Steven668866's picture
upload checkpoint-2000/trainer_state.json
a2faf3d verified
Raw
History Blame Contribute Delete
351 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.017987795280901907,
"eval_steps": 500,
"global_step": 2000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 8.993897640450955e-06,
"grad_norm": 624.5777587890625,
"learning_rate": 0.0,
"loss": 17.9534,
"step": 1
},
{
"epoch": 1.798779528090191e-05,
"grad_norm": 1787.1439208984375,
"learning_rate": 9.090909090909091e-08,
"loss": 17.9833,
"step": 2
},
{
"epoch": 2.6981692921352862e-05,
"grad_norm": 543.7571411132812,
"learning_rate": 1.8181818181818183e-07,
"loss": 17.4185,
"step": 3
},
{
"epoch": 3.597559056180382e-05,
"grad_norm": 1464.9036865234375,
"learning_rate": 2.7272727272727274e-07,
"loss": 17.0435,
"step": 4
},
{
"epoch": 4.496948820225477e-05,
"grad_norm": 433.85675048828125,
"learning_rate": 3.6363636363636366e-07,
"loss": 17.6478,
"step": 5
},
{
"epoch": 5.3963385842705725e-05,
"grad_norm": 618.152587890625,
"learning_rate": 4.545454545454545e-07,
"loss": 17.6259,
"step": 6
},
{
"epoch": 6.295728348315668e-05,
"grad_norm": 723.3334350585938,
"learning_rate": 5.454545454545455e-07,
"loss": 18.6397,
"step": 7
},
{
"epoch": 7.195118112360764e-05,
"grad_norm": 842.21875,
"learning_rate": 6.363636363636364e-07,
"loss": 17.0937,
"step": 8
},
{
"epoch": 8.094507876405858e-05,
"grad_norm": 575.5238647460938,
"learning_rate": 7.272727272727273e-07,
"loss": 17.2475,
"step": 9
},
{
"epoch": 8.993897640450954e-05,
"grad_norm": 1246.06787109375,
"learning_rate": 8.181818181818183e-07,
"loss": 17.6788,
"step": 10
},
{
"epoch": 9.89328740449605e-05,
"grad_norm": 703.3920288085938,
"learning_rate": 9.09090909090909e-07,
"loss": 17.6445,
"step": 11
},
{
"epoch": 0.00010792677168541145,
"grad_norm": 789.861328125,
"learning_rate": 1.0000000000000002e-06,
"loss": 18.0761,
"step": 12
},
{
"epoch": 0.0001169206693258624,
"grad_norm": 1688.398193359375,
"learning_rate": 1.090909090909091e-06,
"loss": 17.1804,
"step": 13
},
{
"epoch": 0.00012591456696631336,
"grad_norm": 970.5971069335938,
"learning_rate": 1.1818181818181818e-06,
"loss": 17.8007,
"step": 14
},
{
"epoch": 0.00013490846460676432,
"grad_norm": 484.73492431640625,
"learning_rate": 1.2727272727272728e-06,
"loss": 17.3304,
"step": 15
},
{
"epoch": 0.00014390236224721528,
"grad_norm": 1148.683349609375,
"learning_rate": 1.3636363636363636e-06,
"loss": 18.1804,
"step": 16
},
{
"epoch": 0.0001528962598876662,
"grad_norm": 1276.140380859375,
"learning_rate": 1.4545454545454546e-06,
"loss": 17.8103,
"step": 17
},
{
"epoch": 0.00016189015752811716,
"grad_norm": 1091.2808837890625,
"learning_rate": 1.5454545454545457e-06,
"loss": 17.7373,
"step": 18
},
{
"epoch": 0.00017088405516856812,
"grad_norm": 1217.447021484375,
"learning_rate": 1.6363636363636367e-06,
"loss": 17.1021,
"step": 19
},
{
"epoch": 0.00017987795280901907,
"grad_norm": 1050.6259765625,
"learning_rate": 1.7272727272727273e-06,
"loss": 17.8152,
"step": 20
},
{
"epoch": 0.00018887185044947003,
"grad_norm": 1252.3167724609375,
"learning_rate": 1.818181818181818e-06,
"loss": 17.0886,
"step": 21
},
{
"epoch": 0.000197865748089921,
"grad_norm": 1881.7572021484375,
"learning_rate": 1.9090909090909095e-06,
"loss": 16.6945,
"step": 22
},
{
"epoch": 0.00020685964573037194,
"grad_norm": 969.9746704101562,
"learning_rate": 2.0000000000000003e-06,
"loss": 17.8223,
"step": 23
},
{
"epoch": 0.0002158535433708229,
"grad_norm": 1115.3941650390625,
"learning_rate": 2.090909090909091e-06,
"loss": 16.8948,
"step": 24
},
{
"epoch": 0.00022484744101127386,
"grad_norm": 1183.36328125,
"learning_rate": 2.181818181818182e-06,
"loss": 16.6381,
"step": 25
},
{
"epoch": 0.0002338413386517248,
"grad_norm": 530.5995483398438,
"learning_rate": 2.2727272727272728e-06,
"loss": 17.2105,
"step": 26
},
{
"epoch": 0.00024283523629217577,
"grad_norm": 1587.966552734375,
"learning_rate": 2.3636363636363636e-06,
"loss": 17.1436,
"step": 27
},
{
"epoch": 0.0002518291339326267,
"grad_norm": 523.8456420898438,
"learning_rate": 2.4545454545454544e-06,
"loss": 17.1746,
"step": 28
},
{
"epoch": 0.0002608230315730777,
"grad_norm": 452.8838806152344,
"learning_rate": 2.5454545454545456e-06,
"loss": 16.8128,
"step": 29
},
{
"epoch": 0.00026981692921352864,
"grad_norm": 1319.495849609375,
"learning_rate": 2.6363636363636364e-06,
"loss": 17.1558,
"step": 30
},
{
"epoch": 0.0002788108268539796,
"grad_norm": 343.67791748046875,
"learning_rate": 2.7272727272727272e-06,
"loss": 16.7303,
"step": 31
},
{
"epoch": 0.00028780472449443055,
"grad_norm": 670.576904296875,
"learning_rate": 2.8181818181818185e-06,
"loss": 16.2864,
"step": 32
},
{
"epoch": 0.0002967986221348815,
"grad_norm": 586.6473388671875,
"learning_rate": 2.9090909090909093e-06,
"loss": 16.3865,
"step": 33
},
{
"epoch": 0.0003057925197753324,
"grad_norm": 401.2021789550781,
"learning_rate": 3e-06,
"loss": 16.2615,
"step": 34
},
{
"epoch": 0.00031478641741578337,
"grad_norm": 550.7803955078125,
"learning_rate": 3.0909090909090913e-06,
"loss": 16.0867,
"step": 35
},
{
"epoch": 0.0003237803150562343,
"grad_norm": 449.146728515625,
"learning_rate": 3.1818181818181817e-06,
"loss": 16.7781,
"step": 36
},
{
"epoch": 0.0003327742126966853,
"grad_norm": 236.89505004882812,
"learning_rate": 3.2727272727272733e-06,
"loss": 16.7936,
"step": 37
},
{
"epoch": 0.00034176811033713624,
"grad_norm": 260.0072021484375,
"learning_rate": 3.363636363636364e-06,
"loss": 16.7107,
"step": 38
},
{
"epoch": 0.0003507620079775872,
"grad_norm": 362.1142578125,
"learning_rate": 3.4545454545454545e-06,
"loss": 15.8317,
"step": 39
},
{
"epoch": 0.00035975590561803815,
"grad_norm": 176.87826538085938,
"learning_rate": 3.5454545454545454e-06,
"loss": 16.213,
"step": 40
},
{
"epoch": 0.0003687498032584891,
"grad_norm": 311.5877685546875,
"learning_rate": 3.636363636363636e-06,
"loss": 17.5375,
"step": 41
},
{
"epoch": 0.00037774370089894006,
"grad_norm": 157.58517456054688,
"learning_rate": 3.727272727272727e-06,
"loss": 15.7588,
"step": 42
},
{
"epoch": 0.000386737598539391,
"grad_norm": 180.6936492919922,
"learning_rate": 3.818181818181819e-06,
"loss": 16.0854,
"step": 43
},
{
"epoch": 0.000395731496179842,
"grad_norm": 256.3274841308594,
"learning_rate": 3.90909090909091e-06,
"loss": 15.3464,
"step": 44
},
{
"epoch": 0.00040472539382029293,
"grad_norm": 831.1060791015625,
"learning_rate": 4.000000000000001e-06,
"loss": 16.4678,
"step": 45
},
{
"epoch": 0.0004137192914607439,
"grad_norm": 295.3250732421875,
"learning_rate": 4.0909090909090915e-06,
"loss": 16.0141,
"step": 46
},
{
"epoch": 0.00042271318910119484,
"grad_norm": 260.3460388183594,
"learning_rate": 4.181818181818182e-06,
"loss": 16.0095,
"step": 47
},
{
"epoch": 0.0004317070867416458,
"grad_norm": 753.9771118164062,
"learning_rate": 4.272727272727273e-06,
"loss": 15.0963,
"step": 48
},
{
"epoch": 0.00044070098438209676,
"grad_norm": 602.2676391601562,
"learning_rate": 4.363636363636364e-06,
"loss": 15.3355,
"step": 49
},
{
"epoch": 0.0004496948820225477,
"grad_norm": 550.3294677734375,
"learning_rate": 4.454545454545455e-06,
"loss": 15.6943,
"step": 50
},
{
"epoch": 0.00045868877966299867,
"grad_norm": 525.8038330078125,
"learning_rate": 4.5454545454545455e-06,
"loss": 16.3719,
"step": 51
},
{
"epoch": 0.0004676826773034496,
"grad_norm": 331.14923095703125,
"learning_rate": 4.636363636363636e-06,
"loss": 15.1211,
"step": 52
},
{
"epoch": 0.0004766765749439006,
"grad_norm": 355.77093505859375,
"learning_rate": 4.727272727272727e-06,
"loss": 15.1063,
"step": 53
},
{
"epoch": 0.00048567047258435154,
"grad_norm": 3031.103515625,
"learning_rate": 4.818181818181818e-06,
"loss": 15.4299,
"step": 54
},
{
"epoch": 0.0004946643702248024,
"grad_norm": 418.1381530761719,
"learning_rate": 4.909090909090909e-06,
"loss": 15.4191,
"step": 55
},
{
"epoch": 0.0005036582678652535,
"grad_norm": 911.3734130859375,
"learning_rate": 5e-06,
"loss": 15.5616,
"step": 56
},
{
"epoch": 0.0005126521655057044,
"grad_norm": 185.1163330078125,
"learning_rate": 5.090909090909091e-06,
"loss": 14.8612,
"step": 57
},
{
"epoch": 0.0005216460631461554,
"grad_norm": 627.5923461914062,
"learning_rate": 5.181818181818182e-06,
"loss": 14.6524,
"step": 58
},
{
"epoch": 0.0005306399607866063,
"grad_norm": 139.53121948242188,
"learning_rate": 5.272727272727273e-06,
"loss": 15.3519,
"step": 59
},
{
"epoch": 0.0005396338584270573,
"grad_norm": 830.2334594726562,
"learning_rate": 5.363636363636364e-06,
"loss": 14.6705,
"step": 60
},
{
"epoch": 0.0005486277560675082,
"grad_norm": 1547.2772216796875,
"learning_rate": 5.4545454545454545e-06,
"loss": 14.7645,
"step": 61
},
{
"epoch": 0.0005576216537079592,
"grad_norm": 1675.0274658203125,
"learning_rate": 5.545454545454546e-06,
"loss": 14.1208,
"step": 62
},
{
"epoch": 0.0005666155513484101,
"grad_norm": 405.97027587890625,
"learning_rate": 5.636363636363637e-06,
"loss": 15.0928,
"step": 63
},
{
"epoch": 0.0005756094489888611,
"grad_norm": 611.3431396484375,
"learning_rate": 5.727272727272728e-06,
"loss": 14.0582,
"step": 64
},
{
"epoch": 0.000584603346629312,
"grad_norm": 3743.0625,
"learning_rate": 5.8181818181818185e-06,
"loss": 14.564,
"step": 65
},
{
"epoch": 0.000593597244269763,
"grad_norm": 2080.346923828125,
"learning_rate": 5.909090909090909e-06,
"loss": 15.3672,
"step": 66
},
{
"epoch": 0.0006025911419102139,
"grad_norm": 847.2843627929688,
"learning_rate": 6e-06,
"loss": 13.6193,
"step": 67
},
{
"epoch": 0.0006115850395506648,
"grad_norm": 710.4701538085938,
"learning_rate": 6.090909090909091e-06,
"loss": 13.9402,
"step": 68
},
{
"epoch": 0.0006205789371911158,
"grad_norm": 565.0715942382812,
"learning_rate": 6.181818181818183e-06,
"loss": 15.2938,
"step": 69
},
{
"epoch": 0.0006295728348315667,
"grad_norm": 715.6685180664062,
"learning_rate": 6.2727272727272734e-06,
"loss": 14.6494,
"step": 70
},
{
"epoch": 0.0006385667324720177,
"grad_norm": 686.0399780273438,
"learning_rate": 6.363636363636363e-06,
"loss": 13.6058,
"step": 71
},
{
"epoch": 0.0006475606301124686,
"grad_norm": 903.3383178710938,
"learning_rate": 6.454545454545455e-06,
"loss": 14.07,
"step": 72
},
{
"epoch": 0.0006565545277529197,
"grad_norm": 281.6870422363281,
"learning_rate": 6.545454545454547e-06,
"loss": 14.7523,
"step": 73
},
{
"epoch": 0.0006655484253933706,
"grad_norm": 806.2594604492188,
"learning_rate": 6.636363636363637e-06,
"loss": 12.7969,
"step": 74
},
{
"epoch": 0.0006745423230338216,
"grad_norm": 399.9375,
"learning_rate": 6.727272727272728e-06,
"loss": 14.0856,
"step": 75
},
{
"epoch": 0.0006835362206742725,
"grad_norm": 641.3349609375,
"learning_rate": 6.818181818181818e-06,
"loss": 12.93,
"step": 76
},
{
"epoch": 0.0006925301183147235,
"grad_norm": 395.2854309082031,
"learning_rate": 6.909090909090909e-06,
"loss": 13.2835,
"step": 77
},
{
"epoch": 0.0007015240159551744,
"grad_norm": 838.3772583007812,
"learning_rate": 7.000000000000001e-06,
"loss": 12.6786,
"step": 78
},
{
"epoch": 0.0007105179135956254,
"grad_norm": 196.3134002685547,
"learning_rate": 7.090909090909091e-06,
"loss": 12.5385,
"step": 79
},
{
"epoch": 0.0007195118112360763,
"grad_norm": 1469.98046875,
"learning_rate": 7.181818181818182e-06,
"loss": 12.5102,
"step": 80
},
{
"epoch": 0.0007285057088765273,
"grad_norm": 538.8250122070312,
"learning_rate": 7.272727272727272e-06,
"loss": 12.4218,
"step": 81
},
{
"epoch": 0.0007374996065169782,
"grad_norm": 384.11102294921875,
"learning_rate": 7.363636363636364e-06,
"loss": 12.2247,
"step": 82
},
{
"epoch": 0.0007464935041574292,
"grad_norm": 225.18310546875,
"learning_rate": 7.454545454545454e-06,
"loss": 11.9204,
"step": 83
},
{
"epoch": 0.0007554874017978801,
"grad_norm": 209.81674194335938,
"learning_rate": 7.545454545454546e-06,
"loss": 12.9126,
"step": 84
},
{
"epoch": 0.0007644812994383311,
"grad_norm": 237.86912536621094,
"learning_rate": 7.636363636363638e-06,
"loss": 11.8694,
"step": 85
},
{
"epoch": 0.000773475197078782,
"grad_norm": 231.4341278076172,
"learning_rate": 7.727272727272727e-06,
"loss": 11.5688,
"step": 86
},
{
"epoch": 0.000782469094719233,
"grad_norm": 245.27500915527344,
"learning_rate": 7.81818181818182e-06,
"loss": 11.2395,
"step": 87
},
{
"epoch": 0.000791462992359684,
"grad_norm": 375.0210876464844,
"learning_rate": 7.909090909090909e-06,
"loss": 11.0175,
"step": 88
},
{
"epoch": 0.000800456890000135,
"grad_norm": 345.97308349609375,
"learning_rate": 8.000000000000001e-06,
"loss": 10.774,
"step": 89
},
{
"epoch": 0.0008094507876405859,
"grad_norm": 206.9285125732422,
"learning_rate": 8.09090909090909e-06,
"loss": 10.9838,
"step": 90
},
{
"epoch": 0.0008184446852810368,
"grad_norm": 290.99908447265625,
"learning_rate": 8.181818181818183e-06,
"loss": 10.8066,
"step": 91
},
{
"epoch": 0.0008274385829214878,
"grad_norm": 455.3750915527344,
"learning_rate": 8.272727272727274e-06,
"loss": 10.1416,
"step": 92
},
{
"epoch": 0.0008364324805619387,
"grad_norm": 345.7048645019531,
"learning_rate": 8.363636363636365e-06,
"loss": 10.3245,
"step": 93
},
{
"epoch": 0.0008454263782023897,
"grad_norm": 171.3020782470703,
"learning_rate": 8.454545454545455e-06,
"loss": 9.6277,
"step": 94
},
{
"epoch": 0.0008544202758428406,
"grad_norm": 267.947265625,
"learning_rate": 8.545454545454546e-06,
"loss": 10.3132,
"step": 95
},
{
"epoch": 0.0008634141734832916,
"grad_norm": 387.4131164550781,
"learning_rate": 8.636363636363637e-06,
"loss": 10.5936,
"step": 96
},
{
"epoch": 0.0008724080711237425,
"grad_norm": 417.14166259765625,
"learning_rate": 8.727272727272728e-06,
"loss": 9.4039,
"step": 97
},
{
"epoch": 0.0008814019687641935,
"grad_norm": 212.96841430664062,
"learning_rate": 8.818181818181819e-06,
"loss": 8.9306,
"step": 98
},
{
"epoch": 0.0008903958664046444,
"grad_norm": 179.31338500976562,
"learning_rate": 8.90909090909091e-06,
"loss": 8.1835,
"step": 99
},
{
"epoch": 0.0008993897640450954,
"grad_norm": 259.7271423339844,
"learning_rate": 9e-06,
"loss": 9.3655,
"step": 100
},
{
"epoch": 0.0009083836616855463,
"grad_norm": 553.3068237304688,
"learning_rate": 9.090909090909091e-06,
"loss": 7.9431,
"step": 101
},
{
"epoch": 0.0009173775593259973,
"grad_norm": 1172.50341796875,
"learning_rate": 9.181818181818182e-06,
"loss": 7.8488,
"step": 102
},
{
"epoch": 0.0009263714569664482,
"grad_norm": 241.0426025390625,
"learning_rate": 9.272727272727273e-06,
"loss": 6.5786,
"step": 103
},
{
"epoch": 0.0009353653546068993,
"grad_norm": 540.9775390625,
"learning_rate": 9.363636363636365e-06,
"loss": 8.4175,
"step": 104
},
{
"epoch": 0.0009443592522473502,
"grad_norm": 241.18971252441406,
"learning_rate": 9.454545454545454e-06,
"loss": 6.1545,
"step": 105
},
{
"epoch": 0.0009533531498878012,
"grad_norm": 370.6540222167969,
"learning_rate": 9.545454545454547e-06,
"loss": 5.4816,
"step": 106
},
{
"epoch": 0.0009623470475282521,
"grad_norm": 316.3355712890625,
"learning_rate": 9.636363636363636e-06,
"loss": 5.4708,
"step": 107
},
{
"epoch": 0.0009713409451687031,
"grad_norm": 335.52239990234375,
"learning_rate": 9.727272727272728e-06,
"loss": 5.9066,
"step": 108
},
{
"epoch": 0.000980334842809154,
"grad_norm": 212.37066650390625,
"learning_rate": 9.818181818181818e-06,
"loss": 6.5996,
"step": 109
},
{
"epoch": 0.0009893287404496049,
"grad_norm": 158.48117065429688,
"learning_rate": 9.90909090909091e-06,
"loss": 5.0258,
"step": 110
},
{
"epoch": 0.000998322638090056,
"grad_norm": 406.5316162109375,
"learning_rate": 1e-05,
"loss": 5.2208,
"step": 111
},
{
"epoch": 0.001007316535730507,
"grad_norm": 220.3340301513672,
"learning_rate": 1.0090909090909092e-05,
"loss": 5.7037,
"step": 112
},
{
"epoch": 0.001016310433370958,
"grad_norm": 117.0799789428711,
"learning_rate": 1.0181818181818182e-05,
"loss": 4.5769,
"step": 113
},
{
"epoch": 0.0010253043310114087,
"grad_norm": 86.51113891601562,
"learning_rate": 1.0272727272727273e-05,
"loss": 4.5308,
"step": 114
},
{
"epoch": 0.0010342982286518597,
"grad_norm": 119.74539184570312,
"learning_rate": 1.0363636363636364e-05,
"loss": 5.0915,
"step": 115
},
{
"epoch": 0.0010432921262923107,
"grad_norm": 61.35383224487305,
"learning_rate": 1.0454545454545455e-05,
"loss": 4.4312,
"step": 116
},
{
"epoch": 0.0010522860239327615,
"grad_norm": 58.55775451660156,
"learning_rate": 1.0545454545454546e-05,
"loss": 4.2491,
"step": 117
},
{
"epoch": 0.0010612799215732125,
"grad_norm": 120.29472351074219,
"learning_rate": 1.0636363636363638e-05,
"loss": 4.385,
"step": 118
},
{
"epoch": 0.0010702738192136635,
"grad_norm": 73.53257751464844,
"learning_rate": 1.0727272727272727e-05,
"loss": 4.9063,
"step": 119
},
{
"epoch": 0.0010792677168541146,
"grad_norm": 110.20689392089844,
"learning_rate": 1.081818181818182e-05,
"loss": 4.8728,
"step": 120
},
{
"epoch": 0.0010882616144945653,
"grad_norm": 85.38801574707031,
"learning_rate": 1.0909090909090909e-05,
"loss": 5.0187,
"step": 121
},
{
"epoch": 0.0010972555121350164,
"grad_norm": 153.42276000976562,
"learning_rate": 1.1000000000000001e-05,
"loss": 4.186,
"step": 122
},
{
"epoch": 0.0011062494097754674,
"grad_norm": 64.13314819335938,
"learning_rate": 1.1090909090909092e-05,
"loss": 4.0873,
"step": 123
},
{
"epoch": 0.0011152433074159184,
"grad_norm": 72.71946716308594,
"learning_rate": 1.1181818181818183e-05,
"loss": 4.5826,
"step": 124
},
{
"epoch": 0.0011242372050563692,
"grad_norm": 157.0845947265625,
"learning_rate": 1.1272727272727274e-05,
"loss": 4.8166,
"step": 125
},
{
"epoch": 0.0011332311026968202,
"grad_norm": 87.81234741210938,
"learning_rate": 1.1363636363636365e-05,
"loss": 4.6643,
"step": 126
},
{
"epoch": 0.0011422250003372712,
"grad_norm": 77.61016845703125,
"learning_rate": 1.1454545454545455e-05,
"loss": 4.7834,
"step": 127
},
{
"epoch": 0.0011512188979777222,
"grad_norm": 92.09603118896484,
"learning_rate": 1.1545454545454545e-05,
"loss": 5.0563,
"step": 128
},
{
"epoch": 0.001160212795618173,
"grad_norm": 72.69209289550781,
"learning_rate": 1.1636363636363637e-05,
"loss": 3.8202,
"step": 129
},
{
"epoch": 0.001169206693258624,
"grad_norm": 62.451534271240234,
"learning_rate": 1.1727272727272728e-05,
"loss": 3.8198,
"step": 130
},
{
"epoch": 0.001178200590899075,
"grad_norm": 136.73240661621094,
"learning_rate": 1.1818181818181819e-05,
"loss": 4.7253,
"step": 131
},
{
"epoch": 0.001187194488539526,
"grad_norm": 186.7635040283203,
"learning_rate": 1.190909090909091e-05,
"loss": 4.0171,
"step": 132
},
{
"epoch": 0.0011961883861799768,
"grad_norm": 150.6204833984375,
"learning_rate": 1.2e-05,
"loss": 3.9884,
"step": 133
},
{
"epoch": 0.0012051822838204278,
"grad_norm": 144.12637329101562,
"learning_rate": 1.2090909090909091e-05,
"loss": 3.6953,
"step": 134
},
{
"epoch": 0.0012141761814608788,
"grad_norm": 66.29656982421875,
"learning_rate": 1.2181818181818182e-05,
"loss": 4.5341,
"step": 135
},
{
"epoch": 0.0012231700791013296,
"grad_norm": 122.22705078125,
"learning_rate": 1.2272727272727273e-05,
"loss": 5.4335,
"step": 136
},
{
"epoch": 0.0012321639767417807,
"grad_norm": 120.67552185058594,
"learning_rate": 1.2363636363636365e-05,
"loss": 3.7019,
"step": 137
},
{
"epoch": 0.0012411578743822317,
"grad_norm": 60.150630950927734,
"learning_rate": 1.2454545454545454e-05,
"loss": 3.3658,
"step": 138
},
{
"epoch": 0.0012501517720226827,
"grad_norm": 148.12303161621094,
"learning_rate": 1.2545454545454547e-05,
"loss": 3.3753,
"step": 139
},
{
"epoch": 0.0012591456696631335,
"grad_norm": 113.60433959960938,
"learning_rate": 1.2636363636363638e-05,
"loss": 3.999,
"step": 140
},
{
"epoch": 0.0012681395673035845,
"grad_norm": 149.37197875976562,
"learning_rate": 1.2727272727272727e-05,
"loss": 4.0167,
"step": 141
},
{
"epoch": 0.0012771334649440355,
"grad_norm": 93.42738342285156,
"learning_rate": 1.2818181818181818e-05,
"loss": 3.5332,
"step": 142
},
{
"epoch": 0.0012861273625844865,
"grad_norm": 125.68160247802734,
"learning_rate": 1.290909090909091e-05,
"loss": 4.0157,
"step": 143
},
{
"epoch": 0.0012951212602249373,
"grad_norm": 153.78225708007812,
"learning_rate": 1.3000000000000001e-05,
"loss": 3.9976,
"step": 144
},
{
"epoch": 0.0013041151578653883,
"grad_norm": 180.440673828125,
"learning_rate": 1.3090909090909093e-05,
"loss": 3.8925,
"step": 145
},
{
"epoch": 0.0013131090555058393,
"grad_norm": 95.41625213623047,
"learning_rate": 1.318181818181818e-05,
"loss": 3.7465,
"step": 146
},
{
"epoch": 0.0013221029531462903,
"grad_norm": 122.04198455810547,
"learning_rate": 1.3272727272727273e-05,
"loss": 3.8933,
"step": 147
},
{
"epoch": 0.0013310968507867411,
"grad_norm": 107.50264739990234,
"learning_rate": 1.3363636363636364e-05,
"loss": 4.019,
"step": 148
},
{
"epoch": 0.0013400907484271921,
"grad_norm": 98.88459777832031,
"learning_rate": 1.3454545454545457e-05,
"loss": 4.0481,
"step": 149
},
{
"epoch": 0.0013490846460676431,
"grad_norm": 160.04849243164062,
"learning_rate": 1.3545454545454547e-05,
"loss": 3.3143,
"step": 150
},
{
"epoch": 0.0013580785437080941,
"grad_norm": 92.48059844970703,
"learning_rate": 1.3636363636363637e-05,
"loss": 3.3451,
"step": 151
},
{
"epoch": 0.001367072441348545,
"grad_norm": 160.9433135986328,
"learning_rate": 1.3727272727272727e-05,
"loss": 3.67,
"step": 152
},
{
"epoch": 0.001376066338988996,
"grad_norm": 155.68333435058594,
"learning_rate": 1.3818181818181818e-05,
"loss": 3.6389,
"step": 153
},
{
"epoch": 0.001385060236629447,
"grad_norm": 243.73287963867188,
"learning_rate": 1.390909090909091e-05,
"loss": 3.9784,
"step": 154
},
{
"epoch": 0.001394054134269898,
"grad_norm": 247.50204467773438,
"learning_rate": 1.4000000000000001e-05,
"loss": 4.2037,
"step": 155
},
{
"epoch": 0.0014030480319103488,
"grad_norm": 137.28292846679688,
"learning_rate": 1.409090909090909e-05,
"loss": 3.4788,
"step": 156
},
{
"epoch": 0.0014120419295507998,
"grad_norm": 143.1851806640625,
"learning_rate": 1.4181818181818181e-05,
"loss": 3.0819,
"step": 157
},
{
"epoch": 0.0014210358271912508,
"grad_norm": 239.31680297851562,
"learning_rate": 1.4272727272727274e-05,
"loss": 3.7093,
"step": 158
},
{
"epoch": 0.0014300297248317016,
"grad_norm": 223.9289093017578,
"learning_rate": 1.4363636363636365e-05,
"loss": 4.1168,
"step": 159
},
{
"epoch": 0.0014390236224721526,
"grad_norm": 140.57806396484375,
"learning_rate": 1.4454545454545454e-05,
"loss": 4.0792,
"step": 160
},
{
"epoch": 0.0014480175201126036,
"grad_norm": 187.22354125976562,
"learning_rate": 1.4545454545454545e-05,
"loss": 3.3268,
"step": 161
},
{
"epoch": 0.0014570114177530546,
"grad_norm": 206.4626922607422,
"learning_rate": 1.4636363636363637e-05,
"loss": 3.9791,
"step": 162
},
{
"epoch": 0.0014660053153935054,
"grad_norm": 608.0723266601562,
"learning_rate": 1.4727272727272728e-05,
"loss": 3.7737,
"step": 163
},
{
"epoch": 0.0014749992130339564,
"grad_norm": 162.69952392578125,
"learning_rate": 1.481818181818182e-05,
"loss": 3.4041,
"step": 164
},
{
"epoch": 0.0014839931106744074,
"grad_norm": 98.44976043701172,
"learning_rate": 1.4909090909090908e-05,
"loss": 3.2322,
"step": 165
},
{
"epoch": 0.0014929870083148584,
"grad_norm": 110.34259796142578,
"learning_rate": 1.5e-05,
"loss": 3.4905,
"step": 166
},
{
"epoch": 0.0015019809059553092,
"grad_norm": 157.68157958984375,
"learning_rate": 1.5090909090909091e-05,
"loss": 3.7214,
"step": 167
},
{
"epoch": 0.0015109748035957602,
"grad_norm": 149.73728942871094,
"learning_rate": 1.5181818181818184e-05,
"loss": 3.937,
"step": 168
},
{
"epoch": 0.0015199687012362113,
"grad_norm": 156.5905303955078,
"learning_rate": 1.5272727272727276e-05,
"loss": 2.9693,
"step": 169
},
{
"epoch": 0.0015289625988766623,
"grad_norm": 192.3663330078125,
"learning_rate": 1.5363636363636362e-05,
"loss": 3.0357,
"step": 170
},
{
"epoch": 0.001537956496517113,
"grad_norm": 106.77263641357422,
"learning_rate": 1.5454545454545454e-05,
"loss": 4.5128,
"step": 171
},
{
"epoch": 0.001546950394157564,
"grad_norm": 225.660888671875,
"learning_rate": 1.5545454545454547e-05,
"loss": 3.5855,
"step": 172
},
{
"epoch": 0.001555944291798015,
"grad_norm": 179.28977966308594,
"learning_rate": 1.563636363636364e-05,
"loss": 3.9906,
"step": 173
},
{
"epoch": 0.001564938189438466,
"grad_norm": 120.28265380859375,
"learning_rate": 1.572727272727273e-05,
"loss": 3.1352,
"step": 174
},
{
"epoch": 0.0015739320870789169,
"grad_norm": 107.6299819946289,
"learning_rate": 1.5818181818181818e-05,
"loss": 3.7572,
"step": 175
},
{
"epoch": 0.001582925984719368,
"grad_norm": 257.49505615234375,
"learning_rate": 1.590909090909091e-05,
"loss": 3.3607,
"step": 176
},
{
"epoch": 0.001591919882359819,
"grad_norm": 246.9903564453125,
"learning_rate": 1.6000000000000003e-05,
"loss": 3.3219,
"step": 177
},
{
"epoch": 0.00160091378000027,
"grad_norm": 137.95632934570312,
"learning_rate": 1.6090909090909092e-05,
"loss": 3.1963,
"step": 178
},
{
"epoch": 0.0016099076776407207,
"grad_norm": 79.78209686279297,
"learning_rate": 1.618181818181818e-05,
"loss": 3.9131,
"step": 179
},
{
"epoch": 0.0016189015752811717,
"grad_norm": 92.68106079101562,
"learning_rate": 1.6272727272727273e-05,
"loss": 3.792,
"step": 180
},
{
"epoch": 0.0016278954729216227,
"grad_norm": 286.2545166015625,
"learning_rate": 1.6363636363636366e-05,
"loss": 3.1824,
"step": 181
},
{
"epoch": 0.0016368893705620735,
"grad_norm": 374.82830810546875,
"learning_rate": 1.6454545454545455e-05,
"loss": 3.6738,
"step": 182
},
{
"epoch": 0.0016458832682025245,
"grad_norm": 263.861083984375,
"learning_rate": 1.6545454545454548e-05,
"loss": 3.6791,
"step": 183
},
{
"epoch": 0.0016548771658429755,
"grad_norm": 279.99603271484375,
"learning_rate": 1.6636363636363637e-05,
"loss": 3.0861,
"step": 184
},
{
"epoch": 0.0016638710634834266,
"grad_norm": 330.89990234375,
"learning_rate": 1.672727272727273e-05,
"loss": 3.091,
"step": 185
},
{
"epoch": 0.0016728649611238774,
"grad_norm": 168.16578674316406,
"learning_rate": 1.6818181818181818e-05,
"loss": 3.1189,
"step": 186
},
{
"epoch": 0.0016818588587643284,
"grad_norm": 136.7516632080078,
"learning_rate": 1.690909090909091e-05,
"loss": 3.5708,
"step": 187
},
{
"epoch": 0.0016908527564047794,
"grad_norm": 371.912109375,
"learning_rate": 1.7000000000000003e-05,
"loss": 3.5552,
"step": 188
},
{
"epoch": 0.0016998466540452304,
"grad_norm": 133.0180206298828,
"learning_rate": 1.7090909090909092e-05,
"loss": 3.1068,
"step": 189
},
{
"epoch": 0.0017088405516856812,
"grad_norm": 122.30583953857422,
"learning_rate": 1.718181818181818e-05,
"loss": 3.2482,
"step": 190
},
{
"epoch": 0.0017178344493261322,
"grad_norm": 139.85595703125,
"learning_rate": 1.7272727272727274e-05,
"loss": 3.1804,
"step": 191
},
{
"epoch": 0.0017268283469665832,
"grad_norm": 85.83197021484375,
"learning_rate": 1.7363636363636366e-05,
"loss": 3.5153,
"step": 192
},
{
"epoch": 0.0017358222446070342,
"grad_norm": 82.42160034179688,
"learning_rate": 1.7454545454545456e-05,
"loss": 3.5141,
"step": 193
},
{
"epoch": 0.001744816142247485,
"grad_norm": 77.44760131835938,
"learning_rate": 1.7545454545454545e-05,
"loss": 3.0701,
"step": 194
},
{
"epoch": 0.001753810039887936,
"grad_norm": 111.4917984008789,
"learning_rate": 1.7636363636363637e-05,
"loss": 3.075,
"step": 195
},
{
"epoch": 0.001762803937528387,
"grad_norm": 85.65130615234375,
"learning_rate": 1.772727272727273e-05,
"loss": 3.1711,
"step": 196
},
{
"epoch": 0.001771797835168838,
"grad_norm": 64.50166320800781,
"learning_rate": 1.781818181818182e-05,
"loss": 3.0664,
"step": 197
},
{
"epoch": 0.0017807917328092888,
"grad_norm": 251.89498901367188,
"learning_rate": 1.7909090909090908e-05,
"loss": 3.7399,
"step": 198
},
{
"epoch": 0.0017897856304497398,
"grad_norm": 157.12791442871094,
"learning_rate": 1.8e-05,
"loss": 3.4729,
"step": 199
},
{
"epoch": 0.0017987795280901909,
"grad_norm": 152.71597290039062,
"learning_rate": 1.8090909090909093e-05,
"loss": 3.0034,
"step": 200
},
{
"epoch": 0.0018077734257306419,
"grad_norm": 263.5816650390625,
"learning_rate": 1.8181818181818182e-05,
"loss": 3.4371,
"step": 201
},
{
"epoch": 0.0018167673233710927,
"grad_norm": 119.90428924560547,
"learning_rate": 1.8272727272727275e-05,
"loss": 3.3518,
"step": 202
},
{
"epoch": 0.0018257612210115437,
"grad_norm": 190.151123046875,
"learning_rate": 1.8363636363636364e-05,
"loss": 3.7226,
"step": 203
},
{
"epoch": 0.0018347551186519947,
"grad_norm": 180.32394409179688,
"learning_rate": 1.8454545454545456e-05,
"loss": 3.1088,
"step": 204
},
{
"epoch": 0.0018437490162924455,
"grad_norm": 70.09622955322266,
"learning_rate": 1.8545454545454545e-05,
"loss": 3.1764,
"step": 205
},
{
"epoch": 0.0018527429139328965,
"grad_norm": 135.13621520996094,
"learning_rate": 1.8636363636363638e-05,
"loss": 3.3676,
"step": 206
},
{
"epoch": 0.0018617368115733475,
"grad_norm": 225.05950927734375,
"learning_rate": 1.872727272727273e-05,
"loss": 3.66,
"step": 207
},
{
"epoch": 0.0018707307092137985,
"grad_norm": 122.18893432617188,
"learning_rate": 1.881818181818182e-05,
"loss": 2.9396,
"step": 208
},
{
"epoch": 0.0018797246068542493,
"grad_norm": 85.67851257324219,
"learning_rate": 1.890909090909091e-05,
"loss": 2.81,
"step": 209
},
{
"epoch": 0.0018887185044947003,
"grad_norm": 112.75260162353516,
"learning_rate": 1.9e-05,
"loss": 3.1241,
"step": 210
},
{
"epoch": 0.0018977124021351513,
"grad_norm": 76.12918853759766,
"learning_rate": 1.9090909090909094e-05,
"loss": 3.0039,
"step": 211
},
{
"epoch": 0.0019067062997756023,
"grad_norm": 103.438232421875,
"learning_rate": 1.9181818181818183e-05,
"loss": 3.3159,
"step": 212
},
{
"epoch": 0.0019157001974160531,
"grad_norm": 64.7052993774414,
"learning_rate": 1.9272727272727272e-05,
"loss": 2.8556,
"step": 213
},
{
"epoch": 0.0019246940950565041,
"grad_norm": 67.88910675048828,
"learning_rate": 1.9363636363636364e-05,
"loss": 3.1821,
"step": 214
},
{
"epoch": 0.0019336879926969551,
"grad_norm": 138.4066925048828,
"learning_rate": 1.9454545454545457e-05,
"loss": 3.1407,
"step": 215
},
{
"epoch": 0.0019426818903374062,
"grad_norm": 48.77667999267578,
"learning_rate": 1.9545454545454546e-05,
"loss": 3.2218,
"step": 216
},
{
"epoch": 0.001951675787977857,
"grad_norm": 129.01211547851562,
"learning_rate": 1.9636363636363635e-05,
"loss": 3.1226,
"step": 217
},
{
"epoch": 0.001960669685618308,
"grad_norm": 137.63018798828125,
"learning_rate": 1.9727272727272728e-05,
"loss": 3.1247,
"step": 218
},
{
"epoch": 0.001969663583258759,
"grad_norm": 69.55642700195312,
"learning_rate": 1.981818181818182e-05,
"loss": 3.2127,
"step": 219
},
{
"epoch": 0.0019786574808992098,
"grad_norm": 139.25526428222656,
"learning_rate": 1.990909090909091e-05,
"loss": 3.0208,
"step": 220
},
{
"epoch": 0.001987651378539661,
"grad_norm": 95.1934814453125,
"learning_rate": 2e-05,
"loss": 3.1067,
"step": 221
},
{
"epoch": 0.001996645276180112,
"grad_norm": 92.59369659423828,
"learning_rate": 2.009090909090909e-05,
"loss": 3.3906,
"step": 222
},
{
"epoch": 0.0020056391738205626,
"grad_norm": 111.0574951171875,
"learning_rate": 2.0181818181818183e-05,
"loss": 3.7034,
"step": 223
},
{
"epoch": 0.002014633071461014,
"grad_norm": 189.96214294433594,
"learning_rate": 2.0272727272727272e-05,
"loss": 3.1057,
"step": 224
},
{
"epoch": 0.0020236269691014646,
"grad_norm": 61.11247634887695,
"learning_rate": 2.0363636363636365e-05,
"loss": 3.2343,
"step": 225
},
{
"epoch": 0.002032620866741916,
"grad_norm": 35.40169906616211,
"learning_rate": 2.0454545454545457e-05,
"loss": 2.7804,
"step": 226
},
{
"epoch": 0.0020416147643823666,
"grad_norm": 52.55931091308594,
"learning_rate": 2.0545454545454546e-05,
"loss": 3.3604,
"step": 227
},
{
"epoch": 0.0020506086620228174,
"grad_norm": 72.27637481689453,
"learning_rate": 2.0636363636363636e-05,
"loss": 3.1553,
"step": 228
},
{
"epoch": 0.0020596025596632686,
"grad_norm": 40.75813674926758,
"learning_rate": 2.0727272727272728e-05,
"loss": 3.083,
"step": 229
},
{
"epoch": 0.0020685964573037194,
"grad_norm": 91.76944732666016,
"learning_rate": 2.081818181818182e-05,
"loss": 2.9126,
"step": 230
},
{
"epoch": 0.0020775903549441702,
"grad_norm": 91.73649597167969,
"learning_rate": 2.090909090909091e-05,
"loss": 2.8581,
"step": 231
},
{
"epoch": 0.0020865842525846215,
"grad_norm": 57.67620086669922,
"learning_rate": 2.1e-05,
"loss": 2.9455,
"step": 232
},
{
"epoch": 0.0020955781502250722,
"grad_norm": 31.565837860107422,
"learning_rate": 2.109090909090909e-05,
"loss": 2.7875,
"step": 233
},
{
"epoch": 0.002104572047865523,
"grad_norm": 61.9347038269043,
"learning_rate": 2.1181818181818184e-05,
"loss": 2.9887,
"step": 234
},
{
"epoch": 0.0021135659455059743,
"grad_norm": 68.58342742919922,
"learning_rate": 2.1272727272727276e-05,
"loss": 2.8304,
"step": 235
},
{
"epoch": 0.002122559843146425,
"grad_norm": 69.1456069946289,
"learning_rate": 2.1363636363636362e-05,
"loss": 3.029,
"step": 236
},
{
"epoch": 0.0021315537407868763,
"grad_norm": 96.83798217773438,
"learning_rate": 2.1454545454545455e-05,
"loss": 3.1787,
"step": 237
},
{
"epoch": 0.002140547638427327,
"grad_norm": 69.93182373046875,
"learning_rate": 2.1545454545454547e-05,
"loss": 2.9907,
"step": 238
},
{
"epoch": 0.002149541536067778,
"grad_norm": 98.68118286132812,
"learning_rate": 2.163636363636364e-05,
"loss": 2.9288,
"step": 239
},
{
"epoch": 0.002158535433708229,
"grad_norm": 59.91753387451172,
"learning_rate": 2.172727272727273e-05,
"loss": 3.9726,
"step": 240
},
{
"epoch": 0.00216752933134868,
"grad_norm": 195.9169921875,
"learning_rate": 2.1818181818181818e-05,
"loss": 3.191,
"step": 241
},
{
"epoch": 0.0021765232289891307,
"grad_norm": 115.80036163330078,
"learning_rate": 2.190909090909091e-05,
"loss": 3.262,
"step": 242
},
{
"epoch": 0.002185517126629582,
"grad_norm": 192.85691833496094,
"learning_rate": 2.2000000000000003e-05,
"loss": 3.0146,
"step": 243
},
{
"epoch": 0.0021945110242700327,
"grad_norm": 239.3960723876953,
"learning_rate": 2.2090909090909092e-05,
"loss": 3.0813,
"step": 244
},
{
"epoch": 0.002203504921910484,
"grad_norm": 80.44845581054688,
"learning_rate": 2.2181818181818184e-05,
"loss": 2.8697,
"step": 245
},
{
"epoch": 0.0022124988195509347,
"grad_norm": 215.9846649169922,
"learning_rate": 2.2272727272727274e-05,
"loss": 3.1669,
"step": 246
},
{
"epoch": 0.0022214927171913855,
"grad_norm": 343.25738525390625,
"learning_rate": 2.2363636363636366e-05,
"loss": 3.1174,
"step": 247
},
{
"epoch": 0.0022304866148318368,
"grad_norm": 384.26153564453125,
"learning_rate": 2.2454545454545455e-05,
"loss": 3.008,
"step": 248
},
{
"epoch": 0.0022394805124722876,
"grad_norm": 124.88306427001953,
"learning_rate": 2.2545454545454548e-05,
"loss": 2.9283,
"step": 249
},
{
"epoch": 0.0022484744101127383,
"grad_norm": 103.40723419189453,
"learning_rate": 2.2636363636363637e-05,
"loss": 2.8574,
"step": 250
},
{
"epoch": 0.0022574683077531896,
"grad_norm": 160.85443115234375,
"learning_rate": 2.272727272727273e-05,
"loss": 3.5287,
"step": 251
},
{
"epoch": 0.0022664622053936404,
"grad_norm": 45.96371078491211,
"learning_rate": 2.281818181818182e-05,
"loss": 2.8872,
"step": 252
},
{
"epoch": 0.002275456103034091,
"grad_norm": 90.8829116821289,
"learning_rate": 2.290909090909091e-05,
"loss": 3.2123,
"step": 253
},
{
"epoch": 0.0022844500006745424,
"grad_norm": 91.640869140625,
"learning_rate": 2.3000000000000003e-05,
"loss": 2.7109,
"step": 254
},
{
"epoch": 0.002293443898314993,
"grad_norm": 47.6387939453125,
"learning_rate": 2.309090909090909e-05,
"loss": 3.4242,
"step": 255
},
{
"epoch": 0.0023024377959554444,
"grad_norm": 108.52681732177734,
"learning_rate": 2.318181818181818e-05,
"loss": 3.0449,
"step": 256
},
{
"epoch": 0.002311431693595895,
"grad_norm": 62.707340240478516,
"learning_rate": 2.3272727272727274e-05,
"loss": 3.1075,
"step": 257
},
{
"epoch": 0.002320425591236346,
"grad_norm": 115.00871276855469,
"learning_rate": 2.3363636363636367e-05,
"loss": 2.7261,
"step": 258
},
{
"epoch": 0.0023294194888767972,
"grad_norm": 136.74339294433594,
"learning_rate": 2.3454545454545456e-05,
"loss": 3.1472,
"step": 259
},
{
"epoch": 0.002338413386517248,
"grad_norm": 100.91593933105469,
"learning_rate": 2.3545454545454545e-05,
"loss": 2.7715,
"step": 260
},
{
"epoch": 0.002347407284157699,
"grad_norm": 35.69953536987305,
"learning_rate": 2.3636363636363637e-05,
"loss": 2.691,
"step": 261
},
{
"epoch": 0.00235640118179815,
"grad_norm": 35.81983184814453,
"learning_rate": 2.372727272727273e-05,
"loss": 2.8004,
"step": 262
},
{
"epoch": 0.002365395079438601,
"grad_norm": 72.92234802246094,
"learning_rate": 2.381818181818182e-05,
"loss": 2.7828,
"step": 263
},
{
"epoch": 0.002374388977079052,
"grad_norm": 62.81526565551758,
"learning_rate": 2.390909090909091e-05,
"loss": 2.9331,
"step": 264
},
{
"epoch": 0.002383382874719503,
"grad_norm": 265.2322692871094,
"learning_rate": 2.4e-05,
"loss": 2.9601,
"step": 265
},
{
"epoch": 0.0023923767723599536,
"grad_norm": 134.36927795410156,
"learning_rate": 2.4090909090909093e-05,
"loss": 2.8622,
"step": 266
},
{
"epoch": 0.002401370670000405,
"grad_norm": 72.66773223876953,
"learning_rate": 2.4181818181818182e-05,
"loss": 2.9547,
"step": 267
},
{
"epoch": 0.0024103645676408557,
"grad_norm": 79.63045501708984,
"learning_rate": 2.4272727272727275e-05,
"loss": 2.8842,
"step": 268
},
{
"epoch": 0.0024193584652813065,
"grad_norm": 34.807342529296875,
"learning_rate": 2.4363636363636364e-05,
"loss": 2.9106,
"step": 269
},
{
"epoch": 0.0024283523629217577,
"grad_norm": 64.01009368896484,
"learning_rate": 2.4454545454545456e-05,
"loss": 3.0516,
"step": 270
},
{
"epoch": 0.0024373462605622085,
"grad_norm": 123.88774108886719,
"learning_rate": 2.4545454545454545e-05,
"loss": 3.1205,
"step": 271
},
{
"epoch": 0.0024463401582026593,
"grad_norm": 32.183597564697266,
"learning_rate": 2.4636363636363638e-05,
"loss": 2.7969,
"step": 272
},
{
"epoch": 0.0024553340558431105,
"grad_norm": 34.47355270385742,
"learning_rate": 2.472727272727273e-05,
"loss": 2.6981,
"step": 273
},
{
"epoch": 0.0024643279534835613,
"grad_norm": 35.006256103515625,
"learning_rate": 2.481818181818182e-05,
"loss": 2.6709,
"step": 274
},
{
"epoch": 0.0024733218511240125,
"grad_norm": 42.49452590942383,
"learning_rate": 2.490909090909091e-05,
"loss": 2.6843,
"step": 275
},
{
"epoch": 0.0024823157487644633,
"grad_norm": 29.628705978393555,
"learning_rate": 2.5e-05,
"loss": 2.7749,
"step": 276
},
{
"epoch": 0.002491309646404914,
"grad_norm": 33.55371856689453,
"learning_rate": 2.5090909090909094e-05,
"loss": 2.7034,
"step": 277
},
{
"epoch": 0.0025003035440453653,
"grad_norm": 33.486297607421875,
"learning_rate": 2.5181818181818183e-05,
"loss": 2.7372,
"step": 278
},
{
"epoch": 0.002509297441685816,
"grad_norm": 26.911794662475586,
"learning_rate": 2.5272727272727275e-05,
"loss": 2.8555,
"step": 279
},
{
"epoch": 0.002518291339326267,
"grad_norm": 28.314470291137695,
"learning_rate": 2.5363636363636368e-05,
"loss": 2.8832,
"step": 280
},
{
"epoch": 0.002527285236966718,
"grad_norm": 86.84230041503906,
"learning_rate": 2.5454545454545454e-05,
"loss": 2.7332,
"step": 281
},
{
"epoch": 0.002536279134607169,
"grad_norm": 73.28656005859375,
"learning_rate": 2.5545454545454546e-05,
"loss": 2.7645,
"step": 282
},
{
"epoch": 0.00254527303224762,
"grad_norm": 50.48367691040039,
"learning_rate": 2.5636363636363635e-05,
"loss": 2.7545,
"step": 283
},
{
"epoch": 0.002554266929888071,
"grad_norm": 79.51824188232422,
"learning_rate": 2.5727272727272728e-05,
"loss": 2.7741,
"step": 284
},
{
"epoch": 0.0025632608275285218,
"grad_norm": 96.66320037841797,
"learning_rate": 2.581818181818182e-05,
"loss": 2.8547,
"step": 285
},
{
"epoch": 0.002572254725168973,
"grad_norm": 41.353294372558594,
"learning_rate": 2.590909090909091e-05,
"loss": 2.7092,
"step": 286
},
{
"epoch": 0.002581248622809424,
"grad_norm": 68.37921905517578,
"learning_rate": 2.6000000000000002e-05,
"loss": 2.6085,
"step": 287
},
{
"epoch": 0.0025902425204498746,
"grad_norm": 51.16245651245117,
"learning_rate": 2.6090909090909094e-05,
"loss": 2.7847,
"step": 288
},
{
"epoch": 0.002599236418090326,
"grad_norm": 34.79716491699219,
"learning_rate": 2.6181818181818187e-05,
"loss": 2.7926,
"step": 289
},
{
"epoch": 0.0026082303157307766,
"grad_norm": 349.29339599609375,
"learning_rate": 2.6272727272727276e-05,
"loss": 2.6578,
"step": 290
},
{
"epoch": 0.002617224213371228,
"grad_norm": 299.22015380859375,
"learning_rate": 2.636363636363636e-05,
"loss": 3.0245,
"step": 291
},
{
"epoch": 0.0026262181110116786,
"grad_norm": 99.52041625976562,
"learning_rate": 2.6454545454545454e-05,
"loss": 2.8194,
"step": 292
},
{
"epoch": 0.0026352120086521294,
"grad_norm": 52.22737503051758,
"learning_rate": 2.6545454545454547e-05,
"loss": 2.8177,
"step": 293
},
{
"epoch": 0.0026442059062925806,
"grad_norm": 96.27494049072266,
"learning_rate": 2.6636363636363636e-05,
"loss": 3.0188,
"step": 294
},
{
"epoch": 0.0026531998039330314,
"grad_norm": 197.97125244140625,
"learning_rate": 2.6727272727272728e-05,
"loss": 2.8892,
"step": 295
},
{
"epoch": 0.0026621937015734822,
"grad_norm": 473.30499267578125,
"learning_rate": 2.681818181818182e-05,
"loss": 2.6927,
"step": 296
},
{
"epoch": 0.0026711875992139335,
"grad_norm": 295.10626220703125,
"learning_rate": 2.6909090909090913e-05,
"loss": 2.7727,
"step": 297
},
{
"epoch": 0.0026801814968543843,
"grad_norm": 419.4664001464844,
"learning_rate": 2.7000000000000002e-05,
"loss": 2.9912,
"step": 298
},
{
"epoch": 0.002689175394494835,
"grad_norm": 899.4752197265625,
"learning_rate": 2.7090909090909095e-05,
"loss": 2.9282,
"step": 299
},
{
"epoch": 0.0026981692921352863,
"grad_norm": 511.102783203125,
"learning_rate": 2.718181818181818e-05,
"loss": 3.1278,
"step": 300
},
{
"epoch": 0.002707163189775737,
"grad_norm": 355.4637145996094,
"learning_rate": 2.7272727272727273e-05,
"loss": 2.7197,
"step": 301
},
{
"epoch": 0.0027161570874161883,
"grad_norm": 87.07572937011719,
"learning_rate": 2.7363636363636362e-05,
"loss": 2.9893,
"step": 302
},
{
"epoch": 0.002725150985056639,
"grad_norm": 351.97991943359375,
"learning_rate": 2.7454545454545455e-05,
"loss": 3.0283,
"step": 303
},
{
"epoch": 0.00273414488269709,
"grad_norm": 347.8600158691406,
"learning_rate": 2.7545454545454547e-05,
"loss": 2.8555,
"step": 304
},
{
"epoch": 0.002743138780337541,
"grad_norm": 146.10769653320312,
"learning_rate": 2.7636363636363636e-05,
"loss": 2.9316,
"step": 305
},
{
"epoch": 0.002752132677977992,
"grad_norm": 42.5224609375,
"learning_rate": 2.772727272727273e-05,
"loss": 2.7596,
"step": 306
},
{
"epoch": 0.0027611265756184427,
"grad_norm": 39.49795150756836,
"learning_rate": 2.781818181818182e-05,
"loss": 2.7516,
"step": 307
},
{
"epoch": 0.002770120473258894,
"grad_norm": 22.18710708618164,
"learning_rate": 2.7909090909090914e-05,
"loss": 2.6042,
"step": 308
},
{
"epoch": 0.0027791143708993447,
"grad_norm": 26.737184524536133,
"learning_rate": 2.8000000000000003e-05,
"loss": 2.7415,
"step": 309
},
{
"epoch": 0.002788108268539796,
"grad_norm": 14.636752128601074,
"learning_rate": 2.809090909090909e-05,
"loss": 2.8259,
"step": 310
},
{
"epoch": 0.0027971021661802467,
"grad_norm": 43.52849578857422,
"learning_rate": 2.818181818181818e-05,
"loss": 2.722,
"step": 311
},
{
"epoch": 0.0028060960638206975,
"grad_norm": 70.25341033935547,
"learning_rate": 2.8272727272727274e-05,
"loss": 2.6344,
"step": 312
},
{
"epoch": 0.0028150899614611488,
"grad_norm": 56.68631362915039,
"learning_rate": 2.8363636363636363e-05,
"loss": 2.7435,
"step": 313
},
{
"epoch": 0.0028240838591015996,
"grad_norm": 31.749832153320312,
"learning_rate": 2.8454545454545455e-05,
"loss": 2.5849,
"step": 314
},
{
"epoch": 0.0028330777567420504,
"grad_norm": 50.61017990112305,
"learning_rate": 2.8545454545454548e-05,
"loss": 2.6884,
"step": 315
},
{
"epoch": 0.0028420716543825016,
"grad_norm": 17.99905014038086,
"learning_rate": 2.863636363636364e-05,
"loss": 2.6863,
"step": 316
},
{
"epoch": 0.0028510655520229524,
"grad_norm": 52.78300476074219,
"learning_rate": 2.872727272727273e-05,
"loss": 2.5679,
"step": 317
},
{
"epoch": 0.002860059449663403,
"grad_norm": 21.387653350830078,
"learning_rate": 2.8818181818181822e-05,
"loss": 2.6698,
"step": 318
},
{
"epoch": 0.0028690533473038544,
"grad_norm": 21.18263053894043,
"learning_rate": 2.8909090909090908e-05,
"loss": 2.7708,
"step": 319
},
{
"epoch": 0.002878047244944305,
"grad_norm": 33.6380615234375,
"learning_rate": 2.9e-05,
"loss": 2.6107,
"step": 320
},
{
"epoch": 0.0028870411425847564,
"grad_norm": 26.449613571166992,
"learning_rate": 2.909090909090909e-05,
"loss": 2.6022,
"step": 321
},
{
"epoch": 0.002896035040225207,
"grad_norm": 67.54065704345703,
"learning_rate": 2.9181818181818182e-05,
"loss": 2.8179,
"step": 322
},
{
"epoch": 0.002905028937865658,
"grad_norm": 94.57304382324219,
"learning_rate": 2.9272727272727274e-05,
"loss": 2.7015,
"step": 323
},
{
"epoch": 0.0029140228355061092,
"grad_norm": 39.76059341430664,
"learning_rate": 2.9363636363636367e-05,
"loss": 2.6519,
"step": 324
},
{
"epoch": 0.00292301673314656,
"grad_norm": 14.038307189941406,
"learning_rate": 2.9454545454545456e-05,
"loss": 2.6765,
"step": 325
},
{
"epoch": 0.002932010630787011,
"grad_norm": 150.26202392578125,
"learning_rate": 2.954545454545455e-05,
"loss": 2.8133,
"step": 326
},
{
"epoch": 0.002941004528427462,
"grad_norm": 87.42321014404297,
"learning_rate": 2.963636363636364e-05,
"loss": 2.7432,
"step": 327
},
{
"epoch": 0.002949998426067913,
"grad_norm": 80.30967712402344,
"learning_rate": 2.972727272727273e-05,
"loss": 2.8271,
"step": 328
},
{
"epoch": 0.002958992323708364,
"grad_norm": 30.537057876586914,
"learning_rate": 2.9818181818181816e-05,
"loss": 2.6839,
"step": 329
},
{
"epoch": 0.002967986221348815,
"grad_norm": 35.343265533447266,
"learning_rate": 2.9909090909090908e-05,
"loss": 2.7018,
"step": 330
},
{
"epoch": 0.0029769801189892657,
"grad_norm": 70.98170471191406,
"learning_rate": 3e-05,
"loss": 2.6828,
"step": 331
},
{
"epoch": 0.002985974016629717,
"grad_norm": 28.612892150878906,
"learning_rate": 3.0090909090909093e-05,
"loss": 2.5517,
"step": 332
},
{
"epoch": 0.0029949679142701677,
"grad_norm": 60.58148193359375,
"learning_rate": 3.0181818181818182e-05,
"loss": 2.6384,
"step": 333
},
{
"epoch": 0.0030039618119106185,
"grad_norm": 100.5984115600586,
"learning_rate": 3.0272727272727275e-05,
"loss": 2.7623,
"step": 334
},
{
"epoch": 0.0030129557095510697,
"grad_norm": 136.9729766845703,
"learning_rate": 3.0363636363636367e-05,
"loss": 2.6914,
"step": 335
},
{
"epoch": 0.0030219496071915205,
"grad_norm": 34.682979583740234,
"learning_rate": 3.0454545454545456e-05,
"loss": 2.7192,
"step": 336
},
{
"epoch": 0.0030309435048319717,
"grad_norm": 77.67842864990234,
"learning_rate": 3.054545454545455e-05,
"loss": 2.4685,
"step": 337
},
{
"epoch": 0.0030399374024724225,
"grad_norm": 96.45396423339844,
"learning_rate": 3.063636363636364e-05,
"loss": 2.5484,
"step": 338
},
{
"epoch": 0.0030489313001128733,
"grad_norm": 52.39377975463867,
"learning_rate": 3.0727272727272724e-05,
"loss": 2.5122,
"step": 339
},
{
"epoch": 0.0030579251977533245,
"grad_norm": 48.01491928100586,
"learning_rate": 3.0818181818181816e-05,
"loss": 2.6634,
"step": 340
},
{
"epoch": 0.0030669190953937753,
"grad_norm": 63.227745056152344,
"learning_rate": 3.090909090909091e-05,
"loss": 2.5317,
"step": 341
},
{
"epoch": 0.003075912993034226,
"grad_norm": 46.99110794067383,
"learning_rate": 3.1e-05,
"loss": 2.6247,
"step": 342
},
{
"epoch": 0.0030849068906746773,
"grad_norm": 39.3392448425293,
"learning_rate": 3.1090909090909094e-05,
"loss": 2.6409,
"step": 343
},
{
"epoch": 0.003093900788315128,
"grad_norm": 123.8673324584961,
"learning_rate": 3.1181818181818186e-05,
"loss": 2.5517,
"step": 344
},
{
"epoch": 0.003102894685955579,
"grad_norm": 40.15580368041992,
"learning_rate": 3.127272727272728e-05,
"loss": 2.6248,
"step": 345
},
{
"epoch": 0.00311188858359603,
"grad_norm": 22.070646286010742,
"learning_rate": 3.1363636363636365e-05,
"loss": 2.6096,
"step": 346
},
{
"epoch": 0.003120882481236481,
"grad_norm": 17.01295280456543,
"learning_rate": 3.145454545454546e-05,
"loss": 2.5886,
"step": 347
},
{
"epoch": 0.003129876378876932,
"grad_norm": 502.77716064453125,
"learning_rate": 3.154545454545454e-05,
"loss": 2.7185,
"step": 348
},
{
"epoch": 0.003138870276517383,
"grad_norm": 220.7124786376953,
"learning_rate": 3.1636363636363635e-05,
"loss": 2.652,
"step": 349
},
{
"epoch": 0.0031478641741578338,
"grad_norm": 40.6119384765625,
"learning_rate": 3.172727272727273e-05,
"loss": 2.5447,
"step": 350
},
{
"epoch": 0.003156858071798285,
"grad_norm": 28.14447021484375,
"learning_rate": 3.181818181818182e-05,
"loss": 2.6028,
"step": 351
},
{
"epoch": 0.003165851969438736,
"grad_norm": 100.99872589111328,
"learning_rate": 3.190909090909091e-05,
"loss": 2.4951,
"step": 352
},
{
"epoch": 0.0031748458670791866,
"grad_norm": 184.5791778564453,
"learning_rate": 3.2000000000000005e-05,
"loss": 2.6375,
"step": 353
},
{
"epoch": 0.003183839764719638,
"grad_norm": 66.14623260498047,
"learning_rate": 3.209090909090909e-05,
"loss": 2.6241,
"step": 354
},
{
"epoch": 0.0031928336623600886,
"grad_norm": 89.26773071289062,
"learning_rate": 3.2181818181818184e-05,
"loss": 2.6124,
"step": 355
},
{
"epoch": 0.00320182756000054,
"grad_norm": 33.29132843017578,
"learning_rate": 3.2272727272727276e-05,
"loss": 2.6274,
"step": 356
},
{
"epoch": 0.0032108214576409906,
"grad_norm": 40.92216873168945,
"learning_rate": 3.236363636363636e-05,
"loss": 2.6266,
"step": 357
},
{
"epoch": 0.0032198153552814414,
"grad_norm": 77.89151000976562,
"learning_rate": 3.2454545454545454e-05,
"loss": 2.5986,
"step": 358
},
{
"epoch": 0.0032288092529218927,
"grad_norm": 23.291912078857422,
"learning_rate": 3.254545454545455e-05,
"loss": 2.5735,
"step": 359
},
{
"epoch": 0.0032378031505623434,
"grad_norm": 41.3476448059082,
"learning_rate": 3.263636363636364e-05,
"loss": 2.5837,
"step": 360
},
{
"epoch": 0.0032467970482027942,
"grad_norm": 54.72057342529297,
"learning_rate": 3.272727272727273e-05,
"loss": 2.6716,
"step": 361
},
{
"epoch": 0.0032557909458432455,
"grad_norm": 41.65103530883789,
"learning_rate": 3.281818181818182e-05,
"loss": 2.7165,
"step": 362
},
{
"epoch": 0.0032647848434836963,
"grad_norm": 28.426193237304688,
"learning_rate": 3.290909090909091e-05,
"loss": 2.5034,
"step": 363
},
{
"epoch": 0.003273778741124147,
"grad_norm": 26.71957015991211,
"learning_rate": 3.3e-05,
"loss": 2.6379,
"step": 364
},
{
"epoch": 0.0032827726387645983,
"grad_norm": 79.43277740478516,
"learning_rate": 3.3090909090909095e-05,
"loss": 2.6091,
"step": 365
},
{
"epoch": 0.003291766536405049,
"grad_norm": 23.48211669921875,
"learning_rate": 3.318181818181819e-05,
"loss": 2.6034,
"step": 366
},
{
"epoch": 0.0033007604340455003,
"grad_norm": 18.23671531677246,
"learning_rate": 3.327272727272727e-05,
"loss": 2.5584,
"step": 367
},
{
"epoch": 0.003309754331685951,
"grad_norm": 11.874900817871094,
"learning_rate": 3.3363636363636366e-05,
"loss": 2.5881,
"step": 368
},
{
"epoch": 0.003318748229326402,
"grad_norm": 18.68777847290039,
"learning_rate": 3.345454545454546e-05,
"loss": 2.627,
"step": 369
},
{
"epoch": 0.003327742126966853,
"grad_norm": 32.40439987182617,
"learning_rate": 3.3545454545454544e-05,
"loss": 2.5897,
"step": 370
},
{
"epoch": 0.003336736024607304,
"grad_norm": 36.62127685546875,
"learning_rate": 3.3636363636363636e-05,
"loss": 2.5898,
"step": 371
},
{
"epoch": 0.0033457299222477547,
"grad_norm": 12.973684310913086,
"learning_rate": 3.372727272727273e-05,
"loss": 2.6294,
"step": 372
},
{
"epoch": 0.003354723819888206,
"grad_norm": 8.09320068359375,
"learning_rate": 3.381818181818182e-05,
"loss": 2.5505,
"step": 373
},
{
"epoch": 0.0033637177175286567,
"grad_norm": 11.420305252075195,
"learning_rate": 3.3909090909090914e-05,
"loss": 2.5784,
"step": 374
},
{
"epoch": 0.003372711615169108,
"grad_norm": 12.049633026123047,
"learning_rate": 3.4000000000000007e-05,
"loss": 2.6101,
"step": 375
},
{
"epoch": 0.0033817055128095587,
"grad_norm": 14.55691909790039,
"learning_rate": 3.409090909090909e-05,
"loss": 2.4904,
"step": 376
},
{
"epoch": 0.0033906994104500095,
"grad_norm": 14.35408878326416,
"learning_rate": 3.4181818181818185e-05,
"loss": 2.5721,
"step": 377
},
{
"epoch": 0.0033996933080904608,
"grad_norm": 16.358585357666016,
"learning_rate": 3.427272727272727e-05,
"loss": 2.5846,
"step": 378
},
{
"epoch": 0.0034086872057309116,
"grad_norm": 13.180251121520996,
"learning_rate": 3.436363636363636e-05,
"loss": 2.5942,
"step": 379
},
{
"epoch": 0.0034176811033713624,
"grad_norm": 11.884993553161621,
"learning_rate": 3.4454545454545455e-05,
"loss": 2.5825,
"step": 380
},
{
"epoch": 0.0034266750010118136,
"grad_norm": 21.903844833374023,
"learning_rate": 3.454545454545455e-05,
"loss": 2.6786,
"step": 381
},
{
"epoch": 0.0034356688986522644,
"grad_norm": 10.655231475830078,
"learning_rate": 3.463636363636364e-05,
"loss": 2.5477,
"step": 382
},
{
"epoch": 0.0034446627962927156,
"grad_norm": 8.134358406066895,
"learning_rate": 3.472727272727273e-05,
"loss": 2.6264,
"step": 383
},
{
"epoch": 0.0034536566939331664,
"grad_norm": 10.995036125183105,
"learning_rate": 3.481818181818182e-05,
"loss": 2.6314,
"step": 384
},
{
"epoch": 0.003462650591573617,
"grad_norm": 11.038241386413574,
"learning_rate": 3.490909090909091e-05,
"loss": 2.499,
"step": 385
},
{
"epoch": 0.0034716444892140684,
"grad_norm": 4.939131259918213,
"learning_rate": 3.5e-05,
"loss": 2.5725,
"step": 386
},
{
"epoch": 0.003480638386854519,
"grad_norm": 8.1621732711792,
"learning_rate": 3.509090909090909e-05,
"loss": 2.5686,
"step": 387
},
{
"epoch": 0.00348963228449497,
"grad_norm": 8.263984680175781,
"learning_rate": 3.518181818181818e-05,
"loss": 2.5254,
"step": 388
},
{
"epoch": 0.0034986261821354212,
"grad_norm": 23.33096694946289,
"learning_rate": 3.5272727272727274e-05,
"loss": 2.5912,
"step": 389
},
{
"epoch": 0.003507620079775872,
"grad_norm": 18.071269989013672,
"learning_rate": 3.536363636363637e-05,
"loss": 2.5056,
"step": 390
},
{
"epoch": 0.003516613977416323,
"grad_norm": 18.0259952545166,
"learning_rate": 3.545454545454546e-05,
"loss": 2.4462,
"step": 391
},
{
"epoch": 0.003525607875056774,
"grad_norm": 10.46397590637207,
"learning_rate": 3.5545454545454545e-05,
"loss": 2.527,
"step": 392
},
{
"epoch": 0.003534601772697225,
"grad_norm": 6.038736343383789,
"learning_rate": 3.563636363636364e-05,
"loss": 2.5446,
"step": 393
},
{
"epoch": 0.003543595670337676,
"grad_norm": 25.153114318847656,
"learning_rate": 3.572727272727273e-05,
"loss": 2.7408,
"step": 394
},
{
"epoch": 0.003552589567978127,
"grad_norm": 19.05910873413086,
"learning_rate": 3.5818181818181816e-05,
"loss": 2.5632,
"step": 395
},
{
"epoch": 0.0035615834656185777,
"grad_norm": 20.161680221557617,
"learning_rate": 3.590909090909091e-05,
"loss": 2.4902,
"step": 396
},
{
"epoch": 0.003570577363259029,
"grad_norm": 12.120384216308594,
"learning_rate": 3.6e-05,
"loss": 2.4759,
"step": 397
},
{
"epoch": 0.0035795712608994797,
"grad_norm": 9.16979694366455,
"learning_rate": 3.6090909090909093e-05,
"loss": 2.6742,
"step": 398
},
{
"epoch": 0.0035885651585399305,
"grad_norm": 5.7415571212768555,
"learning_rate": 3.6181818181818186e-05,
"loss": 2.5058,
"step": 399
},
{
"epoch": 0.0035975590561803817,
"grad_norm": 8.898907661437988,
"learning_rate": 3.627272727272727e-05,
"loss": 2.5235,
"step": 400
},
{
"epoch": 0.0036065529538208325,
"grad_norm": 10.796667098999023,
"learning_rate": 3.6363636363636364e-05,
"loss": 2.4847,
"step": 401
},
{
"epoch": 0.0036155468514612837,
"grad_norm": 13.13328742980957,
"learning_rate": 3.645454545454546e-05,
"loss": 2.6087,
"step": 402
},
{
"epoch": 0.0036245407491017345,
"grad_norm": 5.896069526672363,
"learning_rate": 3.654545454545455e-05,
"loss": 2.5116,
"step": 403
},
{
"epoch": 0.0036335346467421853,
"grad_norm": 12.354032516479492,
"learning_rate": 3.663636363636364e-05,
"loss": 2.4865,
"step": 404
},
{
"epoch": 0.0036425285443826365,
"grad_norm": 9.572775840759277,
"learning_rate": 3.672727272727273e-05,
"loss": 2.5348,
"step": 405
},
{
"epoch": 0.0036515224420230873,
"grad_norm": 18.182926177978516,
"learning_rate": 3.681818181818182e-05,
"loss": 2.641,
"step": 406
},
{
"epoch": 0.003660516339663538,
"grad_norm": 16.372573852539062,
"learning_rate": 3.690909090909091e-05,
"loss": 2.4032,
"step": 407
},
{
"epoch": 0.0036695102373039894,
"grad_norm": 6.369616985321045,
"learning_rate": 3.7e-05,
"loss": 2.5007,
"step": 408
},
{
"epoch": 0.00367850413494444,
"grad_norm": 5.87434720993042,
"learning_rate": 3.709090909090909e-05,
"loss": 2.5089,
"step": 409
},
{
"epoch": 0.003687498032584891,
"grad_norm": 11.947864532470703,
"learning_rate": 3.718181818181818e-05,
"loss": 2.5573,
"step": 410
},
{
"epoch": 0.003696491930225342,
"grad_norm": 13.243257522583008,
"learning_rate": 3.7272727272727276e-05,
"loss": 2.4119,
"step": 411
},
{
"epoch": 0.003705485827865793,
"grad_norm": 96.9281005859375,
"learning_rate": 3.736363636363637e-05,
"loss": 2.5608,
"step": 412
},
{
"epoch": 0.003714479725506244,
"grad_norm": 15.962468147277832,
"learning_rate": 3.745454545454546e-05,
"loss": 2.5225,
"step": 413
},
{
"epoch": 0.003723473623146695,
"grad_norm": 5.415879726409912,
"learning_rate": 3.7545454545454546e-05,
"loss": 2.4449,
"step": 414
},
{
"epoch": 0.0037324675207871458,
"grad_norm": 24.821794509887695,
"learning_rate": 3.763636363636364e-05,
"loss": 2.5894,
"step": 415
},
{
"epoch": 0.003741461418427597,
"grad_norm": 12.479686737060547,
"learning_rate": 3.7727272727272725e-05,
"loss": 2.6217,
"step": 416
},
{
"epoch": 0.003750455316068048,
"grad_norm": 6.780008792877197,
"learning_rate": 3.781818181818182e-05,
"loss": 2.4162,
"step": 417
},
{
"epoch": 0.0037594492137084986,
"grad_norm": 317.8321228027344,
"learning_rate": 3.790909090909091e-05,
"loss": 2.5253,
"step": 418
},
{
"epoch": 0.00376844311134895,
"grad_norm": 30.4754581451416,
"learning_rate": 3.8e-05,
"loss": 2.6148,
"step": 419
},
{
"epoch": 0.0037774370089894006,
"grad_norm": 16.09713363647461,
"learning_rate": 3.8090909090909095e-05,
"loss": 2.5931,
"step": 420
},
{
"epoch": 0.003786430906629852,
"grad_norm": 11.552228927612305,
"learning_rate": 3.818181818181819e-05,
"loss": 2.5226,
"step": 421
},
{
"epoch": 0.0037954248042703026,
"grad_norm": 20.038066864013672,
"learning_rate": 3.827272727272728e-05,
"loss": 2.4769,
"step": 422
},
{
"epoch": 0.0038044187019107534,
"grad_norm": 10.651211738586426,
"learning_rate": 3.8363636363636365e-05,
"loss": 2.5161,
"step": 423
},
{
"epoch": 0.0038134125995512047,
"grad_norm": 227.53170776367188,
"learning_rate": 3.845454545454545e-05,
"loss": 2.5614,
"step": 424
},
{
"epoch": 0.0038224064971916554,
"grad_norm": 24.951213836669922,
"learning_rate": 3.8545454545454544e-05,
"loss": 2.5896,
"step": 425
},
{
"epoch": 0.0038314003948321062,
"grad_norm": 13.660984992980957,
"learning_rate": 3.8636363636363636e-05,
"loss": 2.4454,
"step": 426
},
{
"epoch": 0.0038403942924725575,
"grad_norm": 6.037498474121094,
"learning_rate": 3.872727272727273e-05,
"loss": 2.5347,
"step": 427
},
{
"epoch": 0.0038493881901130083,
"grad_norm": 5.03461217880249,
"learning_rate": 3.881818181818182e-05,
"loss": 2.5223,
"step": 428
},
{
"epoch": 0.0038583820877534595,
"grad_norm": 6.264018535614014,
"learning_rate": 3.8909090909090914e-05,
"loss": 2.518,
"step": 429
},
{
"epoch": 0.0038673759853939103,
"grad_norm": 4.939131736755371,
"learning_rate": 3.9000000000000006e-05,
"loss": 2.3866,
"step": 430
},
{
"epoch": 0.003876369883034361,
"grad_norm": 6.09281587600708,
"learning_rate": 3.909090909090909e-05,
"loss": 2.5,
"step": 431
},
{
"epoch": 0.0038853637806748123,
"grad_norm": 8.432394027709961,
"learning_rate": 3.9181818181818184e-05,
"loss": 2.5482,
"step": 432
},
{
"epoch": 0.003894357678315263,
"grad_norm": 7.908455848693848,
"learning_rate": 3.927272727272727e-05,
"loss": 2.39,
"step": 433
},
{
"epoch": 0.003903351575955714,
"grad_norm": 5.349055767059326,
"learning_rate": 3.936363636363636e-05,
"loss": 2.5227,
"step": 434
},
{
"epoch": 0.003912345473596165,
"grad_norm": 8.727982521057129,
"learning_rate": 3.9454545454545455e-05,
"loss": 2.5689,
"step": 435
},
{
"epoch": 0.003921339371236616,
"grad_norm": 12.721184730529785,
"learning_rate": 3.954545454545455e-05,
"loss": 2.5216,
"step": 436
},
{
"epoch": 0.003930333268877067,
"grad_norm": 10.756434440612793,
"learning_rate": 3.963636363636364e-05,
"loss": 2.5229,
"step": 437
},
{
"epoch": 0.003939327166517518,
"grad_norm": 84.30569458007812,
"learning_rate": 3.972727272727273e-05,
"loss": 2.5331,
"step": 438
},
{
"epoch": 0.003948321064157969,
"grad_norm": 15.8399658203125,
"learning_rate": 3.981818181818182e-05,
"loss": 2.4312,
"step": 439
},
{
"epoch": 0.0039573149617984195,
"grad_norm": 8.966629028320312,
"learning_rate": 3.990909090909091e-05,
"loss": 2.5436,
"step": 440
},
{
"epoch": 0.003966308859438871,
"grad_norm": 7.498310089111328,
"learning_rate": 4e-05,
"loss": 2.555,
"step": 441
},
{
"epoch": 0.003975302757079322,
"grad_norm": 11.625467300415039,
"learning_rate": 4.009090909090909e-05,
"loss": 2.4952,
"step": 442
},
{
"epoch": 0.003984296654719772,
"grad_norm": 7.1807661056518555,
"learning_rate": 4.018181818181818e-05,
"loss": 2.4346,
"step": 443
},
{
"epoch": 0.003993290552360224,
"grad_norm": 185.90045166015625,
"learning_rate": 4.0272727272727274e-05,
"loss": 2.554,
"step": 444
},
{
"epoch": 0.004002284450000675,
"grad_norm": 22.226383209228516,
"learning_rate": 4.0363636363636367e-05,
"loss": 2.4563,
"step": 445
},
{
"epoch": 0.004011278347641125,
"grad_norm": 5.050864219665527,
"learning_rate": 4.045454545454546e-05,
"loss": 2.4988,
"step": 446
},
{
"epoch": 0.004020272245281576,
"grad_norm": 5.204012870788574,
"learning_rate": 4.0545454545454545e-05,
"loss": 2.393,
"step": 447
},
{
"epoch": 0.004029266142922028,
"grad_norm": 4.5905866622924805,
"learning_rate": 4.063636363636364e-05,
"loss": 2.4568,
"step": 448
},
{
"epoch": 0.004038260040562478,
"grad_norm": 6.743342399597168,
"learning_rate": 4.072727272727273e-05,
"loss": 2.386,
"step": 449
},
{
"epoch": 0.004047253938202929,
"grad_norm": 7.234485149383545,
"learning_rate": 4.081818181818182e-05,
"loss": 2.4349,
"step": 450
},
{
"epoch": 0.00405624783584338,
"grad_norm": 11.250282287597656,
"learning_rate": 4.0909090909090915e-05,
"loss": 2.4458,
"step": 451
},
{
"epoch": 0.004065241733483832,
"grad_norm": 6.4098286628723145,
"learning_rate": 4.1e-05,
"loss": 2.4528,
"step": 452
},
{
"epoch": 0.004074235631124282,
"grad_norm": 5.250341415405273,
"learning_rate": 4.109090909090909e-05,
"loss": 2.5672,
"step": 453
},
{
"epoch": 0.004083229528764733,
"grad_norm": 5.866156578063965,
"learning_rate": 4.1181818181818185e-05,
"loss": 2.5054,
"step": 454
},
{
"epoch": 0.0040922234264051845,
"grad_norm": 6.161355495452881,
"learning_rate": 4.127272727272727e-05,
"loss": 2.442,
"step": 455
},
{
"epoch": 0.004101217324045635,
"grad_norm": 4.930241107940674,
"learning_rate": 4.1363636363636364e-05,
"loss": 2.3989,
"step": 456
},
{
"epoch": 0.004110211221686086,
"grad_norm": 6.114069938659668,
"learning_rate": 4.1454545454545456e-05,
"loss": 2.5388,
"step": 457
},
{
"epoch": 0.004119205119326537,
"grad_norm": 5.344629764556885,
"learning_rate": 4.154545454545455e-05,
"loss": 2.4695,
"step": 458
},
{
"epoch": 0.004128199016966988,
"grad_norm": 6.4735002517700195,
"learning_rate": 4.163636363636364e-05,
"loss": 2.4764,
"step": 459
},
{
"epoch": 0.004137192914607439,
"grad_norm": 4.274179458618164,
"learning_rate": 4.1727272727272734e-05,
"loss": 2.4462,
"step": 460
},
{
"epoch": 0.00414618681224789,
"grad_norm": 5.283795356750488,
"learning_rate": 4.181818181818182e-05,
"loss": 2.5101,
"step": 461
},
{
"epoch": 0.0041551807098883405,
"grad_norm": 3.78226375579834,
"learning_rate": 4.190909090909091e-05,
"loss": 2.3994,
"step": 462
},
{
"epoch": 0.004164174607528792,
"grad_norm": 3.4593284130096436,
"learning_rate": 4.2e-05,
"loss": 2.4868,
"step": 463
},
{
"epoch": 0.004173168505169243,
"grad_norm": 3.445216178894043,
"learning_rate": 4.209090909090909e-05,
"loss": 2.435,
"step": 464
},
{
"epoch": 0.004182162402809693,
"grad_norm": 11.3976411819458,
"learning_rate": 4.218181818181818e-05,
"loss": 2.361,
"step": 465
},
{
"epoch": 0.0041911563004501445,
"grad_norm": 7.932605266571045,
"learning_rate": 4.2272727272727275e-05,
"loss": 2.3981,
"step": 466
},
{
"epoch": 0.004200150198090596,
"grad_norm": 5.212550163269043,
"learning_rate": 4.236363636363637e-05,
"loss": 2.3614,
"step": 467
},
{
"epoch": 0.004209144095731046,
"grad_norm": 4.533825397491455,
"learning_rate": 4.245454545454546e-05,
"loss": 2.3672,
"step": 468
},
{
"epoch": 0.004218137993371497,
"grad_norm": 4.939641952514648,
"learning_rate": 4.254545454545455e-05,
"loss": 2.4213,
"step": 469
},
{
"epoch": 0.0042271318910119485,
"grad_norm": 5.2856364250183105,
"learning_rate": 4.263636363636364e-05,
"loss": 2.486,
"step": 470
},
{
"epoch": 0.0042361257886524,
"grad_norm": 2.2887861728668213,
"learning_rate": 4.2727272727272724e-05,
"loss": 2.4456,
"step": 471
},
{
"epoch": 0.00424511968629285,
"grad_norm": 8.504525184631348,
"learning_rate": 4.281818181818182e-05,
"loss": 2.5319,
"step": 472
},
{
"epoch": 0.004254113583933301,
"grad_norm": 5.444023132324219,
"learning_rate": 4.290909090909091e-05,
"loss": 2.4518,
"step": 473
},
{
"epoch": 0.004263107481573753,
"grad_norm": 3.858370780944824,
"learning_rate": 4.3e-05,
"loss": 2.4155,
"step": 474
},
{
"epoch": 0.004272101379214203,
"grad_norm": 6.466838836669922,
"learning_rate": 4.3090909090909094e-05,
"loss": 2.5241,
"step": 475
},
{
"epoch": 0.004281095276854654,
"grad_norm": 5.948734760284424,
"learning_rate": 4.318181818181819e-05,
"loss": 2.5933,
"step": 476
},
{
"epoch": 0.004290089174495105,
"grad_norm": 4.375283718109131,
"learning_rate": 4.327272727272728e-05,
"loss": 2.4722,
"step": 477
},
{
"epoch": 0.004299083072135556,
"grad_norm": 5.097462177276611,
"learning_rate": 4.3363636363636365e-05,
"loss": 2.4599,
"step": 478
},
{
"epoch": 0.004308076969776007,
"grad_norm": 6.355868339538574,
"learning_rate": 4.345454545454546e-05,
"loss": 2.4053,
"step": 479
},
{
"epoch": 0.004317070867416458,
"grad_norm": 6.523740291595459,
"learning_rate": 4.354545454545454e-05,
"loss": 2.4362,
"step": 480
},
{
"epoch": 0.004326064765056909,
"grad_norm": 11.158427238464355,
"learning_rate": 4.3636363636363636e-05,
"loss": 2.5399,
"step": 481
},
{
"epoch": 0.00433505866269736,
"grad_norm": 6.568932056427002,
"learning_rate": 4.372727272727273e-05,
"loss": 2.5359,
"step": 482
},
{
"epoch": 0.004344052560337811,
"grad_norm": 5.511691570281982,
"learning_rate": 4.381818181818182e-05,
"loss": 2.4101,
"step": 483
},
{
"epoch": 0.004353046457978261,
"grad_norm": 4.5280890464782715,
"learning_rate": 4.390909090909091e-05,
"loss": 2.5112,
"step": 484
},
{
"epoch": 0.004362040355618713,
"grad_norm": 2.9648256301879883,
"learning_rate": 4.4000000000000006e-05,
"loss": 2.3865,
"step": 485
},
{
"epoch": 0.004371034253259164,
"grad_norm": 2.3086631298065186,
"learning_rate": 4.409090909090909e-05,
"loss": 2.3406,
"step": 486
},
{
"epoch": 0.004380028150899614,
"grad_norm": 4.144700527191162,
"learning_rate": 4.4181818181818184e-05,
"loss": 2.3723,
"step": 487
},
{
"epoch": 0.004389022048540065,
"grad_norm": 6.331112384796143,
"learning_rate": 4.4272727272727276e-05,
"loss": 2.5191,
"step": 488
},
{
"epoch": 0.004398015946180517,
"grad_norm": 2.395456552505493,
"learning_rate": 4.436363636363637e-05,
"loss": 2.4398,
"step": 489
},
{
"epoch": 0.004407009843820968,
"grad_norm": 11.222804069519043,
"learning_rate": 4.4454545454545455e-05,
"loss": 2.5287,
"step": 490
},
{
"epoch": 0.004416003741461418,
"grad_norm": 7.065359115600586,
"learning_rate": 4.454545454545455e-05,
"loss": 2.3943,
"step": 491
},
{
"epoch": 0.0044249976391018695,
"grad_norm": 2.2675955295562744,
"learning_rate": 4.463636363636364e-05,
"loss": 2.4372,
"step": 492
},
{
"epoch": 0.004433991536742321,
"grad_norm": 4.432316303253174,
"learning_rate": 4.472727272727273e-05,
"loss": 2.5259,
"step": 493
},
{
"epoch": 0.004442985434382771,
"grad_norm": 3.251943826675415,
"learning_rate": 4.481818181818182e-05,
"loss": 2.4694,
"step": 494
},
{
"epoch": 0.004451979332023222,
"grad_norm": 5.685604095458984,
"learning_rate": 4.490909090909091e-05,
"loss": 2.5528,
"step": 495
},
{
"epoch": 0.0044609732296636735,
"grad_norm": 5.714761734008789,
"learning_rate": 4.5e-05,
"loss": 2.5311,
"step": 496
},
{
"epoch": 0.004469967127304124,
"grad_norm": 6.29110860824585,
"learning_rate": 4.5090909090909095e-05,
"loss": 2.4541,
"step": 497
},
{
"epoch": 0.004478961024944575,
"grad_norm": 26.393896102905273,
"learning_rate": 4.518181818181819e-05,
"loss": 2.3905,
"step": 498
},
{
"epoch": 0.004487954922585026,
"grad_norm": 29.4892635345459,
"learning_rate": 4.5272727272727274e-05,
"loss": 2.5434,
"step": 499
},
{
"epoch": 0.004496948820225477,
"grad_norm": 9.63700008392334,
"learning_rate": 4.5363636363636366e-05,
"loss": 2.4066,
"step": 500
},
{
"epoch": 0.004505942717865928,
"grad_norm": 9.156270027160645,
"learning_rate": 4.545454545454546e-05,
"loss": 2.4767,
"step": 501
},
{
"epoch": 0.004514936615506379,
"grad_norm": 137.2318878173828,
"learning_rate": 4.5545454545454544e-05,
"loss": 2.4479,
"step": 502
},
{
"epoch": 0.0045239305131468295,
"grad_norm": 195.78118896484375,
"learning_rate": 4.563636363636364e-05,
"loss": 2.6008,
"step": 503
},
{
"epoch": 0.004532924410787281,
"grad_norm": 35.147743225097656,
"learning_rate": 4.572727272727273e-05,
"loss": 2.4295,
"step": 504
},
{
"epoch": 0.004541918308427732,
"grad_norm": 5.983067989349365,
"learning_rate": 4.581818181818182e-05,
"loss": 2.4059,
"step": 505
},
{
"epoch": 0.004550912206068182,
"grad_norm": 20.23373794555664,
"learning_rate": 4.5909090909090914e-05,
"loss": 2.44,
"step": 506
},
{
"epoch": 0.0045599061037086335,
"grad_norm": 4.173520088195801,
"learning_rate": 4.600000000000001e-05,
"loss": 2.3349,
"step": 507
},
{
"epoch": 0.004568900001349085,
"grad_norm": 9.3209228515625,
"learning_rate": 4.609090909090909e-05,
"loss": 2.4971,
"step": 508
},
{
"epoch": 0.004577893898989536,
"grad_norm": 7.05019998550415,
"learning_rate": 4.618181818181818e-05,
"loss": 2.502,
"step": 509
},
{
"epoch": 0.004586887796629986,
"grad_norm": 4.440962314605713,
"learning_rate": 4.627272727272727e-05,
"loss": 2.3926,
"step": 510
},
{
"epoch": 0.004595881694270438,
"grad_norm": 4.18477201461792,
"learning_rate": 4.636363636363636e-05,
"loss": 2.5137,
"step": 511
},
{
"epoch": 0.004604875591910889,
"grad_norm": 7.083902835845947,
"learning_rate": 4.6454545454545456e-05,
"loss": 2.4064,
"step": 512
},
{
"epoch": 0.004613869489551339,
"grad_norm": 48.055904388427734,
"learning_rate": 4.654545454545455e-05,
"loss": 2.3294,
"step": 513
},
{
"epoch": 0.00462286338719179,
"grad_norm": 5.847009181976318,
"learning_rate": 4.663636363636364e-05,
"loss": 2.3623,
"step": 514
},
{
"epoch": 0.004631857284832242,
"grad_norm": 6.544424533843994,
"learning_rate": 4.672727272727273e-05,
"loss": 2.4563,
"step": 515
},
{
"epoch": 0.004640851182472692,
"grad_norm": 3.1533761024475098,
"learning_rate": 4.681818181818182e-05,
"loss": 2.421,
"step": 516
},
{
"epoch": 0.004649845080113143,
"grad_norm": 4.543294429779053,
"learning_rate": 4.690909090909091e-05,
"loss": 2.3646,
"step": 517
},
{
"epoch": 0.0046588389777535944,
"grad_norm": 4.146971702575684,
"learning_rate": 4.7e-05,
"loss": 2.3599,
"step": 518
},
{
"epoch": 0.004667832875394045,
"grad_norm": 2.5263845920562744,
"learning_rate": 4.709090909090909e-05,
"loss": 2.447,
"step": 519
},
{
"epoch": 0.004676826773034496,
"grad_norm": 3.316958427429199,
"learning_rate": 4.718181818181818e-05,
"loss": 2.5234,
"step": 520
},
{
"epoch": 0.004685820670674947,
"grad_norm": 3.836557626724243,
"learning_rate": 4.7272727272727275e-05,
"loss": 2.3482,
"step": 521
},
{
"epoch": 0.004694814568315398,
"grad_norm": 4.416954517364502,
"learning_rate": 4.736363636363637e-05,
"loss": 2.4686,
"step": 522
},
{
"epoch": 0.004703808465955849,
"grad_norm": 132.07891845703125,
"learning_rate": 4.745454545454546e-05,
"loss": 2.4657,
"step": 523
},
{
"epoch": 0.0047128023635963,
"grad_norm": 759.2149047851562,
"learning_rate": 4.7545454545454545e-05,
"loss": 2.7605,
"step": 524
},
{
"epoch": 0.0047217962612367504,
"grad_norm": 186.16273498535156,
"learning_rate": 4.763636363636364e-05,
"loss": 2.5167,
"step": 525
},
{
"epoch": 0.004730790158877202,
"grad_norm": 21.49215316772461,
"learning_rate": 4.772727272727273e-05,
"loss": 2.4627,
"step": 526
},
{
"epoch": 0.004739784056517653,
"grad_norm": 5.0005083084106445,
"learning_rate": 4.781818181818182e-05,
"loss": 2.4019,
"step": 527
},
{
"epoch": 0.004748777954158104,
"grad_norm": 2.7786004543304443,
"learning_rate": 4.790909090909091e-05,
"loss": 2.4485,
"step": 528
},
{
"epoch": 0.0047577718517985545,
"grad_norm": 2.6912715435028076,
"learning_rate": 4.8e-05,
"loss": 2.4206,
"step": 529
},
{
"epoch": 0.004766765749439006,
"grad_norm": 3.564931631088257,
"learning_rate": 4.8090909090909094e-05,
"loss": 2.4458,
"step": 530
},
{
"epoch": 0.004775759647079457,
"grad_norm": 4.527584075927734,
"learning_rate": 4.8181818181818186e-05,
"loss": 2.3245,
"step": 531
},
{
"epoch": 0.004784753544719907,
"grad_norm": 10.07675838470459,
"learning_rate": 4.827272727272727e-05,
"loss": 2.4126,
"step": 532
},
{
"epoch": 0.0047937474423603585,
"grad_norm": 6.41603422164917,
"learning_rate": 4.8363636363636364e-05,
"loss": 2.3938,
"step": 533
},
{
"epoch": 0.00480274134000081,
"grad_norm": 3.609283685684204,
"learning_rate": 4.845454545454546e-05,
"loss": 2.3916,
"step": 534
},
{
"epoch": 0.00481173523764126,
"grad_norm": 1.7395579814910889,
"learning_rate": 4.854545454545455e-05,
"loss": 2.5133,
"step": 535
},
{
"epoch": 0.004820729135281711,
"grad_norm": 2.8066775798797607,
"learning_rate": 4.863636363636364e-05,
"loss": 2.4421,
"step": 536
},
{
"epoch": 0.004829723032922163,
"grad_norm": 8.41027545928955,
"learning_rate": 4.872727272727273e-05,
"loss": 2.3096,
"step": 537
},
{
"epoch": 0.004838716930562613,
"grad_norm": 5.382160663604736,
"learning_rate": 4.881818181818182e-05,
"loss": 2.4843,
"step": 538
},
{
"epoch": 0.004847710828203064,
"grad_norm": 3.2162673473358154,
"learning_rate": 4.890909090909091e-05,
"loss": 2.4602,
"step": 539
},
{
"epoch": 0.004856704725843515,
"grad_norm": 2.2050821781158447,
"learning_rate": 4.9e-05,
"loss": 2.4471,
"step": 540
},
{
"epoch": 0.004865698623483966,
"grad_norm": 3.3169596195220947,
"learning_rate": 4.909090909090909e-05,
"loss": 2.5143,
"step": 541
},
{
"epoch": 0.004874692521124417,
"grad_norm": 5.980579376220703,
"learning_rate": 4.9181818181818183e-05,
"loss": 2.5125,
"step": 542
},
{
"epoch": 0.004883686418764868,
"grad_norm": 7.0032196044921875,
"learning_rate": 4.9272727272727276e-05,
"loss": 2.3135,
"step": 543
},
{
"epoch": 0.0048926803164053186,
"grad_norm": 2.2724649906158447,
"learning_rate": 4.936363636363637e-05,
"loss": 2.4153,
"step": 544
},
{
"epoch": 0.00490167421404577,
"grad_norm": 53.67974853515625,
"learning_rate": 4.945454545454546e-05,
"loss": 2.5726,
"step": 545
},
{
"epoch": 0.004910668111686221,
"grad_norm": 4.723302364349365,
"learning_rate": 4.9545454545454553e-05,
"loss": 2.3774,
"step": 546
},
{
"epoch": 0.004919662009326672,
"grad_norm": 2.6171138286590576,
"learning_rate": 4.963636363636364e-05,
"loss": 2.4464,
"step": 547
},
{
"epoch": 0.004928655906967123,
"grad_norm": 2.6987946033477783,
"learning_rate": 4.9727272727272725e-05,
"loss": 2.466,
"step": 548
},
{
"epoch": 0.004937649804607574,
"grad_norm": 10.407817840576172,
"learning_rate": 4.981818181818182e-05,
"loss": 2.4244,
"step": 549
},
{
"epoch": 0.004946643702248025,
"grad_norm": 4.350247859954834,
"learning_rate": 4.990909090909091e-05,
"loss": 2.4616,
"step": 550
},
{
"epoch": 0.004955637599888475,
"grad_norm": 6.2453107833862305,
"learning_rate": 5e-05,
"loss": 2.4426,
"step": 551
},
{
"epoch": 0.004964631497528927,
"grad_norm": 37.29060745239258,
"learning_rate": 4.9999998870263466e-05,
"loss": 2.3924,
"step": 552
},
{
"epoch": 0.004973625395169378,
"grad_norm": 74.37647247314453,
"learning_rate": 4.999999548105396e-05,
"loss": 2.4737,
"step": 553
},
{
"epoch": 0.004982619292809828,
"grad_norm": 14.813892364501953,
"learning_rate": 4.999998983237179e-05,
"loss": 2.4317,
"step": 554
},
{
"epoch": 0.0049916131904502795,
"grad_norm": 8.05070972442627,
"learning_rate": 4.9999981924217466e-05,
"loss": 2.5713,
"step": 555
},
{
"epoch": 0.005000607088090731,
"grad_norm": 6.338493824005127,
"learning_rate": 4.99999717565917e-05,
"loss": 2.396,
"step": 556
},
{
"epoch": 0.005009600985731181,
"grad_norm": 18.641860961914062,
"learning_rate": 4.999995932949543e-05,
"loss": 2.4908,
"step": 557
},
{
"epoch": 0.005018594883371632,
"grad_norm": 82.1209945678711,
"learning_rate": 4.999994464292975e-05,
"loss": 2.516,
"step": 558
},
{
"epoch": 0.0050275887810120835,
"grad_norm": 8.401792526245117,
"learning_rate": 4.9999927696896e-05,
"loss": 2.4075,
"step": 559
},
{
"epoch": 0.005036582678652534,
"grad_norm": 3.3167531490325928,
"learning_rate": 4.999990849139571e-05,
"loss": 2.5289,
"step": 560
},
{
"epoch": 0.005045576576292985,
"grad_norm": 3.5516357421875,
"learning_rate": 4.9999887026430626e-05,
"loss": 2.5607,
"step": 561
},
{
"epoch": 0.005054570473933436,
"grad_norm": 5.763249397277832,
"learning_rate": 4.999986330200268e-05,
"loss": 2.3415,
"step": 562
},
{
"epoch": 0.0050635643715738875,
"grad_norm": 5.255375385284424,
"learning_rate": 4.999983731811402e-05,
"loss": 2.4213,
"step": 563
},
{
"epoch": 0.005072558269214338,
"grad_norm": 4.464290618896484,
"learning_rate": 4.999980907476698e-05,
"loss": 2.4283,
"step": 564
},
{
"epoch": 0.005081552166854789,
"grad_norm": 4.298441410064697,
"learning_rate": 4.999977857196413e-05,
"loss": 2.3682,
"step": 565
},
{
"epoch": 0.00509054606449524,
"grad_norm": 6.778446197509766,
"learning_rate": 4.9999745809708215e-05,
"loss": 2.3273,
"step": 566
},
{
"epoch": 0.005099539962135691,
"grad_norm": 868.2053833007812,
"learning_rate": 4.999971078800221e-05,
"loss": 2.9579,
"step": 567
},
{
"epoch": 0.005108533859776142,
"grad_norm": 1063.8876953125,
"learning_rate": 4.9999673506849263e-05,
"loss": 2.9642,
"step": 568
},
{
"epoch": 0.005117527757416593,
"grad_norm": 12.19197940826416,
"learning_rate": 4.999963396625276e-05,
"loss": 2.4405,
"step": 569
},
{
"epoch": 0.0051265216550570435,
"grad_norm": 24.950546264648438,
"learning_rate": 4.9999592166216256e-05,
"loss": 2.4331,
"step": 570
},
{
"epoch": 0.005135515552697495,
"grad_norm": 943.664794921875,
"learning_rate": 4.9999548106743545e-05,
"loss": 3.0833,
"step": 571
},
{
"epoch": 0.005144509450337946,
"grad_norm": 1022.8340454101562,
"learning_rate": 4.99995017878386e-05,
"loss": 3.0055,
"step": 572
},
{
"epoch": 0.005153503347978396,
"grad_norm": 29.700937271118164,
"learning_rate": 4.999945320950562e-05,
"loss": 2.42,
"step": 573
},
{
"epoch": 0.005162497245618848,
"grad_norm": 3.144054889678955,
"learning_rate": 4.9999402371748984e-05,
"loss": 2.3299,
"step": 574
},
{
"epoch": 0.005171491143259299,
"grad_norm": 1.2397363185882568,
"learning_rate": 4.999934927457328e-05,
"loss": 2.4631,
"step": 575
},
{
"epoch": 0.005180485040899749,
"grad_norm": 5.255294322967529,
"learning_rate": 4.999929391798332e-05,
"loss": 2.3657,
"step": 576
},
{
"epoch": 0.0051894789385402,
"grad_norm": 5.393404006958008,
"learning_rate": 4.9999236301984104e-05,
"loss": 2.5346,
"step": 577
},
{
"epoch": 0.005198472836180652,
"grad_norm": 3.3912456035614014,
"learning_rate": 4.999917642658084e-05,
"loss": 2.4842,
"step": 578
},
{
"epoch": 0.005207466733821102,
"grad_norm": 1.8318792581558228,
"learning_rate": 4.999911429177893e-05,
"loss": 2.4364,
"step": 579
},
{
"epoch": 0.005216460631461553,
"grad_norm": 26.505447387695312,
"learning_rate": 4.9999049897584e-05,
"loss": 2.313,
"step": 580
},
{
"epoch": 0.005225454529102004,
"grad_norm": 3.3377368450164795,
"learning_rate": 4.9998983244001874e-05,
"loss": 2.3303,
"step": 581
},
{
"epoch": 0.005234448426742456,
"grad_norm": 6.476102352142334,
"learning_rate": 4.999891433103856e-05,
"loss": 2.4122,
"step": 582
},
{
"epoch": 0.005243442324382906,
"grad_norm": 5.718252182006836,
"learning_rate": 4.99988431587003e-05,
"loss": 2.4131,
"step": 583
},
{
"epoch": 0.005252436222023357,
"grad_norm": 20.037609100341797,
"learning_rate": 4.999876972699352e-05,
"loss": 2.5064,
"step": 584
},
{
"epoch": 0.0052614301196638085,
"grad_norm": 10.646677017211914,
"learning_rate": 4.9998694035924866e-05,
"loss": 2.4485,
"step": 585
},
{
"epoch": 0.005270424017304259,
"grad_norm": 3.244764804840088,
"learning_rate": 4.999861608550116e-05,
"loss": 2.4147,
"step": 586
},
{
"epoch": 0.00527941791494471,
"grad_norm": 2.5757675170898438,
"learning_rate": 4.999853587572947e-05,
"loss": 2.3996,
"step": 587
},
{
"epoch": 0.005288411812585161,
"grad_norm": 2.946089744567871,
"learning_rate": 4.9998453406617025e-05,
"loss": 2.4282,
"step": 588
},
{
"epoch": 0.005297405710225612,
"grad_norm": 4.174030303955078,
"learning_rate": 4.999836867817129e-05,
"loss": 2.5452,
"step": 589
},
{
"epoch": 0.005306399607866063,
"grad_norm": 2.7353034019470215,
"learning_rate": 4.999828169039992e-05,
"loss": 2.4936,
"step": 590
},
{
"epoch": 0.005315393505506514,
"grad_norm": 2.9000155925750732,
"learning_rate": 4.9998192443310776e-05,
"loss": 2.4056,
"step": 591
},
{
"epoch": 0.0053243874031469645,
"grad_norm": 2.387249231338501,
"learning_rate": 4.999810093691192e-05,
"loss": 2.5121,
"step": 592
},
{
"epoch": 0.005333381300787416,
"grad_norm": 1.5600703954696655,
"learning_rate": 4.9998007171211634e-05,
"loss": 2.2865,
"step": 593
},
{
"epoch": 0.005342375198427867,
"grad_norm": 3.0678749084472656,
"learning_rate": 4.999791114621839e-05,
"loss": 2.3634,
"step": 594
},
{
"epoch": 0.005351369096068317,
"grad_norm": 1.6179578304290771,
"learning_rate": 4.999781286194085e-05,
"loss": 2.3452,
"step": 595
},
{
"epoch": 0.0053603629937087685,
"grad_norm": 2.9978394508361816,
"learning_rate": 4.999771231838792e-05,
"loss": 2.4297,
"step": 596
},
{
"epoch": 0.00536935689134922,
"grad_norm": 2.1894114017486572,
"learning_rate": 4.999760951556867e-05,
"loss": 2.4439,
"step": 597
},
{
"epoch": 0.00537835078898967,
"grad_norm": 1.9343936443328857,
"learning_rate": 4.9997504453492395e-05,
"loss": 2.3994,
"step": 598
},
{
"epoch": 0.005387344686630121,
"grad_norm": 1.511374592781067,
"learning_rate": 4.99973971321686e-05,
"loss": 2.3782,
"step": 599
},
{
"epoch": 0.0053963385842705726,
"grad_norm": 2.0215413570404053,
"learning_rate": 4.999728755160697e-05,
"loss": 2.3963,
"step": 600
},
{
"epoch": 0.005405332481911024,
"grad_norm": 2.0300323963165283,
"learning_rate": 4.999717571181742e-05,
"loss": 2.381,
"step": 601
},
{
"epoch": 0.005414326379551474,
"grad_norm": 1.4765416383743286,
"learning_rate": 4.999706161281005e-05,
"loss": 2.5344,
"step": 602
},
{
"epoch": 0.005423320277191925,
"grad_norm": 1.7686407566070557,
"learning_rate": 4.999694525459518e-05,
"loss": 2.3815,
"step": 603
},
{
"epoch": 0.005432314174832377,
"grad_norm": 2.907012939453125,
"learning_rate": 4.9996826637183326e-05,
"loss": 2.4307,
"step": 604
},
{
"epoch": 0.005441308072472827,
"grad_norm": 1.7695300579071045,
"learning_rate": 4.9996705760585194e-05,
"loss": 2.4209,
"step": 605
},
{
"epoch": 0.005450301970113278,
"grad_norm": 2.1694586277008057,
"learning_rate": 4.9996582624811725e-05,
"loss": 2.3793,
"step": 606
},
{
"epoch": 0.005459295867753729,
"grad_norm": 2.7593190670013428,
"learning_rate": 4.9996457229874046e-05,
"loss": 2.4173,
"step": 607
},
{
"epoch": 0.00546828976539418,
"grad_norm": 3.392613649368286,
"learning_rate": 4.999632957578349e-05,
"loss": 2.4865,
"step": 608
},
{
"epoch": 0.005477283663034631,
"grad_norm": 4.626733303070068,
"learning_rate": 4.999619966255158e-05,
"loss": 2.4389,
"step": 609
},
{
"epoch": 0.005486277560675082,
"grad_norm": 1.0661565065383911,
"learning_rate": 4.9996067490190066e-05,
"loss": 2.4327,
"step": 610
},
{
"epoch": 0.005495271458315533,
"grad_norm": 1.619776725769043,
"learning_rate": 4.999593305871091e-05,
"loss": 2.4746,
"step": 611
},
{
"epoch": 0.005504265355955984,
"grad_norm": 1.2350534200668335,
"learning_rate": 4.9995796368126234e-05,
"loss": 2.3905,
"step": 612
},
{
"epoch": 0.005513259253596435,
"grad_norm": 2.012613296508789,
"learning_rate": 4.999565741844841e-05,
"loss": 2.4483,
"step": 613
},
{
"epoch": 0.005522253151236885,
"grad_norm": 1.6660722494125366,
"learning_rate": 4.9995516209689996e-05,
"loss": 2.2968,
"step": 614
},
{
"epoch": 0.005531247048877337,
"grad_norm": 1.593167781829834,
"learning_rate": 4.999537274186375e-05,
"loss": 2.4306,
"step": 615
},
{
"epoch": 0.005540240946517788,
"grad_norm": 1.9116402864456177,
"learning_rate": 4.999522701498263e-05,
"loss": 2.371,
"step": 616
},
{
"epoch": 0.005549234844158238,
"grad_norm": 1.4229307174682617,
"learning_rate": 4.9995079029059824e-05,
"loss": 2.3387,
"step": 617
},
{
"epoch": 0.0055582287417986894,
"grad_norm": 2.0541815757751465,
"learning_rate": 4.9994928784108694e-05,
"loss": 2.3974,
"step": 618
},
{
"epoch": 0.005567222639439141,
"grad_norm": 1.4975587129592896,
"learning_rate": 4.999477628014282e-05,
"loss": 2.3433,
"step": 619
},
{
"epoch": 0.005576216537079592,
"grad_norm": 2.6742827892303467,
"learning_rate": 4.999462151717599e-05,
"loss": 2.3205,
"step": 620
},
{
"epoch": 0.005585210434720042,
"grad_norm": 2.0126726627349854,
"learning_rate": 4.9994464495222196e-05,
"loss": 2.2967,
"step": 621
},
{
"epoch": 0.0055942043323604935,
"grad_norm": 1.6105343103408813,
"learning_rate": 4.999430521429561e-05,
"loss": 2.4415,
"step": 622
},
{
"epoch": 0.005603198230000945,
"grad_norm": 2.3501596450805664,
"learning_rate": 4.9994143674410655e-05,
"loss": 2.3561,
"step": 623
},
{
"epoch": 0.005612192127641395,
"grad_norm": 2.1336519718170166,
"learning_rate": 4.99939798755819e-05,
"loss": 2.3675,
"step": 624
},
{
"epoch": 0.005621186025281846,
"grad_norm": 2.012542247772217,
"learning_rate": 4.999381381782417e-05,
"loss": 2.4565,
"step": 625
},
{
"epoch": 0.0056301799229222975,
"grad_norm": 1.5894619226455688,
"learning_rate": 4.999364550115248e-05,
"loss": 2.3818,
"step": 626
},
{
"epoch": 0.005639173820562748,
"grad_norm": 1.5222396850585938,
"learning_rate": 4.999347492558202e-05,
"loss": 2.308,
"step": 627
},
{
"epoch": 0.005648167718203199,
"grad_norm": 2.7180862426757812,
"learning_rate": 4.999330209112822e-05,
"loss": 2.5078,
"step": 628
},
{
"epoch": 0.00565716161584365,
"grad_norm": 1.638213038444519,
"learning_rate": 4.99931269978067e-05,
"loss": 2.4567,
"step": 629
},
{
"epoch": 0.005666155513484101,
"grad_norm": 2.0418717861175537,
"learning_rate": 4.999294964563328e-05,
"loss": 2.4698,
"step": 630
},
{
"epoch": 0.005675149411124552,
"grad_norm": 2.8325579166412354,
"learning_rate": 4.9992770034623994e-05,
"loss": 2.273,
"step": 631
},
{
"epoch": 0.005684143308765003,
"grad_norm": 2.4228262901306152,
"learning_rate": 4.9992588164795064e-05,
"loss": 2.3302,
"step": 632
},
{
"epoch": 0.0056931372064054535,
"grad_norm": 2.3301501274108887,
"learning_rate": 4.9992404036162944e-05,
"loss": 2.3703,
"step": 633
},
{
"epoch": 0.005702131104045905,
"grad_norm": 2.912381172180176,
"learning_rate": 4.9992217648744254e-05,
"loss": 2.4629,
"step": 634
},
{
"epoch": 0.005711125001686356,
"grad_norm": 2.8964321613311768,
"learning_rate": 4.999202900255586e-05,
"loss": 2.5102,
"step": 635
},
{
"epoch": 0.005720118899326806,
"grad_norm": 2.1238880157470703,
"learning_rate": 4.999183809761481e-05,
"loss": 2.3765,
"step": 636
},
{
"epoch": 0.0057291127969672576,
"grad_norm": 1.820768117904663,
"learning_rate": 4.999164493393835e-05,
"loss": 2.4077,
"step": 637
},
{
"epoch": 0.005738106694607709,
"grad_norm": 1.8464839458465576,
"learning_rate": 4.999144951154394e-05,
"loss": 2.4221,
"step": 638
},
{
"epoch": 0.00574710059224816,
"grad_norm": 1.9260097742080688,
"learning_rate": 4.999125183044924e-05,
"loss": 2.2966,
"step": 639
},
{
"epoch": 0.00575609448988861,
"grad_norm": 2.3923349380493164,
"learning_rate": 4.9991051890672114e-05,
"loss": 2.3924,
"step": 640
},
{
"epoch": 0.005765088387529062,
"grad_norm": 1.808197021484375,
"learning_rate": 4.999084969223063e-05,
"loss": 2.4124,
"step": 641
},
{
"epoch": 0.005774082285169513,
"grad_norm": 2.1454949378967285,
"learning_rate": 4.999064523514309e-05,
"loss": 2.4176,
"step": 642
},
{
"epoch": 0.005783076182809963,
"grad_norm": 1.7350833415985107,
"learning_rate": 4.999043851942793e-05,
"loss": 2.3737,
"step": 643
},
{
"epoch": 0.005792070080450414,
"grad_norm": 2.89910888671875,
"learning_rate": 4.999022954510386e-05,
"loss": 2.5145,
"step": 644
},
{
"epoch": 0.005801063978090866,
"grad_norm": 2.385995388031006,
"learning_rate": 4.999001831218977e-05,
"loss": 2.3942,
"step": 645
},
{
"epoch": 0.005810057875731316,
"grad_norm": 38.989044189453125,
"learning_rate": 4.9989804820704735e-05,
"loss": 2.3923,
"step": 646
},
{
"epoch": 0.005819051773371767,
"grad_norm": 105.62820434570312,
"learning_rate": 4.998958907066805e-05,
"loss": 2.4966,
"step": 647
},
{
"epoch": 0.0058280456710122185,
"grad_norm": 10.19742202758789,
"learning_rate": 4.998937106209923e-05,
"loss": 2.3541,
"step": 648
},
{
"epoch": 0.005837039568652669,
"grad_norm": 2.646786689758301,
"learning_rate": 4.998915079501798e-05,
"loss": 2.3183,
"step": 649
},
{
"epoch": 0.00584603346629312,
"grad_norm": 2.7256410121917725,
"learning_rate": 4.998892826944418e-05,
"loss": 2.5195,
"step": 650
},
{
"epoch": 0.005855027363933571,
"grad_norm": 1.814576268196106,
"learning_rate": 4.998870348539797e-05,
"loss": 2.4928,
"step": 651
},
{
"epoch": 0.005864021261574022,
"grad_norm": 65.62152862548828,
"learning_rate": 4.998847644289965e-05,
"loss": 2.346,
"step": 652
},
{
"epoch": 0.005873015159214473,
"grad_norm": 156.02301025390625,
"learning_rate": 4.998824714196974e-05,
"loss": 2.5314,
"step": 653
},
{
"epoch": 0.005882009056854924,
"grad_norm": 18.26453971862793,
"learning_rate": 4.9988015582628974e-05,
"loss": 2.4169,
"step": 654
},
{
"epoch": 0.005891002954495375,
"grad_norm": 1.6023188829421997,
"learning_rate": 4.998778176489827e-05,
"loss": 2.3979,
"step": 655
},
{
"epoch": 0.005899996852135826,
"grad_norm": 3.3964672088623047,
"learning_rate": 4.998754568879876e-05,
"loss": 2.3552,
"step": 656
},
{
"epoch": 0.005908990749776277,
"grad_norm": 318.99041748046875,
"learning_rate": 4.9987307354351796e-05,
"loss": 2.4107,
"step": 657
},
{
"epoch": 0.005917984647416728,
"grad_norm": 1.386103868484497,
"learning_rate": 4.998706676157891e-05,
"loss": 2.3398,
"step": 658
},
{
"epoch": 0.0059269785450571785,
"grad_norm": 1.273573875427246,
"learning_rate": 4.9986823910501834e-05,
"loss": 2.3497,
"step": 659
},
{
"epoch": 0.00593597244269763,
"grad_norm": 1.5252665281295776,
"learning_rate": 4.998657880114254e-05,
"loss": 2.3604,
"step": 660
},
{
"epoch": 0.005944966340338081,
"grad_norm": 7.399944305419922,
"learning_rate": 4.9986331433523156e-05,
"loss": 2.3463,
"step": 661
},
{
"epoch": 0.005953960237978531,
"grad_norm": 348.9754943847656,
"learning_rate": 4.9986081807666055e-05,
"loss": 2.5269,
"step": 662
},
{
"epoch": 0.0059629541356189825,
"grad_norm": 1212.12109375,
"learning_rate": 4.9985829923593786e-05,
"loss": 2.9826,
"step": 663
},
{
"epoch": 0.005971948033259434,
"grad_norm": 250.246826171875,
"learning_rate": 4.998557578132912e-05,
"loss": 2.4467,
"step": 664
},
{
"epoch": 0.005980941930899884,
"grad_norm": 2.0885791778564453,
"learning_rate": 4.998531938089503e-05,
"loss": 2.4626,
"step": 665
},
{
"epoch": 0.005989935828540335,
"grad_norm": 1.5140365362167358,
"learning_rate": 4.998506072231469e-05,
"loss": 2.3628,
"step": 666
},
{
"epoch": 0.005998929726180787,
"grad_norm": 1.3232437372207642,
"learning_rate": 4.998479980561148e-05,
"loss": 2.3055,
"step": 667
},
{
"epoch": 0.006007923623821237,
"grad_norm": 1.0808279514312744,
"learning_rate": 4.9984536630808964e-05,
"loss": 2.3661,
"step": 668
},
{
"epoch": 0.006016917521461688,
"grad_norm": 1.9327250719070435,
"learning_rate": 4.998427119793094e-05,
"loss": 2.3704,
"step": 669
},
{
"epoch": 0.006025911419102139,
"grad_norm": 4.848217964172363,
"learning_rate": 4.99840035070014e-05,
"loss": 2.4662,
"step": 670
},
{
"epoch": 0.00603490531674259,
"grad_norm": 4.094484806060791,
"learning_rate": 4.9983733558044534e-05,
"loss": 2.4309,
"step": 671
},
{
"epoch": 0.006043899214383041,
"grad_norm": 2.386913537979126,
"learning_rate": 4.9983461351084735e-05,
"loss": 2.3768,
"step": 672
},
{
"epoch": 0.006052893112023492,
"grad_norm": 2.4711503982543945,
"learning_rate": 4.9983186886146606e-05,
"loss": 2.4355,
"step": 673
},
{
"epoch": 0.0060618870096639434,
"grad_norm": 1.4953632354736328,
"learning_rate": 4.998291016325496e-05,
"loss": 2.2473,
"step": 674
},
{
"epoch": 0.006070880907304394,
"grad_norm": 3.127026081085205,
"learning_rate": 4.99826311824348e-05,
"loss": 2.3246,
"step": 675
},
{
"epoch": 0.006079874804944845,
"grad_norm": 3.4192206859588623,
"learning_rate": 4.998234994371135e-05,
"loss": 2.3357,
"step": 676
},
{
"epoch": 0.006088868702585296,
"grad_norm": 1.615310788154602,
"learning_rate": 4.998206644711002e-05,
"loss": 2.3901,
"step": 677
},
{
"epoch": 0.006097862600225747,
"grad_norm": 1.7600175142288208,
"learning_rate": 4.998178069265643e-05,
"loss": 2.3917,
"step": 678
},
{
"epoch": 0.006106856497866198,
"grad_norm": 1.5513185262680054,
"learning_rate": 4.998149268037641e-05,
"loss": 2.3612,
"step": 679
},
{
"epoch": 0.006115850395506649,
"grad_norm": 1.5127142667770386,
"learning_rate": 4.9981202410295984e-05,
"loss": 2.3874,
"step": 680
},
{
"epoch": 0.006124844293147099,
"grad_norm": 4.144172668457031,
"learning_rate": 4.99809098824414e-05,
"loss": 2.4896,
"step": 681
},
{
"epoch": 0.006133838190787551,
"grad_norm": 2.2086148262023926,
"learning_rate": 4.9980615096839076e-05,
"loss": 2.418,
"step": 682
},
{
"epoch": 0.006142832088428002,
"grad_norm": 19.644895553588867,
"learning_rate": 4.9980318053515686e-05,
"loss": 2.5861,
"step": 683
},
{
"epoch": 0.006151825986068452,
"grad_norm": 5.292164325714111,
"learning_rate": 4.9980018752498035e-05,
"loss": 2.3599,
"step": 684
},
{
"epoch": 0.0061608198837089035,
"grad_norm": 59.39234161376953,
"learning_rate": 4.997971719381321e-05,
"loss": 2.4926,
"step": 685
},
{
"epoch": 0.006169813781349355,
"grad_norm": 164.944580078125,
"learning_rate": 4.9979413377488446e-05,
"loss": 2.426,
"step": 686
},
{
"epoch": 0.006178807678989805,
"grad_norm": 2.1656153202056885,
"learning_rate": 4.997910730355121e-05,
"loss": 2.3332,
"step": 687
},
{
"epoch": 0.006187801576630256,
"grad_norm": 361.9581604003906,
"learning_rate": 4.997879897202916e-05,
"loss": 2.4859,
"step": 688
},
{
"epoch": 0.0061967954742707075,
"grad_norm": 81.85458374023438,
"learning_rate": 4.9978488382950165e-05,
"loss": 2.4685,
"step": 689
},
{
"epoch": 0.006205789371911158,
"grad_norm": 5.230372428894043,
"learning_rate": 4.9978175536342286e-05,
"loss": 2.419,
"step": 690
},
{
"epoch": 0.006214783269551609,
"grad_norm": 2.3871922492980957,
"learning_rate": 4.997786043223382e-05,
"loss": 2.3261,
"step": 691
},
{
"epoch": 0.00622377716719206,
"grad_norm": 7.315066814422607,
"learning_rate": 4.997754307065322e-05,
"loss": 2.4856,
"step": 692
},
{
"epoch": 0.0062327710648325116,
"grad_norm": 4.792977333068848,
"learning_rate": 4.997722345162918e-05,
"loss": 2.373,
"step": 693
},
{
"epoch": 0.006241764962472962,
"grad_norm": 4.047577381134033,
"learning_rate": 4.997690157519059e-05,
"loss": 2.3665,
"step": 694
},
{
"epoch": 0.006250758860113413,
"grad_norm": 1.6489315032958984,
"learning_rate": 4.997657744136654e-05,
"loss": 2.2563,
"step": 695
},
{
"epoch": 0.006259752757753864,
"grad_norm": 1.2394626140594482,
"learning_rate": 4.997625105018633e-05,
"loss": 2.2744,
"step": 696
},
{
"epoch": 0.006268746655394315,
"grad_norm": 1.6913032531738281,
"learning_rate": 4.997592240167945e-05,
"loss": 2.3004,
"step": 697
},
{
"epoch": 0.006277740553034766,
"grad_norm": 1.1033881902694702,
"learning_rate": 4.99755914958756e-05,
"loss": 2.3581,
"step": 698
},
{
"epoch": 0.006286734450675217,
"grad_norm": 1.4236403703689575,
"learning_rate": 4.99752583328047e-05,
"loss": 2.3757,
"step": 699
},
{
"epoch": 0.0062957283483156675,
"grad_norm": 1.4072418212890625,
"learning_rate": 4.9974922912496846e-05,
"loss": 2.367,
"step": 700
},
{
"epoch": 0.006304722245956119,
"grad_norm": 1.4323331117630005,
"learning_rate": 4.997458523498236e-05,
"loss": 2.368,
"step": 701
},
{
"epoch": 0.00631371614359657,
"grad_norm": 3.138033390045166,
"learning_rate": 4.9974245300291763e-05,
"loss": 2.4071,
"step": 702
},
{
"epoch": 0.00632271004123702,
"grad_norm": 1.720033049583435,
"learning_rate": 4.997390310845578e-05,
"loss": 2.4112,
"step": 703
},
{
"epoch": 0.006331703938877472,
"grad_norm": 1.968177318572998,
"learning_rate": 4.9973558659505325e-05,
"loss": 2.3203,
"step": 704
},
{
"epoch": 0.006340697836517923,
"grad_norm": 2.108110189437866,
"learning_rate": 4.997321195347154e-05,
"loss": 2.4735,
"step": 705
},
{
"epoch": 0.006349691734158373,
"grad_norm": 1.26445734500885,
"learning_rate": 4.997286299038576e-05,
"loss": 2.527,
"step": 706
},
{
"epoch": 0.006358685631798824,
"grad_norm": 2.095703601837158,
"learning_rate": 4.997251177027952e-05,
"loss": 2.4369,
"step": 707
},
{
"epoch": 0.006367679529439276,
"grad_norm": 1.7823727130889893,
"learning_rate": 4.9972158293184565e-05,
"loss": 2.3546,
"step": 708
},
{
"epoch": 0.006376673427079726,
"grad_norm": 1.430592656135559,
"learning_rate": 4.997180255913284e-05,
"loss": 2.4506,
"step": 709
},
{
"epoch": 0.006385667324720177,
"grad_norm": 2.089142084121704,
"learning_rate": 4.99714445681565e-05,
"loss": 2.3887,
"step": 710
},
{
"epoch": 0.0063946612223606284,
"grad_norm": 1.1167904138565063,
"learning_rate": 4.99710843202879e-05,
"loss": 2.3572,
"step": 711
},
{
"epoch": 0.00640365512000108,
"grad_norm": 1.7042675018310547,
"learning_rate": 4.9970721815559585e-05,
"loss": 2.3336,
"step": 712
},
{
"epoch": 0.00641264901764153,
"grad_norm": 1.5516256093978882,
"learning_rate": 4.9970357054004335e-05,
"loss": 2.3428,
"step": 713
},
{
"epoch": 0.006421642915281981,
"grad_norm": 0.7995926737785339,
"learning_rate": 4.996999003565511e-05,
"loss": 2.3922,
"step": 714
},
{
"epoch": 0.0064306368129224325,
"grad_norm": 1.3005210161209106,
"learning_rate": 4.9969620760545074e-05,
"loss": 2.3788,
"step": 715
},
{
"epoch": 0.006439630710562883,
"grad_norm": 1.9167686700820923,
"learning_rate": 4.996924922870762e-05,
"loss": 2.4571,
"step": 716
},
{
"epoch": 0.006448624608203334,
"grad_norm": 1.8720837831497192,
"learning_rate": 4.996887544017631e-05,
"loss": 2.3795,
"step": 717
},
{
"epoch": 0.006457618505843785,
"grad_norm": 1.3605936765670776,
"learning_rate": 4.996849939498493e-05,
"loss": 2.3193,
"step": 718
},
{
"epoch": 0.006466612403484236,
"grad_norm": 1.6093920469284058,
"learning_rate": 4.9968121093167476e-05,
"loss": 2.4547,
"step": 719
},
{
"epoch": 0.006475606301124687,
"grad_norm": 1.8580682277679443,
"learning_rate": 4.9967740534758126e-05,
"loss": 2.3832,
"step": 720
},
{
"epoch": 0.006484600198765138,
"grad_norm": 1.1695489883422852,
"learning_rate": 4.9967357719791286e-05,
"loss": 2.3307,
"step": 721
},
{
"epoch": 0.0064935940964055885,
"grad_norm": 1.8113250732421875,
"learning_rate": 4.9966972648301536e-05,
"loss": 2.4475,
"step": 722
},
{
"epoch": 0.00650258799404604,
"grad_norm": 1.3906854391098022,
"learning_rate": 4.99665853203237e-05,
"loss": 2.3736,
"step": 723
},
{
"epoch": 0.006511581891686491,
"grad_norm": 1.8299096822738647,
"learning_rate": 4.996619573589277e-05,
"loss": 2.3624,
"step": 724
},
{
"epoch": 0.006520575789326941,
"grad_norm": 2.2322638034820557,
"learning_rate": 4.9965803895043964e-05,
"loss": 2.4239,
"step": 725
},
{
"epoch": 0.0065295696869673925,
"grad_norm": 1.2125953435897827,
"learning_rate": 4.996540979781269e-05,
"loss": 2.3379,
"step": 726
},
{
"epoch": 0.006538563584607844,
"grad_norm": 2.416262149810791,
"learning_rate": 4.996501344423456e-05,
"loss": 2.4828,
"step": 727
},
{
"epoch": 0.006547557482248294,
"grad_norm": 1.60369873046875,
"learning_rate": 4.996461483434542e-05,
"loss": 2.4416,
"step": 728
},
{
"epoch": 0.006556551379888745,
"grad_norm": 1.5307985544204712,
"learning_rate": 4.996421396818127e-05,
"loss": 2.4244,
"step": 729
},
{
"epoch": 0.0065655452775291966,
"grad_norm": 0.987632691860199,
"learning_rate": 4.996381084577836e-05,
"loss": 2.3704,
"step": 730
},
{
"epoch": 0.006574539175169648,
"grad_norm": 0.9207761883735657,
"learning_rate": 4.996340546717311e-05,
"loss": 2.3548,
"step": 731
},
{
"epoch": 0.006583533072810098,
"grad_norm": 1.4703799486160278,
"learning_rate": 4.996299783240217e-05,
"loss": 2.5025,
"step": 732
},
{
"epoch": 0.006592526970450549,
"grad_norm": 1.1804099082946777,
"learning_rate": 4.996258794150236e-05,
"loss": 2.4147,
"step": 733
},
{
"epoch": 0.006601520868091001,
"grad_norm": 2.1551332473754883,
"learning_rate": 4.9962175794510744e-05,
"loss": 2.4238,
"step": 734
},
{
"epoch": 0.006610514765731451,
"grad_norm": 1.1678895950317383,
"learning_rate": 4.996176139146457e-05,
"loss": 2.4213,
"step": 735
},
{
"epoch": 0.006619508663371902,
"grad_norm": 2.0252041816711426,
"learning_rate": 4.9961344732401294e-05,
"loss": 2.4356,
"step": 736
},
{
"epoch": 0.006628502561012353,
"grad_norm": 0.9819325804710388,
"learning_rate": 4.9960925817358564e-05,
"loss": 2.3012,
"step": 737
},
{
"epoch": 0.006637496458652804,
"grad_norm": 1.136311650276184,
"learning_rate": 4.996050464637423e-05,
"loss": 2.3593,
"step": 738
},
{
"epoch": 0.006646490356293255,
"grad_norm": 0.9213350415229797,
"learning_rate": 4.996008121948639e-05,
"loss": 2.3494,
"step": 739
},
{
"epoch": 0.006655484253933706,
"grad_norm": 1.3651710748672485,
"learning_rate": 4.995965553673329e-05,
"loss": 2.5314,
"step": 740
},
{
"epoch": 0.006664478151574157,
"grad_norm": 1.3020514249801636,
"learning_rate": 4.995922759815339e-05,
"loss": 2.2532,
"step": 741
},
{
"epoch": 0.006673472049214608,
"grad_norm": 0.775888204574585,
"learning_rate": 4.99587974037854e-05,
"loss": 2.3791,
"step": 742
},
{
"epoch": 0.006682465946855059,
"grad_norm": 1.7162033319473267,
"learning_rate": 4.995836495366818e-05,
"loss": 2.3848,
"step": 743
},
{
"epoch": 0.006691459844495509,
"grad_norm": 1.0520635843276978,
"learning_rate": 4.9957930247840824e-05,
"loss": 2.3923,
"step": 744
},
{
"epoch": 0.006700453742135961,
"grad_norm": 1.6035592555999756,
"learning_rate": 4.99574932863426e-05,
"loss": 2.5291,
"step": 745
},
{
"epoch": 0.006709447639776412,
"grad_norm": 1.0390980243682861,
"learning_rate": 4.9957054069213025e-05,
"loss": 2.3656,
"step": 746
},
{
"epoch": 0.006718441537416863,
"grad_norm": 1.1373931169509888,
"learning_rate": 4.995661259649179e-05,
"loss": 2.3262,
"step": 747
},
{
"epoch": 0.0067274354350573135,
"grad_norm": 1.5664631128311157,
"learning_rate": 4.9956168868218787e-05,
"loss": 2.3749,
"step": 748
},
{
"epoch": 0.006736429332697765,
"grad_norm": 1.1574504375457764,
"learning_rate": 4.9955722884434114e-05,
"loss": 2.3114,
"step": 749
},
{
"epoch": 0.006745423230338216,
"grad_norm": 0.8074120283126831,
"learning_rate": 4.99552746451781e-05,
"loss": 2.4724,
"step": 750
},
{
"epoch": 0.006754417127978666,
"grad_norm": 0.8828337788581848,
"learning_rate": 4.995482415049123e-05,
"loss": 2.3239,
"step": 751
},
{
"epoch": 0.0067634110256191175,
"grad_norm": 1.0525522232055664,
"learning_rate": 4.995437140041424e-05,
"loss": 2.3863,
"step": 752
},
{
"epoch": 0.006772404923259569,
"grad_norm": 2.016676902770996,
"learning_rate": 4.995391639498803e-05,
"loss": 2.402,
"step": 753
},
{
"epoch": 0.006781398820900019,
"grad_norm": 1.6840852499008179,
"learning_rate": 4.995345913425374e-05,
"loss": 2.3384,
"step": 754
},
{
"epoch": 0.00679039271854047,
"grad_norm": 1.2580803632736206,
"learning_rate": 4.99529996182527e-05,
"loss": 2.3912,
"step": 755
},
{
"epoch": 0.0067993866161809215,
"grad_norm": 2.412668466567993,
"learning_rate": 4.995253784702642e-05,
"loss": 2.3987,
"step": 756
},
{
"epoch": 0.006808380513821372,
"grad_norm": 2.4408586025238037,
"learning_rate": 4.995207382061666e-05,
"loss": 2.3739,
"step": 757
},
{
"epoch": 0.006817374411461823,
"grad_norm": 2.665259599685669,
"learning_rate": 4.995160753906534e-05,
"loss": 2.5046,
"step": 758
},
{
"epoch": 0.006826368309102274,
"grad_norm": 2.536120653152466,
"learning_rate": 4.9951139002414596e-05,
"loss": 2.4292,
"step": 759
},
{
"epoch": 0.006835362206742725,
"grad_norm": 1.1645842790603638,
"learning_rate": 4.995066821070679e-05,
"loss": 2.373,
"step": 760
},
{
"epoch": 0.006844356104383176,
"grad_norm": 2.160353899002075,
"learning_rate": 4.995019516398447e-05,
"loss": 2.5006,
"step": 761
},
{
"epoch": 0.006853350002023627,
"grad_norm": 2.830785036087036,
"learning_rate": 4.994971986229038e-05,
"loss": 2.471,
"step": 762
},
{
"epoch": 0.0068623438996640775,
"grad_norm": 2.361668825149536,
"learning_rate": 4.9949242305667486e-05,
"loss": 2.3206,
"step": 763
},
{
"epoch": 0.006871337797304529,
"grad_norm": 1.6988106966018677,
"learning_rate": 4.994876249415894e-05,
"loss": 2.4014,
"step": 764
},
{
"epoch": 0.00688033169494498,
"grad_norm": 0.9352366924285889,
"learning_rate": 4.9948280427808115e-05,
"loss": 2.402,
"step": 765
},
{
"epoch": 0.006889325592585431,
"grad_norm": 1.5733873844146729,
"learning_rate": 4.994779610665858e-05,
"loss": 2.421,
"step": 766
},
{
"epoch": 0.006898319490225882,
"grad_norm": 1.1752417087554932,
"learning_rate": 4.9947309530754104e-05,
"loss": 2.2774,
"step": 767
},
{
"epoch": 0.006907313387866333,
"grad_norm": 1.1936308145523071,
"learning_rate": 4.9946820700138666e-05,
"loss": 2.2928,
"step": 768
},
{
"epoch": 0.006916307285506784,
"grad_norm": 1.9820133447647095,
"learning_rate": 4.994632961485644e-05,
"loss": 2.3744,
"step": 769
},
{
"epoch": 0.006925301183147234,
"grad_norm": 2.30741810798645,
"learning_rate": 4.9945836274951805e-05,
"loss": 2.3732,
"step": 770
},
{
"epoch": 0.006934295080787686,
"grad_norm": 2.1270976066589355,
"learning_rate": 4.994534068046937e-05,
"loss": 2.3706,
"step": 771
},
{
"epoch": 0.006943288978428137,
"grad_norm": 1.8759336471557617,
"learning_rate": 4.994484283145391e-05,
"loss": 2.4928,
"step": 772
},
{
"epoch": 0.006952282876068587,
"grad_norm": 1.6399668455123901,
"learning_rate": 4.994434272795042e-05,
"loss": 2.4569,
"step": 773
},
{
"epoch": 0.006961276773709038,
"grad_norm": 1.5422440767288208,
"learning_rate": 4.99438403700041e-05,
"loss": 2.4189,
"step": 774
},
{
"epoch": 0.00697027067134949,
"grad_norm": 2.132363796234131,
"learning_rate": 4.994333575766036e-05,
"loss": 2.4798,
"step": 775
},
{
"epoch": 0.00697926456898994,
"grad_norm": 1.4157288074493408,
"learning_rate": 4.99428288909648e-05,
"loss": 2.4164,
"step": 776
},
{
"epoch": 0.006988258466630391,
"grad_norm": 1.1873482465744019,
"learning_rate": 4.994231976996323e-05,
"loss": 2.413,
"step": 777
},
{
"epoch": 0.0069972523642708425,
"grad_norm": 0.9601787328720093,
"learning_rate": 4.994180839470166e-05,
"loss": 2.3641,
"step": 778
},
{
"epoch": 0.007006246261911293,
"grad_norm": 1.6519640684127808,
"learning_rate": 4.994129476522632e-05,
"loss": 2.3609,
"step": 779
},
{
"epoch": 0.007015240159551744,
"grad_norm": 1.522507667541504,
"learning_rate": 4.994077888158362e-05,
"loss": 2.2939,
"step": 780
},
{
"epoch": 0.007024234057192195,
"grad_norm": 1.2533833980560303,
"learning_rate": 4.994026074382019e-05,
"loss": 2.3539,
"step": 781
},
{
"epoch": 0.007033227954832646,
"grad_norm": 1.0521180629730225,
"learning_rate": 4.9939740351982856e-05,
"loss": 2.3217,
"step": 782
},
{
"epoch": 0.007042221852473097,
"grad_norm": 1.5616939067840576,
"learning_rate": 4.993921770611866e-05,
"loss": 2.3804,
"step": 783
},
{
"epoch": 0.007051215750113548,
"grad_norm": 0.9750461578369141,
"learning_rate": 4.9938692806274814e-05,
"loss": 2.328,
"step": 784
},
{
"epoch": 0.007060209647753999,
"grad_norm": 1.0921258926391602,
"learning_rate": 4.9938165652498787e-05,
"loss": 2.2191,
"step": 785
},
{
"epoch": 0.00706920354539445,
"grad_norm": 1.053160548210144,
"learning_rate": 4.993763624483821e-05,
"loss": 2.3964,
"step": 786
},
{
"epoch": 0.007078197443034901,
"grad_norm": 2.779226303100586,
"learning_rate": 4.993710458334092e-05,
"loss": 2.4296,
"step": 787
},
{
"epoch": 0.007087191340675352,
"grad_norm": 1.360906720161438,
"learning_rate": 4.9936570668054984e-05,
"loss": 2.2808,
"step": 788
},
{
"epoch": 0.0070961852383158025,
"grad_norm": 1.4774829149246216,
"learning_rate": 4.993603449902865e-05,
"loss": 2.2954,
"step": 789
},
{
"epoch": 0.007105179135956254,
"grad_norm": 1.9357240200042725,
"learning_rate": 4.9935496076310376e-05,
"loss": 2.296,
"step": 790
},
{
"epoch": 0.007114173033596705,
"grad_norm": 2.325702667236328,
"learning_rate": 4.9934955399948824e-05,
"loss": 2.339,
"step": 791
},
{
"epoch": 0.007123166931237155,
"grad_norm": 1.0266544818878174,
"learning_rate": 4.993441246999287e-05,
"loss": 2.3156,
"step": 792
},
{
"epoch": 0.0071321608288776065,
"grad_norm": 1.1733542680740356,
"learning_rate": 4.993386728649156e-05,
"loss": 2.3455,
"step": 793
},
{
"epoch": 0.007141154726518058,
"grad_norm": 1.7009702920913696,
"learning_rate": 4.9933319849494185e-05,
"loss": 2.3087,
"step": 794
},
{
"epoch": 0.007150148624158508,
"grad_norm": 1.306169867515564,
"learning_rate": 4.9932770159050225e-05,
"loss": 2.3605,
"step": 795
},
{
"epoch": 0.007159142521798959,
"grad_norm": 2.564422845840454,
"learning_rate": 4.9932218215209346e-05,
"loss": 2.4546,
"step": 796
},
{
"epoch": 0.007168136419439411,
"grad_norm": 1.0364539623260498,
"learning_rate": 4.993166401802144e-05,
"loss": 2.4305,
"step": 797
},
{
"epoch": 0.007177130317079861,
"grad_norm": 1.7015596628189087,
"learning_rate": 4.99311075675366e-05,
"loss": 2.3859,
"step": 798
},
{
"epoch": 0.007186124214720312,
"grad_norm": 1.8677973747253418,
"learning_rate": 4.9930548863805105e-05,
"loss": 2.466,
"step": 799
},
{
"epoch": 0.007195118112360763,
"grad_norm": 1.9033832550048828,
"learning_rate": 4.992998790687746e-05,
"loss": 2.3308,
"step": 800
},
{
"epoch": 0.007204112010001214,
"grad_norm": 0.9066283702850342,
"learning_rate": 4.992942469680436e-05,
"loss": 2.282,
"step": 801
},
{
"epoch": 0.007213105907641665,
"grad_norm": 1.59651517868042,
"learning_rate": 4.99288592336367e-05,
"loss": 2.4003,
"step": 802
},
{
"epoch": 0.007222099805282116,
"grad_norm": 1.3170840740203857,
"learning_rate": 4.992829151742561e-05,
"loss": 2.3541,
"step": 803
},
{
"epoch": 0.0072310937029225674,
"grad_norm": 1.7328461408615112,
"learning_rate": 4.9927721548222374e-05,
"loss": 2.3629,
"step": 804
},
{
"epoch": 0.007240087600563018,
"grad_norm": 1.3761106729507446,
"learning_rate": 4.992714932607851e-05,
"loss": 2.3344,
"step": 805
},
{
"epoch": 0.007249081498203469,
"grad_norm": 1.4148802757263184,
"learning_rate": 4.992657485104575e-05,
"loss": 2.2926,
"step": 806
},
{
"epoch": 0.00725807539584392,
"grad_norm": 1.0844882726669312,
"learning_rate": 4.9925998123175996e-05,
"loss": 2.4877,
"step": 807
},
{
"epoch": 0.007267069293484371,
"grad_norm": 2.088745355606079,
"learning_rate": 4.992541914252138e-05,
"loss": 2.3863,
"step": 808
},
{
"epoch": 0.007276063191124822,
"grad_norm": 1.5522487163543701,
"learning_rate": 4.992483790913423e-05,
"loss": 2.315,
"step": 809
},
{
"epoch": 0.007285057088765273,
"grad_norm": 1.7547898292541504,
"learning_rate": 4.992425442306708e-05,
"loss": 2.3906,
"step": 810
},
{
"epoch": 0.0072940509864057234,
"grad_norm": 1.3132327795028687,
"learning_rate": 4.992366868437266e-05,
"loss": 2.3774,
"step": 811
},
{
"epoch": 0.007303044884046175,
"grad_norm": 1.0966075658798218,
"learning_rate": 4.992308069310391e-05,
"loss": 2.394,
"step": 812
},
{
"epoch": 0.007312038781686626,
"grad_norm": 1.5437568426132202,
"learning_rate": 4.992249044931396e-05,
"loss": 2.3382,
"step": 813
},
{
"epoch": 0.007321032679327076,
"grad_norm": 0.9832726120948792,
"learning_rate": 4.992189795305618e-05,
"loss": 2.3258,
"step": 814
},
{
"epoch": 0.0073300265769675275,
"grad_norm": 0.9568870067596436,
"learning_rate": 4.9921303204384104e-05,
"loss": 2.463,
"step": 815
},
{
"epoch": 0.007339020474607979,
"grad_norm": 0.9940560460090637,
"learning_rate": 4.99207062033515e-05,
"loss": 2.4384,
"step": 816
},
{
"epoch": 0.007348014372248429,
"grad_norm": 0.773494303226471,
"learning_rate": 4.992010695001229e-05,
"loss": 2.4385,
"step": 817
},
{
"epoch": 0.00735700826988888,
"grad_norm": 0.9313816428184509,
"learning_rate": 4.9919505444420666e-05,
"loss": 2.3089,
"step": 818
},
{
"epoch": 0.0073660021675293315,
"grad_norm": 1.4051645994186401,
"learning_rate": 4.991890168663098e-05,
"loss": 2.4094,
"step": 819
},
{
"epoch": 0.007374996065169782,
"grad_norm": 0.708573043346405,
"learning_rate": 4.99182956766978e-05,
"loss": 2.3811,
"step": 820
},
{
"epoch": 0.007383989962810233,
"grad_norm": 1.4287480115890503,
"learning_rate": 4.99176874146759e-05,
"loss": 2.3966,
"step": 821
},
{
"epoch": 0.007392983860450684,
"grad_norm": 0.9948827028274536,
"learning_rate": 4.991707690062025e-05,
"loss": 2.2247,
"step": 822
},
{
"epoch": 0.0074019777580911356,
"grad_norm": 0.7052820324897766,
"learning_rate": 4.9916464134586024e-05,
"loss": 2.4127,
"step": 823
},
{
"epoch": 0.007410971655731586,
"grad_norm": 0.8648499250411987,
"learning_rate": 4.991584911662861e-05,
"loss": 2.3283,
"step": 824
},
{
"epoch": 0.007419965553372037,
"grad_norm": 1.2007049322128296,
"learning_rate": 4.991523184680359e-05,
"loss": 2.3553,
"step": 825
},
{
"epoch": 0.007428959451012488,
"grad_norm": 1.5342097282409668,
"learning_rate": 4.991461232516675e-05,
"loss": 2.4664,
"step": 826
},
{
"epoch": 0.007437953348652939,
"grad_norm": 1.252152681350708,
"learning_rate": 4.991399055177408e-05,
"loss": 2.3259,
"step": 827
},
{
"epoch": 0.00744694724629339,
"grad_norm": 1.6909875869750977,
"learning_rate": 4.991336652668179e-05,
"loss": 2.5274,
"step": 828
},
{
"epoch": 0.007455941143933841,
"grad_norm": 1.041088342666626,
"learning_rate": 4.991274024994626e-05,
"loss": 2.5211,
"step": 829
},
{
"epoch": 0.0074649350415742916,
"grad_norm": 1.0081087350845337,
"learning_rate": 4.99121117216241e-05,
"loss": 2.2854,
"step": 830
},
{
"epoch": 0.007473928939214743,
"grad_norm": 0.7612619996070862,
"learning_rate": 4.9911480941772115e-05,
"loss": 2.4526,
"step": 831
},
{
"epoch": 0.007482922836855194,
"grad_norm": 1.263737440109253,
"learning_rate": 4.991084791044732e-05,
"loss": 2.3774,
"step": 832
},
{
"epoch": 0.007491916734495644,
"grad_norm": 1.2397956848144531,
"learning_rate": 4.991021262770692e-05,
"loss": 2.4083,
"step": 833
},
{
"epoch": 0.007500910632136096,
"grad_norm": 1.0493193864822388,
"learning_rate": 4.9909575093608334e-05,
"loss": 2.3255,
"step": 834
},
{
"epoch": 0.007509904529776547,
"grad_norm": 1.3190057277679443,
"learning_rate": 4.990893530820918e-05,
"loss": 2.3486,
"step": 835
},
{
"epoch": 0.007518898427416997,
"grad_norm": 1.0106217861175537,
"learning_rate": 4.9908293271567286e-05,
"loss": 2.2965,
"step": 836
},
{
"epoch": 0.007527892325057448,
"grad_norm": 2.0340452194213867,
"learning_rate": 4.990764898374067e-05,
"loss": 2.344,
"step": 837
},
{
"epoch": 0.0075368862226979,
"grad_norm": 0.9287651777267456,
"learning_rate": 4.990700244478758e-05,
"loss": 2.2849,
"step": 838
},
{
"epoch": 0.00754588012033835,
"grad_norm": 1.7290759086608887,
"learning_rate": 4.990635365476643e-05,
"loss": 2.3609,
"step": 839
},
{
"epoch": 0.007554874017978801,
"grad_norm": 1.1983683109283447,
"learning_rate": 4.990570261373585e-05,
"loss": 2.4809,
"step": 840
},
{
"epoch": 0.0075638679156192525,
"grad_norm": 1.5039162635803223,
"learning_rate": 4.9905049321754706e-05,
"loss": 2.3046,
"step": 841
},
{
"epoch": 0.007572861813259704,
"grad_norm": 1.4246561527252197,
"learning_rate": 4.990439377888203e-05,
"loss": 2.3694,
"step": 842
},
{
"epoch": 0.007581855710900154,
"grad_norm": 1.9078946113586426,
"learning_rate": 4.990373598517707e-05,
"loss": 2.3287,
"step": 843
},
{
"epoch": 0.007590849608540605,
"grad_norm": 2.58236026763916,
"learning_rate": 4.9903075940699265e-05,
"loss": 2.4055,
"step": 844
},
{
"epoch": 0.0075998435061810565,
"grad_norm": 1.0287704467773438,
"learning_rate": 4.990241364550829e-05,
"loss": 2.2513,
"step": 845
},
{
"epoch": 0.007608837403821507,
"grad_norm": 1.243863821029663,
"learning_rate": 4.990174909966399e-05,
"loss": 2.3509,
"step": 846
},
{
"epoch": 0.007617831301461958,
"grad_norm": 2.5628280639648438,
"learning_rate": 4.990108230322643e-05,
"loss": 2.278,
"step": 847
},
{
"epoch": 0.007626825199102409,
"grad_norm": 1.5202381610870361,
"learning_rate": 4.9900413256255876e-05,
"loss": 2.3038,
"step": 848
},
{
"epoch": 0.00763581909674286,
"grad_norm": 2.9225621223449707,
"learning_rate": 4.989974195881279e-05,
"loss": 2.427,
"step": 849
},
{
"epoch": 0.007644812994383311,
"grad_norm": 1.8009926080703735,
"learning_rate": 4.989906841095784e-05,
"loss": 2.4446,
"step": 850
},
{
"epoch": 0.007653806892023762,
"grad_norm": 1.1386442184448242,
"learning_rate": 4.989839261275191e-05,
"loss": 2.3672,
"step": 851
},
{
"epoch": 0.0076628007896642125,
"grad_norm": 2.892897605895996,
"learning_rate": 4.989771456425607e-05,
"loss": 2.3714,
"step": 852
},
{
"epoch": 0.007671794687304664,
"grad_norm": 2.243138074874878,
"learning_rate": 4.989703426553161e-05,
"loss": 2.2661,
"step": 853
},
{
"epoch": 0.007680788584945115,
"grad_norm": 4.217054843902588,
"learning_rate": 4.989635171664e-05,
"loss": 2.3475,
"step": 854
},
{
"epoch": 0.007689782482585565,
"grad_norm": 1.5125004053115845,
"learning_rate": 4.989566691764295e-05,
"loss": 2.316,
"step": 855
},
{
"epoch": 0.0076987763802260165,
"grad_norm": 1.1293413639068604,
"learning_rate": 4.989497986860234e-05,
"loss": 2.2831,
"step": 856
},
{
"epoch": 0.007707770277866468,
"grad_norm": 1.9106346368789673,
"learning_rate": 4.989429056958026e-05,
"loss": 2.426,
"step": 857
},
{
"epoch": 0.007716764175506919,
"grad_norm": 0.665691077709198,
"learning_rate": 4.989359902063901e-05,
"loss": 2.374,
"step": 858
},
{
"epoch": 0.007725758073147369,
"grad_norm": 0.8827734589576721,
"learning_rate": 4.98929052218411e-05,
"loss": 2.2281,
"step": 859
},
{
"epoch": 0.007734751970787821,
"grad_norm": 1.4260919094085693,
"learning_rate": 4.989220917324922e-05,
"loss": 2.3352,
"step": 860
},
{
"epoch": 0.007743745868428272,
"grad_norm": 2.3282699584960938,
"learning_rate": 4.9891510874926293e-05,
"loss": 2.3059,
"step": 861
},
{
"epoch": 0.007752739766068722,
"grad_norm": 2.224483013153076,
"learning_rate": 4.989081032693542e-05,
"loss": 2.3128,
"step": 862
},
{
"epoch": 0.007761733663709173,
"grad_norm": 0.88776034116745,
"learning_rate": 4.989010752933992e-05,
"loss": 2.3525,
"step": 863
},
{
"epoch": 0.007770727561349625,
"grad_norm": 2.1006646156311035,
"learning_rate": 4.988940248220332e-05,
"loss": 2.3066,
"step": 864
},
{
"epoch": 0.007779721458990075,
"grad_norm": 1.7027292251586914,
"learning_rate": 4.9888695185589325e-05,
"loss": 2.3422,
"step": 865
},
{
"epoch": 0.007788715356630526,
"grad_norm": 0.925018846988678,
"learning_rate": 4.988798563956186e-05,
"loss": 2.327,
"step": 866
},
{
"epoch": 0.007797709254270977,
"grad_norm": 1.78342604637146,
"learning_rate": 4.988727384418507e-05,
"loss": 2.3251,
"step": 867
},
{
"epoch": 0.007806703151911428,
"grad_norm": 1.8967210054397583,
"learning_rate": 4.9886559799523275e-05,
"loss": 2.4215,
"step": 868
},
{
"epoch": 0.00781569704955188,
"grad_norm": 1.4762319326400757,
"learning_rate": 4.9885843505641014e-05,
"loss": 2.4299,
"step": 869
},
{
"epoch": 0.00782469094719233,
"grad_norm": 1.4177857637405396,
"learning_rate": 4.988512496260301e-05,
"loss": 2.2442,
"step": 870
},
{
"epoch": 0.00783368484483278,
"grad_norm": 3.988291025161743,
"learning_rate": 4.988440417047423e-05,
"loss": 2.3613,
"step": 871
},
{
"epoch": 0.007842678742473233,
"grad_norm": 2.524430751800537,
"learning_rate": 4.988368112931979e-05,
"loss": 2.3907,
"step": 872
},
{
"epoch": 0.007851672640113683,
"grad_norm": 2.3439295291900635,
"learning_rate": 4.988295583920506e-05,
"loss": 2.4879,
"step": 873
},
{
"epoch": 0.007860666537754133,
"grad_norm": 1.3418093919754028,
"learning_rate": 4.988222830019559e-05,
"loss": 2.3817,
"step": 874
},
{
"epoch": 0.007869660435394586,
"grad_norm": 2.416830539703369,
"learning_rate": 4.988149851235712e-05,
"loss": 2.3644,
"step": 875
},
{
"epoch": 0.007878654333035036,
"grad_norm": 5.460973739624023,
"learning_rate": 4.988076647575562e-05,
"loss": 2.4638,
"step": 876
},
{
"epoch": 0.007887648230675486,
"grad_norm": 7.149078369140625,
"learning_rate": 4.9880032190457226e-05,
"loss": 2.3455,
"step": 877
},
{
"epoch": 0.007896642128315938,
"grad_norm": 7.11635160446167,
"learning_rate": 4.9879295656528334e-05,
"loss": 2.55,
"step": 878
},
{
"epoch": 0.007905636025956389,
"grad_norm": 6.192065238952637,
"learning_rate": 4.9878556874035504e-05,
"loss": 2.3529,
"step": 879
},
{
"epoch": 0.007914629923596839,
"grad_norm": 3.8169753551483154,
"learning_rate": 4.9877815843045495e-05,
"loss": 2.3865,
"step": 880
},
{
"epoch": 0.007923623821237291,
"grad_norm": 1.9059431552886963,
"learning_rate": 4.9877072563625285e-05,
"loss": 2.2723,
"step": 881
},
{
"epoch": 0.007932617718877742,
"grad_norm": 1.1022893190383911,
"learning_rate": 4.987632703584206e-05,
"loss": 2.452,
"step": 882
},
{
"epoch": 0.007941611616518192,
"grad_norm": 1.7090600728988647,
"learning_rate": 4.987557925976318e-05,
"loss": 2.3762,
"step": 883
},
{
"epoch": 0.007950605514158644,
"grad_norm": 1.3606576919555664,
"learning_rate": 4.987482923545625e-05,
"loss": 2.3087,
"step": 884
},
{
"epoch": 0.007959599411799094,
"grad_norm": 1.358298897743225,
"learning_rate": 4.987407696298904e-05,
"loss": 2.3496,
"step": 885
},
{
"epoch": 0.007968593309439545,
"grad_norm": 1.6232304573059082,
"learning_rate": 4.987332244242955e-05,
"loss": 2.3479,
"step": 886
},
{
"epoch": 0.007977587207079997,
"grad_norm": 1.5596123933792114,
"learning_rate": 4.9872565673845974e-05,
"loss": 2.4425,
"step": 887
},
{
"epoch": 0.007986581104720447,
"grad_norm": 1.5620065927505493,
"learning_rate": 4.98718066573067e-05,
"loss": 2.3169,
"step": 888
},
{
"epoch": 0.007995575002360897,
"grad_norm": 2.1944029331207275,
"learning_rate": 4.987104539288032e-05,
"loss": 2.42,
"step": 889
},
{
"epoch": 0.00800456890000135,
"grad_norm": 1.7498599290847778,
"learning_rate": 4.9870281880635656e-05,
"loss": 2.3805,
"step": 890
},
{
"epoch": 0.0080135627976418,
"grad_norm": 1.612972378730774,
"learning_rate": 4.98695161206417e-05,
"loss": 2.3094,
"step": 891
},
{
"epoch": 0.00802255669528225,
"grad_norm": 2.5815799236297607,
"learning_rate": 4.986874811296767e-05,
"loss": 2.4157,
"step": 892
},
{
"epoch": 0.008031550592922702,
"grad_norm": 17.66390037536621,
"learning_rate": 4.9867977857682965e-05,
"loss": 2.4277,
"step": 893
},
{
"epoch": 0.008040544490563153,
"grad_norm": 46.72271728515625,
"learning_rate": 4.9867205354857206e-05,
"loss": 2.3533,
"step": 894
},
{
"epoch": 0.008049538388203603,
"grad_norm": 2.8205180168151855,
"learning_rate": 4.9866430604560214e-05,
"loss": 2.372,
"step": 895
},
{
"epoch": 0.008058532285844055,
"grad_norm": 4.408201217651367,
"learning_rate": 4.986565360686201e-05,
"loss": 2.4591,
"step": 896
},
{
"epoch": 0.008067526183484506,
"grad_norm": 1.0639312267303467,
"learning_rate": 4.986487436183282e-05,
"loss": 2.3197,
"step": 897
},
{
"epoch": 0.008076520081124956,
"grad_norm": 1.9649598598480225,
"learning_rate": 4.986409286954306e-05,
"loss": 2.2815,
"step": 898
},
{
"epoch": 0.008085513978765408,
"grad_norm": 2.717100143432617,
"learning_rate": 4.986330913006337e-05,
"loss": 2.4578,
"step": 899
},
{
"epoch": 0.008094507876405858,
"grad_norm": 2.211270332336426,
"learning_rate": 4.986252314346458e-05,
"loss": 2.2714,
"step": 900
},
{
"epoch": 0.008103501774046309,
"grad_norm": 2.953294515609741,
"learning_rate": 4.986173490981773e-05,
"loss": 2.3534,
"step": 901
},
{
"epoch": 0.00811249567168676,
"grad_norm": 20.0116024017334,
"learning_rate": 4.986094442919406e-05,
"loss": 2.246,
"step": 902
},
{
"epoch": 0.008121489569327211,
"grad_norm": 56.963191986083984,
"learning_rate": 4.9860151701665004e-05,
"loss": 2.3506,
"step": 903
},
{
"epoch": 0.008130483466967663,
"grad_norm": 2.4784321784973145,
"learning_rate": 4.985935672730222e-05,
"loss": 2.3257,
"step": 904
},
{
"epoch": 0.008139477364608114,
"grad_norm": 1.7748329639434814,
"learning_rate": 4.985855950617755e-05,
"loss": 2.3586,
"step": 905
},
{
"epoch": 0.008148471262248564,
"grad_norm": 2.1759116649627686,
"learning_rate": 4.985776003836304e-05,
"loss": 2.3651,
"step": 906
},
{
"epoch": 0.008157465159889016,
"grad_norm": 1.564903974533081,
"learning_rate": 4.985695832393097e-05,
"loss": 2.408,
"step": 907
},
{
"epoch": 0.008166459057529466,
"grad_norm": 0.7693034410476685,
"learning_rate": 4.985615436295377e-05,
"loss": 2.2915,
"step": 908
},
{
"epoch": 0.008175452955169917,
"grad_norm": 3.041149854660034,
"learning_rate": 4.9855348155504104e-05,
"loss": 2.3744,
"step": 909
},
{
"epoch": 0.008184446852810369,
"grad_norm": 2.2738850116729736,
"learning_rate": 4.9854539701654855e-05,
"loss": 2.3491,
"step": 910
},
{
"epoch": 0.00819344075045082,
"grad_norm": 2.5227854251861572,
"learning_rate": 4.9853729001479075e-05,
"loss": 2.373,
"step": 911
},
{
"epoch": 0.00820243464809127,
"grad_norm": 1.8525545597076416,
"learning_rate": 4.9852916055050035e-05,
"loss": 2.3316,
"step": 912
},
{
"epoch": 0.008211428545731722,
"grad_norm": 1.6783733367919922,
"learning_rate": 4.985210086244121e-05,
"loss": 2.3053,
"step": 913
},
{
"epoch": 0.008220422443372172,
"grad_norm": 3.593388319015503,
"learning_rate": 4.985128342372629e-05,
"loss": 2.4132,
"step": 914
},
{
"epoch": 0.008229416341012622,
"grad_norm": 1.6809263229370117,
"learning_rate": 4.985046373897913e-05,
"loss": 2.388,
"step": 915
},
{
"epoch": 0.008238410238653075,
"grad_norm": 1.7394527196884155,
"learning_rate": 4.984964180827383e-05,
"loss": 2.2643,
"step": 916
},
{
"epoch": 0.008247404136293525,
"grad_norm": 1.6720887422561646,
"learning_rate": 4.984881763168466e-05,
"loss": 2.3783,
"step": 917
},
{
"epoch": 0.008256398033933975,
"grad_norm": 114.69852447509766,
"learning_rate": 4.9847991209286124e-05,
"loss": 2.3095,
"step": 918
},
{
"epoch": 0.008265391931574427,
"grad_norm": 2.773820400238037,
"learning_rate": 4.9847162541152906e-05,
"loss": 2.3849,
"step": 919
},
{
"epoch": 0.008274385829214878,
"grad_norm": 2.6709611415863037,
"learning_rate": 4.98463316273599e-05,
"loss": 2.2671,
"step": 920
},
{
"epoch": 0.008283379726855328,
"grad_norm": 2.1583499908447266,
"learning_rate": 4.984549846798221e-05,
"loss": 2.3633,
"step": 921
},
{
"epoch": 0.00829237362449578,
"grad_norm": 2.721205472946167,
"learning_rate": 4.984466306309512e-05,
"loss": 2.3699,
"step": 922
},
{
"epoch": 0.00830136752213623,
"grad_norm": 3.0230939388275146,
"learning_rate": 4.9843825412774145e-05,
"loss": 2.3702,
"step": 923
},
{
"epoch": 0.008310361419776681,
"grad_norm": 2.07593035697937,
"learning_rate": 4.9842985517094996e-05,
"loss": 2.2955,
"step": 924
},
{
"epoch": 0.008319355317417133,
"grad_norm": 8.249615669250488,
"learning_rate": 4.984214337613357e-05,
"loss": 2.3203,
"step": 925
},
{
"epoch": 0.008328349215057583,
"grad_norm": 3.146702766418457,
"learning_rate": 4.9841298989965984e-05,
"loss": 2.4488,
"step": 926
},
{
"epoch": 0.008337343112698034,
"grad_norm": 3.03059458732605,
"learning_rate": 4.984045235866855e-05,
"loss": 2.3512,
"step": 927
},
{
"epoch": 0.008346337010338486,
"grad_norm": 2.2177159786224365,
"learning_rate": 4.983960348231779e-05,
"loss": 2.3528,
"step": 928
},
{
"epoch": 0.008355330907978936,
"grad_norm": 6.031072616577148,
"learning_rate": 4.983875236099042e-05,
"loss": 2.4123,
"step": 929
},
{
"epoch": 0.008364324805619387,
"grad_norm": 4.340119361877441,
"learning_rate": 4.983789899476337e-05,
"loss": 2.4107,
"step": 930
},
{
"epoch": 0.008373318703259839,
"grad_norm": 14.878398895263672,
"learning_rate": 4.9837043383713753e-05,
"loss": 2.4057,
"step": 931
},
{
"epoch": 0.008382312600900289,
"grad_norm": 8.246759414672852,
"learning_rate": 4.9836185527918924e-05,
"loss": 2.3712,
"step": 932
},
{
"epoch": 0.00839130649854074,
"grad_norm": 7.280088901519775,
"learning_rate": 4.983532542745638e-05,
"loss": 2.4236,
"step": 933
},
{
"epoch": 0.008400300396181191,
"grad_norm": 4.351801872253418,
"learning_rate": 4.983446308240388e-05,
"loss": 2.3233,
"step": 934
},
{
"epoch": 0.008409294293821642,
"grad_norm": 2.8788869380950928,
"learning_rate": 4.983359849283936e-05,
"loss": 2.3607,
"step": 935
},
{
"epoch": 0.008418288191462092,
"grad_norm": 49.51881408691406,
"learning_rate": 4.9832731658840956e-05,
"loss": 2.2114,
"step": 936
},
{
"epoch": 0.008427282089102544,
"grad_norm": 16.22137451171875,
"learning_rate": 4.983186258048701e-05,
"loss": 2.3123,
"step": 937
},
{
"epoch": 0.008436275986742995,
"grad_norm": 1.5916885137557983,
"learning_rate": 4.9830991257856074e-05,
"loss": 2.332,
"step": 938
},
{
"epoch": 0.008445269884383445,
"grad_norm": 2.1934919357299805,
"learning_rate": 4.983011769102689e-05,
"loss": 2.3972,
"step": 939
},
{
"epoch": 0.008454263782023897,
"grad_norm": 1.4318190813064575,
"learning_rate": 4.982924188007841e-05,
"loss": 2.3014,
"step": 940
},
{
"epoch": 0.008463257679664347,
"grad_norm": 3.1829514503479004,
"learning_rate": 4.982836382508981e-05,
"loss": 2.538,
"step": 941
},
{
"epoch": 0.0084722515773048,
"grad_norm": 1.1255837678909302,
"learning_rate": 4.9827483526140414e-05,
"loss": 2.3257,
"step": 942
},
{
"epoch": 0.00848124547494525,
"grad_norm": 2.0071754455566406,
"learning_rate": 4.98266009833098e-05,
"loss": 2.4001,
"step": 943
},
{
"epoch": 0.0084902393725857,
"grad_norm": 1.5771818161010742,
"learning_rate": 4.982571619667774e-05,
"loss": 2.3978,
"step": 944
},
{
"epoch": 0.008499233270226152,
"grad_norm": 2.3359062671661377,
"learning_rate": 4.9824829166324186e-05,
"loss": 2.3576,
"step": 945
},
{
"epoch": 0.008508227167866603,
"grad_norm": 85.4216537475586,
"learning_rate": 4.982393989232931e-05,
"loss": 2.3332,
"step": 946
},
{
"epoch": 0.008517221065507053,
"grad_norm": 3.0369982719421387,
"learning_rate": 4.982304837477348e-05,
"loss": 2.3304,
"step": 947
},
{
"epoch": 0.008526214963147505,
"grad_norm": 2.0758142471313477,
"learning_rate": 4.982215461373728e-05,
"loss": 2.3491,
"step": 948
},
{
"epoch": 0.008535208860787956,
"grad_norm": 2.0540716648101807,
"learning_rate": 4.982125860930148e-05,
"loss": 2.4209,
"step": 949
},
{
"epoch": 0.008544202758428406,
"grad_norm": 2.7230679988861084,
"learning_rate": 4.982036036154706e-05,
"loss": 2.4095,
"step": 950
},
{
"epoch": 0.008553196656068858,
"grad_norm": 2.2635655403137207,
"learning_rate": 4.981945987055521e-05,
"loss": 2.3302,
"step": 951
},
{
"epoch": 0.008562190553709308,
"grad_norm": 2.8918027877807617,
"learning_rate": 4.9818557136407304e-05,
"loss": 2.2959,
"step": 952
},
{
"epoch": 0.008571184451349759,
"grad_norm": 0.9320359230041504,
"learning_rate": 4.9817652159184946e-05,
"loss": 2.297,
"step": 953
},
{
"epoch": 0.00858017834899021,
"grad_norm": 1.637445092201233,
"learning_rate": 4.9816744938969904e-05,
"loss": 2.3192,
"step": 954
},
{
"epoch": 0.008589172246630661,
"grad_norm": 0.8960732221603394,
"learning_rate": 4.98158354758442e-05,
"loss": 2.351,
"step": 955
},
{
"epoch": 0.008598166144271112,
"grad_norm": 0.9512754082679749,
"learning_rate": 4.9814923769890004e-05,
"loss": 2.3218,
"step": 956
},
{
"epoch": 0.008607160041911564,
"grad_norm": 1.2916690111160278,
"learning_rate": 4.9814009821189734e-05,
"loss": 2.4091,
"step": 957
},
{
"epoch": 0.008616153939552014,
"grad_norm": 0.800326406955719,
"learning_rate": 4.981309362982598e-05,
"loss": 2.3718,
"step": 958
},
{
"epoch": 0.008625147837192464,
"grad_norm": 1.2276074886322021,
"learning_rate": 4.981217519588155e-05,
"loss": 2.4268,
"step": 959
},
{
"epoch": 0.008634141734832916,
"grad_norm": 1.0263639688491821,
"learning_rate": 4.981125451943944e-05,
"loss": 2.3521,
"step": 960
},
{
"epoch": 0.008643135632473367,
"grad_norm": 2.415966749191284,
"learning_rate": 4.981033160058289e-05,
"loss": 2.383,
"step": 961
},
{
"epoch": 0.008652129530113817,
"grad_norm": 1.3266791105270386,
"learning_rate": 4.9809406439395286e-05,
"loss": 2.3436,
"step": 962
},
{
"epoch": 0.00866112342775427,
"grad_norm": 2.955108404159546,
"learning_rate": 4.980847903596025e-05,
"loss": 2.3794,
"step": 963
},
{
"epoch": 0.00867011732539472,
"grad_norm": 1.9091037511825562,
"learning_rate": 4.98075493903616e-05,
"loss": 2.375,
"step": 964
},
{
"epoch": 0.00867911122303517,
"grad_norm": 6.714507579803467,
"learning_rate": 4.9806617502683356e-05,
"loss": 2.3218,
"step": 965
},
{
"epoch": 0.008688105120675622,
"grad_norm": 1.7211779356002808,
"learning_rate": 4.9805683373009746e-05,
"loss": 2.3044,
"step": 966
},
{
"epoch": 0.008697099018316072,
"grad_norm": 1.9710533618927002,
"learning_rate": 4.980474700142519e-05,
"loss": 2.3285,
"step": 967
},
{
"epoch": 0.008706092915956523,
"grad_norm": 1.4440377950668335,
"learning_rate": 4.980380838801431e-05,
"loss": 2.326,
"step": 968
},
{
"epoch": 0.008715086813596975,
"grad_norm": 2.5308713912963867,
"learning_rate": 4.980286753286195e-05,
"loss": 2.3517,
"step": 969
},
{
"epoch": 0.008724080711237425,
"grad_norm": 2.4353814125061035,
"learning_rate": 4.980192443605314e-05,
"loss": 2.3247,
"step": 970
},
{
"epoch": 0.008733074608877876,
"grad_norm": 1.2030874490737915,
"learning_rate": 4.9800979097673106e-05,
"loss": 2.2785,
"step": 971
},
{
"epoch": 0.008742068506518328,
"grad_norm": 4.3923540115356445,
"learning_rate": 4.98000315178073e-05,
"loss": 2.4156,
"step": 972
},
{
"epoch": 0.008751062404158778,
"grad_norm": 2.5101020336151123,
"learning_rate": 4.979908169654136e-05,
"loss": 2.3458,
"step": 973
},
{
"epoch": 0.008760056301799228,
"grad_norm": 2.055736780166626,
"learning_rate": 4.9798129633961124e-05,
"loss": 2.366,
"step": 974
},
{
"epoch": 0.00876905019943968,
"grad_norm": 1.0032880306243896,
"learning_rate": 4.979717533015264e-05,
"loss": 2.2664,
"step": 975
},
{
"epoch": 0.00877804409708013,
"grad_norm": 2.2960662841796875,
"learning_rate": 4.979621878520216e-05,
"loss": 2.4005,
"step": 976
},
{
"epoch": 0.008787037994720581,
"grad_norm": 1.7479488849639893,
"learning_rate": 4.979525999919614e-05,
"loss": 2.2448,
"step": 977
},
{
"epoch": 0.008796031892361033,
"grad_norm": 1.5676523447036743,
"learning_rate": 4.9794298972221225e-05,
"loss": 2.3536,
"step": 978
},
{
"epoch": 0.008805025790001484,
"grad_norm": 78.5261459350586,
"learning_rate": 4.979333570436427e-05,
"loss": 2.2746,
"step": 979
},
{
"epoch": 0.008814019687641936,
"grad_norm": 9.62118911743164,
"learning_rate": 4.979237019571235e-05,
"loss": 2.2938,
"step": 980
},
{
"epoch": 0.008823013585282386,
"grad_norm": 2.8171589374542236,
"learning_rate": 4.979140244635271e-05,
"loss": 2.3666,
"step": 981
},
{
"epoch": 0.008832007482922836,
"grad_norm": 1.6373629570007324,
"learning_rate": 4.979043245637282e-05,
"loss": 2.244,
"step": 982
},
{
"epoch": 0.008841001380563289,
"grad_norm": 2.18770694732666,
"learning_rate": 4.978946022586035e-05,
"loss": 2.376,
"step": 983
},
{
"epoch": 0.008849995278203739,
"grad_norm": 2.425063371658325,
"learning_rate": 4.978848575490316e-05,
"loss": 2.272,
"step": 984
},
{
"epoch": 0.00885898917584419,
"grad_norm": 1.0710437297821045,
"learning_rate": 4.9787509043589334e-05,
"loss": 2.3817,
"step": 985
},
{
"epoch": 0.008867983073484641,
"grad_norm": 1.5350664854049683,
"learning_rate": 4.9786530092007135e-05,
"loss": 2.3173,
"step": 986
},
{
"epoch": 0.008876976971125092,
"grad_norm": 2.1459872722625732,
"learning_rate": 4.9785548900245045e-05,
"loss": 2.3475,
"step": 987
},
{
"epoch": 0.008885970868765542,
"grad_norm": 2.152179718017578,
"learning_rate": 4.978456546839174e-05,
"loss": 2.3343,
"step": 988
},
{
"epoch": 0.008894964766405994,
"grad_norm": 1.4224385023117065,
"learning_rate": 4.978357979653611e-05,
"loss": 2.3542,
"step": 989
},
{
"epoch": 0.008903958664046445,
"grad_norm": 1.3504154682159424,
"learning_rate": 4.9782591884767224e-05,
"loss": 2.3621,
"step": 990
},
{
"epoch": 0.008912952561686895,
"grad_norm": 7.234045505523682,
"learning_rate": 4.978160173317438e-05,
"loss": 2.377,
"step": 991
},
{
"epoch": 0.008921946459327347,
"grad_norm": 2.711791753768921,
"learning_rate": 4.978060934184707e-05,
"loss": 2.406,
"step": 992
},
{
"epoch": 0.008930940356967797,
"grad_norm": 1.7958492040634155,
"learning_rate": 4.9779614710874974e-05,
"loss": 2.3133,
"step": 993
},
{
"epoch": 0.008939934254608248,
"grad_norm": 4.097232818603516,
"learning_rate": 4.977861784034799e-05,
"loss": 2.4167,
"step": 994
},
{
"epoch": 0.0089489281522487,
"grad_norm": 1.4877700805664062,
"learning_rate": 4.977761873035622e-05,
"loss": 2.4374,
"step": 995
},
{
"epoch": 0.00895792204988915,
"grad_norm": 19.301212310791016,
"learning_rate": 4.977661738098995e-05,
"loss": 2.4801,
"step": 996
},
{
"epoch": 0.0089669159475296,
"grad_norm": 1.8256185054779053,
"learning_rate": 4.977561379233969e-05,
"loss": 2.4585,
"step": 997
},
{
"epoch": 0.008975909845170053,
"grad_norm": 3.464038848876953,
"learning_rate": 4.9774607964496144e-05,
"loss": 2.3755,
"step": 998
},
{
"epoch": 0.008984903742810503,
"grad_norm": 2.2096097469329834,
"learning_rate": 4.9773599897550214e-05,
"loss": 2.4156,
"step": 999
},
{
"epoch": 0.008993897640450953,
"grad_norm": 4.295968532562256,
"learning_rate": 4.977258959159301e-05,
"loss": 2.3988,
"step": 1000
},
{
"epoch": 0.009002891538091405,
"grad_norm": 1.1036388874053955,
"learning_rate": 4.9771577046715846e-05,
"loss": 2.3178,
"step": 1001
},
{
"epoch": 0.009011885435731856,
"grad_norm": 1.7228236198425293,
"learning_rate": 4.9770562263010225e-05,
"loss": 2.2841,
"step": 1002
},
{
"epoch": 0.009020879333372306,
"grad_norm": 2.547356128692627,
"learning_rate": 4.976954524056787e-05,
"loss": 2.4064,
"step": 1003
},
{
"epoch": 0.009029873231012758,
"grad_norm": 2.5266928672790527,
"learning_rate": 4.976852597948069e-05,
"loss": 2.3287,
"step": 1004
},
{
"epoch": 0.009038867128653209,
"grad_norm": 2.128464460372925,
"learning_rate": 4.9767504479840816e-05,
"loss": 2.3066,
"step": 1005
},
{
"epoch": 0.009047861026293659,
"grad_norm": 1.5903282165527344,
"learning_rate": 4.976648074174056e-05,
"loss": 2.3114,
"step": 1006
},
{
"epoch": 0.009056854923934111,
"grad_norm": 284.1065368652344,
"learning_rate": 4.976545476527246e-05,
"loss": 2.4002,
"step": 1007
},
{
"epoch": 0.009065848821574561,
"grad_norm": 446.20068359375,
"learning_rate": 4.976442655052922e-05,
"loss": 2.5204,
"step": 1008
},
{
"epoch": 0.009074842719215012,
"grad_norm": 4.401348114013672,
"learning_rate": 4.976339609760379e-05,
"loss": 2.3228,
"step": 1009
},
{
"epoch": 0.009083836616855464,
"grad_norm": 1.9144341945648193,
"learning_rate": 4.9762363406589295e-05,
"loss": 2.3604,
"step": 1010
},
{
"epoch": 0.009092830514495914,
"grad_norm": 18.24340057373047,
"learning_rate": 4.976132847757906e-05,
"loss": 2.3804,
"step": 1011
},
{
"epoch": 0.009101824412136365,
"grad_norm": 1.3318119049072266,
"learning_rate": 4.976029131066664e-05,
"loss": 2.2884,
"step": 1012
},
{
"epoch": 0.009110818309776817,
"grad_norm": 1.1708160638809204,
"learning_rate": 4.975925190594575e-05,
"loss": 2.3304,
"step": 1013
},
{
"epoch": 0.009119812207417267,
"grad_norm": 1.9248275756835938,
"learning_rate": 4.9758210263510345e-05,
"loss": 2.4356,
"step": 1014
},
{
"epoch": 0.00912880610505772,
"grad_norm": 1.1926578283309937,
"learning_rate": 4.9757166383454565e-05,
"loss": 2.3249,
"step": 1015
},
{
"epoch": 0.00913780000269817,
"grad_norm": 1.5307071208953857,
"learning_rate": 4.975612026587275e-05,
"loss": 2.3416,
"step": 1016
},
{
"epoch": 0.00914679390033862,
"grad_norm": 1.9043503999710083,
"learning_rate": 4.9755071910859456e-05,
"loss": 2.4471,
"step": 1017
},
{
"epoch": 0.009155787797979072,
"grad_norm": 1.5771394968032837,
"learning_rate": 4.975402131850942e-05,
"loss": 2.3845,
"step": 1018
},
{
"epoch": 0.009164781695619522,
"grad_norm": 1.0784991979599,
"learning_rate": 4.9752968488917604e-05,
"loss": 2.2247,
"step": 1019
},
{
"epoch": 0.009173775593259973,
"grad_norm": 11.075843811035156,
"learning_rate": 4.975191342217917e-05,
"loss": 2.2718,
"step": 1020
},
{
"epoch": 0.009182769490900425,
"grad_norm": 11.83606243133545,
"learning_rate": 4.975085611838944e-05,
"loss": 2.3569,
"step": 1021
},
{
"epoch": 0.009191763388540875,
"grad_norm": 6.07887601852417,
"learning_rate": 4.9749796577644e-05,
"loss": 2.4011,
"step": 1022
},
{
"epoch": 0.009200757286181326,
"grad_norm": 1.940321445465088,
"learning_rate": 4.974873480003861e-05,
"loss": 2.2614,
"step": 1023
},
{
"epoch": 0.009209751183821778,
"grad_norm": 3.0341317653656006,
"learning_rate": 4.974767078566922e-05,
"loss": 2.4608,
"step": 1024
},
{
"epoch": 0.009218745081462228,
"grad_norm": 2.68550181388855,
"learning_rate": 4.9746604534632004e-05,
"loss": 2.2984,
"step": 1025
},
{
"epoch": 0.009227738979102678,
"grad_norm": 4.981675624847412,
"learning_rate": 4.9745536047023324e-05,
"loss": 2.3595,
"step": 1026
},
{
"epoch": 0.00923673287674313,
"grad_norm": 3.624321222305298,
"learning_rate": 4.974446532293975e-05,
"loss": 2.3652,
"step": 1027
},
{
"epoch": 0.00924572677438358,
"grad_norm": 2.262476682662964,
"learning_rate": 4.9743392362478046e-05,
"loss": 2.3697,
"step": 1028
},
{
"epoch": 0.009254720672024031,
"grad_norm": 2.0064961910247803,
"learning_rate": 4.9742317165735196e-05,
"loss": 2.3023,
"step": 1029
},
{
"epoch": 0.009263714569664483,
"grad_norm": 1.8211779594421387,
"learning_rate": 4.9741239732808377e-05,
"loss": 2.3624,
"step": 1030
},
{
"epoch": 0.009272708467304934,
"grad_norm": 1.7825077772140503,
"learning_rate": 4.974016006379495e-05,
"loss": 2.3577,
"step": 1031
},
{
"epoch": 0.009281702364945384,
"grad_norm": 2.8979403972625732,
"learning_rate": 4.9739078158792514e-05,
"loss": 2.4009,
"step": 1032
},
{
"epoch": 0.009290696262585836,
"grad_norm": 3.5438995361328125,
"learning_rate": 4.973799401789884e-05,
"loss": 2.393,
"step": 1033
},
{
"epoch": 0.009299690160226286,
"grad_norm": 2.3994574546813965,
"learning_rate": 4.9736907641211904e-05,
"loss": 2.3648,
"step": 1034
},
{
"epoch": 0.009308684057866737,
"grad_norm": 1.2672652006149292,
"learning_rate": 4.97358190288299e-05,
"loss": 2.2841,
"step": 1035
},
{
"epoch": 0.009317677955507189,
"grad_norm": 1.8701341152191162,
"learning_rate": 4.973472818085122e-05,
"loss": 2.341,
"step": 1036
},
{
"epoch": 0.00932667185314764,
"grad_norm": 2.4897868633270264,
"learning_rate": 4.973363509737445e-05,
"loss": 2.3821,
"step": 1037
},
{
"epoch": 0.00933566575078809,
"grad_norm": 2.0629661083221436,
"learning_rate": 4.973253977849838e-05,
"loss": 2.3637,
"step": 1038
},
{
"epoch": 0.009344659648428542,
"grad_norm": 1.559417963027954,
"learning_rate": 4.9731442224322e-05,
"loss": 2.2815,
"step": 1039
},
{
"epoch": 0.009353653546068992,
"grad_norm": 1.5687872171401978,
"learning_rate": 4.973034243494452e-05,
"loss": 2.3517,
"step": 1040
},
{
"epoch": 0.009362647443709442,
"grad_norm": 1.0377593040466309,
"learning_rate": 4.9729240410465316e-05,
"loss": 2.2756,
"step": 1041
},
{
"epoch": 0.009371641341349895,
"grad_norm": 1.5999455451965332,
"learning_rate": 4.9728136150984e-05,
"loss": 2.3488,
"step": 1042
},
{
"epoch": 0.009380635238990345,
"grad_norm": 1.3381636142730713,
"learning_rate": 4.972702965660039e-05,
"loss": 2.3803,
"step": 1043
},
{
"epoch": 0.009389629136630795,
"grad_norm": 1.8739471435546875,
"learning_rate": 4.972592092741446e-05,
"loss": 2.3593,
"step": 1044
},
{
"epoch": 0.009398623034271247,
"grad_norm": 1.6740717887878418,
"learning_rate": 4.972480996352644e-05,
"loss": 2.3002,
"step": 1045
},
{
"epoch": 0.009407616931911698,
"grad_norm": 1.1257169246673584,
"learning_rate": 4.972369676503672e-05,
"loss": 2.349,
"step": 1046
},
{
"epoch": 0.009416610829552148,
"grad_norm": 37.121726989746094,
"learning_rate": 4.9722581332045906e-05,
"loss": 2.3289,
"step": 1047
},
{
"epoch": 0.0094256047271926,
"grad_norm": 1.7633475065231323,
"learning_rate": 4.9721463664654835e-05,
"loss": 2.375,
"step": 1048
},
{
"epoch": 0.00943459862483305,
"grad_norm": 2.242678642272949,
"learning_rate": 4.97203437629645e-05,
"loss": 2.2794,
"step": 1049
},
{
"epoch": 0.009443592522473501,
"grad_norm": 2.0801889896392822,
"learning_rate": 4.971922162707612e-05,
"loss": 2.2789,
"step": 1050
},
{
"epoch": 0.009452586420113953,
"grad_norm": 1.684820294380188,
"learning_rate": 4.971809725709112e-05,
"loss": 2.4247,
"step": 1051
},
{
"epoch": 0.009461580317754403,
"grad_norm": 2.8138086795806885,
"learning_rate": 4.971697065311111e-05,
"loss": 2.3125,
"step": 1052
},
{
"epoch": 0.009470574215394855,
"grad_norm": 1.4332460165023804,
"learning_rate": 4.971584181523792e-05,
"loss": 2.3261,
"step": 1053
},
{
"epoch": 0.009479568113035306,
"grad_norm": 1.1635377407073975,
"learning_rate": 4.9714710743573566e-05,
"loss": 2.3102,
"step": 1054
},
{
"epoch": 0.009488562010675756,
"grad_norm": 12.210381507873535,
"learning_rate": 4.971357743822027e-05,
"loss": 2.3955,
"step": 1055
},
{
"epoch": 0.009497555908316208,
"grad_norm": 2.8512823581695557,
"learning_rate": 4.971244189928047e-05,
"loss": 2.2482,
"step": 1056
},
{
"epoch": 0.009506549805956659,
"grad_norm": 2.8980586528778076,
"learning_rate": 4.971130412685679e-05,
"loss": 2.2319,
"step": 1057
},
{
"epoch": 0.009515543703597109,
"grad_norm": 2.3892319202423096,
"learning_rate": 4.971016412105206e-05,
"loss": 2.4365,
"step": 1058
},
{
"epoch": 0.009524537601237561,
"grad_norm": 1.2553093433380127,
"learning_rate": 4.970902188196931e-05,
"loss": 2.2569,
"step": 1059
},
{
"epoch": 0.009533531498878011,
"grad_norm": 1.172286868095398,
"learning_rate": 4.970787740971178e-05,
"loss": 2.2426,
"step": 1060
},
{
"epoch": 0.009542525396518462,
"grad_norm": 1.9281389713287354,
"learning_rate": 4.97067307043829e-05,
"loss": 2.3148,
"step": 1061
},
{
"epoch": 0.009551519294158914,
"grad_norm": 5.392895221710205,
"learning_rate": 4.970558176608631e-05,
"loss": 2.3564,
"step": 1062
},
{
"epoch": 0.009560513191799364,
"grad_norm": 111.55917358398438,
"learning_rate": 4.970443059492585e-05,
"loss": 2.6691,
"step": 1063
},
{
"epoch": 0.009569507089439815,
"grad_norm": 79.27272033691406,
"learning_rate": 4.9703277191005556e-05,
"loss": 2.7401,
"step": 1064
},
{
"epoch": 0.009578500987080267,
"grad_norm": 64.21418762207031,
"learning_rate": 4.9702121554429695e-05,
"loss": 2.7139,
"step": 1065
},
{
"epoch": 0.009587494884720717,
"grad_norm": 238.8323211669922,
"learning_rate": 4.970096368530268e-05,
"loss": 2.5298,
"step": 1066
},
{
"epoch": 0.009596488782361167,
"grad_norm": 5.2633185386657715,
"learning_rate": 4.969980358372918e-05,
"loss": 2.4147,
"step": 1067
},
{
"epoch": 0.00960548268000162,
"grad_norm": 3.5609946250915527,
"learning_rate": 4.9698641249814036e-05,
"loss": 2.2809,
"step": 1068
},
{
"epoch": 0.00961447657764207,
"grad_norm": 186.07591247558594,
"learning_rate": 4.96974766836623e-05,
"loss": 2.438,
"step": 1069
},
{
"epoch": 0.00962347047528252,
"grad_norm": 37.068077087402344,
"learning_rate": 4.9696309885379225e-05,
"loss": 2.3861,
"step": 1070
},
{
"epoch": 0.009632464372922972,
"grad_norm": 4.245080947875977,
"learning_rate": 4.9695140855070254e-05,
"loss": 2.3331,
"step": 1071
},
{
"epoch": 0.009641458270563423,
"grad_norm": 1.162161111831665,
"learning_rate": 4.9693969592841054e-05,
"loss": 2.2964,
"step": 1072
},
{
"epoch": 0.009650452168203873,
"grad_norm": 1.679726481437683,
"learning_rate": 4.969279609879749e-05,
"loss": 2.3182,
"step": 1073
},
{
"epoch": 0.009659446065844325,
"grad_norm": 322.6358642578125,
"learning_rate": 4.96916203730456e-05,
"loss": 2.2659,
"step": 1074
},
{
"epoch": 0.009668439963484775,
"grad_norm": 1.6927112340927124,
"learning_rate": 4.969044241569166e-05,
"loss": 2.3038,
"step": 1075
},
{
"epoch": 0.009677433861125226,
"grad_norm": 106.5965805053711,
"learning_rate": 4.968926222684213e-05,
"loss": 2.3294,
"step": 1076
},
{
"epoch": 0.009686427758765678,
"grad_norm": 5.916990756988525,
"learning_rate": 4.968807980660367e-05,
"loss": 2.3532,
"step": 1077
},
{
"epoch": 0.009695421656406128,
"grad_norm": 1.4046456813812256,
"learning_rate": 4.968689515508316e-05,
"loss": 2.3392,
"step": 1078
},
{
"epoch": 0.009704415554046579,
"grad_norm": 2.997627019882202,
"learning_rate": 4.968570827238764e-05,
"loss": 2.3581,
"step": 1079
},
{
"epoch": 0.00971340945168703,
"grad_norm": 2.466237783432007,
"learning_rate": 4.968451915862441e-05,
"loss": 2.3702,
"step": 1080
},
{
"epoch": 0.009722403349327481,
"grad_norm": 6.010172367095947,
"learning_rate": 4.968332781390092e-05,
"loss": 2.339,
"step": 1081
},
{
"epoch": 0.009731397246967931,
"grad_norm": 1.6601020097732544,
"learning_rate": 4.968213423832485e-05,
"loss": 2.306,
"step": 1082
},
{
"epoch": 0.009740391144608384,
"grad_norm": 1.492484450340271,
"learning_rate": 4.9680938432004074e-05,
"loss": 2.3155,
"step": 1083
},
{
"epoch": 0.009749385042248834,
"grad_norm": 1.6578973531723022,
"learning_rate": 4.967974039504666e-05,
"loss": 2.2692,
"step": 1084
},
{
"epoch": 0.009758378939889284,
"grad_norm": 1.943566918373108,
"learning_rate": 4.9678540127560904e-05,
"loss": 2.4199,
"step": 1085
},
{
"epoch": 0.009767372837529736,
"grad_norm": 1.309987187385559,
"learning_rate": 4.9677337629655264e-05,
"loss": 2.2808,
"step": 1086
},
{
"epoch": 0.009776366735170187,
"grad_norm": 1.8059741258621216,
"learning_rate": 4.967613290143843e-05,
"loss": 2.3359,
"step": 1087
},
{
"epoch": 0.009785360632810637,
"grad_norm": 1.6679922342300415,
"learning_rate": 4.967492594301928e-05,
"loss": 2.3139,
"step": 1088
},
{
"epoch": 0.00979435453045109,
"grad_norm": 1.5452219247817993,
"learning_rate": 4.9673716754506906e-05,
"loss": 2.3277,
"step": 1089
},
{
"epoch": 0.00980334842809154,
"grad_norm": 0.9614936709403992,
"learning_rate": 4.967250533601059e-05,
"loss": 2.2946,
"step": 1090
},
{
"epoch": 0.009812342325731992,
"grad_norm": 1.8186993598937988,
"learning_rate": 4.967129168763981e-05,
"loss": 2.3634,
"step": 1091
},
{
"epoch": 0.009821336223372442,
"grad_norm": 3.445974826812744,
"learning_rate": 4.967007580950427e-05,
"loss": 2.3921,
"step": 1092
},
{
"epoch": 0.009830330121012892,
"grad_norm": 214.7694549560547,
"learning_rate": 4.9668857701713844e-05,
"loss": 2.3131,
"step": 1093
},
{
"epoch": 0.009839324018653344,
"grad_norm": 129.35501098632812,
"learning_rate": 4.9667637364378625e-05,
"loss": 2.3979,
"step": 1094
},
{
"epoch": 0.009848317916293795,
"grad_norm": 22.181787490844727,
"learning_rate": 4.966641479760892e-05,
"loss": 2.3343,
"step": 1095
},
{
"epoch": 0.009857311813934245,
"grad_norm": 2.1598572731018066,
"learning_rate": 4.966519000151522e-05,
"loss": 2.3796,
"step": 1096
},
{
"epoch": 0.009866305711574697,
"grad_norm": 2.760488748550415,
"learning_rate": 4.96639629762082e-05,
"loss": 2.3354,
"step": 1097
},
{
"epoch": 0.009875299609215148,
"grad_norm": 23.494564056396484,
"learning_rate": 4.966273372179878e-05,
"loss": 2.301,
"step": 1098
},
{
"epoch": 0.009884293506855598,
"grad_norm": 958.7348022460938,
"learning_rate": 4.966150223839805e-05,
"loss": 2.6469,
"step": 1099
},
{
"epoch": 0.00989328740449605,
"grad_norm": 3.113955497741699,
"learning_rate": 4.966026852611731e-05,
"loss": 2.4272,
"step": 1100
},
{
"epoch": 0.0099022813021365,
"grad_norm": 52.463993072509766,
"learning_rate": 4.965903258506806e-05,
"loss": 2.3236,
"step": 1101
},
{
"epoch": 0.00991127519977695,
"grad_norm": 1.7886731624603271,
"learning_rate": 4.9657794415362014e-05,
"loss": 2.2811,
"step": 1102
},
{
"epoch": 0.009920269097417403,
"grad_norm": 2.48513126373291,
"learning_rate": 4.965655401711106e-05,
"loss": 2.2932,
"step": 1103
},
{
"epoch": 0.009929262995057853,
"grad_norm": 1.5509709119796753,
"learning_rate": 4.965531139042731e-05,
"loss": 2.3416,
"step": 1104
},
{
"epoch": 0.009938256892698304,
"grad_norm": 4.904428482055664,
"learning_rate": 4.9654066535423075e-05,
"loss": 2.3302,
"step": 1105
},
{
"epoch": 0.009947250790338756,
"grad_norm": 2.888819932937622,
"learning_rate": 4.9652819452210865e-05,
"loss": 2.3515,
"step": 1106
},
{
"epoch": 0.009956244687979206,
"grad_norm": 1.615666389465332,
"learning_rate": 4.9651570140903384e-05,
"loss": 2.2589,
"step": 1107
},
{
"epoch": 0.009965238585619656,
"grad_norm": 1.6766654253005981,
"learning_rate": 4.965031860161355e-05,
"loss": 2.3282,
"step": 1108
},
{
"epoch": 0.009974232483260109,
"grad_norm": 2.3949713706970215,
"learning_rate": 4.9649064834454474e-05,
"loss": 2.2685,
"step": 1109
},
{
"epoch": 0.009983226380900559,
"grad_norm": 1.4269980192184448,
"learning_rate": 4.9647808839539455e-05,
"loss": 2.3944,
"step": 1110
},
{
"epoch": 0.00999222027854101,
"grad_norm": 1.2341117858886719,
"learning_rate": 4.9646550616982034e-05,
"loss": 2.2753,
"step": 1111
},
{
"epoch": 0.010001214176181461,
"grad_norm": 1.9939247369766235,
"learning_rate": 4.9645290166895906e-05,
"loss": 2.3253,
"step": 1112
},
{
"epoch": 0.010010208073821912,
"grad_norm": 1.176496982574463,
"learning_rate": 4.9644027489395005e-05,
"loss": 2.3194,
"step": 1113
},
{
"epoch": 0.010019201971462362,
"grad_norm": 2.368574380874634,
"learning_rate": 4.964276258459345e-05,
"loss": 2.3987,
"step": 1114
},
{
"epoch": 0.010028195869102814,
"grad_norm": 1.1555325984954834,
"learning_rate": 4.9641495452605545e-05,
"loss": 2.2783,
"step": 1115
},
{
"epoch": 0.010037189766743265,
"grad_norm": 11.08785343170166,
"learning_rate": 4.964022609354583e-05,
"loss": 2.3416,
"step": 1116
},
{
"epoch": 0.010046183664383715,
"grad_norm": 3.6222121715545654,
"learning_rate": 4.9638954507529014e-05,
"loss": 2.3072,
"step": 1117
},
{
"epoch": 0.010055177562024167,
"grad_norm": 5.869036674499512,
"learning_rate": 4.963768069467003e-05,
"loss": 2.3855,
"step": 1118
},
{
"epoch": 0.010064171459664617,
"grad_norm": 1.9966332912445068,
"learning_rate": 4.9636404655084e-05,
"loss": 2.3254,
"step": 1119
},
{
"epoch": 0.010073165357305068,
"grad_norm": 4.625076770782471,
"learning_rate": 4.9635126388886256e-05,
"loss": 2.3371,
"step": 1120
},
{
"epoch": 0.01008215925494552,
"grad_norm": 3.0359678268432617,
"learning_rate": 4.963384589619233e-05,
"loss": 2.2586,
"step": 1121
},
{
"epoch": 0.01009115315258597,
"grad_norm": 1.8090620040893555,
"learning_rate": 4.963256317711794e-05,
"loss": 2.2716,
"step": 1122
},
{
"epoch": 0.01010014705022642,
"grad_norm": 1.4668354988098145,
"learning_rate": 4.963127823177902e-05,
"loss": 2.3089,
"step": 1123
},
{
"epoch": 0.010109140947866873,
"grad_norm": 2.646360158920288,
"learning_rate": 4.96299910602917e-05,
"loss": 2.3216,
"step": 1124
},
{
"epoch": 0.010118134845507323,
"grad_norm": 1.7903180122375488,
"learning_rate": 4.9628701662772325e-05,
"loss": 2.2168,
"step": 1125
},
{
"epoch": 0.010127128743147775,
"grad_norm": 1.1596676111221313,
"learning_rate": 4.962741003933742e-05,
"loss": 2.2366,
"step": 1126
},
{
"epoch": 0.010136122640788225,
"grad_norm": 2.070632219314575,
"learning_rate": 4.9626116190103724e-05,
"loss": 2.3454,
"step": 1127
},
{
"epoch": 0.010145116538428676,
"grad_norm": 0.9789115786552429,
"learning_rate": 4.9624820115188174e-05,
"loss": 2.2597,
"step": 1128
},
{
"epoch": 0.010154110436069128,
"grad_norm": 1.8018876314163208,
"learning_rate": 4.9623521814707894e-05,
"loss": 2.305,
"step": 1129
},
{
"epoch": 0.010163104333709578,
"grad_norm": 0.8252420425415039,
"learning_rate": 4.962222128878025e-05,
"loss": 2.2355,
"step": 1130
},
{
"epoch": 0.010172098231350029,
"grad_norm": 1.1904879808425903,
"learning_rate": 4.962091853752275e-05,
"loss": 2.3075,
"step": 1131
},
{
"epoch": 0.01018109212899048,
"grad_norm": 1.8671382665634155,
"learning_rate": 4.961961356105317e-05,
"loss": 2.2624,
"step": 1132
},
{
"epoch": 0.010190086026630931,
"grad_norm": 1.6544936895370483,
"learning_rate": 4.961830635948942e-05,
"loss": 2.3444,
"step": 1133
},
{
"epoch": 0.010199079924271381,
"grad_norm": 2.3454082012176514,
"learning_rate": 4.9616996932949666e-05,
"loss": 2.4088,
"step": 1134
},
{
"epoch": 0.010208073821911834,
"grad_norm": 1.764322280883789,
"learning_rate": 4.961568528155224e-05,
"loss": 2.2591,
"step": 1135
},
{
"epoch": 0.010217067719552284,
"grad_norm": 1.2431902885437012,
"learning_rate": 4.9614371405415696e-05,
"loss": 2.3133,
"step": 1136
},
{
"epoch": 0.010226061617192734,
"grad_norm": 1.7182575464248657,
"learning_rate": 4.9613055304658776e-05,
"loss": 2.2835,
"step": 1137
},
{
"epoch": 0.010235055514833186,
"grad_norm": 16.691743850708008,
"learning_rate": 4.961173697940042e-05,
"loss": 2.2866,
"step": 1138
},
{
"epoch": 0.010244049412473637,
"grad_norm": 1.0919997692108154,
"learning_rate": 4.9610416429759797e-05,
"loss": 2.3138,
"step": 1139
},
{
"epoch": 0.010253043310114087,
"grad_norm": 2.4553284645080566,
"learning_rate": 4.960909365585624e-05,
"loss": 2.4208,
"step": 1140
},
{
"epoch": 0.01026203720775454,
"grad_norm": 2.8286502361297607,
"learning_rate": 4.9607768657809306e-05,
"loss": 2.3583,
"step": 1141
},
{
"epoch": 0.01027103110539499,
"grad_norm": 1.5299936532974243,
"learning_rate": 4.960644143573875e-05,
"loss": 2.3979,
"step": 1142
},
{
"epoch": 0.01028002500303544,
"grad_norm": 1.3973337411880493,
"learning_rate": 4.960511198976451e-05,
"loss": 2.2907,
"step": 1143
},
{
"epoch": 0.010289018900675892,
"grad_norm": 1.4424344301223755,
"learning_rate": 4.960378032000676e-05,
"loss": 2.2768,
"step": 1144
},
{
"epoch": 0.010298012798316342,
"grad_norm": 1.1546728610992432,
"learning_rate": 4.960244642658585e-05,
"loss": 2.2212,
"step": 1145
},
{
"epoch": 0.010307006695956793,
"grad_norm": 3.8030622005462646,
"learning_rate": 4.960111030962232e-05,
"loss": 2.3394,
"step": 1146
},
{
"epoch": 0.010316000593597245,
"grad_norm": 2.399183750152588,
"learning_rate": 4.959977196923694e-05,
"loss": 2.3322,
"step": 1147
},
{
"epoch": 0.010324994491237695,
"grad_norm": 3.7612507343292236,
"learning_rate": 4.959843140555067e-05,
"loss": 2.3258,
"step": 1148
},
{
"epoch": 0.010333988388878146,
"grad_norm": 2.4283108711242676,
"learning_rate": 4.959708861868466e-05,
"loss": 2.3865,
"step": 1149
},
{
"epoch": 0.010342982286518598,
"grad_norm": 2.4122281074523926,
"learning_rate": 4.959574360876028e-05,
"loss": 2.3813,
"step": 1150
},
{
"epoch": 0.010351976184159048,
"grad_norm": 2.301945447921753,
"learning_rate": 4.959439637589909e-05,
"loss": 2.3788,
"step": 1151
},
{
"epoch": 0.010360970081799498,
"grad_norm": 2.392803907394409,
"learning_rate": 4.959304692022284e-05,
"loss": 2.3019,
"step": 1152
},
{
"epoch": 0.01036996397943995,
"grad_norm": 1.652524709701538,
"learning_rate": 4.9591695241853495e-05,
"loss": 2.2857,
"step": 1153
},
{
"epoch": 0.0103789578770804,
"grad_norm": 1.127707600593567,
"learning_rate": 4.959034134091323e-05,
"loss": 2.2377,
"step": 1154
},
{
"epoch": 0.010387951774720851,
"grad_norm": 2.509152412414551,
"learning_rate": 4.95889852175244e-05,
"loss": 2.3167,
"step": 1155
},
{
"epoch": 0.010396945672361303,
"grad_norm": 2.5287764072418213,
"learning_rate": 4.958762687180956e-05,
"loss": 2.3174,
"step": 1156
},
{
"epoch": 0.010405939570001754,
"grad_norm": 2.6051721572875977,
"learning_rate": 4.958626630389149e-05,
"loss": 2.3245,
"step": 1157
},
{
"epoch": 0.010414933467642204,
"grad_norm": 3.279592275619507,
"learning_rate": 4.958490351389316e-05,
"loss": 2.3171,
"step": 1158
},
{
"epoch": 0.010423927365282656,
"grad_norm": 3.551753520965576,
"learning_rate": 4.958353850193773e-05,
"loss": 2.2767,
"step": 1159
},
{
"epoch": 0.010432921262923106,
"grad_norm": 19.09039306640625,
"learning_rate": 4.958217126814857e-05,
"loss": 2.3301,
"step": 1160
},
{
"epoch": 0.010441915160563557,
"grad_norm": 1.6319879293441772,
"learning_rate": 4.958080181264925e-05,
"loss": 2.3,
"step": 1161
},
{
"epoch": 0.010450909058204009,
"grad_norm": 1.3658466339111328,
"learning_rate": 4.9579430135563534e-05,
"loss": 2.3458,
"step": 1162
},
{
"epoch": 0.01045990295584446,
"grad_norm": 1.20742666721344,
"learning_rate": 4.95780562370154e-05,
"loss": 2.2795,
"step": 1163
},
{
"epoch": 0.010468896853484911,
"grad_norm": 1.6854474544525146,
"learning_rate": 4.957668011712902e-05,
"loss": 2.3564,
"step": 1164
},
{
"epoch": 0.010477890751125362,
"grad_norm": 2.292989492416382,
"learning_rate": 4.957530177602876e-05,
"loss": 2.3595,
"step": 1165
},
{
"epoch": 0.010486884648765812,
"grad_norm": 1.4159866571426392,
"learning_rate": 4.9573921213839194e-05,
"loss": 2.3254,
"step": 1166
},
{
"epoch": 0.010495878546406264,
"grad_norm": 1.691332221031189,
"learning_rate": 4.9572538430685094e-05,
"loss": 2.3564,
"step": 1167
},
{
"epoch": 0.010504872444046714,
"grad_norm": 2.640218734741211,
"learning_rate": 4.957115342669144e-05,
"loss": 2.3182,
"step": 1168
},
{
"epoch": 0.010513866341687165,
"grad_norm": 1.481858253479004,
"learning_rate": 4.956976620198342e-05,
"loss": 2.2696,
"step": 1169
},
{
"epoch": 0.010522860239327617,
"grad_norm": 1.0896230936050415,
"learning_rate": 4.956837675668637e-05,
"loss": 2.2795,
"step": 1170
},
{
"epoch": 0.010531854136968067,
"grad_norm": 1.335352897644043,
"learning_rate": 4.9566985090925904e-05,
"loss": 2.3225,
"step": 1171
},
{
"epoch": 0.010540848034608518,
"grad_norm": 8.702324867248535,
"learning_rate": 4.956559120482779e-05,
"loss": 2.3123,
"step": 1172
},
{
"epoch": 0.01054984193224897,
"grad_norm": 2.045257329940796,
"learning_rate": 4.9564195098518e-05,
"loss": 2.3944,
"step": 1173
},
{
"epoch": 0.01055883582988942,
"grad_norm": 1.743530511856079,
"learning_rate": 4.956279677212271e-05,
"loss": 2.3436,
"step": 1174
},
{
"epoch": 0.01056782972752987,
"grad_norm": 2.2260138988494873,
"learning_rate": 4.956139622576831e-05,
"loss": 2.4495,
"step": 1175
},
{
"epoch": 0.010576823625170323,
"grad_norm": 1.0863250494003296,
"learning_rate": 4.9559993459581375e-05,
"loss": 2.3514,
"step": 1176
},
{
"epoch": 0.010585817522810773,
"grad_norm": 1.050481915473938,
"learning_rate": 4.9558588473688685e-05,
"loss": 2.3664,
"step": 1177
},
{
"epoch": 0.010594811420451223,
"grad_norm": 2.1504368782043457,
"learning_rate": 4.9557181268217227e-05,
"loss": 2.3724,
"step": 1178
},
{
"epoch": 0.010603805318091675,
"grad_norm": 3.4603984355926514,
"learning_rate": 4.955577184329416e-05,
"loss": 2.3355,
"step": 1179
},
{
"epoch": 0.010612799215732126,
"grad_norm": 1.2105674743652344,
"learning_rate": 4.955436019904689e-05,
"loss": 2.3539,
"step": 1180
},
{
"epoch": 0.010621793113372576,
"grad_norm": 1.112475037574768,
"learning_rate": 4.9552946335603e-05,
"loss": 2.2702,
"step": 1181
},
{
"epoch": 0.010630787011013028,
"grad_norm": 1.2488195896148682,
"learning_rate": 4.9551530253090263e-05,
"loss": 2.2923,
"step": 1182
},
{
"epoch": 0.010639780908653479,
"grad_norm": 1.3281227350234985,
"learning_rate": 4.9550111951636666e-05,
"loss": 2.2355,
"step": 1183
},
{
"epoch": 0.010648774806293929,
"grad_norm": 2.6969902515411377,
"learning_rate": 4.954869143137039e-05,
"loss": 2.288,
"step": 1184
},
{
"epoch": 0.010657768703934381,
"grad_norm": 1.9387229681015015,
"learning_rate": 4.9547268692419834e-05,
"loss": 2.3412,
"step": 1185
},
{
"epoch": 0.010666762601574831,
"grad_norm": 0.9136784672737122,
"learning_rate": 4.954584373491357e-05,
"loss": 2.2832,
"step": 1186
},
{
"epoch": 0.010675756499215282,
"grad_norm": 1.5051130056381226,
"learning_rate": 4.9544416558980376e-05,
"loss": 2.3421,
"step": 1187
},
{
"epoch": 0.010684750396855734,
"grad_norm": 0.7083832621574402,
"learning_rate": 4.9542987164749264e-05,
"loss": 2.2245,
"step": 1188
},
{
"epoch": 0.010693744294496184,
"grad_norm": 0.9482129812240601,
"learning_rate": 4.95415555523494e-05,
"loss": 2.2884,
"step": 1189
},
{
"epoch": 0.010702738192136635,
"grad_norm": 1.8723399639129639,
"learning_rate": 4.954012172191018e-05,
"loss": 2.3801,
"step": 1190
},
{
"epoch": 0.010711732089777087,
"grad_norm": 1.4367190599441528,
"learning_rate": 4.95386856735612e-05,
"loss": 2.2496,
"step": 1191
},
{
"epoch": 0.010720725987417537,
"grad_norm": 1.2865971326828003,
"learning_rate": 4.953724740743223e-05,
"loss": 2.3046,
"step": 1192
},
{
"epoch": 0.010729719885057987,
"grad_norm": 1.185848593711853,
"learning_rate": 4.9535806923653275e-05,
"loss": 2.2834,
"step": 1193
},
{
"epoch": 0.01073871378269844,
"grad_norm": 1.4188623428344727,
"learning_rate": 4.953436422235451e-05,
"loss": 2.2954,
"step": 1194
},
{
"epoch": 0.01074770768033889,
"grad_norm": 1.5299527645111084,
"learning_rate": 4.953291930366634e-05,
"loss": 2.2884,
"step": 1195
},
{
"epoch": 0.01075670157797934,
"grad_norm": 1.187511682510376,
"learning_rate": 4.953147216771935e-05,
"loss": 2.359,
"step": 1196
},
{
"epoch": 0.010765695475619792,
"grad_norm": 1.9731473922729492,
"learning_rate": 4.953002281464432e-05,
"loss": 2.2737,
"step": 1197
},
{
"epoch": 0.010774689373260243,
"grad_norm": 0.9455071687698364,
"learning_rate": 4.952857124457226e-05,
"loss": 2.2927,
"step": 1198
},
{
"epoch": 0.010783683270900695,
"grad_norm": 1.5308231115341187,
"learning_rate": 4.952711745763434e-05,
"loss": 2.3734,
"step": 1199
},
{
"epoch": 0.010792677168541145,
"grad_norm": 1.7899839878082275,
"learning_rate": 4.952566145396197e-05,
"loss": 2.1991,
"step": 1200
},
{
"epoch": 0.010801671066181595,
"grad_norm": 1.5891908407211304,
"learning_rate": 4.952420323368673e-05,
"loss": 2.4113,
"step": 1201
},
{
"epoch": 0.010810664963822048,
"grad_norm": 1.4638893604278564,
"learning_rate": 4.952274279694042e-05,
"loss": 2.3024,
"step": 1202
},
{
"epoch": 0.010819658861462498,
"grad_norm": 2.119044542312622,
"learning_rate": 4.952128014385503e-05,
"loss": 2.308,
"step": 1203
},
{
"epoch": 0.010828652759102948,
"grad_norm": 1.6738792657852173,
"learning_rate": 4.9519815274562756e-05,
"loss": 2.2302,
"step": 1204
},
{
"epoch": 0.0108376466567434,
"grad_norm": 1.505308747291565,
"learning_rate": 4.9518348189195984e-05,
"loss": 2.3468,
"step": 1205
},
{
"epoch": 0.01084664055438385,
"grad_norm": 1.9311424493789673,
"learning_rate": 4.9516878887887305e-05,
"loss": 2.224,
"step": 1206
},
{
"epoch": 0.010855634452024301,
"grad_norm": 2.1454856395721436,
"learning_rate": 4.9515407370769526e-05,
"loss": 2.338,
"step": 1207
},
{
"epoch": 0.010864628349664753,
"grad_norm": 1.1918364763259888,
"learning_rate": 4.951393363797564e-05,
"loss": 2.3401,
"step": 1208
},
{
"epoch": 0.010873622247305204,
"grad_norm": 1.2202930450439453,
"learning_rate": 4.951245768963883e-05,
"loss": 2.3299,
"step": 1209
},
{
"epoch": 0.010882616144945654,
"grad_norm": 1.1868610382080078,
"learning_rate": 4.95109795258925e-05,
"loss": 2.3391,
"step": 1210
},
{
"epoch": 0.010891610042586106,
"grad_norm": 1.0417895317077637,
"learning_rate": 4.9509499146870236e-05,
"loss": 2.2816,
"step": 1211
},
{
"epoch": 0.010900603940226556,
"grad_norm": 2.5350186824798584,
"learning_rate": 4.9508016552705835e-05,
"loss": 2.3242,
"step": 1212
},
{
"epoch": 0.010909597837867007,
"grad_norm": 1.461430311203003,
"learning_rate": 4.9506531743533316e-05,
"loss": 2.34,
"step": 1213
},
{
"epoch": 0.010918591735507459,
"grad_norm": 3.288555145263672,
"learning_rate": 4.9505044719486836e-05,
"loss": 2.4029,
"step": 1214
},
{
"epoch": 0.01092758563314791,
"grad_norm": 2.890031337738037,
"learning_rate": 4.9503555480700815e-05,
"loss": 2.3105,
"step": 1215
},
{
"epoch": 0.01093657953078836,
"grad_norm": 5.223935127258301,
"learning_rate": 4.9502064027309836e-05,
"loss": 2.2961,
"step": 1216
},
{
"epoch": 0.010945573428428812,
"grad_norm": 96.04720306396484,
"learning_rate": 4.950057035944871e-05,
"loss": 2.4028,
"step": 1217
},
{
"epoch": 0.010954567326069262,
"grad_norm": 2.35675311088562,
"learning_rate": 4.949907447725242e-05,
"loss": 2.3561,
"step": 1218
},
{
"epoch": 0.010963561223709712,
"grad_norm": 2.025588274002075,
"learning_rate": 4.949757638085617e-05,
"loss": 2.3425,
"step": 1219
},
{
"epoch": 0.010972555121350164,
"grad_norm": 1.438686490058899,
"learning_rate": 4.9496076070395346e-05,
"loss": 2.3929,
"step": 1220
},
{
"epoch": 0.010981549018990615,
"grad_norm": 2.8650848865509033,
"learning_rate": 4.949457354600556e-05,
"loss": 2.2802,
"step": 1221
},
{
"epoch": 0.010990542916631065,
"grad_norm": 1.5126683712005615,
"learning_rate": 4.94930688078226e-05,
"loss": 2.2844,
"step": 1222
},
{
"epoch": 0.010999536814271517,
"grad_norm": 1.913718819618225,
"learning_rate": 4.949156185598246e-05,
"loss": 2.3641,
"step": 1223
},
{
"epoch": 0.011008530711911968,
"grad_norm": 1.2559486627578735,
"learning_rate": 4.949005269062134e-05,
"loss": 2.233,
"step": 1224
},
{
"epoch": 0.011017524609552418,
"grad_norm": 1.4811545610427856,
"learning_rate": 4.948854131187564e-05,
"loss": 2.3706,
"step": 1225
},
{
"epoch": 0.01102651850719287,
"grad_norm": 1.481111764907837,
"learning_rate": 4.948702771988195e-05,
"loss": 2.3131,
"step": 1226
},
{
"epoch": 0.01103551240483332,
"grad_norm": 1.9949358701705933,
"learning_rate": 4.948551191477707e-05,
"loss": 2.3826,
"step": 1227
},
{
"epoch": 0.01104450630247377,
"grad_norm": 2.204502582550049,
"learning_rate": 4.9483993896697996e-05,
"loss": 2.342,
"step": 1228
},
{
"epoch": 0.011053500200114223,
"grad_norm": 2.171661376953125,
"learning_rate": 4.948247366578192e-05,
"loss": 2.3162,
"step": 1229
},
{
"epoch": 0.011062494097754673,
"grad_norm": 2.7246861457824707,
"learning_rate": 4.9480951222166254e-05,
"loss": 2.3667,
"step": 1230
},
{
"epoch": 0.011071487995395124,
"grad_norm": 1.9337729215621948,
"learning_rate": 4.9479426565988585e-05,
"loss": 2.3223,
"step": 1231
},
{
"epoch": 0.011080481893035576,
"grad_norm": 2.4414193630218506,
"learning_rate": 4.947789969738671e-05,
"loss": 2.3164,
"step": 1232
},
{
"epoch": 0.011089475790676026,
"grad_norm": 2.173196315765381,
"learning_rate": 4.9476370616498617e-05,
"loss": 2.2918,
"step": 1233
},
{
"epoch": 0.011098469688316476,
"grad_norm": 2.7107467651367188,
"learning_rate": 4.947483932346252e-05,
"loss": 2.3572,
"step": 1234
},
{
"epoch": 0.011107463585956929,
"grad_norm": 1.258103370666504,
"learning_rate": 4.9473305818416806e-05,
"loss": 2.2838,
"step": 1235
},
{
"epoch": 0.011116457483597379,
"grad_norm": 2.1360626220703125,
"learning_rate": 4.947177010150007e-05,
"loss": 2.3383,
"step": 1236
},
{
"epoch": 0.011125451381237831,
"grad_norm": 1.0039118528366089,
"learning_rate": 4.9470232172851114e-05,
"loss": 2.3596,
"step": 1237
},
{
"epoch": 0.011134445278878281,
"grad_norm": 2.0521559715270996,
"learning_rate": 4.9468692032608926e-05,
"loss": 2.3033,
"step": 1238
},
{
"epoch": 0.011143439176518732,
"grad_norm": 1.234839677810669,
"learning_rate": 4.946714968091271e-05,
"loss": 2.246,
"step": 1239
},
{
"epoch": 0.011152433074159184,
"grad_norm": 1.0917353630065918,
"learning_rate": 4.946560511790186e-05,
"loss": 2.2894,
"step": 1240
},
{
"epoch": 0.011161426971799634,
"grad_norm": 1.0826351642608643,
"learning_rate": 4.946405834371597e-05,
"loss": 2.3392,
"step": 1241
},
{
"epoch": 0.011170420869440085,
"grad_norm": 1.1655303239822388,
"learning_rate": 4.9462509358494844e-05,
"loss": 2.3199,
"step": 1242
},
{
"epoch": 0.011179414767080537,
"grad_norm": 1.2636141777038574,
"learning_rate": 4.946095816237847e-05,
"loss": 2.3306,
"step": 1243
},
{
"epoch": 0.011188408664720987,
"grad_norm": 2.016878604888916,
"learning_rate": 4.945940475550703e-05,
"loss": 2.3402,
"step": 1244
},
{
"epoch": 0.011197402562361437,
"grad_norm": 1.3355700969696045,
"learning_rate": 4.945784913802095e-05,
"loss": 2.243,
"step": 1245
},
{
"epoch": 0.01120639646000189,
"grad_norm": 2.1973514556884766,
"learning_rate": 4.9456291310060806e-05,
"loss": 2.3478,
"step": 1246
},
{
"epoch": 0.01121539035764234,
"grad_norm": 1.1568211317062378,
"learning_rate": 4.9454731271767395e-05,
"loss": 2.3208,
"step": 1247
},
{
"epoch": 0.01122438425528279,
"grad_norm": 1.5626015663146973,
"learning_rate": 4.945316902328171e-05,
"loss": 2.2441,
"step": 1248
},
{
"epoch": 0.011233378152923242,
"grad_norm": 2.322676420211792,
"learning_rate": 4.9451604564744944e-05,
"loss": 2.3715,
"step": 1249
},
{
"epoch": 0.011242372050563693,
"grad_norm": 2.1260242462158203,
"learning_rate": 4.9450037896298495e-05,
"loss": 2.341,
"step": 1250
},
{
"epoch": 0.011251365948204143,
"grad_norm": 1.714827537536621,
"learning_rate": 4.9448469018083965e-05,
"loss": 2.2521,
"step": 1251
},
{
"epoch": 0.011260359845844595,
"grad_norm": 6.027016639709473,
"learning_rate": 4.944689793024313e-05,
"loss": 2.2911,
"step": 1252
},
{
"epoch": 0.011269353743485045,
"grad_norm": 2.1543612480163574,
"learning_rate": 4.9445324632918e-05,
"loss": 2.3697,
"step": 1253
},
{
"epoch": 0.011278347641125496,
"grad_norm": 1.6868714094161987,
"learning_rate": 4.944374912625075e-05,
"loss": 2.3858,
"step": 1254
},
{
"epoch": 0.011287341538765948,
"grad_norm": 1.0732194185256958,
"learning_rate": 4.944217141038379e-05,
"loss": 2.3034,
"step": 1255
},
{
"epoch": 0.011296335436406398,
"grad_norm": 1.4649441242218018,
"learning_rate": 4.9440591485459706e-05,
"loss": 2.3401,
"step": 1256
},
{
"epoch": 0.011305329334046849,
"grad_norm": 1.239201545715332,
"learning_rate": 4.943900935162128e-05,
"loss": 2.3145,
"step": 1257
},
{
"epoch": 0.0113143232316873,
"grad_norm": 1.889357566833496,
"learning_rate": 4.9437425009011526e-05,
"loss": 2.2848,
"step": 1258
},
{
"epoch": 0.011323317129327751,
"grad_norm": 1.7509350776672363,
"learning_rate": 4.9435838457773606e-05,
"loss": 2.2887,
"step": 1259
},
{
"epoch": 0.011332311026968201,
"grad_norm": 2.5500247478485107,
"learning_rate": 4.943424969805094e-05,
"loss": 2.302,
"step": 1260
},
{
"epoch": 0.011341304924608653,
"grad_norm": 2.0146055221557617,
"learning_rate": 4.94326587299871e-05,
"loss": 2.3183,
"step": 1261
},
{
"epoch": 0.011350298822249104,
"grad_norm": 2.1095030307769775,
"learning_rate": 4.9431065553725874e-05,
"loss": 2.293,
"step": 1262
},
{
"epoch": 0.011359292719889554,
"grad_norm": 0.8967338800430298,
"learning_rate": 4.942947016941126e-05,
"loss": 2.2293,
"step": 1263
},
{
"epoch": 0.011368286617530006,
"grad_norm": 1.6413196325302124,
"learning_rate": 4.942787257718745e-05,
"loss": 2.2868,
"step": 1264
},
{
"epoch": 0.011377280515170457,
"grad_norm": 0.9218994379043579,
"learning_rate": 4.9426272777198836e-05,
"loss": 2.3297,
"step": 1265
},
{
"epoch": 0.011386274412810907,
"grad_norm": 2.4999451637268066,
"learning_rate": 4.9424670769589984e-05,
"loss": 2.2538,
"step": 1266
},
{
"epoch": 0.011395268310451359,
"grad_norm": 1.861364483833313,
"learning_rate": 4.9423066554505706e-05,
"loss": 2.3073,
"step": 1267
},
{
"epoch": 0.01140426220809181,
"grad_norm": 2.0318455696105957,
"learning_rate": 4.942146013209097e-05,
"loss": 2.3113,
"step": 1268
},
{
"epoch": 0.01141325610573226,
"grad_norm": 1.3827846050262451,
"learning_rate": 4.9419851502490976e-05,
"loss": 2.2541,
"step": 1269
},
{
"epoch": 0.011422250003372712,
"grad_norm": 1.509055733680725,
"learning_rate": 4.9418240665851106e-05,
"loss": 2.3673,
"step": 1270
},
{
"epoch": 0.011431243901013162,
"grad_norm": 1.5659303665161133,
"learning_rate": 4.941662762231695e-05,
"loss": 2.3634,
"step": 1271
},
{
"epoch": 0.011440237798653613,
"grad_norm": 1.633475661277771,
"learning_rate": 4.941501237203428e-05,
"loss": 2.3204,
"step": 1272
},
{
"epoch": 0.011449231696294065,
"grad_norm": 1.589223027229309,
"learning_rate": 4.9413394915149094e-05,
"loss": 2.3345,
"step": 1273
},
{
"epoch": 0.011458225593934515,
"grad_norm": 1.170649528503418,
"learning_rate": 4.941177525180757e-05,
"loss": 2.2869,
"step": 1274
},
{
"epoch": 0.011467219491574967,
"grad_norm": 0.9948899745941162,
"learning_rate": 4.94101533821561e-05,
"loss": 2.2896,
"step": 1275
},
{
"epoch": 0.011476213389215418,
"grad_norm": 0.8960604667663574,
"learning_rate": 4.9408529306341255e-05,
"loss": 2.2698,
"step": 1276
},
{
"epoch": 0.011485207286855868,
"grad_norm": 2.304053783416748,
"learning_rate": 4.940690302450982e-05,
"loss": 2.3175,
"step": 1277
},
{
"epoch": 0.01149420118449632,
"grad_norm": 1.649854063987732,
"learning_rate": 4.940527453680878e-05,
"loss": 2.3434,
"step": 1278
},
{
"epoch": 0.01150319508213677,
"grad_norm": 1.9188902378082275,
"learning_rate": 4.940364384338532e-05,
"loss": 2.3023,
"step": 1279
},
{
"epoch": 0.01151218897977722,
"grad_norm": 1.5064107179641724,
"learning_rate": 4.94020109443868e-05,
"loss": 2.3428,
"step": 1280
},
{
"epoch": 0.011521182877417673,
"grad_norm": 1.1684613227844238,
"learning_rate": 4.9400375839960826e-05,
"loss": 2.274,
"step": 1281
},
{
"epoch": 0.011530176775058123,
"grad_norm": 1.879447340965271,
"learning_rate": 4.9398738530255164e-05,
"loss": 2.311,
"step": 1282
},
{
"epoch": 0.011539170672698574,
"grad_norm": 5.295740127563477,
"learning_rate": 4.9397099015417794e-05,
"loss": 2.3564,
"step": 1283
},
{
"epoch": 0.011548164570339026,
"grad_norm": 1.2326210737228394,
"learning_rate": 4.9395457295596895e-05,
"loss": 2.2755,
"step": 1284
},
{
"epoch": 0.011557158467979476,
"grad_norm": 1.1373960971832275,
"learning_rate": 4.939381337094084e-05,
"loss": 2.2385,
"step": 1285
},
{
"epoch": 0.011566152365619926,
"grad_norm": 1.6129990816116333,
"learning_rate": 4.9392167241598206e-05,
"loss": 2.3477,
"step": 1286
},
{
"epoch": 0.011575146263260378,
"grad_norm": 1.6414514780044556,
"learning_rate": 4.939051890771777e-05,
"loss": 2.3078,
"step": 1287
},
{
"epoch": 0.011584140160900829,
"grad_norm": 1.6887080669403076,
"learning_rate": 4.938886836944851e-05,
"loss": 2.2672,
"step": 1288
},
{
"epoch": 0.01159313405854128,
"grad_norm": 1.3660883903503418,
"learning_rate": 4.938721562693959e-05,
"loss": 2.3015,
"step": 1289
},
{
"epoch": 0.011602127956181731,
"grad_norm": 0.9072462916374207,
"learning_rate": 4.93855606803404e-05,
"loss": 2.2724,
"step": 1290
},
{
"epoch": 0.011611121853822182,
"grad_norm": 1.0197970867156982,
"learning_rate": 4.938390352980049e-05,
"loss": 2.322,
"step": 1291
},
{
"epoch": 0.011620115751462632,
"grad_norm": 1.6257226467132568,
"learning_rate": 4.938224417546965e-05,
"loss": 2.292,
"step": 1292
},
{
"epoch": 0.011629109649103084,
"grad_norm": 0.8640071749687195,
"learning_rate": 4.938058261749784e-05,
"loss": 2.3107,
"step": 1293
},
{
"epoch": 0.011638103546743534,
"grad_norm": 2.1623289585113525,
"learning_rate": 4.937891885603524e-05,
"loss": 2.3685,
"step": 1294
},
{
"epoch": 0.011647097444383985,
"grad_norm": 2.623000144958496,
"learning_rate": 4.9377252891232205e-05,
"loss": 2.4302,
"step": 1295
},
{
"epoch": 0.011656091342024437,
"grad_norm": 4.7325873374938965,
"learning_rate": 4.937558472323932e-05,
"loss": 2.3178,
"step": 1296
},
{
"epoch": 0.011665085239664887,
"grad_norm": 2.6786139011383057,
"learning_rate": 4.937391435220733e-05,
"loss": 2.3095,
"step": 1297
},
{
"epoch": 0.011674079137305338,
"grad_norm": 1.3070074319839478,
"learning_rate": 4.937224177828722e-05,
"loss": 2.3799,
"step": 1298
},
{
"epoch": 0.01168307303494579,
"grad_norm": 4.2565693855285645,
"learning_rate": 4.937056700163015e-05,
"loss": 2.3069,
"step": 1299
},
{
"epoch": 0.01169206693258624,
"grad_norm": 2.372837543487549,
"learning_rate": 4.936889002238749e-05,
"loss": 2.2814,
"step": 1300
},
{
"epoch": 0.01170106083022669,
"grad_norm": 4.036974906921387,
"learning_rate": 4.936721084071079e-05,
"loss": 2.3156,
"step": 1301
},
{
"epoch": 0.011710054727867143,
"grad_norm": 1.7612476348876953,
"learning_rate": 4.9365529456751823e-05,
"loss": 2.3362,
"step": 1302
},
{
"epoch": 0.011719048625507593,
"grad_norm": 1.62795889377594,
"learning_rate": 4.936384587066255e-05,
"loss": 2.3077,
"step": 1303
},
{
"epoch": 0.011728042523148043,
"grad_norm": 1.8024780750274658,
"learning_rate": 4.936216008259513e-05,
"loss": 2.2378,
"step": 1304
},
{
"epoch": 0.011737036420788495,
"grad_norm": 1.9589920043945312,
"learning_rate": 4.9360472092701914e-05,
"loss": 2.3331,
"step": 1305
},
{
"epoch": 0.011746030318428946,
"grad_norm": 5.74002742767334,
"learning_rate": 4.935878190113548e-05,
"loss": 2.272,
"step": 1306
},
{
"epoch": 0.011755024216069396,
"grad_norm": 1.2023650407791138,
"learning_rate": 4.935708950804857e-05,
"loss": 2.3192,
"step": 1307
},
{
"epoch": 0.011764018113709848,
"grad_norm": 2.29329252243042,
"learning_rate": 4.935539491359415e-05,
"loss": 2.4965,
"step": 1308
},
{
"epoch": 0.011773012011350299,
"grad_norm": 5.815248489379883,
"learning_rate": 4.935369811792536e-05,
"loss": 2.2611,
"step": 1309
},
{
"epoch": 0.01178200590899075,
"grad_norm": 1.0783790349960327,
"learning_rate": 4.935199912119558e-05,
"loss": 2.315,
"step": 1310
},
{
"epoch": 0.011790999806631201,
"grad_norm": 1.660359263420105,
"learning_rate": 4.935029792355834e-05,
"loss": 2.2553,
"step": 1311
},
{
"epoch": 0.011799993704271651,
"grad_norm": 1.8036712408065796,
"learning_rate": 4.93485945251674e-05,
"loss": 2.333,
"step": 1312
},
{
"epoch": 0.011808987601912103,
"grad_norm": 1.599664330482483,
"learning_rate": 4.934688892617672e-05,
"loss": 2.3116,
"step": 1313
},
{
"epoch": 0.011817981499552554,
"grad_norm": 4.0947957038879395,
"learning_rate": 4.934518112674044e-05,
"loss": 2.3851,
"step": 1314
},
{
"epoch": 0.011826975397193004,
"grad_norm": 1.2490711212158203,
"learning_rate": 4.9343471127012916e-05,
"loss": 2.3138,
"step": 1315
},
{
"epoch": 0.011835969294833456,
"grad_norm": 1.5987191200256348,
"learning_rate": 4.9341758927148685e-05,
"loss": 2.2805,
"step": 1316
},
{
"epoch": 0.011844963192473907,
"grad_norm": 1.4530214071273804,
"learning_rate": 4.934004452730251e-05,
"loss": 2.2581,
"step": 1317
},
{
"epoch": 0.011853957090114357,
"grad_norm": 1.4896512031555176,
"learning_rate": 4.933832792762932e-05,
"loss": 2.2328,
"step": 1318
},
{
"epoch": 0.011862950987754809,
"grad_norm": 1.3908450603485107,
"learning_rate": 4.933660912828427e-05,
"loss": 2.2771,
"step": 1319
},
{
"epoch": 0.01187194488539526,
"grad_norm": 1.1553617715835571,
"learning_rate": 4.93348881294227e-05,
"loss": 2.2609,
"step": 1320
},
{
"epoch": 0.01188093878303571,
"grad_norm": 3.907634735107422,
"learning_rate": 4.933316493120015e-05,
"loss": 2.3196,
"step": 1321
},
{
"epoch": 0.011889932680676162,
"grad_norm": 1.322037935256958,
"learning_rate": 4.933143953377236e-05,
"loss": 2.2876,
"step": 1322
},
{
"epoch": 0.011898926578316612,
"grad_norm": 2.2371411323547363,
"learning_rate": 4.932971193729528e-05,
"loss": 2.2772,
"step": 1323
},
{
"epoch": 0.011907920475957063,
"grad_norm": 1.2199677228927612,
"learning_rate": 4.9327982141925036e-05,
"loss": 2.3,
"step": 1324
},
{
"epoch": 0.011916914373597515,
"grad_norm": 1.5293182134628296,
"learning_rate": 4.9326250147817975e-05,
"loss": 2.3158,
"step": 1325
},
{
"epoch": 0.011925908271237965,
"grad_norm": 0.74586421251297,
"learning_rate": 4.932451595513062e-05,
"loss": 2.3088,
"step": 1326
},
{
"epoch": 0.011934902168878415,
"grad_norm": 1.2650678157806396,
"learning_rate": 4.932277956401972e-05,
"loss": 2.329,
"step": 1327
},
{
"epoch": 0.011943896066518868,
"grad_norm": 1.5978035926818848,
"learning_rate": 4.93210409746422e-05,
"loss": 2.2952,
"step": 1328
},
{
"epoch": 0.011952889964159318,
"grad_norm": 1.2295321226119995,
"learning_rate": 4.931930018715519e-05,
"loss": 2.2475,
"step": 1329
},
{
"epoch": 0.011961883861799768,
"grad_norm": 1.0471268892288208,
"learning_rate": 4.931755720171603e-05,
"loss": 2.2794,
"step": 1330
},
{
"epoch": 0.01197087775944022,
"grad_norm": 2.31353497505188,
"learning_rate": 4.931581201848224e-05,
"loss": 2.356,
"step": 1331
},
{
"epoch": 0.01197987165708067,
"grad_norm": 1.8516645431518555,
"learning_rate": 4.931406463761154e-05,
"loss": 2.3354,
"step": 1332
},
{
"epoch": 0.011988865554721121,
"grad_norm": 1.6288126707077026,
"learning_rate": 4.931231505926188e-05,
"loss": 2.2762,
"step": 1333
},
{
"epoch": 0.011997859452361573,
"grad_norm": 1.9435045719146729,
"learning_rate": 4.931056328359137e-05,
"loss": 2.3321,
"step": 1334
},
{
"epoch": 0.012006853350002024,
"grad_norm": 1.2276501655578613,
"learning_rate": 4.930880931075834e-05,
"loss": 2.3042,
"step": 1335
},
{
"epoch": 0.012015847247642474,
"grad_norm": 1.3298304080963135,
"learning_rate": 4.93070531409213e-05,
"loss": 2.3203,
"step": 1336
},
{
"epoch": 0.012024841145282926,
"grad_norm": 0.9763184785842896,
"learning_rate": 4.930529477423898e-05,
"loss": 2.243,
"step": 1337
},
{
"epoch": 0.012033835042923376,
"grad_norm": 1.9828107357025146,
"learning_rate": 4.9303534210870295e-05,
"loss": 2.3426,
"step": 1338
},
{
"epoch": 0.012042828940563827,
"grad_norm": 2.406665802001953,
"learning_rate": 4.930177145097438e-05,
"loss": 2.3129,
"step": 1339
},
{
"epoch": 0.012051822838204279,
"grad_norm": 1.736283302307129,
"learning_rate": 4.9300006494710516e-05,
"loss": 2.266,
"step": 1340
},
{
"epoch": 0.01206081673584473,
"grad_norm": 1.6486119031906128,
"learning_rate": 4.929823934223825e-05,
"loss": 2.2789,
"step": 1341
},
{
"epoch": 0.01206981063348518,
"grad_norm": 2.0691566467285156,
"learning_rate": 4.929646999371729e-05,
"loss": 2.3114,
"step": 1342
},
{
"epoch": 0.012078804531125632,
"grad_norm": 1.2073266506195068,
"learning_rate": 4.929469844930753e-05,
"loss": 2.344,
"step": 1343
},
{
"epoch": 0.012087798428766082,
"grad_norm": 1.0628546476364136,
"learning_rate": 4.9292924709169095e-05,
"loss": 2.305,
"step": 1344
},
{
"epoch": 0.012096792326406532,
"grad_norm": 0.9728950262069702,
"learning_rate": 4.929114877346229e-05,
"loss": 2.265,
"step": 1345
},
{
"epoch": 0.012105786224046984,
"grad_norm": 1.0010732412338257,
"learning_rate": 4.928937064234763e-05,
"loss": 2.2663,
"step": 1346
},
{
"epoch": 0.012114780121687435,
"grad_norm": 1.1296560764312744,
"learning_rate": 4.928759031598581e-05,
"loss": 2.371,
"step": 1347
},
{
"epoch": 0.012123774019327887,
"grad_norm": 1.5296412706375122,
"learning_rate": 4.928580779453774e-05,
"loss": 2.2256,
"step": 1348
},
{
"epoch": 0.012132767916968337,
"grad_norm": 1.653023362159729,
"learning_rate": 4.928402307816451e-05,
"loss": 2.3913,
"step": 1349
},
{
"epoch": 0.012141761814608788,
"grad_norm": 3.8353521823883057,
"learning_rate": 4.928223616702744e-05,
"loss": 2.4941,
"step": 1350
},
{
"epoch": 0.01215075571224924,
"grad_norm": 2.857924699783325,
"learning_rate": 4.928044706128803e-05,
"loss": 2.421,
"step": 1351
},
{
"epoch": 0.01215974960988969,
"grad_norm": 2.749598979949951,
"learning_rate": 4.927865576110795e-05,
"loss": 2.348,
"step": 1352
},
{
"epoch": 0.01216874350753014,
"grad_norm": 1.8370927572250366,
"learning_rate": 4.9276862266649114e-05,
"loss": 2.3351,
"step": 1353
},
{
"epoch": 0.012177737405170592,
"grad_norm": 0.851827085018158,
"learning_rate": 4.9275066578073626e-05,
"loss": 2.3425,
"step": 1354
},
{
"epoch": 0.012186731302811043,
"grad_norm": 0.9431597590446472,
"learning_rate": 4.927326869554376e-05,
"loss": 2.3447,
"step": 1355
},
{
"epoch": 0.012195725200451493,
"grad_norm": 1.717712640762329,
"learning_rate": 4.9271468619222016e-05,
"loss": 2.2972,
"step": 1356
},
{
"epoch": 0.012204719098091945,
"grad_norm": 2.3689746856689453,
"learning_rate": 4.926966634927108e-05,
"loss": 2.2116,
"step": 1357
},
{
"epoch": 0.012213712995732396,
"grad_norm": 1.7761175632476807,
"learning_rate": 4.926786188585384e-05,
"loss": 2.3788,
"step": 1358
},
{
"epoch": 0.012222706893372846,
"grad_norm": 1.5584611892700195,
"learning_rate": 4.9266055229133386e-05,
"loss": 2.3209,
"step": 1359
},
{
"epoch": 0.012231700791013298,
"grad_norm": 1.228891134262085,
"learning_rate": 4.9264246379272996e-05,
"loss": 2.3825,
"step": 1360
},
{
"epoch": 0.012240694688653748,
"grad_norm": 4.562474727630615,
"learning_rate": 4.9262435336436154e-05,
"loss": 2.2612,
"step": 1361
},
{
"epoch": 0.012249688586294199,
"grad_norm": 1.772830605506897,
"learning_rate": 4.9260622100786534e-05,
"loss": 2.2911,
"step": 1362
},
{
"epoch": 0.012258682483934651,
"grad_norm": 2.08131742477417,
"learning_rate": 4.9258806672488025e-05,
"loss": 2.3708,
"step": 1363
},
{
"epoch": 0.012267676381575101,
"grad_norm": 1.5537928342819214,
"learning_rate": 4.9256989051704696e-05,
"loss": 2.2242,
"step": 1364
},
{
"epoch": 0.012276670279215552,
"grad_norm": 3.594076633453369,
"learning_rate": 4.925516923860083e-05,
"loss": 2.3497,
"step": 1365
},
{
"epoch": 0.012285664176856004,
"grad_norm": 1.2255253791809082,
"learning_rate": 4.925334723334089e-05,
"loss": 2.3622,
"step": 1366
},
{
"epoch": 0.012294658074496454,
"grad_norm": 1.7830429077148438,
"learning_rate": 4.925152303608955e-05,
"loss": 2.321,
"step": 1367
},
{
"epoch": 0.012303651972136904,
"grad_norm": 0.9183161854743958,
"learning_rate": 4.924969664701168e-05,
"loss": 2.288,
"step": 1368
},
{
"epoch": 0.012312645869777357,
"grad_norm": 0.824897289276123,
"learning_rate": 4.9247868066272345e-05,
"loss": 2.3118,
"step": 1369
},
{
"epoch": 0.012321639767417807,
"grad_norm": 0.9935973286628723,
"learning_rate": 4.924603729403681e-05,
"loss": 2.2862,
"step": 1370
},
{
"epoch": 0.012330633665058257,
"grad_norm": 1.8484669923782349,
"learning_rate": 4.9244204330470544e-05,
"loss": 2.3656,
"step": 1371
},
{
"epoch": 0.01233962756269871,
"grad_norm": 1.6523222923278809,
"learning_rate": 4.924236917573921e-05,
"loss": 2.3523,
"step": 1372
},
{
"epoch": 0.01234862146033916,
"grad_norm": 1.3153334856033325,
"learning_rate": 4.924053183000865e-05,
"loss": 2.3487,
"step": 1373
},
{
"epoch": 0.01235761535797961,
"grad_norm": 2.817600727081299,
"learning_rate": 4.923869229344494e-05,
"loss": 2.3055,
"step": 1374
},
{
"epoch": 0.012366609255620062,
"grad_norm": 1.8938976526260376,
"learning_rate": 4.923685056621432e-05,
"loss": 2.3513,
"step": 1375
},
{
"epoch": 0.012375603153260513,
"grad_norm": 1.8110777139663696,
"learning_rate": 4.923500664848326e-05,
"loss": 2.3197,
"step": 1376
},
{
"epoch": 0.012384597050900963,
"grad_norm": 1.3235276937484741,
"learning_rate": 4.92331605404184e-05,
"loss": 2.3093,
"step": 1377
},
{
"epoch": 0.012393590948541415,
"grad_norm": 1.1013685464859009,
"learning_rate": 4.92313122421866e-05,
"loss": 2.3095,
"step": 1378
},
{
"epoch": 0.012402584846181865,
"grad_norm": 3.155381679534912,
"learning_rate": 4.922946175395489e-05,
"loss": 2.2927,
"step": 1379
},
{
"epoch": 0.012411578743822316,
"grad_norm": 1.7338447570800781,
"learning_rate": 4.922760907589052e-05,
"loss": 2.3123,
"step": 1380
},
{
"epoch": 0.012420572641462768,
"grad_norm": 1.9885655641555786,
"learning_rate": 4.922575420816095e-05,
"loss": 2.374,
"step": 1381
},
{
"epoch": 0.012429566539103218,
"grad_norm": 1.0699632167816162,
"learning_rate": 4.9223897150933804e-05,
"loss": 2.3657,
"step": 1382
},
{
"epoch": 0.012438560436743669,
"grad_norm": 2.571831703186035,
"learning_rate": 4.922203790437693e-05,
"loss": 2.3649,
"step": 1383
},
{
"epoch": 0.01244755433438412,
"grad_norm": 2.8510687351226807,
"learning_rate": 4.922017646865835e-05,
"loss": 2.3133,
"step": 1384
},
{
"epoch": 0.012456548232024571,
"grad_norm": 2.8858659267425537,
"learning_rate": 4.9218312843946314e-05,
"loss": 2.3696,
"step": 1385
},
{
"epoch": 0.012465542129665023,
"grad_norm": 2.153928518295288,
"learning_rate": 4.921644703040925e-05,
"loss": 2.2591,
"step": 1386
},
{
"epoch": 0.012474536027305473,
"grad_norm": 1.840311884880066,
"learning_rate": 4.9214579028215776e-05,
"loss": 2.3336,
"step": 1387
},
{
"epoch": 0.012483529924945924,
"grad_norm": 1.258995532989502,
"learning_rate": 4.9212708837534746e-05,
"loss": 2.2697,
"step": 1388
},
{
"epoch": 0.012492523822586376,
"grad_norm": 1.5013561248779297,
"learning_rate": 4.9210836458535165e-05,
"loss": 2.3315,
"step": 1389
},
{
"epoch": 0.012501517720226826,
"grad_norm": 2.9934911727905273,
"learning_rate": 4.9208961891386265e-05,
"loss": 2.3513,
"step": 1390
},
{
"epoch": 0.012510511617867277,
"grad_norm": 1.2505568265914917,
"learning_rate": 4.920708513625746e-05,
"loss": 2.304,
"step": 1391
},
{
"epoch": 0.012519505515507729,
"grad_norm": 3.9782907962799072,
"learning_rate": 4.9205206193318374e-05,
"loss": 2.3204,
"step": 1392
},
{
"epoch": 0.012528499413148179,
"grad_norm": 4.457254409790039,
"learning_rate": 4.920332506273883e-05,
"loss": 2.2693,
"step": 1393
},
{
"epoch": 0.01253749331078863,
"grad_norm": 1.9486520290374756,
"learning_rate": 4.9201441744688835e-05,
"loss": 2.298,
"step": 1394
},
{
"epoch": 0.012546487208429082,
"grad_norm": 1.8635858297348022,
"learning_rate": 4.91995562393386e-05,
"loss": 2.296,
"step": 1395
},
{
"epoch": 0.012555481106069532,
"grad_norm": 1.4830387830734253,
"learning_rate": 4.9197668546858536e-05,
"loss": 2.3269,
"step": 1396
},
{
"epoch": 0.012564475003709982,
"grad_norm": 0.9352790117263794,
"learning_rate": 4.919577866741926e-05,
"loss": 2.288,
"step": 1397
},
{
"epoch": 0.012573468901350434,
"grad_norm": 1.602815866470337,
"learning_rate": 4.919388660119156e-05,
"loss": 2.2405,
"step": 1398
},
{
"epoch": 0.012582462798990885,
"grad_norm": 1.2547131776809692,
"learning_rate": 4.919199234834646e-05,
"loss": 2.2738,
"step": 1399
},
{
"epoch": 0.012591456696631335,
"grad_norm": 1.011612057685852,
"learning_rate": 4.919009590905514e-05,
"loss": 2.3585,
"step": 1400
},
{
"epoch": 0.012600450594271787,
"grad_norm": 0.7901346683502197,
"learning_rate": 4.9188197283489015e-05,
"loss": 2.2434,
"step": 1401
},
{
"epoch": 0.012609444491912238,
"grad_norm": 1.0523431301116943,
"learning_rate": 4.918629647181966e-05,
"loss": 2.2917,
"step": 1402
},
{
"epoch": 0.012618438389552688,
"grad_norm": 1.2329866886138916,
"learning_rate": 4.9184393474218894e-05,
"loss": 2.3502,
"step": 1403
},
{
"epoch": 0.01262743228719314,
"grad_norm": 1.7527052164077759,
"learning_rate": 4.9182488290858686e-05,
"loss": 2.3325,
"step": 1404
},
{
"epoch": 0.01263642618483359,
"grad_norm": 0.9112148284912109,
"learning_rate": 4.9180580921911236e-05,
"loss": 2.3672,
"step": 1405
},
{
"epoch": 0.01264542008247404,
"grad_norm": 1.489206314086914,
"learning_rate": 4.917867136754893e-05,
"loss": 2.2949,
"step": 1406
},
{
"epoch": 0.012654413980114493,
"grad_norm": 2.510047674179077,
"learning_rate": 4.9176759627944346e-05,
"loss": 2.2474,
"step": 1407
},
{
"epoch": 0.012663407877754943,
"grad_norm": 1.6785236597061157,
"learning_rate": 4.917484570327027e-05,
"loss": 2.3413,
"step": 1408
},
{
"epoch": 0.012672401775395394,
"grad_norm": 0.9878116846084595,
"learning_rate": 4.917292959369968e-05,
"loss": 2.2201,
"step": 1409
},
{
"epoch": 0.012681395673035846,
"grad_norm": 1.2093920707702637,
"learning_rate": 4.917101129940575e-05,
"loss": 2.3082,
"step": 1410
},
{
"epoch": 0.012690389570676296,
"grad_norm": 1.625248670578003,
"learning_rate": 4.916909082056185e-05,
"loss": 2.3316,
"step": 1411
},
{
"epoch": 0.012699383468316746,
"grad_norm": 1.675307035446167,
"learning_rate": 4.916716815734156e-05,
"loss": 2.355,
"step": 1412
},
{
"epoch": 0.012708377365957198,
"grad_norm": 1.44845449924469,
"learning_rate": 4.9165243309918634e-05,
"loss": 2.3005,
"step": 1413
},
{
"epoch": 0.012717371263597649,
"grad_norm": 1.645935297012329,
"learning_rate": 4.9163316278467056e-05,
"loss": 2.3009,
"step": 1414
},
{
"epoch": 0.0127263651612381,
"grad_norm": 1.3288215398788452,
"learning_rate": 4.916138706316097e-05,
"loss": 2.3164,
"step": 1415
},
{
"epoch": 0.012735359058878551,
"grad_norm": 2.2171990871429443,
"learning_rate": 4.9159455664174756e-05,
"loss": 2.3462,
"step": 1416
},
{
"epoch": 0.012744352956519002,
"grad_norm": 1.350297451019287,
"learning_rate": 4.915752208168295e-05,
"loss": 2.363,
"step": 1417
},
{
"epoch": 0.012753346854159452,
"grad_norm": 0.9702437520027161,
"learning_rate": 4.915558631586032e-05,
"loss": 2.2286,
"step": 1418
},
{
"epoch": 0.012762340751799904,
"grad_norm": 1.020673155784607,
"learning_rate": 4.9153648366881826e-05,
"loss": 2.2715,
"step": 1419
},
{
"epoch": 0.012771334649440354,
"grad_norm": 0.8472902178764343,
"learning_rate": 4.9151708234922603e-05,
"loss": 2.2092,
"step": 1420
},
{
"epoch": 0.012780328547080807,
"grad_norm": 1.295495867729187,
"learning_rate": 4.9149765920158005e-05,
"loss": 2.3591,
"step": 1421
},
{
"epoch": 0.012789322444721257,
"grad_norm": 1.7508115768432617,
"learning_rate": 4.914782142276357e-05,
"loss": 2.3541,
"step": 1422
},
{
"epoch": 0.012798316342361707,
"grad_norm": 1.0777674913406372,
"learning_rate": 4.9145874742915056e-05,
"loss": 2.3678,
"step": 1423
},
{
"epoch": 0.01280731024000216,
"grad_norm": 1.720782995223999,
"learning_rate": 4.9143925880788386e-05,
"loss": 2.3476,
"step": 1424
},
{
"epoch": 0.01281630413764261,
"grad_norm": 2.026348829269409,
"learning_rate": 4.9141974836559704e-05,
"loss": 2.343,
"step": 1425
},
{
"epoch": 0.01282529803528306,
"grad_norm": 1.1215814352035522,
"learning_rate": 4.9140021610405326e-05,
"loss": 2.2349,
"step": 1426
},
{
"epoch": 0.012834291932923512,
"grad_norm": 0.973241925239563,
"learning_rate": 4.913806620250181e-05,
"loss": 2.2499,
"step": 1427
},
{
"epoch": 0.012843285830563963,
"grad_norm": 1.0042587518692017,
"learning_rate": 4.9136108613025866e-05,
"loss": 2.2721,
"step": 1428
},
{
"epoch": 0.012852279728204413,
"grad_norm": 1.5908631086349487,
"learning_rate": 4.913414884215442e-05,
"loss": 2.2816,
"step": 1429
},
{
"epoch": 0.012861273625844865,
"grad_norm": 1.480427622795105,
"learning_rate": 4.9132186890064605e-05,
"loss": 2.3384,
"step": 1430
},
{
"epoch": 0.012870267523485315,
"grad_norm": 1.840626835823059,
"learning_rate": 4.913022275693372e-05,
"loss": 2.3776,
"step": 1431
},
{
"epoch": 0.012879261421125766,
"grad_norm": 2.2271974086761475,
"learning_rate": 4.91282564429393e-05,
"loss": 2.2949,
"step": 1432
},
{
"epoch": 0.012888255318766218,
"grad_norm": 1.788423776626587,
"learning_rate": 4.9126287948259045e-05,
"loss": 2.3391,
"step": 1433
},
{
"epoch": 0.012897249216406668,
"grad_norm": 2.0390164852142334,
"learning_rate": 4.912431727307088e-05,
"loss": 2.3828,
"step": 1434
},
{
"epoch": 0.012906243114047119,
"grad_norm": 1.7438021898269653,
"learning_rate": 4.9122344417552904e-05,
"loss": 2.3585,
"step": 1435
},
{
"epoch": 0.01291523701168757,
"grad_norm": 1.5729365348815918,
"learning_rate": 4.912036938188342e-05,
"loss": 2.2889,
"step": 1436
},
{
"epoch": 0.012924230909328021,
"grad_norm": 1.4151298999786377,
"learning_rate": 4.911839216624093e-05,
"loss": 2.2024,
"step": 1437
},
{
"epoch": 0.012933224806968471,
"grad_norm": 1.3459993600845337,
"learning_rate": 4.911641277080413e-05,
"loss": 2.257,
"step": 1438
},
{
"epoch": 0.012942218704608923,
"grad_norm": 1.3020594120025635,
"learning_rate": 4.911443119575193e-05,
"loss": 2.3676,
"step": 1439
},
{
"epoch": 0.012951212602249374,
"grad_norm": 0.9443081021308899,
"learning_rate": 4.911244744126341e-05,
"loss": 2.2889,
"step": 1440
},
{
"epoch": 0.012960206499889824,
"grad_norm": 1.1419777870178223,
"learning_rate": 4.911046150751786e-05,
"loss": 2.3973,
"step": 1441
},
{
"epoch": 0.012969200397530276,
"grad_norm": 0.7631124258041382,
"learning_rate": 4.910847339469477e-05,
"loss": 2.1734,
"step": 1442
},
{
"epoch": 0.012978194295170727,
"grad_norm": 1.3308863639831543,
"learning_rate": 4.9106483102973824e-05,
"loss": 2.3016,
"step": 1443
},
{
"epoch": 0.012987188192811177,
"grad_norm": 0.6941406726837158,
"learning_rate": 4.9104490632534894e-05,
"loss": 2.2565,
"step": 1444
},
{
"epoch": 0.012996182090451629,
"grad_norm": 1.0932955741882324,
"learning_rate": 4.910249598355806e-05,
"loss": 2.2987,
"step": 1445
},
{
"epoch": 0.01300517598809208,
"grad_norm": 0.8946712613105774,
"learning_rate": 4.910049915622361e-05,
"loss": 2.2626,
"step": 1446
},
{
"epoch": 0.01301416988573253,
"grad_norm": 1.3275916576385498,
"learning_rate": 4.9098500150712e-05,
"loss": 2.3549,
"step": 1447
},
{
"epoch": 0.013023163783372982,
"grad_norm": 0.7752215266227722,
"learning_rate": 4.9096498967203905e-05,
"loss": 2.241,
"step": 1448
},
{
"epoch": 0.013032157681013432,
"grad_norm": 1.0660349130630493,
"learning_rate": 4.909449560588019e-05,
"loss": 2.291,
"step": 1449
},
{
"epoch": 0.013041151578653883,
"grad_norm": 1.3548780679702759,
"learning_rate": 4.9092490066921916e-05,
"loss": 2.3213,
"step": 1450
},
{
"epoch": 0.013050145476294335,
"grad_norm": 1.3699947595596313,
"learning_rate": 4.9090482350510336e-05,
"loss": 2.2725,
"step": 1451
},
{
"epoch": 0.013059139373934785,
"grad_norm": 1.3277685642242432,
"learning_rate": 4.9088472456826904e-05,
"loss": 2.2799,
"step": 1452
},
{
"epoch": 0.013068133271575235,
"grad_norm": 1.296507716178894,
"learning_rate": 4.908646038605329e-05,
"loss": 2.3449,
"step": 1453
},
{
"epoch": 0.013077127169215687,
"grad_norm": 1.3192994594573975,
"learning_rate": 4.908444613837132e-05,
"loss": 2.2762,
"step": 1454
},
{
"epoch": 0.013086121066856138,
"grad_norm": 1.0131722688674927,
"learning_rate": 4.908242971396305e-05,
"loss": 2.2564,
"step": 1455
},
{
"epoch": 0.013095114964496588,
"grad_norm": 1.6245476007461548,
"learning_rate": 4.908041111301073e-05,
"loss": 2.3634,
"step": 1456
},
{
"epoch": 0.01310410886213704,
"grad_norm": 1.0569870471954346,
"learning_rate": 4.907839033569679e-05,
"loss": 2.3689,
"step": 1457
},
{
"epoch": 0.01311310275977749,
"grad_norm": 0.6416109800338745,
"learning_rate": 4.907636738220386e-05,
"loss": 2.2693,
"step": 1458
},
{
"epoch": 0.013122096657417943,
"grad_norm": 0.8165781497955322,
"learning_rate": 4.907434225271478e-05,
"loss": 2.2533,
"step": 1459
},
{
"epoch": 0.013131090555058393,
"grad_norm": 1.4047127962112427,
"learning_rate": 4.907231494741258e-05,
"loss": 2.3631,
"step": 1460
},
{
"epoch": 0.013140084452698843,
"grad_norm": 1.3052713871002197,
"learning_rate": 4.907028546648048e-05,
"loss": 2.2588,
"step": 1461
},
{
"epoch": 0.013149078350339296,
"grad_norm": 1.8216019868850708,
"learning_rate": 4.9068253810101915e-05,
"loss": 2.3131,
"step": 1462
},
{
"epoch": 0.013158072247979746,
"grad_norm": 1.5135239362716675,
"learning_rate": 4.906621997846049e-05,
"loss": 2.3277,
"step": 1463
},
{
"epoch": 0.013167066145620196,
"grad_norm": 0.896512508392334,
"learning_rate": 4.906418397174002e-05,
"loss": 2.3493,
"step": 1464
},
{
"epoch": 0.013176060043260648,
"grad_norm": 7.013914585113525,
"learning_rate": 4.9062145790124526e-05,
"loss": 2.2922,
"step": 1465
},
{
"epoch": 0.013185053940901099,
"grad_norm": 1.9243741035461426,
"learning_rate": 4.9060105433798215e-05,
"loss": 2.3571,
"step": 1466
},
{
"epoch": 0.013194047838541549,
"grad_norm": 1.5016390085220337,
"learning_rate": 4.905806290294549e-05,
"loss": 2.3823,
"step": 1467
},
{
"epoch": 0.013203041736182001,
"grad_norm": 1.5999473333358765,
"learning_rate": 4.905601819775095e-05,
"loss": 2.3691,
"step": 1468
},
{
"epoch": 0.013212035633822452,
"grad_norm": 1.295243263244629,
"learning_rate": 4.90539713183994e-05,
"loss": 2.2992,
"step": 1469
},
{
"epoch": 0.013221029531462902,
"grad_norm": 2.173689842224121,
"learning_rate": 4.9051922265075834e-05,
"loss": 2.2334,
"step": 1470
},
{
"epoch": 0.013230023429103354,
"grad_norm": 1.455386757850647,
"learning_rate": 4.904987103796543e-05,
"loss": 2.2295,
"step": 1471
},
{
"epoch": 0.013239017326743804,
"grad_norm": 1.8496041297912598,
"learning_rate": 4.904781763725359e-05,
"loss": 2.3184,
"step": 1472
},
{
"epoch": 0.013248011224384255,
"grad_norm": 0.9949283003807068,
"learning_rate": 4.90457620631259e-05,
"loss": 2.2462,
"step": 1473
},
{
"epoch": 0.013257005122024707,
"grad_norm": 5.564322471618652,
"learning_rate": 4.904370431576813e-05,
"loss": 2.395,
"step": 1474
},
{
"epoch": 0.013265999019665157,
"grad_norm": 1.4406638145446777,
"learning_rate": 4.904164439536626e-05,
"loss": 2.3453,
"step": 1475
},
{
"epoch": 0.013274992917305608,
"grad_norm": 1.9769619703292847,
"learning_rate": 4.9039582302106465e-05,
"loss": 2.2512,
"step": 1476
},
{
"epoch": 0.01328398681494606,
"grad_norm": 2.204035520553589,
"learning_rate": 4.9037518036175126e-05,
"loss": 2.3678,
"step": 1477
},
{
"epoch": 0.01329298071258651,
"grad_norm": 0.815660834312439,
"learning_rate": 4.903545159775879e-05,
"loss": 2.2468,
"step": 1478
},
{
"epoch": 0.01330197461022696,
"grad_norm": 5.241702556610107,
"learning_rate": 4.903338298704422e-05,
"loss": 2.3255,
"step": 1479
},
{
"epoch": 0.013310968507867412,
"grad_norm": 1.594947099685669,
"learning_rate": 4.9031312204218396e-05,
"loss": 2.3213,
"step": 1480
},
{
"epoch": 0.013319962405507863,
"grad_norm": 2.852376937866211,
"learning_rate": 4.902923924946844e-05,
"loss": 2.3545,
"step": 1481
},
{
"epoch": 0.013328956303148313,
"grad_norm": 1.9794753789901733,
"learning_rate": 4.902716412298174e-05,
"loss": 2.296,
"step": 1482
},
{
"epoch": 0.013337950200788765,
"grad_norm": 1.202060341835022,
"learning_rate": 4.902508682494582e-05,
"loss": 2.3212,
"step": 1483
},
{
"epoch": 0.013346944098429216,
"grad_norm": 1.2876771688461304,
"learning_rate": 4.9023007355548436e-05,
"loss": 2.2472,
"step": 1484
},
{
"epoch": 0.013355937996069666,
"grad_norm": 8.125530242919922,
"learning_rate": 4.9020925714977514e-05,
"loss": 2.4545,
"step": 1485
},
{
"epoch": 0.013364931893710118,
"grad_norm": 2.8215906620025635,
"learning_rate": 4.901884190342121e-05,
"loss": 2.3704,
"step": 1486
},
{
"epoch": 0.013373925791350568,
"grad_norm": 1.6175618171691895,
"learning_rate": 4.901675592106783e-05,
"loss": 2.3454,
"step": 1487
},
{
"epoch": 0.013382919688991019,
"grad_norm": 1.2598589658737183,
"learning_rate": 4.9014667768105936e-05,
"loss": 2.3632,
"step": 1488
},
{
"epoch": 0.013391913586631471,
"grad_norm": 1.0440917015075684,
"learning_rate": 4.901257744472423e-05,
"loss": 2.3011,
"step": 1489
},
{
"epoch": 0.013400907484271921,
"grad_norm": 0.7207644581794739,
"learning_rate": 4.901048495111164e-05,
"loss": 2.2114,
"step": 1490
},
{
"epoch": 0.013409901381912372,
"grad_norm": 1.503522515296936,
"learning_rate": 4.9008390287457276e-05,
"loss": 2.2791,
"step": 1491
},
{
"epoch": 0.013418895279552824,
"grad_norm": 1.039432406425476,
"learning_rate": 4.9006293453950455e-05,
"loss": 2.2587,
"step": 1492
},
{
"epoch": 0.013427889177193274,
"grad_norm": 1.23433518409729,
"learning_rate": 4.900419445078069e-05,
"loss": 2.322,
"step": 1493
},
{
"epoch": 0.013436883074833726,
"grad_norm": 1.8497687578201294,
"learning_rate": 4.900209327813769e-05,
"loss": 2.2566,
"step": 1494
},
{
"epoch": 0.013445876972474177,
"grad_norm": 1.0818395614624023,
"learning_rate": 4.8999989936211356e-05,
"loss": 2.2234,
"step": 1495
},
{
"epoch": 0.013454870870114627,
"grad_norm": 1.1293376684188843,
"learning_rate": 4.899788442519177e-05,
"loss": 2.2535,
"step": 1496
},
{
"epoch": 0.013463864767755079,
"grad_norm": 1.1049357652664185,
"learning_rate": 4.8995776745269254e-05,
"loss": 2.235,
"step": 1497
},
{
"epoch": 0.01347285866539553,
"grad_norm": 1.1415754556655884,
"learning_rate": 4.8993666896634274e-05,
"loss": 2.3691,
"step": 1498
},
{
"epoch": 0.01348185256303598,
"grad_norm": 1.2333590984344482,
"learning_rate": 4.899155487947752e-05,
"loss": 2.2423,
"step": 1499
},
{
"epoch": 0.013490846460676432,
"grad_norm": 1.201111078262329,
"learning_rate": 4.898944069398989e-05,
"loss": 2.3714,
"step": 1500
},
{
"epoch": 0.013499840358316882,
"grad_norm": 1.2897664308547974,
"learning_rate": 4.898732434036244e-05,
"loss": 2.322,
"step": 1501
},
{
"epoch": 0.013508834255957333,
"grad_norm": 1.4544669389724731,
"learning_rate": 4.898520581878645e-05,
"loss": 2.3104,
"step": 1502
},
{
"epoch": 0.013517828153597785,
"grad_norm": 1.377790093421936,
"learning_rate": 4.89830851294534e-05,
"loss": 2.268,
"step": 1503
},
{
"epoch": 0.013526822051238235,
"grad_norm": 2.3503384590148926,
"learning_rate": 4.898096227255494e-05,
"loss": 2.3085,
"step": 1504
},
{
"epoch": 0.013535815948878685,
"grad_norm": 1.3760664463043213,
"learning_rate": 4.897883724828295e-05,
"loss": 2.2564,
"step": 1505
},
{
"epoch": 0.013544809846519137,
"grad_norm": 1.2119131088256836,
"learning_rate": 4.8976710056829476e-05,
"loss": 2.3232,
"step": 1506
},
{
"epoch": 0.013553803744159588,
"grad_norm": 0.9780412912368774,
"learning_rate": 4.8974580698386765e-05,
"loss": 2.2853,
"step": 1507
},
{
"epoch": 0.013562797641800038,
"grad_norm": 1.0036619901657104,
"learning_rate": 4.8972449173147276e-05,
"loss": 2.2166,
"step": 1508
},
{
"epoch": 0.01357179153944049,
"grad_norm": 0.927261471748352,
"learning_rate": 4.897031548130365e-05,
"loss": 2.3161,
"step": 1509
},
{
"epoch": 0.01358078543708094,
"grad_norm": 1.3687413930892944,
"learning_rate": 4.896817962304874e-05,
"loss": 2.3037,
"step": 1510
},
{
"epoch": 0.013589779334721391,
"grad_norm": 0.7189801931381226,
"learning_rate": 4.896604159857556e-05,
"loss": 2.3202,
"step": 1511
},
{
"epoch": 0.013598773232361843,
"grad_norm": 0.8859209418296814,
"learning_rate": 4.8963901408077364e-05,
"loss": 2.2335,
"step": 1512
},
{
"epoch": 0.013607767130002293,
"grad_norm": 1.1698659658432007,
"learning_rate": 4.896175905174757e-05,
"loss": 2.3372,
"step": 1513
},
{
"epoch": 0.013616761027642744,
"grad_norm": 1.1946779489517212,
"learning_rate": 4.8959614529779796e-05,
"loss": 2.2648,
"step": 1514
},
{
"epoch": 0.013625754925283196,
"grad_norm": 1.2864960432052612,
"learning_rate": 4.895746784236787e-05,
"loss": 2.34,
"step": 1515
},
{
"epoch": 0.013634748822923646,
"grad_norm": 1.3298262357711792,
"learning_rate": 4.895531898970581e-05,
"loss": 2.2857,
"step": 1516
},
{
"epoch": 0.013643742720564097,
"grad_norm": 1.0850532054901123,
"learning_rate": 4.895316797198781e-05,
"loss": 2.3234,
"step": 1517
},
{
"epoch": 0.013652736618204549,
"grad_norm": 1.1582186222076416,
"learning_rate": 4.89510147894083e-05,
"loss": 2.2708,
"step": 1518
},
{
"epoch": 0.013661730515844999,
"grad_norm": 0.9520682692527771,
"learning_rate": 4.8948859442161874e-05,
"loss": 2.3413,
"step": 1519
},
{
"epoch": 0.01367072441348545,
"grad_norm": 1.1450742483139038,
"learning_rate": 4.894670193044332e-05,
"loss": 2.3047,
"step": 1520
},
{
"epoch": 0.013679718311125902,
"grad_norm": 1.4632195234298706,
"learning_rate": 4.894454225444764e-05,
"loss": 2.325,
"step": 1521
},
{
"epoch": 0.013688712208766352,
"grad_norm": 1.5279765129089355,
"learning_rate": 4.894238041437003e-05,
"loss": 2.3494,
"step": 1522
},
{
"epoch": 0.013697706106406802,
"grad_norm": 1.4543125629425049,
"learning_rate": 4.894021641040585e-05,
"loss": 2.3153,
"step": 1523
},
{
"epoch": 0.013706700004047254,
"grad_norm": 1.0561600923538208,
"learning_rate": 4.893805024275071e-05,
"loss": 2.3385,
"step": 1524
},
{
"epoch": 0.013715693901687705,
"grad_norm": 1.064774751663208,
"learning_rate": 4.8935881911600366e-05,
"loss": 2.2873,
"step": 1525
},
{
"epoch": 0.013724687799328155,
"grad_norm": 1.2532635927200317,
"learning_rate": 4.89337114171508e-05,
"loss": 2.3004,
"step": 1526
},
{
"epoch": 0.013733681696968607,
"grad_norm": 2.683222770690918,
"learning_rate": 4.893153875959817e-05,
"loss": 2.3142,
"step": 1527
},
{
"epoch": 0.013742675594609058,
"grad_norm": 1.316571831703186,
"learning_rate": 4.8929363939138854e-05,
"loss": 2.3365,
"step": 1528
},
{
"epoch": 0.013751669492249508,
"grad_norm": 1.5955365896224976,
"learning_rate": 4.89271869559694e-05,
"loss": 2.2927,
"step": 1529
},
{
"epoch": 0.01376066338988996,
"grad_norm": 1.0240486860275269,
"learning_rate": 4.8925007810286546e-05,
"loss": 2.2553,
"step": 1530
},
{
"epoch": 0.01376965728753041,
"grad_norm": 0.8262383937835693,
"learning_rate": 4.8922826502287275e-05,
"loss": 2.2031,
"step": 1531
},
{
"epoch": 0.013778651185170862,
"grad_norm": 0.7110381722450256,
"learning_rate": 4.8920643032168703e-05,
"loss": 2.3212,
"step": 1532
},
{
"epoch": 0.013787645082811313,
"grad_norm": 1.495509386062622,
"learning_rate": 4.891845740012818e-05,
"loss": 2.3231,
"step": 1533
},
{
"epoch": 0.013796638980451763,
"grad_norm": 1.0226750373840332,
"learning_rate": 4.891626960636323e-05,
"loss": 2.2767,
"step": 1534
},
{
"epoch": 0.013805632878092215,
"grad_norm": 0.8253571391105652,
"learning_rate": 4.8914079651071606e-05,
"loss": 2.2549,
"step": 1535
},
{
"epoch": 0.013814626775732666,
"grad_norm": 1.2516210079193115,
"learning_rate": 4.891188753445122e-05,
"loss": 2.294,
"step": 1536
},
{
"epoch": 0.013823620673373116,
"grad_norm": 0.818970263004303,
"learning_rate": 4.89096932567002e-05,
"loss": 2.2824,
"step": 1537
},
{
"epoch": 0.013832614571013568,
"grad_norm": 0.9809755086898804,
"learning_rate": 4.890749681801684e-05,
"loss": 2.2402,
"step": 1538
},
{
"epoch": 0.013841608468654018,
"grad_norm": 1.5677928924560547,
"learning_rate": 4.890529821859968e-05,
"loss": 2.3347,
"step": 1539
},
{
"epoch": 0.013850602366294469,
"grad_norm": 0.8905024528503418,
"learning_rate": 4.890309745864741e-05,
"loss": 2.2615,
"step": 1540
},
{
"epoch": 0.01385959626393492,
"grad_norm": 0.9016208052635193,
"learning_rate": 4.8900894538358944e-05,
"loss": 2.2615,
"step": 1541
},
{
"epoch": 0.013868590161575371,
"grad_norm": 1.0534595251083374,
"learning_rate": 4.8898689457933363e-05,
"loss": 2.2869,
"step": 1542
},
{
"epoch": 0.013877584059215822,
"grad_norm": 1.4061229228973389,
"learning_rate": 4.889648221756999e-05,
"loss": 2.3209,
"step": 1543
},
{
"epoch": 0.013886577956856274,
"grad_norm": 1.2702196836471558,
"learning_rate": 4.889427281746828e-05,
"loss": 2.2897,
"step": 1544
},
{
"epoch": 0.013895571854496724,
"grad_norm": 1.3807512521743774,
"learning_rate": 4.8892061257827924e-05,
"loss": 2.2773,
"step": 1545
},
{
"epoch": 0.013904565752137174,
"grad_norm": 0.9990711808204651,
"learning_rate": 4.888984753884882e-05,
"loss": 2.2724,
"step": 1546
},
{
"epoch": 0.013913559649777626,
"grad_norm": 1.0630393028259277,
"learning_rate": 4.888763166073101e-05,
"loss": 2.2663,
"step": 1547
},
{
"epoch": 0.013922553547418077,
"grad_norm": 0.831275224685669,
"learning_rate": 4.8885413623674804e-05,
"loss": 2.3437,
"step": 1548
},
{
"epoch": 0.013931547445058527,
"grad_norm": 1.5377510786056519,
"learning_rate": 4.8883193427880625e-05,
"loss": 2.2938,
"step": 1549
},
{
"epoch": 0.01394054134269898,
"grad_norm": 0.9480425715446472,
"learning_rate": 4.8880971073549154e-05,
"loss": 2.3141,
"step": 1550
},
{
"epoch": 0.01394953524033943,
"grad_norm": 1.0191166400909424,
"learning_rate": 4.887874656088124e-05,
"loss": 2.2693,
"step": 1551
},
{
"epoch": 0.01395852913797988,
"grad_norm": 1.0550453662872314,
"learning_rate": 4.8876519890077926e-05,
"loss": 2.2977,
"step": 1552
},
{
"epoch": 0.013967523035620332,
"grad_norm": 0.9050216674804688,
"learning_rate": 4.887429106134047e-05,
"loss": 2.279,
"step": 1553
},
{
"epoch": 0.013976516933260782,
"grad_norm": 1.4496594667434692,
"learning_rate": 4.88720600748703e-05,
"loss": 2.3405,
"step": 1554
},
{
"epoch": 0.013985510830901233,
"grad_norm": 1.1739283800125122,
"learning_rate": 4.8869826930869045e-05,
"loss": 2.2464,
"step": 1555
},
{
"epoch": 0.013994504728541685,
"grad_norm": 0.8753737807273865,
"learning_rate": 4.886759162953856e-05,
"loss": 2.2698,
"step": 1556
},
{
"epoch": 0.014003498626182135,
"grad_norm": 1.2392778396606445,
"learning_rate": 4.886535417108084e-05,
"loss": 2.2212,
"step": 1557
},
{
"epoch": 0.014012492523822586,
"grad_norm": 0.8766874670982361,
"learning_rate": 4.886311455569811e-05,
"loss": 2.3099,
"step": 1558
},
{
"epoch": 0.014021486421463038,
"grad_norm": 1.513805627822876,
"learning_rate": 4.88608727835928e-05,
"loss": 2.3002,
"step": 1559
},
{
"epoch": 0.014030480319103488,
"grad_norm": 0.7318100929260254,
"learning_rate": 4.88586288549675e-05,
"loss": 2.2333,
"step": 1560
},
{
"epoch": 0.014039474216743938,
"grad_norm": 1.0444992780685425,
"learning_rate": 4.8856382770025034e-05,
"loss": 2.3115,
"step": 1561
},
{
"epoch": 0.01404846811438439,
"grad_norm": 1.5187288522720337,
"learning_rate": 4.885413452896838e-05,
"loss": 2.2705,
"step": 1562
},
{
"epoch": 0.014057462012024841,
"grad_norm": 1.7314192056655884,
"learning_rate": 4.885188413200075e-05,
"loss": 2.2405,
"step": 1563
},
{
"epoch": 0.014066455909665291,
"grad_norm": 1.0305639505386353,
"learning_rate": 4.884963157932552e-05,
"loss": 2.235,
"step": 1564
},
{
"epoch": 0.014075449807305743,
"grad_norm": 1.9407317638397217,
"learning_rate": 4.8847376871146265e-05,
"loss": 2.331,
"step": 1565
},
{
"epoch": 0.014084443704946194,
"grad_norm": 1.0195425748825073,
"learning_rate": 4.884512000766679e-05,
"loss": 2.3955,
"step": 1566
},
{
"epoch": 0.014093437602586644,
"grad_norm": 0.9120759963989258,
"learning_rate": 4.884286098909104e-05,
"loss": 2.2323,
"step": 1567
},
{
"epoch": 0.014102431500227096,
"grad_norm": 1.0566129684448242,
"learning_rate": 4.884059981562321e-05,
"loss": 2.2677,
"step": 1568
},
{
"epoch": 0.014111425397867547,
"grad_norm": 1.2720448970794678,
"learning_rate": 4.883833648746763e-05,
"loss": 2.2322,
"step": 1569
},
{
"epoch": 0.014120419295507999,
"grad_norm": 1.0273852348327637,
"learning_rate": 4.8836071004828886e-05,
"loss": 2.172,
"step": 1570
},
{
"epoch": 0.014129413193148449,
"grad_norm": 1.0984374284744263,
"learning_rate": 4.8833803367911717e-05,
"loss": 2.2965,
"step": 1571
},
{
"epoch": 0.0141384070907889,
"grad_norm": 1.053094506263733,
"learning_rate": 4.883153357692107e-05,
"loss": 2.3454,
"step": 1572
},
{
"epoch": 0.014147400988429351,
"grad_norm": 1.8766710758209229,
"learning_rate": 4.882926163206209e-05,
"loss": 2.2171,
"step": 1573
},
{
"epoch": 0.014156394886069802,
"grad_norm": 1.362017035484314,
"learning_rate": 4.8826987533540114e-05,
"loss": 2.3389,
"step": 1574
},
{
"epoch": 0.014165388783710252,
"grad_norm": 1.2835819721221924,
"learning_rate": 4.882471128156066e-05,
"loss": 2.3169,
"step": 1575
},
{
"epoch": 0.014174382681350704,
"grad_norm": 1.1079405546188354,
"learning_rate": 4.882243287632947e-05,
"loss": 2.3226,
"step": 1576
},
{
"epoch": 0.014183376578991155,
"grad_norm": 1.3148959875106812,
"learning_rate": 4.882015231805245e-05,
"loss": 2.3007,
"step": 1577
},
{
"epoch": 0.014192370476631605,
"grad_norm": 1.1894735097885132,
"learning_rate": 4.881786960693572e-05,
"loss": 2.3892,
"step": 1578
},
{
"epoch": 0.014201364374272057,
"grad_norm": 1.3177655935287476,
"learning_rate": 4.8815584743185593e-05,
"loss": 2.2949,
"step": 1579
},
{
"epoch": 0.014210358271912507,
"grad_norm": 1.26853346824646,
"learning_rate": 4.881329772700857e-05,
"loss": 2.3299,
"step": 1580
},
{
"epoch": 0.014219352169552958,
"grad_norm": 1.016809344291687,
"learning_rate": 4.881100855861134e-05,
"loss": 2.3164,
"step": 1581
},
{
"epoch": 0.01422834606719341,
"grad_norm": 0.8750474452972412,
"learning_rate": 4.8808717238200815e-05,
"loss": 2.3421,
"step": 1582
},
{
"epoch": 0.01423733996483386,
"grad_norm": 1.335290789604187,
"learning_rate": 4.880642376598407e-05,
"loss": 2.2986,
"step": 1583
},
{
"epoch": 0.01424633386247431,
"grad_norm": 1.4852817058563232,
"learning_rate": 4.8804128142168374e-05,
"loss": 2.2835,
"step": 1584
},
{
"epoch": 0.014255327760114763,
"grad_norm": 0.7631269097328186,
"learning_rate": 4.880183036696123e-05,
"loss": 2.2489,
"step": 1585
},
{
"epoch": 0.014264321657755213,
"grad_norm": 1.0256742238998413,
"learning_rate": 4.8799530440570284e-05,
"loss": 2.2861,
"step": 1586
},
{
"epoch": 0.014273315555395663,
"grad_norm": 1.2452783584594727,
"learning_rate": 4.8797228363203416e-05,
"loss": 2.2486,
"step": 1587
},
{
"epoch": 0.014282309453036116,
"grad_norm": 0.9505096077919006,
"learning_rate": 4.879492413506869e-05,
"loss": 2.2661,
"step": 1588
},
{
"epoch": 0.014291303350676566,
"grad_norm": 1.1314697265625,
"learning_rate": 4.879261775637434e-05,
"loss": 2.402,
"step": 1589
},
{
"epoch": 0.014300297248317016,
"grad_norm": 0.9363741278648376,
"learning_rate": 4.879030922732882e-05,
"loss": 2.2796,
"step": 1590
},
{
"epoch": 0.014309291145957468,
"grad_norm": 0.8884981274604797,
"learning_rate": 4.878799854814079e-05,
"loss": 2.2422,
"step": 1591
},
{
"epoch": 0.014318285043597919,
"grad_norm": 0.917630136013031,
"learning_rate": 4.8785685719019064e-05,
"loss": 2.2508,
"step": 1592
},
{
"epoch": 0.014327278941238369,
"grad_norm": 1.1392261981964111,
"learning_rate": 4.878337074017268e-05,
"loss": 2.2778,
"step": 1593
},
{
"epoch": 0.014336272838878821,
"grad_norm": 0.961719810962677,
"learning_rate": 4.878105361181088e-05,
"loss": 2.25,
"step": 1594
},
{
"epoch": 0.014345266736519272,
"grad_norm": 1.5704357624053955,
"learning_rate": 4.877873433414306e-05,
"loss": 2.2166,
"step": 1595
},
{
"epoch": 0.014354260634159722,
"grad_norm": 1.2725164890289307,
"learning_rate": 4.877641290737884e-05,
"loss": 2.2487,
"step": 1596
},
{
"epoch": 0.014363254531800174,
"grad_norm": 1.314294457435608,
"learning_rate": 4.8774089331728034e-05,
"loss": 2.3055,
"step": 1597
},
{
"epoch": 0.014372248429440624,
"grad_norm": 1.0970916748046875,
"learning_rate": 4.877176360740064e-05,
"loss": 2.2969,
"step": 1598
},
{
"epoch": 0.014381242327081075,
"grad_norm": 0.7976924777030945,
"learning_rate": 4.8769435734606864e-05,
"loss": 2.3035,
"step": 1599
},
{
"epoch": 0.014390236224721527,
"grad_norm": 1.0993620157241821,
"learning_rate": 4.8767105713557085e-05,
"loss": 2.302,
"step": 1600
},
{
"epoch": 0.014399230122361977,
"grad_norm": 1.098041296005249,
"learning_rate": 4.8764773544461886e-05,
"loss": 2.3381,
"step": 1601
},
{
"epoch": 0.014408224020002428,
"grad_norm": 0.9061229825019836,
"learning_rate": 4.8762439227532056e-05,
"loss": 2.2633,
"step": 1602
},
{
"epoch": 0.01441721791764288,
"grad_norm": 0.6966328620910645,
"learning_rate": 4.876010276297857e-05,
"loss": 2.2389,
"step": 1603
},
{
"epoch": 0.01442621181528333,
"grad_norm": 1.0640064477920532,
"learning_rate": 4.8757764151012575e-05,
"loss": 2.2952,
"step": 1604
},
{
"epoch": 0.014435205712923782,
"grad_norm": 0.7718282341957092,
"learning_rate": 4.875542339184546e-05,
"loss": 2.2782,
"step": 1605
},
{
"epoch": 0.014444199610564232,
"grad_norm": 1.0525380373001099,
"learning_rate": 4.875308048568875e-05,
"loss": 2.2899,
"step": 1606
},
{
"epoch": 0.014453193508204683,
"grad_norm": 0.8518139123916626,
"learning_rate": 4.875073543275422e-05,
"loss": 2.1997,
"step": 1607
},
{
"epoch": 0.014462187405845135,
"grad_norm": 0.6957027316093445,
"learning_rate": 4.874838823325381e-05,
"loss": 2.2464,
"step": 1608
},
{
"epoch": 0.014471181303485585,
"grad_norm": 0.7280006408691406,
"learning_rate": 4.874603888739964e-05,
"loss": 2.1706,
"step": 1609
},
{
"epoch": 0.014480175201126036,
"grad_norm": 1.2138913869857788,
"learning_rate": 4.874368739540405e-05,
"loss": 2.4041,
"step": 1610
},
{
"epoch": 0.014489169098766488,
"grad_norm": 1.1543684005737305,
"learning_rate": 4.874133375747957e-05,
"loss": 2.2338,
"step": 1611
},
{
"epoch": 0.014498162996406938,
"grad_norm": 1.113680124282837,
"learning_rate": 4.873897797383893e-05,
"loss": 2.2884,
"step": 1612
},
{
"epoch": 0.014507156894047388,
"grad_norm": 1.1381075382232666,
"learning_rate": 4.873662004469501e-05,
"loss": 2.2889,
"step": 1613
},
{
"epoch": 0.01451615079168784,
"grad_norm": 0.8906194567680359,
"learning_rate": 4.8734259970260954e-05,
"loss": 2.2815,
"step": 1614
},
{
"epoch": 0.014525144689328291,
"grad_norm": 1.0186251401901245,
"learning_rate": 4.8731897750750047e-05,
"loss": 2.2484,
"step": 1615
},
{
"epoch": 0.014534138586968741,
"grad_norm": 2.067753314971924,
"learning_rate": 4.872953338637577e-05,
"loss": 2.2734,
"step": 1616
},
{
"epoch": 0.014543132484609193,
"grad_norm": 1.6547949314117432,
"learning_rate": 4.872716687735183e-05,
"loss": 2.3562,
"step": 1617
},
{
"epoch": 0.014552126382249644,
"grad_norm": 1.2500650882720947,
"learning_rate": 4.872479822389211e-05,
"loss": 2.3027,
"step": 1618
},
{
"epoch": 0.014561120279890094,
"grad_norm": 2.5716488361358643,
"learning_rate": 4.8722427426210675e-05,
"loss": 2.2621,
"step": 1619
},
{
"epoch": 0.014570114177530546,
"grad_norm": 1.515515923500061,
"learning_rate": 4.87200544845218e-05,
"loss": 2.3118,
"step": 1620
},
{
"epoch": 0.014579108075170997,
"grad_norm": 1.1795251369476318,
"learning_rate": 4.871767939903996e-05,
"loss": 2.3822,
"step": 1621
},
{
"epoch": 0.014588101972811447,
"grad_norm": 1.2546466588974,
"learning_rate": 4.871530216997979e-05,
"loss": 2.3991,
"step": 1622
},
{
"epoch": 0.014597095870451899,
"grad_norm": 1.5921902656555176,
"learning_rate": 4.871292279755616e-05,
"loss": 2.3393,
"step": 1623
},
{
"epoch": 0.01460608976809235,
"grad_norm": 1.4096040725708008,
"learning_rate": 4.8710541281984105e-05,
"loss": 2.2775,
"step": 1624
},
{
"epoch": 0.0146150836657328,
"grad_norm": 0.7142196297645569,
"learning_rate": 4.870815762347887e-05,
"loss": 2.2274,
"step": 1625
},
{
"epoch": 0.014624077563373252,
"grad_norm": 1.1841425895690918,
"learning_rate": 4.870577182225589e-05,
"loss": 2.1523,
"step": 1626
},
{
"epoch": 0.014633071461013702,
"grad_norm": 1.2558950185775757,
"learning_rate": 4.870338387853078e-05,
"loss": 2.3211,
"step": 1627
},
{
"epoch": 0.014642065358654152,
"grad_norm": 0.8260988593101501,
"learning_rate": 4.870099379251937e-05,
"loss": 2.2266,
"step": 1628
},
{
"epoch": 0.014651059256294605,
"grad_norm": 1.4167429208755493,
"learning_rate": 4.8698601564437675e-05,
"loss": 2.2658,
"step": 1629
},
{
"epoch": 0.014660053153935055,
"grad_norm": 1.6624324321746826,
"learning_rate": 4.869620719450189e-05,
"loss": 2.2527,
"step": 1630
},
{
"epoch": 0.014669047051575505,
"grad_norm": 1.6745294332504272,
"learning_rate": 4.8693810682928423e-05,
"loss": 2.2864,
"step": 1631
},
{
"epoch": 0.014678040949215957,
"grad_norm": 1.4230668544769287,
"learning_rate": 4.869141202993386e-05,
"loss": 2.3429,
"step": 1632
},
{
"epoch": 0.014687034846856408,
"grad_norm": 123.09016418457031,
"learning_rate": 4.8689011235735006e-05,
"loss": 2.3163,
"step": 1633
},
{
"epoch": 0.014696028744496858,
"grad_norm": 66.58489990234375,
"learning_rate": 4.868660830054883e-05,
"loss": 2.3473,
"step": 1634
},
{
"epoch": 0.01470502264213731,
"grad_norm": 1.6765671968460083,
"learning_rate": 4.8684203224592506e-05,
"loss": 2.312,
"step": 1635
},
{
"epoch": 0.01471401653977776,
"grad_norm": 44.038726806640625,
"learning_rate": 4.86817960080834e-05,
"loss": 2.2916,
"step": 1636
},
{
"epoch": 0.014723010437418211,
"grad_norm": 1.037361979484558,
"learning_rate": 4.867938665123909e-05,
"loss": 2.3141,
"step": 1637
},
{
"epoch": 0.014732004335058663,
"grad_norm": 1.3186761140823364,
"learning_rate": 4.8676975154277314e-05,
"loss": 2.322,
"step": 1638
},
{
"epoch": 0.014740998232699113,
"grad_norm": 1.5475441217422485,
"learning_rate": 4.867456151741603e-05,
"loss": 2.3099,
"step": 1639
},
{
"epoch": 0.014749992130339564,
"grad_norm": 1.6584702730178833,
"learning_rate": 4.8672145740873374e-05,
"loss": 2.3032,
"step": 1640
},
{
"epoch": 0.014758986027980016,
"grad_norm": 1.1423327922821045,
"learning_rate": 4.866972782486768e-05,
"loss": 2.3606,
"step": 1641
},
{
"epoch": 0.014767979925620466,
"grad_norm": 1.4102569818496704,
"learning_rate": 4.866730776961749e-05,
"loss": 2.3171,
"step": 1642
},
{
"epoch": 0.014776973823260918,
"grad_norm": 4.442851543426514,
"learning_rate": 4.8664885575341506e-05,
"loss": 2.3512,
"step": 1643
},
{
"epoch": 0.014785967720901369,
"grad_norm": 1.1988860368728638,
"learning_rate": 4.8662461242258664e-05,
"loss": 2.3322,
"step": 1644
},
{
"epoch": 0.014794961618541819,
"grad_norm": 1.2642006874084473,
"learning_rate": 4.8660034770588046e-05,
"loss": 2.2811,
"step": 1645
},
{
"epoch": 0.014803955516182271,
"grad_norm": 0.9310919046401978,
"learning_rate": 4.8657606160548986e-05,
"loss": 2.3692,
"step": 1646
},
{
"epoch": 0.014812949413822721,
"grad_norm": 1.4296855926513672,
"learning_rate": 4.8655175412360954e-05,
"loss": 2.3736,
"step": 1647
},
{
"epoch": 0.014821943311463172,
"grad_norm": 1.0482155084609985,
"learning_rate": 4.8652742526243644e-05,
"loss": 2.3192,
"step": 1648
},
{
"epoch": 0.014830937209103624,
"grad_norm": 0.7324535846710205,
"learning_rate": 4.8650307502416945e-05,
"loss": 2.2758,
"step": 1649
},
{
"epoch": 0.014839931106744074,
"grad_norm": 1.776188850402832,
"learning_rate": 4.864787034110093e-05,
"loss": 2.2996,
"step": 1650
},
{
"epoch": 0.014848925004384525,
"grad_norm": 1.8748433589935303,
"learning_rate": 4.864543104251587e-05,
"loss": 2.311,
"step": 1651
},
{
"epoch": 0.014857918902024977,
"grad_norm": 0.8143200874328613,
"learning_rate": 4.864298960688222e-05,
"loss": 2.254,
"step": 1652
},
{
"epoch": 0.014866912799665427,
"grad_norm": 1.3252731561660767,
"learning_rate": 4.864054603442062e-05,
"loss": 2.243,
"step": 1653
},
{
"epoch": 0.014875906697305877,
"grad_norm": 1.9002904891967773,
"learning_rate": 4.863810032535195e-05,
"loss": 2.3136,
"step": 1654
},
{
"epoch": 0.01488490059494633,
"grad_norm": 1.0166923999786377,
"learning_rate": 4.863565247989723e-05,
"loss": 2.3262,
"step": 1655
},
{
"epoch": 0.01489389449258678,
"grad_norm": 1.0350232124328613,
"learning_rate": 4.863320249827769e-05,
"loss": 2.338,
"step": 1656
},
{
"epoch": 0.01490288839022723,
"grad_norm": 1.132045865058899,
"learning_rate": 4.863075038071477e-05,
"loss": 2.2896,
"step": 1657
},
{
"epoch": 0.014911882287867682,
"grad_norm": 1.8662331104278564,
"learning_rate": 4.862829612743008e-05,
"loss": 2.4228,
"step": 1658
},
{
"epoch": 0.014920876185508133,
"grad_norm": 2.886127233505249,
"learning_rate": 4.862583973864544e-05,
"loss": 2.3359,
"step": 1659
},
{
"epoch": 0.014929870083148583,
"grad_norm": 1.339038610458374,
"learning_rate": 4.8623381214582855e-05,
"loss": 2.3592,
"step": 1660
},
{
"epoch": 0.014938863980789035,
"grad_norm": 1.4620894193649292,
"learning_rate": 4.862092055546451e-05,
"loss": 2.3012,
"step": 1661
},
{
"epoch": 0.014947857878429486,
"grad_norm": 1.0699290037155151,
"learning_rate": 4.861845776151281e-05,
"loss": 2.3676,
"step": 1662
},
{
"epoch": 0.014956851776069936,
"grad_norm": 1.0566192865371704,
"learning_rate": 4.861599283295035e-05,
"loss": 2.276,
"step": 1663
},
{
"epoch": 0.014965845673710388,
"grad_norm": 1.0292271375656128,
"learning_rate": 4.861352576999989e-05,
"loss": 2.2916,
"step": 1664
},
{
"epoch": 0.014974839571350838,
"grad_norm": 1.2903679609298706,
"learning_rate": 4.86110565728844e-05,
"loss": 2.3391,
"step": 1665
},
{
"epoch": 0.014983833468991289,
"grad_norm": 1.051493763923645,
"learning_rate": 4.8608585241827046e-05,
"loss": 2.3237,
"step": 1666
},
{
"epoch": 0.01499282736663174,
"grad_norm": 1.1401419639587402,
"learning_rate": 4.8606111777051186e-05,
"loss": 2.3291,
"step": 1667
},
{
"epoch": 0.015001821264272191,
"grad_norm": 1.4332388639450073,
"learning_rate": 4.860363617878037e-05,
"loss": 2.262,
"step": 1668
},
{
"epoch": 0.015010815161912642,
"grad_norm": 0.9544563889503479,
"learning_rate": 4.8601158447238346e-05,
"loss": 2.2888,
"step": 1669
},
{
"epoch": 0.015019809059553094,
"grad_norm": 0.9660168290138245,
"learning_rate": 4.859867858264905e-05,
"loss": 2.2314,
"step": 1670
},
{
"epoch": 0.015028802957193544,
"grad_norm": 0.9694017171859741,
"learning_rate": 4.8596196585236595e-05,
"loss": 2.2047,
"step": 1671
},
{
"epoch": 0.015037796854833994,
"grad_norm": 1.5287903547286987,
"learning_rate": 4.85937124552253e-05,
"loss": 2.3677,
"step": 1672
},
{
"epoch": 0.015046790752474446,
"grad_norm": 0.8434200286865234,
"learning_rate": 4.8591226192839696e-05,
"loss": 2.2631,
"step": 1673
},
{
"epoch": 0.015055784650114897,
"grad_norm": 0.9062039852142334,
"learning_rate": 4.8588737798304476e-05,
"loss": 2.258,
"step": 1674
},
{
"epoch": 0.015064778547755347,
"grad_norm": 1.3914508819580078,
"learning_rate": 4.8586247271844546e-05,
"loss": 2.306,
"step": 1675
},
{
"epoch": 0.0150737724453958,
"grad_norm": 1.364774465560913,
"learning_rate": 4.858375461368499e-05,
"loss": 2.2833,
"step": 1676
},
{
"epoch": 0.01508276634303625,
"grad_norm": 0.8096304535865784,
"learning_rate": 4.858125982405108e-05,
"loss": 2.2704,
"step": 1677
},
{
"epoch": 0.0150917602406767,
"grad_norm": 0.9544671773910522,
"learning_rate": 4.857876290316833e-05,
"loss": 2.2346,
"step": 1678
},
{
"epoch": 0.015100754138317152,
"grad_norm": 1.6853290796279907,
"learning_rate": 4.857626385126237e-05,
"loss": 2.3401,
"step": 1679
},
{
"epoch": 0.015109748035957602,
"grad_norm": 1.2222957611083984,
"learning_rate": 4.857376266855909e-05,
"loss": 2.2968,
"step": 1680
},
{
"epoch": 0.015118741933598055,
"grad_norm": 1.121512770652771,
"learning_rate": 4.8571259355284515e-05,
"loss": 2.3299,
"step": 1681
},
{
"epoch": 0.015127735831238505,
"grad_norm": 1.0393519401550293,
"learning_rate": 4.856875391166492e-05,
"loss": 2.2729,
"step": 1682
},
{
"epoch": 0.015136729728878955,
"grad_norm": 0.924287736415863,
"learning_rate": 4.856624633792673e-05,
"loss": 2.2261,
"step": 1683
},
{
"epoch": 0.015145723626519407,
"grad_norm": 1.5494250059127808,
"learning_rate": 4.856373663429657e-05,
"loss": 2.2968,
"step": 1684
},
{
"epoch": 0.015154717524159858,
"grad_norm": 1.256866455078125,
"learning_rate": 4.856122480100129e-05,
"loss": 2.3281,
"step": 1685
},
{
"epoch": 0.015163711421800308,
"grad_norm": 2.0648908615112305,
"learning_rate": 4.8558710838267885e-05,
"loss": 2.2923,
"step": 1686
},
{
"epoch": 0.01517270531944076,
"grad_norm": 5.36793851852417,
"learning_rate": 4.855619474632357e-05,
"loss": 2.3124,
"step": 1687
},
{
"epoch": 0.01518169921708121,
"grad_norm": 0.9082876443862915,
"learning_rate": 4.855367652539574e-05,
"loss": 2.1946,
"step": 1688
},
{
"epoch": 0.015190693114721661,
"grad_norm": 2.38844895362854,
"learning_rate": 4.8551156175712e-05,
"loss": 2.29,
"step": 1689
},
{
"epoch": 0.015199687012362113,
"grad_norm": 3.0658109188079834,
"learning_rate": 4.854863369750014e-05,
"loss": 2.2714,
"step": 1690
},
{
"epoch": 0.015208680910002563,
"grad_norm": 1.5102766752243042,
"learning_rate": 4.854610909098812e-05,
"loss": 2.2177,
"step": 1691
},
{
"epoch": 0.015217674807643014,
"grad_norm": 0.9167364239692688,
"learning_rate": 4.854358235640413e-05,
"loss": 2.243,
"step": 1692
},
{
"epoch": 0.015226668705283466,
"grad_norm": 1.907095193862915,
"learning_rate": 4.8541053493976516e-05,
"loss": 2.34,
"step": 1693
},
{
"epoch": 0.015235662602923916,
"grad_norm": 0.6987553238868713,
"learning_rate": 4.8538522503933846e-05,
"loss": 2.2872,
"step": 1694
},
{
"epoch": 0.015244656500564367,
"grad_norm": 0.8857932090759277,
"learning_rate": 4.853598938650487e-05,
"loss": 2.2709,
"step": 1695
},
{
"epoch": 0.015253650398204819,
"grad_norm": 1.0220707654953003,
"learning_rate": 4.853345414191852e-05,
"loss": 2.3388,
"step": 1696
},
{
"epoch": 0.015262644295845269,
"grad_norm": 0.9603571891784668,
"learning_rate": 4.853091677040395e-05,
"loss": 2.2508,
"step": 1697
},
{
"epoch": 0.01527163819348572,
"grad_norm": 1.0661171674728394,
"learning_rate": 4.852837727219044e-05,
"loss": 2.3104,
"step": 1698
},
{
"epoch": 0.015280632091126171,
"grad_norm": 1.0047944784164429,
"learning_rate": 4.852583564750755e-05,
"loss": 2.346,
"step": 1699
},
{
"epoch": 0.015289625988766622,
"grad_norm": 1.9199333190917969,
"learning_rate": 4.8523291896584977e-05,
"loss": 2.3277,
"step": 1700
},
{
"epoch": 0.015298619886407072,
"grad_norm": 1.2526764869689941,
"learning_rate": 4.8520746019652605e-05,
"loss": 2.2917,
"step": 1701
},
{
"epoch": 0.015307613784047524,
"grad_norm": 1.4083847999572754,
"learning_rate": 4.851819801694056e-05,
"loss": 2.2945,
"step": 1702
},
{
"epoch": 0.015316607681687975,
"grad_norm": 1.1006942987442017,
"learning_rate": 4.85156478886791e-05,
"loss": 2.2471,
"step": 1703
},
{
"epoch": 0.015325601579328425,
"grad_norm": 0.9158415794372559,
"learning_rate": 4.851309563509872e-05,
"loss": 2.2687,
"step": 1704
},
{
"epoch": 0.015334595476968877,
"grad_norm": 1.537262201309204,
"learning_rate": 4.851054125643008e-05,
"loss": 2.2725,
"step": 1705
},
{
"epoch": 0.015343589374609327,
"grad_norm": 0.9294925332069397,
"learning_rate": 4.850798475290403e-05,
"loss": 2.3882,
"step": 1706
},
{
"epoch": 0.015352583272249778,
"grad_norm": 0.7939320802688599,
"learning_rate": 4.850542612475165e-05,
"loss": 2.2375,
"step": 1707
},
{
"epoch": 0.01536157716989023,
"grad_norm": 0.7283051013946533,
"learning_rate": 4.8502865372204184e-05,
"loss": 2.247,
"step": 1708
},
{
"epoch": 0.01537057106753068,
"grad_norm": 1.1219340562820435,
"learning_rate": 4.850030249549305e-05,
"loss": 2.2645,
"step": 1709
},
{
"epoch": 0.01537956496517113,
"grad_norm": 0.6406038403511047,
"learning_rate": 4.849773749484989e-05,
"loss": 2.264,
"step": 1710
},
{
"epoch": 0.015388558862811583,
"grad_norm": 1.3027560710906982,
"learning_rate": 4.849517037050652e-05,
"loss": 2.2326,
"step": 1711
},
{
"epoch": 0.015397552760452033,
"grad_norm": 1.0084408521652222,
"learning_rate": 4.849260112269497e-05,
"loss": 2.3088,
"step": 1712
},
{
"epoch": 0.015406546658092483,
"grad_norm": 0.5874270796775818,
"learning_rate": 4.849002975164743e-05,
"loss": 2.2642,
"step": 1713
},
{
"epoch": 0.015415540555732936,
"grad_norm": 1.212501883506775,
"learning_rate": 4.8487456257596305e-05,
"loss": 2.2844,
"step": 1714
},
{
"epoch": 0.015424534453373386,
"grad_norm": 0.9356610774993896,
"learning_rate": 4.8484880640774176e-05,
"loss": 2.2396,
"step": 1715
},
{
"epoch": 0.015433528351013838,
"grad_norm": 1.258927583694458,
"learning_rate": 4.848230290141383e-05,
"loss": 2.2819,
"step": 1716
},
{
"epoch": 0.015442522248654288,
"grad_norm": 0.7848899960517883,
"learning_rate": 4.8479723039748246e-05,
"loss": 2.2881,
"step": 1717
},
{
"epoch": 0.015451516146294739,
"grad_norm": 1.0405381917953491,
"learning_rate": 4.847714105601059e-05,
"loss": 2.2882,
"step": 1718
},
{
"epoch": 0.01546051004393519,
"grad_norm": 0.8394889831542969,
"learning_rate": 4.84745569504342e-05,
"loss": 2.2942,
"step": 1719
},
{
"epoch": 0.015469503941575641,
"grad_norm": 0.628038227558136,
"learning_rate": 4.8471970723252645e-05,
"loss": 2.3152,
"step": 1720
},
{
"epoch": 0.015478497839216092,
"grad_norm": 1.528188943862915,
"learning_rate": 4.846938237469966e-05,
"loss": 2.345,
"step": 1721
},
{
"epoch": 0.015487491736856544,
"grad_norm": 1.0436937808990479,
"learning_rate": 4.8466791905009165e-05,
"loss": 2.2968,
"step": 1722
},
{
"epoch": 0.015496485634496994,
"grad_norm": 1.2217814922332764,
"learning_rate": 4.84641993144153e-05,
"loss": 2.3549,
"step": 1723
},
{
"epoch": 0.015505479532137444,
"grad_norm": 1.5608245134353638,
"learning_rate": 4.8461604603152376e-05,
"loss": 2.3443,
"step": 1724
},
{
"epoch": 0.015514473429777896,
"grad_norm": 0.9040743112564087,
"learning_rate": 4.84590077714549e-05,
"loss": 2.2734,
"step": 1725
},
{
"epoch": 0.015523467327418347,
"grad_norm": 1.0167921781539917,
"learning_rate": 4.8456408819557564e-05,
"loss": 2.2381,
"step": 1726
},
{
"epoch": 0.015532461225058797,
"grad_norm": 0.7634522318840027,
"learning_rate": 4.845380774769527e-05,
"loss": 2.2552,
"step": 1727
},
{
"epoch": 0.01554145512269925,
"grad_norm": 1.1253379583358765,
"learning_rate": 4.845120455610309e-05,
"loss": 2.3381,
"step": 1728
},
{
"epoch": 0.0155504490203397,
"grad_norm": 1.342286229133606,
"learning_rate": 4.84485992450163e-05,
"loss": 2.2358,
"step": 1729
},
{
"epoch": 0.01555944291798015,
"grad_norm": 1.160684585571289,
"learning_rate": 4.844599181467037e-05,
"loss": 2.3146,
"step": 1730
},
{
"epoch": 0.015568436815620602,
"grad_norm": 0.8372664451599121,
"learning_rate": 4.844338226530095e-05,
"loss": 2.2624,
"step": 1731
},
{
"epoch": 0.015577430713261052,
"grad_norm": 0.6305686235427856,
"learning_rate": 4.8440770597143906e-05,
"loss": 2.2075,
"step": 1732
},
{
"epoch": 0.015586424610901503,
"grad_norm": 1.278459072113037,
"learning_rate": 4.843815681043525e-05,
"loss": 2.2647,
"step": 1733
},
{
"epoch": 0.015595418508541955,
"grad_norm": 1.4778318405151367,
"learning_rate": 4.843554090541123e-05,
"loss": 2.2812,
"step": 1734
},
{
"epoch": 0.015604412406182405,
"grad_norm": 0.843654215335846,
"learning_rate": 4.843292288230826e-05,
"loss": 2.2355,
"step": 1735
},
{
"epoch": 0.015613406303822856,
"grad_norm": 1.2035799026489258,
"learning_rate": 4.843030274136297e-05,
"loss": 2.2931,
"step": 1736
},
{
"epoch": 0.015622400201463308,
"grad_norm": 0.9506595134735107,
"learning_rate": 4.842768048281216e-05,
"loss": 2.2913,
"step": 1737
},
{
"epoch": 0.01563139409910376,
"grad_norm": 1.4588580131530762,
"learning_rate": 4.8425056106892805e-05,
"loss": 2.3381,
"step": 1738
},
{
"epoch": 0.01564038799674421,
"grad_norm": 1.281014084815979,
"learning_rate": 4.842242961384211e-05,
"loss": 2.2844,
"step": 1739
},
{
"epoch": 0.01564938189438466,
"grad_norm": 1.2635339498519897,
"learning_rate": 4.841980100389747e-05,
"loss": 2.3184,
"step": 1740
},
{
"epoch": 0.01565837579202511,
"grad_norm": 0.9929319024085999,
"learning_rate": 4.841717027729643e-05,
"loss": 2.3244,
"step": 1741
},
{
"epoch": 0.01566736968966556,
"grad_norm": 0.9049252867698669,
"learning_rate": 4.841453743427678e-05,
"loss": 2.3104,
"step": 1742
},
{
"epoch": 0.01567636358730601,
"grad_norm": 1.3217833042144775,
"learning_rate": 4.8411902475076435e-05,
"loss": 2.3355,
"step": 1743
},
{
"epoch": 0.015685357484946465,
"grad_norm": 1.3570677042007446,
"learning_rate": 4.8409265399933574e-05,
"loss": 2.2917,
"step": 1744
},
{
"epoch": 0.015694351382586916,
"grad_norm": 1.055814266204834,
"learning_rate": 4.8406626209086515e-05,
"loss": 2.3087,
"step": 1745
},
{
"epoch": 0.015703345280227366,
"grad_norm": 0.8690663576126099,
"learning_rate": 4.8403984902773796e-05,
"loss": 2.3487,
"step": 1746
},
{
"epoch": 0.015712339177867816,
"grad_norm": 1.463718056678772,
"learning_rate": 4.840134148123413e-05,
"loss": 2.3123,
"step": 1747
},
{
"epoch": 0.015721333075508267,
"grad_norm": 1.0361788272857666,
"learning_rate": 4.839869594470642e-05,
"loss": 2.2572,
"step": 1748
},
{
"epoch": 0.015730326973148717,
"grad_norm": 1.4331319332122803,
"learning_rate": 4.839604829342978e-05,
"loss": 2.25,
"step": 1749
},
{
"epoch": 0.01573932087078917,
"grad_norm": 1.7820483446121216,
"learning_rate": 4.83933985276435e-05,
"loss": 2.3103,
"step": 1750
},
{
"epoch": 0.01574831476842962,
"grad_norm": 1.5386706590652466,
"learning_rate": 4.839074664758704e-05,
"loss": 2.2559,
"step": 1751
},
{
"epoch": 0.015757308666070072,
"grad_norm": 1.9421716928482056,
"learning_rate": 4.8388092653500105e-05,
"loss": 2.3133,
"step": 1752
},
{
"epoch": 0.015766302563710522,
"grad_norm": 0.7473291754722595,
"learning_rate": 4.838543654562255e-05,
"loss": 2.2426,
"step": 1753
},
{
"epoch": 0.015775296461350972,
"grad_norm": 2.3026628494262695,
"learning_rate": 4.838277832419442e-05,
"loss": 2.4159,
"step": 1754
},
{
"epoch": 0.015784290358991423,
"grad_norm": 1.0959279537200928,
"learning_rate": 4.838011798945598e-05,
"loss": 2.2783,
"step": 1755
},
{
"epoch": 0.015793284256631877,
"grad_norm": 1.1502763032913208,
"learning_rate": 4.837745554164765e-05,
"loss": 2.2898,
"step": 1756
},
{
"epoch": 0.015802278154272327,
"grad_norm": 0.7148979902267456,
"learning_rate": 4.8374790981010075e-05,
"loss": 2.1665,
"step": 1757
},
{
"epoch": 0.015811272051912777,
"grad_norm": 1.093146562576294,
"learning_rate": 4.8372124307784075e-05,
"loss": 2.2296,
"step": 1758
},
{
"epoch": 0.015820265949553228,
"grad_norm": 0.8140142560005188,
"learning_rate": 4.836945552221065e-05,
"loss": 2.2643,
"step": 1759
},
{
"epoch": 0.015829259847193678,
"grad_norm": 0.9954135417938232,
"learning_rate": 4.8366784624531e-05,
"loss": 2.2318,
"step": 1760
},
{
"epoch": 0.01583825374483413,
"grad_norm": 1.4276254177093506,
"learning_rate": 4.8364111614986527e-05,
"loss": 2.2922,
"step": 1761
},
{
"epoch": 0.015847247642474582,
"grad_norm": 0.871088445186615,
"learning_rate": 4.836143649381881e-05,
"loss": 2.3666,
"step": 1762
},
{
"epoch": 0.015856241540115033,
"grad_norm": 1.3774094581604004,
"learning_rate": 4.8358759261269626e-05,
"loss": 2.3687,
"step": 1763
},
{
"epoch": 0.015865235437755483,
"grad_norm": 0.8023905754089355,
"learning_rate": 4.8356079917580944e-05,
"loss": 2.2467,
"step": 1764
},
{
"epoch": 0.015874229335395933,
"grad_norm": 1.524170994758606,
"learning_rate": 4.8353398462994916e-05,
"loss": 2.3557,
"step": 1765
},
{
"epoch": 0.015883223233036384,
"grad_norm": 1.3545712232589722,
"learning_rate": 4.835071489775388e-05,
"loss": 2.2675,
"step": 1766
},
{
"epoch": 0.015892217130676834,
"grad_norm": 0.9542168974876404,
"learning_rate": 4.834802922210039e-05,
"loss": 2.2941,
"step": 1767
},
{
"epoch": 0.015901211028317288,
"grad_norm": 1.1913856267929077,
"learning_rate": 4.834534143627717e-05,
"loss": 2.2461,
"step": 1768
},
{
"epoch": 0.01591020492595774,
"grad_norm": 0.9957913160324097,
"learning_rate": 4.8342651540527136e-05,
"loss": 2.2682,
"step": 1769
},
{
"epoch": 0.01591919882359819,
"grad_norm": 1.2226651906967163,
"learning_rate": 4.8339959535093395e-05,
"loss": 2.3396,
"step": 1770
},
{
"epoch": 0.01592819272123864,
"grad_norm": 2.445849657058716,
"learning_rate": 4.8337265420219245e-05,
"loss": 2.2851,
"step": 1771
},
{
"epoch": 0.01593718661887909,
"grad_norm": 2.062366247177124,
"learning_rate": 4.83345691961482e-05,
"loss": 2.2925,
"step": 1772
},
{
"epoch": 0.015946180516519543,
"grad_norm": 1.8406349420547485,
"learning_rate": 4.8331870863123905e-05,
"loss": 2.2293,
"step": 1773
},
{
"epoch": 0.015955174414159994,
"grad_norm": 0.8395392894744873,
"learning_rate": 4.832917042139027e-05,
"loss": 2.3894,
"step": 1774
},
{
"epoch": 0.015964168311800444,
"grad_norm": 1.0644644498825073,
"learning_rate": 4.8326467871191325e-05,
"loss": 2.3345,
"step": 1775
},
{
"epoch": 0.015973162209440894,
"grad_norm": 0.9206362962722778,
"learning_rate": 4.8323763212771354e-05,
"loss": 2.2374,
"step": 1776
},
{
"epoch": 0.015982156107081345,
"grad_norm": 0.7812778353691101,
"learning_rate": 4.832105644637478e-05,
"loss": 2.254,
"step": 1777
},
{
"epoch": 0.015991150004721795,
"grad_norm": 1.4544132947921753,
"learning_rate": 4.831834757224624e-05,
"loss": 2.2713,
"step": 1778
},
{
"epoch": 0.01600014390236225,
"grad_norm": 1.4938353300094604,
"learning_rate": 4.8315636590630566e-05,
"loss": 2.3449,
"step": 1779
},
{
"epoch": 0.0160091378000027,
"grad_norm": 0.861350953578949,
"learning_rate": 4.831292350177277e-05,
"loss": 2.2377,
"step": 1780
},
{
"epoch": 0.01601813169764315,
"grad_norm": 0.9754862785339355,
"learning_rate": 4.831020830591806e-05,
"loss": 2.3128,
"step": 1781
},
{
"epoch": 0.0160271255952836,
"grad_norm": 0.8671904802322388,
"learning_rate": 4.830749100331183e-05,
"loss": 2.2829,
"step": 1782
},
{
"epoch": 0.01603611949292405,
"grad_norm": 1.046902060508728,
"learning_rate": 4.830477159419966e-05,
"loss": 2.2218,
"step": 1783
},
{
"epoch": 0.0160451133905645,
"grad_norm": 1.9188002347946167,
"learning_rate": 4.830205007882735e-05,
"loss": 2.2688,
"step": 1784
},
{
"epoch": 0.016054107288204954,
"grad_norm": 1.4355274438858032,
"learning_rate": 4.829932645744084e-05,
"loss": 2.3428,
"step": 1785
},
{
"epoch": 0.016063101185845405,
"grad_norm": 1.168654441833496,
"learning_rate": 4.829660073028631e-05,
"loss": 2.2224,
"step": 1786
},
{
"epoch": 0.016072095083485855,
"grad_norm": 1.3810064792633057,
"learning_rate": 4.8293872897610096e-05,
"loss": 2.2886,
"step": 1787
},
{
"epoch": 0.016081088981126306,
"grad_norm": 1.2109360694885254,
"learning_rate": 4.829114295965874e-05,
"loss": 2.2377,
"step": 1788
},
{
"epoch": 0.016090082878766756,
"grad_norm": 1.4794343709945679,
"learning_rate": 4.828841091667897e-05,
"loss": 2.2686,
"step": 1789
},
{
"epoch": 0.016099076776407206,
"grad_norm": 1.6576998233795166,
"learning_rate": 4.8285676768917704e-05,
"loss": 2.2041,
"step": 1790
},
{
"epoch": 0.01610807067404766,
"grad_norm": 1.1506435871124268,
"learning_rate": 4.828294051662206e-05,
"loss": 2.2795,
"step": 1791
},
{
"epoch": 0.01611706457168811,
"grad_norm": 1.8207429647445679,
"learning_rate": 4.828020216003932e-05,
"loss": 2.2831,
"step": 1792
},
{
"epoch": 0.01612605846932856,
"grad_norm": 1.3562047481536865,
"learning_rate": 4.827746169941698e-05,
"loss": 2.2682,
"step": 1793
},
{
"epoch": 0.01613505236696901,
"grad_norm": 1.6682422161102295,
"learning_rate": 4.827471913500274e-05,
"loss": 2.3876,
"step": 1794
},
{
"epoch": 0.01614404626460946,
"grad_norm": 1.1676632165908813,
"learning_rate": 4.827197446704444e-05,
"loss": 2.2827,
"step": 1795
},
{
"epoch": 0.016153040162249912,
"grad_norm": 0.7973687052726746,
"learning_rate": 4.826922769579016e-05,
"loss": 2.2901,
"step": 1796
},
{
"epoch": 0.016162034059890366,
"grad_norm": 1.4986774921417236,
"learning_rate": 4.826647882148815e-05,
"loss": 2.2922,
"step": 1797
},
{
"epoch": 0.016171027957530816,
"grad_norm": 1.4185477495193481,
"learning_rate": 4.826372784438684e-05,
"loss": 2.3027,
"step": 1798
},
{
"epoch": 0.016180021855171266,
"grad_norm": 1.560782790184021,
"learning_rate": 4.826097476473486e-05,
"loss": 2.3667,
"step": 1799
},
{
"epoch": 0.016189015752811717,
"grad_norm": 1.1071298122406006,
"learning_rate": 4.8258219582781046e-05,
"loss": 2.2883,
"step": 1800
},
{
"epoch": 0.016198009650452167,
"grad_norm": 1.3657180070877075,
"learning_rate": 4.825546229877439e-05,
"loss": 2.3031,
"step": 1801
},
{
"epoch": 0.016207003548092618,
"grad_norm": 1.2129443883895874,
"learning_rate": 4.82527029129641e-05,
"loss": 2.2703,
"step": 1802
},
{
"epoch": 0.01621599744573307,
"grad_norm": 0.9238573312759399,
"learning_rate": 4.824994142559957e-05,
"loss": 2.2827,
"step": 1803
},
{
"epoch": 0.01622499134337352,
"grad_norm": 1.260115146636963,
"learning_rate": 4.824717783693038e-05,
"loss": 2.3157,
"step": 1804
},
{
"epoch": 0.016233985241013972,
"grad_norm": 1.4081841707229614,
"learning_rate": 4.8244412147206284e-05,
"loss": 2.2942,
"step": 1805
},
{
"epoch": 0.016242979138654422,
"grad_norm": 0.9430930614471436,
"learning_rate": 4.824164435667727e-05,
"loss": 2.2714,
"step": 1806
},
{
"epoch": 0.016251973036294873,
"grad_norm": 1.180860996246338,
"learning_rate": 4.8238874465593464e-05,
"loss": 2.319,
"step": 1807
},
{
"epoch": 0.016260966933935327,
"grad_norm": 1.352108359336853,
"learning_rate": 4.823610247420521e-05,
"loss": 2.3218,
"step": 1808
},
{
"epoch": 0.016269960831575777,
"grad_norm": 2.0896854400634766,
"learning_rate": 4.8233328382763057e-05,
"loss": 2.3172,
"step": 1809
},
{
"epoch": 0.016278954729216227,
"grad_norm": 1.1523449420928955,
"learning_rate": 4.8230552191517694e-05,
"loss": 2.2646,
"step": 1810
},
{
"epoch": 0.016287948626856678,
"grad_norm": 1.7392079830169678,
"learning_rate": 4.8227773900720056e-05,
"loss": 2.2986,
"step": 1811
},
{
"epoch": 0.016296942524497128,
"grad_norm": 0.9787827134132385,
"learning_rate": 4.822499351062123e-05,
"loss": 2.2402,
"step": 1812
},
{
"epoch": 0.01630593642213758,
"grad_norm": 1.443446397781372,
"learning_rate": 4.8222211021472506e-05,
"loss": 2.3217,
"step": 1813
},
{
"epoch": 0.016314930319778032,
"grad_norm": 0.7974103689193726,
"learning_rate": 4.821942643352536e-05,
"loss": 2.2411,
"step": 1814
},
{
"epoch": 0.016323924217418483,
"grad_norm": 1.7783901691436768,
"learning_rate": 4.821663974703147e-05,
"loss": 2.4581,
"step": 1815
},
{
"epoch": 0.016332918115058933,
"grad_norm": 0.8028185963630676,
"learning_rate": 4.821385096224268e-05,
"loss": 2.2644,
"step": 1816
},
{
"epoch": 0.016341912012699383,
"grad_norm": 0.8400540351867676,
"learning_rate": 4.821106007941105e-05,
"loss": 2.2547,
"step": 1817
},
{
"epoch": 0.016350905910339834,
"grad_norm": 1.2375069856643677,
"learning_rate": 4.820826709878881e-05,
"loss": 2.2746,
"step": 1818
},
{
"epoch": 0.016359899807980284,
"grad_norm": 0.670217752456665,
"learning_rate": 4.820547202062837e-05,
"loss": 2.2338,
"step": 1819
},
{
"epoch": 0.016368893705620738,
"grad_norm": 1.0312367677688599,
"learning_rate": 4.8202674845182384e-05,
"loss": 2.2935,
"step": 1820
},
{
"epoch": 0.016377887603261188,
"grad_norm": 0.7886951565742493,
"learning_rate": 4.819987557270363e-05,
"loss": 2.1898,
"step": 1821
},
{
"epoch": 0.01638688150090164,
"grad_norm": 0.848126232624054,
"learning_rate": 4.819707420344511e-05,
"loss": 2.2792,
"step": 1822
},
{
"epoch": 0.01639587539854209,
"grad_norm": 1.2181836366653442,
"learning_rate": 4.819427073766001e-05,
"loss": 2.298,
"step": 1823
},
{
"epoch": 0.01640486929618254,
"grad_norm": 1.0177326202392578,
"learning_rate": 4.8191465175601706e-05,
"loss": 2.3741,
"step": 1824
},
{
"epoch": 0.01641386319382299,
"grad_norm": 1.1135579347610474,
"learning_rate": 4.818865751752376e-05,
"loss": 2.3023,
"step": 1825
},
{
"epoch": 0.016422857091463443,
"grad_norm": 0.6524264216423035,
"learning_rate": 4.818584776367993e-05,
"loss": 2.2265,
"step": 1826
},
{
"epoch": 0.016431850989103894,
"grad_norm": 1.0768200159072876,
"learning_rate": 4.8183035914324136e-05,
"loss": 2.2416,
"step": 1827
},
{
"epoch": 0.016440844886744344,
"grad_norm": 1.1255426406860352,
"learning_rate": 4.818022196971054e-05,
"loss": 2.3167,
"step": 1828
},
{
"epoch": 0.016449838784384795,
"grad_norm": 0.659188985824585,
"learning_rate": 4.817740593009344e-05,
"loss": 2.2551,
"step": 1829
},
{
"epoch": 0.016458832682025245,
"grad_norm": 1.033416748046875,
"learning_rate": 4.8174587795727364e-05,
"loss": 2.2822,
"step": 1830
},
{
"epoch": 0.016467826579665695,
"grad_norm": 1.2327380180358887,
"learning_rate": 4.8171767566867e-05,
"loss": 2.2626,
"step": 1831
},
{
"epoch": 0.01647682047730615,
"grad_norm": 1.0146552324295044,
"learning_rate": 4.816894524376725e-05,
"loss": 2.3406,
"step": 1832
},
{
"epoch": 0.0164858143749466,
"grad_norm": 0.994263231754303,
"learning_rate": 4.816612082668318e-05,
"loss": 2.2253,
"step": 1833
},
{
"epoch": 0.01649480827258705,
"grad_norm": 0.9802228808403015,
"learning_rate": 4.816329431587006e-05,
"loss": 2.2665,
"step": 1834
},
{
"epoch": 0.0165038021702275,
"grad_norm": 0.693946897983551,
"learning_rate": 4.816046571158335e-05,
"loss": 2.2454,
"step": 1835
},
{
"epoch": 0.01651279606786795,
"grad_norm": 1.3491873741149902,
"learning_rate": 4.8157635014078694e-05,
"loss": 2.2501,
"step": 1836
},
{
"epoch": 0.0165217899655084,
"grad_norm": 1.929073691368103,
"learning_rate": 4.8154802223611935e-05,
"loss": 2.369,
"step": 1837
},
{
"epoch": 0.016530783863148855,
"grad_norm": 1.6979625225067139,
"learning_rate": 4.815196734043909e-05,
"loss": 2.3639,
"step": 1838
},
{
"epoch": 0.016539777760789305,
"grad_norm": 1.1496517658233643,
"learning_rate": 4.8149130364816364e-05,
"loss": 2.2529,
"step": 1839
},
{
"epoch": 0.016548771658429755,
"grad_norm": 1.0107629299163818,
"learning_rate": 4.8146291297000176e-05,
"loss": 2.3402,
"step": 1840
},
{
"epoch": 0.016557765556070206,
"grad_norm": 1.1265355348587036,
"learning_rate": 4.8143450137247114e-05,
"loss": 2.3118,
"step": 1841
},
{
"epoch": 0.016566759453710656,
"grad_norm": 1.4963544607162476,
"learning_rate": 4.814060688581396e-05,
"loss": 2.2731,
"step": 1842
},
{
"epoch": 0.016575753351351107,
"grad_norm": 1.1520501375198364,
"learning_rate": 4.813776154295767e-05,
"loss": 2.2518,
"step": 1843
},
{
"epoch": 0.01658474724899156,
"grad_norm": 1.1177328824996948,
"learning_rate": 4.813491410893543e-05,
"loss": 2.1875,
"step": 1844
},
{
"epoch": 0.01659374114663201,
"grad_norm": 0.7695892453193665,
"learning_rate": 4.813206458400455e-05,
"loss": 2.2137,
"step": 1845
},
{
"epoch": 0.01660273504427246,
"grad_norm": 1.1445571184158325,
"learning_rate": 4.812921296842261e-05,
"loss": 2.2929,
"step": 1846
},
{
"epoch": 0.01661172894191291,
"grad_norm": 0.9211652874946594,
"learning_rate": 4.81263592624473e-05,
"loss": 2.2681,
"step": 1847
},
{
"epoch": 0.016620722839553362,
"grad_norm": 1.8441261053085327,
"learning_rate": 4.812350346633655e-05,
"loss": 2.3464,
"step": 1848
},
{
"epoch": 0.016629716737193816,
"grad_norm": 0.8590016961097717,
"learning_rate": 4.812064558034847e-05,
"loss": 2.2257,
"step": 1849
},
{
"epoch": 0.016638710634834266,
"grad_norm": 0.8752726912498474,
"learning_rate": 4.811778560474135e-05,
"loss": 2.231,
"step": 1850
},
{
"epoch": 0.016647704532474716,
"grad_norm": 1.0980899333953857,
"learning_rate": 4.811492353977366e-05,
"loss": 2.3283,
"step": 1851
},
{
"epoch": 0.016656698430115167,
"grad_norm": 1.6323179006576538,
"learning_rate": 4.811205938570408e-05,
"loss": 2.1941,
"step": 1852
},
{
"epoch": 0.016665692327755617,
"grad_norm": 1.0961986780166626,
"learning_rate": 4.810919314279146e-05,
"loss": 2.2665,
"step": 1853
},
{
"epoch": 0.016674686225396067,
"grad_norm": 0.9938058853149414,
"learning_rate": 4.810632481129487e-05,
"loss": 2.281,
"step": 1854
},
{
"epoch": 0.01668368012303652,
"grad_norm": 0.6960379481315613,
"learning_rate": 4.8103454391473514e-05,
"loss": 2.2964,
"step": 1855
},
{
"epoch": 0.01669267402067697,
"grad_norm": 0.9237589240074158,
"learning_rate": 4.810058188358685e-05,
"loss": 2.3102,
"step": 1856
},
{
"epoch": 0.016701667918317422,
"grad_norm": 0.6621735692024231,
"learning_rate": 4.809770728789447e-05,
"loss": 2.2571,
"step": 1857
},
{
"epoch": 0.016710661815957872,
"grad_norm": 0.8308791518211365,
"learning_rate": 4.8094830604656185e-05,
"loss": 2.3512,
"step": 1858
},
{
"epoch": 0.016719655713598323,
"grad_norm": 1.0479060411453247,
"learning_rate": 4.8091951834131984e-05,
"loss": 2.3564,
"step": 1859
},
{
"epoch": 0.016728649611238773,
"grad_norm": 1.0854172706604004,
"learning_rate": 4.8089070976582054e-05,
"loss": 2.2815,
"step": 1860
},
{
"epoch": 0.016737643508879227,
"grad_norm": 1.0308998823165894,
"learning_rate": 4.808618803226675e-05,
"loss": 2.2354,
"step": 1861
},
{
"epoch": 0.016746637406519677,
"grad_norm": 1.7791664600372314,
"learning_rate": 4.808330300144665e-05,
"loss": 2.2741,
"step": 1862
},
{
"epoch": 0.016755631304160128,
"grad_norm": 1.4389320611953735,
"learning_rate": 4.8080415884382474e-05,
"loss": 2.2824,
"step": 1863
},
{
"epoch": 0.016764625201800578,
"grad_norm": 1.377921223640442,
"learning_rate": 4.8077526681335176e-05,
"loss": 2.2444,
"step": 1864
},
{
"epoch": 0.01677361909944103,
"grad_norm": 0.8582924008369446,
"learning_rate": 4.807463539256588e-05,
"loss": 2.2474,
"step": 1865
},
{
"epoch": 0.01678261299708148,
"grad_norm": 0.9905087947845459,
"learning_rate": 4.807174201833589e-05,
"loss": 2.2381,
"step": 1866
},
{
"epoch": 0.016791606894721933,
"grad_norm": 1.1994690895080566,
"learning_rate": 4.8068846558906696e-05,
"loss": 2.2416,
"step": 1867
},
{
"epoch": 0.016800600792362383,
"grad_norm": 0.7634406089782715,
"learning_rate": 4.806594901454001e-05,
"loss": 2.2251,
"step": 1868
},
{
"epoch": 0.016809594690002833,
"grad_norm": 0.8624756336212158,
"learning_rate": 4.806304938549769e-05,
"loss": 2.3237,
"step": 1869
},
{
"epoch": 0.016818588587643284,
"grad_norm": 0.7684710025787354,
"learning_rate": 4.806014767204181e-05,
"loss": 2.1954,
"step": 1870
},
{
"epoch": 0.016827582485283734,
"grad_norm": 1.9442144632339478,
"learning_rate": 4.805724387443462e-05,
"loss": 2.2917,
"step": 1871
},
{
"epoch": 0.016836576382924184,
"grad_norm": 1.1653467416763306,
"learning_rate": 4.8054337992938566e-05,
"loss": 2.243,
"step": 1872
},
{
"epoch": 0.016845570280564638,
"grad_norm": 1.2989853620529175,
"learning_rate": 4.805143002781628e-05,
"loss": 2.2672,
"step": 1873
},
{
"epoch": 0.01685456417820509,
"grad_norm": 0.5724433064460754,
"learning_rate": 4.804851997933057e-05,
"loss": 2.2178,
"step": 1874
},
{
"epoch": 0.01686355807584554,
"grad_norm": 0.9845910668373108,
"learning_rate": 4.8045607847744454e-05,
"loss": 2.303,
"step": 1875
},
{
"epoch": 0.01687255197348599,
"grad_norm": 0.8734267950057983,
"learning_rate": 4.804269363332112e-05,
"loss": 2.2648,
"step": 1876
},
{
"epoch": 0.01688154587112644,
"grad_norm": 0.7489177584648132,
"learning_rate": 4.8039777336323957e-05,
"loss": 2.2902,
"step": 1877
},
{
"epoch": 0.01689053976876689,
"grad_norm": 1.152745246887207,
"learning_rate": 4.803685895701654e-05,
"loss": 2.3479,
"step": 1878
},
{
"epoch": 0.016899533666407344,
"grad_norm": 0.8649763464927673,
"learning_rate": 4.803393849566261e-05,
"loss": 2.2638,
"step": 1879
},
{
"epoch": 0.016908527564047794,
"grad_norm": 0.957438588142395,
"learning_rate": 4.803101595252614e-05,
"loss": 2.3078,
"step": 1880
},
{
"epoch": 0.016917521461688245,
"grad_norm": 0.749431848526001,
"learning_rate": 4.802809132787125e-05,
"loss": 2.1633,
"step": 1881
},
{
"epoch": 0.016926515359328695,
"grad_norm": 0.8619750738143921,
"learning_rate": 4.8025164621962284e-05,
"loss": 2.2075,
"step": 1882
},
{
"epoch": 0.016935509256969145,
"grad_norm": 1.1175601482391357,
"learning_rate": 4.802223583506372e-05,
"loss": 2.2293,
"step": 1883
},
{
"epoch": 0.0169445031546096,
"grad_norm": 1.3970891237258911,
"learning_rate": 4.8019304967440294e-05,
"loss": 2.2848,
"step": 1884
},
{
"epoch": 0.01695349705225005,
"grad_norm": 0.780992329120636,
"learning_rate": 4.801637201935687e-05,
"loss": 2.1801,
"step": 1885
},
{
"epoch": 0.0169624909498905,
"grad_norm": 0.994679868221283,
"learning_rate": 4.801343699107854e-05,
"loss": 2.309,
"step": 1886
},
{
"epoch": 0.01697148484753095,
"grad_norm": 0.9754831194877625,
"learning_rate": 4.801049988287055e-05,
"loss": 2.2761,
"step": 1887
},
{
"epoch": 0.0169804787451714,
"grad_norm": 1.1257673501968384,
"learning_rate": 4.8007560694998375e-05,
"loss": 2.2436,
"step": 1888
},
{
"epoch": 0.01698947264281185,
"grad_norm": 0.8542078733444214,
"learning_rate": 4.800461942772765e-05,
"loss": 2.2935,
"step": 1889
},
{
"epoch": 0.016998466540452305,
"grad_norm": 0.7948276400566101,
"learning_rate": 4.80016760813242e-05,
"loss": 2.2553,
"step": 1890
},
{
"epoch": 0.017007460438092755,
"grad_norm": 1.023480772972107,
"learning_rate": 4.7998730656054035e-05,
"loss": 2.2589,
"step": 1891
},
{
"epoch": 0.017016454335733205,
"grad_norm": 0.5797092914581299,
"learning_rate": 4.7995783152183366e-05,
"loss": 2.272,
"step": 1892
},
{
"epoch": 0.017025448233373656,
"grad_norm": 0.8182694315910339,
"learning_rate": 4.799283356997859e-05,
"loss": 2.2563,
"step": 1893
},
{
"epoch": 0.017034442131014106,
"grad_norm": 1.257016658782959,
"learning_rate": 4.798988190970628e-05,
"loss": 2.2578,
"step": 1894
},
{
"epoch": 0.017043436028654557,
"grad_norm": 1.1566047668457031,
"learning_rate": 4.79869281716332e-05,
"loss": 2.2644,
"step": 1895
},
{
"epoch": 0.01705242992629501,
"grad_norm": 1.223054051399231,
"learning_rate": 4.7983972356026315e-05,
"loss": 2.3659,
"step": 1896
},
{
"epoch": 0.01706142382393546,
"grad_norm": 0.6454828977584839,
"learning_rate": 4.798101446315276e-05,
"loss": 2.2161,
"step": 1897
},
{
"epoch": 0.01707041772157591,
"grad_norm": 0.9303452968597412,
"learning_rate": 4.797805449327988e-05,
"loss": 2.2114,
"step": 1898
},
{
"epoch": 0.01707941161921636,
"grad_norm": 1.0017653703689575,
"learning_rate": 4.797509244667518e-05,
"loss": 2.2505,
"step": 1899
},
{
"epoch": 0.017088405516856812,
"grad_norm": 0.8761176466941833,
"learning_rate": 4.797212832360637e-05,
"loss": 2.3295,
"step": 1900
},
{
"epoch": 0.017097399414497262,
"grad_norm": 1.3909958600997925,
"learning_rate": 4.796916212434135e-05,
"loss": 2.2344,
"step": 1901
},
{
"epoch": 0.017106393312137716,
"grad_norm": 1.188353419303894,
"learning_rate": 4.7966193849148194e-05,
"loss": 2.3737,
"step": 1902
},
{
"epoch": 0.017115387209778166,
"grad_norm": 0.9582924842834473,
"learning_rate": 4.7963223498295176e-05,
"loss": 2.2416,
"step": 1903
},
{
"epoch": 0.017124381107418617,
"grad_norm": 0.6712927222251892,
"learning_rate": 4.796025107205075e-05,
"loss": 2.2281,
"step": 1904
},
{
"epoch": 0.017133375005059067,
"grad_norm": 1.0080678462982178,
"learning_rate": 4.7957276570683566e-05,
"loss": 2.2302,
"step": 1905
},
{
"epoch": 0.017142368902699517,
"grad_norm": 0.8767439126968384,
"learning_rate": 4.7954299994462454e-05,
"loss": 2.3459,
"step": 1906
},
{
"epoch": 0.017151362800339968,
"grad_norm": 1.0786638259887695,
"learning_rate": 4.795132134365643e-05,
"loss": 2.2993,
"step": 1907
},
{
"epoch": 0.01716035669798042,
"grad_norm": 1.2587511539459229,
"learning_rate": 4.79483406185347e-05,
"loss": 2.2827,
"step": 1908
},
{
"epoch": 0.017169350595620872,
"grad_norm": 1.4085206985473633,
"learning_rate": 4.7945357819366675e-05,
"loss": 2.2252,
"step": 1909
},
{
"epoch": 0.017178344493261322,
"grad_norm": 1.057995080947876,
"learning_rate": 4.794237294642192e-05,
"loss": 2.3504,
"step": 1910
},
{
"epoch": 0.017187338390901773,
"grad_norm": 1.0978472232818604,
"learning_rate": 4.793938599997021e-05,
"loss": 2.3538,
"step": 1911
},
{
"epoch": 0.017196332288542223,
"grad_norm": 0.9442919492721558,
"learning_rate": 4.7936396980281496e-05,
"loss": 2.3188,
"step": 1912
},
{
"epoch": 0.017205326186182673,
"grad_norm": 0.8239895105361938,
"learning_rate": 4.793340588762593e-05,
"loss": 2.2172,
"step": 1913
},
{
"epoch": 0.017214320083823127,
"grad_norm": 1.4184783697128296,
"learning_rate": 4.793041272227384e-05,
"loss": 2.283,
"step": 1914
},
{
"epoch": 0.017223313981463578,
"grad_norm": 0.6340658068656921,
"learning_rate": 4.792741748449575e-05,
"loss": 2.1854,
"step": 1915
},
{
"epoch": 0.017232307879104028,
"grad_norm": 1.1299843788146973,
"learning_rate": 4.792442017456237e-05,
"loss": 2.3031,
"step": 1916
},
{
"epoch": 0.01724130177674448,
"grad_norm": 1.4297564029693604,
"learning_rate": 4.792142079274457e-05,
"loss": 2.2117,
"step": 1917
},
{
"epoch": 0.01725029567438493,
"grad_norm": 0.6740981340408325,
"learning_rate": 4.791841933931346e-05,
"loss": 2.1921,
"step": 1918
},
{
"epoch": 0.017259289572025382,
"grad_norm": 1.3782836198806763,
"learning_rate": 4.7915415814540296e-05,
"loss": 2.2988,
"step": 1919
},
{
"epoch": 0.017268283469665833,
"grad_norm": 2.1514110565185547,
"learning_rate": 4.791241021869653e-05,
"loss": 2.292,
"step": 1920
},
{
"epoch": 0.017277277367306283,
"grad_norm": 0.9755523204803467,
"learning_rate": 4.79094025520538e-05,
"loss": 2.1852,
"step": 1921
},
{
"epoch": 0.017286271264946734,
"grad_norm": 0.5889203548431396,
"learning_rate": 4.790639281488396e-05,
"loss": 2.1865,
"step": 1922
},
{
"epoch": 0.017295265162587184,
"grad_norm": 0.8238359093666077,
"learning_rate": 4.7903381007459e-05,
"loss": 2.2827,
"step": 1923
},
{
"epoch": 0.017304259060227634,
"grad_norm": 1.3062047958374023,
"learning_rate": 4.790036713005114e-05,
"loss": 2.3012,
"step": 1924
},
{
"epoch": 0.017313252957868088,
"grad_norm": 0.6288627982139587,
"learning_rate": 4.7897351182932756e-05,
"loss": 2.1774,
"step": 1925
},
{
"epoch": 0.01732224685550854,
"grad_norm": 1.2543439865112305,
"learning_rate": 4.789433316637644e-05,
"loss": 2.3061,
"step": 1926
},
{
"epoch": 0.01733124075314899,
"grad_norm": 1.6748862266540527,
"learning_rate": 4.7891313080654954e-05,
"loss": 2.2744,
"step": 1927
},
{
"epoch": 0.01734023465078944,
"grad_norm": 1.1074962615966797,
"learning_rate": 4.788829092604124e-05,
"loss": 2.2932,
"step": 1928
},
{
"epoch": 0.01734922854842989,
"grad_norm": 1.0261276960372925,
"learning_rate": 4.788526670280845e-05,
"loss": 2.254,
"step": 1929
},
{
"epoch": 0.01735822244607034,
"grad_norm": 1.3492799997329712,
"learning_rate": 4.7882240411229905e-05,
"loss": 2.3212,
"step": 1930
},
{
"epoch": 0.017367216343710794,
"grad_norm": 0.8169590830802917,
"learning_rate": 4.787921205157912e-05,
"loss": 2.1939,
"step": 1931
},
{
"epoch": 0.017376210241351244,
"grad_norm": 2.293365240097046,
"learning_rate": 4.78761816241298e-05,
"loss": 2.3158,
"step": 1932
},
{
"epoch": 0.017385204138991694,
"grad_norm": 0.652710497379303,
"learning_rate": 4.7873149129155816e-05,
"loss": 2.2172,
"step": 1933
},
{
"epoch": 0.017394198036632145,
"grad_norm": 0.9734316468238831,
"learning_rate": 4.787011456693125e-05,
"loss": 2.2232,
"step": 1934
},
{
"epoch": 0.017403191934272595,
"grad_norm": 1.9363571405410767,
"learning_rate": 4.786707793773037e-05,
"loss": 2.3442,
"step": 1935
},
{
"epoch": 0.017412185831913046,
"grad_norm": 0.9665070176124573,
"learning_rate": 4.786403924182761e-05,
"loss": 2.2754,
"step": 1936
},
{
"epoch": 0.0174211797295535,
"grad_norm": 1.055876612663269,
"learning_rate": 4.786099847949761e-05,
"loss": 2.2133,
"step": 1937
},
{
"epoch": 0.01743017362719395,
"grad_norm": 1.0245312452316284,
"learning_rate": 4.78579556510152e-05,
"loss": 2.2364,
"step": 1938
},
{
"epoch": 0.0174391675248344,
"grad_norm": 0.9658739566802979,
"learning_rate": 4.785491075665537e-05,
"loss": 2.124,
"step": 1939
},
{
"epoch": 0.01744816142247485,
"grad_norm": 1.2974826097488403,
"learning_rate": 4.7851863796693326e-05,
"loss": 2.2864,
"step": 1940
},
{
"epoch": 0.0174571553201153,
"grad_norm": 1.042313814163208,
"learning_rate": 4.784881477140445e-05,
"loss": 2.2755,
"step": 1941
},
{
"epoch": 0.01746614921775575,
"grad_norm": 1.1423113346099854,
"learning_rate": 4.7845763681064296e-05,
"loss": 2.2599,
"step": 1942
},
{
"epoch": 0.017475143115396205,
"grad_norm": 0.994827151298523,
"learning_rate": 4.784271052594863e-05,
"loss": 2.2505,
"step": 1943
},
{
"epoch": 0.017484137013036655,
"grad_norm": 1.4856171607971191,
"learning_rate": 4.7839655306333395e-05,
"loss": 2.2876,
"step": 1944
},
{
"epoch": 0.017493130910677106,
"grad_norm": 1.0082606077194214,
"learning_rate": 4.783659802249472e-05,
"loss": 2.315,
"step": 1945
},
{
"epoch": 0.017502124808317556,
"grad_norm": 1.24881911277771,
"learning_rate": 4.7833538674708905e-05,
"loss": 2.3216,
"step": 1946
},
{
"epoch": 0.017511118705958006,
"grad_norm": 0.852787435054779,
"learning_rate": 4.783047726325246e-05,
"loss": 2.1809,
"step": 1947
},
{
"epoch": 0.017520112603598457,
"grad_norm": 0.8327879905700684,
"learning_rate": 4.7827413788402077e-05,
"loss": 2.2645,
"step": 1948
},
{
"epoch": 0.01752910650123891,
"grad_norm": 1.182180643081665,
"learning_rate": 4.7824348250434615e-05,
"loss": 2.3767,
"step": 1949
},
{
"epoch": 0.01753810039887936,
"grad_norm": 1.3073945045471191,
"learning_rate": 4.782128064962714e-05,
"loss": 2.284,
"step": 1950
},
{
"epoch": 0.01754709429651981,
"grad_norm": 0.8969705700874329,
"learning_rate": 4.78182109862569e-05,
"loss": 2.2808,
"step": 1951
},
{
"epoch": 0.01755608819416026,
"grad_norm": 1.233926773071289,
"learning_rate": 4.781513926060135e-05,
"loss": 2.2481,
"step": 1952
},
{
"epoch": 0.017565082091800712,
"grad_norm": 0.8716261982917786,
"learning_rate": 4.7812065472938075e-05,
"loss": 2.1913,
"step": 1953
},
{
"epoch": 0.017574075989441162,
"grad_norm": 1.4345555305480957,
"learning_rate": 4.780898962354489e-05,
"loss": 2.1927,
"step": 1954
},
{
"epoch": 0.017583069887081616,
"grad_norm": 0.8084832429885864,
"learning_rate": 4.780591171269979e-05,
"loss": 2.3321,
"step": 1955
},
{
"epoch": 0.017592063784722067,
"grad_norm": 0.8963560461997986,
"learning_rate": 4.7802831740680955e-05,
"loss": 2.3104,
"step": 1956
},
{
"epoch": 0.017601057682362517,
"grad_norm": 0.7294359803199768,
"learning_rate": 4.779974970776675e-05,
"loss": 2.2292,
"step": 1957
},
{
"epoch": 0.017610051580002967,
"grad_norm": 0.9283105134963989,
"learning_rate": 4.779666561423573e-05,
"loss": 2.3325,
"step": 1958
},
{
"epoch": 0.017619045477643418,
"grad_norm": 1.5552716255187988,
"learning_rate": 4.779357946036661e-05,
"loss": 2.307,
"step": 1959
},
{
"epoch": 0.01762803937528387,
"grad_norm": 1.5735383033752441,
"learning_rate": 4.7790491246438345e-05,
"loss": 2.2355,
"step": 1960
},
{
"epoch": 0.017637033272924322,
"grad_norm": 1.307526707649231,
"learning_rate": 4.7787400972730024e-05,
"loss": 2.2477,
"step": 1961
},
{
"epoch": 0.017646027170564772,
"grad_norm": 1.1407091617584229,
"learning_rate": 4.7784308639520945e-05,
"loss": 2.2384,
"step": 1962
},
{
"epoch": 0.017655021068205223,
"grad_norm": 1.6965885162353516,
"learning_rate": 4.77812142470906e-05,
"loss": 2.1858,
"step": 1963
},
{
"epoch": 0.017664014965845673,
"grad_norm": 1.4773070812225342,
"learning_rate": 4.777811779571864e-05,
"loss": 2.2493,
"step": 1964
},
{
"epoch": 0.017673008863486123,
"grad_norm": 1.2642821073532104,
"learning_rate": 4.777501928568494e-05,
"loss": 2.2737,
"step": 1965
},
{
"epoch": 0.017682002761126577,
"grad_norm": 1.8974300622940063,
"learning_rate": 4.777191871726952e-05,
"loss": 2.1704,
"step": 1966
},
{
"epoch": 0.017690996658767028,
"grad_norm": 4.7509870529174805,
"learning_rate": 4.7768816090752614e-05,
"loss": 2.3157,
"step": 1967
},
{
"epoch": 0.017699990556407478,
"grad_norm": 1.0007964372634888,
"learning_rate": 4.776571140641464e-05,
"loss": 2.2045,
"step": 1968
},
{
"epoch": 0.017708984454047928,
"grad_norm": 1.1411465406417847,
"learning_rate": 4.776260466453618e-05,
"loss": 2.2924,
"step": 1969
},
{
"epoch": 0.01771797835168838,
"grad_norm": 0.6316106915473938,
"learning_rate": 4.775949586539803e-05,
"loss": 2.1816,
"step": 1970
},
{
"epoch": 0.01772697224932883,
"grad_norm": 0.6891862750053406,
"learning_rate": 4.7756385009281165e-05,
"loss": 2.2516,
"step": 1971
},
{
"epoch": 0.017735966146969283,
"grad_norm": 1.0875741243362427,
"learning_rate": 4.775327209646673e-05,
"loss": 2.1976,
"step": 1972
},
{
"epoch": 0.017744960044609733,
"grad_norm": 1.7854807376861572,
"learning_rate": 4.775015712723607e-05,
"loss": 2.3174,
"step": 1973
},
{
"epoch": 0.017753953942250184,
"grad_norm": 1.4643340110778809,
"learning_rate": 4.774704010187072e-05,
"loss": 2.2586,
"step": 1974
},
{
"epoch": 0.017762947839890634,
"grad_norm": 0.8734381794929504,
"learning_rate": 4.7743921020652384e-05,
"loss": 2.2768,
"step": 1975
},
{
"epoch": 0.017771941737531084,
"grad_norm": 0.7068848013877869,
"learning_rate": 4.774079988386296e-05,
"loss": 2.2247,
"step": 1976
},
{
"epoch": 0.017780935635171535,
"grad_norm": 1.885035753250122,
"learning_rate": 4.773767669178454e-05,
"loss": 2.3304,
"step": 1977
},
{
"epoch": 0.01778992953281199,
"grad_norm": 1.7003952264785767,
"learning_rate": 4.773455144469939e-05,
"loss": 2.2169,
"step": 1978
},
{
"epoch": 0.01779892343045244,
"grad_norm": 0.9710990786552429,
"learning_rate": 4.7731424142889966e-05,
"loss": 2.2928,
"step": 1979
},
{
"epoch": 0.01780791732809289,
"grad_norm": 1.119415044784546,
"learning_rate": 4.7728294786638914e-05,
"loss": 2.2104,
"step": 1980
},
{
"epoch": 0.01781691122573334,
"grad_norm": 1.5172535181045532,
"learning_rate": 4.7725163376229064e-05,
"loss": 2.2277,
"step": 1981
},
{
"epoch": 0.01782590512337379,
"grad_norm": 1.0494482517242432,
"learning_rate": 4.772202991194342e-05,
"loss": 2.2825,
"step": 1982
},
{
"epoch": 0.01783489902101424,
"grad_norm": 3.9221458435058594,
"learning_rate": 4.77188943940652e-05,
"loss": 2.1774,
"step": 1983
},
{
"epoch": 0.017843892918654694,
"grad_norm": 1.619274377822876,
"learning_rate": 4.771575682287776e-05,
"loss": 2.2868,
"step": 1984
},
{
"epoch": 0.017852886816295144,
"grad_norm": 2.6348698139190674,
"learning_rate": 4.771261719866469e-05,
"loss": 2.3313,
"step": 1985
},
{
"epoch": 0.017861880713935595,
"grad_norm": 1.3733265399932861,
"learning_rate": 4.770947552170974e-05,
"loss": 2.1807,
"step": 1986
},
{
"epoch": 0.017870874611576045,
"grad_norm": 1.2614948749542236,
"learning_rate": 4.770633179229686e-05,
"loss": 2.279,
"step": 1987
},
{
"epoch": 0.017879868509216496,
"grad_norm": 1.1199651956558228,
"learning_rate": 4.770318601071016e-05,
"loss": 2.3157,
"step": 1988
},
{
"epoch": 0.017888862406856946,
"grad_norm": 0.6340201497077942,
"learning_rate": 4.7700038177233965e-05,
"loss": 2.2394,
"step": 1989
},
{
"epoch": 0.0178978563044974,
"grad_norm": 1.096011996269226,
"learning_rate": 4.769688829215278e-05,
"loss": 2.2629,
"step": 1990
},
{
"epoch": 0.01790685020213785,
"grad_norm": 0.7549166679382324,
"learning_rate": 4.769373635575126e-05,
"loss": 2.2123,
"step": 1991
},
{
"epoch": 0.0179158440997783,
"grad_norm": 1.0398750305175781,
"learning_rate": 4.7690582368314304e-05,
"loss": 2.2757,
"step": 1992
},
{
"epoch": 0.01792483799741875,
"grad_norm": 0.9786906242370605,
"learning_rate": 4.768742633012695e-05,
"loss": 2.295,
"step": 1993
},
{
"epoch": 0.0179338318950592,
"grad_norm": 1.1991000175476074,
"learning_rate": 4.768426824147445e-05,
"loss": 2.2628,
"step": 1994
},
{
"epoch": 0.017942825792699655,
"grad_norm": 1.0576684474945068,
"learning_rate": 4.768110810264221e-05,
"loss": 2.2412,
"step": 1995
},
{
"epoch": 0.017951819690340105,
"grad_norm": 0.6487658619880676,
"learning_rate": 4.767794591391585e-05,
"loss": 2.2332,
"step": 1996
},
{
"epoch": 0.017960813587980556,
"grad_norm": 0.7374446988105774,
"learning_rate": 4.767478167558117e-05,
"loss": 2.3014,
"step": 1997
},
{
"epoch": 0.017969807485621006,
"grad_norm": 1.2109500169754028,
"learning_rate": 4.767161538792414e-05,
"loss": 2.249,
"step": 1998
},
{
"epoch": 0.017978801383261456,
"grad_norm": 0.9672105312347412,
"learning_rate": 4.766844705123093e-05,
"loss": 2.3134,
"step": 1999
},
{
"epoch": 0.017987795280901907,
"grad_norm": 0.717119038105011,
"learning_rate": 4.766527666578791e-05,
"loss": 2.2366,
"step": 2000
}
],
"logging_steps": 1.0,
"max_steps": 11000,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2.3767912808448e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}