alphamath_round2 / trainer_state.json
jayoohwang's picture
Upload folder using huggingface_hub
549d1f7 verified
Raw
History Blame Contribute Delete
160 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 7.677543186180422,
"eval_steps": 500,
"global_step": 1000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"grad_norm": 2.164377212524414,
"learning_rate": 1.2121212121212122e-06,
"loss": 0.3481,
"step": 1
},
{
"epoch": 0.02,
"grad_norm": 2.2210371494293213,
"learning_rate": 2.4242424242424244e-06,
"loss": 0.3518,
"step": 2
},
{
"epoch": 0.02,
"grad_norm": 2.2581589221954346,
"learning_rate": 3.6363636363636366e-06,
"loss": 0.3417,
"step": 3
},
{
"epoch": 0.03,
"grad_norm": 2.166820764541626,
"learning_rate": 4.848484848484849e-06,
"loss": 0.3522,
"step": 4
},
{
"epoch": 0.04,
"grad_norm": 2.031558036804199,
"learning_rate": 6.060606060606061e-06,
"loss": 0.3281,
"step": 5
},
{
"epoch": 0.05,
"grad_norm": 2.1338369846343994,
"learning_rate": 7.272727272727273e-06,
"loss": 0.3362,
"step": 6
},
{
"epoch": 0.05,
"grad_norm": 1.8426257371902466,
"learning_rate": 8.484848484848486e-06,
"loss": 0.326,
"step": 7
},
{
"epoch": 0.06,
"grad_norm": 2.0699379444122314,
"learning_rate": 9.696969696969698e-06,
"loss": 0.2957,
"step": 8
},
{
"epoch": 0.07,
"grad_norm": 1.3295575380325317,
"learning_rate": 1.0909090909090909e-05,
"loss": 0.2559,
"step": 9
},
{
"epoch": 0.08,
"grad_norm": 1.1346920728683472,
"learning_rate": 1.2121212121212122e-05,
"loss": 0.2361,
"step": 10
},
{
"epoch": 0.08,
"grad_norm": 0.9885814785957336,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.2214,
"step": 11
},
{
"epoch": 0.09,
"grad_norm": 0.7244998812675476,
"learning_rate": 1.4545454545454546e-05,
"loss": 0.1963,
"step": 12
},
{
"epoch": 0.1,
"grad_norm": 0.5507056713104248,
"learning_rate": 1.575757575757576e-05,
"loss": 0.1835,
"step": 13
},
{
"epoch": 0.11,
"grad_norm": 0.46277302503585815,
"learning_rate": 1.6969696969696972e-05,
"loss": 0.1763,
"step": 14
},
{
"epoch": 0.12,
"grad_norm": 0.5373420119285583,
"learning_rate": 1.8181818181818182e-05,
"loss": 0.1851,
"step": 15
},
{
"epoch": 0.12,
"grad_norm": 0.5908621549606323,
"learning_rate": 1.9393939393939395e-05,
"loss": 0.1839,
"step": 16
},
{
"epoch": 0.13,
"grad_norm": 0.4572305381298065,
"learning_rate": 2.0606060606060608e-05,
"loss": 0.1791,
"step": 17
},
{
"epoch": 0.14,
"grad_norm": 0.4026539921760559,
"learning_rate": 2.1818181818181818e-05,
"loss": 0.1862,
"step": 18
},
{
"epoch": 0.15,
"grad_norm": 0.3413575291633606,
"learning_rate": 2.3030303030303034e-05,
"loss": 0.1784,
"step": 19
},
{
"epoch": 0.15,
"grad_norm": 0.4631746709346771,
"learning_rate": 2.4242424242424244e-05,
"loss": 0.1812,
"step": 20
},
{
"epoch": 0.16,
"grad_norm": 0.4660058319568634,
"learning_rate": 2.5454545454545457e-05,
"loss": 0.1862,
"step": 21
},
{
"epoch": 0.17,
"grad_norm": 0.3439246118068695,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.1771,
"step": 22
},
{
"epoch": 0.18,
"grad_norm": 0.30352091789245605,
"learning_rate": 2.7878787878787883e-05,
"loss": 0.1743,
"step": 23
},
{
"epoch": 0.18,
"grad_norm": 0.47577032446861267,
"learning_rate": 2.9090909090909093e-05,
"loss": 0.1728,
"step": 24
},
{
"epoch": 0.19,
"grad_norm": 0.2906971275806427,
"learning_rate": 3.0303030303030306e-05,
"loss": 0.1771,
"step": 25
},
{
"epoch": 0.2,
"grad_norm": 0.26286426186561584,
"learning_rate": 3.151515151515152e-05,
"loss": 0.1617,
"step": 26
},
{
"epoch": 0.21,
"grad_norm": 0.2660459280014038,
"learning_rate": 3.272727272727273e-05,
"loss": 0.1612,
"step": 27
},
{
"epoch": 0.21,
"grad_norm": 0.2895466685295105,
"learning_rate": 3.3939393939393945e-05,
"loss": 0.1622,
"step": 28
},
{
"epoch": 0.22,
"grad_norm": 0.2738928198814392,
"learning_rate": 3.515151515151515e-05,
"loss": 0.1603,
"step": 29
},
{
"epoch": 0.23,
"grad_norm": 0.2555297613143921,
"learning_rate": 3.6363636363636364e-05,
"loss": 0.1626,
"step": 30
},
{
"epoch": 0.24,
"grad_norm": 0.25712355971336365,
"learning_rate": 3.7575757575757584e-05,
"loss": 0.1641,
"step": 31
},
{
"epoch": 0.25,
"grad_norm": 0.26491039991378784,
"learning_rate": 3.878787878787879e-05,
"loss": 0.1575,
"step": 32
},
{
"epoch": 0.25,
"grad_norm": 0.2663068473339081,
"learning_rate": 4e-05,
"loss": 0.1519,
"step": 33
},
{
"epoch": 0.26,
"grad_norm": 0.25683528184890747,
"learning_rate": 3.9999902671405465e-05,
"loss": 0.1527,
"step": 34
},
{
"epoch": 0.27,
"grad_norm": 0.2544044852256775,
"learning_rate": 3.999961068656914e-05,
"loss": 0.15,
"step": 35
},
{
"epoch": 0.28,
"grad_norm": 0.2390272617340088,
"learning_rate": 3.999912404833288e-05,
"loss": 0.1472,
"step": 36
},
{
"epoch": 0.28,
"grad_norm": 0.23785777390003204,
"learning_rate": 3.999844276143305e-05,
"loss": 0.1491,
"step": 37
},
{
"epoch": 0.29,
"grad_norm": 0.2470083385705948,
"learning_rate": 3.999756683250054e-05,
"loss": 0.1484,
"step": 38
},
{
"epoch": 0.3,
"grad_norm": 0.22296255826950073,
"learning_rate": 3.999649627006063e-05,
"loss": 0.1445,
"step": 39
},
{
"epoch": 0.31,
"grad_norm": 0.28281041979789734,
"learning_rate": 3.9995231084532964e-05,
"loss": 0.1406,
"step": 40
},
{
"epoch": 0.31,
"grad_norm": 0.26012447476387024,
"learning_rate": 3.99937712882314e-05,
"loss": 0.1397,
"step": 41
},
{
"epoch": 0.32,
"grad_norm": 0.23696556687355042,
"learning_rate": 3.9992116895363946e-05,
"loss": 0.1435,
"step": 42
},
{
"epoch": 0.33,
"grad_norm": 0.2600996494293213,
"learning_rate": 3.999026792203256e-05,
"loss": 0.1404,
"step": 43
},
{
"epoch": 0.34,
"grad_norm": 0.2460625171661377,
"learning_rate": 3.9988224386233055e-05,
"loss": 0.1437,
"step": 44
},
{
"epoch": 0.35,
"grad_norm": 0.2851013243198395,
"learning_rate": 3.9985986307854865e-05,
"loss": 0.1396,
"step": 45
},
{
"epoch": 0.35,
"grad_norm": 0.30484041571617126,
"learning_rate": 3.99835537086809e-05,
"loss": 0.1347,
"step": 46
},
{
"epoch": 0.36,
"grad_norm": 0.24361257255077362,
"learning_rate": 3.9980926612387304e-05,
"loss": 0.1318,
"step": 47
},
{
"epoch": 0.37,
"grad_norm": 0.23862454295158386,
"learning_rate": 3.997810504454324e-05,
"loss": 0.1371,
"step": 48
},
{
"epoch": 0.38,
"grad_norm": 0.27349838614463806,
"learning_rate": 3.9975089032610626e-05,
"loss": 0.1358,
"step": 49
},
{
"epoch": 0.38,
"grad_norm": 0.23924770951271057,
"learning_rate": 3.997187860594387e-05,
"loss": 0.1336,
"step": 50
},
{
"epoch": 0.39,
"grad_norm": 0.2943211495876312,
"learning_rate": 3.996847379578963e-05,
"loss": 0.1301,
"step": 51
},
{
"epoch": 0.4,
"grad_norm": 0.23755450546741486,
"learning_rate": 3.9964874635286436e-05,
"loss": 0.1271,
"step": 52
},
{
"epoch": 0.41,
"grad_norm": 0.23102109134197235,
"learning_rate": 3.9961081159464395e-05,
"loss": 0.1332,
"step": 53
},
{
"epoch": 0.41,
"grad_norm": 0.26369133591651917,
"learning_rate": 3.995709340524489e-05,
"loss": 0.1339,
"step": 54
},
{
"epoch": 0.42,
"grad_norm": 0.221548929810524,
"learning_rate": 3.995291141144017e-05,
"loss": 0.1267,
"step": 55
},
{
"epoch": 0.43,
"grad_norm": 0.24895070493221283,
"learning_rate": 3.994853521875299e-05,
"loss": 0.1283,
"step": 56
},
{
"epoch": 0.44,
"grad_norm": 0.2894270122051239,
"learning_rate": 3.994396486977621e-05,
"loss": 0.1289,
"step": 57
},
{
"epoch": 0.45,
"grad_norm": 0.28780320286750793,
"learning_rate": 3.993920040899242e-05,
"loss": 0.1322,
"step": 58
},
{
"epoch": 0.45,
"grad_norm": 0.2660049498081207,
"learning_rate": 3.9934241882773414e-05,
"loss": 0.1246,
"step": 59
},
{
"epoch": 0.46,
"grad_norm": 0.23252670466899872,
"learning_rate": 3.992908933937986e-05,
"loss": 0.1225,
"step": 60
},
{
"epoch": 0.47,
"grad_norm": 0.2243393510580063,
"learning_rate": 3.992374282896072e-05,
"loss": 0.1261,
"step": 61
},
{
"epoch": 0.48,
"grad_norm": 0.22698526084423065,
"learning_rate": 3.991820240355284e-05,
"loss": 0.1326,
"step": 62
},
{
"epoch": 0.48,
"grad_norm": 0.2564685642719269,
"learning_rate": 3.991246811708039e-05,
"loss": 0.1241,
"step": 63
},
{
"epoch": 0.49,
"grad_norm": 0.23229768872261047,
"learning_rate": 3.990654002535439e-05,
"loss": 0.1225,
"step": 64
},
{
"epoch": 0.5,
"grad_norm": 0.23054753243923187,
"learning_rate": 3.99004181860721e-05,
"loss": 0.1175,
"step": 65
},
{
"epoch": 0.51,
"grad_norm": 0.2098405361175537,
"learning_rate": 3.989410265881655e-05,
"loss": 0.1191,
"step": 66
},
{
"epoch": 0.51,
"grad_norm": 0.25725990533828735,
"learning_rate": 3.9887593505055865e-05,
"loss": 0.1195,
"step": 67
},
{
"epoch": 0.52,
"grad_norm": 0.21651968359947205,
"learning_rate": 3.9880890788142716e-05,
"loss": 0.1201,
"step": 68
},
{
"epoch": 0.53,
"grad_norm": 0.2460426688194275,
"learning_rate": 3.987399457331371e-05,
"loss": 0.1182,
"step": 69
},
{
"epoch": 0.54,
"grad_norm": 0.24313770234584808,
"learning_rate": 3.986690492768875e-05,
"loss": 0.1156,
"step": 70
},
{
"epoch": 0.55,
"grad_norm": 0.2558782994747162,
"learning_rate": 3.985962192027034e-05,
"loss": 0.1203,
"step": 71
},
{
"epoch": 0.55,
"grad_norm": 0.24763956665992737,
"learning_rate": 3.985214562194299e-05,
"loss": 0.119,
"step": 72
},
{
"epoch": 0.56,
"grad_norm": 0.2278430312871933,
"learning_rate": 3.984447610547243e-05,
"loss": 0.1175,
"step": 73
},
{
"epoch": 0.57,
"grad_norm": 0.2235199362039566,
"learning_rate": 3.9836613445505015e-05,
"loss": 0.1198,
"step": 74
},
{
"epoch": 0.58,
"grad_norm": 0.25674423575401306,
"learning_rate": 3.982855771856689e-05,
"loss": 0.1153,
"step": 75
},
{
"epoch": 0.58,
"grad_norm": 0.25616034865379333,
"learning_rate": 3.982030900306333e-05,
"loss": 0.1143,
"step": 76
},
{
"epoch": 0.59,
"grad_norm": 0.2404707819223404,
"learning_rate": 3.9811867379277915e-05,
"loss": 0.1163,
"step": 77
},
{
"epoch": 0.6,
"grad_norm": 0.24204260110855103,
"learning_rate": 3.980323292937178e-05,
"loss": 0.1138,
"step": 78
},
{
"epoch": 0.61,
"grad_norm": 0.26231199502944946,
"learning_rate": 3.979440573738282e-05,
"loss": 0.1133,
"step": 79
},
{
"epoch": 0.61,
"grad_norm": 0.26181137561798096,
"learning_rate": 3.978538588922484e-05,
"loss": 0.1095,
"step": 80
},
{
"epoch": 0.62,
"grad_norm": 0.2599407434463501,
"learning_rate": 3.977617347268677e-05,
"loss": 0.1165,
"step": 81
},
{
"epoch": 0.63,
"grad_norm": 0.25307539105415344,
"learning_rate": 3.976676857743175e-05,
"loss": 0.1105,
"step": 82
},
{
"epoch": 0.64,
"grad_norm": 0.2505055069923401,
"learning_rate": 3.975717129499632e-05,
"loss": 0.1154,
"step": 83
},
{
"epoch": 0.64,
"grad_norm": 0.21991747617721558,
"learning_rate": 3.9747381718789475e-05,
"loss": 0.109,
"step": 84
},
{
"epoch": 0.65,
"grad_norm": 0.21642076969146729,
"learning_rate": 3.9737399944091784e-05,
"loss": 0.1118,
"step": 85
},
{
"epoch": 0.66,
"grad_norm": 0.28659364581108093,
"learning_rate": 3.972722606805445e-05,
"loss": 0.1072,
"step": 86
},
{
"epoch": 0.67,
"grad_norm": 0.20825989544391632,
"learning_rate": 3.971686018969839e-05,
"loss": 0.1095,
"step": 87
},
{
"epoch": 0.68,
"grad_norm": 0.2419530600309372,
"learning_rate": 3.9706302409913235e-05,
"loss": 0.109,
"step": 88
},
{
"epoch": 0.68,
"grad_norm": 0.27549922466278076,
"learning_rate": 3.969555283145637e-05,
"loss": 0.1132,
"step": 89
},
{
"epoch": 0.69,
"grad_norm": 0.24143557250499725,
"learning_rate": 3.968461155895194e-05,
"loss": 0.1104,
"step": 90
},
{
"epoch": 0.7,
"grad_norm": 0.22690293192863464,
"learning_rate": 3.96734786988898e-05,
"loss": 0.1102,
"step": 91
},
{
"epoch": 0.71,
"grad_norm": 0.24802416563034058,
"learning_rate": 3.966215435962452e-05,
"loss": 0.1127,
"step": 92
},
{
"epoch": 0.71,
"grad_norm": 0.21746200323104858,
"learning_rate": 3.965063865137431e-05,
"loss": 0.11,
"step": 93
},
{
"epoch": 0.72,
"grad_norm": 0.2443949431180954,
"learning_rate": 3.963893168621993e-05,
"loss": 0.1027,
"step": 94
},
{
"epoch": 0.73,
"grad_norm": 0.25307193398475647,
"learning_rate": 3.962703357810362e-05,
"loss": 0.1071,
"step": 95
},
{
"epoch": 0.74,
"grad_norm": 0.2298382967710495,
"learning_rate": 3.9614944442828004e-05,
"loss": 0.1087,
"step": 96
},
{
"epoch": 0.74,
"grad_norm": 0.28927159309387207,
"learning_rate": 3.9602664398054935e-05,
"loss": 0.107,
"step": 97
},
{
"epoch": 0.75,
"grad_norm": 0.24768786132335663,
"learning_rate": 3.959019356330437e-05,
"loss": 0.1037,
"step": 98
},
{
"epoch": 0.76,
"grad_norm": 0.21881985664367676,
"learning_rate": 3.9577532059953174e-05,
"loss": 0.1053,
"step": 99
},
{
"epoch": 0.77,
"grad_norm": 0.2732795774936676,
"learning_rate": 3.956468001123399e-05,
"loss": 0.1104,
"step": 100
},
{
"epoch": 0.78,
"grad_norm": 0.25283458828926086,
"learning_rate": 3.9551637542234e-05,
"loss": 0.0992,
"step": 101
},
{
"epoch": 0.78,
"grad_norm": 0.216854527592659,
"learning_rate": 3.9538404779893726e-05,
"loss": 0.0992,
"step": 102
},
{
"epoch": 0.79,
"grad_norm": 0.2594543397426605,
"learning_rate": 3.952498185300578e-05,
"loss": 0.1026,
"step": 103
},
{
"epoch": 0.8,
"grad_norm": 0.2466372549533844,
"learning_rate": 3.9511368892213625e-05,
"loss": 0.1044,
"step": 104
},
{
"epoch": 0.81,
"grad_norm": 0.22496414184570312,
"learning_rate": 3.949756603001029e-05,
"loss": 0.0969,
"step": 105
},
{
"epoch": 0.81,
"grad_norm": 0.2625040113925934,
"learning_rate": 3.94835734007371e-05,
"loss": 0.1013,
"step": 106
},
{
"epoch": 0.82,
"grad_norm": 0.2332611382007599,
"learning_rate": 3.946939114058234e-05,
"loss": 0.1021,
"step": 107
},
{
"epoch": 0.83,
"grad_norm": 0.23123830556869507,
"learning_rate": 3.945501938757996e-05,
"loss": 0.1036,
"step": 108
},
{
"epoch": 0.84,
"grad_norm": 0.23309484124183655,
"learning_rate": 3.944045828160822e-05,
"loss": 0.1053,
"step": 109
},
{
"epoch": 0.84,
"grad_norm": 0.26369422674179077,
"learning_rate": 3.942570796438831e-05,
"loss": 0.1081,
"step": 110
},
{
"epoch": 0.85,
"grad_norm": 0.2477720081806183,
"learning_rate": 3.9410768579482986e-05,
"loss": 0.1053,
"step": 111
},
{
"epoch": 0.86,
"grad_norm": 0.25222164392471313,
"learning_rate": 3.9395640272295196e-05,
"loss": 0.1017,
"step": 112
},
{
"epoch": 0.87,
"grad_norm": 0.23728714883327484,
"learning_rate": 3.938032319006662e-05,
"loss": 0.0996,
"step": 113
},
{
"epoch": 0.88,
"grad_norm": 0.21224068105220795,
"learning_rate": 3.9364817481876264e-05,
"loss": 0.099,
"step": 114
},
{
"epoch": 0.88,
"grad_norm": 0.2237294763326645,
"learning_rate": 3.934912329863902e-05,
"loss": 0.101,
"step": 115
},
{
"epoch": 0.89,
"grad_norm": 0.23548083007335663,
"learning_rate": 3.933324079310415e-05,
"loss": 0.1,
"step": 116
},
{
"epoch": 0.9,
"grad_norm": 0.2109685242176056,
"learning_rate": 3.931717011985387e-05,
"loss": 0.0978,
"step": 117
},
{
"epoch": 0.91,
"grad_norm": 0.21971383690834045,
"learning_rate": 3.9300911435301757e-05,
"loss": 0.0984,
"step": 118
},
{
"epoch": 0.91,
"grad_norm": 0.27964138984680176,
"learning_rate": 3.9284464897691326e-05,
"loss": 0.0981,
"step": 119
},
{
"epoch": 0.92,
"grad_norm": 0.23407186567783356,
"learning_rate": 3.9267830667094403e-05,
"loss": 0.0953,
"step": 120
},
{
"epoch": 0.93,
"grad_norm": 0.21624523401260376,
"learning_rate": 3.925100890540962e-05,
"loss": 0.1015,
"step": 121
},
{
"epoch": 0.94,
"grad_norm": 0.22755053639411926,
"learning_rate": 3.9233999776360816e-05,
"loss": 0.0972,
"step": 122
},
{
"epoch": 0.94,
"grad_norm": 0.21811529994010925,
"learning_rate": 3.921680344549546e-05,
"loss": 0.101,
"step": 123
},
{
"epoch": 0.95,
"grad_norm": 0.209341898560524,
"learning_rate": 3.919942008018302e-05,
"loss": 0.0997,
"step": 124
},
{
"epoch": 0.96,
"grad_norm": 0.20779836177825928,
"learning_rate": 3.9181849849613354e-05,
"loss": 0.0897,
"step": 125
},
{
"epoch": 0.97,
"grad_norm": 0.25512444972991943,
"learning_rate": 3.916409292479503e-05,
"loss": 0.0973,
"step": 126
},
{
"epoch": 0.98,
"grad_norm": 0.23092181980609894,
"learning_rate": 3.9146149478553715e-05,
"loss": 0.0985,
"step": 127
},
{
"epoch": 0.98,
"grad_norm": 0.21427614986896515,
"learning_rate": 3.912801968553045e-05,
"loss": 0.0981,
"step": 128
},
{
"epoch": 0.99,
"grad_norm": 0.22677119076251984,
"learning_rate": 3.910970372217996e-05,
"loss": 0.0972,
"step": 129
},
{
"epoch": 1.0,
"grad_norm": 0.25555670261383057,
"learning_rate": 3.909120176676894e-05,
"loss": 0.0946,
"step": 130
},
{
"epoch": 1.01,
"grad_norm": 0.18584538996219635,
"learning_rate": 3.9072513999374314e-05,
"loss": 0.0676,
"step": 131
},
{
"epoch": 1.01,
"grad_norm": 0.21039831638336182,
"learning_rate": 3.905364060188151e-05,
"loss": 0.064,
"step": 132
},
{
"epoch": 1.02,
"grad_norm": 0.16928508877754211,
"learning_rate": 3.903458175798265e-05,
"loss": 0.0587,
"step": 133
},
{
"epoch": 1.03,
"grad_norm": 0.16559255123138428,
"learning_rate": 3.901533765317478e-05,
"loss": 0.0598,
"step": 134
},
{
"epoch": 1.04,
"grad_norm": 0.20140372216701508,
"learning_rate": 3.899590847475806e-05,
"loss": 0.0577,
"step": 135
},
{
"epoch": 1.04,
"grad_norm": 0.18609680235385895,
"learning_rate": 3.897629441183397e-05,
"loss": 0.058,
"step": 136
},
{
"epoch": 1.05,
"grad_norm": 0.1916918158531189,
"learning_rate": 3.895649565530342e-05,
"loss": 0.0596,
"step": 137
},
{
"epoch": 1.06,
"grad_norm": 0.1645829826593399,
"learning_rate": 3.893651239786491e-05,
"loss": 0.0548,
"step": 138
},
{
"epoch": 1.07,
"grad_norm": 0.19628936052322388,
"learning_rate": 3.8916344834012695e-05,
"loss": 0.0553,
"step": 139
},
{
"epoch": 1.07,
"grad_norm": 0.20016969740390778,
"learning_rate": 3.8895993160034836e-05,
"loss": 0.0565,
"step": 140
},
{
"epoch": 1.08,
"grad_norm": 0.1823544204235077,
"learning_rate": 3.887545757401131e-05,
"loss": 0.0547,
"step": 141
},
{
"epoch": 1.09,
"grad_norm": 0.1844291090965271,
"learning_rate": 3.885473827581209e-05,
"loss": 0.0553,
"step": 142
},
{
"epoch": 1.1,
"grad_norm": 0.20922432839870453,
"learning_rate": 3.883383546709521e-05,
"loss": 0.0585,
"step": 143
},
{
"epoch": 1.11,
"grad_norm": 0.20714524388313293,
"learning_rate": 3.881274935130474e-05,
"loss": 0.0568,
"step": 144
},
{
"epoch": 1.11,
"grad_norm": 0.1846606284379959,
"learning_rate": 3.87914801336689e-05,
"loss": 0.06,
"step": 145
},
{
"epoch": 1.12,
"grad_norm": 0.16825266182422638,
"learning_rate": 3.8770028021197995e-05,
"loss": 0.0551,
"step": 146
},
{
"epoch": 1.13,
"grad_norm": 0.2070164680480957,
"learning_rate": 3.8748393222682425e-05,
"loss": 0.058,
"step": 147
},
{
"epoch": 1.14,
"grad_norm": 0.1790006011724472,
"learning_rate": 3.8726575948690624e-05,
"loss": 0.0581,
"step": 148
},
{
"epoch": 1.14,
"grad_norm": 0.1863609105348587,
"learning_rate": 3.870457641156708e-05,
"loss": 0.0597,
"step": 149
},
{
"epoch": 1.15,
"grad_norm": 0.1952793151140213,
"learning_rate": 3.868239482543016e-05,
"loss": 0.0539,
"step": 150
},
{
"epoch": 1.16,
"grad_norm": 0.1799939125776291,
"learning_rate": 3.8660031406170164e-05,
"loss": 0.0551,
"step": 151
},
{
"epoch": 1.17,
"grad_norm": 0.1753685474395752,
"learning_rate": 3.863748637144709e-05,
"loss": 0.0544,
"step": 152
},
{
"epoch": 1.17,
"grad_norm": 0.1807538866996765,
"learning_rate": 3.861475994068859e-05,
"loss": 0.0572,
"step": 153
},
{
"epoch": 1.18,
"grad_norm": 0.19420287013053894,
"learning_rate": 3.859185233508783e-05,
"loss": 0.0542,
"step": 154
},
{
"epoch": 1.19,
"grad_norm": 0.16617827117443085,
"learning_rate": 3.856876377760129e-05,
"loss": 0.053,
"step": 155
},
{
"epoch": 1.2,
"grad_norm": 0.20015448331832886,
"learning_rate": 3.854549449294669e-05,
"loss": 0.0563,
"step": 156
},
{
"epoch": 1.21,
"grad_norm": 0.17849506437778473,
"learning_rate": 3.852204470760069e-05,
"loss": 0.0569,
"step": 157
},
{
"epoch": 1.21,
"grad_norm": 0.18749572336673737,
"learning_rate": 3.8498414649796754e-05,
"loss": 0.0546,
"step": 158
},
{
"epoch": 1.22,
"grad_norm": 0.17191903293132782,
"learning_rate": 3.8474604549522914e-05,
"loss": 0.0542,
"step": 159
},
{
"epoch": 1.23,
"grad_norm": 0.19160284101963043,
"learning_rate": 3.845061463851954e-05,
"loss": 0.0557,
"step": 160
},
{
"epoch": 1.24,
"grad_norm": 0.1951061636209488,
"learning_rate": 3.842644515027704e-05,
"loss": 0.0576,
"step": 161
},
{
"epoch": 1.24,
"grad_norm": 0.16996340453624725,
"learning_rate": 3.840209632003368e-05,
"loss": 0.0545,
"step": 162
},
{
"epoch": 1.25,
"grad_norm": 0.15638357400894165,
"learning_rate": 3.8377568384773176e-05,
"loss": 0.0572,
"step": 163
},
{
"epoch": 1.26,
"grad_norm": 0.17367182672023773,
"learning_rate": 3.835286158322248e-05,
"loss": 0.0569,
"step": 164
},
{
"epoch": 1.27,
"grad_norm": 0.17908230423927307,
"learning_rate": 3.8327976155849425e-05,
"loss": 0.0561,
"step": 165
},
{
"epoch": 1.27,
"grad_norm": 0.17450128495693207,
"learning_rate": 3.830291234486037e-05,
"loss": 0.0561,
"step": 166
},
{
"epoch": 1.28,
"grad_norm": 0.17404121160507202,
"learning_rate": 3.827767039419788e-05,
"loss": 0.054,
"step": 167
},
{
"epoch": 1.29,
"grad_norm": 0.2045254111289978,
"learning_rate": 3.8252250549538296e-05,
"loss": 0.056,
"step": 168
},
{
"epoch": 1.3,
"grad_norm": 0.16833621263504028,
"learning_rate": 3.82266530582894e-05,
"loss": 0.0518,
"step": 169
},
{
"epoch": 1.31,
"grad_norm": 0.17673517763614655,
"learning_rate": 3.820087816958798e-05,
"loss": 0.0579,
"step": 170
},
{
"epoch": 1.31,
"grad_norm": 0.17191103100776672,
"learning_rate": 3.8174926134297396e-05,
"loss": 0.0509,
"step": 171
},
{
"epoch": 1.32,
"grad_norm": 0.21835172176361084,
"learning_rate": 3.814879720500517e-05,
"loss": 0.0565,
"step": 172
},
{
"epoch": 1.33,
"grad_norm": 0.18357177078723907,
"learning_rate": 3.81224916360205e-05,
"loss": 0.0559,
"step": 173
},
{
"epoch": 1.34,
"grad_norm": 0.1742638200521469,
"learning_rate": 3.8096009683371776e-05,
"loss": 0.054,
"step": 174
},
{
"epoch": 1.34,
"grad_norm": 0.1941666156053543,
"learning_rate": 3.806935160480414e-05,
"loss": 0.0521,
"step": 175
},
{
"epoch": 1.35,
"grad_norm": 0.18287934362888336,
"learning_rate": 3.8042517659776914e-05,
"loss": 0.055,
"step": 176
},
{
"epoch": 1.36,
"grad_norm": 0.16756708920001984,
"learning_rate": 3.801550810946111e-05,
"loss": 0.0552,
"step": 177
},
{
"epoch": 1.37,
"grad_norm": 0.1827043890953064,
"learning_rate": 3.79883232167369e-05,
"loss": 0.0557,
"step": 178
},
{
"epoch": 1.37,
"grad_norm": 0.1814430207014084,
"learning_rate": 3.7960963246191004e-05,
"loss": 0.0573,
"step": 179
},
{
"epoch": 1.38,
"grad_norm": 0.16552063822746277,
"learning_rate": 3.793342846411419e-05,
"loss": 0.0559,
"step": 180
},
{
"epoch": 1.39,
"grad_norm": 0.1673465520143509,
"learning_rate": 3.79057191384986e-05,
"loss": 0.0558,
"step": 181
},
{
"epoch": 1.4,
"grad_norm": 0.17430761456489563,
"learning_rate": 3.787783553903523e-05,
"loss": 0.0538,
"step": 182
},
{
"epoch": 1.4,
"grad_norm": 0.17373594641685486,
"learning_rate": 3.784977793711122e-05,
"loss": 0.0527,
"step": 183
},
{
"epoch": 1.41,
"grad_norm": 0.17581802606582642,
"learning_rate": 3.7821546605807264e-05,
"loss": 0.0536,
"step": 184
},
{
"epoch": 1.42,
"grad_norm": 0.17668214440345764,
"learning_rate": 3.7793141819894955e-05,
"loss": 0.0555,
"step": 185
},
{
"epoch": 1.43,
"grad_norm": 0.16610494256019592,
"learning_rate": 3.776456385583406e-05,
"loss": 0.0541,
"step": 186
},
{
"epoch": 1.44,
"grad_norm": 0.17192181944847107,
"learning_rate": 3.773581299176991e-05,
"loss": 0.0562,
"step": 187
},
{
"epoch": 1.44,
"grad_norm": 0.27887117862701416,
"learning_rate": 3.7706889507530615e-05,
"loss": 0.0551,
"step": 188
},
{
"epoch": 1.45,
"grad_norm": 0.17390647530555725,
"learning_rate": 3.767779368462438e-05,
"loss": 0.0573,
"step": 189
},
{
"epoch": 1.46,
"grad_norm": 0.16879643499851227,
"learning_rate": 3.764852580623676e-05,
"loss": 0.0557,
"step": 190
},
{
"epoch": 1.47,
"grad_norm": 0.1760348677635193,
"learning_rate": 3.761908615722791e-05,
"loss": 0.0534,
"step": 191
},
{
"epoch": 1.47,
"grad_norm": 0.17846859991550446,
"learning_rate": 3.758947502412978e-05,
"loss": 0.0556,
"step": 192
},
{
"epoch": 1.48,
"grad_norm": 0.17464794218540192,
"learning_rate": 3.755969269514339e-05,
"loss": 0.0552,
"step": 193
},
{
"epoch": 1.49,
"grad_norm": 0.1696479320526123,
"learning_rate": 3.752973946013594e-05,
"loss": 0.054,
"step": 194
},
{
"epoch": 1.5,
"grad_norm": 0.16251127421855927,
"learning_rate": 3.749961561063806e-05,
"loss": 0.0542,
"step": 195
},
{
"epoch": 1.5,
"grad_norm": 0.17379574477672577,
"learning_rate": 3.7469321439840955e-05,
"loss": 0.0551,
"step": 196
},
{
"epoch": 1.51,
"grad_norm": 0.1821696162223816,
"learning_rate": 3.743885724259353e-05,
"loss": 0.0568,
"step": 197
},
{
"epoch": 1.52,
"grad_norm": 0.20694509148597717,
"learning_rate": 3.7408223315399516e-05,
"loss": 0.0548,
"step": 198
},
{
"epoch": 1.53,
"grad_norm": 0.1554410755634308,
"learning_rate": 3.7377419956414646e-05,
"loss": 0.0526,
"step": 199
},
{
"epoch": 1.54,
"grad_norm": 0.16154886782169342,
"learning_rate": 3.734644746544368e-05,
"loss": 0.0529,
"step": 200
},
{
"epoch": 1.54,
"grad_norm": 0.1738775670528412,
"learning_rate": 3.7315306143937496e-05,
"loss": 0.0558,
"step": 201
},
{
"epoch": 1.55,
"grad_norm": 0.17919109761714935,
"learning_rate": 3.728399629499023e-05,
"loss": 0.0543,
"step": 202
},
{
"epoch": 1.56,
"grad_norm": 0.17479391396045685,
"learning_rate": 3.7252518223336214e-05,
"loss": 0.0568,
"step": 203
},
{
"epoch": 1.57,
"grad_norm": 0.16831263899803162,
"learning_rate": 3.722087223534711e-05,
"loss": 0.0553,
"step": 204
},
{
"epoch": 1.57,
"grad_norm": 0.15777549147605896,
"learning_rate": 3.7189058639028873e-05,
"loss": 0.0535,
"step": 205
},
{
"epoch": 1.58,
"grad_norm": 0.19188426434993744,
"learning_rate": 3.7157077744018766e-05,
"loss": 0.0539,
"step": 206
},
{
"epoch": 1.59,
"grad_norm": 0.160503089427948,
"learning_rate": 3.712492986158234e-05,
"loss": 0.0548,
"step": 207
},
{
"epoch": 1.6,
"grad_norm": 0.18297429382801056,
"learning_rate": 3.7092615304610414e-05,
"loss": 0.0533,
"step": 208
},
{
"epoch": 1.6,
"grad_norm": 0.16605788469314575,
"learning_rate": 3.706013438761604e-05,
"loss": 0.051,
"step": 209
},
{
"epoch": 1.61,
"grad_norm": 0.18309757113456726,
"learning_rate": 3.70274874267314e-05,
"loss": 0.0536,
"step": 210
},
{
"epoch": 1.62,
"grad_norm": 0.17336522042751312,
"learning_rate": 3.6994674739704795e-05,
"loss": 0.0578,
"step": 211
},
{
"epoch": 1.63,
"grad_norm": 0.16661109030246735,
"learning_rate": 3.696169664589748e-05,
"loss": 0.0534,
"step": 212
},
{
"epoch": 1.64,
"grad_norm": 0.16480354964733124,
"learning_rate": 3.6928553466280615e-05,
"loss": 0.0526,
"step": 213
},
{
"epoch": 1.64,
"grad_norm": 0.1613331288099289,
"learning_rate": 3.6895245523432116e-05,
"loss": 0.0552,
"step": 214
},
{
"epoch": 1.65,
"grad_norm": 0.18282514810562134,
"learning_rate": 3.68617731415335e-05,
"loss": 0.0524,
"step": 215
},
{
"epoch": 1.66,
"grad_norm": 0.17357532680034637,
"learning_rate": 3.6828136646366755e-05,
"loss": 0.0539,
"step": 216
},
{
"epoch": 1.67,
"grad_norm": 0.18777236342430115,
"learning_rate": 3.679433636531117e-05,
"loss": 0.0541,
"step": 217
},
{
"epoch": 1.67,
"grad_norm": 0.15832771360874176,
"learning_rate": 3.676037262734012e-05,
"loss": 0.0537,
"step": 218
},
{
"epoch": 1.68,
"grad_norm": 0.17728379368782043,
"learning_rate": 3.6726245763017895e-05,
"loss": 0.0523,
"step": 219
},
{
"epoch": 1.69,
"grad_norm": 0.15883496403694153,
"learning_rate": 3.669195610449647e-05,
"loss": 0.0528,
"step": 220
},
{
"epoch": 1.7,
"grad_norm": 0.15950903296470642,
"learning_rate": 3.665750398551228e-05,
"loss": 0.0536,
"step": 221
},
{
"epoch": 1.7,
"grad_norm": 0.1801331788301468,
"learning_rate": 3.662288974138294e-05,
"loss": 0.0528,
"step": 222
},
{
"epoch": 1.71,
"grad_norm": 0.1617787629365921,
"learning_rate": 3.658811370900404e-05,
"loss": 0.0545,
"step": 223
},
{
"epoch": 1.72,
"grad_norm": 0.17085997760295868,
"learning_rate": 3.6553176226845807e-05,
"loss": 0.0538,
"step": 224
},
{
"epoch": 1.73,
"grad_norm": 0.17167842388153076,
"learning_rate": 3.6518077634949845e-05,
"loss": 0.0543,
"step": 225
},
{
"epoch": 1.74,
"grad_norm": 0.17927636206150055,
"learning_rate": 3.648281827492582e-05,
"loss": 0.0543,
"step": 226
},
{
"epoch": 1.74,
"grad_norm": 0.17141342163085938,
"learning_rate": 3.644739848994813e-05,
"loss": 0.0571,
"step": 227
},
{
"epoch": 1.75,
"grad_norm": 0.15626104176044464,
"learning_rate": 3.6411818624752554e-05,
"loss": 0.049,
"step": 228
},
{
"epoch": 1.76,
"grad_norm": 0.1727394014596939,
"learning_rate": 3.637607902563292e-05,
"loss": 0.0504,
"step": 229
},
{
"epoch": 1.77,
"grad_norm": 0.16465209424495697,
"learning_rate": 3.634018004043773e-05,
"loss": 0.051,
"step": 230
},
{
"epoch": 1.77,
"grad_norm": 0.1781434714794159,
"learning_rate": 3.630412201856676e-05,
"loss": 0.056,
"step": 231
},
{
"epoch": 1.78,
"grad_norm": 0.19913175702095032,
"learning_rate": 3.626790531096767e-05,
"loss": 0.0556,
"step": 232
},
{
"epoch": 1.79,
"grad_norm": 0.19266557693481445,
"learning_rate": 3.623153027013258e-05,
"loss": 0.0558,
"step": 233
},
{
"epoch": 1.8,
"grad_norm": 0.2734900116920471,
"learning_rate": 3.619499725009466e-05,
"loss": 0.0536,
"step": 234
},
{
"epoch": 1.8,
"grad_norm": 0.16416223347187042,
"learning_rate": 3.615830660642464e-05,
"loss": 0.0517,
"step": 235
},
{
"epoch": 1.81,
"grad_norm": 0.16683989763259888,
"learning_rate": 3.612145869622741e-05,
"loss": 0.0557,
"step": 236
},
{
"epoch": 1.82,
"grad_norm": 0.17539849877357483,
"learning_rate": 3.608445387813851e-05,
"loss": 0.0545,
"step": 237
},
{
"epoch": 1.83,
"grad_norm": 0.1626010239124298,
"learning_rate": 3.604729251232063e-05,
"loss": 0.0528,
"step": 238
},
{
"epoch": 1.83,
"grad_norm": 0.1908058077096939,
"learning_rate": 3.60099749604601e-05,
"loss": 0.0541,
"step": 239
},
{
"epoch": 1.84,
"grad_norm": 0.1676451712846756,
"learning_rate": 3.597250158576344e-05,
"loss": 0.0537,
"step": 240
},
{
"epoch": 1.85,
"grad_norm": 0.17698313295841217,
"learning_rate": 3.593487275295371e-05,
"loss": 0.052,
"step": 241
},
{
"epoch": 1.86,
"grad_norm": 0.16062568128108978,
"learning_rate": 3.589708882826707e-05,
"loss": 0.0519,
"step": 242
},
{
"epoch": 1.87,
"grad_norm": 0.17272892594337463,
"learning_rate": 3.585915017944914e-05,
"loss": 0.0558,
"step": 243
},
{
"epoch": 1.87,
"grad_norm": 0.187282532453537,
"learning_rate": 3.582105717575146e-05,
"loss": 0.054,
"step": 244
},
{
"epoch": 1.88,
"grad_norm": 0.15364833176136017,
"learning_rate": 3.578281018792788e-05,
"loss": 0.0528,
"step": 245
},
{
"epoch": 1.89,
"grad_norm": 0.16929613053798676,
"learning_rate": 3.574440958823095e-05,
"loss": 0.0546,
"step": 246
},
{
"epoch": 1.9,
"grad_norm": 0.16617649793624878,
"learning_rate": 3.5705855750408325e-05,
"loss": 0.0568,
"step": 247
},
{
"epoch": 1.9,
"grad_norm": 0.1865209937095642,
"learning_rate": 3.566714904969907e-05,
"loss": 0.054,
"step": 248
},
{
"epoch": 1.91,
"grad_norm": 0.15715570747852325,
"learning_rate": 3.562828986283009e-05,
"loss": 0.0498,
"step": 249
},
{
"epoch": 1.92,
"grad_norm": 0.17273031175136566,
"learning_rate": 3.558927856801237e-05,
"loss": 0.0501,
"step": 250
},
{
"epoch": 1.93,
"grad_norm": 0.16318467259407043,
"learning_rate": 3.555011554493736e-05,
"loss": 0.0516,
"step": 251
},
{
"epoch": 1.93,
"grad_norm": 0.17008540034294128,
"learning_rate": 3.551080117477326e-05,
"loss": 0.0526,
"step": 252
},
{
"epoch": 1.94,
"grad_norm": 0.16984976828098297,
"learning_rate": 3.5471335840161324e-05,
"loss": 0.0533,
"step": 253
},
{
"epoch": 1.95,
"grad_norm": 0.16334734857082367,
"learning_rate": 3.543171992521209e-05,
"loss": 0.0506,
"step": 254
},
{
"epoch": 1.96,
"grad_norm": 0.15950210392475128,
"learning_rate": 3.539195381550169e-05,
"loss": 0.0495,
"step": 255
},
{
"epoch": 1.97,
"grad_norm": 0.1574394404888153,
"learning_rate": 3.535203789806809e-05,
"loss": 0.0527,
"step": 256
},
{
"epoch": 1.97,
"grad_norm": 0.160975381731987,
"learning_rate": 3.53119725614073e-05,
"loss": 0.0522,
"step": 257
},
{
"epoch": 1.98,
"grad_norm": 0.1672753542661667,
"learning_rate": 3.527175819546961e-05,
"loss": 0.0523,
"step": 258
},
{
"epoch": 1.99,
"grad_norm": 0.16033169627189636,
"learning_rate": 3.5231395191655796e-05,
"loss": 0.0514,
"step": 259
},
{
"epoch": 2.0,
"grad_norm": 0.1621590554714203,
"learning_rate": 3.5190883942813294e-05,
"loss": 0.0529,
"step": 260
},
{
"epoch": 2.0,
"grad_norm": 0.1404048502445221,
"learning_rate": 3.5150224843232405e-05,
"loss": 0.0402,
"step": 261
},
{
"epoch": 2.01,
"grad_norm": 0.11581116914749146,
"learning_rate": 3.5109418288642424e-05,
"loss": 0.0289,
"step": 262
},
{
"epoch": 2.02,
"grad_norm": 0.12721800804138184,
"learning_rate": 3.506846467620781e-05,
"loss": 0.0297,
"step": 263
},
{
"epoch": 2.03,
"grad_norm": 0.12351004034280777,
"learning_rate": 3.5027364404524326e-05,
"loss": 0.0292,
"step": 264
},
{
"epoch": 2.03,
"grad_norm": 0.11635002493858337,
"learning_rate": 3.4986117873615125e-05,
"loss": 0.0284,
"step": 265
},
{
"epoch": 2.04,
"grad_norm": 0.12261362373828888,
"learning_rate": 3.494472548492691e-05,
"loss": 0.029,
"step": 266
},
{
"epoch": 2.05,
"grad_norm": 0.12506607174873352,
"learning_rate": 3.4903187641325974e-05,
"loss": 0.0279,
"step": 267
},
{
"epoch": 2.06,
"grad_norm": 0.13922998309135437,
"learning_rate": 3.486150474709432e-05,
"loss": 0.0283,
"step": 268
},
{
"epoch": 2.07,
"grad_norm": 0.1253671497106552,
"learning_rate": 3.481967720792569e-05,
"loss": 0.0267,
"step": 269
},
{
"epoch": 2.07,
"grad_norm": 0.1219751164317131,
"learning_rate": 3.477770543092164e-05,
"loss": 0.027,
"step": 270
},
{
"epoch": 2.08,
"grad_norm": 0.13030819594860077,
"learning_rate": 3.473558982458759e-05,
"loss": 0.0278,
"step": 271
},
{
"epoch": 2.09,
"grad_norm": 0.13874004781246185,
"learning_rate": 3.469333079882881e-05,
"loss": 0.028,
"step": 272
},
{
"epoch": 2.1,
"grad_norm": 0.13344524800777435,
"learning_rate": 3.4650928764946456e-05,
"loss": 0.0267,
"step": 273
},
{
"epoch": 2.1,
"grad_norm": 0.13453194499015808,
"learning_rate": 3.4608384135633565e-05,
"loss": 0.0283,
"step": 274
},
{
"epoch": 2.11,
"grad_norm": 0.13323816657066345,
"learning_rate": 3.456569732497104e-05,
"loss": 0.0272,
"step": 275
},
{
"epoch": 2.12,
"grad_norm": 0.11938144266605377,
"learning_rate": 3.452286874842361e-05,
"loss": 0.0259,
"step": 276
},
{
"epoch": 2.13,
"grad_norm": 0.1395045667886734,
"learning_rate": 3.447989882283578e-05,
"loss": 0.0278,
"step": 277
},
{
"epoch": 2.13,
"grad_norm": 0.13324624300003052,
"learning_rate": 3.4436787966427804e-05,
"loss": 0.0271,
"step": 278
},
{
"epoch": 2.14,
"grad_norm": 0.14945518970489502,
"learning_rate": 3.43935365987916e-05,
"loss": 0.0274,
"step": 279
},
{
"epoch": 2.15,
"grad_norm": 0.13168294727802277,
"learning_rate": 3.435014514088662e-05,
"loss": 0.0279,
"step": 280
},
{
"epoch": 2.16,
"grad_norm": 0.12740759551525116,
"learning_rate": 3.430661401503586e-05,
"loss": 0.029,
"step": 281
},
{
"epoch": 2.17,
"grad_norm": 0.14633053541183472,
"learning_rate": 3.426294364492163e-05,
"loss": 0.0285,
"step": 282
},
{
"epoch": 2.17,
"grad_norm": 0.12179867923259735,
"learning_rate": 3.421913445558151e-05,
"loss": 0.0262,
"step": 283
},
{
"epoch": 2.18,
"grad_norm": 0.13524915277957916,
"learning_rate": 3.417518687340417e-05,
"loss": 0.0326,
"step": 284
},
{
"epoch": 2.19,
"grad_norm": 0.13417084515094757,
"learning_rate": 3.413110132612527e-05,
"loss": 0.0276,
"step": 285
},
{
"epoch": 2.2,
"grad_norm": 0.12195637077093124,
"learning_rate": 3.408687824282324e-05,
"loss": 0.0271,
"step": 286
},
{
"epoch": 2.2,
"grad_norm": 0.13557040691375732,
"learning_rate": 3.404251805391513e-05,
"loss": 0.0287,
"step": 287
},
{
"epoch": 2.21,
"grad_norm": 0.128498375415802,
"learning_rate": 3.3998021191152425e-05,
"loss": 0.0284,
"step": 288
},
{
"epoch": 2.22,
"grad_norm": 0.1270822286605835,
"learning_rate": 3.395338808761684e-05,
"loss": 0.0268,
"step": 289
},
{
"epoch": 2.23,
"grad_norm": 0.12966714799404144,
"learning_rate": 3.3908619177716086e-05,
"loss": 0.0281,
"step": 290
},
{
"epoch": 2.23,
"grad_norm": 0.14126142859458923,
"learning_rate": 3.3863714897179694e-05,
"loss": 0.0285,
"step": 291
},
{
"epoch": 2.24,
"grad_norm": 0.14122140407562256,
"learning_rate": 3.381867568305469e-05,
"loss": 0.0276,
"step": 292
},
{
"epoch": 2.25,
"grad_norm": 0.134867325425148,
"learning_rate": 3.377350197370143e-05,
"loss": 0.0296,
"step": 293
},
{
"epoch": 2.26,
"grad_norm": 0.12314282357692719,
"learning_rate": 3.372819420878928e-05,
"loss": 0.0285,
"step": 294
},
{
"epoch": 2.26,
"grad_norm": 0.13711945712566376,
"learning_rate": 3.368275282929234e-05,
"loss": 0.0295,
"step": 295
},
{
"epoch": 2.27,
"grad_norm": 0.13677950203418732,
"learning_rate": 3.363717827748517e-05,
"loss": 0.0289,
"step": 296
},
{
"epoch": 2.28,
"grad_norm": 0.14102821052074432,
"learning_rate": 3.359147099693849e-05,
"loss": 0.0296,
"step": 297
},
{
"epoch": 2.29,
"grad_norm": 0.12496308982372284,
"learning_rate": 3.354563143251483e-05,
"loss": 0.0293,
"step": 298
},
{
"epoch": 2.3,
"grad_norm": 0.12004269659519196,
"learning_rate": 3.349966003036421e-05,
"loss": 0.0295,
"step": 299
},
{
"epoch": 2.3,
"grad_norm": 0.12187572568655014,
"learning_rate": 3.3453557237919856e-05,
"loss": 0.0282,
"step": 300
},
{
"epoch": 2.31,
"grad_norm": 0.12873053550720215,
"learning_rate": 3.3407323503893745e-05,
"loss": 0.0292,
"step": 301
},
{
"epoch": 2.32,
"grad_norm": 0.12278107553720474,
"learning_rate": 3.336095927827233e-05,
"loss": 0.0283,
"step": 302
},
{
"epoch": 2.33,
"grad_norm": 0.1376417875289917,
"learning_rate": 3.331446501231208e-05,
"loss": 0.0297,
"step": 303
},
{
"epoch": 2.33,
"grad_norm": 0.1267479807138443,
"learning_rate": 3.326784115853518e-05,
"loss": 0.0285,
"step": 304
},
{
"epoch": 2.34,
"grad_norm": 0.11991389095783234,
"learning_rate": 3.322108817072503e-05,
"loss": 0.0295,
"step": 305
},
{
"epoch": 2.35,
"grad_norm": 0.13724927604198456,
"learning_rate": 3.317420650392187e-05,
"loss": 0.0299,
"step": 306
},
{
"epoch": 2.36,
"grad_norm": 0.12436146289110184,
"learning_rate": 3.3127196614418414e-05,
"loss": 0.0293,
"step": 307
},
{
"epoch": 2.36,
"grad_norm": 0.12342892587184906,
"learning_rate": 3.308005895975528e-05,
"loss": 0.0297,
"step": 308
},
{
"epoch": 2.37,
"grad_norm": 0.1250998079776764,
"learning_rate": 3.303279399871665e-05,
"loss": 0.0289,
"step": 309
},
{
"epoch": 2.38,
"grad_norm": 0.12870967388153076,
"learning_rate": 3.298540219132573e-05,
"loss": 0.0292,
"step": 310
},
{
"epoch": 2.39,
"grad_norm": 0.11160054057836533,
"learning_rate": 3.293788399884034e-05,
"loss": 0.0296,
"step": 311
},
{
"epoch": 2.4,
"grad_norm": 0.11962537467479706,
"learning_rate": 3.289023988374837e-05,
"loss": 0.0285,
"step": 312
},
{
"epoch": 2.4,
"grad_norm": 0.11754218488931656,
"learning_rate": 3.284247030976328e-05,
"loss": 0.0291,
"step": 313
},
{
"epoch": 2.41,
"grad_norm": 0.11548895388841629,
"learning_rate": 3.279457574181963e-05,
"loss": 0.0279,
"step": 314
},
{
"epoch": 2.42,
"grad_norm": 0.14263701438903809,
"learning_rate": 3.274655664606851e-05,
"loss": 0.0301,
"step": 315
},
{
"epoch": 2.43,
"grad_norm": 0.11611343175172806,
"learning_rate": 3.2698413489873035e-05,
"loss": 0.0284,
"step": 316
},
{
"epoch": 2.43,
"grad_norm": 0.1324465423822403,
"learning_rate": 3.265014674180378e-05,
"loss": 0.0313,
"step": 317
},
{
"epoch": 2.44,
"grad_norm": 0.12152358889579773,
"learning_rate": 3.260175687163423e-05,
"loss": 0.0323,
"step": 318
},
{
"epoch": 2.45,
"grad_norm": 0.1221373900771141,
"learning_rate": 3.255324435033617e-05,
"loss": 0.0299,
"step": 319
},
{
"epoch": 2.46,
"grad_norm": 0.11649201065301895,
"learning_rate": 3.250460965007515e-05,
"loss": 0.0291,
"step": 320
},
{
"epoch": 2.46,
"grad_norm": 0.13117162883281708,
"learning_rate": 3.2455853244205894e-05,
"loss": 0.0316,
"step": 321
},
{
"epoch": 2.47,
"grad_norm": 0.12085507810115814,
"learning_rate": 3.240697560726764e-05,
"loss": 0.0302,
"step": 322
},
{
"epoch": 2.48,
"grad_norm": 0.11401225626468658,
"learning_rate": 3.235797721497956e-05,
"loss": 0.0275,
"step": 323
},
{
"epoch": 2.49,
"grad_norm": 0.12400329858064651,
"learning_rate": 3.23088585442361e-05,
"loss": 0.0318,
"step": 324
},
{
"epoch": 2.5,
"grad_norm": 0.11465146392583847,
"learning_rate": 3.22596200731024e-05,
"loss": 0.0284,
"step": 325
},
{
"epoch": 2.5,
"grad_norm": 0.13097919523715973,
"learning_rate": 3.221026228080958e-05,
"loss": 0.0299,
"step": 326
},
{
"epoch": 2.51,
"grad_norm": 0.119114451110363,
"learning_rate": 3.216078564775008e-05,
"loss": 0.0306,
"step": 327
},
{
"epoch": 2.52,
"grad_norm": 0.11645802855491638,
"learning_rate": 3.211119065547302e-05,
"loss": 0.0289,
"step": 328
},
{
"epoch": 2.53,
"grad_norm": 0.12686465680599213,
"learning_rate": 3.206147778667951e-05,
"loss": 0.029,
"step": 329
},
{
"epoch": 2.53,
"grad_norm": 0.1158125177025795,
"learning_rate": 3.201164752521788e-05,
"loss": 0.0296,
"step": 330
},
{
"epoch": 2.54,
"grad_norm": 0.11673195660114288,
"learning_rate": 3.1961700356079094e-05,
"loss": 0.0315,
"step": 331
},
{
"epoch": 2.55,
"grad_norm": 0.13741804659366608,
"learning_rate": 3.191163676539191e-05,
"loss": 0.0309,
"step": 332
},
{
"epoch": 2.56,
"grad_norm": 0.1236562728881836,
"learning_rate": 3.186145724041822e-05,
"loss": 0.0295,
"step": 333
},
{
"epoch": 2.56,
"grad_norm": 0.12199486047029495,
"learning_rate": 3.181116226954831e-05,
"loss": 0.0311,
"step": 334
},
{
"epoch": 2.57,
"grad_norm": 0.1242741197347641,
"learning_rate": 3.176075234229603e-05,
"loss": 0.0309,
"step": 335
},
{
"epoch": 2.58,
"grad_norm": 0.11496566236019135,
"learning_rate": 3.171022794929414e-05,
"loss": 0.0304,
"step": 336
},
{
"epoch": 2.59,
"grad_norm": 0.11559294909238815,
"learning_rate": 3.1659589582289446e-05,
"loss": 0.029,
"step": 337
},
{
"epoch": 2.6,
"grad_norm": 0.12180111557245255,
"learning_rate": 3.160883773413806e-05,
"loss": 0.0305,
"step": 338
},
{
"epoch": 2.6,
"grad_norm": 0.11350613832473755,
"learning_rate": 3.155797289880059e-05,
"loss": 0.0296,
"step": 339
},
{
"epoch": 2.61,
"grad_norm": 0.12661539018154144,
"learning_rate": 3.1506995571337324e-05,
"loss": 0.0302,
"step": 340
},
{
"epoch": 2.62,
"grad_norm": 0.11439922451972961,
"learning_rate": 3.145590624790344e-05,
"loss": 0.0312,
"step": 341
},
{
"epoch": 2.63,
"grad_norm": 0.13127104938030243,
"learning_rate": 3.140470542574411e-05,
"loss": 0.0294,
"step": 342
},
{
"epoch": 2.63,
"grad_norm": 0.1237630695104599,
"learning_rate": 3.135339360318977e-05,
"loss": 0.0297,
"step": 343
},
{
"epoch": 2.64,
"grad_norm": 0.12231829017400742,
"learning_rate": 3.130197127965117e-05,
"loss": 0.0296,
"step": 344
},
{
"epoch": 2.65,
"grad_norm": 0.15128518640995026,
"learning_rate": 3.125043895561456e-05,
"loss": 0.0295,
"step": 345
},
{
"epoch": 2.66,
"grad_norm": 0.13165183365345,
"learning_rate": 3.1198797132636796e-05,
"loss": 0.0297,
"step": 346
},
{
"epoch": 2.66,
"grad_norm": 0.10592794418334961,
"learning_rate": 3.114704631334049e-05,
"loss": 0.0295,
"step": 347
},
{
"epoch": 2.67,
"grad_norm": 0.12173110991716385,
"learning_rate": 3.1095187001409095e-05,
"loss": 0.0304,
"step": 348
},
{
"epoch": 2.68,
"grad_norm": 0.11305344849824905,
"learning_rate": 3.104321970158201e-05,
"loss": 0.0303,
"step": 349
},
{
"epoch": 2.69,
"grad_norm": 0.11418797075748444,
"learning_rate": 3.099114491964964e-05,
"loss": 0.0299,
"step": 350
},
{
"epoch": 2.69,
"grad_norm": 0.11991377919912338,
"learning_rate": 3.093896316244855e-05,
"loss": 0.0309,
"step": 351
},
{
"epoch": 2.7,
"grad_norm": 0.11982237547636032,
"learning_rate": 3.088667493785641e-05,
"loss": 0.0297,
"step": 352
},
{
"epoch": 2.71,
"grad_norm": 0.1160060241818428,
"learning_rate": 3.083428075478719e-05,
"loss": 0.0296,
"step": 353
},
{
"epoch": 2.72,
"grad_norm": 0.12493564933538437,
"learning_rate": 3.078178112318611e-05,
"loss": 0.0329,
"step": 354
},
{
"epoch": 2.73,
"grad_norm": 0.12926845252513885,
"learning_rate": 3.0729176554024686e-05,
"loss": 0.0292,
"step": 355
},
{
"epoch": 2.73,
"grad_norm": 0.11742337048053741,
"learning_rate": 3.067646755929582e-05,
"loss": 0.0316,
"step": 356
},
{
"epoch": 2.74,
"grad_norm": 0.11704780161380768,
"learning_rate": 3.0623654652008724e-05,
"loss": 0.0293,
"step": 357
},
{
"epoch": 2.75,
"grad_norm": 0.11220533400774002,
"learning_rate": 3.057073834618403e-05,
"loss": 0.0309,
"step": 358
},
{
"epoch": 2.76,
"grad_norm": 0.12217779457569122,
"learning_rate": 3.051771915684868e-05,
"loss": 0.03,
"step": 359
},
{
"epoch": 2.76,
"grad_norm": 0.1118379682302475,
"learning_rate": 3.0464597600031013e-05,
"loss": 0.0294,
"step": 360
},
{
"epoch": 2.77,
"grad_norm": 0.12424726784229279,
"learning_rate": 3.041137419275566e-05,
"loss": 0.0307,
"step": 361
},
{
"epoch": 2.78,
"grad_norm": 0.12416715919971466,
"learning_rate": 3.035804945303857e-05,
"loss": 0.0298,
"step": 362
},
{
"epoch": 2.79,
"grad_norm": 0.12165214121341705,
"learning_rate": 3.030462389988194e-05,
"loss": 0.0305,
"step": 363
},
{
"epoch": 2.79,
"grad_norm": 0.1300431489944458,
"learning_rate": 3.025109805326917e-05,
"loss": 0.0307,
"step": 364
},
{
"epoch": 2.8,
"grad_norm": 0.1264806091785431,
"learning_rate": 3.01974724341598e-05,
"loss": 0.0283,
"step": 365
},
{
"epoch": 2.81,
"grad_norm": 0.11748187243938446,
"learning_rate": 3.0143747564484446e-05,
"loss": 0.029,
"step": 366
},
{
"epoch": 2.82,
"grad_norm": 0.11787963658571243,
"learning_rate": 3.0089923967139713e-05,
"loss": 0.0298,
"step": 367
},
{
"epoch": 2.83,
"grad_norm": 0.12340479344129562,
"learning_rate": 3.0036002165983106e-05,
"loss": 0.0306,
"step": 368
},
{
"epoch": 2.83,
"grad_norm": 0.1214551031589508,
"learning_rate": 2.9981982685827947e-05,
"loss": 0.0294,
"step": 369
},
{
"epoch": 2.84,
"grad_norm": 0.12566496431827545,
"learning_rate": 2.9927866052438243e-05,
"loss": 0.03,
"step": 370
},
{
"epoch": 2.85,
"grad_norm": 0.12087432295084,
"learning_rate": 2.9873652792523565e-05,
"loss": 0.0303,
"step": 371
},
{
"epoch": 2.86,
"grad_norm": 0.11541420966386795,
"learning_rate": 2.981934343373397e-05,
"loss": 0.03,
"step": 372
},
{
"epoch": 2.86,
"grad_norm": 0.12600594758987427,
"learning_rate": 2.97649385046548e-05,
"loss": 0.0324,
"step": 373
},
{
"epoch": 2.87,
"grad_norm": 0.11920228600502014,
"learning_rate": 2.9710438534801597e-05,
"loss": 0.0302,
"step": 374
},
{
"epoch": 2.88,
"grad_norm": 0.11343642324209213,
"learning_rate": 2.96558440546149e-05,
"loss": 0.0293,
"step": 375
},
{
"epoch": 2.89,
"grad_norm": 0.13571327924728394,
"learning_rate": 2.9601155595455116e-05,
"loss": 0.0306,
"step": 376
},
{
"epoch": 2.89,
"grad_norm": 0.10728243738412857,
"learning_rate": 2.9546373689597324e-05,
"loss": 0.0287,
"step": 377
},
{
"epoch": 2.9,
"grad_norm": 0.1138627678155899,
"learning_rate": 2.9491498870226117e-05,
"loss": 0.0299,
"step": 378
},
{
"epoch": 2.91,
"grad_norm": 0.11929310858249664,
"learning_rate": 2.9436531671430403e-05,
"loss": 0.0325,
"step": 379
},
{
"epoch": 2.92,
"grad_norm": 0.17112714052200317,
"learning_rate": 2.93814726281982e-05,
"loss": 0.0304,
"step": 380
},
{
"epoch": 2.93,
"grad_norm": 0.11112922430038452,
"learning_rate": 2.9326322276411446e-05,
"loss": 0.0293,
"step": 381
},
{
"epoch": 2.93,
"grad_norm": 0.10914931446313858,
"learning_rate": 2.927108115284075e-05,
"loss": 0.0297,
"step": 382
},
{
"epoch": 2.94,
"grad_norm": 0.14038969576358795,
"learning_rate": 2.921574979514021e-05,
"loss": 0.031,
"step": 383
},
{
"epoch": 2.95,
"grad_norm": 0.11356016993522644,
"learning_rate": 2.9160328741842158e-05,
"loss": 0.0306,
"step": 384
},
{
"epoch": 2.96,
"grad_norm": 0.11667381227016449,
"learning_rate": 2.9104818532351905e-05,
"loss": 0.0293,
"step": 385
},
{
"epoch": 2.96,
"grad_norm": 0.12412907183170319,
"learning_rate": 2.9049219706942532e-05,
"loss": 0.0307,
"step": 386
},
{
"epoch": 2.97,
"grad_norm": 0.12285000830888748,
"learning_rate": 2.8993532806749582e-05,
"loss": 0.0293,
"step": 387
},
{
"epoch": 2.98,
"grad_norm": 0.11109296977519989,
"learning_rate": 2.8937758373765832e-05,
"loss": 0.0307,
"step": 388
},
{
"epoch": 2.99,
"grad_norm": 0.11916177719831467,
"learning_rate": 2.8881896950836003e-05,
"loss": 0.0297,
"step": 389
},
{
"epoch": 2.99,
"grad_norm": 0.11571844667196274,
"learning_rate": 2.8825949081651464e-05,
"loss": 0.031,
"step": 390
},
{
"epoch": 3.0,
"grad_norm": 0.09632351249456406,
"learning_rate": 2.876991531074497e-05,
"loss": 0.025,
"step": 391
},
{
"epoch": 3.01,
"grad_norm": 0.08864370733499527,
"learning_rate": 2.871379618348534e-05,
"loss": 0.0194,
"step": 392
},
{
"epoch": 3.02,
"grad_norm": 0.08486244082450867,
"learning_rate": 2.8657592246072144e-05,
"loss": 0.0179,
"step": 393
},
{
"epoch": 3.02,
"grad_norm": 0.07573611289262772,
"learning_rate": 2.8601304045530414e-05,
"loss": 0.0171,
"step": 394
},
{
"epoch": 3.03,
"grad_norm": 0.09659112244844437,
"learning_rate": 2.8544932129705287e-05,
"loss": 0.0181,
"step": 395
},
{
"epoch": 3.04,
"grad_norm": 0.09413903206586838,
"learning_rate": 2.84884770472567e-05,
"loss": 0.0179,
"step": 396
},
{
"epoch": 3.05,
"grad_norm": 0.09005770087242126,
"learning_rate": 2.843193934765404e-05,
"loss": 0.018,
"step": 397
},
{
"epoch": 3.06,
"grad_norm": 0.09371140599250793,
"learning_rate": 2.837531958117078e-05,
"loss": 0.0179,
"step": 398
},
{
"epoch": 3.06,
"grad_norm": 0.09308820962905884,
"learning_rate": 2.831861829887917e-05,
"loss": 0.0162,
"step": 399
},
{
"epoch": 3.07,
"grad_norm": 0.09369007498025894,
"learning_rate": 2.82618360526448e-05,
"loss": 0.0167,
"step": 400
},
{
"epoch": 3.08,
"grad_norm": 0.08827667683362961,
"learning_rate": 2.8204973395121298e-05,
"loss": 0.0179,
"step": 401
},
{
"epoch": 3.09,
"grad_norm": 0.08312886953353882,
"learning_rate": 2.814803087974492e-05,
"loss": 0.0162,
"step": 402
},
{
"epoch": 3.09,
"grad_norm": 0.10320934653282166,
"learning_rate": 2.809100906072917e-05,
"loss": 0.0201,
"step": 403
},
{
"epoch": 3.1,
"grad_norm": 0.09035062044858932,
"learning_rate": 2.8033908493059394e-05,
"loss": 0.0182,
"step": 404
},
{
"epoch": 3.11,
"grad_norm": 0.09151903539896011,
"learning_rate": 2.7976729732487386e-05,
"loss": 0.0161,
"step": 405
},
{
"epoch": 3.12,
"grad_norm": 0.0930718183517456,
"learning_rate": 2.7919473335525995e-05,
"loss": 0.0173,
"step": 406
},
{
"epoch": 3.12,
"grad_norm": 0.10269258171319962,
"learning_rate": 2.7862139859443683e-05,
"loss": 0.0182,
"step": 407
},
{
"epoch": 3.13,
"grad_norm": 0.11026918143033981,
"learning_rate": 2.7804729862259114e-05,
"loss": 0.0173,
"step": 408
},
{
"epoch": 3.14,
"grad_norm": 0.09837930649518967,
"learning_rate": 2.7747243902735723e-05,
"loss": 0.0176,
"step": 409
},
{
"epoch": 3.15,
"grad_norm": 0.08965995907783508,
"learning_rate": 2.7689682540376267e-05,
"loss": 0.0174,
"step": 410
},
{
"epoch": 3.16,
"grad_norm": 0.09735357761383057,
"learning_rate": 2.7632046335417412e-05,
"loss": 0.0182,
"step": 411
},
{
"epoch": 3.16,
"grad_norm": 0.11039672046899796,
"learning_rate": 2.7574335848824226e-05,
"loss": 0.0177,
"step": 412
},
{
"epoch": 3.17,
"grad_norm": 0.10187050700187683,
"learning_rate": 2.7516551642284765e-05,
"loss": 0.0177,
"step": 413
},
{
"epoch": 3.18,
"grad_norm": 0.08932174742221832,
"learning_rate": 2.74586942782046e-05,
"loss": 0.0173,
"step": 414
},
{
"epoch": 3.19,
"grad_norm": 0.09614111483097076,
"learning_rate": 2.7400764319701312e-05,
"loss": 0.0177,
"step": 415
},
{
"epoch": 3.19,
"grad_norm": 0.10016363114118576,
"learning_rate": 2.7342762330599048e-05,
"loss": 0.0179,
"step": 416
},
{
"epoch": 3.2,
"grad_norm": 0.1256435066461563,
"learning_rate": 2.7284688875423025e-05,
"loss": 0.0187,
"step": 417
},
{
"epoch": 3.21,
"grad_norm": 0.09919959306716919,
"learning_rate": 2.7226544519394015e-05,
"loss": 0.0171,
"step": 418
},
{
"epoch": 3.22,
"grad_norm": 0.09221014380455017,
"learning_rate": 2.7168329828422856e-05,
"loss": 0.0185,
"step": 419
},
{
"epoch": 3.22,
"grad_norm": 0.09211818128824234,
"learning_rate": 2.7110045369104967e-05,
"loss": 0.0177,
"step": 420
},
{
"epoch": 3.23,
"grad_norm": 0.08906693011522293,
"learning_rate": 2.7051691708714784e-05,
"loss": 0.0175,
"step": 421
},
{
"epoch": 3.24,
"grad_norm": 0.09074138104915619,
"learning_rate": 2.6993269415200294e-05,
"loss": 0.0191,
"step": 422
},
{
"epoch": 3.25,
"grad_norm": 0.09929236024618149,
"learning_rate": 2.6934779057177466e-05,
"loss": 0.0178,
"step": 423
},
{
"epoch": 3.26,
"grad_norm": 0.09517091512680054,
"learning_rate": 2.6876221203924723e-05,
"loss": 0.0188,
"step": 424
},
{
"epoch": 3.26,
"grad_norm": 0.10076611489057541,
"learning_rate": 2.6817596425377437e-05,
"loss": 0.0178,
"step": 425
},
{
"epoch": 3.27,
"grad_norm": 0.09757237136363983,
"learning_rate": 2.6758905292122327e-05,
"loss": 0.019,
"step": 426
},
{
"epoch": 3.28,
"grad_norm": 0.11040135473012924,
"learning_rate": 2.6700148375391958e-05,
"loss": 0.0202,
"step": 427
},
{
"epoch": 3.29,
"grad_norm": 0.09049519151449203,
"learning_rate": 2.6641326247059124e-05,
"loss": 0.0189,
"step": 428
},
{
"epoch": 3.29,
"grad_norm": 0.1085498183965683,
"learning_rate": 2.658243947963134e-05,
"loss": 0.0194,
"step": 429
},
{
"epoch": 3.3,
"grad_norm": 0.08746053278446198,
"learning_rate": 2.6523488646245242e-05,
"loss": 0.0176,
"step": 430
},
{
"epoch": 3.31,
"grad_norm": 0.09314620494842529,
"learning_rate": 2.6464474320660998e-05,
"loss": 0.0182,
"step": 431
},
{
"epoch": 3.32,
"grad_norm": 0.12029460072517395,
"learning_rate": 2.6405397077256752e-05,
"loss": 0.0188,
"step": 432
},
{
"epoch": 3.32,
"grad_norm": 0.0930788666009903,
"learning_rate": 2.6346257491023015e-05,
"loss": 0.0192,
"step": 433
},
{
"epoch": 3.33,
"grad_norm": 0.09339287132024765,
"learning_rate": 2.6287056137557054e-05,
"loss": 0.0195,
"step": 434
},
{
"epoch": 3.34,
"grad_norm": 0.08442039042711258,
"learning_rate": 2.622779359305733e-05,
"loss": 0.0179,
"step": 435
},
{
"epoch": 3.35,
"grad_norm": 0.09663324803113937,
"learning_rate": 2.6168470434317858e-05,
"loss": 0.0196,
"step": 436
},
{
"epoch": 3.36,
"grad_norm": 0.09552859514951706,
"learning_rate": 2.6109087238722604e-05,
"loss": 0.0189,
"step": 437
},
{
"epoch": 3.36,
"grad_norm": 0.09758557379245758,
"learning_rate": 2.6049644584239867e-05,
"loss": 0.0192,
"step": 438
},
{
"epoch": 3.37,
"grad_norm": 0.09988658130168915,
"learning_rate": 2.599014304941665e-05,
"loss": 0.0198,
"step": 439
},
{
"epoch": 3.38,
"grad_norm": 0.08330779522657394,
"learning_rate": 2.593058321337302e-05,
"loss": 0.0188,
"step": 440
},
{
"epoch": 3.39,
"grad_norm": 0.09017366170883179,
"learning_rate": 2.5870965655796497e-05,
"loss": 0.0187,
"step": 441
},
{
"epoch": 3.39,
"grad_norm": 0.09778042137622833,
"learning_rate": 2.5811290956936388e-05,
"loss": 0.0193,
"step": 442
},
{
"epoch": 3.4,
"grad_norm": 0.10484529286623001,
"learning_rate": 2.5751559697598155e-05,
"loss": 0.0187,
"step": 443
},
{
"epoch": 3.41,
"grad_norm": 0.09345745295286179,
"learning_rate": 2.5691772459137747e-05,
"loss": 0.0194,
"step": 444
},
{
"epoch": 3.42,
"grad_norm": 0.09034112095832825,
"learning_rate": 2.5631929823455945e-05,
"loss": 0.0196,
"step": 445
},
{
"epoch": 3.42,
"grad_norm": 0.09179243445396423,
"learning_rate": 2.557203237299272e-05,
"loss": 0.0194,
"step": 446
},
{
"epoch": 3.43,
"grad_norm": 0.0940314382314682,
"learning_rate": 2.5512080690721543e-05,
"loss": 0.019,
"step": 447
},
{
"epoch": 3.44,
"grad_norm": 0.09003882110118866,
"learning_rate": 2.5452075360143703e-05,
"loss": 0.0194,
"step": 448
},
{
"epoch": 3.45,
"grad_norm": 0.10961990803480148,
"learning_rate": 2.5392016965282647e-05,
"loss": 0.0193,
"step": 449
},
{
"epoch": 3.45,
"grad_norm": 0.09360860288143158,
"learning_rate": 2.5331906090678303e-05,
"loss": 0.0205,
"step": 450
},
{
"epoch": 3.46,
"grad_norm": 0.0873427763581276,
"learning_rate": 2.5271743321381354e-05,
"loss": 0.019,
"step": 451
},
{
"epoch": 3.47,
"grad_norm": 0.08723457157611847,
"learning_rate": 2.5211529242947586e-05,
"loss": 0.0192,
"step": 452
},
{
"epoch": 3.48,
"grad_norm": 0.0958879366517067,
"learning_rate": 2.5151264441432158e-05,
"loss": 0.0192,
"step": 453
},
{
"epoch": 3.49,
"grad_norm": 0.11000870913267136,
"learning_rate": 2.5090949503383906e-05,
"loss": 0.0211,
"step": 454
},
{
"epoch": 3.49,
"grad_norm": 0.0878106877207756,
"learning_rate": 2.5030585015839658e-05,
"loss": 0.0192,
"step": 455
},
{
"epoch": 3.5,
"grad_norm": 0.09739448130130768,
"learning_rate": 2.4970171566318477e-05,
"loss": 0.0192,
"step": 456
},
{
"epoch": 3.51,
"grad_norm": 0.10094361007213593,
"learning_rate": 2.4909709742815986e-05,
"loss": 0.0207,
"step": 457
},
{
"epoch": 3.52,
"grad_norm": 0.09005150198936462,
"learning_rate": 2.4849200133798608e-05,
"loss": 0.0207,
"step": 458
},
{
"epoch": 3.52,
"grad_norm": 0.09568016976118088,
"learning_rate": 2.4788643328197863e-05,
"loss": 0.0195,
"step": 459
},
{
"epoch": 3.53,
"grad_norm": 0.08716892451047897,
"learning_rate": 2.472803991540463e-05,
"loss": 0.019,
"step": 460
},
{
"epoch": 3.54,
"grad_norm": 0.08845728635787964,
"learning_rate": 2.466739048526341e-05,
"loss": 0.019,
"step": 461
},
{
"epoch": 3.55,
"grad_norm": 0.08424665778875351,
"learning_rate": 2.4606695628066584e-05,
"loss": 0.0189,
"step": 462
},
{
"epoch": 3.55,
"grad_norm": 0.08149096369743347,
"learning_rate": 2.454595593454866e-05,
"loss": 0.0188,
"step": 463
},
{
"epoch": 3.56,
"grad_norm": 0.09580852091312408,
"learning_rate": 2.448517199588055e-05,
"loss": 0.0215,
"step": 464
},
{
"epoch": 3.57,
"grad_norm": 0.08978178352117538,
"learning_rate": 2.442434440366377e-05,
"loss": 0.0203,
"step": 465
},
{
"epoch": 3.58,
"grad_norm": 0.08793075382709503,
"learning_rate": 2.436347374992474e-05,
"loss": 0.0204,
"step": 466
},
{
"epoch": 3.59,
"grad_norm": 0.08886060863733292,
"learning_rate": 2.430256062710897e-05,
"loss": 0.0203,
"step": 467
},
{
"epoch": 3.59,
"grad_norm": 0.1001208946108818,
"learning_rate": 2.424160562807533e-05,
"loss": 0.0204,
"step": 468
},
{
"epoch": 3.6,
"grad_norm": 0.08480829745531082,
"learning_rate": 2.4180609346090247e-05,
"loss": 0.0194,
"step": 469
},
{
"epoch": 3.61,
"grad_norm": 0.09243600815534592,
"learning_rate": 2.4119572374821968e-05,
"loss": 0.0194,
"step": 470
},
{
"epoch": 3.62,
"grad_norm": 0.08637120574712753,
"learning_rate": 2.4058495308334758e-05,
"loss": 0.0201,
"step": 471
},
{
"epoch": 3.62,
"grad_norm": 0.09526148438453674,
"learning_rate": 2.3997378741083127e-05,
"loss": 0.0203,
"step": 472
},
{
"epoch": 3.63,
"grad_norm": 0.08574876189231873,
"learning_rate": 2.3936223267906017e-05,
"loss": 0.0197,
"step": 473
},
{
"epoch": 3.64,
"grad_norm": 0.08735029399394989,
"learning_rate": 2.387502948402106e-05,
"loss": 0.019,
"step": 474
},
{
"epoch": 3.65,
"grad_norm": 0.08287892490625381,
"learning_rate": 2.381379798501876e-05,
"loss": 0.02,
"step": 475
},
{
"epoch": 3.65,
"grad_norm": 0.08337916433811188,
"learning_rate": 2.375252936685669e-05,
"loss": 0.0199,
"step": 476
},
{
"epoch": 3.66,
"grad_norm": 0.09577334672212601,
"learning_rate": 2.3691224225853685e-05,
"loss": 0.0194,
"step": 477
},
{
"epoch": 3.67,
"grad_norm": 0.08564145117998123,
"learning_rate": 2.362988315868409e-05,
"loss": 0.0195,
"step": 478
},
{
"epoch": 3.68,
"grad_norm": 0.09480710327625275,
"learning_rate": 2.3568506762371875e-05,
"loss": 0.0209,
"step": 479
},
{
"epoch": 3.69,
"grad_norm": 0.09078356623649597,
"learning_rate": 2.350709563428488e-05,
"loss": 0.0203,
"step": 480
},
{
"epoch": 3.69,
"grad_norm": 0.08901910483837128,
"learning_rate": 2.344565037212899e-05,
"loss": 0.0199,
"step": 481
},
{
"epoch": 3.7,
"grad_norm": 0.08397212624549866,
"learning_rate": 2.33841715739423e-05,
"loss": 0.0194,
"step": 482
},
{
"epoch": 3.71,
"grad_norm": 0.08882483094930649,
"learning_rate": 2.3322659838089313e-05,
"loss": 0.0202,
"step": 483
},
{
"epoch": 3.72,
"grad_norm": 0.08408450335264206,
"learning_rate": 2.3261115763255108e-05,
"loss": 0.0202,
"step": 484
},
{
"epoch": 3.72,
"grad_norm": 0.09042160212993622,
"learning_rate": 2.3199539948439512e-05,
"loss": 0.0203,
"step": 485
},
{
"epoch": 3.73,
"grad_norm": 0.08511916548013687,
"learning_rate": 2.3137932992951298e-05,
"loss": 0.0188,
"step": 486
},
{
"epoch": 3.74,
"grad_norm": 0.09483899176120758,
"learning_rate": 2.3076295496402277e-05,
"loss": 0.0197,
"step": 487
},
{
"epoch": 3.75,
"grad_norm": 0.08442514389753342,
"learning_rate": 2.301462805870155e-05,
"loss": 0.0201,
"step": 488
},
{
"epoch": 3.75,
"grad_norm": 0.08056911826133728,
"learning_rate": 2.2952931280049628e-05,
"loss": 0.0192,
"step": 489
},
{
"epoch": 3.76,
"grad_norm": 0.08947514742612839,
"learning_rate": 2.289120576093258e-05,
"loss": 0.02,
"step": 490
},
{
"epoch": 3.77,
"grad_norm": 0.08950494229793549,
"learning_rate": 2.2829452102116206e-05,
"loss": 0.0194,
"step": 491
},
{
"epoch": 3.78,
"grad_norm": 0.08909059315919876,
"learning_rate": 2.2767670904640197e-05,
"loss": 0.0201,
"step": 492
},
{
"epoch": 3.79,
"grad_norm": 0.09698917716741562,
"learning_rate": 2.2705862769812252e-05,
"loss": 0.0201,
"step": 493
},
{
"epoch": 3.79,
"grad_norm": 0.09862850606441498,
"learning_rate": 2.2644028299202275e-05,
"loss": 0.0198,
"step": 494
},
{
"epoch": 3.8,
"grad_norm": 0.10406886786222458,
"learning_rate": 2.258216809463646e-05,
"loss": 0.0203,
"step": 495
},
{
"epoch": 3.81,
"grad_norm": 0.08993860334157944,
"learning_rate": 2.25202827581915e-05,
"loss": 0.0192,
"step": 496
},
{
"epoch": 3.82,
"grad_norm": 0.0922846645116806,
"learning_rate": 2.2458372892188675e-05,
"loss": 0.0202,
"step": 497
},
{
"epoch": 3.82,
"grad_norm": 0.08428039401769638,
"learning_rate": 2.2396439099188e-05,
"loss": 0.0203,
"step": 498
},
{
"epoch": 3.83,
"grad_norm": 0.08886083960533142,
"learning_rate": 2.2334481981982385e-05,
"loss": 0.0205,
"step": 499
},
{
"epoch": 3.84,
"grad_norm": 0.0839906558394432,
"learning_rate": 2.2272502143591742e-05,
"loss": 0.0195,
"step": 500
},
{
"epoch": 3.85,
"grad_norm": 0.08571367710828781,
"learning_rate": 2.221050018725713e-05,
"loss": 0.0198,
"step": 501
},
{
"epoch": 3.85,
"grad_norm": 0.08495412021875381,
"learning_rate": 2.2148476716434874e-05,
"loss": 0.0198,
"step": 502
},
{
"epoch": 3.86,
"grad_norm": 0.09099718183279037,
"learning_rate": 2.208643233479069e-05,
"loss": 0.0193,
"step": 503
},
{
"epoch": 3.87,
"grad_norm": 0.09915336966514587,
"learning_rate": 2.2024367646193838e-05,
"loss": 0.0197,
"step": 504
},
{
"epoch": 3.88,
"grad_norm": 0.09666319936513901,
"learning_rate": 2.19622832547112e-05,
"loss": 0.0208,
"step": 505
},
{
"epoch": 3.88,
"grad_norm": 0.09048108011484146,
"learning_rate": 2.1900179764601442e-05,
"loss": 0.0207,
"step": 506
},
{
"epoch": 3.89,
"grad_norm": 0.08322761207818985,
"learning_rate": 2.1838057780309088e-05,
"loss": 0.0195,
"step": 507
},
{
"epoch": 3.9,
"grad_norm": 0.08545359969139099,
"learning_rate": 2.1775917906458698e-05,
"loss": 0.0202,
"step": 508
},
{
"epoch": 3.91,
"grad_norm": 0.1045001745223999,
"learning_rate": 2.1713760747848922e-05,
"loss": 0.0208,
"step": 509
},
{
"epoch": 3.92,
"grad_norm": 0.09209699928760529,
"learning_rate": 2.165158690944665e-05,
"loss": 0.0207,
"step": 510
},
{
"epoch": 3.92,
"grad_norm": 0.08286304026842117,
"learning_rate": 2.1589396996381115e-05,
"loss": 0.0194,
"step": 511
},
{
"epoch": 3.93,
"grad_norm": 0.10599466413259506,
"learning_rate": 2.1527191613937993e-05,
"loss": 0.0202,
"step": 512
},
{
"epoch": 3.94,
"grad_norm": 0.09873639047145844,
"learning_rate": 2.1464971367553535e-05,
"loss": 0.0211,
"step": 513
},
{
"epoch": 3.95,
"grad_norm": 0.09936564415693283,
"learning_rate": 2.1402736862808654e-05,
"loss": 0.0207,
"step": 514
},
{
"epoch": 3.95,
"grad_norm": 0.09293445199728012,
"learning_rate": 2.1340488705423035e-05,
"loss": 0.0204,
"step": 515
},
{
"epoch": 3.96,
"grad_norm": 0.0853753536939621,
"learning_rate": 2.1278227501249253e-05,
"loss": 0.0202,
"step": 516
},
{
"epoch": 3.97,
"grad_norm": 0.08694113045930862,
"learning_rate": 2.1215953856266846e-05,
"loss": 0.0195,
"step": 517
},
{
"epoch": 3.98,
"grad_norm": 0.0869363471865654,
"learning_rate": 2.1153668376576452e-05,
"loss": 0.0204,
"step": 518
},
{
"epoch": 3.98,
"grad_norm": 0.09057950973510742,
"learning_rate": 2.1091371668393895e-05,
"loss": 0.0202,
"step": 519
},
{
"epoch": 3.99,
"grad_norm": 0.08620808273553848,
"learning_rate": 2.102906433804428e-05,
"loss": 0.0194,
"step": 520
},
{
"epoch": 4.0,
"grad_norm": 0.08564823865890503,
"learning_rate": 2.0966746991956093e-05,
"loss": 0.0197,
"step": 521
},
{
"epoch": 4.01,
"grad_norm": 0.06662225723266602,
"learning_rate": 2.0904420236655304e-05,
"loss": 0.0127,
"step": 522
},
{
"epoch": 4.02,
"grad_norm": 0.06877375394105911,
"learning_rate": 2.0842084678759465e-05,
"loss": 0.0124,
"step": 523
},
{
"epoch": 4.02,
"grad_norm": 0.06511253863573074,
"learning_rate": 2.0779740924971806e-05,
"loss": 0.0119,
"step": 524
},
{
"epoch": 4.03,
"grad_norm": 0.0667959600687027,
"learning_rate": 2.0717389582075308e-05,
"loss": 0.0127,
"step": 525
},
{
"epoch": 4.04,
"grad_norm": 0.06564877927303314,
"learning_rate": 2.065503125692683e-05,
"loss": 0.0124,
"step": 526
},
{
"epoch": 4.05,
"grad_norm": 0.06754275411367416,
"learning_rate": 2.0592666556451197e-05,
"loss": 0.0128,
"step": 527
},
{
"epoch": 4.05,
"grad_norm": 0.06806235015392303,
"learning_rate": 2.0530296087635263e-05,
"loss": 0.0119,
"step": 528
},
{
"epoch": 4.06,
"grad_norm": 0.0656081959605217,
"learning_rate": 2.0467920457522043e-05,
"loss": 0.0124,
"step": 529
},
{
"epoch": 4.07,
"grad_norm": 0.07134638726711273,
"learning_rate": 2.0405540273204777e-05,
"loss": 0.0119,
"step": 530
},
{
"epoch": 4.08,
"grad_norm": 0.06481177359819412,
"learning_rate": 2.0343156141821025e-05,
"loss": 0.0119,
"step": 531
},
{
"epoch": 4.08,
"grad_norm": 0.0694548636674881,
"learning_rate": 2.028076867054677e-05,
"loss": 0.0128,
"step": 532
},
{
"epoch": 4.09,
"grad_norm": 0.07217177748680115,
"learning_rate": 2.021837846659051e-05,
"loss": 0.0125,
"step": 533
},
{
"epoch": 4.1,
"grad_norm": 0.07400055229663849,
"learning_rate": 2.0155986137187326e-05,
"loss": 0.0121,
"step": 534
},
{
"epoch": 4.11,
"grad_norm": 0.07027433812618256,
"learning_rate": 2.0093592289592997e-05,
"loss": 0.0117,
"step": 535
},
{
"epoch": 4.12,
"grad_norm": 0.06789272278547287,
"learning_rate": 2.0031197531078064e-05,
"loss": 0.0122,
"step": 536
},
{
"epoch": 4.12,
"grad_norm": 0.071243517100811,
"learning_rate": 1.9968802468921942e-05,
"loss": 0.0123,
"step": 537
},
{
"epoch": 4.13,
"grad_norm": 0.07798586040735245,
"learning_rate": 1.990640771040701e-05,
"loss": 0.0128,
"step": 538
},
{
"epoch": 4.14,
"grad_norm": 0.07501710951328278,
"learning_rate": 1.9844013862812674e-05,
"loss": 0.0133,
"step": 539
},
{
"epoch": 4.15,
"grad_norm": 0.06893974542617798,
"learning_rate": 1.978162153340949e-05,
"loss": 0.0121,
"step": 540
},
{
"epoch": 4.15,
"grad_norm": 0.08002519607543945,
"learning_rate": 1.9719231329453232e-05,
"loss": 0.0129,
"step": 541
},
{
"epoch": 4.16,
"grad_norm": 0.07770746201276779,
"learning_rate": 1.9656843858178988e-05,
"loss": 0.0127,
"step": 542
},
{
"epoch": 4.17,
"grad_norm": 0.07314487546682358,
"learning_rate": 1.9594459726795233e-05,
"loss": 0.013,
"step": 543
},
{
"epoch": 4.18,
"grad_norm": 0.07891873270273209,
"learning_rate": 1.9532079542477963e-05,
"loss": 0.0131,
"step": 544
},
{
"epoch": 4.18,
"grad_norm": 0.07237127423286438,
"learning_rate": 1.946970391236474e-05,
"loss": 0.0121,
"step": 545
},
{
"epoch": 4.19,
"grad_norm": 0.06825929880142212,
"learning_rate": 1.9407333443548806e-05,
"loss": 0.0126,
"step": 546
},
{
"epoch": 4.2,
"grad_norm": 0.06899861991405487,
"learning_rate": 1.9344968743073173e-05,
"loss": 0.0127,
"step": 547
},
{
"epoch": 4.21,
"grad_norm": 0.07207786291837692,
"learning_rate": 1.92826104179247e-05,
"loss": 0.0125,
"step": 548
},
{
"epoch": 4.21,
"grad_norm": 0.06783226877450943,
"learning_rate": 1.9220259075028204e-05,
"loss": 0.0123,
"step": 549
},
{
"epoch": 4.22,
"grad_norm": 0.06856781244277954,
"learning_rate": 1.9157915321240534e-05,
"loss": 0.0126,
"step": 550
},
{
"epoch": 4.23,
"grad_norm": 0.07011223584413528,
"learning_rate": 1.9095579763344696e-05,
"loss": 0.0131,
"step": 551
},
{
"epoch": 4.24,
"grad_norm": 0.06701693683862686,
"learning_rate": 1.9033253008043914e-05,
"loss": 0.0121,
"step": 552
},
{
"epoch": 4.25,
"grad_norm": 0.07344800978899002,
"learning_rate": 1.8970935661955727e-05,
"loss": 0.0133,
"step": 553
},
{
"epoch": 4.25,
"grad_norm": 0.07029430568218231,
"learning_rate": 1.8908628331606108e-05,
"loss": 0.0131,
"step": 554
},
{
"epoch": 4.26,
"grad_norm": 0.06867130845785141,
"learning_rate": 1.8846331623423555e-05,
"loss": 0.013,
"step": 555
},
{
"epoch": 4.27,
"grad_norm": 0.0735846534371376,
"learning_rate": 1.878404614373316e-05,
"loss": 0.0122,
"step": 556
},
{
"epoch": 4.28,
"grad_norm": 0.07147074490785599,
"learning_rate": 1.872177249875075e-05,
"loss": 0.0133,
"step": 557
},
{
"epoch": 4.28,
"grad_norm": 0.06911338865756989,
"learning_rate": 1.8659511294576968e-05,
"loss": 0.0131,
"step": 558
},
{
"epoch": 4.29,
"grad_norm": 0.07107849419116974,
"learning_rate": 1.859726313719135e-05,
"loss": 0.0129,
"step": 559
},
{
"epoch": 4.3,
"grad_norm": 0.07813014090061188,
"learning_rate": 1.853502863244647e-05,
"loss": 0.0141,
"step": 560
},
{
"epoch": 4.31,
"grad_norm": 0.06971003860235214,
"learning_rate": 1.847280838606202e-05,
"loss": 0.013,
"step": 561
},
{
"epoch": 4.31,
"grad_norm": 0.070152148604393,
"learning_rate": 1.8410603003618895e-05,
"loss": 0.0129,
"step": 562
},
{
"epoch": 4.32,
"grad_norm": 0.0758005678653717,
"learning_rate": 1.8348413090553356e-05,
"loss": 0.013,
"step": 563
},
{
"epoch": 4.33,
"grad_norm": 0.07347214221954346,
"learning_rate": 1.8286239252151084e-05,
"loss": 0.0135,
"step": 564
},
{
"epoch": 4.34,
"grad_norm": 0.06794369220733643,
"learning_rate": 1.8224082093541306e-05,
"loss": 0.0134,
"step": 565
},
{
"epoch": 4.35,
"grad_norm": 0.06773723661899567,
"learning_rate": 1.8161942219690915e-05,
"loss": 0.0135,
"step": 566
},
{
"epoch": 4.35,
"grad_norm": 0.06954654306173325,
"learning_rate": 1.8099820235398564e-05,
"loss": 0.0132,
"step": 567
},
{
"epoch": 4.36,
"grad_norm": 0.070184126496315,
"learning_rate": 1.80377167452888e-05,
"loss": 0.0139,
"step": 568
},
{
"epoch": 4.37,
"grad_norm": 0.07064472138881683,
"learning_rate": 1.7975632353806166e-05,
"loss": 0.0137,
"step": 569
},
{
"epoch": 4.38,
"grad_norm": 0.07439540326595306,
"learning_rate": 1.791356766520931e-05,
"loss": 0.0134,
"step": 570
},
{
"epoch": 4.38,
"grad_norm": 0.07122640311717987,
"learning_rate": 1.7851523283565136e-05,
"loss": 0.0134,
"step": 571
},
{
"epoch": 4.39,
"grad_norm": 0.07025646418333054,
"learning_rate": 1.778949981274288e-05,
"loss": 0.0132,
"step": 572
},
{
"epoch": 4.4,
"grad_norm": 0.06943293660879135,
"learning_rate": 1.772749785640826e-05,
"loss": 0.0139,
"step": 573
},
{
"epoch": 4.41,
"grad_norm": 0.07135516405105591,
"learning_rate": 1.7665518018017622e-05,
"loss": 0.014,
"step": 574
},
{
"epoch": 4.41,
"grad_norm": 0.07798086851835251,
"learning_rate": 1.7603560900812003e-05,
"loss": 0.0146,
"step": 575
},
{
"epoch": 4.42,
"grad_norm": 0.06879287213087082,
"learning_rate": 1.754162710781133e-05,
"loss": 0.0127,
"step": 576
},
{
"epoch": 4.43,
"grad_norm": 0.06986016780138016,
"learning_rate": 1.7479717241808504e-05,
"loss": 0.0131,
"step": 577
},
{
"epoch": 4.44,
"grad_norm": 0.06606993824243546,
"learning_rate": 1.7417831905363542e-05,
"loss": 0.0133,
"step": 578
},
{
"epoch": 4.45,
"grad_norm": 0.06856866180896759,
"learning_rate": 1.7355971700797732e-05,
"loss": 0.0138,
"step": 579
},
{
"epoch": 4.45,
"grad_norm": 0.0680074468255043,
"learning_rate": 1.729413723018775e-05,
"loss": 0.0134,
"step": 580
},
{
"epoch": 4.46,
"grad_norm": 0.07777101546525955,
"learning_rate": 1.7232329095359813e-05,
"loss": 0.0139,
"step": 581
},
{
"epoch": 4.47,
"grad_norm": 0.07146996259689331,
"learning_rate": 1.71705478978838e-05,
"loss": 0.0147,
"step": 582
},
{
"epoch": 4.48,
"grad_norm": 0.06847533583641052,
"learning_rate": 1.710879423906743e-05,
"loss": 0.0132,
"step": 583
},
{
"epoch": 4.48,
"grad_norm": 0.07164774090051651,
"learning_rate": 1.704706871995038e-05,
"loss": 0.0138,
"step": 584
},
{
"epoch": 4.49,
"grad_norm": 0.07443536818027496,
"learning_rate": 1.6985371941298457e-05,
"loss": 0.0137,
"step": 585
},
{
"epoch": 4.5,
"grad_norm": 0.07483036816120148,
"learning_rate": 1.692370450359773e-05,
"loss": 0.0148,
"step": 586
},
{
"epoch": 4.51,
"grad_norm": 0.06895185261964798,
"learning_rate": 1.686206700704871e-05,
"loss": 0.0138,
"step": 587
},
{
"epoch": 4.51,
"grad_norm": 0.07155775278806686,
"learning_rate": 1.6800460051560485e-05,
"loss": 0.0139,
"step": 588
},
{
"epoch": 4.52,
"grad_norm": 0.08851808309555054,
"learning_rate": 1.6738884236744895e-05,
"loss": 0.0146,
"step": 589
},
{
"epoch": 4.53,
"grad_norm": 0.08131503313779831,
"learning_rate": 1.6677340161910697e-05,
"loss": 0.0139,
"step": 590
},
{
"epoch": 4.54,
"grad_norm": 0.06781022995710373,
"learning_rate": 1.6615828426057712e-05,
"loss": 0.014,
"step": 591
},
{
"epoch": 4.55,
"grad_norm": 0.06458062678575516,
"learning_rate": 1.6554349627871018e-05,
"loss": 0.0134,
"step": 592
},
{
"epoch": 4.55,
"grad_norm": 0.0716966986656189,
"learning_rate": 1.6492904365715122e-05,
"loss": 0.0139,
"step": 593
},
{
"epoch": 4.56,
"grad_norm": 0.06861037015914917,
"learning_rate": 1.6431493237628132e-05,
"loss": 0.0139,
"step": 594
},
{
"epoch": 4.57,
"grad_norm": 0.07515101134777069,
"learning_rate": 1.6370116841315913e-05,
"loss": 0.0136,
"step": 595
},
{
"epoch": 4.58,
"grad_norm": 0.06479217112064362,
"learning_rate": 1.6308775774146318e-05,
"loss": 0.0131,
"step": 596
},
{
"epoch": 4.58,
"grad_norm": 0.0692434012889862,
"learning_rate": 1.624747063314332e-05,
"loss": 0.0133,
"step": 597
},
{
"epoch": 4.59,
"grad_norm": 0.08492613583803177,
"learning_rate": 1.618620201498124e-05,
"loss": 0.0144,
"step": 598
},
{
"epoch": 4.6,
"grad_norm": 0.07428327947854996,
"learning_rate": 1.6124970515978943e-05,
"loss": 0.0134,
"step": 599
},
{
"epoch": 4.61,
"grad_norm": 0.07502056658267975,
"learning_rate": 1.6063776732093996e-05,
"loss": 0.0146,
"step": 600
},
{
"epoch": 4.61,
"grad_norm": 0.07609640806913376,
"learning_rate": 1.6002621258916887e-05,
"loss": 0.0143,
"step": 601
},
{
"epoch": 4.62,
"grad_norm": 0.07405232638120651,
"learning_rate": 1.594150469166525e-05,
"loss": 0.0142,
"step": 602
},
{
"epoch": 4.63,
"grad_norm": 0.07452607154846191,
"learning_rate": 1.5880427625178035e-05,
"loss": 0.0138,
"step": 603
},
{
"epoch": 4.64,
"grad_norm": 0.06582068651914597,
"learning_rate": 1.5819390653909756e-05,
"loss": 0.014,
"step": 604
},
{
"epoch": 4.64,
"grad_norm": 0.06651710718870163,
"learning_rate": 1.5758394371924678e-05,
"loss": 0.0133,
"step": 605
},
{
"epoch": 4.65,
"grad_norm": 0.07143042236566544,
"learning_rate": 1.5697439372891032e-05,
"loss": 0.0135,
"step": 606
},
{
"epoch": 4.66,
"grad_norm": 0.11468907445669174,
"learning_rate": 1.5636526250075266e-05,
"loss": 0.0136,
"step": 607
},
{
"epoch": 4.67,
"grad_norm": 0.06521432846784592,
"learning_rate": 1.557565559633623e-05,
"loss": 0.0139,
"step": 608
},
{
"epoch": 4.68,
"grad_norm": 0.06903672218322754,
"learning_rate": 1.5514828004119452e-05,
"loss": 0.014,
"step": 609
},
{
"epoch": 4.68,
"grad_norm": 0.0764835774898529,
"learning_rate": 1.5454044065451342e-05,
"loss": 0.0144,
"step": 610
},
{
"epoch": 4.69,
"grad_norm": 0.06870891898870468,
"learning_rate": 1.5393304371933426e-05,
"loss": 0.0142,
"step": 611
},
{
"epoch": 4.7,
"grad_norm": 0.07378222048282623,
"learning_rate": 1.5332609514736593e-05,
"loss": 0.0143,
"step": 612
},
{
"epoch": 4.71,
"grad_norm": 0.0660475343465805,
"learning_rate": 1.5271960084595375e-05,
"loss": 0.0142,
"step": 613
},
{
"epoch": 4.71,
"grad_norm": 0.07662945985794067,
"learning_rate": 1.5211356671802142e-05,
"loss": 0.0138,
"step": 614
},
{
"epoch": 4.72,
"grad_norm": 0.06944695860147476,
"learning_rate": 1.5150799866201396e-05,
"loss": 0.0144,
"step": 615
},
{
"epoch": 4.73,
"grad_norm": 0.07750552147626877,
"learning_rate": 1.5090290257184019e-05,
"loss": 0.0148,
"step": 616
},
{
"epoch": 4.74,
"grad_norm": 0.06265722215175629,
"learning_rate": 1.502982843368152e-05,
"loss": 0.0134,
"step": 617
},
{
"epoch": 4.74,
"grad_norm": 0.0708547905087471,
"learning_rate": 1.4969414984160344e-05,
"loss": 0.0142,
"step": 618
},
{
"epoch": 4.75,
"grad_norm": 0.07105150818824768,
"learning_rate": 1.4909050496616102e-05,
"loss": 0.0147,
"step": 619
},
{
"epoch": 4.76,
"grad_norm": 0.0647650808095932,
"learning_rate": 1.4848735558567852e-05,
"loss": 0.0136,
"step": 620
},
{
"epoch": 4.77,
"grad_norm": 0.07723703980445862,
"learning_rate": 1.4788470757052419e-05,
"loss": 0.0144,
"step": 621
},
{
"epoch": 4.78,
"grad_norm": 0.0752875953912735,
"learning_rate": 1.4728256678618652e-05,
"loss": 0.0163,
"step": 622
},
{
"epoch": 4.78,
"grad_norm": 0.07296160608530045,
"learning_rate": 1.4668093909321704e-05,
"loss": 0.0152,
"step": 623
},
{
"epoch": 4.79,
"grad_norm": 0.06530386954545975,
"learning_rate": 1.4607983034717356e-05,
"loss": 0.0145,
"step": 624
},
{
"epoch": 4.8,
"grad_norm": 0.07218363881111145,
"learning_rate": 1.4547924639856304e-05,
"loss": 0.0144,
"step": 625
},
{
"epoch": 4.81,
"grad_norm": 0.06703885644674301,
"learning_rate": 1.4487919309278459e-05,
"loss": 0.0138,
"step": 626
},
{
"epoch": 4.81,
"grad_norm": 0.0659651830792427,
"learning_rate": 1.442796762700728e-05,
"loss": 0.0144,
"step": 627
},
{
"epoch": 4.82,
"grad_norm": 0.06553196161985397,
"learning_rate": 1.4368070176544055e-05,
"loss": 0.0145,
"step": 628
},
{
"epoch": 4.83,
"grad_norm": 0.0666075050830841,
"learning_rate": 1.4308227540862265e-05,
"loss": 0.0139,
"step": 629
},
{
"epoch": 4.84,
"grad_norm": 0.06991173326969147,
"learning_rate": 1.4248440302401853e-05,
"loss": 0.0149,
"step": 630
},
{
"epoch": 4.84,
"grad_norm": 0.06509461998939514,
"learning_rate": 1.4188709043063616e-05,
"loss": 0.0141,
"step": 631
},
{
"epoch": 4.85,
"grad_norm": 0.06537767499685287,
"learning_rate": 1.4129034344203508e-05,
"loss": 0.0141,
"step": 632
},
{
"epoch": 4.86,
"grad_norm": 0.07116088271141052,
"learning_rate": 1.4069416786626987e-05,
"loss": 0.0142,
"step": 633
},
{
"epoch": 4.87,
"grad_norm": 0.06058025732636452,
"learning_rate": 1.4009856950583358e-05,
"loss": 0.014,
"step": 634
},
{
"epoch": 4.88,
"grad_norm": 0.06636171042919159,
"learning_rate": 1.395035541576014e-05,
"loss": 0.0139,
"step": 635
},
{
"epoch": 4.88,
"grad_norm": 0.06787938624620438,
"learning_rate": 1.38909127612774e-05,
"loss": 0.0142,
"step": 636
},
{
"epoch": 4.89,
"grad_norm": 0.07677444815635681,
"learning_rate": 1.3831529565682145e-05,
"loss": 0.0144,
"step": 637
},
{
"epoch": 4.9,
"grad_norm": 0.07003045827150345,
"learning_rate": 1.3772206406942675e-05,
"loss": 0.0139,
"step": 638
},
{
"epoch": 4.91,
"grad_norm": 0.07242821902036667,
"learning_rate": 1.3712943862442956e-05,
"loss": 0.0145,
"step": 639
},
{
"epoch": 4.91,
"grad_norm": 0.06925047934055328,
"learning_rate": 1.3653742508976993e-05,
"loss": 0.0144,
"step": 640
},
{
"epoch": 4.92,
"grad_norm": 0.06887952238321304,
"learning_rate": 1.3594602922743252e-05,
"loss": 0.0141,
"step": 641
},
{
"epoch": 4.93,
"grad_norm": 0.06720682233572006,
"learning_rate": 1.3535525679339005e-05,
"loss": 0.0137,
"step": 642
},
{
"epoch": 4.94,
"grad_norm": 0.06664611399173737,
"learning_rate": 1.3476511353754763e-05,
"loss": 0.0137,
"step": 643
},
{
"epoch": 4.94,
"grad_norm": 0.07690813392400742,
"learning_rate": 1.3417560520368666e-05,
"loss": 0.0152,
"step": 644
},
{
"epoch": 4.95,
"grad_norm": 0.06674012541770935,
"learning_rate": 1.3358673752940883e-05,
"loss": 0.014,
"step": 645
},
{
"epoch": 4.96,
"grad_norm": 0.06875568628311157,
"learning_rate": 1.329985162460805e-05,
"loss": 0.0144,
"step": 646
},
{
"epoch": 4.97,
"grad_norm": 0.07102454453706741,
"learning_rate": 1.3241094707877673e-05,
"loss": 0.0145,
"step": 647
},
{
"epoch": 4.98,
"grad_norm": 0.06280366331338882,
"learning_rate": 1.3182403574622572e-05,
"loss": 0.0134,
"step": 648
},
{
"epoch": 4.98,
"grad_norm": 0.07154522091150284,
"learning_rate": 1.3123778796075285e-05,
"loss": 0.0145,
"step": 649
},
{
"epoch": 4.99,
"grad_norm": 0.07398602366447449,
"learning_rate": 1.3065220942822548e-05,
"loss": 0.0148,
"step": 650
},
{
"epoch": 5.0,
"grad_norm": 0.06619401276111603,
"learning_rate": 1.3006730584799715e-05,
"loss": 0.0141,
"step": 651
},
{
"epoch": 5.01,
"grad_norm": 0.07948291301727295,
"learning_rate": 1.2948308291285223e-05,
"loss": 0.0119,
"step": 652
},
{
"epoch": 5.01,
"grad_norm": 0.05343073979020119,
"learning_rate": 1.2889954630895041e-05,
"loss": 0.009,
"step": 653
},
{
"epoch": 5.02,
"grad_norm": 0.05225706100463867,
"learning_rate": 1.2831670171577148e-05,
"loss": 0.0087,
"step": 654
},
{
"epoch": 5.03,
"grad_norm": 0.05560747906565666,
"learning_rate": 1.277345548060599e-05,
"loss": 0.0092,
"step": 655
},
{
"epoch": 5.04,
"grad_norm": 0.053426165133714676,
"learning_rate": 1.2715311124576974e-05,
"loss": 0.0089,
"step": 656
},
{
"epoch": 5.04,
"grad_norm": 0.05312531813979149,
"learning_rate": 1.265723766940095e-05,
"loss": 0.0091,
"step": 657
},
{
"epoch": 5.05,
"grad_norm": 0.051174722611904144,
"learning_rate": 1.2599235680298698e-05,
"loss": 0.0088,
"step": 658
},
{
"epoch": 5.06,
"grad_norm": 0.05924823880195618,
"learning_rate": 1.254130572179541e-05,
"loss": 0.009,
"step": 659
},
{
"epoch": 5.07,
"grad_norm": 0.05730690434575081,
"learning_rate": 1.2483448357715242e-05,
"loss": 0.0089,
"step": 660
},
{
"epoch": 5.07,
"grad_norm": 0.06067182496190071,
"learning_rate": 1.2425664151175783e-05,
"loss": 0.0091,
"step": 661
},
{
"epoch": 5.08,
"grad_norm": 0.05133102834224701,
"learning_rate": 1.2367953664582593e-05,
"loss": 0.0093,
"step": 662
},
{
"epoch": 5.09,
"grad_norm": 0.05600114166736603,
"learning_rate": 1.2310317459623734e-05,
"loss": 0.0092,
"step": 663
},
{
"epoch": 5.1,
"grad_norm": 0.05484625697135925,
"learning_rate": 1.225275609726428e-05,
"loss": 0.0085,
"step": 664
},
{
"epoch": 5.11,
"grad_norm": 0.05700346827507019,
"learning_rate": 1.219527013774089e-05,
"loss": 0.0092,
"step": 665
},
{
"epoch": 5.11,
"grad_norm": 0.05756707116961479,
"learning_rate": 1.2137860140556319e-05,
"loss": 0.0087,
"step": 666
},
{
"epoch": 5.12,
"grad_norm": 0.0635422095656395,
"learning_rate": 1.2080526664474003e-05,
"loss": 0.0097,
"step": 667
},
{
"epoch": 5.13,
"grad_norm": 0.05950229987502098,
"learning_rate": 1.2023270267512619e-05,
"loss": 0.0091,
"step": 668
},
{
"epoch": 5.14,
"grad_norm": 0.055414970964193344,
"learning_rate": 1.1966091506940616e-05,
"loss": 0.0087,
"step": 669
},
{
"epoch": 5.14,
"grad_norm": 0.06001156568527222,
"learning_rate": 1.1908990939270835e-05,
"loss": 0.0094,
"step": 670
},
{
"epoch": 5.15,
"grad_norm": 0.06438879668712616,
"learning_rate": 1.1851969120255085e-05,
"loss": 0.0092,
"step": 671
},
{
"epoch": 5.16,
"grad_norm": 0.06288009136915207,
"learning_rate": 1.179502660487871e-05,
"loss": 0.0091,
"step": 672
},
{
"epoch": 5.17,
"grad_norm": 0.05976157262921333,
"learning_rate": 1.1738163947355209e-05,
"loss": 0.0093,
"step": 673
},
{
"epoch": 5.17,
"grad_norm": 0.0599808543920517,
"learning_rate": 1.168138170112084e-05,
"loss": 0.0093,
"step": 674
},
{
"epoch": 5.18,
"grad_norm": 0.0639665350317955,
"learning_rate": 1.162468041882922e-05,
"loss": 0.0095,
"step": 675
},
{
"epoch": 5.19,
"grad_norm": 0.055945783853530884,
"learning_rate": 1.1568060652345962e-05,
"loss": 0.0086,
"step": 676
},
{
"epoch": 5.2,
"grad_norm": 0.061186812818050385,
"learning_rate": 1.1511522952743306e-05,
"loss": 0.0092,
"step": 677
},
{
"epoch": 5.21,
"grad_norm": 0.06328795105218887,
"learning_rate": 1.1455067870294718e-05,
"loss": 0.0089,
"step": 678
},
{
"epoch": 5.21,
"grad_norm": 0.05885428935289383,
"learning_rate": 1.1398695954469598e-05,
"loss": 0.0093,
"step": 679
},
{
"epoch": 5.22,
"grad_norm": 0.06359921395778656,
"learning_rate": 1.1342407753927862e-05,
"loss": 0.0096,
"step": 680
},
{
"epoch": 5.23,
"grad_norm": 0.06182555481791496,
"learning_rate": 1.1286203816514667e-05,
"loss": 0.0093,
"step": 681
},
{
"epoch": 5.24,
"grad_norm": 0.06308590620756149,
"learning_rate": 1.123008468925503e-05,
"loss": 0.0096,
"step": 682
},
{
"epoch": 5.24,
"grad_norm": 0.06224444881081581,
"learning_rate": 1.1174050918348543e-05,
"loss": 0.0098,
"step": 683
},
{
"epoch": 5.25,
"grad_norm": 0.06342342495918274,
"learning_rate": 1.1118103049164004e-05,
"loss": 0.0102,
"step": 684
},
{
"epoch": 5.26,
"grad_norm": 0.061808157712221146,
"learning_rate": 1.1062241626234171e-05,
"loss": 0.009,
"step": 685
},
{
"epoch": 5.27,
"grad_norm": 0.05987200140953064,
"learning_rate": 1.1006467193250421e-05,
"loss": 0.0097,
"step": 686
},
{
"epoch": 5.27,
"grad_norm": 0.05916636437177658,
"learning_rate": 1.0950780293057476e-05,
"loss": 0.0094,
"step": 687
},
{
"epoch": 5.28,
"grad_norm": 0.06484086811542511,
"learning_rate": 1.08951814676481e-05,
"loss": 0.0097,
"step": 688
},
{
"epoch": 5.29,
"grad_norm": 0.06139872595667839,
"learning_rate": 1.0839671258157856e-05,
"loss": 0.0101,
"step": 689
},
{
"epoch": 5.3,
"grad_norm": 0.06333613395690918,
"learning_rate": 1.0784250204859798e-05,
"loss": 0.0089,
"step": 690
},
{
"epoch": 5.31,
"grad_norm": 0.061021000146865845,
"learning_rate": 1.0728918847159258e-05,
"loss": 0.0098,
"step": 691
},
{
"epoch": 5.31,
"grad_norm": 0.06266611814498901,
"learning_rate": 1.067367772358856e-05,
"loss": 0.0099,
"step": 692
},
{
"epoch": 5.32,
"grad_norm": 0.06387548893690109,
"learning_rate": 1.0618527371801798e-05,
"loss": 0.0098,
"step": 693
},
{
"epoch": 5.33,
"grad_norm": 0.06081629917025566,
"learning_rate": 1.0563468328569603e-05,
"loss": 0.0098,
"step": 694
},
{
"epoch": 5.34,
"grad_norm": 0.06264431774616241,
"learning_rate": 1.0508501129773888e-05,
"loss": 0.0101,
"step": 695
},
{
"epoch": 5.34,
"grad_norm": 0.06257108598947525,
"learning_rate": 1.045362631040268e-05,
"loss": 0.0098,
"step": 696
},
{
"epoch": 5.35,
"grad_norm": 0.05395418778061867,
"learning_rate": 1.0398844404544897e-05,
"loss": 0.0096,
"step": 697
},
{
"epoch": 5.36,
"grad_norm": 0.0616784542798996,
"learning_rate": 1.0344155945385106e-05,
"loss": 0.0101,
"step": 698
},
{
"epoch": 5.37,
"grad_norm": 0.06094779074192047,
"learning_rate": 1.0289561465198405e-05,
"loss": 0.0098,
"step": 699
},
{
"epoch": 5.37,
"grad_norm": 0.06029227748513222,
"learning_rate": 1.0235061495345206e-05,
"loss": 0.0099,
"step": 700
},
{
"epoch": 5.38,
"grad_norm": 0.059027206152677536,
"learning_rate": 1.0180656566266039e-05,
"loss": 0.0095,
"step": 701
},
{
"epoch": 5.39,
"grad_norm": 0.05769060552120209,
"learning_rate": 1.012634720747644e-05,
"loss": 0.0093,
"step": 702
},
{
"epoch": 5.4,
"grad_norm": 0.0685461089015007,
"learning_rate": 1.0072133947561763e-05,
"loss": 0.0102,
"step": 703
},
{
"epoch": 5.4,
"grad_norm": 0.061841391026973724,
"learning_rate": 1.001801731417205e-05,
"loss": 0.0101,
"step": 704
},
{
"epoch": 5.41,
"grad_norm": 0.06308455020189285,
"learning_rate": 9.963997834016897e-06,
"loss": 0.0095,
"step": 705
},
{
"epoch": 5.42,
"grad_norm": 0.06159483268857002,
"learning_rate": 9.910076032860299e-06,
"loss": 0.0102,
"step": 706
},
{
"epoch": 5.43,
"grad_norm": 0.057194169610738754,
"learning_rate": 9.856252435515566e-06,
"loss": 0.0096,
"step": 707
},
{
"epoch": 5.44,
"grad_norm": 0.060037873685359955,
"learning_rate": 9.80252756584021e-06,
"loss": 0.0097,
"step": 708
},
{
"epoch": 5.44,
"grad_norm": 0.05958379805088043,
"learning_rate": 9.748901946730838e-06,
"loss": 0.0099,
"step": 709
},
{
"epoch": 5.45,
"grad_norm": 0.06121755763888359,
"learning_rate": 9.695376100118066e-06,
"loss": 0.0102,
"step": 710
},
{
"epoch": 5.46,
"grad_norm": 0.061053842306137085,
"learning_rate": 9.641950546961441e-06,
"loss": 0.0099,
"step": 711
},
{
"epoch": 5.47,
"grad_norm": 0.06203647330403328,
"learning_rate": 9.588625807244347e-06,
"loss": 0.0101,
"step": 712
},
{
"epoch": 5.47,
"grad_norm": 0.06038645654916763,
"learning_rate": 9.535402399968994e-06,
"loss": 0.01,
"step": 713
},
{
"epoch": 5.48,
"grad_norm": 0.06025896593928337,
"learning_rate": 9.48228084315132e-06,
"loss": 0.0104,
"step": 714
},
{
"epoch": 5.49,
"grad_norm": 0.05739811435341835,
"learning_rate": 9.429261653815972e-06,
"loss": 0.0102,
"step": 715
},
{
"epoch": 5.5,
"grad_norm": 0.057722482830286026,
"learning_rate": 9.376345347991276e-06,
"loss": 0.0101,
"step": 716
},
{
"epoch": 5.5,
"grad_norm": 0.16650202870368958,
"learning_rate": 9.323532440704196e-06,
"loss": 0.0106,
"step": 717
},
{
"epoch": 5.51,
"grad_norm": 0.06275496631860733,
"learning_rate": 9.27082344597532e-06,
"loss": 0.0106,
"step": 718
},
{
"epoch": 5.52,
"grad_norm": 0.06536982953548431,
"learning_rate": 9.2182188768139e-06,
"loss": 0.0099,
"step": 719
},
{
"epoch": 5.53,
"grad_norm": 0.06038543954491615,
"learning_rate": 9.165719245212812e-06,
"loss": 0.0104,
"step": 720
},
{
"epoch": 5.54,
"grad_norm": 0.05523300915956497,
"learning_rate": 9.113325062143592e-06,
"loss": 0.0099,
"step": 721
},
{
"epoch": 5.54,
"grad_norm": 0.0561201237142086,
"learning_rate": 9.061036837551467e-06,
"loss": 0.0099,
"step": 722
},
{
"epoch": 5.55,
"grad_norm": 0.059468839317560196,
"learning_rate": 9.008855080350363e-06,
"loss": 0.0104,
"step": 723
},
{
"epoch": 5.56,
"grad_norm": 0.05762971192598343,
"learning_rate": 8.956780298417997e-06,
"loss": 0.01,
"step": 724
},
{
"epoch": 5.57,
"grad_norm": 0.058596912771463394,
"learning_rate": 8.904812998590904e-06,
"loss": 0.01,
"step": 725
},
{
"epoch": 5.57,
"grad_norm": 0.05691235885024071,
"learning_rate": 8.852953686659513e-06,
"loss": 0.0097,
"step": 726
},
{
"epoch": 5.58,
"grad_norm": 0.0685015544295311,
"learning_rate": 8.801202867363207e-06,
"loss": 0.0106,
"step": 727
},
{
"epoch": 5.59,
"grad_norm": 0.05923609063029289,
"learning_rate": 8.749561044385453e-06,
"loss": 0.0106,
"step": 728
},
{
"epoch": 5.6,
"grad_norm": 0.060426920652389526,
"learning_rate": 8.698028720348835e-06,
"loss": 0.0105,
"step": 729
},
{
"epoch": 5.6,
"grad_norm": 0.06719065457582474,
"learning_rate": 8.646606396810234e-06,
"loss": 0.0109,
"step": 730
},
{
"epoch": 5.61,
"grad_norm": 0.06025005131959915,
"learning_rate": 8.595294574255893e-06,
"loss": 0.0104,
"step": 731
},
{
"epoch": 5.62,
"grad_norm": 0.05733707547187805,
"learning_rate": 8.544093752096572e-06,
"loss": 0.0099,
"step": 732
},
{
"epoch": 5.63,
"grad_norm": 0.06899592280387878,
"learning_rate": 8.49300442866268e-06,
"loss": 0.0114,
"step": 733
},
{
"epoch": 5.64,
"grad_norm": 0.0604536272585392,
"learning_rate": 8.442027101199414e-06,
"loss": 0.0106,
"step": 734
},
{
"epoch": 5.64,
"grad_norm": 0.05913786217570305,
"learning_rate": 8.39116226586195e-06,
"loss": 0.0105,
"step": 735
},
{
"epoch": 5.65,
"grad_norm": 0.058253396302461624,
"learning_rate": 8.340410417710562e-06,
"loss": 0.0103,
"step": 736
},
{
"epoch": 5.66,
"grad_norm": 0.06156881898641586,
"learning_rate": 8.289772050705869e-06,
"loss": 0.0106,
"step": 737
},
{
"epoch": 5.67,
"grad_norm": 0.06145387887954712,
"learning_rate": 8.239247657703974e-06,
"loss": 0.0106,
"step": 738
},
{
"epoch": 5.67,
"grad_norm": 0.060531727969646454,
"learning_rate": 8.188837730451706e-06,
"loss": 0.0107,
"step": 739
},
{
"epoch": 5.68,
"grad_norm": 0.06284269690513611,
"learning_rate": 8.138542759581783e-06,
"loss": 0.0102,
"step": 740
},
{
"epoch": 5.69,
"grad_norm": 0.061300307512283325,
"learning_rate": 8.088363234608097e-06,
"loss": 0.0112,
"step": 741
},
{
"epoch": 5.7,
"grad_norm": 0.05871857330203056,
"learning_rate": 8.03829964392091e-06,
"loss": 0.0104,
"step": 742
},
{
"epoch": 5.7,
"grad_norm": 0.06662297248840332,
"learning_rate": 7.988352474782118e-06,
"loss": 0.0104,
"step": 743
},
{
"epoch": 5.71,
"grad_norm": 0.057630885392427444,
"learning_rate": 7.938522213320503e-06,
"loss": 0.0099,
"step": 744
},
{
"epoch": 5.72,
"grad_norm": 0.06206595152616501,
"learning_rate": 7.888809344526988e-06,
"loss": 0.0106,
"step": 745
},
{
"epoch": 5.73,
"grad_norm": 0.062052857130765915,
"learning_rate": 7.839214352249933e-06,
"loss": 0.0107,
"step": 746
},
{
"epoch": 5.74,
"grad_norm": 0.05820086970925331,
"learning_rate": 7.789737719190433e-06,
"loss": 0.0102,
"step": 747
},
{
"epoch": 5.74,
"grad_norm": 0.056203752756118774,
"learning_rate": 7.740379926897602e-06,
"loss": 0.0101,
"step": 748
},
{
"epoch": 5.75,
"grad_norm": 0.06261992454528809,
"learning_rate": 7.691141455763903e-06,
"loss": 0.0113,
"step": 749
},
{
"epoch": 5.76,
"grad_norm": 0.057759836316108704,
"learning_rate": 7.642022785020455e-06,
"loss": 0.0104,
"step": 750
},
{
"epoch": 5.77,
"grad_norm": 0.056587763130664825,
"learning_rate": 7.5930243927323645e-06,
"loss": 0.0099,
"step": 751
},
{
"epoch": 5.77,
"grad_norm": 0.057172466069459915,
"learning_rate": 7.544146755794104e-06,
"loss": 0.0108,
"step": 752
},
{
"epoch": 5.78,
"grad_norm": 0.06228978931903839,
"learning_rate": 7.495390349924847e-06,
"loss": 0.0108,
"step": 753
},
{
"epoch": 5.79,
"grad_norm": 0.06040652096271515,
"learning_rate": 7.446755649663835e-06,
"loss": 0.0109,
"step": 754
},
{
"epoch": 5.8,
"grad_norm": 0.06216125562787056,
"learning_rate": 7.3982431283657805e-06,
"loss": 0.011,
"step": 755
},
{
"epoch": 5.8,
"grad_norm": 0.0665709525346756,
"learning_rate": 7.349853258196226e-06,
"loss": 0.0107,
"step": 756
},
{
"epoch": 5.81,
"grad_norm": 0.05716193839907646,
"learning_rate": 7.30158651012697e-06,
"loss": 0.0104,
"step": 757
},
{
"epoch": 5.82,
"grad_norm": 0.05837683007121086,
"learning_rate": 7.253443353931497e-06,
"loss": 0.0102,
"step": 758
},
{
"epoch": 5.83,
"grad_norm": 0.06044989079236984,
"learning_rate": 7.2054242581803776e-06,
"loss": 0.01,
"step": 759
},
{
"epoch": 5.83,
"grad_norm": 0.05887800455093384,
"learning_rate": 7.157529690236724e-06,
"loss": 0.0104,
"step": 760
},
{
"epoch": 5.84,
"grad_norm": 0.057960037142038345,
"learning_rate": 7.109760116251638e-06,
"loss": 0.0104,
"step": 761
},
{
"epoch": 5.85,
"grad_norm": 0.058715853840112686,
"learning_rate": 7.06211600115966e-06,
"loss": 0.0109,
"step": 762
},
{
"epoch": 5.86,
"grad_norm": 0.28305932879447937,
"learning_rate": 7.014597808674268e-06,
"loss": 0.0109,
"step": 763
},
{
"epoch": 5.87,
"grad_norm": 0.0576317235827446,
"learning_rate": 6.9672060012833554e-06,
"loss": 0.0103,
"step": 764
},
{
"epoch": 5.87,
"grad_norm": 0.06595731526613235,
"learning_rate": 6.919941040244728e-06,
"loss": 0.0126,
"step": 765
},
{
"epoch": 5.88,
"grad_norm": 0.06269898265600204,
"learning_rate": 6.872803385581594e-06,
"loss": 0.0106,
"step": 766
},
{
"epoch": 5.89,
"grad_norm": 0.056339289993047714,
"learning_rate": 6.825793496078135e-06,
"loss": 0.0102,
"step": 767
},
{
"epoch": 5.9,
"grad_norm": 0.06980512291193008,
"learning_rate": 6.778911829274986e-06,
"loss": 0.0106,
"step": 768
},
{
"epoch": 5.9,
"grad_norm": 0.06072579324245453,
"learning_rate": 6.732158841464826e-06,
"loss": 0.0104,
"step": 769
},
{
"epoch": 5.91,
"grad_norm": 0.05551678687334061,
"learning_rate": 6.685534987687918e-06,
"loss": 0.0106,
"step": 770
},
{
"epoch": 5.92,
"grad_norm": 0.05761561915278435,
"learning_rate": 6.639040721727674e-06,
"loss": 0.0102,
"step": 771
},
{
"epoch": 5.93,
"grad_norm": 0.07146608084440231,
"learning_rate": 6.5926764961062585e-06,
"loss": 0.0104,
"step": 772
},
{
"epoch": 5.93,
"grad_norm": 0.05681779980659485,
"learning_rate": 6.546442762080149e-06,
"loss": 0.0102,
"step": 773
},
{
"epoch": 5.94,
"grad_norm": 0.0556565523147583,
"learning_rate": 6.500339969635794e-06,
"loss": 0.01,
"step": 774
},
{
"epoch": 5.95,
"grad_norm": 0.06225784868001938,
"learning_rate": 6.4543685674851834e-06,
"loss": 0.0113,
"step": 775
},
{
"epoch": 5.96,
"grad_norm": 0.06186867505311966,
"learning_rate": 6.408529003061512e-06,
"loss": 0.0113,
"step": 776
},
{
"epoch": 5.97,
"grad_norm": 0.06227191537618637,
"learning_rate": 6.362821722514827e-06,
"loss": 0.0108,
"step": 777
},
{
"epoch": 5.97,
"grad_norm": 0.06345570832490921,
"learning_rate": 6.317247170707666e-06,
"loss": 0.0111,
"step": 778
},
{
"epoch": 5.98,
"grad_norm": 0.06830250471830368,
"learning_rate": 6.271805791210723e-06,
"loss": 0.0107,
"step": 779
},
{
"epoch": 5.99,
"grad_norm": 0.06165685877203941,
"learning_rate": 6.226498026298569e-06,
"loss": 0.0103,
"step": 780
},
{
"epoch": 6.0,
"grad_norm": 0.05917961150407791,
"learning_rate": 6.18132431694531e-06,
"loss": 0.0103,
"step": 781
},
{
"epoch": 6.0,
"grad_norm": 0.05313841626048088,
"learning_rate": 6.1362851028203095e-06,
"loss": 0.0087,
"step": 782
},
{
"epoch": 6.01,
"grad_norm": 0.05500919744372368,
"learning_rate": 6.091380822283914e-06,
"loss": 0.0071,
"step": 783
},
{
"epoch": 6.02,
"grad_norm": 0.04682999849319458,
"learning_rate": 6.046611912383171e-06,
"loss": 0.0072,
"step": 784
},
{
"epoch": 6.03,
"grad_norm": 0.04729432985186577,
"learning_rate": 6.0019788088475815e-06,
"loss": 0.0071,
"step": 785
},
{
"epoch": 6.03,
"grad_norm": 0.054288025945425034,
"learning_rate": 5.957481946084876e-06,
"loss": 0.0076,
"step": 786
},
{
"epoch": 6.04,
"grad_norm": 0.04912056028842926,
"learning_rate": 5.913121757176763e-06,
"loss": 0.0072,
"step": 787
},
{
"epoch": 6.05,
"grad_norm": 0.04554147273302078,
"learning_rate": 5.868898673874728e-06,
"loss": 0.0071,
"step": 788
},
{
"epoch": 6.06,
"grad_norm": 0.05164865776896477,
"learning_rate": 5.824813126595832e-06,
"loss": 0.008,
"step": 789
},
{
"epoch": 6.07,
"grad_norm": 0.04785354807972908,
"learning_rate": 5.780865544418499e-06,
"loss": 0.007,
"step": 790
},
{
"epoch": 6.07,
"grad_norm": 0.048543281853199005,
"learning_rate": 5.737056355078374e-06,
"loss": 0.007,
"step": 791
},
{
"epoch": 6.08,
"grad_norm": 0.0490332767367363,
"learning_rate": 5.6933859849641416e-06,
"loss": 0.0073,
"step": 792
},
{
"epoch": 6.09,
"grad_norm": 0.04716874659061432,
"learning_rate": 5.6498548591133725e-06,
"loss": 0.0069,
"step": 793
},
{
"epoch": 6.1,
"grad_norm": 0.048864495009183884,
"learning_rate": 5.606463401208408e-06,
"loss": 0.0067,
"step": 794
},
{
"epoch": 6.1,
"grad_norm": 0.04809195548295975,
"learning_rate": 5.563212033572201e-06,
"loss": 0.0072,
"step": 795
},
{
"epoch": 6.11,
"grad_norm": 0.0525406189262867,
"learning_rate": 5.520101177164228e-06,
"loss": 0.0074,
"step": 796
},
{
"epoch": 6.12,
"grad_norm": 0.04640812799334526,
"learning_rate": 5.4771312515764e-06,
"loss": 0.0067,
"step": 797
},
{
"epoch": 6.13,
"grad_norm": 0.051228053867816925,
"learning_rate": 5.434302675028964e-06,
"loss": 0.0072,
"step": 798
},
{
"epoch": 6.13,
"grad_norm": 0.050089213997125626,
"learning_rate": 5.391615864366438e-06,
"loss": 0.0068,
"step": 799
},
{
"epoch": 6.14,
"grad_norm": 0.05177585408091545,
"learning_rate": 5.349071235053554e-06,
"loss": 0.0072,
"step": 800
},
{
"epoch": 6.15,
"grad_norm": 0.053178057074546814,
"learning_rate": 5.306669201171197e-06,
"loss": 0.0073,
"step": 801
},
{
"epoch": 6.16,
"grad_norm": 0.04784577339887619,
"learning_rate": 5.264410175412414e-06,
"loss": 0.0069,
"step": 802
},
{
"epoch": 6.17,
"grad_norm": 0.05324474722146988,
"learning_rate": 5.222294569078366e-06,
"loss": 0.0073,
"step": 803
},
{
"epoch": 6.17,
"grad_norm": 0.05342203751206398,
"learning_rate": 5.180322792074317e-06,
"loss": 0.0073,
"step": 804
},
{
"epoch": 6.18,
"grad_norm": 0.05281771719455719,
"learning_rate": 5.138495252905684e-06,
"loss": 0.0078,
"step": 805
},
{
"epoch": 6.19,
"grad_norm": 0.05480753630399704,
"learning_rate": 5.09681235867403e-06,
"loss": 0.0073,
"step": 806
},
{
"epoch": 6.2,
"grad_norm": 0.05215248838067055,
"learning_rate": 5.055274515073094e-06,
"loss": 0.0072,
"step": 807
},
{
"epoch": 6.2,
"grad_norm": 0.05400507152080536,
"learning_rate": 5.013882126384877e-06,
"loss": 0.0071,
"step": 808
},
{
"epoch": 6.21,
"grad_norm": 0.05833006650209427,
"learning_rate": 4.9726355954756786e-06,
"loss": 0.0079,
"step": 809
},
{
"epoch": 6.22,
"grad_norm": 0.0508585087954998,
"learning_rate": 4.931535323792189e-06,
"loss": 0.007,
"step": 810
},
{
"epoch": 6.23,
"grad_norm": 0.052157700061798096,
"learning_rate": 4.890581711357582e-06,
"loss": 0.0072,
"step": 811
},
{
"epoch": 6.23,
"grad_norm": 0.053852394223213196,
"learning_rate": 4.849775156767598e-06,
"loss": 0.0074,
"step": 812
},
{
"epoch": 6.24,
"grad_norm": 0.05250995606184006,
"learning_rate": 4.809116057186711e-06,
"loss": 0.0071,
"step": 813
},
{
"epoch": 6.25,
"grad_norm": 0.055634334683418274,
"learning_rate": 4.7686048083442105e-06,
"loss": 0.0077,
"step": 814
},
{
"epoch": 6.26,
"grad_norm": 0.05276574566960335,
"learning_rate": 4.7282418045303954e-06,
"loss": 0.0073,
"step": 815
},
{
"epoch": 6.26,
"grad_norm": 0.06131422892212868,
"learning_rate": 4.688027438592706e-06,
"loss": 0.007,
"step": 816
},
{
"epoch": 6.27,
"grad_norm": 0.05558731034398079,
"learning_rate": 4.64796210193192e-06,
"loss": 0.0073,
"step": 817
},
{
"epoch": 6.28,
"grad_norm": 0.05414123833179474,
"learning_rate": 4.608046184498318e-06,
"loss": 0.0074,
"step": 818
},
{
"epoch": 6.29,
"grad_norm": 0.058134667575359344,
"learning_rate": 4.568280074787919e-06,
"loss": 0.0072,
"step": 819
},
{
"epoch": 6.3,
"grad_norm": 0.060313113033771515,
"learning_rate": 4.528664159838681e-06,
"loss": 0.0081,
"step": 820
},
{
"epoch": 6.3,
"grad_norm": 0.054234303534030914,
"learning_rate": 4.489198825226742e-06,
"loss": 0.0074,
"step": 821
},
{
"epoch": 6.31,
"grad_norm": 0.06013951450586319,
"learning_rate": 4.449884455062643e-06,
"loss": 0.0078,
"step": 822
},
{
"epoch": 6.32,
"grad_norm": 0.05726248770952225,
"learning_rate": 4.410721431987639e-06,
"loss": 0.0077,
"step": 823
},
{
"epoch": 6.33,
"grad_norm": 0.05651944503188133,
"learning_rate": 4.3717101371699155e-06,
"loss": 0.0074,
"step": 824
},
{
"epoch": 6.33,
"grad_norm": 0.05372058227658272,
"learning_rate": 4.3328509503009286e-06,
"loss": 0.0071,
"step": 825
},
{
"epoch": 6.34,
"grad_norm": 0.05767044425010681,
"learning_rate": 4.2941442495916805e-06,
"loss": 0.0075,
"step": 826
},
{
"epoch": 6.35,
"grad_norm": 0.05186377093195915,
"learning_rate": 4.255590411769057e-06,
"loss": 0.0073,
"step": 827
},
{
"epoch": 6.36,
"grad_norm": 0.056691158562898636,
"learning_rate": 4.217189812072131e-06,
"loss": 0.0074,
"step": 828
},
{
"epoch": 6.36,
"grad_norm": 0.057784903794527054,
"learning_rate": 4.1789428242485485e-06,
"loss": 0.0075,
"step": 829
},
{
"epoch": 6.37,
"grad_norm": 0.06017537787556648,
"learning_rate": 4.1408498205508654e-06,
"loss": 0.0076,
"step": 830
},
{
"epoch": 6.38,
"grad_norm": 0.059429608285427094,
"learning_rate": 4.102911171732933e-06,
"loss": 0.0078,
"step": 831
},
{
"epoch": 6.39,
"grad_norm": 0.055457279086112976,
"learning_rate": 4.065127247046294e-06,
"loss": 0.0075,
"step": 832
},
{
"epoch": 6.4,
"grad_norm": 0.05795317515730858,
"learning_rate": 4.0274984142365724e-06,
"loss": 0.0077,
"step": 833
},
{
"epoch": 6.4,
"grad_norm": 0.07282472401857376,
"learning_rate": 3.990025039539902e-06,
"loss": 0.0075,
"step": 834
},
{
"epoch": 6.41,
"grad_norm": 0.05700072646141052,
"learning_rate": 3.952707487679379e-06,
"loss": 0.0076,
"step": 835
},
{
"epoch": 6.42,
"grad_norm": 0.05585889890789986,
"learning_rate": 3.915546121861491e-06,
"loss": 0.0076,
"step": 836
},
{
"epoch": 6.43,
"grad_norm": 0.05375934764742851,
"learning_rate": 3.878541303772587e-06,
"loss": 0.0071,
"step": 837
},
{
"epoch": 6.43,
"grad_norm": 0.06071506068110466,
"learning_rate": 3.8416933935753675e-06,
"loss": 0.0079,
"step": 838
},
{
"epoch": 6.44,
"grad_norm": 0.05467228963971138,
"learning_rate": 3.8050027499053534e-06,
"loss": 0.0074,
"step": 839
},
{
"epoch": 6.45,
"grad_norm": 0.05671767145395279,
"learning_rate": 3.7684697298674234e-06,
"loss": 0.0076,
"step": 840
},
{
"epoch": 6.46,
"grad_norm": 0.05431463569402695,
"learning_rate": 3.732094689032333e-06,
"loss": 0.008,
"step": 841
},
{
"epoch": 6.46,
"grad_norm": 0.05200907960534096,
"learning_rate": 3.6958779814332448e-06,
"loss": 0.0069,
"step": 842
},
{
"epoch": 6.47,
"grad_norm": 0.05386371910572052,
"learning_rate": 3.659819959562274e-06,
"loss": 0.0074,
"step": 843
},
{
"epoch": 6.48,
"grad_norm": 0.0579906590282917,
"learning_rate": 3.6239209743670878e-06,
"loss": 0.0075,
"step": 844
},
{
"epoch": 6.49,
"grad_norm": 0.05794265866279602,
"learning_rate": 3.588181375247457e-06,
"loss": 0.0077,
"step": 845
},
{
"epoch": 6.5,
"grad_norm": 0.0597093403339386,
"learning_rate": 3.5526015100518785e-06,
"loss": 0.0078,
"step": 846
},
{
"epoch": 6.5,
"grad_norm": 0.06104637682437897,
"learning_rate": 3.5171817250741815e-06,
"loss": 0.008,
"step": 847
},
{
"epoch": 6.51,
"grad_norm": 0.053399961441755295,
"learning_rate": 3.4819223650501543e-06,
"loss": 0.0074,
"step": 848
},
{
"epoch": 6.52,
"grad_norm": 0.06024890020489693,
"learning_rate": 3.4468237731541976e-06,
"loss": 0.008,
"step": 849
},
{
"epoch": 6.53,
"grad_norm": 0.054280877113342285,
"learning_rate": 3.411886290995965e-06,
"loss": 0.0075,
"step": 850
},
{
"epoch": 6.53,
"grad_norm": 0.05808375030755997,
"learning_rate": 3.377110258617062e-06,
"loss": 0.0082,
"step": 851
},
{
"epoch": 6.54,
"grad_norm": 0.06474139541387558,
"learning_rate": 3.342496014487733e-06,
"loss": 0.0074,
"step": 852
},
{
"epoch": 6.55,
"grad_norm": 0.05558295547962189,
"learning_rate": 3.3080438955035365e-06,
"loss": 0.0079,
"step": 853
},
{
"epoch": 6.56,
"grad_norm": 0.054287388920784,
"learning_rate": 3.2737542369821117e-06,
"loss": 0.0074,
"step": 854
},
{
"epoch": 6.56,
"grad_norm": 0.062351033091545105,
"learning_rate": 3.2396273726598925e-06,
"loss": 0.0085,
"step": 855
},
{
"epoch": 6.57,
"grad_norm": 0.05515924468636513,
"learning_rate": 3.2056636346888403e-06,
"loss": 0.0078,
"step": 856
},
{
"epoch": 6.58,
"grad_norm": 0.05551682412624359,
"learning_rate": 3.1718633536332486e-06,
"loss": 0.0079,
"step": 857
},
{
"epoch": 6.59,
"grad_norm": 0.054728493094444275,
"learning_rate": 3.1382268584665043e-06,
"loss": 0.0079,
"step": 858
},
{
"epoch": 6.6,
"grad_norm": 0.054819852113723755,
"learning_rate": 3.104754476567886e-06,
"loss": 0.0075,
"step": 859
},
{
"epoch": 6.6,
"grad_norm": 0.057974983006715775,
"learning_rate": 3.0714465337193864e-06,
"loss": 0.0076,
"step": 860
},
{
"epoch": 6.61,
"grad_norm": 0.05336621403694153,
"learning_rate": 3.03830335410253e-06,
"loss": 0.0074,
"step": 861
},
{
"epoch": 6.62,
"grad_norm": 0.05819125473499298,
"learning_rate": 3.0053252602952177e-06,
"loss": 0.0079,
"step": 862
},
{
"epoch": 6.63,
"grad_norm": 0.06345003843307495,
"learning_rate": 2.9725125732686065e-06,
"loss": 0.0081,
"step": 863
},
{
"epoch": 6.63,
"grad_norm": 0.05852939561009407,
"learning_rate": 2.9398656123839696e-06,
"loss": 0.0078,
"step": 864
},
{
"epoch": 6.64,
"grad_norm": 0.05667982995510101,
"learning_rate": 2.9073846953895877e-06,
"loss": 0.0078,
"step": 865
},
{
"epoch": 6.65,
"grad_norm": 0.059139229357242584,
"learning_rate": 2.875070138417666e-06,
"loss": 0.0077,
"step": 866
},
{
"epoch": 6.66,
"grad_norm": 0.05454858019948006,
"learning_rate": 2.8429222559812376e-06,
"loss": 0.0074,
"step": 867
},
{
"epoch": 6.66,
"grad_norm": 0.05825025215744972,
"learning_rate": 2.810941360971127e-06,
"loss": 0.0078,
"step": 868
},
{
"epoch": 6.67,
"grad_norm": 0.05783124268054962,
"learning_rate": 2.7791277646528893e-06,
"loss": 0.0079,
"step": 869
},
{
"epoch": 6.68,
"grad_norm": 0.055337242782115936,
"learning_rate": 2.7474817766637896e-06,
"loss": 0.0079,
"step": 870
},
{
"epoch": 6.69,
"grad_norm": 0.057424288243055344,
"learning_rate": 2.716003705009782e-06,
"loss": 0.0076,
"step": 871
},
{
"epoch": 6.69,
"grad_norm": 0.05621185526251793,
"learning_rate": 2.6846938560625124e-06,
"loss": 0.008,
"step": 872
},
{
"epoch": 6.7,
"grad_norm": 0.057556215673685074,
"learning_rate": 2.653552534556334e-06,
"loss": 0.0082,
"step": 873
},
{
"epoch": 6.71,
"grad_norm": 0.061952680349349976,
"learning_rate": 2.6225800435853567e-06,
"loss": 0.0083,
"step": 874
},
{
"epoch": 6.72,
"grad_norm": 0.06053425744175911,
"learning_rate": 2.5917766846004846e-06,
"loss": 0.008,
"step": 875
},
{
"epoch": 6.73,
"grad_norm": 0.056171003729104996,
"learning_rate": 2.561142757406476e-06,
"loss": 0.0076,
"step": 876
},
{
"epoch": 6.73,
"grad_norm": 0.058878615498542786,
"learning_rate": 2.53067856015905e-06,
"loss": 0.008,
"step": 877
},
{
"epoch": 6.74,
"grad_norm": 0.059263814240694046,
"learning_rate": 2.500384389361943e-06,
"loss": 0.0079,
"step": 878
},
{
"epoch": 6.75,
"grad_norm": 0.05734783038496971,
"learning_rate": 2.4702605398640668e-06,
"loss": 0.0075,
"step": 879
},
{
"epoch": 6.76,
"grad_norm": 0.05672736093401909,
"learning_rate": 2.440307304856615e-06,
"loss": 0.0078,
"step": 880
},
{
"epoch": 6.76,
"grad_norm": 0.055851615965366364,
"learning_rate": 2.410524975870221e-06,
"loss": 0.0076,
"step": 881
},
{
"epoch": 6.77,
"grad_norm": 0.05455206707119942,
"learning_rate": 2.3809138427720946e-06,
"loss": 0.0079,
"step": 882
},
{
"epoch": 6.78,
"grad_norm": 0.05759782716631889,
"learning_rate": 2.3514741937632433e-06,
"loss": 0.0079,
"step": 883
},
{
"epoch": 6.79,
"grad_norm": 0.056483786553144455,
"learning_rate": 2.3222063153756237e-06,
"loss": 0.0075,
"step": 884
},
{
"epoch": 6.79,
"grad_norm": 0.059396564960479736,
"learning_rate": 2.293110492469388e-06,
"loss": 0.0079,
"step": 885
},
{
"epoch": 6.8,
"grad_norm": 0.058641623705625534,
"learning_rate": 2.264187008230092e-06,
"loss": 0.0078,
"step": 886
},
{
"epoch": 6.81,
"grad_norm": 0.0548095777630806,
"learning_rate": 2.235436144165939e-06,
"loss": 0.0079,
"step": 887
},
{
"epoch": 6.82,
"grad_norm": 0.05954607576131821,
"learning_rate": 2.2068581801050557e-06,
"loss": 0.0082,
"step": 888
},
{
"epoch": 6.83,
"grad_norm": 0.05823349207639694,
"learning_rate": 2.178453394192741e-06,
"loss": 0.0084,
"step": 889
},
{
"epoch": 6.83,
"grad_norm": 0.055938269942998886,
"learning_rate": 2.1502220628887916e-06,
"loss": 0.0079,
"step": 890
},
{
"epoch": 6.84,
"grad_norm": 0.06316045671701431,
"learning_rate": 2.1221644609647773e-06,
"loss": 0.0086,
"step": 891
},
{
"epoch": 6.85,
"grad_norm": 0.06259404867887497,
"learning_rate": 2.094280861501403e-06,
"loss": 0.0082,
"step": 892
},
{
"epoch": 6.86,
"grad_norm": 0.05663475766777992,
"learning_rate": 2.0665715358858195e-06,
"loss": 0.0075,
"step": 893
},
{
"epoch": 6.86,
"grad_norm": 0.06532254070043564,
"learning_rate": 2.039036753809003e-06,
"loss": 0.0085,
"step": 894
},
{
"epoch": 6.87,
"grad_norm": 0.05870622768998146,
"learning_rate": 2.01167678326311e-06,
"loss": 0.008,
"step": 895
},
{
"epoch": 6.88,
"grad_norm": 0.05485890060663223,
"learning_rate": 1.984491890538893e-06,
"loss": 0.0079,
"step": 896
},
{
"epoch": 6.89,
"grad_norm": 0.06565118581056595,
"learning_rate": 1.9574823402230913e-06,
"loss": 0.0086,
"step": 897
},
{
"epoch": 6.89,
"grad_norm": 0.05567467212677002,
"learning_rate": 1.930648395195862e-06,
"loss": 0.0081,
"step": 898
},
{
"epoch": 6.9,
"grad_norm": 0.060943204909563065,
"learning_rate": 1.9039903166282258e-06,
"loss": 0.0082,
"step": 899
},
{
"epoch": 6.91,
"grad_norm": 0.056582383811473846,
"learning_rate": 1.87750836397951e-06,
"loss": 0.0083,
"step": 900
},
{
"epoch": 6.92,
"grad_norm": 0.05313001945614815,
"learning_rate": 1.8512027949948353e-06,
"loss": 0.0076,
"step": 901
},
{
"epoch": 6.93,
"grad_norm": 0.056625016033649445,
"learning_rate": 1.825073865702607e-06,
"loss": 0.008,
"step": 902
},
{
"epoch": 6.93,
"grad_norm": 0.06274149566888809,
"learning_rate": 1.7991218304120272e-06,
"loss": 0.0083,
"step": 903
},
{
"epoch": 6.94,
"grad_norm": 0.05565561726689339,
"learning_rate": 1.773346941710603e-06,
"loss": 0.0082,
"step": 904
},
{
"epoch": 6.95,
"grad_norm": 0.059408511966466904,
"learning_rate": 1.747749450461711e-06,
"loss": 0.0081,
"step": 905
},
{
"epoch": 6.96,
"grad_norm": 0.057353563606739044,
"learning_rate": 1.722329605802129e-06,
"loss": 0.0079,
"step": 906
},
{
"epoch": 6.96,
"grad_norm": 0.05773783475160599,
"learning_rate": 1.6970876551396309e-06,
"loss": 0.0081,
"step": 907
},
{
"epoch": 6.97,
"grad_norm": 0.05856630578637123,
"learning_rate": 1.6720238441505787e-06,
"loss": 0.008,
"step": 908
},
{
"epoch": 6.98,
"grad_norm": 0.057708289474248886,
"learning_rate": 1.6471384167775228e-06,
"loss": 0.008,
"step": 909
},
{
"epoch": 6.99,
"grad_norm": 0.06603064388036728,
"learning_rate": 1.6224316152268315e-06,
"loss": 0.0081,
"step": 910
},
{
"epoch": 6.99,
"grad_norm": 0.05872678756713867,
"learning_rate": 1.5979036799663283e-06,
"loss": 0.0082,
"step": 911
},
{
"epoch": 7.0,
"grad_norm": 0.056233812123537064,
"learning_rate": 1.5735548497229603e-06,
"loss": 0.0076,
"step": 912
},
{
"epoch": 7.01,
"grad_norm": 0.0477331280708313,
"learning_rate": 1.5493853614804687e-06,
"loss": 0.006,
"step": 913
},
{
"epoch": 7.02,
"grad_norm": 0.05013463646173477,
"learning_rate": 1.5253954504770873e-06,
"loss": 0.0065,
"step": 914
},
{
"epoch": 7.02,
"grad_norm": 0.04779525101184845,
"learning_rate": 1.5015853502032495e-06,
"loss": 0.0065,
"step": 915
},
{
"epoch": 7.03,
"grad_norm": 0.05848625302314758,
"learning_rate": 1.4779552923993156e-06,
"loss": 0.0066,
"step": 916
},
{
"epoch": 7.04,
"grad_norm": 0.04815210774540901,
"learning_rate": 1.454505507053312e-06,
"loss": 0.0064,
"step": 917
},
{
"epoch": 7.05,
"grad_norm": 0.044312819838523865,
"learning_rate": 1.4312362223987086e-06,
"loss": 0.0059,
"step": 918
},
{
"epoch": 7.06,
"grad_norm": 0.05249709263443947,
"learning_rate": 1.4081476649121827e-06,
"loss": 0.0062,
"step": 919
},
{
"epoch": 7.06,
"grad_norm": 0.04766643047332764,
"learning_rate": 1.3852400593114168e-06,
"loss": 0.0061,
"step": 920
},
{
"epoch": 7.07,
"grad_norm": 0.04619971662759781,
"learning_rate": 1.3625136285529151e-06,
"loss": 0.0061,
"step": 921
},
{
"epoch": 7.08,
"grad_norm": 0.045929890125989914,
"learning_rate": 1.339968593829839e-06,
"loss": 0.0062,
"step": 922
},
{
"epoch": 7.09,
"grad_norm": 0.050117433071136475,
"learning_rate": 1.3176051745698383e-06,
"loss": 0.0065,
"step": 923
},
{
"epoch": 7.09,
"grad_norm": 0.048868823796510696,
"learning_rate": 1.2954235884329314e-06,
"loss": 0.0067,
"step": 924
},
{
"epoch": 7.1,
"grad_norm": 0.046368200331926346,
"learning_rate": 1.2734240513093776e-06,
"loss": 0.0061,
"step": 925
},
{
"epoch": 7.11,
"grad_norm": 0.047504231333732605,
"learning_rate": 1.2516067773175822e-06,
"loss": 0.0058,
"step": 926
},
{
"epoch": 7.12,
"grad_norm": 0.04716791212558746,
"learning_rate": 1.229971978802007e-06,
"loss": 0.0061,
"step": 927
},
{
"epoch": 7.12,
"grad_norm": 0.047293853014707565,
"learning_rate": 1.2085198663310993e-06,
"loss": 0.0063,
"step": 928
},
{
"epoch": 7.13,
"grad_norm": 0.045351818203926086,
"learning_rate": 1.1872506486952639e-06,
"loss": 0.0063,
"step": 929
},
{
"epoch": 7.14,
"grad_norm": 0.04756508022546768,
"learning_rate": 1.1661645329047987e-06,
"loss": 0.0062,
"step": 930
},
{
"epoch": 7.15,
"grad_norm": 0.05072380602359772,
"learning_rate": 1.1452617241879072e-06,
"loss": 0.0061,
"step": 931
},
{
"epoch": 7.16,
"grad_norm": 0.051270999014377594,
"learning_rate": 1.1245424259886906e-06,
"loss": 0.0063,
"step": 932
},
{
"epoch": 7.16,
"grad_norm": 0.049795448780059814,
"learning_rate": 1.1040068399651683e-06,
"loss": 0.0061,
"step": 933
},
{
"epoch": 7.17,
"grad_norm": 0.05117611214518547,
"learning_rate": 1.0836551659873073e-06,
"loss": 0.0064,
"step": 934
},
{
"epoch": 7.18,
"grad_norm": 0.05119408667087555,
"learning_rate": 1.0634876021350937e-06,
"loss": 0.0066,
"step": 935
},
{
"epoch": 7.19,
"grad_norm": 0.05104885250329971,
"learning_rate": 1.0435043446965886e-06,
"loss": 0.0064,
"step": 936
},
{
"epoch": 7.19,
"grad_norm": 0.047611530870199203,
"learning_rate": 1.0237055881660328e-06,
"loss": 0.0063,
"step": 937
},
{
"epoch": 7.2,
"grad_norm": 0.052856262773275375,
"learning_rate": 1.0040915252419415e-06,
"loss": 0.006,
"step": 938
},
{
"epoch": 7.21,
"grad_norm": 0.05077509582042694,
"learning_rate": 9.84662346825227e-07,
"loss": 0.0063,
"step": 939
},
{
"epoch": 7.22,
"grad_norm": 0.05041491985321045,
"learning_rate": 9.654182420173553e-07,
"loss": 0.0058,
"step": 940
},
{
"epoch": 7.22,
"grad_norm": 0.04933353513479233,
"learning_rate": 9.463593981184904e-07,
"loss": 0.0062,
"step": 941
},
{
"epoch": 7.23,
"grad_norm": 0.05266598239541054,
"learning_rate": 9.274860006256858e-07,
"loss": 0.0065,
"step": 942
},
{
"epoch": 7.24,
"grad_norm": 0.05019863694906235,
"learning_rate": 9.087982332310674e-07,
"loss": 0.0064,
"step": 943
},
{
"epoch": 7.25,
"grad_norm": 0.0492464043200016,
"learning_rate": 8.902962778200464e-07,
"loss": 0.0061,
"step": 944
},
{
"epoch": 7.26,
"grad_norm": 0.0491764098405838,
"learning_rate": 8.719803144695516e-07,
"loss": 0.0073,
"step": 945
},
{
"epoch": 7.26,
"grad_norm": 0.050660282373428345,
"learning_rate": 8.538505214462845e-07,
"loss": 0.006,
"step": 946
},
{
"epoch": 7.27,
"grad_norm": 0.053767457604408264,
"learning_rate": 8.359070752049713e-07,
"loss": 0.0062,
"step": 947
},
{
"epoch": 7.28,
"grad_norm": 0.05090208351612091,
"learning_rate": 8.181501503866495e-07,
"loss": 0.0063,
"step": 948
},
{
"epoch": 7.29,
"grad_norm": 0.04544226825237274,
"learning_rate": 8.005799198169817e-07,
"loss": 0.0058,
"step": 949
},
{
"epoch": 7.29,
"grad_norm": 0.0457618348300457,
"learning_rate": 7.831965545045439e-07,
"loss": 0.0062,
"step": 950
},
{
"epoch": 7.3,
"grad_norm": 0.046685926616191864,
"learning_rate": 7.660002236391873e-07,
"loss": 0.0059,
"step": 951
},
{
"epoch": 7.31,
"grad_norm": 0.05885709077119827,
"learning_rate": 7.489910945903856e-07,
"loss": 0.0065,
"step": 952
},
{
"epoch": 7.32,
"grad_norm": 0.052940983325242996,
"learning_rate": 7.321693329056012e-07,
"loss": 0.0064,
"step": 953
},
{
"epoch": 7.32,
"grad_norm": 0.04693511128425598,
"learning_rate": 7.155351023086776e-07,
"loss": 0.006,
"step": 954
},
{
"epoch": 7.33,
"grad_norm": 0.0517946220934391,
"learning_rate": 6.990885646982449e-07,
"loss": 0.0064,
"step": 955
},
{
"epoch": 7.34,
"grad_norm": 0.04994283244013786,
"learning_rate": 6.828298801461364e-07,
"loss": 0.0063,
"step": 956
},
{
"epoch": 7.35,
"grad_norm": 0.04989714175462723,
"learning_rate": 6.667592068958484e-07,
"loss": 0.0062,
"step": 957
},
{
"epoch": 7.36,
"grad_norm": 0.05148443952202797,
"learning_rate": 6.508767013609852e-07,
"loss": 0.006,
"step": 958
},
{
"epoch": 7.36,
"grad_norm": 0.05187150835990906,
"learning_rate": 6.351825181237359e-07,
"loss": 0.0069,
"step": 959
},
{
"epoch": 7.37,
"grad_norm": 0.05286801606416702,
"learning_rate": 6.19676809933385e-07,
"loss": 0.0064,
"step": 960
},
{
"epoch": 7.38,
"grad_norm": 0.05000551417469978,
"learning_rate": 6.043597277048107e-07,
"loss": 0.0061,
"step": 961
},
{
"epoch": 7.39,
"grad_norm": 0.05106586962938309,
"learning_rate": 5.89231420517018e-07,
"loss": 0.0062,
"step": 962
},
{
"epoch": 7.39,
"grad_norm": 0.050964899361133575,
"learning_rate": 5.742920356116988e-07,
"loss": 0.006,
"step": 963
},
{
"epoch": 7.4,
"grad_norm": 0.05159709230065346,
"learning_rate": 5.595417183917851e-07,
"loss": 0.0067,
"step": 964
},
{
"epoch": 7.41,
"grad_norm": 0.05571756884455681,
"learning_rate": 5.449806124200385e-07,
"loss": 0.0062,
"step": 965
},
{
"epoch": 7.42,
"grad_norm": 0.05309474468231201,
"learning_rate": 5.306088594176628e-07,
"loss": 0.0068,
"step": 966
},
{
"epoch": 7.42,
"grad_norm": 0.048645325005054474,
"learning_rate": 5.164265992629069e-07,
"loss": 0.0061,
"step": 967
},
{
"epoch": 7.43,
"grad_norm": 0.04711391031742096,
"learning_rate": 5.024339699897151e-07,
"loss": 0.006,
"step": 968
},
{
"epoch": 7.44,
"grad_norm": 0.04801530763506889,
"learning_rate": 4.886311077863814e-07,
"loss": 0.0062,
"step": 969
},
{
"epoch": 7.45,
"grad_norm": 0.05131622403860092,
"learning_rate": 4.750181469942217e-07,
"loss": 0.0066,
"step": 970
},
{
"epoch": 7.45,
"grad_norm": 0.053375840187072754,
"learning_rate": 4.615952201062768e-07,
"loss": 0.0067,
"step": 971
},
{
"epoch": 7.46,
"grad_norm": 0.048528678715229034,
"learning_rate": 4.4836245776600063e-07,
"loss": 0.0061,
"step": 972
},
{
"epoch": 7.47,
"grad_norm": 0.0490819588303566,
"learning_rate": 4.353199887660142e-07,
"loss": 0.006,
"step": 973
},
{
"epoch": 7.48,
"grad_norm": 0.04894344508647919,
"learning_rate": 4.224679400468312e-07,
"loss": 0.0065,
"step": 974
},
{
"epoch": 7.49,
"grad_norm": 0.048594992607831955,
"learning_rate": 4.098064366956367e-07,
"loss": 0.006,
"step": 975
},
{
"epoch": 7.49,
"grad_norm": 0.05195420980453491,
"learning_rate": 3.973356019450658e-07,
"loss": 0.0059,
"step": 976
},
{
"epoch": 7.5,
"grad_norm": 0.04922621697187424,
"learning_rate": 3.8505555717199826e-07,
"loss": 0.0066,
"step": 977
},
{
"epoch": 7.51,
"grad_norm": 0.04782318323850632,
"learning_rate": 3.7296642189638577e-07,
"loss": 0.0062,
"step": 978
},
{
"epoch": 7.52,
"grad_norm": 0.05116094648838043,
"learning_rate": 3.6106831378007744e-07,
"loss": 0.0066,
"step": 979
},
{
"epoch": 7.52,
"grad_norm": 0.05064942687749863,
"learning_rate": 3.4936134862569415e-07,
"loss": 0.006,
"step": 980
},
{
"epoch": 7.53,
"grad_norm": 0.050393134355545044,
"learning_rate": 3.378456403754782e-07,
"loss": 0.0062,
"step": 981
},
{
"epoch": 7.54,
"grad_norm": 0.05107494071125984,
"learning_rate": 3.2652130111020084e-07,
"loss": 0.0063,
"step": 982
},
{
"epoch": 7.55,
"grad_norm": 0.050017185509204865,
"learning_rate": 3.1538844104806343e-07,
"loss": 0.0059,
"step": 983
},
{
"epoch": 7.55,
"grad_norm": 0.05066869780421257,
"learning_rate": 3.044471685436312e-07,
"loss": 0.0062,
"step": 984
},
{
"epoch": 7.56,
"grad_norm": 0.05147412046790123,
"learning_rate": 2.936975900867678e-07,
"loss": 0.0062,
"step": 985
},
{
"epoch": 7.57,
"grad_norm": 0.04994218796491623,
"learning_rate": 2.8313981030161364e-07,
"loss": 0.006,
"step": 986
},
{
"epoch": 7.58,
"grad_norm": 0.051188819110393524,
"learning_rate": 2.7277393194555357e-07,
"loss": 0.0064,
"step": 987
},
{
"epoch": 7.59,
"grad_norm": 0.0506623201072216,
"learning_rate": 2.626000559082242e-07,
"loss": 0.0062,
"step": 988
},
{
"epoch": 7.59,
"grad_norm": 0.04762762412428856,
"learning_rate": 2.5261828121053047e-07,
"loss": 0.0063,
"step": 989
},
{
"epoch": 7.6,
"grad_norm": 0.05141177028417587,
"learning_rate": 2.428287050036837e-07,
"loss": 0.006,
"step": 990
},
{
"epoch": 7.61,
"grad_norm": 0.05564945563673973,
"learning_rate": 2.3323142256825193e-07,
"loss": 0.0065,
"step": 991
},
{
"epoch": 7.62,
"grad_norm": 0.04980726167559624,
"learning_rate": 2.2382652731323783e-07,
"loss": 0.0063,
"step": 992
},
{
"epoch": 7.62,
"grad_norm": 0.05145712196826935,
"learning_rate": 2.1461411077516648e-07,
"loss": 0.0064,
"step": 993
},
{
"epoch": 7.63,
"grad_norm": 0.04716255143284798,
"learning_rate": 2.055942626171903e-07,
"loss": 0.0062,
"step": 994
},
{
"epoch": 7.64,
"grad_norm": 0.055020444095134735,
"learning_rate": 1.9676707062822543e-07,
"loss": 0.0065,
"step": 995
},
{
"epoch": 7.65,
"grad_norm": 0.048993561416864395,
"learning_rate": 1.8813262072209014e-07,
"loss": 0.0059,
"step": 996
},
{
"epoch": 7.65,
"grad_norm": 0.05176731199026108,
"learning_rate": 1.7969099693667225e-07,
"loss": 0.0061,
"step": 997
},
{
"epoch": 7.66,
"grad_norm": 0.0494525283575058,
"learning_rate": 1.714422814331096e-07,
"loss": 0.0063,
"step": 998
},
{
"epoch": 7.67,
"grad_norm": 0.05234352499246597,
"learning_rate": 1.6338655449499086e-07,
"loss": 0.0064,
"step": 999
},
{
"epoch": 7.68,
"grad_norm": 0.05156205594539642,
"learning_rate": 1.5552389452757166e-07,
"loss": 0.0065,
"step": 1000
}
],
"logging_steps": 1.0,
"max_steps": 1040,
"num_input_tokens_seen": 0,
"num_train_epochs": 8,
"save_steps": 100,
"total_flos": 0.0,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}