.6351912617683411,
"learning_rate": 1.0220125786163524e-06,
"loss": 0.6816,
"step": 2025
},
{
"epoch": 0.7963836477987422,
"grad_norm": 0.7194613218307495,
"learning_rate": 1.0200471698113209e-06,
"loss": 0.7289,
"step": 2026
},
{
"epoch": 0.7967767295597484,
"grad_norm": 0.6231948733329773,
"learning_rate": 1.0180817610062895e-06,
"loss": 0.6969,
"step": 2027
},
{
"epoch": 0.7971698113207547,
"grad_norm": 0.7831196784973145,
"learning_rate": 1.016116352201258e-06,
"loss": 0.6332,
"step": 2028
},
{
"epoch": 0.797562893081761,
"grad_norm": 0.7124795913696289,
"learning_rate": 1.0141509433962265e-06,
"loss": 0.7302,
"step": 2029
},
{
"epoch": 0.7979559748427673,
"grad_norm": 0.6532600522041321,
"learning_rate": 1.012185534591195e-06,
"loss": 0.624,
"step": 2030
},
{
"epoch": 0.7983490566037735,
"grad_norm": 0.6349525451660156,
"learning_rate": 1.0102201257861636e-06,
"loss": 0.696,
"step": 2031
},
{
"epoch": 0.7987421383647799,
"grad_norm": 0.8113916516304016,
"learning_rate": 1.0082547169811322e-06,
"loss": 0.771,
"step": 2032
},
{
"epoch": 0.7991352201257862,
"grad_norm": 0.7549936771392822,
"learning_rate": 1.0062893081761007e-06,
"loss": 0.7341,
"step": 2033
},
{
"epoch": 0.7995283018867925,
"grad_norm": 0.6913676261901855,
"learning_rate": 1.0043238993710693e-06,
"loss": 0.6641,
"step": 2034
},
{
"epoch": 0.7999213836477987,
"grad_norm": 0.7270733118057251,
"learning_rate": 1.002358490566038e-06,
"loss": 0.6226,
"step": 2035
},
{
"epoch": 0.800314465408805,
"grad_norm": 0.8471736311912537,
"learning_rate": 1.0003930817610064e-06,
"loss": 0.7563,
"step": 2036
},
{
"epoch": 0.8007075471698113,
"grad_norm": 0.8240023851394653,
"learning_rate": 9.98427672955975e-07,
"loss": 0.688,
"step": 2037
},
{
"epoch": 0.8011006289308176,
"grad_norm": 0.8322147130966187,
"learning_rate": 9.964622641509434e-07,
"loss": 0.7874,
"step": 2038
},
{
"epoch": 0.8014937106918238,
"grad_norm": 0.649107038974762,
"learning_rate": 9.94496855345912e-07,
"loss": 0.6521,
"step": 2039
},
{
"epoch": 0.8018867924528302,
"grad_norm": 0.7060580253601074,
"learning_rate": 9.925314465408805e-07,
"loss": 0.5976,
"step": 2040
},
{
"epoch": 0.8022798742138365,
"grad_norm": 0.5053736567497253,
"learning_rate": 9.90566037735849e-07,
"loss": 0.6136,
"step": 2041
},
{
"epoch": 0.8026729559748428,
"grad_norm": 0.834518313407898,
"learning_rate": 9.886006289308177e-07,
"loss": 0.7267,
"step": 2042
},
{
"epoch": 0.8030660377358491,
"grad_norm": 0.7801578044891357,
"learning_rate": 9.866352201257864e-07,
"loss": 0.7327,
"step": 2043
},
{
"epoch": 0.8034591194968553,
"grad_norm": 1.050073504447937,
"learning_rate": 9.846698113207548e-07,
"loss": 0.8484,
"step": 2044
},
{
"epoch": 0.8038522012578616,
"grad_norm": 0.7547958493232727,
"learning_rate": 9.827044025157234e-07,
"loss": 0.7649,
"step": 2045
},
{
"epoch": 0.8042452830188679,
"grad_norm": 0.6827205419540405,
"learning_rate": 9.807389937106919e-07,
"loss": 0.6729,
"step": 2046
},
{
"epoch": 0.8046383647798742,
"grad_norm": 0.9065365791320801,
"learning_rate": 9.787735849056605e-07,
"loss": 0.8091,
"step": 2047
},
{
"epoch": 0.8050314465408805,
"grad_norm": 0.7954641580581665,
"learning_rate": 9.76808176100629e-07,
"loss": 0.6702,
"step": 2048
},
{
"epoch": 0.8054245283018868,
"grad_norm": 0.8450998067855835,
"learning_rate": 9.748427672955975e-07,
"loss": 0.7217,
"step": 2049
},
{
"epoch": 0.8058176100628931,
"grad_norm": 0.550790548324585,
"learning_rate": 9.728773584905662e-07,
"loss": 0.6655,
"step": 2050
},
{
"epoch": 0.8062106918238994,
"grad_norm": 0.9165398478507996,
"learning_rate": 9.709119496855348e-07,
"loss": 0.7446,
"step": 2051
},
{
"epoch": 0.8066037735849056,
"grad_norm": 0.6817429661750793,
"learning_rate": 9.689465408805032e-07,
"loss": 0.6947,
"step": 2052
},
{
"epoch": 0.8069968553459119,
"grad_norm": 0.7346860766410828,
"learning_rate": 9.669811320754719e-07,
"loss": 0.6498,
"step": 2053
},
{
"epoch": 0.8073899371069182,
"grad_norm": 0.843814492225647,
"learning_rate": 9.650157232704403e-07,
"loss": 0.6902,
"step": 2054
},
{
"epoch": 0.8077830188679245,
"grad_norm": 0.7756029367446899,
"learning_rate": 9.63050314465409e-07,
"loss": 0.6871,
"step": 2055
},
{
"epoch": 0.8081761006289309,
"grad_norm": 0.83280348777771,
"learning_rate": 9.610849056603774e-07,
"loss": 0.6338,
"step": 2056
},
{
"epoch": 0.8085691823899371,
"grad_norm": 0.4901805818080902,
"learning_rate": 9.59119496855346e-07,
"loss": 0.5479,
"step": 2057
},
{
"epoch": 0.8089622641509434,
"grad_norm": 0.632866621017456,
"learning_rate": 9.571540880503146e-07,
"loss": 0.6495,
"step": 2058
},
{
"epoch": 0.8093553459119497,
"grad_norm": 0.8583084344863892,
"learning_rate": 9.551886792452833e-07,
"loss": 0.7517,
"step": 2059
},
{
"epoch": 0.809748427672956,
"grad_norm": 0.602715015411377,
"learning_rate": 9.532232704402517e-07,
"loss": 0.6033,
"step": 2060
},
{
"epoch": 0.8101415094339622,
"grad_norm": 0.7460941672325134,
"learning_rate": 9.512578616352203e-07,
"loss": 0.6783,
"step": 2061
},
{
"epoch": 0.8105345911949685,
"grad_norm": 0.849700927734375,
"learning_rate": 9.492924528301887e-07,
"loss": 0.7837,
"step": 2062
},
{
"epoch": 0.8109276729559748,
"grad_norm": 0.9711039066314697,
"learning_rate": 9.473270440251574e-07,
"loss": 0.8469,
"step": 2063
},
{
"epoch": 0.8113207547169812,
"grad_norm": 0.6572835445404053,
"learning_rate": 9.453616352201259e-07,
"loss": 0.6526,
"step": 2064
},
{
"epoch": 0.8117138364779874,
"grad_norm": 0.6990302801132202,
"learning_rate": 9.433962264150944e-07,
"loss": 0.6887,
"step": 2065
},
{
"epoch": 0.8121069182389937,
"grad_norm": 0.5969560742378235,
"learning_rate": 9.41430817610063e-07,
"loss": 0.6442,
"step": 2066
},
{
"epoch": 0.8125,
"grad_norm": 0.5617697238922119,
"learning_rate": 9.394654088050316e-07,
"loss": 0.6594,
"step": 2067
},
{
"epoch": 0.8128930817610063,
"grad_norm": 0.6697828769683838,
"learning_rate": 9.375000000000001e-07,
"loss": 0.6802,
"step": 2068
},
{
"epoch": 0.8132861635220126,
"grad_norm": 0.6873515844345093,
"learning_rate": 9.355345911949686e-07,
"loss": 0.561,
"step": 2069
},
{
"epoch": 0.8136792452830188,
"grad_norm": 0.7622373700141907,
"learning_rate": 9.335691823899372e-07,
"loss": 0.681,
"step": 2070
},
{
"epoch": 0.8140723270440252,
"grad_norm": 0.7938657402992249,
"learning_rate": 9.316037735849057e-07,
"loss": 0.7458,
"step": 2071
},
{
"epoch": 0.8144654088050315,
"grad_norm": 0.7434559464454651,
"learning_rate": 9.296383647798742e-07,
"loss": 0.7386,
"step": 2072
},
{
"epoch": 0.8148584905660378,
"grad_norm": 0.4549093544483185,
"learning_rate": 9.276729559748428e-07,
"loss": 0.5219,
"step": 2073
},
{
"epoch": 0.815251572327044,
"grad_norm": 0.7378183603286743,
"learning_rate": 9.257075471698114e-07,
"loss": 0.6254,
"step": 2074
},
{
"epoch": 0.8156446540880503,
"grad_norm": 0.7158591747283936,
"learning_rate": 9.237421383647798e-07,
"loss": 0.6313,
"step": 2075
},
{
"epoch": 0.8160377358490566,
"grad_norm": 0.886826753616333,
"learning_rate": 9.217767295597485e-07,
"loss": 0.7092,
"step": 2076
},
{
"epoch": 0.8164308176100629,
"grad_norm": 0.7937220335006714,
"learning_rate": 9.19811320754717e-07,
"loss": 0.7277,
"step": 2077
},
{
"epoch": 0.8168238993710691,
"grad_norm": 0.7406826019287109,
"learning_rate": 9.178459119496856e-07,
"loss": 0.6782,
"step": 2078
},
{
"epoch": 0.8172169811320755,
"grad_norm": 0.6834436058998108,
"learning_rate": 9.158805031446541e-07,
"loss": 0.6097,
"step": 2079
},
{
"epoch": 0.8176100628930818,
"grad_norm": 0.7956271171569824,
"learning_rate": 9.139150943396227e-07,
"loss": 0.7198,
"step": 2080
},
{
"epoch": 0.8180031446540881,
"grad_norm": 0.7943224906921387,
"learning_rate": 9.119496855345912e-07,
"loss": 0.7303,
"step": 2081
},
{
"epoch": 0.8183962264150944,
"grad_norm": 0.9182903170585632,
"learning_rate": 9.099842767295599e-07,
"loss": 0.8661,
"step": 2082
},
{
"epoch": 0.8187893081761006,
"grad_norm": 0.7442007660865784,
"learning_rate": 9.080188679245283e-07,
"loss": 0.6719,
"step": 2083
},
{
"epoch": 0.8191823899371069,
"grad_norm": 0.6482791304588318,
"learning_rate": 9.060534591194969e-07,
"loss": 0.652,
"step": 2084
},
{
"epoch": 0.8195754716981132,
"grad_norm": 0.7246862649917603,
"learning_rate": 9.040880503144654e-07,
"loss": 0.6819,
"step": 2085
},
{
"epoch": 0.8199685534591195,
"grad_norm": 0.6830928921699524,
"learning_rate": 9.021226415094341e-07,
"loss": 0.6654,
"step": 2086
},
{
"epoch": 0.8203616352201258,
"grad_norm": 0.7986692786216736,
"learning_rate": 9.001572327044025e-07,
"loss": 0.6924,
"step": 2087
},
{
"epoch": 0.8207547169811321,
"grad_norm": 0.9282518625259399,
"learning_rate": 8.981918238993711e-07,
"loss": 0.7483,
"step": 2088
},
{
"epoch": 0.8211477987421384,
"grad_norm": 0.5931207537651062,
"learning_rate": 8.962264150943397e-07,
"loss": 0.5927,
"step": 2089
},
{
"epoch": 0.8215408805031447,
"grad_norm": 0.7713273167610168,
"learning_rate": 8.942610062893083e-07,
"loss": 0.7413,
"step": 2090
},
{
"epoch": 0.8219339622641509,
"grad_norm": 0.7351595759391785,
"learning_rate": 8.922955974842767e-07,
"loss": 0.6971,
"step": 2091
},
{
"epoch": 0.8223270440251572,
"grad_norm": 0.7585707306861877,
"learning_rate": 8.903301886792454e-07,
"loss": 0.7081,
"step": 2092
},
{
"epoch": 0.8227201257861635,
"grad_norm": 0.824413537979126,
"learning_rate": 8.883647798742139e-07,
"loss": 0.7028,
"step": 2093
},
{
"epoch": 0.8231132075471698,
"grad_norm": 0.8870236277580261,
"learning_rate": 8.863993710691825e-07,
"loss": 0.7643,
"step": 2094
},
{
"epoch": 0.8235062893081762,
"grad_norm": 0.7966058850288391,
"learning_rate": 8.844339622641509e-07,
"loss": 0.7181,
"step": 2095
},
{
"epoch": 0.8238993710691824,
"grad_norm": 0.6055851578712463,
"learning_rate": 8.824685534591196e-07,
"loss": 0.6648,
"step": 2096
},
{
"epoch": 0.8242924528301887,
"grad_norm": 0.4398571848869324,
"learning_rate": 8.805031446540881e-07,
"loss": 0.5864,
"step": 2097
},
{
"epoch": 0.824685534591195,
"grad_norm": 0.8281417489051819,
"learning_rate": 8.785377358490567e-07,
"loss": 0.7391,
"step": 2098
},
{
"epoch": 0.8250786163522013,
"grad_norm": 0.5791189074516296,
"learning_rate": 8.765723270440252e-07,
"loss": 0.688,
"step": 2099
},
{
"epoch": 0.8254716981132075,
"grad_norm": 0.6722317337989807,
"learning_rate": 8.746069182389938e-07,
"loss": 0.7361,
"step": 2100
},
{
"epoch": 0.8258647798742138,
"grad_norm": 0.670451283454895,
"learning_rate": 8.726415094339623e-07,
"loss": 0.6855,
"step": 2101
},
{
"epoch": 0.8262578616352201,
"grad_norm": 0.6406052112579346,
"learning_rate": 8.706761006289309e-07,
"loss": 0.7218,
"step": 2102
},
{
"epoch": 0.8266509433962265,
"grad_norm": 0.7226956486701965,
"learning_rate": 8.687106918238994e-07,
"loss": 0.7148,
"step": 2103
},
{
"epoch": 0.8270440251572327,
"grad_norm": 0.6525507569313049,
"learning_rate": 8.66745283018868e-07,
"loss": 0.6497,
"step": 2104
},
{
"epoch": 0.827437106918239,
"grad_norm": 0.7424986958503723,
"learning_rate": 8.647798742138364e-07,
"loss": 0.6357,
"step": 2105
},
{
"epoch": 0.8278301886792453,
"grad_norm": 0.633202314376831,
"learning_rate": 8.628144654088051e-07,
"loss": 0.6255,
"step": 2106
},
{
"epoch": 0.8282232704402516,
"grad_norm": 0.9444885849952698,
"learning_rate": 8.608490566037736e-07,
"loss": 0.7743,
"step": 2107
},
{
"epoch": 0.8286163522012578,
"grad_norm": 0.7263575196266174,
"learning_rate": 8.588836477987422e-07,
"loss": 0.6861,
"step": 2108
},
{
"epoch": 0.8290094339622641,
"grad_norm": 0.657383382320404,
"learning_rate": 8.569182389937107e-07,
"loss": 0.6805,
"step": 2109
},
{
"epoch": 0.8294025157232704,
"grad_norm": 0.6440843343734741,
"learning_rate": 8.549528301886793e-07,
"loss": 0.6363,
"step": 2110
},
{
"epoch": 0.8297955974842768,
"grad_norm": 0.9268834590911865,
"learning_rate": 8.529874213836478e-07,
"loss": 0.8014,
"step": 2111
},
{
"epoch": 0.8301886792452831,
"grad_norm": 0.8338631987571716,
"learning_rate": 8.510220125786165e-07,
"loss": 0.7562,
"step": 2112
},
{
"epoch": 0.8305817610062893,
"grad_norm": 0.7695666551589966,
"learning_rate": 8.490566037735849e-07,
"loss": 0.6395,
"step": 2113
},
{
"epoch": 0.8309748427672956,
"grad_norm": 0.7504798173904419,
"learning_rate": 8.470911949685535e-07,
"loss": 0.6812,
"step": 2114
},
{
"epoch": 0.8313679245283019,
"grad_norm": 0.7264564633369446,
"learning_rate": 8.45125786163522e-07,
"loss": 0.642,
"step": 2115
},
{
"epoch": 0.8317610062893082,
"grad_norm": 0.9987094402313232,
"learning_rate": 8.431603773584907e-07,
"loss": 0.8151,
"step": 2116
},
{
"epoch": 0.8321540880503144,
"grad_norm": 0.5091885924339294,
"learning_rate": 8.411949685534591e-07,
"loss": 0.5454,
"step": 2117
},
{
"epoch": 0.8325471698113207,
"grad_norm": 0.6013506054878235,
"learning_rate": 8.392295597484277e-07,
"loss": 0.5931,
"step": 2118
},
{
"epoch": 0.8329402515723271,
"grad_norm": 0.7142112851142883,
"learning_rate": 8.372641509433963e-07,
"loss": 0.658,
"step": 2119
},
{
"epoch": 0.8333333333333334,
"grad_norm": 0.6570302248001099,
"learning_rate": 8.352987421383649e-07,
"loss": 0.5923,
"step": 2120
},
{
"epoch": 0.8337264150943396,
"grad_norm": 0.928011953830719,
"learning_rate": 8.333333333333333e-07,
"loss": 0.8181,
"step": 2121
},
{
"epoch": 0.8341194968553459,
"grad_norm": 0.8374940156936646,
"learning_rate": 8.31367924528302e-07,
"loss": 0.728,
"step": 2122
},
{
"epoch": 0.8345125786163522,
"grad_norm": 0.7601545453071594,
"learning_rate": 8.294025157232705e-07,
"loss": 0.6691,
"step": 2123
},
{
"epoch": 0.8349056603773585,
"grad_norm": 0.6093471646308899,
"learning_rate": 8.274371069182391e-07,
"loss": 0.6224,
"step": 2124
},
{
"epoch": 0.8352987421383647,
"grad_norm": 0.678017258644104,
"learning_rate": 8.254716981132076e-07,
"loss": 0.7063,
"step": 2125
},
{
"epoch": 0.835691823899371,
"grad_norm": 0.8472076058387756,
"learning_rate": 8.235062893081762e-07,
"loss": 0.8,
"step": 2126
},
{
"epoch": 0.8360849056603774,
"grad_norm": 0.6381399631500244,
"learning_rate": 8.215408805031447e-07,
"loss": 0.6809,
"step": 2127
},
{
"epoch": 0.8364779874213837,
"grad_norm": 0.7611330151557922,
"learning_rate": 8.195754716981134e-07,
"loss": 0.7484,
"step": 2128
},
{
"epoch": 0.83687106918239,
"grad_norm": 0.7774641513824463,
"learning_rate": 8.176100628930818e-07,
"loss": 0.6965,
"step": 2129
},
{
"epoch": 0.8372641509433962,
"grad_norm": 0.7421042323112488,
"learning_rate": 8.156446540880504e-07,
"loss": 0.7106,
"step": 2130
},
{
"epoch": 0.8376572327044025,
"grad_norm": 0.6298864483833313,
"learning_rate": 8.136792452830189e-07,
"loss": 0.6703,
"step": 2131
},
{
"epoch": 0.8380503144654088,
"grad_norm": 0.9404190182685852,
"learning_rate": 8.117138364779875e-07,
"loss": 0.6199,
"step": 2132
},
{
"epoch": 0.8384433962264151,
"grad_norm": 0.7531468272209167,
"learning_rate": 8.09748427672956e-07,
"loss": 0.6202,
"step": 2133
},
{
"epoch": 0.8388364779874213,
"grad_norm": 0.6355054974555969,
"learning_rate": 8.077830188679246e-07,
"loss": 0.6564,
"step": 2134
},
{
"epoch": 0.8392295597484277,
"grad_norm": 0.7885834574699402,
"learning_rate": 8.058176100628932e-07,
"loss": 0.776,
"step": 2135
},
{
"epoch": 0.839622641509434,
"grad_norm": 0.592938244342804,
"learning_rate": 8.038522012578617e-07,
"loss": 0.6511,
"step": 2136
},
{
"epoch": 0.8400157232704403,
"grad_norm": 0.6714280247688293,
"learning_rate": 8.018867924528302e-07,
"loss": 0.6371,
"step": 2137
},
{
"epoch": 0.8404088050314465,
"grad_norm": 0.8623645901679993,
"learning_rate": 7.999213836477989e-07,
"loss": 0.7181,
"step": 2138
},
{
"epoch": 0.8408018867924528,
"grad_norm": 0.625275194644928,
"learning_rate": 7.979559748427673e-07,
"loss": 0.7197,
"step": 2139
},
{
"epoch": 0.8411949685534591,
"grad_norm": 0.6897543668746948,
"learning_rate": 7.959905660377359e-07,
"loss": 0.7083,
"step": 2140
},
{
"epoch": 0.8415880503144654,
"grad_norm": 0.6384572982788086,
"learning_rate": 7.940251572327044e-07,
"loss": 0.6453,
"step": 2141
},
{
"epoch": 0.8419811320754716,
"grad_norm": 0.6829888820648193,
"learning_rate": 7.920597484276731e-07,
"loss": 0.7482,
"step": 2142
},
{
"epoch": 0.842374213836478,
"grad_norm": 0.8479306697845459,
"learning_rate": 7.900943396226415e-07,
"loss": 0.7228,
"step": 2143
},
{
"epoch": 0.8427672955974843,
"grad_norm": 0.677605390548706,
"learning_rate": 7.881289308176101e-07,
"loss": 0.6638,
"step": 2144
},
{
"epoch": 0.8431603773584906,
"grad_norm": 0.8834080696105957,
"learning_rate": 7.861635220125787e-07,
"loss": 0.7296,
"step": 2145
},
{
"epoch": 0.8435534591194969,
"grad_norm": 0.7259169816970825,
"learning_rate": 7.841981132075473e-07,
"loss": 0.7983,
"step": 2146
},
{
"epoch": 0.8439465408805031,
"grad_norm": 0.7068047523498535,
"learning_rate": 7.822327044025157e-07,
"loss": 0.7087,
"step": 2147
},
{
"epoch": 0.8443396226415094,
"grad_norm": 0.7905950546264648,
"learning_rate": 7.802672955974844e-07,
"loss": 0.6978,
"step": 2148
},
{
"epoch": 0.8447327044025157,
"grad_norm": 0.9114578366279602,
"learning_rate": 7.783018867924529e-07,
"loss": 0.7656,
"step": 2149
},
{
"epoch": 0.845125786163522,
"grad_norm": 0.845358669757843,
"learning_rate": 7.763364779874215e-07,
"loss": 0.7706,
"step": 2150
},
{
"epoch": 0.8455188679245284,
"grad_norm": 0.8917364478111267,
"learning_rate": 7.743710691823899e-07,
"loss": 0.7563,
"step": 2151
},
{
"epoch": 0.8459119496855346,
"grad_norm": 0.5324864983558655,
"learning_rate": 7.724056603773586e-07,
"loss": 0.5914,
"step": 2152
},
{
"epoch": 0.8463050314465409,
"grad_norm": 0.7455214262008667,
"learning_rate": 7.704402515723271e-07,
"loss": 0.7793,
"step": 2153
},
{
"epoch": 0.8466981132075472,
"grad_norm": 0.6343944072723389,
"learning_rate": 7.684748427672957e-07,
"loss": 0.6285,
"step": 2154
},
{
"epoch": 0.8470911949685535,
"grad_norm": 0.9082866311073303,
"learning_rate": 7.665094339622642e-07,
"loss": 0.7603,
"step": 2155
},
{
"epoch": 0.8474842767295597,
"grad_norm": 0.6421903967857361,
"learning_rate": 7.645440251572328e-07,
"loss": 0.6855,
"step": 2156
},
{
"epoch": 0.847877358490566,
"grad_norm": 0.651233434677124,
"learning_rate": 7.625786163522013e-07,
"loss": 0.6672,
"step": 2157
},
{
"epoch": 0.8482704402515723,
"grad_norm": 0.5363133549690247,
"learning_rate": 7.6061320754717e-07,
"loss": 0.5588,
"step": 2158
},
{
"epoch": 0.8486635220125787,
"grad_norm": 0.8857976198196411,
"learning_rate": 7.586477987421384e-07,
"loss": 0.8063,
"step": 2159
},
{
"epoch": 0.8490566037735849,
"grad_norm": 0.5353593826293945,
"learning_rate": 7.56682389937107e-07,
"loss": 0.6345,
"step": 2160
},
{
"epoch": 0.8494496855345912,
"grad_norm": 0.925854504108429,
"learning_rate": 7.547169811320755e-07,
"loss": 0.7935,
"step": 2161
},
{
"epoch": 0.8498427672955975,
"grad_norm": 0.9320954084396362,
"learning_rate": 7.527515723270441e-07,
"loss": 0.7528,
"step": 2162
},
{
"epoch": 0.8502358490566038,
"grad_norm": 0.9282365441322327,
"learning_rate": 7.507861635220126e-07,
"loss": 0.7432,
"step": 2163
},
{
"epoch": 0.85062893081761,
"grad_norm": 0.7289482951164246,
"learning_rate": 7.488207547169812e-07,
"loss": 0.6461,
"step": 2164
},
{
"epoch": 0.8510220125786163,
"grad_norm": 0.7657983303070068,
"learning_rate": 7.468553459119498e-07,
"loss": 0.6829,
"step": 2165
},
{
"epoch": 0.8514150943396226,
"grad_norm": 0.5177061557769775,
"learning_rate": 7.448899371069183e-07,
"loss": 0.6401,
"step": 2166
},
{
"epoch": 0.851808176100629,
"grad_norm": 0.8209033608436584,
"learning_rate": 7.429245283018868e-07,
"loss": 0.676,
"step": 2167
},
{
"epoch": 0.8522012578616353,
"grad_norm": 0.6597095727920532,
"learning_rate": 7.409591194968555e-07,
"loss": 0.6927,
"step": 2168
},
{
"epoch": 0.8525943396226415,
"grad_norm": 0.42412251234054565,
"learning_rate": 7.389937106918239e-07,
"loss": 0.5245,
"step": 2169
},
{
"epoch": 0.8529874213836478,
"grad_norm": 0.5578562617301941,
"learning_rate": 7.370283018867925e-07,
"loss": 0.6927,
"step": 2170
},
{
"epoch": 0.8533805031446541,
"grad_norm": 0.7341641187667847,
"learning_rate": 7.35062893081761e-07,
"loss": 0.725,
"step": 2171
},
{
"epoch": 0.8537735849056604,
"grad_norm": 0.7412542104721069,
"learning_rate": 7.330974842767297e-07,
"loss": 0.6853,
"step": 2172
},
{
"epoch": 0.8541666666666666,
"grad_norm": 0.5031754374504089,
"learning_rate": 7.311320754716981e-07,
"loss": 0.59,
"step": 2173
},
{
"epoch": 0.8545597484276729,
"grad_norm": 0.7424189448356628,
"learning_rate": 7.291666666666667e-07,
"loss": 0.5995,
"step": 2174
},
{
"epoch": 0.8549528301886793,
"grad_norm": 0.5712892413139343,
"learning_rate": 7.272012578616353e-07,
"loss": 0.6033,
"step": 2175
},
{
"epoch": 0.8553459119496856,
"grad_norm": 0.8861500024795532,
"learning_rate": 7.252358490566039e-07,
"loss": 0.6608,
"step": 2176
},
{
"epoch": 0.8557389937106918,
"grad_norm": 0.8404026031494141,
"learning_rate": 7.232704402515723e-07,
"loss": 0.6948,
"step": 2177
},
{
"epoch": 0.8561320754716981,
"grad_norm": 0.5331982374191284,
"learning_rate": 7.21305031446541e-07,
"loss": 0.6372,
"step": 2178
},
{
"epoch": 0.8565251572327044,
"grad_norm": 0.6254251003265381,
"learning_rate": 7.193396226415095e-07,
"loss": 0.6569,
"step": 2179
},
{
"epoch": 0.8569182389937107,
"grad_norm": 0.8525477647781372,
"learning_rate": 7.173742138364781e-07,
"loss": 0.7988,
"step": 2180
},
{
"epoch": 0.8573113207547169,
"grad_norm": 0.8504541516304016,
"learning_rate": 7.154088050314466e-07,
"loss": 0.7075,
"step": 2181
},
{
"epoch": 0.8577044025157232,
"grad_norm": 0.8278780579566956,
"learning_rate": 7.134433962264152e-07,
"loss": 0.6654,
"step": 2182
},
{
"epoch": 0.8580974842767296,
"grad_norm": 0.678631067276001,
"learning_rate": 7.114779874213837e-07,
"loss": 0.6135,
"step": 2183
},
{
"epoch": 0.8584905660377359,
"grad_norm": 0.6618059873580933,
"learning_rate": 7.095125786163524e-07,
"loss": 0.649,
"step": 2184
},
{
"epoch": 0.8588836477987422,
"grad_norm": 0.7474246025085449,
"learning_rate": 7.075471698113208e-07,
"loss": 0.6835,
"step": 2185
},
{
"epoch": 0.8592767295597484,
"grad_norm": 0.8349992632865906,
"learning_rate": 7.055817610062894e-07,
"loss": 0.6382,
"step": 2186
},
{
"epoch": 0.8596698113207547,
"grad_norm": 0.8552443981170654,
"learning_rate": 7.036163522012579e-07,
"loss": 0.7053,
"step": 2187
},
{
"epoch": 0.860062893081761,
"grad_norm": 0.6273599863052368,
"learning_rate": 7.016509433962266e-07,
"loss": 0.6512,
"step": 2188
},
{
"epoch": 0.8604559748427673,
"grad_norm": 0.8922796249389648,
"learning_rate": 6.99685534591195e-07,
"loss": 0.7006,
"step": 2189
},
{
"epoch": 0.8608490566037735,
"grad_norm": 0.6670902967453003,
"learning_rate": 6.977201257861636e-07,
"loss": 0.6704,
"step": 2190
},
{
"epoch": 0.8612421383647799,
"grad_norm": 0.8337033987045288,
"learning_rate": 6.957547169811322e-07,
"loss": 0.6497,
"step": 2191
},
{
"epoch": 0.8616352201257862,
"grad_norm": 0.6916108131408691,
"learning_rate": 6.937893081761007e-07,
"loss": 0.7192,
"step": 2192
},
{
"epoch": 0.8620283018867925,
"grad_norm": 0.8048521280288696,
"learning_rate": 6.918238993710692e-07,
"loss": 0.7347,
"step": 2193
},
{
"epoch": 0.8624213836477987,
"grad_norm": 0.7148359417915344,
"learning_rate": 6.898584905660379e-07,
"loss": 0.6766,
"step": 2194
},
{
"epoch": 0.862814465408805,
"grad_norm": 0.6825347542762756,
"learning_rate": 6.878930817610064e-07,
"loss": 0.646,
"step": 2195
},
{
"epoch": 0.8632075471698113,
"grad_norm": 0.768984854221344,
"learning_rate": 6.859276729559749e-07,
"loss": 0.7266,
"step": 2196
},
{
"epoch": 0.8636006289308176,
"grad_norm": 0.7933337688446045,
"learning_rate": 6.839622641509434e-07,
"loss": 0.617,
"step": 2197
},
{
"epoch": 0.8639937106918238,
"grad_norm": 0.6067782044410706,
"learning_rate": 6.819968553459121e-07,
"loss": 0.6754,
"step": 2198
},
{
"epoch": 0.8643867924528302,
"grad_norm": 0.7924044132232666,
"learning_rate": 6.800314465408805e-07,
"loss": 0.6211,
"step": 2199
},
{
"epoch": 0.8647798742138365,
"grad_norm": 0.7121582627296448,
"learning_rate": 6.780660377358491e-07,
"loss": 0.7187,
"step": 2200
},
{
"epoch": 0.8651729559748428,
"grad_norm": 0.7485283613204956,
"learning_rate": 6.761006289308177e-07,
"loss": 0.6687,
"step": 2201
},
{
"epoch": 0.8655660377358491,
"grad_norm": 0.8024988770484924,
"learning_rate": 6.741352201257863e-07,
"loss": 0.804,
"step": 2202
},
{
"epoch": 0.8659591194968553,
"grad_norm": 0.5811664462089539,
"learning_rate": 6.721698113207547e-07,
"loss": 0.5707,
"step": 2203
},
{
"epoch": 0.8663522012578616,
"grad_norm": 0.730022668838501,
"learning_rate": 6.702044025157234e-07,
"loss": 0.7007,
"step": 2204
},
{
"epoch": 0.8667452830188679,
"grad_norm": 0.8079362511634827,
"learning_rate": 6.682389937106919e-07,
"loss": 0.783,
"step": 2205
},
{
"epoch": 0.8671383647798742,
"grad_norm": 0.7234131693840027,
"learning_rate": 6.662735849056605e-07,
"loss": 0.6558,
"step": 2206
},
{
"epoch": 0.8675314465408805,
"grad_norm": 0.8410349488258362,
"learning_rate": 6.643081761006289e-07,
"loss": 0.692,
"step": 2207
},
{
"epoch": 0.8679245283018868,
"grad_norm": 0.728874146938324,
"learning_rate": 6.623427672955976e-07,
"loss": 0.7708,
"step": 2208
},
{
"epoch": 0.8683176100628931,
"grad_norm": 0.7635148763656616,
"learning_rate": 6.603773584905661e-07,
"loss": 0.7605,
"step": 2209
},
{
"epoch": 0.8687106918238994,
"grad_norm": 0.8761162757873535,
"learning_rate": 6.584119496855347e-07,
"loss": 0.7369,
"step": 2210
},
{
"epoch": 0.8691037735849056,
"grad_norm": 0.7046441435813904,
"learning_rate": 6.564465408805032e-07,
"loss": 0.6838,
"step": 2211
},
{
"epoch": 0.8694968553459119,
"grad_norm": 0.7394724488258362,
"learning_rate": 6.544811320754718e-07,
"loss": 0.7415,
"step": 2212
},
{
"epoch": 0.8698899371069182,
"grad_norm": 0.7082499861717224,
"learning_rate": 6.525157232704403e-07,
"loss": 0.7039,
"step": 2213
},
{
"epoch": 0.8702830188679245,
"grad_norm": 0.9574195742607117,
"learning_rate": 6.50550314465409e-07,
"loss": 0.7682,
"step": 2214
},
{
"epoch": 0.8706761006289309,
"grad_norm": 0.5450464487075806,
"learning_rate": 6.485849056603774e-07,
"loss": 0.6431,
"step": 2215
},
{
"epoch": 0.8710691823899371,
"grad_norm": 0.6493065357208252,
"learning_rate": 6.46619496855346e-07,
"loss": 0.6252,
"step": 2216
},
{
"epoch": 0.8714622641509434,
"grad_norm": 0.5460614562034607,
"learning_rate": 6.446540880503145e-07,
"loss": 0.6515,
"step": 2217
},
{
"epoch": 0.8718553459119497,
"grad_norm": 0.539410412311554,
"learning_rate": 6.426886792452832e-07,
"loss": 0.5117,
"step": 2218
},
{
"epoch": 0.872248427672956,
"grad_norm": 0.867027759552002,
"learning_rate": 6.407232704402516e-07,
"loss": 0.7623,
"step": 2219
},
{
"epoch": 0.8726415094339622,
"grad_norm": 0.6146731376647949,
"learning_rate": 6.387578616352202e-07,
"loss": 0.6447,
"step": 2220
},
{
"epoch": 0.8730345911949685,
"grad_norm": 0.8140153884887695,
"learning_rate": 6.367924528301888e-07,
"loss": 0.6348,
"step": 2221
},
{
"epoch": 0.8734276729559748,
"grad_norm": 0.829351007938385,
"learning_rate": 6.348270440251573e-07,
"loss": 0.6948,
"step": 2222
},
{
"epoch": 0.8738207547169812,
"grad_norm": 0.8364050984382629,
"learning_rate": 6.328616352201258e-07,
"loss": 0.7408,
"step": 2223
},
{
"epoch": 0.8742138364779874,
"grad_norm": 0.7193739414215088,
"learning_rate": 6.308962264150945e-07,
"loss": 0.637,
"step": 2224
},
{
"epoch": 0.8746069182389937,
"grad_norm": 0.7198230624198914,
"learning_rate": 6.28930817610063e-07,
"loss": 0.6712,
"step": 2225
},
{
"epoch": 0.875,
"grad_norm": 0.8115258812904358,
"learning_rate": 6.269654088050315e-07,
"loss": 0.6478,
"step": 2226
},
{
"epoch": 0.8753930817610063,
"grad_norm": 0.6732747554779053,
"learning_rate": 6.25e-07,
"loss": 0.6956,
"step": 2227
},
{
"epoch": 0.8757861635220126,
"grad_norm": 0.7144453525543213,
"learning_rate": 6.230345911949686e-07,
"loss": 0.689,
"step": 2228
},
{
"epoch": 0.8761792452830188,
"grad_norm": 0.6687198281288147,
"learning_rate": 6.210691823899371e-07,
"loss": 0.6521,
"step": 2229
},
{
"epoch": 0.8765723270440252,
"grad_norm": 0.7080721855163574,
"learning_rate": 6.191037735849057e-07,
"loss": 0.665,
"step": 2230
},
{
"epoch": 0.8769654088050315,
"grad_norm": 0.957785964012146,
"learning_rate": 6.171383647798743e-07,
"loss": 0.7585,
"step": 2231
},
{
"epoch": 0.8773584905660378,
"grad_norm": 0.6964478492736816,
"learning_rate": 6.151729559748428e-07,
"loss": 0.6253,
"step": 2232
},
{
"epoch": 0.877751572327044,
"grad_norm": 0.7985994815826416,
"learning_rate": 6.132075471698113e-07,
"loss": 0.6979,
"step": 2233
},
{
"epoch": 0.8781446540880503,
"grad_norm": 0.8320304155349731,
"learning_rate": 6.1124213836478e-07,
"loss": 0.7583,
"step": 2234
},
{
"epoch": 0.8785377358490566,
"grad_norm": 0.6821916103363037,
"learning_rate": 6.092767295597485e-07,
"loss": 0.6218,
"step": 2235
},
{
"epoch": 0.8789308176100629,
"grad_norm": 0.7088303565979004,
"learning_rate": 6.07311320754717e-07,
"loss": 0.7252,
"step": 2236
},
{
"epoch": 0.8793238993710691,
"grad_norm": 0.7732560634613037,
"learning_rate": 6.053459119496856e-07,
"loss": 0.7052,
"step": 2237
},
{
"epoch": 0.8797169811320755,
"grad_norm": 0.8688510656356812,
"learning_rate": 6.033805031446542e-07,
"loss": 0.699,
"step": 2238
},
{
"epoch": 0.8801100628930818,
"grad_norm": 0.8021149635314941,
"learning_rate": 6.014150943396227e-07,
"loss": 0.6927,
"step": 2239
},
{
"epoch": 0.8805031446540881,
"grad_norm": 0.6642415523529053,
"learning_rate": 5.994496855345912e-07,
"loss": 0.6189,
"step": 2240
},
{
"epoch": 0.8808962264150944,
"grad_norm": 0.9047964811325073,
"learning_rate": 5.974842767295598e-07,
"loss": 0.735,
"step": 2241
},
{
"epoch": 0.8812893081761006,
"grad_norm": 0.6826280951499939,
"learning_rate": 5.955188679245284e-07,
"loss": 0.608,
"step": 2242
},
{
"epoch": 0.8816823899371069,
"grad_norm": 0.5821129679679871,
"learning_rate": 5.935534591194969e-07,
"loss": 0.5951,
"step": 2243
},
{
"epoch": 0.8820754716981132,
"grad_norm": 0.7674536108970642,
"learning_rate": 5.915880503144655e-07,
"loss": 0.6544,
"step": 2244
},
{
"epoch": 0.8824685534591195,
"grad_norm": 0.6841880679130554,
"learning_rate": 5.89622641509434e-07,
"loss": 0.7448,
"step": 2245
},
{
"epoch": 0.8828616352201258,
"grad_norm": 0.5807332992553711,
"learning_rate": 5.876572327044025e-07,
"loss": 0.6755,
"step": 2246
},
{
"epoch": 0.8832547169811321,
"grad_norm": 0.8066653609275818,
"learning_rate": 5.856918238993712e-07,
"loss": 0.7344,
"step": 2247
},
{
"epoch": 0.8836477987421384,
"grad_norm": 0.7158612012863159,
"learning_rate": 5.837264150943397e-07,
"loss": 0.5919,
"step": 2248
},
{
"epoch": 0.8840408805031447,
"grad_norm": 0.7206102609634399,
"learning_rate": 5.817610062893082e-07,
"loss": 0.6393,
"step": 2249
},
{
"epoch": 0.8844339622641509,
"grad_norm": 0.825652539730072,
"learning_rate": 5.797955974842767e-07,
"loss": 0.7383,
"step": 2250
},
{
"epoch": 0.8848270440251572,
"grad_norm": 0.5739361047744751,
"learning_rate": 5.778301886792454e-07,
"loss": 0.5748,
"step": 2251
},
{
"epoch": 0.8852201257861635,
"grad_norm": 0.5491999387741089,
"learning_rate": 5.758647798742139e-07,
"loss": 0.583,
"step": 2252
},
{
"epoch": 0.8856132075471698,
"grad_norm": 0.6294852495193481,
"learning_rate": 5.738993710691824e-07,
"loss": 0.6313,
"step": 2253
},
{
"epoch": 0.8860062893081762,
"grad_norm": 0.6460597515106201,
"learning_rate": 5.71933962264151e-07,
"loss": 0.6604,
"step": 2254
},
{
"epoch": 0.8863993710691824,
"grad_norm": 0.6836321353912354,
"learning_rate": 5.699685534591196e-07,
"loss": 0.6979,
"step": 2255
},
{
"epoch": 0.8867924528301887,
"grad_norm": 0.7384762167930603,
"learning_rate": 5.680031446540881e-07,
"loss": 0.7661,
"step": 2256
},
{
"epoch": 0.887185534591195,
"grad_norm": 0.88250333070755,
"learning_rate": 5.660377358490567e-07,
"loss": 0.8219,
"step": 2257
},
{
"epoch": 0.8875786163522013,
"grad_norm": 0.8016765713691711,
"learning_rate": 5.640723270440252e-07,
"loss": 0.6896,
"step": 2258
},
{
"epoch": 0.8879716981132075,
"grad_norm": 0.5383222103118896,
"learning_rate": 5.621069182389937e-07,
"loss": 0.6212,
"step": 2259
},
{
"epoch": 0.8883647798742138,
"grad_norm": 0.7442513108253479,
"learning_rate": 5.601415094339624e-07,
"loss": 0.6866,
"step": 2260
},
{
"epoch": 0.8887578616352201,
"grad_norm": 0.862188994884491,
"learning_rate": 5.581761006289309e-07,
"loss": 0.7236,
"step": 2261
},
{
"epoch": 0.8891509433962265,
"grad_norm": 0.8897164463996887,
"learning_rate": 5.562106918238994e-07,
"loss": 0.7001,
"step": 2262
},
{
"epoch": 0.8895440251572327,
"grad_norm": 0.6462240815162659,
"learning_rate": 5.542452830188679e-07,
"loss": 0.5953,
"step": 2263
},
{
"epoch": 0.889937106918239,
"grad_norm": 0.8180879354476929,
"learning_rate": 5.522798742138366e-07,
"loss": 0.7578,
"step": 2264
},
{
"epoch": 0.8903301886792453,
"grad_norm": 0.84379643201828,
"learning_rate": 5.503144654088051e-07,
"loss": 0.7339,
"step": 2265
},
{
"epoch": 0.8907232704402516,
"grad_norm": 0.48507997393608093,
"learning_rate": 5.483490566037736e-07,
"loss": 0.5914,
"step": 2266
},
{
"epoch": 0.8911163522012578,
"grad_norm": 0.9045882225036621,
"learning_rate": 5.463836477987422e-07,
"loss": 0.6763,
"step": 2267
},
{
"epoch": 0.8915094339622641,
"grad_norm": 0.8475188612937927,
"learning_rate": 5.444182389937108e-07,
"loss": 0.7516,
"step": 2268
},
{
"epoch": 0.8919025157232704,
"grad_norm": 0.7776294946670532,
"learning_rate": 5.424528301886793e-07,
"loss": 0.7421,
"step": 2269
},
{
"epoch": 0.8922955974842768,
"grad_norm": 0.626072883605957,
"learning_rate": 5.404874213836479e-07,
"loss": 0.6707,
"step": 2270
},
{
"epoch": 0.8926886792452831,
"grad_norm": 0.6033751964569092,
"learning_rate": 5.385220125786164e-07,
"loss": 0.6152,
"step": 2271
},
{
"epoch": 0.8930817610062893,
"grad_norm": 0.6357232928276062,
"learning_rate": 5.36556603773585e-07,
"loss": 0.6588,
"step": 2272
},
{
"epoch": 0.8934748427672956,
"grad_norm": 0.6891751289367676,
"learning_rate": 5.345911949685535e-07,
"loss": 0.6877,
"step": 2273
},
{
"epoch": 0.8938679245283019,
"grad_norm": 0.6049560904502869,
"learning_rate": 5.326257861635221e-07,
"loss": 0.6633,
"step": 2274
},
{
"epoch": 0.8942610062893082,
"grad_norm": 0.7945406436920166,
"learning_rate": 5.306603773584906e-07,
"loss": 0.6687,
"step": 2275
},
{
"epoch": 0.8946540880503144,
"grad_norm": 0.8579593896865845,
"learning_rate": 5.286949685534591e-07,
"loss": 0.6996,
"step": 2276
},
{
"epoch": 0.8950471698113207,
"grad_norm": 0.7623710036277771,
"learning_rate": 5.267295597484278e-07,
"loss": 0.6553,
"step": 2277
},
{
"epoch": 0.8954402515723271,
"grad_norm": 0.6010566353797913,
"learning_rate": 5.247641509433963e-07,
"loss": 0.6554,
"step": 2278
},
{
"epoch": 0.8958333333333334,
"grad_norm": 0.8942103385925293,
"learning_rate": 5.227987421383648e-07,
"loss": 0.824,
"step": 2279
},
{
"epoch": 0.8962264150943396,
"grad_norm": 0.7015033960342407,
"learning_rate": 5.208333333333334e-07,
"loss": 0.6941,
"step": 2280
},
{
"epoch": 0.8966194968553459,
"grad_norm": 0.6229555606842041,
"learning_rate": 5.18867924528302e-07,
"loss": 0.5747,
"step": 2281
},
{
"epoch": 0.8970125786163522,
"grad_norm": 0.7651469111442566,
"learning_rate": 5.169025157232705e-07,
"loss": 0.6854,
"step": 2282
},
{
"epoch": 0.8974056603773585,
"grad_norm": 0.5889945030212402,
"learning_rate": 5.14937106918239e-07,
"loss": 0.6182,
"step": 2283
},
{
"epoch": 0.8977987421383647,
"grad_norm": 0.664434552192688,
"learning_rate": 5.129716981132076e-07,
"loss": 0.6506,
"step": 2284
},
{
"epoch": 0.898191823899371,
"grad_norm": 0.6179591417312622,
"learning_rate": 5.110062893081762e-07,
"loss": 0.578,
"step": 2285
},
{
"epoch": 0.8985849056603774,
"grad_norm": 0.6659973859786987,
"learning_rate": 5.090408805031447e-07,
"loss": 0.6963,
"step": 2286
},
{
"epoch": 0.8989779874213837,
"grad_norm": 0.7498767971992493,
"learning_rate": 5.070754716981133e-07,
"loss": 0.7275,
"step": 2287
},
{
"epoch": 0.89937106918239,
"grad_norm": 0.7169463038444519,
"learning_rate": 5.051100628930818e-07,
"loss": 0.7921,
"step": 2288
},
{
"epoch": 0.8997641509433962,
"grad_norm": 0.6784248352050781,
"learning_rate": 5.031446540880503e-07,
"loss": 0.7273,
"step": 2289
},
{
"epoch": 0.9001572327044025,
"grad_norm": 0.5459941625595093,
"learning_rate": 5.01179245283019e-07,
"loss": 0.5861,
"step": 2290
},
{
"epoch": 0.9005503144654088,
"grad_norm": 0.5371059775352478,
"learning_rate": 4.992138364779875e-07,
"loss": 0.5895,
"step": 2291
},
{
"epoch": 0.9009433962264151,
"grad_norm": 0.8659892678260803,
"learning_rate": 4.97248427672956e-07,
"loss": 0.6904,
"step": 2292
},
{
"epoch": 0.9013364779874213,
"grad_norm": 0.6568877696990967,
"learning_rate": 4.952830188679246e-07,
"loss": 0.5797,
"step": 2293
},
{
"epoch": 0.9017295597484277,
"grad_norm": 0.6748905777931213,
"learning_rate": 4.933176100628932e-07,
"loss": 0.686,
"step": 2294
},
{
"epoch": 0.902122641509434,
"grad_norm": 0.7980628609657288,
"learning_rate": 4.913522012578617e-07,
"loss": 0.6523,
"step": 2295
},
{
"epoch": 0.9025157232704403,
"grad_norm": 0.742996335029602,
"learning_rate": 4.893867924528302e-07,
"loss": 0.6721,
"step": 2296
},
{
"epoch": 0.9029088050314465,
"grad_norm": 0.5274096131324768,
"learning_rate": 4.874213836477988e-07,
"loss": 0.6355,
"step": 2297
},
{
"epoch": 0.9033018867924528,
"grad_norm": 0.5477995872497559,
"learning_rate": 4.854559748427674e-07,
"loss": 0.5836,
"step": 2298
},
{
"epoch": 0.9036949685534591,
"grad_norm": 0.4998157024383545,
"learning_rate": 4.834905660377359e-07,
"loss": 0.5814,
"step": 2299
},
{
"epoch": 0.9040880503144654,
"grad_norm": 0.6170914769172668,
"learning_rate": 4.815251572327045e-07,
"loss": 0.7358,
"step": 2300
},
{
"epoch": 0.9044811320754716,
"grad_norm": 0.8767325282096863,
"learning_rate": 4.79559748427673e-07,
"loss": 0.7571,
"step": 2301
},
{
"epoch": 0.904874213836478,
"grad_norm": 0.8191457390785217,
"learning_rate": 4.775943396226416e-07,
"loss": 0.7429,
"step": 2302
},
{
"epoch": 0.9052672955974843,
"grad_norm": 0.6178702712059021,
"learning_rate": 4.7562893081761016e-07,
"loss": 0.649,
"step": 2303
},
{
"epoch": 0.9056603773584906,
"grad_norm": 0.7363266944885254,
"learning_rate": 4.736635220125787e-07,
"loss": 0.6407,
"step": 2304
},
{
"epoch": 0.9060534591194969,
"grad_norm": 0.7614299654960632,
"learning_rate": 4.716981132075472e-07,
"loss": 0.7721,
"step": 2305
},
{
"epoch": 0.9064465408805031,
"grad_norm": 1.0564371347427368,
"learning_rate": 4.697327044025158e-07,
"loss": 0.8239,
"step": 2306
},
{
"epoch": 0.9068396226415094,
"grad_norm": 0.6215397119522095,
"learning_rate": 4.677672955974843e-07,
"loss": 0.6331,
"step": 2307
},
{
"epoch": 0.9072327044025157,
"grad_norm": 0.8987818360328674,
"learning_rate": 4.6580188679245286e-07,
"loss": 0.7419,
"step": 2308
},
{
"epoch": 0.907625786163522,
"grad_norm": 0.9845703840255737,
"learning_rate": 4.638364779874214e-07,
"loss": 0.7115,
"step": 2309
},
{
"epoch": 0.9080188679245284,
"grad_norm": 0.5923305153846741,
"learning_rate": 4.618710691823899e-07,
"loss": 0.5868,
"step": 2310
},
{
"epoch": 0.9084119496855346,
"grad_norm": 0.777499794960022,
"learning_rate": 4.599056603773585e-07,
"loss": 0.6838,
"step": 2311
},
{
"epoch": 0.9088050314465409,
"grad_norm": 0.7136509418487549,
"learning_rate": 4.5794025157232703e-07,
"loss": 0.699,
"step": 2312
},
{
"epoch": 0.9091981132075472,
"grad_norm": 0.8085744976997375,
"learning_rate": 4.559748427672956e-07,
"loss": 0.7384,
"step": 2313
},
{
"epoch": 0.9095911949685535,
"grad_norm": 0.9945822954177856,
"learning_rate": 4.5400943396226414e-07,
"loss": 0.7372,
"step": 2314
},
{
"epoch": 0.9099842767295597,
"grad_norm": 0.7743339538574219,
"learning_rate": 4.520440251572327e-07,
"loss": 0.7036,
"step": 2315
},
{
"epoch": 0.910377358490566,
"grad_norm": 0.6208356618881226,
"learning_rate": 4.5007861635220125e-07,
"loss": 0.6498,
"step": 2316
},
{
"epoch": 0.9107704402515723,
"grad_norm": 0.5301515460014343,
"learning_rate": 4.4811320754716983e-07,
"loss": 0.5159,
"step": 2317
},
{
"epoch": 0.9111635220125787,
"grad_norm": 0.6418805718421936,
"learning_rate": 4.4614779874213836e-07,
"loss": 0.649,
"step": 2318
},
{
"epoch": 0.9115566037735849,
"grad_norm": 0.9385761618614197,
"learning_rate": 4.4418238993710694e-07,
"loss": 0.8372,
"step": 2319
},
{
"epoch": 0.9119496855345912,
"grad_norm": 0.8436682820320129,
"learning_rate": 4.4221698113207547e-07,
"loss": 0.7915,
"step": 2320
},
{
"epoch": 0.9123427672955975,
"grad_norm": 0.7746303677558899,
"learning_rate": 4.4025157232704405e-07,
"loss": 0.6681,
"step": 2321
},
{
"epoch": 0.9127358490566038,
"grad_norm": 0.7922371029853821,
"learning_rate": 4.382861635220126e-07,
"loss": 0.5936,
"step": 2322
},
{
"epoch": 0.91312893081761,
"grad_norm": 0.8280319571495056,
"learning_rate": 4.3632075471698116e-07,
"loss": 0.7589,
"step": 2323
},
{
"epoch": 0.9135220125786163,
"grad_norm": 0.8240867853164673,
"learning_rate": 4.343553459119497e-07,
"loss": 0.7092,
"step": 2324
},
{
"epoch": 0.9139150943396226,
"grad_norm": 0.5760433077812195,
"learning_rate": 4.323899371069182e-07,
"loss": 0.593,
"step": 2325
},
{
"epoch": 0.914308176100629,
"grad_norm": 0.5271994471549988,
"learning_rate": 4.304245283018868e-07,
"loss": 0.6395,
"step": 2326
},
{
"epoch": 0.9147012578616353,
"grad_norm": 0.7953335046768188,
"learning_rate": 4.2845911949685533e-07,
"loss": 0.6132,
"step": 2327
},
{
"epoch": 0.9150943396226415,
"grad_norm": 0.6854754090309143,
"learning_rate": 4.264937106918239e-07,
"loss": 0.6368,
"step": 2328
},
{
"epoch": 0.9154874213836478,
"grad_norm": 0.7024213075637817,
"learning_rate": 4.2452830188679244e-07,
"loss": 0.6725,
"step": 2329
},
{
"epoch": 0.9158805031446541,
"grad_norm": 0.7252901196479797,
"learning_rate": 4.22562893081761e-07,
"loss": 0.7396,
"step": 2330
},
{
"epoch": 0.9162735849056604,
"grad_norm": 0.7473825216293335,
"learning_rate": 4.2059748427672955e-07,
"loss": 0.7454,
"step": 2331
},
{
"epoch": 0.9166666666666666,
"grad_norm": 0.438187837600708,
"learning_rate": 4.1863207547169814e-07,
"loss": 0.577,
"step": 2332
},
{
"epoch": 0.9170597484276729,
"grad_norm": 0.6881603598594666,
"learning_rate": 4.1666666666666667e-07,
"loss": 0.7153,
"step": 2333
},
{
"epoch": 0.9174528301886793,
"grad_norm": 0.8859202861785889,
"learning_rate": 4.1470125786163525e-07,
"loss": 0.7133,
"step": 2334
},
{
"epoch": 0.9178459119496856,
"grad_norm": 0.8237269520759583,
"learning_rate": 4.127358490566038e-07,
"loss": 0.7325,
"step": 2335
},
{
"epoch": 0.9182389937106918,
"grad_norm": 0.695204496383667,
"learning_rate": 4.1077044025157236e-07,
"loss": 0.6534,
"step": 2336
},
{
"epoch": 0.9186320754716981,
"grad_norm": 0.6794942021369934,
"learning_rate": 4.088050314465409e-07,
"loss": 0.6331,
"step": 2337
},
{
"epoch": 0.9190251572327044,
"grad_norm": 0.6913821697235107,
"learning_rate": 4.0683962264150947e-07,
"loss": 0.6609,
"step": 2338
},
{
"epoch": 0.9194182389937107,
"grad_norm": 0.8759965896606445,
"learning_rate": 4.04874213836478e-07,
"loss": 0.6796,
"step": 2339
},
{
"epoch": 0.9198113207547169,
"grad_norm": 0.7205759286880493,
"learning_rate": 4.029088050314466e-07,
"loss": 0.7416,
"step": 2340
},
{
"epoch": 0.9202044025157232,
"grad_norm": 0.5913090109825134,
"learning_rate": 4.009433962264151e-07,
"loss": 0.6349,
"step": 2341
},
{
"epoch": 0.9205974842767296,
"grad_norm": 0.6801372766494751,
"learning_rate": 3.9897798742138364e-07,
"loss": 0.7255,
"step": 2342
},
{
"epoch": 0.9209905660377359,
"grad_norm": 0.7718689441680908,
"learning_rate": 3.970125786163522e-07,
"loss": 0.7382,
"step": 2343
},
{
"epoch": 0.9213836477987422,
"grad_norm": 0.8258588314056396,
"learning_rate": 3.9504716981132075e-07,
"loss": 0.7838,
"step": 2344
},
{
"epoch": 0.9217767295597484,
"grad_norm": 0.6906368732452393,
"learning_rate": 3.9308176100628933e-07,
"loss": 0.6479,
"step": 2345
},
{
"epoch": 0.9221698113207547,
"grad_norm": 0.6711302995681763,
"learning_rate": 3.9111635220125786e-07,
"loss": 0.6063,
"step": 2346
},
{
"epoch": 0.922562893081761,
"grad_norm": 0.7531667351722717,
"learning_rate": 3.8915094339622644e-07,
"loss": 0.6414,
"step": 2347
},
{
"epoch": 0.9229559748427673,
"grad_norm": 0.7275190949440002,
"learning_rate": 3.8718553459119497e-07,
"loss": 0.699,
"step": 2348
},
{
"epoch": 0.9233490566037735,
"grad_norm": 0.7164285182952881,
"learning_rate": 3.8522012578616355e-07,
"loss": 0.6818,
"step": 2349
},
{
"epoch": 0.9237421383647799,
"grad_norm": 0.809161901473999,
"learning_rate": 3.832547169811321e-07,
"loss": 0.7682,
"step": 2350
},
{
"epoch": 0.9241352201257862,
"grad_norm": 0.7315307259559631,
"learning_rate": 3.8128930817610066e-07,
"loss": 0.6817,
"step": 2351
},
{
"epoch": 0.9245283018867925,
"grad_norm": 0.8704824447631836,
"learning_rate": 3.793238993710692e-07,
"loss": 0.8666,
"step": 2352
},
{
"epoch": 0.9249213836477987,
"grad_norm": 1.2433363199234009,
"learning_rate": 3.773584905660378e-07,
"loss": 0.6985,
"step": 2353
},
{
"epoch": 0.925314465408805,
"grad_norm": 0.7518700361251831,
"learning_rate": 3.753930817610063e-07,
"loss": 0.6293,
"step": 2354
},
{
"epoch": 0.9257075471698113,
"grad_norm": 0.7492860555648804,
"learning_rate": 3.734276729559749e-07,
"loss": 0.6794,
"step": 2355
},
{
"epoch": 0.9261006289308176,
"grad_norm": 0.8851842284202576,
"learning_rate": 3.714622641509434e-07,
"loss": 0.6887,
"step": 2356
},
{
"epoch": 0.9264937106918238,
"grad_norm": 0.5948863625526428,
"learning_rate": 3.6949685534591194e-07,
"loss": 0.6636,
"step": 2357
},
{
"epoch": 0.9268867924528302,
"grad_norm": 0.6122174859046936,
"learning_rate": 3.675314465408805e-07,
"loss": 0.6884,
"step": 2358
},
{
"epoch": 0.9272798742138365,
"grad_norm": 0.6668429374694824,
"learning_rate": 3.6556603773584905e-07,
"loss": 0.6976,
"step": 2359
},
{
"epoch": 0.9276729559748428,
"grad_norm": 0.717311680316925,
"learning_rate": 3.6360062893081764e-07,
"loss": 0.6895,
"step": 2360
},
{
"epoch": 0.9280660377358491,
"grad_norm": 0.7679991722106934,
"learning_rate": 3.6163522012578617e-07,
"loss": 0.7487,
"step": 2361
},
{
"epoch": 0.9284591194968553,
"grad_norm": 0.6769139170646667,
"learning_rate": 3.5966981132075475e-07,
"loss": 0.6782,
"step": 2362
},
{
"epoch": 0.9288522012578616,
"grad_norm": 0.6605120301246643,
"learning_rate": 3.577044025157233e-07,
"loss": 0.6775,
"step": 2363
},
{
"epoch": 0.9292452830188679,
"grad_norm": 0.6043455600738525,
"learning_rate": 3.5573899371069186e-07,
"loss": 0.6935,
"step": 2364
},
{
"epoch": 0.9296383647798742,
"grad_norm": 0.4983665943145752,
"learning_rate": 3.537735849056604e-07,
"loss": 0.6005,
"step": 2365
},
{
"epoch": 0.9300314465408805,
"grad_norm": 0.8987080454826355,
"learning_rate": 3.5180817610062897e-07,
"loss": 0.7004,
"step": 2366
},
{
"epoch": 0.9304245283018868,
"grad_norm": 0.6549533605575562,
"learning_rate": 3.498427672955975e-07,
"loss": 0.5887,
"step": 2367
},
{
"epoch": 0.9308176100628931,
"grad_norm": 0.6867566704750061,
"learning_rate": 3.478773584905661e-07,
"loss": 0.5782,
"step": 2368
},
{
"epoch": 0.9312106918238994,
"grad_norm": 0.756587564945221,
"learning_rate": 3.459119496855346e-07,
"loss": 0.7014,
"step": 2369
},
{
"epoch": 0.9316037735849056,
"grad_norm": 0.7742331624031067,
"learning_rate": 3.439465408805032e-07,
"loss": 0.6985,
"step": 2370
},
{
"epoch": 0.9319968553459119,
"grad_norm": 0.7038583159446716,
"learning_rate": 3.419811320754717e-07,
"loss": 0.7147,
"step": 2371
},
{
"epoch": 0.9323899371069182,
"grad_norm": 0.7191103100776672,
"learning_rate": 3.4001572327044025e-07,
"loss": 0.7206,
"step": 2372
},
{
"epoch": 0.9327830188679245,
"grad_norm": 0.8880930542945862,
"learning_rate": 3.3805031446540883e-07,
"loss": 0.7617,
"step": 2373
},
{
"epoch": 0.9331761006289309,
"grad_norm": 1.020113468170166,
"learning_rate": 3.3608490566037736e-07,
"loss": 0.7974,
"step": 2374
},
{
"epoch": 0.9335691823899371,
"grad_norm": 0.7107282876968384,
"learning_rate": 3.3411949685534594e-07,
"loss": 0.6372,
"step": 2375
},
{
"epoch": 0.9339622641509434,
"grad_norm": 0.8825415968894958,
"learning_rate": 3.3215408805031447e-07,
"loss": 0.8318,
"step": 2376
},
{
"epoch": 0.9343553459119497,
"grad_norm": 0.6801412105560303,
"learning_rate": 3.3018867924528305e-07,
"loss": 0.6851,
"step": 2377
},
{
"epoch": 0.934748427672956,
"grad_norm": 0.6181076765060425,
"learning_rate": 3.282232704402516e-07,
"loss": 0.6158,
"step": 2378
},
{
"epoch": 0.9351415094339622,
"grad_norm": 0.5888661742210388,
"learning_rate": 3.2625786163522016e-07,
"loss": 0.5974,
"step": 2379
},
{
"epoch": 0.9355345911949685,
"grad_norm": 0.6409070491790771,
"learning_rate": 3.242924528301887e-07,
"loss": 0.6611,
"step": 2380
},
{
"epoch": 0.9359276729559748,
"grad_norm": 0.821003258228302,
"learning_rate": 3.223270440251573e-07,
"loss": 0.7035,
"step": 2381
},
{
"epoch": 0.9363207547169812,
"grad_norm": 0.6716895699501038,
"learning_rate": 3.203616352201258e-07,
"loss": 0.6309,
"step": 2382
},
{
"epoch": 0.9367138364779874,
"grad_norm": 0.7499057054519653,
"learning_rate": 3.183962264150944e-07,
"loss": 0.6597,
"step": 2383
},
{
"epoch": 0.9371069182389937,
"grad_norm": 0.7916625738143921,
"learning_rate": 3.164308176100629e-07,
"loss": 0.6998,
"step": 2384
},
{
"epoch": 0.9375,
"grad_norm": 0.782078206539154,
"learning_rate": 3.144654088050315e-07,
"loss": 0.7583,
"step": 2385
},
{
"epoch": 0.9378930817610063,
"grad_norm": 1.0178775787353516,
"learning_rate": 3.125e-07,
"loss": 0.7756,
"step": 2386
},
{
"epoch": 0.9382861635220126,
"grad_norm": 0.8949575424194336,
"learning_rate": 3.1053459119496855e-07,
"loss": 0.7803,
"step": 2387
},
{
"epoch": 0.9386792452830188,
"grad_norm": 0.7660632729530334,
"learning_rate": 3.0856918238993714e-07,
"loss": 0.7054,
"step": 2388
},
{
"epoch": 0.9390723270440252,
"grad_norm": 0.6456524133682251,
"learning_rate": 3.0660377358490567e-07,
"loss": 0.623,
"step": 2389
},
{
"epoch": 0.9394654088050315,
"grad_norm": 0.6632437705993652,
"learning_rate": 3.0463836477987425e-07,
"loss": 0.7106,
"step": 2390
},
{
"epoch": 0.9398584905660378,
"grad_norm": 0.7999177575111389,
"learning_rate": 3.026729559748428e-07,
"loss": 0.7051,
"step": 2391
},
{
"epoch": 0.940251572327044,
"grad_norm": 0.9132054448127747,
"learning_rate": 3.0070754716981136e-07,
"loss": 0.7803,
"step": 2392
},
{
"epoch": 0.9406446540880503,
"grad_norm": 0.8408699035644531,
"learning_rate": 2.987421383647799e-07,
"loss": 0.7463,
"step": 2393
},
{
"epoch": 0.9410377358490566,
"grad_norm": 0.6376796960830688,
"learning_rate": 2.9677672955974847e-07,
"loss": 0.6403,
"step": 2394
},
{
"epoch": 0.9414308176100629,
"grad_norm": 0.573990523815155,
"learning_rate": 2.94811320754717e-07,
"loss": 0.5634,
"step": 2395
},
{
"epoch": 0.9418238993710691,
"grad_norm": 0.6912069916725159,
"learning_rate": 2.928459119496856e-07,
"loss": 0.6109,
"step": 2396
},
{
"epoch": 0.9422169811320755,
"grad_norm": 0.8401463627815247,
"learning_rate": 2.908805031446541e-07,
"loss": 0.6425,
"step": 2397
},
{
"epoch": 0.9426100628930818,
"grad_norm": 0.6724429130554199,
"learning_rate": 2.889150943396227e-07,
"loss": 0.6292,
"step": 2398
},
{
"epoch": 0.9430031446540881,
"grad_norm": 0.6276699304580688,
"learning_rate": 2.869496855345912e-07,
"loss": 0.6661,
"step": 2399
},
{
"epoch": 0.9433962264150944,
"grad_norm": 0.5308483242988586,
"learning_rate": 2.849842767295598e-07,
"loss": 0.6078,
"step": 2400
},
{
"epoch": 0.9437893081761006,
"grad_norm": 0.687053382396698,
"learning_rate": 2.8301886792452833e-07,
"loss": 0.6473,
"step": 2401
},
{
"epoch": 0.9441823899371069,
"grad_norm": 0.8723526000976562,
"learning_rate": 2.8105345911949686e-07,
"loss": 0.6957,
"step": 2402
},
{
"epoch": 0.9445754716981132,
"grad_norm": 0.992062509059906,
"learning_rate": 2.7908805031446544e-07,
"loss": 0.7163,
"step": 2403
},
{
"epoch": 0.9449685534591195,
"grad_norm": 0.643212616443634,
"learning_rate": 2.7712264150943397e-07,
"loss": 0.674,
"step": 2404
},
{
"epoch": 0.9453616352201258,
"grad_norm": 0.7510846853256226,
"learning_rate": 2.7515723270440255e-07,
"loss": 0.6978,
"step": 2405
},
{
"epoch": 0.9457547169811321,
"grad_norm": 0.7666007280349731,
"learning_rate": 2.731918238993711e-07,
"loss": 0.6072,
"step": 2406
},
{
"epoch": 0.9461477987421384,
"grad_norm": 0.8785666823387146,
"learning_rate": 2.7122641509433966e-07,
"loss": 0.6984,
"step": 2407
},
{
"epoch": 0.9465408805031447,
"grad_norm": 0.5799489617347717,
"learning_rate": 2.692610062893082e-07,
"loss": 0.6203,
"step": 2408
},
{
"epoch": 0.9469339622641509,
"grad_norm": 0.6542503237724304,
"learning_rate": 2.672955974842768e-07,
"loss": 0.6688,
"step": 2409
},
{
"epoch": 0.9473270440251572,
"grad_norm": 0.6964982748031616,
"learning_rate": 2.653301886792453e-07,
"loss": 0.6665,
"step": 2410
},
{
"epoch": 0.9477201257861635,
"grad_norm": 0.8040996193885803,
"learning_rate": 2.633647798742139e-07,
"loss": 0.7969,
"step": 2411
},
{
"epoch": 0.9481132075471698,
"grad_norm": 0.7918441891670227,
"learning_rate": 2.613993710691824e-07,
"loss": 0.6233,
"step": 2412
},
{
"epoch": 0.9485062893081762,
"grad_norm": 0.7664849758148193,
"learning_rate": 2.59433962264151e-07,
"loss": 0.6324,
"step": 2413
},
{
"epoch": 0.9488993710691824,
"grad_norm": 0.6251068711280823,
"learning_rate": 2.574685534591195e-07,
"loss": 0.5737,
"step": 2414
},
{
"epoch": 0.9492924528301887,
"grad_norm": 0.7931944131851196,
"learning_rate": 2.555031446540881e-07,
"loss": 0.7975,
"step": 2415
},
{
"epoch": 0.949685534591195,
"grad_norm": 0.7203230261802673,
"learning_rate": 2.5353773584905664e-07,
"loss": 0.6905,
"step": 2416
},
{
"epoch": 0.9500786163522013,
"grad_norm": 0.6621385812759399,
"learning_rate": 2.5157232704402517e-07,
"loss": 0.6588,
"step": 2417
},
{
"epoch": 0.9504716981132075,
"grad_norm": 0.5589357018470764,
"learning_rate": 2.4960691823899375e-07,
"loss": 0.6547,
"step": 2418
},
{
"epoch": 0.9508647798742138,
"grad_norm": 0.7616498470306396,
"learning_rate": 2.476415094339623e-07,
"loss": 0.6937,
"step": 2419
},
{
"epoch": 0.9512578616352201,
"grad_norm": 0.850884735584259,
"learning_rate": 2.4567610062893086e-07,
"loss": 0.6728,
"step": 2420
},
{
"epoch": 0.9516509433962265,
"grad_norm": 0.6656224727630615,
"learning_rate": 2.437106918238994e-07,
"loss": 0.6049,
"step": 2421
},
{
"epoch": 0.9520440251572327,
"grad_norm": 0.8645806908607483,
"learning_rate": 2.4174528301886797e-07,
"loss": 0.7186,
"step": 2422
},
{
"epoch": 0.952437106918239,
"grad_norm": 0.7322193384170532,
"learning_rate": 2.397798742138365e-07,
"loss": 0.559,
"step": 2423
},
{
"epoch": 0.9528301886792453,
"grad_norm": 0.7243507504463196,
"learning_rate": 2.3781446540880508e-07,
"loss": 0.7239,
"step": 2424
},
{
"epoch": 0.9532232704402516,
"grad_norm": 0.6973174810409546,
"learning_rate": 2.358490566037736e-07,
"loss": 0.7207,
"step": 2425
},
{
"epoch": 0.9536163522012578,
"grad_norm": 0.6906088590621948,
"learning_rate": 2.3388364779874214e-07,
"loss": 0.6377,
"step": 2426
},
{
"epoch": 0.9540094339622641,
"grad_norm": 0.6723107695579529,
"learning_rate": 2.319182389937107e-07,
"loss": 0.6757,
"step": 2427
},
{
"epoch": 0.9544025157232704,
"grad_norm": 0.5140407681465149,
"learning_rate": 2.2995283018867925e-07,
"loss": 0.61,
"step": 2428
},
{
"epoch": 0.9547955974842768,
"grad_norm": 0.4904915392398834,
"learning_rate": 2.279874213836478e-07,
"loss": 0.6024,
"step": 2429
},
{
"epoch": 0.9551886792452831,
"grad_norm": 0.7497533559799194,
"learning_rate": 2.2602201257861636e-07,
"loss": 0.7081,
"step": 2430
},
{
"epoch": 0.9555817610062893,
"grad_norm": 0.6081426739692688,
"learning_rate": 2.2405660377358492e-07,
"loss": 0.7305,
"step": 2431
},
{
"epoch": 0.9559748427672956,
"grad_norm": 0.7061208486557007,
"learning_rate": 2.2209119496855347e-07,
"loss": 0.7689,
"step": 2432
},
{
"epoch": 0.9563679245283019,
"grad_norm": 0.9319342970848083,
"learning_rate": 2.2012578616352203e-07,
"loss": 0.8121,
"step": 2433
},
{
"epoch": 0.9567610062893082,
"grad_norm": 0.6911268830299377,
"learning_rate": 2.1816037735849058e-07,
"loss": 0.7218,
"step": 2434
},
{
"epoch": 0.9571540880503144,
"grad_norm": 0.9680249691009521,
"learning_rate": 2.161949685534591e-07,
"loss": 0.7939,
"step": 2435
},
{
"epoch": 0.9575471698113207,
"grad_norm": 0.8764474391937256,
"learning_rate": 2.1422955974842767e-07,
"loss": 0.6988,
"step": 2436
},
{
"epoch": 0.9579402515723271,
"grad_norm": 0.726374626159668,
"learning_rate": 2.1226415094339622e-07,
"loss": 0.6597,
"step": 2437
},
{
"epoch": 0.9583333333333334,
"grad_norm": 0.7107248306274414,
"learning_rate": 2.1029874213836478e-07,
"loss": 0.7306,
"step": 2438
},
{
"epoch": 0.9587264150943396,
"grad_norm": 0.896129310131073,
"learning_rate": 2.0833333333333333e-07,
"loss": 0.7371,
"step": 2439
},
{
"epoch": 0.9591194968553459,
"grad_norm": 0.7122219204902649,
"learning_rate": 2.063679245283019e-07,
"loss": 0.6177,
"step": 2440
},
{
"epoch": 0.9595125786163522,
"grad_norm": 0.7633679509162903,
"learning_rate": 2.0440251572327044e-07,
"loss": 0.629,
"step": 2441
},
{
"epoch": 0.9599056603773585,
"grad_norm": 0.8356642127037048,
"learning_rate": 2.02437106918239e-07,
"loss": 0.7682,
"step": 2442
},
{
"epoch": 0.9602987421383647,
"grad_norm": 0.9026091694831848,
"learning_rate": 2.0047169811320755e-07,
"loss": 0.7515,
"step": 2443
},
{
"epoch": 0.960691823899371,
"grad_norm": 0.6321382522583008,
"learning_rate": 1.985062893081761e-07,
"loss": 0.6265,
"step": 2444
},
{
"epoch": 0.9610849056603774,
"grad_norm": 0.4846247434616089,
"learning_rate": 1.9654088050314467e-07,
"loss": 0.5051,
"step": 2445
},
{
"epoch": 0.9614779874213837,
"grad_norm": 0.9835233092308044,
"learning_rate": 1.9457547169811322e-07,
"loss": 0.7893,
"step": 2446
},
{
"epoch": 0.96187106918239,
"grad_norm": 0.7191046476364136,
"learning_rate": 1.9261006289308178e-07,
"loss": 0.723,
"step": 2447
},
{
"epoch": 0.9622641509433962,
"grad_norm": 0.758245050907135,
"learning_rate": 1.9064465408805033e-07,
"loss": 0.6943,
"step": 2448
},
{
"epoch": 0.9626572327044025,
"grad_norm": 0.9750900864601135,
"learning_rate": 1.886792452830189e-07,
"loss": 0.7382,
"step": 2449
},
{
"epoch": 0.9630503144654088,
"grad_norm": 0.787466824054718,
"learning_rate": 1.8671383647798744e-07,
"loss": 0.7515,
"step": 2450
},
{
"epoch": 0.9634433962264151,
"grad_norm": 0.5801446437835693,
"learning_rate": 1.8474842767295597e-07,
"loss": 0.738,
"step": 2451
},
{
"epoch": 0.9638364779874213,
"grad_norm": 0.73136305809021,
"learning_rate": 1.8278301886792453e-07,
"loss": 0.6926,
"step": 2452
},
{
"epoch": 0.9642295597484277,
"grad_norm": 0.8923184275627136,
"learning_rate": 1.8081761006289308e-07,
"loss": 0.8289,
"step": 2453
},
{
"epoch": 0.964622641509434,
"grad_norm": 0.663041889667511,
"learning_rate": 1.7885220125786164e-07,
"loss": 0.6262,
"step": 2454
},
{
"epoch": 0.9650157232704403,
"grad_norm": 0.8111377954483032,
"learning_rate": 1.768867924528302e-07,
"loss": 0.6666,
"step": 2455
},
{
"epoch": 0.9654088050314465,
"grad_norm": 0.7093123197555542,
"learning_rate": 1.7492138364779875e-07,
"loss": 0.7781,
"step": 2456
},
{
"epoch": 0.9658018867924528,
"grad_norm": 0.7517929673194885,
"learning_rate": 1.729559748427673e-07,
"loss": 0.7445,
"step": 2457
},
{
"epoch": 0.9661949685534591,
"grad_norm": 0.8743553757667542,
"learning_rate": 1.7099056603773586e-07,
"loss": 0.7446,
"step": 2458
},
{
"epoch": 0.9665880503144654,
"grad_norm": 0.8092712759971619,
"learning_rate": 1.6902515723270442e-07,
"loss": 0.7897,
"step": 2459
},
{
"epoch": 0.9669811320754716,
"grad_norm": 1.052727222442627,
"learning_rate": 1.6705974842767297e-07,
"loss": 0.803,
"step": 2460
},
{
"epoch": 0.967374213836478,
"grad_norm": 0.6981686949729919,
"learning_rate": 1.6509433962264153e-07,
"loss": 0.6232,
"step": 2461
},
{
"epoch": 0.9677672955974843,
"grad_norm": 0.7388179898262024,
"learning_rate": 1.6312893081761008e-07,
"loss": 0.6827,
"step": 2462
},
{
"epoch": 0.9681603773584906,
"grad_norm": 0.5581703782081604,
"learning_rate": 1.6116352201257864e-07,
"loss": 0.5865,
"step": 2463
},
{
"epoch": 0.9685534591194969,
"grad_norm": 0.8329030275344849,
"learning_rate": 1.591981132075472e-07,
"loss": 0.6504,
"step": 2464
},
{
"epoch": 0.9689465408805031,
"grad_norm": 0.7310532331466675,
"learning_rate": 1.5723270440251575e-07,
"loss": 0.6884,
"step": 2465
},
{
"epoch": 0.9693396226415094,
"grad_norm": 0.7513861656188965,
"learning_rate": 1.5526729559748428e-07,
"loss": 0.7721,
"step": 2466
},
{
"epoch": 0.9697327044025157,
"grad_norm": 0.6153862476348877,
"learning_rate": 1.5330188679245283e-07,
"loss": 0.6087,
"step": 2467
},
{
"epoch": 0.970125786163522,
"grad_norm": 0.9130159020423889,
"learning_rate": 1.513364779874214e-07,
"loss": 0.6497,
"step": 2468
},
{
"epoch": 0.9705188679245284,
"grad_norm": 0.6084980368614197,
"learning_rate": 1.4937106918238994e-07,
"loss": 0.6271,
"step": 2469
},
{
"epoch": 0.9709119496855346,
"grad_norm": 0.7814450860023499,
"learning_rate": 1.474056603773585e-07,
"loss": 0.6664,
"step": 2470
},
{
"epoch": 0.9713050314465409,
"grad_norm": 0.6484072804450989,
"learning_rate": 1.4544025157232705e-07,
"loss": 0.6536,
"step": 2471
},
{
"epoch": 0.9716981132075472,
"grad_norm": 0.6341542601585388,
"learning_rate": 1.434748427672956e-07,
"loss": 0.6462,
"step": 2472
},
{
"epoch": 0.9720911949685535,
"grad_norm": 0.5989482402801514,
"learning_rate": 1.4150943396226417e-07,
"loss": 0.6156,
"step": 2473
},
{
"epoch": 0.9724842767295597,
"grad_norm": 0.8680738806724548,
"learning_rate": 1.3954402515723272e-07,
"loss": 0.7318,
"step": 2474
},
{
"epoch": 0.972877358490566,
"grad_norm": 0.5970274209976196,
"learning_rate": 1.3757861635220128e-07,
"loss": 0.6366,
"step": 2475
},
{
"epoch": 0.9732704402515723,
"grad_norm": 0.7139399647712708,
"learning_rate": 1.3561320754716983e-07,
"loss": 0.6981,
"step": 2476
},
{
"epoch": 0.9736635220125787,
"grad_norm": 0.9744982719421387,
"learning_rate": 1.336477987421384e-07,
"loss": 0.693,
"step": 2477
},
{
"epoch": 0.9740566037735849,
"grad_norm": 0.6138280034065247,
"learning_rate": 1.3168238993710694e-07,
"loss": 0.657,
"step": 2478
},
{
"epoch": 0.9744496855345912,
"grad_norm": 0.84993577003479,
"learning_rate": 1.297169811320755e-07,
"loss": 0.6833,
"step": 2479
},
{
"epoch": 0.9748427672955975,
"grad_norm": 0.897020161151886,
"learning_rate": 1.2775157232704405e-07,
"loss": 0.744,
"step": 2480
},
{
"epoch": 0.9752358490566038,
"grad_norm": 0.7882517576217651,
"learning_rate": 1.2578616352201258e-07,
"loss": 0.694,
"step": 2481
},
{
"epoch": 0.97562893081761,
"grad_norm": 0.791140079498291,
"learning_rate": 1.2382075471698114e-07,
"loss": 0.7955,
"step": 2482
},
{
"epoch": 0.9760220125786163,
"grad_norm": 0.6419038772583008,
"learning_rate": 1.218553459119497e-07,
"loss": 0.7024,
"step": 2483
},
{
"epoch": 0.9764150943396226,
"grad_norm": 0.6117247939109802,
"learning_rate": 1.1988993710691825e-07,
"loss": 0.6097,
"step": 2484
},
{
"epoch": 0.976808176100629,
"grad_norm": 0.8920881152153015,
"learning_rate": 1.179245283018868e-07,
"loss": 0.7422,
"step": 2485
},
{
"epoch": 0.9772012578616353,
"grad_norm": 0.8514476418495178,
"learning_rate": 1.1595911949685535e-07,
"loss": 0.6788,
"step": 2486
},
{
"epoch": 0.9775943396226415,
"grad_norm": 0.6733038425445557,
"learning_rate": 1.139937106918239e-07,
"loss": 0.6463,
"step": 2487
},
{
"epoch": 0.9779874213836478,
"grad_norm": 0.884686291217804,
"learning_rate": 1.1202830188679246e-07,
"loss": 0.7595,
"step": 2488
},
{
"epoch": 0.9783805031446541,
"grad_norm": 0.782560408115387,
"learning_rate": 1.1006289308176101e-07,
"loss": 0.8177,
"step": 2489
},
{
"epoch": 0.9787735849056604,
"grad_norm": 0.5646007657051086,
"learning_rate": 1.0809748427672956e-07,
"loss": 0.6177,
"step": 2490
},
{
"epoch": 0.9791666666666666,
"grad_norm": 0.6456993222236633,
"learning_rate": 1.0613207547169811e-07,
"loss": 0.6493,
"step": 2491
},
{
"epoch": 0.9795597484276729,
"grad_norm": 0.747462272644043,
"learning_rate": 1.0416666666666667e-07,
"loss": 0.7115,
"step": 2492
},
{
"epoch": 0.9799528301886793,
"grad_norm": 0.6056180000305176,
"learning_rate": 1.0220125786163522e-07,
"loss": 0.7197,
"step": 2493
},
{
"epoch": 0.9803459119496856,
"grad_norm": 0.8783689737319946,
"learning_rate": 1.0023584905660378e-07,
"loss": 0.6714,
"step": 2494
},
{
"epoch": 0.9807389937106918,
"grad_norm": 0.6365591883659363,
"learning_rate": 9.827044025157233e-08,
"loss": 0.6168,
"step": 2495
},
{
"epoch": 0.9811320754716981,
"grad_norm": 0.6840308904647827,
"learning_rate": 9.630503144654089e-08,
"loss": 0.5951,
"step": 2496
},
{
"epoch": 0.9815251572327044,
"grad_norm": 0.6192978024482727,
"learning_rate": 9.433962264150944e-08,
"loss": 0.5952,
"step": 2497
},
{
"epoch": 0.9819182389937107,
"grad_norm": 0.7644177675247192,
"learning_rate": 9.237421383647799e-08,
"loss": 0.6479,
"step": 2498
},
{
"epoch": 0.9823113207547169,
"grad_norm": 0.8905887603759766,
"learning_rate": 9.040880503144654e-08,
"loss": 0.6678,
"step": 2499
},
{
"epoch": 0.9827044025157232,
"grad_norm": 0.6398415565490723,
"learning_rate": 8.84433962264151e-08,
"loss": 0.629,
"step": 2500
},
{
"epoch": 0.9830974842767296,
"grad_norm": 0.8488085269927979,
"learning_rate": 8.647798742138365e-08,
"loss": 0.6816,
"step": 2501
},
{
"epoch": 0.9834905660377359,
"grad_norm": 0.6772359609603882,
"learning_rate": 8.451257861635221e-08,
"loss": 0.6461,
"step": 2502
},
{
"epoch": 0.9838836477987422,
"grad_norm": 0.6087149977684021,
"learning_rate": 8.254716981132076e-08,
"loss": 0.6775,
"step": 2503
},
{
"epoch": 0.9842767295597484,
"grad_norm": 0.5166673064231873,
"learning_rate": 8.058176100628932e-08,
"loss": 0.6442,
"step": 2504
},
{
"epoch": 0.9846698113207547,
"grad_norm": 0.6319252848625183,
"learning_rate": 7.861635220125787e-08,
"loss": 0.7473,
"step": 2505
},
{
"epoch": 0.985062893081761,
"grad_norm": 0.6890890002250671,
"learning_rate": 7.665094339622642e-08,
"loss": 0.7067,
"step": 2506
},
{
"epoch": 0.9854559748427673,
"grad_norm": 0.7080012559890747,
"learning_rate": 7.468553459119497e-08,
"loss": 0.7562,
"step": 2507
},
{
"epoch": 0.9858490566037735,
"grad_norm": 0.8671581745147705,
"learning_rate": 7.272012578616353e-08,
"loss": 0.7545,
"step": 2508
},
{
"epoch": 0.9862421383647799,
"grad_norm": 0.7419705390930176,
"learning_rate": 7.075471698113208e-08,
"loss": 0.6943,
"step": 2509
},
{
"epoch": 0.9866352201257862,
"grad_norm": 1.2231389284133911,
"learning_rate": 6.878930817610064e-08,
"loss": 0.6819,
"step": 2510
},
{
"epoch": 0.9870283018867925,
"grad_norm": 0.48727840185165405,
"learning_rate": 6.68238993710692e-08,
"loss": 0.5791,
"step": 2511
},
{
"epoch": 0.9874213836477987,
"grad_norm": 0.7134295105934143,
"learning_rate": 6.485849056603775e-08,
"loss": 0.5671,
"step": 2512
},
{
"epoch": 0.987814465408805,
"grad_norm": 0.7744563817977905,
"learning_rate": 6.289308176100629e-08,
"loss": 0.7228,
"step": 2513
},
{
"epoch": 0.9882075471698113,
"grad_norm": 0.6920917630195618,
"learning_rate": 6.092767295597485e-08,
"loss": 0.6251,
"step": 2514
},
{
"epoch": 0.9886006289308176,
"grad_norm": 0.9543319344520569,
"learning_rate": 5.89622641509434e-08,
"loss": 0.7806,
"step": 2515
},
{
"epoch": 0.9889937106918238,
"grad_norm": 0.7128175497055054,
"learning_rate": 5.699685534591195e-08,
"loss": 0.6957,
"step": 2516
},
{
"epoch": 0.9893867924528302,
"grad_norm": 0.9007143378257751,
"learning_rate": 5.5031446540880507e-08,
"loss": 0.728,
"step": 2517
},
{
"epoch": 0.9897798742138365,
"grad_norm": 0.7549263834953308,
"learning_rate": 5.3066037735849055e-08,
"loss": 0.6715,
"step": 2518
},
{
"epoch": 0.9901729559748428,
"grad_norm": 0.500179648399353,
"learning_rate": 5.110062893081761e-08,
"loss": 0.6453,
"step": 2519
},
{
"epoch": 0.9905660377358491,
"grad_norm": 0.702854573726654,
"learning_rate": 4.9135220125786166e-08,
"loss": 0.7889,
"step": 2520
},
{
"epoch": 0.9909591194968553,
"grad_norm": 0.6856566071510315,
"learning_rate": 4.716981132075472e-08,
"loss": 0.7197,
"step": 2521
},
{
"epoch": 0.9913522012578616,
"grad_norm": 0.8665456175804138,
"learning_rate": 4.520440251572327e-08,
"loss": 0.7267,
"step": 2522
},
{
"epoch": 0.9917452830188679,
"grad_norm": 0.5870963335037231,
"learning_rate": 4.3238993710691826e-08,
"loss": 0.5935,
"step": 2523
},
{
"epoch": 0.9921383647798742,
"grad_norm": 0.8170610070228577,
"learning_rate": 4.127358490566038e-08,
"loss": 0.5768,
"step": 2524
},
{
"epoch": 0.9925314465408805,
"grad_norm": 0.6386653780937195,
"learning_rate": 3.930817610062894e-08,
"loss": 0.7142,
"step": 2525
},
{
"epoch": 0.9929245283018868,
"grad_norm": 0.5451639294624329,
"learning_rate": 3.7342767295597486e-08,
"loss": 0.6298,
"step": 2526
},
{
"epoch": 0.9933176100628931,
"grad_norm": 0.7032883167266846,
"learning_rate": 3.537735849056604e-08,
"loss": 0.7181,
"step": 2527
},
{
"epoch": 0.9937106918238994,
"grad_norm": 0.9108713865280151,
"learning_rate": 3.34119496855346e-08,
"loss": 0.7771,
"step": 2528
},
{
"epoch": 0.9941037735849056,
"grad_norm": 0.7550582885742188,
"learning_rate": 3.1446540880503146e-08,
"loss": 0.6689,
"step": 2529
},
{
"epoch": 0.9944968553459119,
"grad_norm": 0.5265597701072693,
"learning_rate": 2.94811320754717e-08,
"loss": 0.5793,
"step": 2530
},
{
"epoch": 0.9948899371069182,
"grad_norm": 0.7188118696212769,
"learning_rate": 2.7515723270440253e-08,
"loss": 0.6039,
"step": 2531
},
{
"epoch": 0.9952830188679245,
"grad_norm": 0.6250026226043701,
"learning_rate": 2.5550314465408805e-08,
"loss": 0.6548,
"step": 2532
},
{
"epoch": 0.9956761006289309,
"grad_norm": 0.46753814816474915,
"learning_rate": 2.358490566037736e-08,
"loss": 0.5419,
"step": 2533
},
{
"epoch": 0.9960691823899371,
"grad_norm": 0.9113474488258362,
"learning_rate": 2.1619496855345913e-08,
"loss": 0.6825,
"step": 2534
},
{
"epoch": 0.9964622641509434,
"grad_norm": 0.8646556735038757,
"learning_rate": 1.965408805031447e-08,
"loss": 0.7331,
"step": 2535
},
{
"epoch": 0.9968553459119497,
"grad_norm": 0.47674545645713806,
"learning_rate": 1.768867924528302e-08,
"loss": 0.6338,
"step": 2536
},
{
"epoch": 0.997248427672956,
"grad_norm": 0.7165384888648987,
"learning_rate": 1.5723270440251573e-08,
"loss": 0.6193,
"step": 2537
},
{
"epoch": 0.9976415094339622,
"grad_norm": 0.5947349071502686,
"learning_rate": 1.3757861635220127e-08,
"loss": 0.6346,
"step": 2538
},
{
"epoch": 0.9980345911949685,
"grad_norm": 0.5829725861549377,
"learning_rate": 1.179245283018868e-08,
"loss": 0.6309,
"step": 2539
},
{
"epoch": 0.9984276729559748,
"grad_norm": 0.6105396151542664,
"learning_rate": 9.827044025157234e-09,
"loss": 0.7139,
"step": 2540
},
{
"epoch": 0.9988207547169812,
"grad_norm": 0.741215705871582,
"learning_rate": 7.861635220125786e-09,
"loss": 0.7444,
"step": 2541
},
{
"epoch": 0.9992138364779874,
"grad_norm": 0.6117772459983826,
"learning_rate": 5.89622641509434e-09,
"loss": 0.699,
"step": 2542
},
{
"epoch": 0.9996069182389937,
"grad_norm": 0.8725244998931885,
"learning_rate": 3.930817610062893e-09,
"loss": 0.7484,
"step": 2543
},
{
"epoch": 1.0,
"grad_norm": 0.5320550203323364,
"learning_rate": 1.9654088050314466e-09,
"loss": 0.4575,
"step": 2544
}
],
"logging_steps": 1.0,
"max_steps": 2544,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 0,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.3749063616638747e+19,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}