gsmyrnis's picture
End of training
5b7d28f verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 1065,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0028169014084507044,
"grad_norm": 5.39381152989452,
"learning_rate": 9.345794392523364e-08,
"loss": 0.9214,
"step": 1
},
{
"epoch": 0.005633802816901409,
"grad_norm": 5.710911651711583,
"learning_rate": 1.8691588785046729e-07,
"loss": 0.8778,
"step": 2
},
{
"epoch": 0.008450704225352112,
"grad_norm": 4.807270772834428,
"learning_rate": 2.8037383177570096e-07,
"loss": 0.8196,
"step": 3
},
{
"epoch": 0.011267605633802818,
"grad_norm": 5.480036707909385,
"learning_rate": 3.7383177570093457e-07,
"loss": 0.8791,
"step": 4
},
{
"epoch": 0.014084507042253521,
"grad_norm": 5.318989970133928,
"learning_rate": 4.6728971962616824e-07,
"loss": 0.8627,
"step": 5
},
{
"epoch": 0.016901408450704224,
"grad_norm": 4.809114137429962,
"learning_rate": 5.607476635514019e-07,
"loss": 0.8093,
"step": 6
},
{
"epoch": 0.01971830985915493,
"grad_norm": 5.50714519392,
"learning_rate": 6.542056074766355e-07,
"loss": 0.8247,
"step": 7
},
{
"epoch": 0.022535211267605635,
"grad_norm": 4.502971860173534,
"learning_rate": 7.476635514018691e-07,
"loss": 0.8387,
"step": 8
},
{
"epoch": 0.02535211267605634,
"grad_norm": 4.575448924350796,
"learning_rate": 8.411214953271029e-07,
"loss": 0.8583,
"step": 9
},
{
"epoch": 0.028169014084507043,
"grad_norm": 4.883031846315503,
"learning_rate": 9.345794392523365e-07,
"loss": 0.8986,
"step": 10
},
{
"epoch": 0.030985915492957747,
"grad_norm": 3.558388444517262,
"learning_rate": 1.0280373831775702e-06,
"loss": 0.7943,
"step": 11
},
{
"epoch": 0.03380281690140845,
"grad_norm": 3.3653902333255328,
"learning_rate": 1.1214953271028038e-06,
"loss": 0.703,
"step": 12
},
{
"epoch": 0.036619718309859155,
"grad_norm": 3.5878521578631757,
"learning_rate": 1.2149532710280374e-06,
"loss": 0.8741,
"step": 13
},
{
"epoch": 0.03943661971830986,
"grad_norm": 3.0669396475993276,
"learning_rate": 1.308411214953271e-06,
"loss": 0.7508,
"step": 14
},
{
"epoch": 0.04225352112676056,
"grad_norm": 2.360776473843334,
"learning_rate": 1.4018691588785047e-06,
"loss": 0.7579,
"step": 15
},
{
"epoch": 0.04507042253521127,
"grad_norm": 2.406994148358753,
"learning_rate": 1.4953271028037383e-06,
"loss": 0.7386,
"step": 16
},
{
"epoch": 0.04788732394366197,
"grad_norm": 2.4289659946387627,
"learning_rate": 1.588785046728972e-06,
"loss": 0.8448,
"step": 17
},
{
"epoch": 0.05070422535211268,
"grad_norm": 2.1113226098450655,
"learning_rate": 1.6822429906542057e-06,
"loss": 0.8075,
"step": 18
},
{
"epoch": 0.05352112676056338,
"grad_norm": 2.0005363458129373,
"learning_rate": 1.7757009345794394e-06,
"loss": 0.8043,
"step": 19
},
{
"epoch": 0.056338028169014086,
"grad_norm": 2.0984615538535456,
"learning_rate": 1.869158878504673e-06,
"loss": 0.7654,
"step": 20
},
{
"epoch": 0.059154929577464786,
"grad_norm": 2.454463629584827,
"learning_rate": 1.962616822429907e-06,
"loss": 0.7389,
"step": 21
},
{
"epoch": 0.061971830985915494,
"grad_norm": 2.5756627563324677,
"learning_rate": 2.0560747663551404e-06,
"loss": 0.7537,
"step": 22
},
{
"epoch": 0.0647887323943662,
"grad_norm": 2.3450614128411487,
"learning_rate": 2.149532710280374e-06,
"loss": 0.7403,
"step": 23
},
{
"epoch": 0.0676056338028169,
"grad_norm": 2.21793106800515,
"learning_rate": 2.2429906542056077e-06,
"loss": 0.7722,
"step": 24
},
{
"epoch": 0.07042253521126761,
"grad_norm": 1.7549218172116292,
"learning_rate": 2.3364485981308413e-06,
"loss": 0.7552,
"step": 25
},
{
"epoch": 0.07323943661971831,
"grad_norm": 1.4710972828597284,
"learning_rate": 2.429906542056075e-06,
"loss": 0.7718,
"step": 26
},
{
"epoch": 0.07605633802816901,
"grad_norm": 1.3151046770941317,
"learning_rate": 2.5233644859813085e-06,
"loss": 0.7333,
"step": 27
},
{
"epoch": 0.07887323943661972,
"grad_norm": 1.1938149562869083,
"learning_rate": 2.616822429906542e-06,
"loss": 0.7295,
"step": 28
},
{
"epoch": 0.08169014084507042,
"grad_norm": 1.3259440496917112,
"learning_rate": 2.7102803738317757e-06,
"loss": 0.8544,
"step": 29
},
{
"epoch": 0.08450704225352113,
"grad_norm": 1.249957124569568,
"learning_rate": 2.8037383177570094e-06,
"loss": 0.7156,
"step": 30
},
{
"epoch": 0.08732394366197183,
"grad_norm": 1.184686946843432,
"learning_rate": 2.897196261682243e-06,
"loss": 0.7365,
"step": 31
},
{
"epoch": 0.09014084507042254,
"grad_norm": 1.0223702189060315,
"learning_rate": 2.9906542056074766e-06,
"loss": 0.6805,
"step": 32
},
{
"epoch": 0.09295774647887324,
"grad_norm": 1.1473857275242683,
"learning_rate": 3.08411214953271e-06,
"loss": 0.7861,
"step": 33
},
{
"epoch": 0.09577464788732394,
"grad_norm": 1.030743194655851,
"learning_rate": 3.177570093457944e-06,
"loss": 0.7734,
"step": 34
},
{
"epoch": 0.09859154929577464,
"grad_norm": 1.1418816841824204,
"learning_rate": 3.2710280373831774e-06,
"loss": 0.7042,
"step": 35
},
{
"epoch": 0.10140845070422536,
"grad_norm": 0.8969695635147503,
"learning_rate": 3.3644859813084115e-06,
"loss": 0.6155,
"step": 36
},
{
"epoch": 0.10422535211267606,
"grad_norm": 1.0522251076635003,
"learning_rate": 3.457943925233645e-06,
"loss": 0.7224,
"step": 37
},
{
"epoch": 0.10704225352112676,
"grad_norm": 1.0245713785476949,
"learning_rate": 3.5514018691588787e-06,
"loss": 0.7261,
"step": 38
},
{
"epoch": 0.10985915492957747,
"grad_norm": 0.9472567384114589,
"learning_rate": 3.6448598130841123e-06,
"loss": 0.6792,
"step": 39
},
{
"epoch": 0.11267605633802817,
"grad_norm": 0.9445593106604601,
"learning_rate": 3.738317757009346e-06,
"loss": 0.6477,
"step": 40
},
{
"epoch": 0.11549295774647887,
"grad_norm": 0.9279518039467434,
"learning_rate": 3.8317757009345796e-06,
"loss": 0.7259,
"step": 41
},
{
"epoch": 0.11830985915492957,
"grad_norm": 0.9154015884052205,
"learning_rate": 3.925233644859814e-06,
"loss": 0.6307,
"step": 42
},
{
"epoch": 0.12112676056338029,
"grad_norm": 0.9828247076921207,
"learning_rate": 4.018691588785047e-06,
"loss": 0.7188,
"step": 43
},
{
"epoch": 0.12394366197183099,
"grad_norm": 0.9916044713420874,
"learning_rate": 4.112149532710281e-06,
"loss": 0.7279,
"step": 44
},
{
"epoch": 0.1267605633802817,
"grad_norm": 0.9328317475978831,
"learning_rate": 4.205607476635514e-06,
"loss": 0.6689,
"step": 45
},
{
"epoch": 0.1295774647887324,
"grad_norm": 0.8726360541682836,
"learning_rate": 4.299065420560748e-06,
"loss": 0.71,
"step": 46
},
{
"epoch": 0.1323943661971831,
"grad_norm": 1.0372408545860783,
"learning_rate": 4.392523364485981e-06,
"loss": 0.7656,
"step": 47
},
{
"epoch": 0.1352112676056338,
"grad_norm": 0.9201045709937832,
"learning_rate": 4.485981308411215e-06,
"loss": 0.7125,
"step": 48
},
{
"epoch": 0.13802816901408452,
"grad_norm": 0.8985693053741854,
"learning_rate": 4.579439252336449e-06,
"loss": 0.6673,
"step": 49
},
{
"epoch": 0.14084507042253522,
"grad_norm": 0.8016552690681369,
"learning_rate": 4.6728971962616825e-06,
"loss": 0.6028,
"step": 50
},
{
"epoch": 0.14366197183098592,
"grad_norm": 0.9375805058634695,
"learning_rate": 4.766355140186917e-06,
"loss": 0.7053,
"step": 51
},
{
"epoch": 0.14647887323943662,
"grad_norm": 0.8235300906984858,
"learning_rate": 4.85981308411215e-06,
"loss": 0.709,
"step": 52
},
{
"epoch": 0.14929577464788732,
"grad_norm": 0.8885825883694105,
"learning_rate": 4.953271028037384e-06,
"loss": 0.7004,
"step": 53
},
{
"epoch": 0.15211267605633802,
"grad_norm": 0.8993913153528437,
"learning_rate": 5.046728971962617e-06,
"loss": 0.6802,
"step": 54
},
{
"epoch": 0.15492957746478872,
"grad_norm": 0.8690379834087235,
"learning_rate": 5.14018691588785e-06,
"loss": 0.7182,
"step": 55
},
{
"epoch": 0.15774647887323945,
"grad_norm": 1.0168474666547533,
"learning_rate": 5.233644859813084e-06,
"loss": 0.7477,
"step": 56
},
{
"epoch": 0.16056338028169015,
"grad_norm": 0.8719423698232311,
"learning_rate": 5.3271028037383174e-06,
"loss": 0.74,
"step": 57
},
{
"epoch": 0.16338028169014085,
"grad_norm": 1.026890979433,
"learning_rate": 5.4205607476635515e-06,
"loss": 0.7706,
"step": 58
},
{
"epoch": 0.16619718309859155,
"grad_norm": 0.8955556843274681,
"learning_rate": 5.514018691588785e-06,
"loss": 0.7149,
"step": 59
},
{
"epoch": 0.16901408450704225,
"grad_norm": 0.8694856642141823,
"learning_rate": 5.607476635514019e-06,
"loss": 0.7079,
"step": 60
},
{
"epoch": 0.17183098591549295,
"grad_norm": 0.9902510235542388,
"learning_rate": 5.700934579439253e-06,
"loss": 0.6935,
"step": 61
},
{
"epoch": 0.17464788732394365,
"grad_norm": 0.8421158549635541,
"learning_rate": 5.794392523364486e-06,
"loss": 0.6476,
"step": 62
},
{
"epoch": 0.17746478873239438,
"grad_norm": 0.8618342169352607,
"learning_rate": 5.88785046728972e-06,
"loss": 0.6251,
"step": 63
},
{
"epoch": 0.18028169014084508,
"grad_norm": 0.9627093955885784,
"learning_rate": 5.981308411214953e-06,
"loss": 0.7648,
"step": 64
},
{
"epoch": 0.18309859154929578,
"grad_norm": 0.8615872113332586,
"learning_rate": 6.074766355140187e-06,
"loss": 0.7052,
"step": 65
},
{
"epoch": 0.18591549295774648,
"grad_norm": 0.7938477476235895,
"learning_rate": 6.16822429906542e-06,
"loss": 0.6856,
"step": 66
},
{
"epoch": 0.18873239436619718,
"grad_norm": 0.9314578053407399,
"learning_rate": 6.2616822429906544e-06,
"loss": 0.675,
"step": 67
},
{
"epoch": 0.19154929577464788,
"grad_norm": 0.8414192598073413,
"learning_rate": 6.355140186915888e-06,
"loss": 0.7094,
"step": 68
},
{
"epoch": 0.19436619718309858,
"grad_norm": 0.9088665681264082,
"learning_rate": 6.448598130841122e-06,
"loss": 0.6834,
"step": 69
},
{
"epoch": 0.19718309859154928,
"grad_norm": 0.902387432849795,
"learning_rate": 6.542056074766355e-06,
"loss": 0.7152,
"step": 70
},
{
"epoch": 0.2,
"grad_norm": 0.9642448781976154,
"learning_rate": 6.635514018691589e-06,
"loss": 0.685,
"step": 71
},
{
"epoch": 0.2028169014084507,
"grad_norm": 0.8474244419117023,
"learning_rate": 6.728971962616823e-06,
"loss": 0.6712,
"step": 72
},
{
"epoch": 0.2056338028169014,
"grad_norm": 0.9474565823957127,
"learning_rate": 6.822429906542056e-06,
"loss": 0.6786,
"step": 73
},
{
"epoch": 0.2084507042253521,
"grad_norm": 0.9289574016521713,
"learning_rate": 6.91588785046729e-06,
"loss": 0.7304,
"step": 74
},
{
"epoch": 0.2112676056338028,
"grad_norm": 0.9127008704865647,
"learning_rate": 7.009345794392523e-06,
"loss": 0.7452,
"step": 75
},
{
"epoch": 0.2140845070422535,
"grad_norm": 0.9662686159807351,
"learning_rate": 7.1028037383177574e-06,
"loss": 0.7134,
"step": 76
},
{
"epoch": 0.21690140845070421,
"grad_norm": 1.0772478476012384,
"learning_rate": 7.196261682242991e-06,
"loss": 0.6781,
"step": 77
},
{
"epoch": 0.21971830985915494,
"grad_norm": 1.0550174155184562,
"learning_rate": 7.289719626168225e-06,
"loss": 0.6633,
"step": 78
},
{
"epoch": 0.22253521126760564,
"grad_norm": 0.8802108160521099,
"learning_rate": 7.383177570093458e-06,
"loss": 0.6876,
"step": 79
},
{
"epoch": 0.22535211267605634,
"grad_norm": 0.9429110893196399,
"learning_rate": 7.476635514018692e-06,
"loss": 0.7335,
"step": 80
},
{
"epoch": 0.22816901408450704,
"grad_norm": 0.938383459599816,
"learning_rate": 7.570093457943926e-06,
"loss": 0.7397,
"step": 81
},
{
"epoch": 0.23098591549295774,
"grad_norm": 0.9096159486194378,
"learning_rate": 7.663551401869159e-06,
"loss": 0.6888,
"step": 82
},
{
"epoch": 0.23380281690140844,
"grad_norm": 0.8917639799429912,
"learning_rate": 7.757009345794392e-06,
"loss": 0.7029,
"step": 83
},
{
"epoch": 0.23661971830985915,
"grad_norm": 0.8979990857654783,
"learning_rate": 7.850467289719627e-06,
"loss": 0.768,
"step": 84
},
{
"epoch": 0.23943661971830985,
"grad_norm": 0.8993682822907498,
"learning_rate": 7.94392523364486e-06,
"loss": 0.7967,
"step": 85
},
{
"epoch": 0.24225352112676057,
"grad_norm": 0.9519036462618192,
"learning_rate": 8.037383177570094e-06,
"loss": 0.7236,
"step": 86
},
{
"epoch": 0.24507042253521127,
"grad_norm": 0.9320429116107417,
"learning_rate": 8.130841121495327e-06,
"loss": 0.7031,
"step": 87
},
{
"epoch": 0.24788732394366197,
"grad_norm": 1.1968431066091298,
"learning_rate": 8.224299065420562e-06,
"loss": 0.8184,
"step": 88
},
{
"epoch": 0.2507042253521127,
"grad_norm": 1.015717114563735,
"learning_rate": 8.317757009345795e-06,
"loss": 0.7366,
"step": 89
},
{
"epoch": 0.2535211267605634,
"grad_norm": 0.8628312778604383,
"learning_rate": 8.411214953271028e-06,
"loss": 0.6516,
"step": 90
},
{
"epoch": 0.2563380281690141,
"grad_norm": 0.9207687578294539,
"learning_rate": 8.504672897196263e-06,
"loss": 0.7463,
"step": 91
},
{
"epoch": 0.2591549295774648,
"grad_norm": 0.9076929603375471,
"learning_rate": 8.598130841121496e-06,
"loss": 0.7232,
"step": 92
},
{
"epoch": 0.2619718309859155,
"grad_norm": 0.8551379886884969,
"learning_rate": 8.69158878504673e-06,
"loss": 0.6387,
"step": 93
},
{
"epoch": 0.2647887323943662,
"grad_norm": 0.874389963082252,
"learning_rate": 8.785046728971963e-06,
"loss": 0.5928,
"step": 94
},
{
"epoch": 0.2676056338028169,
"grad_norm": 0.9772206166305732,
"learning_rate": 8.878504672897197e-06,
"loss": 0.7431,
"step": 95
},
{
"epoch": 0.2704225352112676,
"grad_norm": 1.0472604316389014,
"learning_rate": 8.97196261682243e-06,
"loss": 0.6916,
"step": 96
},
{
"epoch": 0.27323943661971833,
"grad_norm": 0.9161522458184038,
"learning_rate": 9.065420560747664e-06,
"loss": 0.7618,
"step": 97
},
{
"epoch": 0.27605633802816903,
"grad_norm": 0.9694215423750022,
"learning_rate": 9.158878504672899e-06,
"loss": 0.7371,
"step": 98
},
{
"epoch": 0.27887323943661974,
"grad_norm": 0.9914139238391411,
"learning_rate": 9.252336448598132e-06,
"loss": 0.7141,
"step": 99
},
{
"epoch": 0.28169014084507044,
"grad_norm": 1.0202196257679186,
"learning_rate": 9.345794392523365e-06,
"loss": 0.7255,
"step": 100
},
{
"epoch": 0.28450704225352114,
"grad_norm": 0.9601713649988124,
"learning_rate": 9.439252336448598e-06,
"loss": 0.7014,
"step": 101
},
{
"epoch": 0.28732394366197184,
"grad_norm": 0.8615212518278053,
"learning_rate": 9.532710280373833e-06,
"loss": 0.6382,
"step": 102
},
{
"epoch": 0.29014084507042254,
"grad_norm": 1.2055501454669908,
"learning_rate": 9.626168224299066e-06,
"loss": 0.7745,
"step": 103
},
{
"epoch": 0.29295774647887324,
"grad_norm": 0.8554711418047789,
"learning_rate": 9.7196261682243e-06,
"loss": 0.678,
"step": 104
},
{
"epoch": 0.29577464788732394,
"grad_norm": 0.9686952732108494,
"learning_rate": 9.813084112149533e-06,
"loss": 0.7142,
"step": 105
},
{
"epoch": 0.29859154929577464,
"grad_norm": 0.8833954097474772,
"learning_rate": 9.906542056074768e-06,
"loss": 0.6931,
"step": 106
},
{
"epoch": 0.30140845070422534,
"grad_norm": 0.8996092526129557,
"learning_rate": 1e-05,
"loss": 0.7937,
"step": 107
},
{
"epoch": 0.30422535211267604,
"grad_norm": 0.9372563478579707,
"learning_rate": 9.999973115104874e-06,
"loss": 0.7803,
"step": 108
},
{
"epoch": 0.30704225352112674,
"grad_norm": 0.8475734380103672,
"learning_rate": 9.999892460708615e-06,
"loss": 0.701,
"step": 109
},
{
"epoch": 0.30985915492957744,
"grad_norm": 0.8534917997572844,
"learning_rate": 9.999758037678576e-06,
"loss": 0.7371,
"step": 110
},
{
"epoch": 0.3126760563380282,
"grad_norm": 0.8434562246377327,
"learning_rate": 9.999569847460335e-06,
"loss": 0.7268,
"step": 111
},
{
"epoch": 0.3154929577464789,
"grad_norm": 0.8281726646550165,
"learning_rate": 9.999327892077683e-06,
"loss": 0.6853,
"step": 112
},
{
"epoch": 0.3183098591549296,
"grad_norm": 0.9271927724711224,
"learning_rate": 9.9990321741326e-06,
"loss": 0.6452,
"step": 113
},
{
"epoch": 0.3211267605633803,
"grad_norm": 0.8848897647327078,
"learning_rate": 9.998682696805224e-06,
"loss": 0.7243,
"step": 114
},
{
"epoch": 0.323943661971831,
"grad_norm": 0.8610699024398826,
"learning_rate": 9.998279463853819e-06,
"loss": 0.6448,
"step": 115
},
{
"epoch": 0.3267605633802817,
"grad_norm": 0.8677582610096554,
"learning_rate": 9.997822479614733e-06,
"loss": 0.6943,
"step": 116
},
{
"epoch": 0.3295774647887324,
"grad_norm": 0.8471875616608381,
"learning_rate": 9.997311749002358e-06,
"loss": 0.685,
"step": 117
},
{
"epoch": 0.3323943661971831,
"grad_norm": 0.8682796024046532,
"learning_rate": 9.996747277509068e-06,
"loss": 0.745,
"step": 118
},
{
"epoch": 0.3352112676056338,
"grad_norm": 0.8341754007010466,
"learning_rate": 9.996129071205167e-06,
"loss": 0.7004,
"step": 119
},
{
"epoch": 0.3380281690140845,
"grad_norm": 0.8707994349697389,
"learning_rate": 9.99545713673882e-06,
"loss": 0.7398,
"step": 120
},
{
"epoch": 0.3408450704225352,
"grad_norm": 0.8082724338122019,
"learning_rate": 9.994731481335979e-06,
"loss": 0.7061,
"step": 121
},
{
"epoch": 0.3436619718309859,
"grad_norm": 0.9249095219534365,
"learning_rate": 9.993952112800314e-06,
"loss": 0.696,
"step": 122
},
{
"epoch": 0.3464788732394366,
"grad_norm": 0.944187877477761,
"learning_rate": 9.993119039513124e-06,
"loss": 0.73,
"step": 123
},
{
"epoch": 0.3492957746478873,
"grad_norm": 1.0057102257733501,
"learning_rate": 9.992232270433239e-06,
"loss": 0.7613,
"step": 124
},
{
"epoch": 0.352112676056338,
"grad_norm": 0.8333808705246333,
"learning_rate": 9.991291815096941e-06,
"loss": 0.7452,
"step": 125
},
{
"epoch": 0.35492957746478876,
"grad_norm": 0.8949365308913656,
"learning_rate": 9.990297683617844e-06,
"loss": 0.7289,
"step": 126
},
{
"epoch": 0.35774647887323946,
"grad_norm": 0.8224952199583537,
"learning_rate": 9.989249886686798e-06,
"loss": 0.697,
"step": 127
},
{
"epoch": 0.36056338028169016,
"grad_norm": 0.8463117609492709,
"learning_rate": 9.988148435571766e-06,
"loss": 0.6907,
"step": 128
},
{
"epoch": 0.36338028169014086,
"grad_norm": 0.8097036868252439,
"learning_rate": 9.98699334211771e-06,
"loss": 0.6631,
"step": 129
},
{
"epoch": 0.36619718309859156,
"grad_norm": 0.8671869864436087,
"learning_rate": 9.985784618746455e-06,
"loss": 0.6995,
"step": 130
},
{
"epoch": 0.36901408450704226,
"grad_norm": 0.8736051125955552,
"learning_rate": 9.984522278456558e-06,
"loss": 0.7605,
"step": 131
},
{
"epoch": 0.37183098591549296,
"grad_norm": 0.8479930342434334,
"learning_rate": 9.983206334823179e-06,
"loss": 0.7521,
"step": 132
},
{
"epoch": 0.37464788732394366,
"grad_norm": 0.8797715865284296,
"learning_rate": 9.981836801997916e-06,
"loss": 0.6507,
"step": 133
},
{
"epoch": 0.37746478873239436,
"grad_norm": 0.8625026171210389,
"learning_rate": 9.98041369470867e-06,
"loss": 0.7442,
"step": 134
},
{
"epoch": 0.38028169014084506,
"grad_norm": 0.9674565189535757,
"learning_rate": 9.978937028259477e-06,
"loss": 0.6997,
"step": 135
},
{
"epoch": 0.38309859154929576,
"grad_norm": 0.8204379275136592,
"learning_rate": 9.977406818530346e-06,
"loss": 0.7128,
"step": 136
},
{
"epoch": 0.38591549295774646,
"grad_norm": 0.9145363476852016,
"learning_rate": 9.975823081977089e-06,
"loss": 0.7915,
"step": 137
},
{
"epoch": 0.38873239436619716,
"grad_norm": 0.8473502868722297,
"learning_rate": 9.974185835631141e-06,
"loss": 0.6332,
"step": 138
},
{
"epoch": 0.39154929577464787,
"grad_norm": 0.8811521109738086,
"learning_rate": 9.972495097099379e-06,
"loss": 0.6646,
"step": 139
},
{
"epoch": 0.39436619718309857,
"grad_norm": 0.8420515822936742,
"learning_rate": 9.970750884563938e-06,
"loss": 0.6463,
"step": 140
},
{
"epoch": 0.3971830985915493,
"grad_norm": 0.8796307650058819,
"learning_rate": 9.968953216782005e-06,
"loss": 0.7702,
"step": 141
},
{
"epoch": 0.4,
"grad_norm": 0.9113098059629853,
"learning_rate": 9.96710211308562e-06,
"loss": 0.6678,
"step": 142
},
{
"epoch": 0.4028169014084507,
"grad_norm": 0.82206028117941,
"learning_rate": 9.965197593381483e-06,
"loss": 0.7489,
"step": 143
},
{
"epoch": 0.4056338028169014,
"grad_norm": 0.880880001482584,
"learning_rate": 9.963239678150712e-06,
"loss": 0.6868,
"step": 144
},
{
"epoch": 0.4084507042253521,
"grad_norm": 0.9060474241316776,
"learning_rate": 9.961228388448648e-06,
"loss": 0.7212,
"step": 145
},
{
"epoch": 0.4112676056338028,
"grad_norm": 0.9310497997261157,
"learning_rate": 9.959163745904613e-06,
"loss": 0.7736,
"step": 146
},
{
"epoch": 0.4140845070422535,
"grad_norm": 0.7931697883942757,
"learning_rate": 9.95704577272169e-06,
"loss": 0.6676,
"step": 147
},
{
"epoch": 0.4169014084507042,
"grad_norm": 0.8932141674301038,
"learning_rate": 9.954874491676474e-06,
"loss": 0.7192,
"step": 148
},
{
"epoch": 0.4197183098591549,
"grad_norm": 0.9415400016839601,
"learning_rate": 9.952649926118827e-06,
"loss": 0.6864,
"step": 149
},
{
"epoch": 0.4225352112676056,
"grad_norm": 0.7916165651672555,
"learning_rate": 9.950372099971635e-06,
"loss": 0.6611,
"step": 150
},
{
"epoch": 0.4253521126760563,
"grad_norm": 0.8634003427333417,
"learning_rate": 9.948041037730546e-06,
"loss": 0.7326,
"step": 151
},
{
"epoch": 0.428169014084507,
"grad_norm": 0.9263824014581405,
"learning_rate": 9.945656764463706e-06,
"loss": 0.7068,
"step": 152
},
{
"epoch": 0.4309859154929577,
"grad_norm": 0.7528011216400118,
"learning_rate": 9.943219305811486e-06,
"loss": 0.662,
"step": 153
},
{
"epoch": 0.43380281690140843,
"grad_norm": 0.8372646453988154,
"learning_rate": 9.940728687986219e-06,
"loss": 0.7213,
"step": 154
},
{
"epoch": 0.43661971830985913,
"grad_norm": 0.9463740009923008,
"learning_rate": 9.938184937771903e-06,
"loss": 0.6919,
"step": 155
},
{
"epoch": 0.4394366197183099,
"grad_norm": 0.959386942798321,
"learning_rate": 9.935588082523917e-06,
"loss": 0.6449,
"step": 156
},
{
"epoch": 0.4422535211267606,
"grad_norm": 0.9250783308719386,
"learning_rate": 9.93293815016874e-06,
"loss": 0.6796,
"step": 157
},
{
"epoch": 0.4450704225352113,
"grad_norm": 0.7946799691080673,
"learning_rate": 9.930235169203628e-06,
"loss": 0.67,
"step": 158
},
{
"epoch": 0.447887323943662,
"grad_norm": 0.8522628286586079,
"learning_rate": 9.927479168696327e-06,
"loss": 0.7309,
"step": 159
},
{
"epoch": 0.4507042253521127,
"grad_norm": 0.837190553799222,
"learning_rate": 9.924670178284751e-06,
"loss": 0.6668,
"step": 160
},
{
"epoch": 0.4535211267605634,
"grad_norm": 1.0436089990244963,
"learning_rate": 9.921808228176667e-06,
"loss": 0.7829,
"step": 161
},
{
"epoch": 0.4563380281690141,
"grad_norm": 0.893929704291238,
"learning_rate": 9.918893349149361e-06,
"loss": 0.7153,
"step": 162
},
{
"epoch": 0.4591549295774648,
"grad_norm": 0.8869097673696764,
"learning_rate": 9.915925572549325e-06,
"loss": 0.6485,
"step": 163
},
{
"epoch": 0.4619718309859155,
"grad_norm": 0.8467721185605459,
"learning_rate": 9.912904930291902e-06,
"loss": 0.7482,
"step": 164
},
{
"epoch": 0.4647887323943662,
"grad_norm": 0.897220133636629,
"learning_rate": 9.909831454860953e-06,
"loss": 0.6422,
"step": 165
},
{
"epoch": 0.4676056338028169,
"grad_norm": 0.8919253252137377,
"learning_rate": 9.906705179308503e-06,
"loss": 0.7032,
"step": 166
},
{
"epoch": 0.4704225352112676,
"grad_norm": 0.8857603615046576,
"learning_rate": 9.903526137254386e-06,
"loss": 0.6246,
"step": 167
},
{
"epoch": 0.4732394366197183,
"grad_norm": 0.8283723940151976,
"learning_rate": 9.900294362885892e-06,
"loss": 0.643,
"step": 168
},
{
"epoch": 0.476056338028169,
"grad_norm": 0.8463508629352864,
"learning_rate": 9.897009890957382e-06,
"loss": 0.7326,
"step": 169
},
{
"epoch": 0.4788732394366197,
"grad_norm": 0.8166043234476047,
"learning_rate": 9.893672756789933e-06,
"loss": 0.6919,
"step": 170
},
{
"epoch": 0.48169014084507045,
"grad_norm": 0.8798248125614349,
"learning_rate": 9.890282996270944e-06,
"loss": 0.7502,
"step": 171
},
{
"epoch": 0.48450704225352115,
"grad_norm": 0.9137410379876206,
"learning_rate": 9.886840645853757e-06,
"loss": 0.717,
"step": 172
},
{
"epoch": 0.48732394366197185,
"grad_norm": 0.8682406299826412,
"learning_rate": 9.883345742557265e-06,
"loss": 0.7093,
"step": 173
},
{
"epoch": 0.49014084507042255,
"grad_norm": 0.768099268451737,
"learning_rate": 9.879798323965512e-06,
"loss": 0.6324,
"step": 174
},
{
"epoch": 0.49295774647887325,
"grad_norm": 0.7951997110898612,
"learning_rate": 9.876198428227288e-06,
"loss": 0.6043,
"step": 175
},
{
"epoch": 0.49577464788732395,
"grad_norm": 0.8150266343962548,
"learning_rate": 9.872546094055719e-06,
"loss": 0.7314,
"step": 176
},
{
"epoch": 0.49859154929577465,
"grad_norm": 0.7567196532773179,
"learning_rate": 9.868841360727856e-06,
"loss": 0.6341,
"step": 177
},
{
"epoch": 0.5014084507042254,
"grad_norm": 0.8057546679078017,
"learning_rate": 9.865084268084247e-06,
"loss": 0.5985,
"step": 178
},
{
"epoch": 0.504225352112676,
"grad_norm": 0.8248072860159059,
"learning_rate": 9.861274856528504e-06,
"loss": 0.6856,
"step": 179
},
{
"epoch": 0.5070422535211268,
"grad_norm": 0.7422143006416313,
"learning_rate": 9.857413167026881e-06,
"loss": 0.6509,
"step": 180
},
{
"epoch": 0.5098591549295775,
"grad_norm": 0.842171137809266,
"learning_rate": 9.853499241107827e-06,
"loss": 0.6759,
"step": 181
},
{
"epoch": 0.5126760563380282,
"grad_norm": 1.0101293356673013,
"learning_rate": 9.849533120861537e-06,
"loss": 0.838,
"step": 182
},
{
"epoch": 0.5154929577464789,
"grad_norm": 0.7530541447107382,
"learning_rate": 9.845514848939499e-06,
"loss": 0.7384,
"step": 183
},
{
"epoch": 0.5183098591549296,
"grad_norm": 0.914908795328647,
"learning_rate": 9.841444468554043e-06,
"loss": 0.7436,
"step": 184
},
{
"epoch": 0.5211267605633803,
"grad_norm": 0.8862498828793486,
"learning_rate": 9.837322023477866e-06,
"loss": 0.7144,
"step": 185
},
{
"epoch": 0.523943661971831,
"grad_norm": 0.8056528376306878,
"learning_rate": 9.833147558043574e-06,
"loss": 0.6498,
"step": 186
},
{
"epoch": 0.5267605633802817,
"grad_norm": 0.8011120670486118,
"learning_rate": 9.82892111714319e-06,
"loss": 0.7161,
"step": 187
},
{
"epoch": 0.5295774647887324,
"grad_norm": 0.7572342622375217,
"learning_rate": 9.824642746227682e-06,
"loss": 0.7064,
"step": 188
},
{
"epoch": 0.532394366197183,
"grad_norm": 0.8039481620735608,
"learning_rate": 9.820312491306471e-06,
"loss": 0.6512,
"step": 189
},
{
"epoch": 0.5352112676056338,
"grad_norm": 0.7738723306767284,
"learning_rate": 9.815930398946942e-06,
"loss": 0.6906,
"step": 190
},
{
"epoch": 0.5380281690140845,
"grad_norm": 0.7165440107878704,
"learning_rate": 9.811496516273925e-06,
"loss": 0.6153,
"step": 191
},
{
"epoch": 0.5408450704225352,
"grad_norm": 0.8319057805907693,
"learning_rate": 9.807010890969214e-06,
"loss": 0.6864,
"step": 192
},
{
"epoch": 0.543661971830986,
"grad_norm": 0.773189524368119,
"learning_rate": 9.80247357127103e-06,
"loss": 0.7204,
"step": 193
},
{
"epoch": 0.5464788732394367,
"grad_norm": 0.7653522714932102,
"learning_rate": 9.797884605973523e-06,
"loss": 0.6266,
"step": 194
},
{
"epoch": 0.5492957746478874,
"grad_norm": 0.76169482878666,
"learning_rate": 9.793244044426233e-06,
"loss": 0.6837,
"step": 195
},
{
"epoch": 0.5521126760563381,
"grad_norm": 0.853170087758016,
"learning_rate": 9.788551936533561e-06,
"loss": 0.7598,
"step": 196
},
{
"epoch": 0.5549295774647888,
"grad_norm": 0.8075207803578713,
"learning_rate": 9.783808332754242e-06,
"loss": 0.733,
"step": 197
},
{
"epoch": 0.5577464788732395,
"grad_norm": 0.8517323289471529,
"learning_rate": 9.779013284100791e-06,
"loss": 0.7236,
"step": 198
},
{
"epoch": 0.5605633802816902,
"grad_norm": 0.7997362496535447,
"learning_rate": 9.77416684213896e-06,
"loss": 0.712,
"step": 199
},
{
"epoch": 0.5633802816901409,
"grad_norm": 0.8430063817076954,
"learning_rate": 9.769269058987181e-06,
"loss": 0.6284,
"step": 200
},
{
"epoch": 0.5661971830985916,
"grad_norm": 0.8121616168773351,
"learning_rate": 9.76431998731601e-06,
"loss": 0.6711,
"step": 201
},
{
"epoch": 0.5690140845070423,
"grad_norm": 0.8827282760216014,
"learning_rate": 9.759319680347556e-06,
"loss": 0.7354,
"step": 202
},
{
"epoch": 0.571830985915493,
"grad_norm": 0.853130964046107,
"learning_rate": 9.75426819185491e-06,
"loss": 0.7771,
"step": 203
},
{
"epoch": 0.5746478873239437,
"grad_norm": 0.9947412584733369,
"learning_rate": 9.74916557616157e-06,
"loss": 0.7,
"step": 204
},
{
"epoch": 0.5774647887323944,
"grad_norm": 0.8769746348104829,
"learning_rate": 9.744011888140847e-06,
"loss": 0.6514,
"step": 205
},
{
"epoch": 0.5802816901408451,
"grad_norm": 0.9184017741530913,
"learning_rate": 9.738807183215288e-06,
"loss": 0.7239,
"step": 206
},
{
"epoch": 0.5830985915492958,
"grad_norm": 0.8842913206665063,
"learning_rate": 9.73355151735607e-06,
"loss": 0.684,
"step": 207
},
{
"epoch": 0.5859154929577465,
"grad_norm": 0.794506224518057,
"learning_rate": 9.728244947082405e-06,
"loss": 0.6926,
"step": 208
},
{
"epoch": 0.5887323943661972,
"grad_norm": 0.8074121706042865,
"learning_rate": 9.722887529460928e-06,
"loss": 0.7134,
"step": 209
},
{
"epoch": 0.5915492957746479,
"grad_norm": 0.8230238825095924,
"learning_rate": 9.717479322105079e-06,
"loss": 0.7206,
"step": 210
},
{
"epoch": 0.5943661971830986,
"grad_norm": 0.9594950928708149,
"learning_rate": 9.712020383174496e-06,
"loss": 0.6946,
"step": 211
},
{
"epoch": 0.5971830985915493,
"grad_norm": 0.7521256870035444,
"learning_rate": 9.70651077137438e-06,
"loss": 0.5711,
"step": 212
},
{
"epoch": 0.6,
"grad_norm": 0.7671581105894907,
"learning_rate": 9.700950545954867e-06,
"loss": 0.6589,
"step": 213
},
{
"epoch": 0.6028169014084507,
"grad_norm": 0.8178947584831048,
"learning_rate": 9.695339766710382e-06,
"loss": 0.723,
"step": 214
},
{
"epoch": 0.6056338028169014,
"grad_norm": 0.8003431837226573,
"learning_rate": 9.689678493979011e-06,
"loss": 0.6493,
"step": 215
},
{
"epoch": 0.6084507042253521,
"grad_norm": 0.7952819058062531,
"learning_rate": 9.683966788641848e-06,
"loss": 0.6962,
"step": 216
},
{
"epoch": 0.6112676056338028,
"grad_norm": 0.7826856935612455,
"learning_rate": 9.678204712122328e-06,
"loss": 0.7371,
"step": 217
},
{
"epoch": 0.6140845070422535,
"grad_norm": 0.7746418412525581,
"learning_rate": 9.672392326385582e-06,
"loss": 0.6902,
"step": 218
},
{
"epoch": 0.6169014084507042,
"grad_norm": 0.940402540251371,
"learning_rate": 9.666529693937763e-06,
"loss": 0.779,
"step": 219
},
{
"epoch": 0.6197183098591549,
"grad_norm": 0.8449206053067433,
"learning_rate": 9.660616877825374e-06,
"loss": 0.6905,
"step": 220
},
{
"epoch": 0.6225352112676056,
"grad_norm": 0.7994133282679672,
"learning_rate": 9.65465394163459e-06,
"loss": 0.7171,
"step": 221
},
{
"epoch": 0.6253521126760564,
"grad_norm": 0.8032156386437383,
"learning_rate": 9.64864094949058e-06,
"loss": 0.6782,
"step": 222
},
{
"epoch": 0.6281690140845071,
"grad_norm": 0.9003451206809043,
"learning_rate": 9.642577966056806e-06,
"loss": 0.7278,
"step": 223
},
{
"epoch": 0.6309859154929578,
"grad_norm": 0.7775350319628389,
"learning_rate": 9.636465056534337e-06,
"loss": 0.6913,
"step": 224
},
{
"epoch": 0.6338028169014085,
"grad_norm": 0.8310388732940904,
"learning_rate": 9.630302286661149e-06,
"loss": 0.7544,
"step": 225
},
{
"epoch": 0.6366197183098592,
"grad_norm": 0.8853987442843421,
"learning_rate": 9.624089722711409e-06,
"loss": 0.6923,
"step": 226
},
{
"epoch": 0.6394366197183099,
"grad_norm": 0.8804755951508887,
"learning_rate": 9.617827431494769e-06,
"loss": 0.6605,
"step": 227
},
{
"epoch": 0.6422535211267606,
"grad_norm": 0.8366059157406054,
"learning_rate": 9.611515480355644e-06,
"loss": 0.7357,
"step": 228
},
{
"epoch": 0.6450704225352113,
"grad_norm": 0.7210382994869629,
"learning_rate": 9.605153937172495e-06,
"loss": 0.6835,
"step": 229
},
{
"epoch": 0.647887323943662,
"grad_norm": 0.8739466441160235,
"learning_rate": 9.598742870357089e-06,
"loss": 0.6739,
"step": 230
},
{
"epoch": 0.6507042253521127,
"grad_norm": 0.8654712248667421,
"learning_rate": 9.592282348853772e-06,
"loss": 0.7532,
"step": 231
},
{
"epoch": 0.6535211267605634,
"grad_norm": 0.9065008902303662,
"learning_rate": 9.585772442138717e-06,
"loss": 0.734,
"step": 232
},
{
"epoch": 0.6563380281690141,
"grad_norm": 0.7871365715275687,
"learning_rate": 9.57921322021919e-06,
"loss": 0.6635,
"step": 233
},
{
"epoch": 0.6591549295774648,
"grad_norm": 0.7642041831022736,
"learning_rate": 9.572604753632789e-06,
"loss": 0.6484,
"step": 234
},
{
"epoch": 0.6619718309859155,
"grad_norm": 0.8271285548719295,
"learning_rate": 9.565947113446685e-06,
"loss": 0.6601,
"step": 235
},
{
"epoch": 0.6647887323943662,
"grad_norm": 0.7538375562561086,
"learning_rate": 9.559240371256861e-06,
"loss": 0.6902,
"step": 236
},
{
"epoch": 0.6676056338028169,
"grad_norm": 0.8043626834134439,
"learning_rate": 9.552484599187344e-06,
"loss": 0.6368,
"step": 237
},
{
"epoch": 0.6704225352112676,
"grad_norm": 0.7529150606741751,
"learning_rate": 9.545679869889422e-06,
"loss": 0.7056,
"step": 238
},
{
"epoch": 0.6732394366197183,
"grad_norm": 0.7284992451418077,
"learning_rate": 9.538826256540866e-06,
"loss": 0.701,
"step": 239
},
{
"epoch": 0.676056338028169,
"grad_norm": 0.7430111850931722,
"learning_rate": 9.53192383284515e-06,
"loss": 0.6279,
"step": 240
},
{
"epoch": 0.6788732394366197,
"grad_norm": 0.8112227328214626,
"learning_rate": 9.524972673030646e-06,
"loss": 0.6795,
"step": 241
},
{
"epoch": 0.6816901408450704,
"grad_norm": 0.7709521327734363,
"learning_rate": 9.517972851849837e-06,
"loss": 0.6938,
"step": 242
},
{
"epoch": 0.6845070422535211,
"grad_norm": 0.839896263111437,
"learning_rate": 9.510924444578505e-06,
"loss": 0.7249,
"step": 243
},
{
"epoch": 0.6873239436619718,
"grad_norm": 0.7232560690742893,
"learning_rate": 9.503827527014927e-06,
"loss": 0.6598,
"step": 244
},
{
"epoch": 0.6901408450704225,
"grad_norm": 0.8123508737767481,
"learning_rate": 9.496682175479058e-06,
"loss": 0.7528,
"step": 245
},
{
"epoch": 0.6929577464788732,
"grad_norm": 0.8462883686229686,
"learning_rate": 9.489488466811706e-06,
"loss": 0.6947,
"step": 246
},
{
"epoch": 0.6957746478873239,
"grad_norm": 0.9431639768896318,
"learning_rate": 9.482246478373713e-06,
"loss": 0.8318,
"step": 247
},
{
"epoch": 0.6985915492957746,
"grad_norm": 0.7818012727406521,
"learning_rate": 9.47495628804512e-06,
"loss": 0.7279,
"step": 248
},
{
"epoch": 0.7014084507042253,
"grad_norm": 0.8562275464337439,
"learning_rate": 9.467617974224326e-06,
"loss": 0.7178,
"step": 249
},
{
"epoch": 0.704225352112676,
"grad_norm": 0.7745218443030316,
"learning_rate": 9.460231615827254e-06,
"loss": 0.6763,
"step": 250
},
{
"epoch": 0.7070422535211267,
"grad_norm": 0.8938034789544281,
"learning_rate": 9.452797292286486e-06,
"loss": 0.7282,
"step": 251
},
{
"epoch": 0.7098591549295775,
"grad_norm": 0.8421964462964258,
"learning_rate": 9.445315083550432e-06,
"loss": 0.6092,
"step": 252
},
{
"epoch": 0.7126760563380282,
"grad_norm": 0.8842902719459901,
"learning_rate": 9.437785070082448e-06,
"loss": 0.7439,
"step": 253
},
{
"epoch": 0.7154929577464789,
"grad_norm": 0.8295883668473504,
"learning_rate": 9.430207332859981e-06,
"loss": 0.6941,
"step": 254
},
{
"epoch": 0.7183098591549296,
"grad_norm": 0.7997799100355183,
"learning_rate": 9.422581953373704e-06,
"loss": 0.6726,
"step": 255
},
{
"epoch": 0.7211267605633803,
"grad_norm": 0.8249888871807052,
"learning_rate": 9.414909013626624e-06,
"loss": 0.7556,
"step": 256
},
{
"epoch": 0.723943661971831,
"grad_norm": 0.7139160556941465,
"learning_rate": 9.407188596133212e-06,
"loss": 0.7071,
"step": 257
},
{
"epoch": 0.7267605633802817,
"grad_norm": 0.8024124317877482,
"learning_rate": 9.399420783918518e-06,
"loss": 0.6568,
"step": 258
},
{
"epoch": 0.7295774647887324,
"grad_norm": 0.8584267549374202,
"learning_rate": 9.391605660517268e-06,
"loss": 0.6367,
"step": 259
},
{
"epoch": 0.7323943661971831,
"grad_norm": 0.7902851358778393,
"learning_rate": 9.38374330997297e-06,
"loss": 0.6593,
"step": 260
},
{
"epoch": 0.7352112676056338,
"grad_norm": 0.8023474180886204,
"learning_rate": 9.375833816837012e-06,
"loss": 0.6464,
"step": 261
},
{
"epoch": 0.7380281690140845,
"grad_norm": 0.8154573367134973,
"learning_rate": 9.367877266167752e-06,
"loss": 0.8018,
"step": 262
},
{
"epoch": 0.7408450704225352,
"grad_norm": 0.8115241369192978,
"learning_rate": 9.3598737435296e-06,
"loss": 0.7639,
"step": 263
},
{
"epoch": 0.7436619718309859,
"grad_norm": 0.7849626692594518,
"learning_rate": 9.351823334992105e-06,
"loss": 0.6817,
"step": 264
},
{
"epoch": 0.7464788732394366,
"grad_norm": 0.9197729948526122,
"learning_rate": 9.343726127129023e-06,
"loss": 0.7335,
"step": 265
},
{
"epoch": 0.7492957746478873,
"grad_norm": 0.7275321442153452,
"learning_rate": 9.335582207017385e-06,
"loss": 0.7261,
"step": 266
},
{
"epoch": 0.752112676056338,
"grad_norm": 0.8163208868320786,
"learning_rate": 9.327391662236569e-06,
"loss": 0.741,
"step": 267
},
{
"epoch": 0.7549295774647887,
"grad_norm": 0.7840261409488222,
"learning_rate": 9.31915458086735e-06,
"loss": 0.6902,
"step": 268
},
{
"epoch": 0.7577464788732394,
"grad_norm": 0.8479200482066225,
"learning_rate": 9.310871051490953e-06,
"loss": 0.6263,
"step": 269
},
{
"epoch": 0.7605633802816901,
"grad_norm": 0.7707125463434206,
"learning_rate": 9.302541163188107e-06,
"loss": 0.7071,
"step": 270
},
{
"epoch": 0.7633802816901408,
"grad_norm": 0.8206470495747352,
"learning_rate": 9.294165005538083e-06,
"loss": 0.7162,
"step": 271
},
{
"epoch": 0.7661971830985915,
"grad_norm": 0.7938247034692177,
"learning_rate": 9.285742668617725e-06,
"loss": 0.7303,
"step": 272
},
{
"epoch": 0.7690140845070422,
"grad_norm": 0.7724192077582562,
"learning_rate": 9.277274243000495e-06,
"loss": 0.6745,
"step": 273
},
{
"epoch": 0.7718309859154929,
"grad_norm": 0.7556279269726222,
"learning_rate": 9.268759819755484e-06,
"loss": 0.6684,
"step": 274
},
{
"epoch": 0.7746478873239436,
"grad_norm": 0.9010158273015102,
"learning_rate": 9.260199490446444e-06,
"loss": 0.6968,
"step": 275
},
{
"epoch": 0.7774647887323943,
"grad_norm": 0.9190880584517591,
"learning_rate": 9.251593347130796e-06,
"loss": 0.6965,
"step": 276
},
{
"epoch": 0.780281690140845,
"grad_norm": 0.7392835129813884,
"learning_rate": 9.242941482358646e-06,
"loss": 0.6955,
"step": 277
},
{
"epoch": 0.7830985915492957,
"grad_norm": 0.7721524689242966,
"learning_rate": 9.234243989171784e-06,
"loss": 0.7349,
"step": 278
},
{
"epoch": 0.7859154929577464,
"grad_norm": 0.700499216910955,
"learning_rate": 9.225500961102685e-06,
"loss": 0.6621,
"step": 279
},
{
"epoch": 0.7887323943661971,
"grad_norm": 0.7596012409463865,
"learning_rate": 9.216712492173509e-06,
"loss": 0.73,
"step": 280
},
{
"epoch": 0.7915492957746478,
"grad_norm": 0.7112477735132746,
"learning_rate": 9.20787867689508e-06,
"loss": 0.6936,
"step": 281
},
{
"epoch": 0.7943661971830986,
"grad_norm": 0.7538333397867995,
"learning_rate": 9.19899961026588e-06,
"loss": 0.7108,
"step": 282
},
{
"epoch": 0.7971830985915493,
"grad_norm": 0.7500840684245914,
"learning_rate": 9.190075387771014e-06,
"loss": 0.6424,
"step": 283
},
{
"epoch": 0.8,
"grad_norm": 0.7767104224997774,
"learning_rate": 9.181106105381201e-06,
"loss": 0.7673,
"step": 284
},
{
"epoch": 0.8028169014084507,
"grad_norm": 0.7051945870953593,
"learning_rate": 9.172091859551727e-06,
"loss": 0.6375,
"step": 285
},
{
"epoch": 0.8056338028169014,
"grad_norm": 0.8187705644445149,
"learning_rate": 9.16303274722141e-06,
"loss": 0.7782,
"step": 286
},
{
"epoch": 0.8084507042253521,
"grad_norm": 0.7805258045768315,
"learning_rate": 9.153928865811567e-06,
"loss": 0.7418,
"step": 287
},
{
"epoch": 0.8112676056338028,
"grad_norm": 0.8128901606684851,
"learning_rate": 9.144780313224955e-06,
"loss": 0.7399,
"step": 288
},
{
"epoch": 0.8140845070422535,
"grad_norm": 0.8154420318456077,
"learning_rate": 9.135587187844727e-06,
"loss": 0.723,
"step": 289
},
{
"epoch": 0.8169014084507042,
"grad_norm": 0.7858069446808291,
"learning_rate": 9.126349588533367e-06,
"loss": 0.68,
"step": 290
},
{
"epoch": 0.819718309859155,
"grad_norm": 0.8223570062612096,
"learning_rate": 9.117067614631629e-06,
"loss": 0.7178,
"step": 291
},
{
"epoch": 0.8225352112676056,
"grad_norm": 0.7361952237465246,
"learning_rate": 9.107741365957473e-06,
"loss": 0.7253,
"step": 292
},
{
"epoch": 0.8253521126760563,
"grad_norm": 0.7410997914967067,
"learning_rate": 9.098370942804984e-06,
"loss": 0.7365,
"step": 293
},
{
"epoch": 0.828169014084507,
"grad_norm": 0.7841577337029458,
"learning_rate": 9.0889564459433e-06,
"loss": 0.6961,
"step": 294
},
{
"epoch": 0.8309859154929577,
"grad_norm": 0.7103875687394275,
"learning_rate": 9.079497976615527e-06,
"loss": 0.7009,
"step": 295
},
{
"epoch": 0.8338028169014085,
"grad_norm": 0.8613504465876265,
"learning_rate": 9.069995636537646e-06,
"loss": 0.8999,
"step": 296
},
{
"epoch": 0.8366197183098592,
"grad_norm": 0.7512880794878137,
"learning_rate": 9.060449527897424e-06,
"loss": 0.7458,
"step": 297
},
{
"epoch": 0.8394366197183099,
"grad_norm": 0.7108104992364117,
"learning_rate": 9.050859753353312e-06,
"loss": 0.6101,
"step": 298
},
{
"epoch": 0.8422535211267606,
"grad_norm": 0.7613675413497479,
"learning_rate": 9.041226416033344e-06,
"loss": 0.6791,
"step": 299
},
{
"epoch": 0.8450704225352113,
"grad_norm": 0.7227340858608536,
"learning_rate": 9.031549619534025e-06,
"loss": 0.62,
"step": 300
},
{
"epoch": 0.847887323943662,
"grad_norm": 0.739015826440164,
"learning_rate": 9.021829467919218e-06,
"loss": 0.6194,
"step": 301
},
{
"epoch": 0.8507042253521127,
"grad_norm": 0.7621930325437081,
"learning_rate": 9.012066065719026e-06,
"loss": 0.7342,
"step": 302
},
{
"epoch": 0.8535211267605634,
"grad_norm": 0.7518088571559766,
"learning_rate": 9.00225951792867e-06,
"loss": 0.734,
"step": 303
},
{
"epoch": 0.856338028169014,
"grad_norm": 0.7619799485709029,
"learning_rate": 8.99240993000735e-06,
"loss": 0.6668,
"step": 304
},
{
"epoch": 0.8591549295774648,
"grad_norm": 0.7512429320136312,
"learning_rate": 8.98251740787712e-06,
"loss": 0.6285,
"step": 305
},
{
"epoch": 0.8619718309859155,
"grad_norm": 0.7687230386597979,
"learning_rate": 8.972582057921753e-06,
"loss": 0.6658,
"step": 306
},
{
"epoch": 0.8647887323943662,
"grad_norm": 0.7939832464335207,
"learning_rate": 8.962603986985582e-06,
"loss": 0.7445,
"step": 307
},
{
"epoch": 0.8676056338028169,
"grad_norm": 0.7384840890732082,
"learning_rate": 8.952583302372361e-06,
"loss": 0.6054,
"step": 308
},
{
"epoch": 0.8704225352112676,
"grad_norm": 0.7383794192616162,
"learning_rate": 8.942520111844117e-06,
"loss": 0.6556,
"step": 309
},
{
"epoch": 0.8732394366197183,
"grad_norm": 0.7927695830739503,
"learning_rate": 8.932414523619979e-06,
"loss": 0.7918,
"step": 310
},
{
"epoch": 0.8760563380281691,
"grad_norm": 0.8017064580913946,
"learning_rate": 8.922266646375012e-06,
"loss": 0.7132,
"step": 311
},
{
"epoch": 0.8788732394366198,
"grad_norm": 0.7792989124184551,
"learning_rate": 8.912076589239071e-06,
"loss": 0.7115,
"step": 312
},
{
"epoch": 0.8816901408450705,
"grad_norm": 0.7369492414275604,
"learning_rate": 8.901844461795597e-06,
"loss": 0.6636,
"step": 313
},
{
"epoch": 0.8845070422535212,
"grad_norm": 0.7168268633853285,
"learning_rate": 8.89157037408046e-06,
"loss": 0.6391,
"step": 314
},
{
"epoch": 0.8873239436619719,
"grad_norm": 0.7684412827406625,
"learning_rate": 8.881254436580771e-06,
"loss": 0.721,
"step": 315
},
{
"epoch": 0.8901408450704226,
"grad_norm": 0.7399940724937712,
"learning_rate": 8.870896760233687e-06,
"loss": 0.7014,
"step": 316
},
{
"epoch": 0.8929577464788733,
"grad_norm": 0.7427553650129637,
"learning_rate": 8.860497456425226e-06,
"loss": 0.7486,
"step": 317
},
{
"epoch": 0.895774647887324,
"grad_norm": 0.7355572590337344,
"learning_rate": 8.850056636989064e-06,
"loss": 0.6474,
"step": 318
},
{
"epoch": 0.8985915492957747,
"grad_norm": 0.7670739435633628,
"learning_rate": 8.839574414205335e-06,
"loss": 0.7253,
"step": 319
},
{
"epoch": 0.9014084507042254,
"grad_norm": 0.7223450642414903,
"learning_rate": 8.829050900799426e-06,
"loss": 0.6343,
"step": 320
},
{
"epoch": 0.9042253521126761,
"grad_norm": 0.8339291046044646,
"learning_rate": 8.818486209940754e-06,
"loss": 0.7616,
"step": 321
},
{
"epoch": 0.9070422535211268,
"grad_norm": 0.769620973035585,
"learning_rate": 8.807880455241562e-06,
"loss": 0.7673,
"step": 322
},
{
"epoch": 0.9098591549295775,
"grad_norm": 0.851306065801614,
"learning_rate": 8.797233750755696e-06,
"loss": 0.7491,
"step": 323
},
{
"epoch": 0.9126760563380282,
"grad_norm": 0.7651622574675689,
"learning_rate": 8.786546210977364e-06,
"loss": 0.7015,
"step": 324
},
{
"epoch": 0.9154929577464789,
"grad_norm": 0.71441222991302,
"learning_rate": 8.775817950839925e-06,
"loss": 0.6956,
"step": 325
},
{
"epoch": 0.9183098591549296,
"grad_norm": 0.7506938193868895,
"learning_rate": 8.765049085714634e-06,
"loss": 0.6509,
"step": 326
},
{
"epoch": 0.9211267605633803,
"grad_norm": 0.8071556078391223,
"learning_rate": 8.754239731409418e-06,
"loss": 0.7149,
"step": 327
},
{
"epoch": 0.923943661971831,
"grad_norm": 0.7818360292624567,
"learning_rate": 8.743390004167618e-06,
"loss": 0.7043,
"step": 328
},
{
"epoch": 0.9267605633802817,
"grad_norm": 0.81081711408685,
"learning_rate": 8.732500020666748e-06,
"loss": 0.6703,
"step": 329
},
{
"epoch": 0.9295774647887324,
"grad_norm": 0.7372182871990494,
"learning_rate": 8.72156989801723e-06,
"loss": 0.7131,
"step": 330
},
{
"epoch": 0.9323943661971831,
"grad_norm": 0.7249976512649938,
"learning_rate": 8.710599753761148e-06,
"loss": 0.5962,
"step": 331
},
{
"epoch": 0.9352112676056338,
"grad_norm": 0.7427619837327385,
"learning_rate": 8.699589705870972e-06,
"loss": 0.6572,
"step": 332
},
{
"epoch": 0.9380281690140845,
"grad_norm": 0.7244013374655381,
"learning_rate": 8.688539872748293e-06,
"loss": 0.7367,
"step": 333
},
{
"epoch": 0.9408450704225352,
"grad_norm": 0.7966894332891572,
"learning_rate": 8.677450373222555e-06,
"loss": 0.72,
"step": 334
},
{
"epoch": 0.9436619718309859,
"grad_norm": 0.6590452238575061,
"learning_rate": 8.666321326549771e-06,
"loss": 0.6308,
"step": 335
},
{
"epoch": 0.9464788732394366,
"grad_norm": 0.7153952945703177,
"learning_rate": 8.655152852411242e-06,
"loss": 0.6736,
"step": 336
},
{
"epoch": 0.9492957746478873,
"grad_norm": 0.6900657816220632,
"learning_rate": 8.643945070912269e-06,
"loss": 0.6529,
"step": 337
},
{
"epoch": 0.952112676056338,
"grad_norm": 0.7842151250405361,
"learning_rate": 8.632698102580866e-06,
"loss": 0.6962,
"step": 338
},
{
"epoch": 0.9549295774647887,
"grad_norm": 0.6970865060911673,
"learning_rate": 8.621412068366455e-06,
"loss": 0.6492,
"step": 339
},
{
"epoch": 0.9577464788732394,
"grad_norm": 0.828143820100817,
"learning_rate": 8.61008708963858e-06,
"loss": 0.6946,
"step": 340
},
{
"epoch": 0.9605633802816902,
"grad_norm": 0.7922646487759859,
"learning_rate": 8.598723288185582e-06,
"loss": 0.7274,
"step": 341
},
{
"epoch": 0.9633802816901409,
"grad_norm": 0.7009049498810189,
"learning_rate": 8.587320786213308e-06,
"loss": 0.653,
"step": 342
},
{
"epoch": 0.9661971830985916,
"grad_norm": 0.7345915900001514,
"learning_rate": 8.575879706343786e-06,
"loss": 0.69,
"step": 343
},
{
"epoch": 0.9690140845070423,
"grad_norm": 0.7390349577087024,
"learning_rate": 8.564400171613907e-06,
"loss": 0.6721,
"step": 344
},
{
"epoch": 0.971830985915493,
"grad_norm": 0.7240412147255116,
"learning_rate": 8.552882305474106e-06,
"loss": 0.7208,
"step": 345
},
{
"epoch": 0.9746478873239437,
"grad_norm": 0.7431098381900512,
"learning_rate": 8.541326231787036e-06,
"loss": 0.7387,
"step": 346
},
{
"epoch": 0.9774647887323944,
"grad_norm": 0.7044683661504436,
"learning_rate": 8.529732074826225e-06,
"loss": 0.6333,
"step": 347
},
{
"epoch": 0.9802816901408451,
"grad_norm": 0.6550673857067976,
"learning_rate": 8.518099959274751e-06,
"loss": 0.6141,
"step": 348
},
{
"epoch": 0.9830985915492958,
"grad_norm": 0.6811939354097548,
"learning_rate": 8.506430010223899e-06,
"loss": 0.695,
"step": 349
},
{
"epoch": 0.9859154929577465,
"grad_norm": 0.7188183598452683,
"learning_rate": 8.494722353171807e-06,
"loss": 0.6578,
"step": 350
},
{
"epoch": 0.9887323943661972,
"grad_norm": 0.753120522013002,
"learning_rate": 8.482977114022133e-06,
"loss": 0.719,
"step": 351
},
{
"epoch": 0.9915492957746479,
"grad_norm": 0.6573052433756366,
"learning_rate": 8.471194419082683e-06,
"loss": 0.6247,
"step": 352
},
{
"epoch": 0.9943661971830986,
"grad_norm": 0.6972129024473817,
"learning_rate": 8.459374395064066e-06,
"loss": 0.7089,
"step": 353
},
{
"epoch": 0.9971830985915493,
"grad_norm": 0.6991471510748611,
"learning_rate": 8.447517169078322e-06,
"loss": 0.7019,
"step": 354
},
{
"epoch": 1.0,
"grad_norm": 0.7539991152104814,
"learning_rate": 8.435622868637562e-06,
"loss": 0.6866,
"step": 355
},
{
"epoch": 1.0028169014084507,
"grad_norm": 0.7309316536347265,
"learning_rate": 8.423691621652597e-06,
"loss": 0.6191,
"step": 356
},
{
"epoch": 1.0056338028169014,
"grad_norm": 0.7327659006791467,
"learning_rate": 8.411723556431555e-06,
"loss": 0.5881,
"step": 357
},
{
"epoch": 1.008450704225352,
"grad_norm": 0.71963119446251,
"learning_rate": 8.399718801678507e-06,
"loss": 0.5709,
"step": 358
},
{
"epoch": 1.0112676056338028,
"grad_norm": 0.6650645225444921,
"learning_rate": 8.38767748649208e-06,
"loss": 0.5872,
"step": 359
},
{
"epoch": 1.0140845070422535,
"grad_norm": 0.6962048521955931,
"learning_rate": 8.375599740364076e-06,
"loss": 0.6494,
"step": 360
},
{
"epoch": 1.0169014084507042,
"grad_norm": 0.732308607955144,
"learning_rate": 8.363485693178068e-06,
"loss": 0.6065,
"step": 361
},
{
"epoch": 1.019718309859155,
"grad_norm": 0.7479553465916133,
"learning_rate": 8.351335475208013e-06,
"loss": 0.5926,
"step": 362
},
{
"epoch": 1.0225352112676056,
"grad_norm": 0.8104642606206092,
"learning_rate": 8.339149217116844e-06,
"loss": 0.6862,
"step": 363
},
{
"epoch": 1.0253521126760563,
"grad_norm": 0.8307345169299164,
"learning_rate": 8.32692704995507e-06,
"loss": 0.5706,
"step": 364
},
{
"epoch": 1.028169014084507,
"grad_norm": 0.7979429932796829,
"learning_rate": 8.314669105159363e-06,
"loss": 0.7012,
"step": 365
},
{
"epoch": 1.0309859154929577,
"grad_norm": 0.7731130165738759,
"learning_rate": 8.302375514551147e-06,
"loss": 0.6017,
"step": 366
},
{
"epoch": 1.0338028169014084,
"grad_norm": 0.808354828090498,
"learning_rate": 8.29004641033518e-06,
"loss": 0.6374,
"step": 367
},
{
"epoch": 1.036619718309859,
"grad_norm": 0.9435485675478585,
"learning_rate": 8.277681925098133e-06,
"loss": 0.6787,
"step": 368
},
{
"epoch": 1.0394366197183098,
"grad_norm": 0.8129322510408442,
"learning_rate": 8.26528219180716e-06,
"loss": 0.5717,
"step": 369
},
{
"epoch": 1.0422535211267605,
"grad_norm": 0.7952569015637464,
"learning_rate": 8.252847343808473e-06,
"loss": 0.6102,
"step": 370
},
{
"epoch": 1.0450704225352112,
"grad_norm": 0.7372836570712195,
"learning_rate": 8.240377514825906e-06,
"loss": 0.6064,
"step": 371
},
{
"epoch": 1.047887323943662,
"grad_norm": 0.700959266674204,
"learning_rate": 8.227872838959478e-06,
"loss": 0.5533,
"step": 372
},
{
"epoch": 1.0507042253521126,
"grad_norm": 0.7392699999637229,
"learning_rate": 8.215333450683945e-06,
"loss": 0.6194,
"step": 373
},
{
"epoch": 1.0535211267605633,
"grad_norm": 0.8016719369071104,
"learning_rate": 8.202759484847366e-06,
"loss": 0.6321,
"step": 374
},
{
"epoch": 1.056338028169014,
"grad_norm": 0.879203616291769,
"learning_rate": 8.19015107666964e-06,
"loss": 0.6167,
"step": 375
},
{
"epoch": 1.0591549295774647,
"grad_norm": 0.7999750323120932,
"learning_rate": 8.177508361741063e-06,
"loss": 0.61,
"step": 376
},
{
"epoch": 1.0619718309859154,
"grad_norm": 1.2853051561846651,
"learning_rate": 8.164831476020856e-06,
"loss": 0.7454,
"step": 377
},
{
"epoch": 1.064788732394366,
"grad_norm": 0.8055398755078872,
"learning_rate": 8.15212055583572e-06,
"loss": 0.6196,
"step": 378
},
{
"epoch": 1.0676056338028168,
"grad_norm": 0.7446253097717447,
"learning_rate": 8.139375737878356e-06,
"loss": 0.6322,
"step": 379
},
{
"epoch": 1.0704225352112675,
"grad_norm": 0.8411611220367352,
"learning_rate": 8.126597159206002e-06,
"loss": 0.6148,
"step": 380
},
{
"epoch": 1.0732394366197182,
"grad_norm": 0.7097418449161869,
"learning_rate": 8.113784957238957e-06,
"loss": 0.5783,
"step": 381
},
{
"epoch": 1.076056338028169,
"grad_norm": 0.9215534743065565,
"learning_rate": 8.100939269759103e-06,
"loss": 0.636,
"step": 382
},
{
"epoch": 1.0788732394366196,
"grad_norm": 0.9640442672292225,
"learning_rate": 8.088060234908425e-06,
"loss": 0.6169,
"step": 383
},
{
"epoch": 1.0816901408450703,
"grad_norm": 0.7562932638842584,
"learning_rate": 8.075147991187521e-06,
"loss": 0.626,
"step": 384
},
{
"epoch": 1.084507042253521,
"grad_norm": 0.7398497359039935,
"learning_rate": 8.062202677454123e-06,
"loss": 0.6072,
"step": 385
},
{
"epoch": 1.0873239436619717,
"grad_norm": 0.7834989845435792,
"learning_rate": 8.04922443292159e-06,
"loss": 0.6587,
"step": 386
},
{
"epoch": 1.0901408450704226,
"grad_norm": 0.8109941310202463,
"learning_rate": 8.036213397157418e-06,
"loss": 0.5767,
"step": 387
},
{
"epoch": 1.0929577464788733,
"grad_norm": 0.867677796061852,
"learning_rate": 8.02316971008174e-06,
"loss": 0.6495,
"step": 388
},
{
"epoch": 1.095774647887324,
"grad_norm": 0.8568914788227452,
"learning_rate": 8.01009351196582e-06,
"loss": 0.6608,
"step": 389
},
{
"epoch": 1.0985915492957747,
"grad_norm": 0.7414767319860525,
"learning_rate": 7.996984943430544e-06,
"loss": 0.6201,
"step": 390
},
{
"epoch": 1.1014084507042254,
"grad_norm": 0.7712950858852734,
"learning_rate": 7.983844145444908e-06,
"loss": 0.6559,
"step": 391
},
{
"epoch": 1.1042253521126761,
"grad_norm": 0.7279417190285391,
"learning_rate": 7.970671259324502e-06,
"loss": 0.6579,
"step": 392
},
{
"epoch": 1.1070422535211268,
"grad_norm": 0.6809679469757443,
"learning_rate": 7.957466426729995e-06,
"loss": 0.5586,
"step": 393
},
{
"epoch": 1.1098591549295775,
"grad_norm": 0.7395336563334385,
"learning_rate": 7.944229789665595e-06,
"loss": 0.6107,
"step": 394
},
{
"epoch": 1.1126760563380282,
"grad_norm": 0.7800851945961862,
"learning_rate": 7.930961490477546e-06,
"loss": 0.6775,
"step": 395
},
{
"epoch": 1.115492957746479,
"grad_norm": 0.7428766032430452,
"learning_rate": 7.917661671852582e-06,
"loss": 0.5768,
"step": 396
},
{
"epoch": 1.1183098591549296,
"grad_norm": 0.748113720663839,
"learning_rate": 7.904330476816391e-06,
"loss": 0.6391,
"step": 397
},
{
"epoch": 1.1211267605633803,
"grad_norm": 0.7602498335490296,
"learning_rate": 7.890968048732091e-06,
"loss": 0.6414,
"step": 398
},
{
"epoch": 1.123943661971831,
"grad_norm": 0.7233012193214933,
"learning_rate": 7.877574531298666e-06,
"loss": 0.6122,
"step": 399
},
{
"epoch": 1.1267605633802817,
"grad_norm": 0.7021500909845154,
"learning_rate": 7.864150068549446e-06,
"loss": 0.5922,
"step": 400
},
{
"epoch": 1.1295774647887324,
"grad_norm": 0.7106078546209356,
"learning_rate": 7.850694804850538e-06,
"loss": 0.6049,
"step": 401
},
{
"epoch": 1.1323943661971831,
"grad_norm": 0.7065107396130941,
"learning_rate": 7.837208884899283e-06,
"loss": 0.5874,
"step": 402
},
{
"epoch": 1.1352112676056338,
"grad_norm": 0.7102875370949132,
"learning_rate": 7.823692453722701e-06,
"loss": 0.6108,
"step": 403
},
{
"epoch": 1.1380281690140845,
"grad_norm": 0.7912610129085115,
"learning_rate": 7.810145656675923e-06,
"loss": 0.604,
"step": 404
},
{
"epoch": 1.1408450704225352,
"grad_norm": 0.737281408366629,
"learning_rate": 7.796568639440635e-06,
"loss": 0.5719,
"step": 405
},
{
"epoch": 1.143661971830986,
"grad_norm": 0.7670922935606069,
"learning_rate": 7.782961548023515e-06,
"loss": 0.5963,
"step": 406
},
{
"epoch": 1.1464788732394366,
"grad_norm": 0.7641025205212992,
"learning_rate": 7.769324528754653e-06,
"loss": 0.6672,
"step": 407
},
{
"epoch": 1.1492957746478873,
"grad_norm": 0.7823576553904255,
"learning_rate": 7.755657728285979e-06,
"loss": 0.6518,
"step": 408
},
{
"epoch": 1.152112676056338,
"grad_norm": 0.768517790974834,
"learning_rate": 7.741961293589693e-06,
"loss": 0.5925,
"step": 409
},
{
"epoch": 1.1549295774647887,
"grad_norm": 0.8178224875850133,
"learning_rate": 7.728235371956678e-06,
"loss": 0.6393,
"step": 410
},
{
"epoch": 1.1577464788732394,
"grad_norm": 0.7593574451798548,
"learning_rate": 7.714480110994922e-06,
"loss": 0.5932,
"step": 411
},
{
"epoch": 1.1605633802816901,
"grad_norm": 0.816042257929934,
"learning_rate": 7.700695658627924e-06,
"loss": 0.6303,
"step": 412
},
{
"epoch": 1.1633802816901408,
"grad_norm": 0.7790772874456501,
"learning_rate": 7.686882163093106e-06,
"loss": 0.6387,
"step": 413
},
{
"epoch": 1.1661971830985915,
"grad_norm": 0.759549985343405,
"learning_rate": 7.673039772940218e-06,
"loss": 0.6733,
"step": 414
},
{
"epoch": 1.1690140845070423,
"grad_norm": 0.8435272915696094,
"learning_rate": 7.659168637029746e-06,
"loss": 0.7193,
"step": 415
},
{
"epoch": 1.171830985915493,
"grad_norm": 0.6766200231780173,
"learning_rate": 7.6452689045313e-06,
"loss": 0.5821,
"step": 416
},
{
"epoch": 1.1746478873239437,
"grad_norm": 0.7660868612697938,
"learning_rate": 7.631340724922023e-06,
"loss": 0.6028,
"step": 417
},
{
"epoch": 1.1774647887323944,
"grad_norm": 0.7708935275497972,
"learning_rate": 7.617384247984973e-06,
"loss": 0.6066,
"step": 418
},
{
"epoch": 1.180281690140845,
"grad_norm": 0.7188801905334284,
"learning_rate": 7.603399623807519e-06,
"loss": 0.6452,
"step": 419
},
{
"epoch": 1.1830985915492958,
"grad_norm": 0.6894638678128713,
"learning_rate": 7.589387002779722e-06,
"loss": 0.5232,
"step": 420
},
{
"epoch": 1.1859154929577465,
"grad_norm": 0.7428721573686932,
"learning_rate": 7.575346535592721e-06,
"loss": 0.5196,
"step": 421
},
{
"epoch": 1.1887323943661972,
"grad_norm": 0.7167711000069285,
"learning_rate": 7.561278373237108e-06,
"loss": 0.5482,
"step": 422
},
{
"epoch": 1.1915492957746479,
"grad_norm": 0.755690004522166,
"learning_rate": 7.547182667001316e-06,
"loss": 0.6087,
"step": 423
},
{
"epoch": 1.1943661971830986,
"grad_norm": 0.7231715291470082,
"learning_rate": 7.5330595684699735e-06,
"loss": 0.6187,
"step": 424
},
{
"epoch": 1.1971830985915493,
"grad_norm": 0.7805865030943964,
"learning_rate": 7.5189092295222945e-06,
"loss": 0.5708,
"step": 425
},
{
"epoch": 1.2,
"grad_norm": 0.7398169564046222,
"learning_rate": 7.504731802330427e-06,
"loss": 0.6753,
"step": 426
},
{
"epoch": 1.2028169014084507,
"grad_norm": 0.7283567187045813,
"learning_rate": 7.49052743935783e-06,
"loss": 0.6255,
"step": 427
},
{
"epoch": 1.2056338028169014,
"grad_norm": 0.7503448710021737,
"learning_rate": 7.476296293357626e-06,
"loss": 0.6657,
"step": 428
},
{
"epoch": 1.208450704225352,
"grad_norm": 0.7434547390985952,
"learning_rate": 7.462038517370962e-06,
"loss": 0.69,
"step": 429
},
{
"epoch": 1.2112676056338028,
"grad_norm": 0.904002334343554,
"learning_rate": 7.4477542647253645e-06,
"loss": 0.7437,
"step": 430
},
{
"epoch": 1.2140845070422535,
"grad_norm": 0.6929989080088196,
"learning_rate": 7.4334436890330845e-06,
"loss": 0.5423,
"step": 431
},
{
"epoch": 1.2169014084507042,
"grad_norm": 0.7444154551200957,
"learning_rate": 7.4191069441894555e-06,
"loss": 0.5752,
"step": 432
},
{
"epoch": 1.2197183098591549,
"grad_norm": 0.8200086939143616,
"learning_rate": 7.404744184371229e-06,
"loss": 0.6005,
"step": 433
},
{
"epoch": 1.2225352112676056,
"grad_norm": 0.7685321740288726,
"learning_rate": 7.39035556403492e-06,
"loss": 0.6407,
"step": 434
},
{
"epoch": 1.2253521126760563,
"grad_norm": 0.7496357556269256,
"learning_rate": 7.375941237915151e-06,
"loss": 0.6213,
"step": 435
},
{
"epoch": 1.228169014084507,
"grad_norm": 0.7143102894255702,
"learning_rate": 7.361501361022978e-06,
"loss": 0.6168,
"step": 436
},
{
"epoch": 1.2309859154929577,
"grad_norm": 0.7309978741831288,
"learning_rate": 7.347036088644232e-06,
"loss": 0.633,
"step": 437
},
{
"epoch": 1.2338028169014084,
"grad_norm": 0.6905063561288732,
"learning_rate": 7.3325455763378465e-06,
"loss": 0.577,
"step": 438
},
{
"epoch": 1.236619718309859,
"grad_norm": 0.7022680079706316,
"learning_rate": 7.318029979934181e-06,
"loss": 0.5884,
"step": 439
},
{
"epoch": 1.2394366197183098,
"grad_norm": 0.7307996255503993,
"learning_rate": 7.303489455533352e-06,
"loss": 0.6065,
"step": 440
},
{
"epoch": 1.2422535211267607,
"grad_norm": 0.8189596586259464,
"learning_rate": 7.288924159503549e-06,
"loss": 0.6288,
"step": 441
},
{
"epoch": 1.2450704225352114,
"grad_norm": 0.7319103016752759,
"learning_rate": 7.274334248479353e-06,
"loss": 0.6402,
"step": 442
},
{
"epoch": 1.247887323943662,
"grad_norm": 0.6812316396156196,
"learning_rate": 7.259719879360054e-06,
"loss": 0.6013,
"step": 443
},
{
"epoch": 1.2507042253521128,
"grad_norm": 0.7654316061184241,
"learning_rate": 7.2450812093079695e-06,
"loss": 0.6302,
"step": 444
},
{
"epoch": 1.2535211267605635,
"grad_norm": 0.7768746107857665,
"learning_rate": 7.2304183957467385e-06,
"loss": 0.6225,
"step": 445
},
{
"epoch": 1.2563380281690142,
"grad_norm": 0.7430991023758472,
"learning_rate": 7.215731596359645e-06,
"loss": 0.6406,
"step": 446
},
{
"epoch": 1.2591549295774649,
"grad_norm": 0.7625029621832455,
"learning_rate": 7.201020969087913e-06,
"loss": 0.6042,
"step": 447
},
{
"epoch": 1.2619718309859156,
"grad_norm": 0.7489522138035615,
"learning_rate": 7.18628667212901e-06,
"loss": 0.6033,
"step": 448
},
{
"epoch": 1.2647887323943663,
"grad_norm": 0.7028238677579237,
"learning_rate": 7.17152886393495e-06,
"loss": 0.5917,
"step": 449
},
{
"epoch": 1.267605633802817,
"grad_norm": 0.8034495427467496,
"learning_rate": 7.15674770321058e-06,
"loss": 0.572,
"step": 450
},
{
"epoch": 1.2704225352112677,
"grad_norm": 0.7066497937705736,
"learning_rate": 7.141943348911885e-06,
"loss": 0.6047,
"step": 451
},
{
"epoch": 1.2732394366197184,
"grad_norm": 0.7798266546615258,
"learning_rate": 7.127115960244269e-06,
"loss": 0.6074,
"step": 452
},
{
"epoch": 1.276056338028169,
"grad_norm": 0.7383424636778058,
"learning_rate": 7.112265696660848e-06,
"loss": 0.5874,
"step": 453
},
{
"epoch": 1.2788732394366198,
"grad_norm": 0.7327825003857283,
"learning_rate": 7.0973927178607335e-06,
"loss": 0.5926,
"step": 454
},
{
"epoch": 1.2816901408450705,
"grad_norm": 0.6688467886867253,
"learning_rate": 7.0824971837873154e-06,
"loss": 0.6032,
"step": 455
},
{
"epoch": 1.2845070422535212,
"grad_norm": 0.71887769142022,
"learning_rate": 7.067579254626543e-06,
"loss": 0.6137,
"step": 456
},
{
"epoch": 1.287323943661972,
"grad_norm": 0.7646631375464373,
"learning_rate": 7.0526390908052e-06,
"loss": 0.5687,
"step": 457
},
{
"epoch": 1.2901408450704226,
"grad_norm": 0.7061537907237097,
"learning_rate": 7.037676852989182e-06,
"loss": 0.5671,
"step": 458
},
{
"epoch": 1.2929577464788733,
"grad_norm": 0.7702181592631588,
"learning_rate": 7.022692702081766e-06,
"loss": 0.5986,
"step": 459
},
{
"epoch": 1.295774647887324,
"grad_norm": 0.7539785898071507,
"learning_rate": 7.007686799221882e-06,
"loss": 0.5724,
"step": 460
},
{
"epoch": 1.2985915492957747,
"grad_norm": 0.800745890577906,
"learning_rate": 6.992659305782381e-06,
"loss": 0.5959,
"step": 461
},
{
"epoch": 1.3014084507042254,
"grad_norm": 0.7599227930304122,
"learning_rate": 6.977610383368296e-06,
"loss": 0.6145,
"step": 462
},
{
"epoch": 1.304225352112676,
"grad_norm": 0.7412143690063574,
"learning_rate": 6.9625401938151085e-06,
"loss": 0.5796,
"step": 463
},
{
"epoch": 1.3070422535211268,
"grad_norm": 0.6670323564590958,
"learning_rate": 6.947448899187004e-06,
"loss": 0.5279,
"step": 464
},
{
"epoch": 1.3098591549295775,
"grad_norm": 0.7448132221148877,
"learning_rate": 6.932336661775132e-06,
"loss": 0.517,
"step": 465
},
{
"epoch": 1.3126760563380282,
"grad_norm": 0.7633666757686683,
"learning_rate": 6.91720364409586e-06,
"loss": 0.5525,
"step": 466
},
{
"epoch": 1.315492957746479,
"grad_norm": 0.7901094664068569,
"learning_rate": 6.902050008889024e-06,
"loss": 0.6333,
"step": 467
},
{
"epoch": 1.3183098591549296,
"grad_norm": 0.7417299545983219,
"learning_rate": 6.886875919116184e-06,
"loss": 0.6238,
"step": 468
},
{
"epoch": 1.3211267605633803,
"grad_norm": 0.8396209754400242,
"learning_rate": 6.871681537958862e-06,
"loss": 0.7036,
"step": 469
},
{
"epoch": 1.323943661971831,
"grad_norm": 0.6793697034990046,
"learning_rate": 6.856467028816797e-06,
"loss": 0.6093,
"step": 470
},
{
"epoch": 1.3267605633802817,
"grad_norm": 0.717578251040186,
"learning_rate": 6.841232555306181e-06,
"loss": 0.6089,
"step": 471
},
{
"epoch": 1.3295774647887324,
"grad_norm": 0.7334395938548597,
"learning_rate": 6.825978281257905e-06,
"loss": 0.6266,
"step": 472
},
{
"epoch": 1.332394366197183,
"grad_norm": 0.7920253133081248,
"learning_rate": 6.810704370715791e-06,
"loss": 0.6245,
"step": 473
},
{
"epoch": 1.3352112676056338,
"grad_norm": 0.6558719992840684,
"learning_rate": 6.795410987934834e-06,
"loss": 0.5258,
"step": 474
},
{
"epoch": 1.3380281690140845,
"grad_norm": 0.6753573860958215,
"learning_rate": 6.780098297379427e-06,
"loss": 0.617,
"step": 475
},
{
"epoch": 1.3408450704225352,
"grad_norm": 0.7579142416021319,
"learning_rate": 6.764766463721605e-06,
"loss": 0.6232,
"step": 476
},
{
"epoch": 1.343661971830986,
"grad_norm": 0.8426339130047485,
"learning_rate": 6.7494156518392625e-06,
"loss": 0.7301,
"step": 477
},
{
"epoch": 1.3464788732394366,
"grad_norm": 0.7100296420326655,
"learning_rate": 6.734046026814388e-06,
"loss": 0.5676,
"step": 478
},
{
"epoch": 1.3492957746478873,
"grad_norm": 0.7140268014097398,
"learning_rate": 6.718657753931284e-06,
"loss": 0.5997,
"step": 479
},
{
"epoch": 1.352112676056338,
"grad_norm": 0.7157881332867808,
"learning_rate": 6.70325099867479e-06,
"loss": 0.5406,
"step": 480
},
{
"epoch": 1.3549295774647887,
"grad_norm": 0.7845113582925589,
"learning_rate": 6.687825926728506e-06,
"loss": 0.6069,
"step": 481
},
{
"epoch": 1.3577464788732394,
"grad_norm": 0.7758617296188998,
"learning_rate": 6.672382703973011e-06,
"loss": 0.6064,
"step": 482
},
{
"epoch": 1.36056338028169,
"grad_norm": 0.6839736903915518,
"learning_rate": 6.65692149648407e-06,
"loss": 0.5659,
"step": 483
},
{
"epoch": 1.3633802816901408,
"grad_norm": 0.8033535577918636,
"learning_rate": 6.641442470530866e-06,
"loss": 0.642,
"step": 484
},
{
"epoch": 1.3661971830985915,
"grad_norm": 0.706908397171264,
"learning_rate": 6.62594579257419e-06,
"loss": 0.5262,
"step": 485
},
{
"epoch": 1.3690140845070422,
"grad_norm": 0.7942995133323388,
"learning_rate": 6.610431629264669e-06,
"loss": 0.6028,
"step": 486
},
{
"epoch": 1.371830985915493,
"grad_norm": 0.7737746271917008,
"learning_rate": 6.594900147440963e-06,
"loss": 0.6341,
"step": 487
},
{
"epoch": 1.3746478873239436,
"grad_norm": 0.7760559302932023,
"learning_rate": 6.579351514127976e-06,
"loss": 0.6546,
"step": 488
},
{
"epoch": 1.3774647887323943,
"grad_norm": 0.7827297911698344,
"learning_rate": 6.56378589653506e-06,
"loss": 0.6294,
"step": 489
},
{
"epoch": 1.380281690140845,
"grad_norm": 0.7071080299842922,
"learning_rate": 6.548203462054211e-06,
"loss": 0.601,
"step": 490
},
{
"epoch": 1.3830985915492957,
"grad_norm": 0.7146539646324296,
"learning_rate": 6.5326043782582785e-06,
"loss": 0.5647,
"step": 491
},
{
"epoch": 1.3859154929577464,
"grad_norm": 0.8087039401070861,
"learning_rate": 6.516988812899154e-06,
"loss": 0.6269,
"step": 492
},
{
"epoch": 1.388732394366197,
"grad_norm": 0.6966320976971616,
"learning_rate": 6.501356933905973e-06,
"loss": 0.5688,
"step": 493
},
{
"epoch": 1.3915492957746478,
"grad_norm": 0.756507465924507,
"learning_rate": 6.485708909383306e-06,
"loss": 0.627,
"step": 494
},
{
"epoch": 1.3943661971830985,
"grad_norm": 0.6901180254887062,
"learning_rate": 6.4700449076093515e-06,
"loss": 0.6092,
"step": 495
},
{
"epoch": 1.3971830985915492,
"grad_norm": 0.7803910885700782,
"learning_rate": 6.454365097034127e-06,
"loss": 0.6495,
"step": 496
},
{
"epoch": 1.4,
"grad_norm": 0.7890702362457062,
"learning_rate": 6.43866964627766e-06,
"loss": 0.6578,
"step": 497
},
{
"epoch": 1.4028169014084506,
"grad_norm": 0.8090766090233742,
"learning_rate": 6.422958724128169e-06,
"loss": 0.6242,
"step": 498
},
{
"epoch": 1.4056338028169013,
"grad_norm": 0.690851480741414,
"learning_rate": 6.40723249954025e-06,
"loss": 0.6112,
"step": 499
},
{
"epoch": 1.408450704225352,
"grad_norm": 0.7180545063167437,
"learning_rate": 6.391491141633064e-06,
"loss": 0.6094,
"step": 500
},
{
"epoch": 1.4112676056338027,
"grad_norm": 0.7761047381057276,
"learning_rate": 6.375734819688514e-06,
"loss": 0.6498,
"step": 501
},
{
"epoch": 1.4140845070422534,
"grad_norm": 0.7086180360990414,
"learning_rate": 6.359963703149424e-06,
"loss": 0.5708,
"step": 502
},
{
"epoch": 1.4169014084507041,
"grad_norm": 0.7137027645516033,
"learning_rate": 6.344177961617719e-06,
"loss": 0.6064,
"step": 503
},
{
"epoch": 1.4197183098591548,
"grad_norm": 0.8064327336753806,
"learning_rate": 6.3283777648526035e-06,
"loss": 0.6317,
"step": 504
},
{
"epoch": 1.4225352112676055,
"grad_norm": 0.7651646342531121,
"learning_rate": 6.312563282768729e-06,
"loss": 0.6288,
"step": 505
},
{
"epoch": 1.4253521126760562,
"grad_norm": 0.6721576291974353,
"learning_rate": 6.296734685434373e-06,
"loss": 0.5563,
"step": 506
},
{
"epoch": 1.428169014084507,
"grad_norm": 0.7300091598809011,
"learning_rate": 6.280892143069607e-06,
"loss": 0.6739,
"step": 507
},
{
"epoch": 1.4309859154929576,
"grad_norm": 0.7631685066742252,
"learning_rate": 6.265035826044467e-06,
"loss": 0.6372,
"step": 508
},
{
"epoch": 1.4338028169014083,
"grad_norm": 0.7434045630709879,
"learning_rate": 6.2491659048771215e-06,
"loss": 0.6554,
"step": 509
},
{
"epoch": 1.436619718309859,
"grad_norm": 0.7167854438600025,
"learning_rate": 6.233282550232036e-06,
"loss": 0.6308,
"step": 510
},
{
"epoch": 1.43943661971831,
"grad_norm": 0.7054646693964248,
"learning_rate": 6.217385932918141e-06,
"loss": 0.5772,
"step": 511
},
{
"epoch": 1.4422535211267606,
"grad_norm": 0.7566300729630193,
"learning_rate": 6.201476223886993e-06,
"loss": 0.6576,
"step": 512
},
{
"epoch": 1.4450704225352113,
"grad_norm": 0.7375259475887378,
"learning_rate": 6.185553594230934e-06,
"loss": 0.6575,
"step": 513
},
{
"epoch": 1.447887323943662,
"grad_norm": 0.7209828197635741,
"learning_rate": 6.169618215181256e-06,
"loss": 0.5953,
"step": 514
},
{
"epoch": 1.4507042253521127,
"grad_norm": 0.70996297479465,
"learning_rate": 6.153670258106356e-06,
"loss": 0.5975,
"step": 515
},
{
"epoch": 1.4535211267605634,
"grad_norm": 0.7185389101707407,
"learning_rate": 6.137709894509898e-06,
"loss": 0.5733,
"step": 516
},
{
"epoch": 1.4563380281690141,
"grad_norm": 0.7916008805778101,
"learning_rate": 6.121737296028959e-06,
"loss": 0.6608,
"step": 517
},
{
"epoch": 1.4591549295774648,
"grad_norm": 0.7055343001343649,
"learning_rate": 6.105752634432195e-06,
"loss": 0.6563,
"step": 518
},
{
"epoch": 1.4619718309859155,
"grad_norm": 0.6662761212311629,
"learning_rate": 6.089756081617987e-06,
"loss": 0.5828,
"step": 519
},
{
"epoch": 1.4647887323943662,
"grad_norm": 0.7271893463200743,
"learning_rate": 6.073747809612591e-06,
"loss": 0.6323,
"step": 520
},
{
"epoch": 1.467605633802817,
"grad_norm": 0.7413143561609734,
"learning_rate": 6.057727990568294e-06,
"loss": 0.6233,
"step": 521
},
{
"epoch": 1.4704225352112676,
"grad_norm": 0.8068322899404378,
"learning_rate": 6.041696796761558e-06,
"loss": 0.6884,
"step": 522
},
{
"epoch": 1.4732394366197183,
"grad_norm": 0.76675704278936,
"learning_rate": 6.025654400591167e-06,
"loss": 0.6514,
"step": 523
},
{
"epoch": 1.476056338028169,
"grad_norm": 0.6978616412021054,
"learning_rate": 6.009600974576378e-06,
"loss": 0.5814,
"step": 524
},
{
"epoch": 1.4788732394366197,
"grad_norm": 0.7367626943759139,
"learning_rate": 5.9935366913550615e-06,
"loss": 0.61,
"step": 525
},
{
"epoch": 1.4816901408450704,
"grad_norm": 0.803924245325719,
"learning_rate": 5.977461723681845e-06,
"loss": 0.6201,
"step": 526
},
{
"epoch": 1.4845070422535211,
"grad_norm": 0.7166646169196906,
"learning_rate": 5.961376244426256e-06,
"loss": 0.5891,
"step": 527
},
{
"epoch": 1.4873239436619718,
"grad_norm": 0.7830895265196487,
"learning_rate": 5.945280426570862e-06,
"loss": 0.6437,
"step": 528
},
{
"epoch": 1.4901408450704225,
"grad_norm": 0.7077569282266005,
"learning_rate": 5.929174443209416e-06,
"loss": 0.5941,
"step": 529
},
{
"epoch": 1.4929577464788732,
"grad_norm": 0.7255908623917363,
"learning_rate": 5.913058467544984e-06,
"loss": 0.6425,
"step": 530
},
{
"epoch": 1.495774647887324,
"grad_norm": 0.6940217061921031,
"learning_rate": 5.8969326728880964e-06,
"loss": 0.615,
"step": 531
},
{
"epoch": 1.4985915492957746,
"grad_norm": 0.741185242757091,
"learning_rate": 5.880797232654869e-06,
"loss": 0.6347,
"step": 532
},
{
"epoch": 1.5014084507042254,
"grad_norm": 0.7501146031536312,
"learning_rate": 5.8646523203651514e-06,
"loss": 0.5906,
"step": 533
},
{
"epoch": 1.504225352112676,
"grad_norm": 0.7363356259610467,
"learning_rate": 5.848498109640652e-06,
"loss": 0.639,
"step": 534
},
{
"epoch": 1.5070422535211268,
"grad_norm": 0.7428030589984295,
"learning_rate": 5.832334774203076e-06,
"loss": 0.6342,
"step": 535
},
{
"epoch": 1.5098591549295775,
"grad_norm": 0.7673144526793103,
"learning_rate": 5.8161624878722545e-06,
"loss": 0.5874,
"step": 536
},
{
"epoch": 1.5126760563380282,
"grad_norm": 0.7650544824750333,
"learning_rate": 5.799981424564275e-06,
"loss": 0.6566,
"step": 537
},
{
"epoch": 1.5154929577464789,
"grad_norm": 0.7489586669750536,
"learning_rate": 5.7837917582896145e-06,
"loss": 0.6858,
"step": 538
},
{
"epoch": 1.5183098591549296,
"grad_norm": 0.7121890274043278,
"learning_rate": 5.767593663151265e-06,
"loss": 0.5805,
"step": 539
},
{
"epoch": 1.5211267605633803,
"grad_norm": 0.745809225911286,
"learning_rate": 5.751387313342863e-06,
"loss": 0.6756,
"step": 540
},
{
"epoch": 1.523943661971831,
"grad_norm": 0.7320056881511781,
"learning_rate": 5.735172883146813e-06,
"loss": 0.588,
"step": 541
},
{
"epoch": 1.5267605633802817,
"grad_norm": 0.7941805383744988,
"learning_rate": 5.718950546932418e-06,
"loss": 0.6574,
"step": 542
},
{
"epoch": 1.5295774647887324,
"grad_norm": 0.7249778426938815,
"learning_rate": 5.702720479154001e-06,
"loss": 0.636,
"step": 543
},
{
"epoch": 1.532394366197183,
"grad_norm": 1.0320714729311038,
"learning_rate": 5.686482854349029e-06,
"loss": 0.6839,
"step": 544
},
{
"epoch": 1.5352112676056338,
"grad_norm": 0.7329898223205343,
"learning_rate": 5.6702378471362394e-06,
"loss": 0.5909,
"step": 545
},
{
"epoch": 1.5380281690140845,
"grad_norm": 0.8189575127373967,
"learning_rate": 5.653985632213758e-06,
"loss": 0.6449,
"step": 546
},
{
"epoch": 1.5408450704225352,
"grad_norm": 0.733175797653608,
"learning_rate": 5.637726384357222e-06,
"loss": 0.613,
"step": 547
},
{
"epoch": 1.543661971830986,
"grad_norm": 0.8383511722961089,
"learning_rate": 5.621460278417901e-06,
"loss": 0.6849,
"step": 548
},
{
"epoch": 1.5464788732394368,
"grad_norm": 0.7237452460851425,
"learning_rate": 5.605187489320815e-06,
"loss": 0.6656,
"step": 549
},
{
"epoch": 1.5492957746478875,
"grad_norm": 0.8208760472428147,
"learning_rate": 5.588908192062858e-06,
"loss": 0.7055,
"step": 550
},
{
"epoch": 1.5521126760563382,
"grad_norm": 0.8069099779678623,
"learning_rate": 5.572622561710906e-06,
"loss": 0.6956,
"step": 551
},
{
"epoch": 1.5549295774647889,
"grad_norm": 0.7668293000678549,
"learning_rate": 5.556330773399948e-06,
"loss": 0.5757,
"step": 552
},
{
"epoch": 1.5577464788732396,
"grad_norm": 0.6893187298629392,
"learning_rate": 5.54003300233119e-06,
"loss": 0.5799,
"step": 553
},
{
"epoch": 1.5605633802816903,
"grad_norm": 0.8113423411383069,
"learning_rate": 5.52372942377018e-06,
"loss": 0.6565,
"step": 554
},
{
"epoch": 1.563380281690141,
"grad_norm": 0.711919002112785,
"learning_rate": 5.507420213044915e-06,
"loss": 0.5681,
"step": 555
},
{
"epoch": 1.5661971830985917,
"grad_norm": 0.7266032009645466,
"learning_rate": 5.491105545543966e-06,
"loss": 0.5502,
"step": 556
},
{
"epoch": 1.5690140845070424,
"grad_norm": 0.7199568163075004,
"learning_rate": 5.474785596714581e-06,
"loss": 0.6467,
"step": 557
},
{
"epoch": 1.571830985915493,
"grad_norm": 0.7273194518859644,
"learning_rate": 5.458460542060807e-06,
"loss": 0.6287,
"step": 558
},
{
"epoch": 1.5746478873239438,
"grad_norm": 0.7195884793791211,
"learning_rate": 5.442130557141595e-06,
"loss": 0.629,
"step": 559
},
{
"epoch": 1.5774647887323945,
"grad_norm": 0.7711408798912917,
"learning_rate": 5.42579581756892e-06,
"loss": 0.6274,
"step": 560
},
{
"epoch": 1.5802816901408452,
"grad_norm": 0.7321868021772925,
"learning_rate": 5.409456499005883e-06,
"loss": 0.6348,
"step": 561
},
{
"epoch": 1.5830985915492959,
"grad_norm": 1.6749686976248583,
"learning_rate": 5.393112777164834e-06,
"loss": 0.6977,
"step": 562
},
{
"epoch": 1.5859154929577466,
"grad_norm": 0.739313690634755,
"learning_rate": 5.376764827805468e-06,
"loss": 0.5653,
"step": 563
},
{
"epoch": 1.5887323943661973,
"grad_norm": 0.6890541921547552,
"learning_rate": 5.36041282673295e-06,
"loss": 0.6276,
"step": 564
},
{
"epoch": 1.591549295774648,
"grad_norm": 0.7082761122225983,
"learning_rate": 5.3440569497960126e-06,
"loss": 0.5671,
"step": 565
},
{
"epoch": 1.5943661971830987,
"grad_norm": 0.7540757628476714,
"learning_rate": 5.32769737288507e-06,
"loss": 0.6114,
"step": 566
},
{
"epoch": 1.5971830985915494,
"grad_norm": 0.7259622552539723,
"learning_rate": 5.3113342719303266e-06,
"loss": 0.5529,
"step": 567
},
{
"epoch": 1.6,
"grad_norm": 0.780097835188365,
"learning_rate": 5.294967822899882e-06,
"loss": 0.6486,
"step": 568
},
{
"epoch": 1.6028169014084508,
"grad_norm": 0.7692463121508737,
"learning_rate": 5.278598201797844e-06,
"loss": 0.6463,
"step": 569
},
{
"epoch": 1.6056338028169015,
"grad_norm": 0.7180689651375777,
"learning_rate": 5.262225584662431e-06,
"loss": 0.6551,
"step": 570
},
{
"epoch": 1.6084507042253522,
"grad_norm": 0.7570672686476732,
"learning_rate": 5.245850147564081e-06,
"loss": 0.6694,
"step": 571
},
{
"epoch": 1.611267605633803,
"grad_norm": 0.8163493286833097,
"learning_rate": 5.229472066603558e-06,
"loss": 0.71,
"step": 572
},
{
"epoch": 1.6140845070422536,
"grad_norm": 0.7627020393879351,
"learning_rate": 5.213091517910056e-06,
"loss": 0.6245,
"step": 573
},
{
"epoch": 1.6169014084507043,
"grad_norm": 0.788407246648653,
"learning_rate": 5.196708677639311e-06,
"loss": 0.623,
"step": 574
},
{
"epoch": 1.619718309859155,
"grad_norm": 0.7596410758168111,
"learning_rate": 5.180323721971696e-06,
"loss": 0.6165,
"step": 575
},
{
"epoch": 1.6225352112676057,
"grad_norm": 0.815723603914329,
"learning_rate": 5.163936827110342e-06,
"loss": 0.6754,
"step": 576
},
{
"epoch": 1.6253521126760564,
"grad_norm": 0.7348793453974766,
"learning_rate": 5.1475481692792235e-06,
"loss": 0.6324,
"step": 577
},
{
"epoch": 1.628169014084507,
"grad_norm": 0.9082203242567118,
"learning_rate": 5.131157924721285e-06,
"loss": 0.5768,
"step": 578
},
{
"epoch": 1.6309859154929578,
"grad_norm": 0.7619495779921246,
"learning_rate": 5.1147662696965254e-06,
"loss": 0.6328,
"step": 579
},
{
"epoch": 1.6338028169014085,
"grad_norm": 0.7410091511557558,
"learning_rate": 5.098373380480114e-06,
"loss": 0.6487,
"step": 580
},
{
"epoch": 1.6366197183098592,
"grad_norm": 0.7060847762225472,
"learning_rate": 5.081979433360498e-06,
"loss": 0.5852,
"step": 581
},
{
"epoch": 1.63943661971831,
"grad_norm": 0.7239693134616567,
"learning_rate": 5.065584604637492e-06,
"loss": 0.5882,
"step": 582
},
{
"epoch": 1.6422535211267606,
"grad_norm": 0.814065239966475,
"learning_rate": 5.0491890706204e-06,
"loss": 0.6785,
"step": 583
},
{
"epoch": 1.6450704225352113,
"grad_norm": 0.801293226086549,
"learning_rate": 5.0327930076261065e-06,
"loss": 0.6055,
"step": 584
},
{
"epoch": 1.647887323943662,
"grad_norm": 0.7336660465183735,
"learning_rate": 5.0163965919771855e-06,
"loss": 0.5906,
"step": 585
},
{
"epoch": 1.6507042253521127,
"grad_norm": 0.7350029739964772,
"learning_rate": 5e-06,
"loss": 0.6355,
"step": 586
},
{
"epoch": 1.6535211267605634,
"grad_norm": 0.6673037142207796,
"learning_rate": 4.983603408022817e-06,
"loss": 0.5716,
"step": 587
},
{
"epoch": 1.656338028169014,
"grad_norm": 0.7854183960568444,
"learning_rate": 4.967206992373894e-06,
"loss": 0.6507,
"step": 588
},
{
"epoch": 1.6591549295774648,
"grad_norm": 0.7320695718469675,
"learning_rate": 4.9508109293796015e-06,
"loss": 0.6632,
"step": 589
},
{
"epoch": 1.6619718309859155,
"grad_norm": 0.766444319869025,
"learning_rate": 4.9344153953625095e-06,
"loss": 0.6353,
"step": 590
},
{
"epoch": 1.6647887323943662,
"grad_norm": 0.7832821327166929,
"learning_rate": 4.918020566639505e-06,
"loss": 0.5952,
"step": 591
},
{
"epoch": 1.667605633802817,
"grad_norm": 0.7342733424551008,
"learning_rate": 4.901626619519888e-06,
"loss": 0.6418,
"step": 592
},
{
"epoch": 1.6704225352112676,
"grad_norm": 0.7538627392333438,
"learning_rate": 4.885233730303475e-06,
"loss": 0.6097,
"step": 593
},
{
"epoch": 1.6732394366197183,
"grad_norm": 0.7286411962944247,
"learning_rate": 4.868842075278717e-06,
"loss": 0.5878,
"step": 594
},
{
"epoch": 1.676056338028169,
"grad_norm": 0.7891328898786883,
"learning_rate": 4.852451830720777e-06,
"loss": 0.5764,
"step": 595
},
{
"epoch": 1.6788732394366197,
"grad_norm": 0.740444597406536,
"learning_rate": 4.83606317288966e-06,
"loss": 0.585,
"step": 596
},
{
"epoch": 1.6816901408450704,
"grad_norm": 0.7170855566165547,
"learning_rate": 4.819676278028305e-06,
"loss": 0.5812,
"step": 597
},
{
"epoch": 1.684507042253521,
"grad_norm": 0.8099731590316525,
"learning_rate": 4.803291322360691e-06,
"loss": 0.658,
"step": 598
},
{
"epoch": 1.6873239436619718,
"grad_norm": 0.7500012004775211,
"learning_rate": 4.7869084820899455e-06,
"loss": 0.6297,
"step": 599
},
{
"epoch": 1.6901408450704225,
"grad_norm": 0.6934277195584079,
"learning_rate": 4.7705279333964435e-06,
"loss": 0.6066,
"step": 600
},
{
"epoch": 1.6929577464788732,
"grad_norm": 0.856885643547699,
"learning_rate": 4.75414985243592e-06,
"loss": 0.6331,
"step": 601
},
{
"epoch": 1.695774647887324,
"grad_norm": 0.7104053919410432,
"learning_rate": 4.73777441533757e-06,
"loss": 0.5909,
"step": 602
},
{
"epoch": 1.6985915492957746,
"grad_norm": 0.7271263834548911,
"learning_rate": 4.721401798202157e-06,
"loss": 0.6467,
"step": 603
},
{
"epoch": 1.7014084507042253,
"grad_norm": 0.8176832728513387,
"learning_rate": 4.70503217710012e-06,
"loss": 0.7502,
"step": 604
},
{
"epoch": 1.704225352112676,
"grad_norm": 0.709756055906634,
"learning_rate": 4.688665728069676e-06,
"loss": 0.6424,
"step": 605
},
{
"epoch": 1.7070422535211267,
"grad_norm": 0.7620569104794489,
"learning_rate": 4.67230262711493e-06,
"loss": 0.6172,
"step": 606
},
{
"epoch": 1.7098591549295774,
"grad_norm": 0.8174091737575135,
"learning_rate": 4.655943050203987e-06,
"loss": 0.5853,
"step": 607
},
{
"epoch": 1.712676056338028,
"grad_norm": 0.7587649322008173,
"learning_rate": 4.63958717326705e-06,
"loss": 0.6383,
"step": 608
},
{
"epoch": 1.7154929577464788,
"grad_norm": 0.7302121202735942,
"learning_rate": 4.623235172194532e-06,
"loss": 0.6067,
"step": 609
},
{
"epoch": 1.7183098591549295,
"grad_norm": 0.752762704354131,
"learning_rate": 4.606887222835168e-06,
"loss": 0.6459,
"step": 610
},
{
"epoch": 1.7211267605633802,
"grad_norm": 0.797352972135618,
"learning_rate": 4.590543500994118e-06,
"loss": 0.6666,
"step": 611
},
{
"epoch": 1.723943661971831,
"grad_norm": 0.7325137557683911,
"learning_rate": 4.574204182431082e-06,
"loss": 0.6217,
"step": 612
},
{
"epoch": 1.7267605633802816,
"grad_norm": 0.7827954004872983,
"learning_rate": 4.557869442858406e-06,
"loss": 0.5668,
"step": 613
},
{
"epoch": 1.7295774647887323,
"grad_norm": 0.7987182588008852,
"learning_rate": 4.541539457939194e-06,
"loss": 0.5635,
"step": 614
},
{
"epoch": 1.732394366197183,
"grad_norm": 0.7547789895804167,
"learning_rate": 4.525214403285421e-06,
"loss": 0.6542,
"step": 615
},
{
"epoch": 1.7352112676056337,
"grad_norm": 0.8107203967195239,
"learning_rate": 4.5088944544560355e-06,
"loss": 0.6506,
"step": 616
},
{
"epoch": 1.7380281690140844,
"grad_norm": 0.7857568585526324,
"learning_rate": 4.4925797869550865e-06,
"loss": 0.6939,
"step": 617
},
{
"epoch": 1.7408450704225351,
"grad_norm": 0.6899620281094992,
"learning_rate": 4.476270576229823e-06,
"loss": 0.543,
"step": 618
},
{
"epoch": 1.7436619718309858,
"grad_norm": 0.7325667426166168,
"learning_rate": 4.459966997668812e-06,
"loss": 0.5554,
"step": 619
},
{
"epoch": 1.7464788732394365,
"grad_norm": 0.7169652401516716,
"learning_rate": 4.443669226600053e-06,
"loss": 0.5877,
"step": 620
},
{
"epoch": 1.7492957746478872,
"grad_norm": 0.7707224813831788,
"learning_rate": 4.427377438289095e-06,
"loss": 0.703,
"step": 621
},
{
"epoch": 1.752112676056338,
"grad_norm": 0.7522757952093085,
"learning_rate": 4.411091807937143e-06,
"loss": 0.6164,
"step": 622
},
{
"epoch": 1.7549295774647886,
"grad_norm": 0.7773676405964031,
"learning_rate": 4.3948125106791854e-06,
"loss": 0.6613,
"step": 623
},
{
"epoch": 1.7577464788732393,
"grad_norm": 0.8050938893918458,
"learning_rate": 4.3785397215821e-06,
"loss": 0.6181,
"step": 624
},
{
"epoch": 1.76056338028169,
"grad_norm": 0.9245792115348718,
"learning_rate": 4.362273615642779e-06,
"loss": 0.6195,
"step": 625
},
{
"epoch": 1.7633802816901407,
"grad_norm": 0.696350239262708,
"learning_rate": 4.346014367786243e-06,
"loss": 0.5982,
"step": 626
},
{
"epoch": 1.7661971830985914,
"grad_norm": 0.7722146841214672,
"learning_rate": 4.329762152863762e-06,
"loss": 0.661,
"step": 627
},
{
"epoch": 1.7690140845070421,
"grad_norm": 0.7713052523645538,
"learning_rate": 4.313517145650973e-06,
"loss": 0.5939,
"step": 628
},
{
"epoch": 1.7718309859154928,
"grad_norm": 0.8249377635295999,
"learning_rate": 4.297279520846002e-06,
"loss": 0.6603,
"step": 629
},
{
"epoch": 1.7746478873239435,
"grad_norm": 0.8847154867126514,
"learning_rate": 4.281049453067584e-06,
"loss": 0.5965,
"step": 630
},
{
"epoch": 1.7774647887323942,
"grad_norm": 0.7325092684105734,
"learning_rate": 4.264827116853189e-06,
"loss": 0.6029,
"step": 631
},
{
"epoch": 1.780281690140845,
"grad_norm": 0.7541904711913734,
"learning_rate": 4.248612686657139e-06,
"loss": 0.6091,
"step": 632
},
{
"epoch": 1.7830985915492956,
"grad_norm": 0.7433264124898937,
"learning_rate": 4.232406336848734e-06,
"loss": 0.6022,
"step": 633
},
{
"epoch": 1.7859154929577463,
"grad_norm": 0.7911613649574278,
"learning_rate": 4.2162082417103855e-06,
"loss": 0.6319,
"step": 634
},
{
"epoch": 1.788732394366197,
"grad_norm": 0.6962140364428401,
"learning_rate": 4.200018575435726e-06,
"loss": 0.6338,
"step": 635
},
{
"epoch": 1.7915492957746477,
"grad_norm": 0.8681521481970501,
"learning_rate": 4.183837512127747e-06,
"loss": 0.649,
"step": 636
},
{
"epoch": 1.7943661971830986,
"grad_norm": 0.64770897691298,
"learning_rate": 4.167665225796925e-06,
"loss": 0.56,
"step": 637
},
{
"epoch": 1.7971830985915493,
"grad_norm": 0.7494952905347187,
"learning_rate": 4.1515018903593485e-06,
"loss": 0.6356,
"step": 638
},
{
"epoch": 1.8,
"grad_norm": 0.7238165155027775,
"learning_rate": 4.135347679634849e-06,
"loss": 0.6387,
"step": 639
},
{
"epoch": 1.8028169014084507,
"grad_norm": 0.7785451847003492,
"learning_rate": 4.119202767345132e-06,
"loss": 0.6425,
"step": 640
},
{
"epoch": 1.8056338028169014,
"grad_norm": 0.7697499616316977,
"learning_rate": 4.103067327111905e-06,
"loss": 0.6415,
"step": 641
},
{
"epoch": 1.8084507042253521,
"grad_norm": 0.8706556510248158,
"learning_rate": 4.086941532455017e-06,
"loss": 0.6575,
"step": 642
},
{
"epoch": 1.8112676056338028,
"grad_norm": 0.7643847624137132,
"learning_rate": 4.070825556790587e-06,
"loss": 0.6373,
"step": 643
},
{
"epoch": 1.8140845070422535,
"grad_norm": 0.7000485049052172,
"learning_rate": 4.05471957342914e-06,
"loss": 0.6102,
"step": 644
},
{
"epoch": 1.8169014084507042,
"grad_norm": 0.7006331724829398,
"learning_rate": 4.0386237555737476e-06,
"loss": 0.6893,
"step": 645
},
{
"epoch": 1.819718309859155,
"grad_norm": 0.7135091596231999,
"learning_rate": 4.022538276318156e-06,
"loss": 0.5599,
"step": 646
},
{
"epoch": 1.8225352112676056,
"grad_norm": 0.7186120810598129,
"learning_rate": 4.006463308644939e-06,
"loss": 0.6411,
"step": 647
},
{
"epoch": 1.8253521126760563,
"grad_norm": 0.72090386197526,
"learning_rate": 3.990399025423622e-06,
"loss": 0.5957,
"step": 648
},
{
"epoch": 1.828169014084507,
"grad_norm": 0.7324797526249632,
"learning_rate": 3.974345599408833e-06,
"loss": 0.6214,
"step": 649
},
{
"epoch": 1.8309859154929577,
"grad_norm": 0.7152179493525903,
"learning_rate": 3.958303203238443e-06,
"loss": 0.5994,
"step": 650
},
{
"epoch": 1.8338028169014085,
"grad_norm": 0.6721169614227414,
"learning_rate": 3.942272009431707e-06,
"loss": 0.5858,
"step": 651
},
{
"epoch": 1.8366197183098592,
"grad_norm": 0.7024310468965755,
"learning_rate": 3.92625219038741e-06,
"loss": 0.6533,
"step": 652
},
{
"epoch": 1.8394366197183099,
"grad_norm": 0.6777544089367332,
"learning_rate": 3.910243918382015e-06,
"loss": 0.5755,
"step": 653
},
{
"epoch": 1.8422535211267606,
"grad_norm": 0.7630416452716073,
"learning_rate": 3.894247365567806e-06,
"loss": 0.6498,
"step": 654
},
{
"epoch": 1.8450704225352113,
"grad_norm": 0.7242343956415039,
"learning_rate": 3.878262703971043e-06,
"loss": 0.6183,
"step": 655
},
{
"epoch": 1.847887323943662,
"grad_norm": 0.7546009817570479,
"learning_rate": 3.8622901054901045e-06,
"loss": 0.5648,
"step": 656
},
{
"epoch": 1.8507042253521127,
"grad_norm": 0.710778044894514,
"learning_rate": 3.846329741893646e-06,
"loss": 0.5615,
"step": 657
},
{
"epoch": 1.8535211267605634,
"grad_norm": 0.7241653075763682,
"learning_rate": 3.830381784818746e-06,
"loss": 0.5854,
"step": 658
},
{
"epoch": 1.856338028169014,
"grad_norm": 0.7251045277047325,
"learning_rate": 3.814446405769069e-06,
"loss": 0.6521,
"step": 659
},
{
"epoch": 1.8591549295774648,
"grad_norm": 0.7617065909179885,
"learning_rate": 3.7985237761130077e-06,
"loss": 0.6095,
"step": 660
},
{
"epoch": 1.8619718309859155,
"grad_norm": 0.7393276040780876,
"learning_rate": 3.7826140670818597e-06,
"loss": 0.6011,
"step": 661
},
{
"epoch": 1.8647887323943662,
"grad_norm": 0.8106406503473126,
"learning_rate": 3.766717449767965e-06,
"loss": 0.5952,
"step": 662
},
{
"epoch": 1.8676056338028169,
"grad_norm": 0.7306467838266344,
"learning_rate": 3.75083409512288e-06,
"loss": 0.5472,
"step": 663
},
{
"epoch": 1.8704225352112676,
"grad_norm": 0.7717881120947158,
"learning_rate": 3.7349641739555342e-06,
"loss": 0.6911,
"step": 664
},
{
"epoch": 1.8732394366197183,
"grad_norm": 0.7245110838917453,
"learning_rate": 3.719107856930395e-06,
"loss": 0.5893,
"step": 665
},
{
"epoch": 1.8760563380281692,
"grad_norm": 0.7023053711110873,
"learning_rate": 3.703265314565629e-06,
"loss": 0.607,
"step": 666
},
{
"epoch": 1.8788732394366199,
"grad_norm": 0.7369624598798884,
"learning_rate": 3.687436717231273e-06,
"loss": 0.6088,
"step": 667
},
{
"epoch": 1.8816901408450706,
"grad_norm": 0.761769778505356,
"learning_rate": 3.6716222351473986e-06,
"loss": 0.6301,
"step": 668
},
{
"epoch": 1.8845070422535213,
"grad_norm": 0.7269704420043447,
"learning_rate": 3.6558220383822824e-06,
"loss": 0.617,
"step": 669
},
{
"epoch": 1.887323943661972,
"grad_norm": 0.7842641739090692,
"learning_rate": 3.6400362968505776e-06,
"loss": 0.5754,
"step": 670
},
{
"epoch": 1.8901408450704227,
"grad_norm": 0.7090787277322763,
"learning_rate": 3.6242651803114876e-06,
"loss": 0.5643,
"step": 671
},
{
"epoch": 1.8929577464788734,
"grad_norm": 0.8121920668052401,
"learning_rate": 3.6085088583669368e-06,
"loss": 0.6903,
"step": 672
},
{
"epoch": 1.895774647887324,
"grad_norm": 0.7887632376046526,
"learning_rate": 3.59276750045975e-06,
"loss": 0.6718,
"step": 673
},
{
"epoch": 1.8985915492957748,
"grad_norm": 0.7626415513119691,
"learning_rate": 3.5770412758718317e-06,
"loss": 0.6656,
"step": 674
},
{
"epoch": 1.9014084507042255,
"grad_norm": 0.697905106081836,
"learning_rate": 3.56133035372234e-06,
"loss": 0.623,
"step": 675
},
{
"epoch": 1.9042253521126762,
"grad_norm": 0.7531071089185666,
"learning_rate": 3.545634902965873e-06,
"loss": 0.592,
"step": 676
},
{
"epoch": 1.9070422535211269,
"grad_norm": 0.7524927476418592,
"learning_rate": 3.52995509239065e-06,
"loss": 0.6335,
"step": 677
},
{
"epoch": 1.9098591549295776,
"grad_norm": 0.804247562794407,
"learning_rate": 3.514291090616696e-06,
"loss": 0.6692,
"step": 678
},
{
"epoch": 1.9126760563380283,
"grad_norm": 0.7610688517320657,
"learning_rate": 3.4986430660940283e-06,
"loss": 0.5432,
"step": 679
},
{
"epoch": 1.915492957746479,
"grad_norm": 0.8240469545199084,
"learning_rate": 3.483011187100847e-06,
"loss": 0.7389,
"step": 680
},
{
"epoch": 1.9183098591549297,
"grad_norm": 0.789740332756857,
"learning_rate": 3.4673956217417228e-06,
"loss": 0.7044,
"step": 681
},
{
"epoch": 1.9211267605633804,
"grad_norm": 0.6593003096580221,
"learning_rate": 3.451796537945791e-06,
"loss": 0.5569,
"step": 682
},
{
"epoch": 1.923943661971831,
"grad_norm": 0.6735471334380218,
"learning_rate": 3.4362141034649434e-06,
"loss": 0.5834,
"step": 683
},
{
"epoch": 1.9267605633802818,
"grad_norm": 0.7845189104147666,
"learning_rate": 3.4206484858720267e-06,
"loss": 0.6282,
"step": 684
},
{
"epoch": 1.9295774647887325,
"grad_norm": 0.7895331554828882,
"learning_rate": 3.4050998525590406e-06,
"loss": 0.5515,
"step": 685
},
{
"epoch": 1.9323943661971832,
"grad_norm": 0.7601388876825378,
"learning_rate": 3.389568370735332e-06,
"loss": 0.6403,
"step": 686
},
{
"epoch": 1.935211267605634,
"grad_norm": 0.7414839816387602,
"learning_rate": 3.374054207425811e-06,
"loss": 0.5768,
"step": 687
},
{
"epoch": 1.9380281690140846,
"grad_norm": 0.727586108347313,
"learning_rate": 3.3585575294691355e-06,
"loss": 0.5856,
"step": 688
},
{
"epoch": 1.9408450704225353,
"grad_norm": 0.7446540281803862,
"learning_rate": 3.3430785035159297e-06,
"loss": 0.6374,
"step": 689
},
{
"epoch": 1.943661971830986,
"grad_norm": 0.7017790257080906,
"learning_rate": 3.327617296026991e-06,
"loss": 0.6344,
"step": 690
},
{
"epoch": 1.9464788732394367,
"grad_norm": 0.8243401164290531,
"learning_rate": 3.3121740732714953e-06,
"loss": 0.6616,
"step": 691
},
{
"epoch": 1.9492957746478874,
"grad_norm": 0.7026711381099482,
"learning_rate": 3.296749001325212e-06,
"loss": 0.5737,
"step": 692
},
{
"epoch": 1.952112676056338,
"grad_norm": 0.689713511673827,
"learning_rate": 3.2813422460687176e-06,
"loss": 0.573,
"step": 693
},
{
"epoch": 1.9549295774647888,
"grad_norm": 0.6772697983969495,
"learning_rate": 3.265953973185613e-06,
"loss": 0.5346,
"step": 694
},
{
"epoch": 1.9577464788732395,
"grad_norm": 0.7092702136331567,
"learning_rate": 3.250584348160738e-06,
"loss": 0.6034,
"step": 695
},
{
"epoch": 1.9605633802816902,
"grad_norm": 0.7683018287013302,
"learning_rate": 3.235233536278396e-06,
"loss": 0.6051,
"step": 696
},
{
"epoch": 1.963380281690141,
"grad_norm": 0.7173404212464849,
"learning_rate": 3.2199017026205744e-06,
"loss": 0.6139,
"step": 697
},
{
"epoch": 1.9661971830985916,
"grad_norm": 0.781035997747711,
"learning_rate": 3.204589012065168e-06,
"loss": 0.6684,
"step": 698
},
{
"epoch": 1.9690140845070423,
"grad_norm": 0.6958474566614261,
"learning_rate": 3.1892956292842103e-06,
"loss": 0.627,
"step": 699
},
{
"epoch": 1.971830985915493,
"grad_norm": 0.740337029417686,
"learning_rate": 3.1740217187420947e-06,
"loss": 0.6597,
"step": 700
},
{
"epoch": 1.9746478873239437,
"grad_norm": 0.6847378780702223,
"learning_rate": 3.158767444693819e-06,
"loss": 0.6148,
"step": 701
},
{
"epoch": 1.9774647887323944,
"grad_norm": 0.715606992756826,
"learning_rate": 3.1435329711832042e-06,
"loss": 0.5912,
"step": 702
},
{
"epoch": 1.980281690140845,
"grad_norm": 0.6889429557937115,
"learning_rate": 3.1283184620411387e-06,
"loss": 0.5801,
"step": 703
},
{
"epoch": 1.9830985915492958,
"grad_norm": 0.7439465480762761,
"learning_rate": 3.1131240808838175e-06,
"loss": 0.6253,
"step": 704
},
{
"epoch": 1.9859154929577465,
"grad_norm": 0.7406110365742725,
"learning_rate": 3.0979499911109768e-06,
"loss": 0.623,
"step": 705
},
{
"epoch": 1.9887323943661972,
"grad_norm": 0.7448811885473077,
"learning_rate": 3.082796355904142e-06,
"loss": 0.5771,
"step": 706
},
{
"epoch": 1.991549295774648,
"grad_norm": 0.7286503648205558,
"learning_rate": 3.06766333822487e-06,
"loss": 0.5972,
"step": 707
},
{
"epoch": 1.9943661971830986,
"grad_norm": 0.7567721937576899,
"learning_rate": 3.052551100812998e-06,
"loss": 0.6863,
"step": 708
},
{
"epoch": 1.9971830985915493,
"grad_norm": 0.6846994910878873,
"learning_rate": 3.0374598061848936e-06,
"loss": 0.5628,
"step": 709
},
{
"epoch": 2.0,
"grad_norm": 0.8874539697229774,
"learning_rate": 3.022389616631706e-06,
"loss": 0.6157,
"step": 710
},
{
"epoch": 2.0028169014084507,
"grad_norm": 2.569736685755078,
"learning_rate": 3.0073406942176214e-06,
"loss": 0.7215,
"step": 711
},
{
"epoch": 2.0056338028169014,
"grad_norm": 0.781170783791145,
"learning_rate": 2.9923132007781206e-06,
"loss": 0.5392,
"step": 712
},
{
"epoch": 2.008450704225352,
"grad_norm": 0.7323283178454171,
"learning_rate": 2.977307297918235e-06,
"loss": 0.5802,
"step": 713
},
{
"epoch": 2.011267605633803,
"grad_norm": 0.7082378256520369,
"learning_rate": 2.9623231470108194e-06,
"loss": 0.5261,
"step": 714
},
{
"epoch": 2.0140845070422535,
"grad_norm": 0.7404544323321716,
"learning_rate": 2.947360909194801e-06,
"loss": 0.5326,
"step": 715
},
{
"epoch": 2.016901408450704,
"grad_norm": 0.8332008982216366,
"learning_rate": 2.9324207453734575e-06,
"loss": 0.6257,
"step": 716
},
{
"epoch": 2.019718309859155,
"grad_norm": 0.7727175457086233,
"learning_rate": 2.917502816212685e-06,
"loss": 0.6265,
"step": 717
},
{
"epoch": 2.0225352112676056,
"grad_norm": 0.6705214828437596,
"learning_rate": 2.902607282139267e-06,
"loss": 0.4832,
"step": 718
},
{
"epoch": 2.0253521126760563,
"grad_norm": 0.745820194371554,
"learning_rate": 2.8877343033391523e-06,
"loss": 0.5358,
"step": 719
},
{
"epoch": 2.028169014084507,
"grad_norm": 0.6712880950400633,
"learning_rate": 2.8728840397557324e-06,
"loss": 0.4593,
"step": 720
},
{
"epoch": 2.0309859154929577,
"grad_norm": 0.7234741402033795,
"learning_rate": 2.8580566510881158e-06,
"loss": 0.5738,
"step": 721
},
{
"epoch": 2.0338028169014084,
"grad_norm": 0.7871830801210608,
"learning_rate": 2.8432522967894217e-06,
"loss": 0.5577,
"step": 722
},
{
"epoch": 2.036619718309859,
"grad_norm": 0.7623729887137076,
"learning_rate": 2.8284711360650517e-06,
"loss": 0.5423,
"step": 723
},
{
"epoch": 2.03943661971831,
"grad_norm": 0.7654196844987124,
"learning_rate": 2.8137133278709917e-06,
"loss": 0.5362,
"step": 724
},
{
"epoch": 2.0422535211267605,
"grad_norm": 0.7890889396144615,
"learning_rate": 2.7989790309120895e-06,
"loss": 0.5604,
"step": 725
},
{
"epoch": 2.045070422535211,
"grad_norm": 0.8280092947316886,
"learning_rate": 2.7842684036403557e-06,
"loss": 0.5466,
"step": 726
},
{
"epoch": 2.047887323943662,
"grad_norm": 0.7384511444156504,
"learning_rate": 2.769581604253262e-06,
"loss": 0.5797,
"step": 727
},
{
"epoch": 2.0507042253521126,
"grad_norm": 0.7594972574065925,
"learning_rate": 2.754918790692031e-06,
"loss": 0.5416,
"step": 728
},
{
"epoch": 2.0535211267605633,
"grad_norm": 0.8190052185681062,
"learning_rate": 2.7402801206399454e-06,
"loss": 0.594,
"step": 729
},
{
"epoch": 2.056338028169014,
"grad_norm": 0.8089065343356978,
"learning_rate": 2.7256657515206487e-06,
"loss": 0.6249,
"step": 730
},
{
"epoch": 2.0591549295774647,
"grad_norm": 0.6795375301061102,
"learning_rate": 2.7110758404964534e-06,
"loss": 0.5078,
"step": 731
},
{
"epoch": 2.0619718309859154,
"grad_norm": 0.7199747851182974,
"learning_rate": 2.6965105444666496e-06,
"loss": 0.5128,
"step": 732
},
{
"epoch": 2.064788732394366,
"grad_norm": 0.7389839114423702,
"learning_rate": 2.6819700200658204e-06,
"loss": 0.5165,
"step": 733
},
{
"epoch": 2.067605633802817,
"grad_norm": 0.7113982746437169,
"learning_rate": 2.667454423662156e-06,
"loss": 0.5283,
"step": 734
},
{
"epoch": 2.0704225352112675,
"grad_norm": 0.8019854200692348,
"learning_rate": 2.6529639113557694e-06,
"loss": 0.5899,
"step": 735
},
{
"epoch": 2.073239436619718,
"grad_norm": 0.7470761591455513,
"learning_rate": 2.638498638977024e-06,
"loss": 0.4994,
"step": 736
},
{
"epoch": 2.076056338028169,
"grad_norm": 0.7463273814271424,
"learning_rate": 2.6240587620848512e-06,
"loss": 0.5425,
"step": 737
},
{
"epoch": 2.0788732394366196,
"grad_norm": 0.7720000642247179,
"learning_rate": 2.6096444359650817e-06,
"loss": 0.5621,
"step": 738
},
{
"epoch": 2.0816901408450703,
"grad_norm": 0.7204053665092582,
"learning_rate": 2.595255815628774e-06,
"loss": 0.5023,
"step": 739
},
{
"epoch": 2.084507042253521,
"grad_norm": 0.7566667785442299,
"learning_rate": 2.580893055810545e-06,
"loss": 0.5514,
"step": 740
},
{
"epoch": 2.0873239436619717,
"grad_norm": 0.6809427689588321,
"learning_rate": 2.5665563109669155e-06,
"loss": 0.5008,
"step": 741
},
{
"epoch": 2.0901408450704224,
"grad_norm": 0.810140586801637,
"learning_rate": 2.5522457352746368e-06,
"loss": 0.5852,
"step": 742
},
{
"epoch": 2.092957746478873,
"grad_norm": 0.7584702735377782,
"learning_rate": 2.5379614826290384e-06,
"loss": 0.5008,
"step": 743
},
{
"epoch": 2.095774647887324,
"grad_norm": 0.7871575310769966,
"learning_rate": 2.5237037066423747e-06,
"loss": 0.6055,
"step": 744
},
{
"epoch": 2.0985915492957745,
"grad_norm": 0.7431957816566821,
"learning_rate": 2.509472560642171e-06,
"loss": 0.579,
"step": 745
},
{
"epoch": 2.101408450704225,
"grad_norm": 0.74509594480476,
"learning_rate": 2.495268197669573e-06,
"loss": 0.5521,
"step": 746
},
{
"epoch": 2.104225352112676,
"grad_norm": 0.7603284772487195,
"learning_rate": 2.481090770477706e-06,
"loss": 0.5488,
"step": 747
},
{
"epoch": 2.1070422535211266,
"grad_norm": 0.7387452720309511,
"learning_rate": 2.466940431530026e-06,
"loss": 0.5107,
"step": 748
},
{
"epoch": 2.1098591549295773,
"grad_norm": 0.774018311134093,
"learning_rate": 2.4528173329986855e-06,
"loss": 0.5619,
"step": 749
},
{
"epoch": 2.112676056338028,
"grad_norm": 0.7094138769934575,
"learning_rate": 2.438721626762892e-06,
"loss": 0.5249,
"step": 750
},
{
"epoch": 2.1154929577464787,
"grad_norm": 0.7857734566073393,
"learning_rate": 2.424653464407281e-06,
"loss": 0.5783,
"step": 751
},
{
"epoch": 2.1183098591549294,
"grad_norm": 0.6681431433578359,
"learning_rate": 2.4106129972202793e-06,
"loss": 0.4836,
"step": 752
},
{
"epoch": 2.12112676056338,
"grad_norm": 0.7059928071944473,
"learning_rate": 2.3966003761924816e-06,
"loss": 0.4662,
"step": 753
},
{
"epoch": 2.123943661971831,
"grad_norm": 0.7193801912300508,
"learning_rate": 2.382615752015028e-06,
"loss": 0.5193,
"step": 754
},
{
"epoch": 2.1267605633802815,
"grad_norm": 0.7102845448999637,
"learning_rate": 2.3686592750779788e-06,
"loss": 0.5067,
"step": 755
},
{
"epoch": 2.129577464788732,
"grad_norm": 0.7427634111028115,
"learning_rate": 2.3547310954687018e-06,
"loss": 0.6256,
"step": 756
},
{
"epoch": 2.132394366197183,
"grad_norm": 0.6979392039426204,
"learning_rate": 2.340831362970257e-06,
"loss": 0.5096,
"step": 757
},
{
"epoch": 2.1352112676056336,
"grad_norm": 0.7444797928330332,
"learning_rate": 2.326960227059784e-06,
"loss": 0.5149,
"step": 758
},
{
"epoch": 2.1380281690140843,
"grad_norm": 0.7150529701936503,
"learning_rate": 2.313117836906897e-06,
"loss": 0.5426,
"step": 759
},
{
"epoch": 2.140845070422535,
"grad_norm": 0.8010874698588896,
"learning_rate": 2.2993043413720784e-06,
"loss": 0.5932,
"step": 760
},
{
"epoch": 2.1436619718309857,
"grad_norm": 0.6872563327495382,
"learning_rate": 2.28551988900508e-06,
"loss": 0.5149,
"step": 761
},
{
"epoch": 2.1464788732394364,
"grad_norm": 0.7460312856986838,
"learning_rate": 2.271764628043324e-06,
"loss": 0.55,
"step": 762
},
{
"epoch": 2.149295774647887,
"grad_norm": 0.7242016051510175,
"learning_rate": 2.258038706410311e-06,
"loss": 0.5533,
"step": 763
},
{
"epoch": 2.152112676056338,
"grad_norm": 0.7495937718335975,
"learning_rate": 2.2443422717140246e-06,
"loss": 0.566,
"step": 764
},
{
"epoch": 2.1549295774647885,
"grad_norm": 0.8090509188921273,
"learning_rate": 2.2306754712453504e-06,
"loss": 0.6159,
"step": 765
},
{
"epoch": 2.1577464788732392,
"grad_norm": 0.7296358155167403,
"learning_rate": 2.217038451976485e-06,
"loss": 0.5482,
"step": 766
},
{
"epoch": 2.1605633802816904,
"grad_norm": 0.728917758383568,
"learning_rate": 2.203431360559365e-06,
"loss": 0.5387,
"step": 767
},
{
"epoch": 2.1633802816901406,
"grad_norm": 0.7271467482036231,
"learning_rate": 2.1898543433240787e-06,
"loss": 0.5074,
"step": 768
},
{
"epoch": 2.1661971830985918,
"grad_norm": 0.7197770997886006,
"learning_rate": 2.1763075462773002e-06,
"loss": 0.5288,
"step": 769
},
{
"epoch": 2.169014084507042,
"grad_norm": 0.7238841621409884,
"learning_rate": 2.1627911151007176e-06,
"loss": 0.5436,
"step": 770
},
{
"epoch": 2.171830985915493,
"grad_norm": 0.7568866047442696,
"learning_rate": 2.149305195149463e-06,
"loss": 0.5536,
"step": 771
},
{
"epoch": 2.1746478873239434,
"grad_norm": 0.765821236668226,
"learning_rate": 2.135849931450555e-06,
"loss": 0.5557,
"step": 772
},
{
"epoch": 2.1774647887323946,
"grad_norm": 0.7794966716146099,
"learning_rate": 2.1224254687013347e-06,
"loss": 0.5536,
"step": 773
},
{
"epoch": 2.1802816901408453,
"grad_norm": 0.7525735089142571,
"learning_rate": 2.1090319512679115e-06,
"loss": 0.511,
"step": 774
},
{
"epoch": 2.183098591549296,
"grad_norm": 0.7288945530927061,
"learning_rate": 2.0956695231836094e-06,
"loss": 0.5095,
"step": 775
},
{
"epoch": 2.1859154929577467,
"grad_norm": 0.81522934972113,
"learning_rate": 2.0823383281474202e-06,
"loss": 0.5456,
"step": 776
},
{
"epoch": 2.1887323943661974,
"grad_norm": 0.8161254562914294,
"learning_rate": 2.0690385095224557e-06,
"loss": 0.5454,
"step": 777
},
{
"epoch": 2.191549295774648,
"grad_norm": 0.7453811173652941,
"learning_rate": 2.0557702103344078e-06,
"loss": 0.5445,
"step": 778
},
{
"epoch": 2.1943661971830988,
"grad_norm": 0.6574151968727797,
"learning_rate": 2.0425335732700075e-06,
"loss": 0.492,
"step": 779
},
{
"epoch": 2.1971830985915495,
"grad_norm": 0.720899834077065,
"learning_rate": 2.0293287406754976e-06,
"loss": 0.4851,
"step": 780
},
{
"epoch": 2.2,
"grad_norm": 0.6987996403755788,
"learning_rate": 2.0161558545550925e-06,
"loss": 0.5257,
"step": 781
},
{
"epoch": 2.202816901408451,
"grad_norm": 0.7902727538260627,
"learning_rate": 2.0030150565694566e-06,
"loss": 0.5831,
"step": 782
},
{
"epoch": 2.2056338028169016,
"grad_norm": 0.786480042851265,
"learning_rate": 1.989906488034181e-06,
"loss": 0.5464,
"step": 783
},
{
"epoch": 2.2084507042253523,
"grad_norm": 0.7459476013930395,
"learning_rate": 1.976830289918261e-06,
"loss": 0.5339,
"step": 784
},
{
"epoch": 2.211267605633803,
"grad_norm": 0.7899232013030847,
"learning_rate": 1.9637866028425835e-06,
"loss": 0.5228,
"step": 785
},
{
"epoch": 2.2140845070422537,
"grad_norm": 0.7298589604614047,
"learning_rate": 1.950775567078411e-06,
"loss": 0.5464,
"step": 786
},
{
"epoch": 2.2169014084507044,
"grad_norm": 0.7390448234363022,
"learning_rate": 1.9377973225458773e-06,
"loss": 0.6104,
"step": 787
},
{
"epoch": 2.219718309859155,
"grad_norm": 0.798187214794037,
"learning_rate": 1.924852008812479e-06,
"loss": 0.5742,
"step": 788
},
{
"epoch": 2.222535211267606,
"grad_norm": 0.705513153122921,
"learning_rate": 1.9119397650915774e-06,
"loss": 0.4954,
"step": 789
},
{
"epoch": 2.2253521126760565,
"grad_norm": 0.7803490529105689,
"learning_rate": 1.8990607302408991e-06,
"loss": 0.5839,
"step": 790
},
{
"epoch": 2.228169014084507,
"grad_norm": 0.772361187559715,
"learning_rate": 1.8862150427610453e-06,
"loss": 0.468,
"step": 791
},
{
"epoch": 2.230985915492958,
"grad_norm": 0.6941194804936315,
"learning_rate": 1.8734028407940003e-06,
"loss": 0.5351,
"step": 792
},
{
"epoch": 2.2338028169014086,
"grad_norm": 0.7124075210532292,
"learning_rate": 1.8606242621216443e-06,
"loss": 0.5628,
"step": 793
},
{
"epoch": 2.2366197183098593,
"grad_norm": 0.7702429591736301,
"learning_rate": 1.84787944416428e-06,
"loss": 0.4947,
"step": 794
},
{
"epoch": 2.23943661971831,
"grad_norm": 0.7404741820420859,
"learning_rate": 1.835168523979144e-06,
"loss": 0.54,
"step": 795
},
{
"epoch": 2.2422535211267607,
"grad_norm": 0.7337881238934444,
"learning_rate": 1.822491638258938e-06,
"loss": 0.5223,
"step": 796
},
{
"epoch": 2.2450704225352114,
"grad_norm": 0.732982524092136,
"learning_rate": 1.8098489233303595e-06,
"loss": 0.5504,
"step": 797
},
{
"epoch": 2.247887323943662,
"grad_norm": 0.7789300885571788,
"learning_rate": 1.7972405151526339e-06,
"loss": 0.5886,
"step": 798
},
{
"epoch": 2.250704225352113,
"grad_norm": 0.7432973879306417,
"learning_rate": 1.7846665493160548e-06,
"loss": 0.5765,
"step": 799
},
{
"epoch": 2.2535211267605635,
"grad_norm": 0.7148013041871587,
"learning_rate": 1.7721271610405232e-06,
"loss": 0.5088,
"step": 800
},
{
"epoch": 2.256338028169014,
"grad_norm": 0.7190924440924693,
"learning_rate": 1.7596224851740938e-06,
"loss": 0.5668,
"step": 801
},
{
"epoch": 2.259154929577465,
"grad_norm": 0.8176217354813979,
"learning_rate": 1.747152656191527e-06,
"loss": 0.5516,
"step": 802
},
{
"epoch": 2.2619718309859156,
"grad_norm": 0.7848830670126005,
"learning_rate": 1.7347178081928407e-06,
"loss": 0.555,
"step": 803
},
{
"epoch": 2.2647887323943663,
"grad_norm": 0.727192900316563,
"learning_rate": 1.7223180749018675e-06,
"loss": 0.5689,
"step": 804
},
{
"epoch": 2.267605633802817,
"grad_norm": 0.7379445481611638,
"learning_rate": 1.7099535896648205e-06,
"loss": 0.4954,
"step": 805
},
{
"epoch": 2.2704225352112677,
"grad_norm": 0.7304515509378274,
"learning_rate": 1.6976244854488545e-06,
"loss": 0.4682,
"step": 806
},
{
"epoch": 2.2732394366197184,
"grad_norm": 0.6864120639105775,
"learning_rate": 1.6853308948406387e-06,
"loss": 0.5088,
"step": 807
},
{
"epoch": 2.276056338028169,
"grad_norm": 0.6814634212226736,
"learning_rate": 1.6730729500449322e-06,
"loss": 0.4825,
"step": 808
},
{
"epoch": 2.27887323943662,
"grad_norm": 0.6828842277477668,
"learning_rate": 1.6608507828831572e-06,
"loss": 0.4605,
"step": 809
},
{
"epoch": 2.2816901408450705,
"grad_norm": 0.7224430516077436,
"learning_rate": 1.648664524791988e-06,
"loss": 0.5593,
"step": 810
},
{
"epoch": 2.284507042253521,
"grad_norm": 0.7151328091297615,
"learning_rate": 1.636514306821933e-06,
"loss": 0.5455,
"step": 811
},
{
"epoch": 2.287323943661972,
"grad_norm": 0.7039701863483498,
"learning_rate": 1.6244002596359255e-06,
"loss": 0.5416,
"step": 812
},
{
"epoch": 2.2901408450704226,
"grad_norm": 0.7326190869630805,
"learning_rate": 1.6123225135079212e-06,
"loss": 0.4614,
"step": 813
},
{
"epoch": 2.2929577464788733,
"grad_norm": 0.7405292155656916,
"learning_rate": 1.6002811983214962e-06,
"loss": 0.4931,
"step": 814
},
{
"epoch": 2.295774647887324,
"grad_norm": 0.7636311159243301,
"learning_rate": 1.5882764435684473e-06,
"loss": 0.5507,
"step": 815
},
{
"epoch": 2.2985915492957747,
"grad_norm": 0.7583190657458188,
"learning_rate": 1.5763083783474048e-06,
"loss": 0.5374,
"step": 816
},
{
"epoch": 2.3014084507042254,
"grad_norm": 0.7285565659472655,
"learning_rate": 1.5643771313624394e-06,
"loss": 0.5595,
"step": 817
},
{
"epoch": 2.304225352112676,
"grad_norm": 0.7347006926328963,
"learning_rate": 1.552482830921681e-06,
"loss": 0.555,
"step": 818
},
{
"epoch": 2.307042253521127,
"grad_norm": 0.7630789601522776,
"learning_rate": 1.5406256049359359e-06,
"loss": 0.5609,
"step": 819
},
{
"epoch": 2.3098591549295775,
"grad_norm": 0.7177620817775812,
"learning_rate": 1.5288055809173174e-06,
"loss": 0.5069,
"step": 820
},
{
"epoch": 2.312676056338028,
"grad_norm": 0.6983432291948535,
"learning_rate": 1.517022885977868e-06,
"loss": 0.532,
"step": 821
},
{
"epoch": 2.315492957746479,
"grad_norm": 0.7460327681895941,
"learning_rate": 1.5052776468281933e-06,
"loss": 0.569,
"step": 822
},
{
"epoch": 2.3183098591549296,
"grad_norm": 0.7790295408392904,
"learning_rate": 1.4935699897761031e-06,
"loss": 0.4965,
"step": 823
},
{
"epoch": 2.3211267605633803,
"grad_norm": 0.7437941519396709,
"learning_rate": 1.4819000407252498e-06,
"loss": 0.5675,
"step": 824
},
{
"epoch": 2.323943661971831,
"grad_norm": 0.6841364101492224,
"learning_rate": 1.4702679251737768e-06,
"loss": 0.5586,
"step": 825
},
{
"epoch": 2.3267605633802817,
"grad_norm": 0.7202178343462267,
"learning_rate": 1.4586737682129653e-06,
"loss": 0.4825,
"step": 826
},
{
"epoch": 2.3295774647887324,
"grad_norm": 0.70160510066986,
"learning_rate": 1.4471176945258947e-06,
"loss": 0.5504,
"step": 827
},
{
"epoch": 2.332394366197183,
"grad_norm": 0.8365294239621358,
"learning_rate": 1.435599828386095e-06,
"loss": 0.5608,
"step": 828
},
{
"epoch": 2.335211267605634,
"grad_norm": 0.7361499428255823,
"learning_rate": 1.4241202936562164e-06,
"loss": 0.5188,
"step": 829
},
{
"epoch": 2.3380281690140845,
"grad_norm": 0.7872741665321432,
"learning_rate": 1.412679213786694e-06,
"loss": 0.563,
"step": 830
},
{
"epoch": 2.340845070422535,
"grad_norm": 0.7530076640574325,
"learning_rate": 1.40127671181442e-06,
"loss": 0.5875,
"step": 831
},
{
"epoch": 2.343661971830986,
"grad_norm": 0.7322800554870551,
"learning_rate": 1.3899129103614229e-06,
"loss": 0.4666,
"step": 832
},
{
"epoch": 2.3464788732394366,
"grad_norm": 0.7673082565463836,
"learning_rate": 1.3785879316335448e-06,
"loss": 0.5516,
"step": 833
},
{
"epoch": 2.3492957746478873,
"grad_norm": 0.6632810241670146,
"learning_rate": 1.3673018974191354e-06,
"loss": 0.448,
"step": 834
},
{
"epoch": 2.352112676056338,
"grad_norm": 2.7516018782482017,
"learning_rate": 1.3560549290877318e-06,
"loss": 0.5398,
"step": 835
},
{
"epoch": 2.3549295774647887,
"grad_norm": 0.7211879094149636,
"learning_rate": 1.3448471475887587e-06,
"loss": 0.5065,
"step": 836
},
{
"epoch": 2.3577464788732394,
"grad_norm": 0.7995856802043136,
"learning_rate": 1.3336786734502294e-06,
"loss": 0.5797,
"step": 837
},
{
"epoch": 2.36056338028169,
"grad_norm": 0.758044778698354,
"learning_rate": 1.3225496267774452e-06,
"loss": 0.562,
"step": 838
},
{
"epoch": 2.363380281690141,
"grad_norm": 0.7575990009876885,
"learning_rate": 1.311460127251708e-06,
"loss": 0.5316,
"step": 839
},
{
"epoch": 2.3661971830985915,
"grad_norm": 0.6736159728507464,
"learning_rate": 1.3004102941290297e-06,
"loss": 0.4474,
"step": 840
},
{
"epoch": 2.369014084507042,
"grad_norm": 0.7205227793745679,
"learning_rate": 1.2894002462388533e-06,
"loss": 0.5218,
"step": 841
},
{
"epoch": 2.371830985915493,
"grad_norm": 0.7217650676264715,
"learning_rate": 1.2784301019827705e-06,
"loss": 0.5204,
"step": 842
},
{
"epoch": 2.3746478873239436,
"grad_norm": 0.7557329810128117,
"learning_rate": 1.2674999793332539e-06,
"loss": 0.5403,
"step": 843
},
{
"epoch": 2.3774647887323943,
"grad_norm": 0.7478007085164625,
"learning_rate": 1.2566099958323824e-06,
"loss": 0.5698,
"step": 844
},
{
"epoch": 2.380281690140845,
"grad_norm": 0.698380036442176,
"learning_rate": 1.2457602685905834e-06,
"loss": 0.5048,
"step": 845
},
{
"epoch": 2.3830985915492957,
"grad_norm": 0.7695914348531823,
"learning_rate": 1.2349509142853672e-06,
"loss": 0.4931,
"step": 846
},
{
"epoch": 2.3859154929577464,
"grad_norm": 0.7762276319195065,
"learning_rate": 1.224182049160077e-06,
"loss": 0.6103,
"step": 847
},
{
"epoch": 2.388732394366197,
"grad_norm": 0.7274222023728415,
"learning_rate": 1.2134537890226366e-06,
"loss": 0.5363,
"step": 848
},
{
"epoch": 2.391549295774648,
"grad_norm": 0.711203424608623,
"learning_rate": 1.202766249244306e-06,
"loss": 0.5136,
"step": 849
},
{
"epoch": 2.3943661971830985,
"grad_norm": 0.6870670016862169,
"learning_rate": 1.1921195447584388e-06,
"loss": 0.5454,
"step": 850
},
{
"epoch": 2.397183098591549,
"grad_norm": 0.7149525528652956,
"learning_rate": 1.1815137900592488e-06,
"loss": 0.5306,
"step": 851
},
{
"epoch": 2.4,
"grad_norm": 0.7638558998644358,
"learning_rate": 1.1709490992005774e-06,
"loss": 0.5352,
"step": 852
},
{
"epoch": 2.4028169014084506,
"grad_norm": 0.7302152693103684,
"learning_rate": 1.1604255857946667e-06,
"loss": 0.5532,
"step": 853
},
{
"epoch": 2.4056338028169013,
"grad_norm": 0.7649138698287055,
"learning_rate": 1.149943363010938e-06,
"loss": 0.522,
"step": 854
},
{
"epoch": 2.408450704225352,
"grad_norm": 0.7093300838883863,
"learning_rate": 1.1395025435747759e-06,
"loss": 0.5058,
"step": 855
},
{
"epoch": 2.4112676056338027,
"grad_norm": 0.7009610002847351,
"learning_rate": 1.1291032397663142e-06,
"loss": 0.5207,
"step": 856
},
{
"epoch": 2.4140845070422534,
"grad_norm": 0.7033284525805,
"learning_rate": 1.1187455634192307e-06,
"loss": 0.5504,
"step": 857
},
{
"epoch": 2.416901408450704,
"grad_norm": 0.7256841908682241,
"learning_rate": 1.108429625919541e-06,
"loss": 0.6407,
"step": 858
},
{
"epoch": 2.419718309859155,
"grad_norm": 0.6662534100506925,
"learning_rate": 1.098155538204404e-06,
"loss": 0.4797,
"step": 859
},
{
"epoch": 2.4225352112676055,
"grad_norm": 0.6908974021921962,
"learning_rate": 1.08792341076093e-06,
"loss": 0.5222,
"step": 860
},
{
"epoch": 2.425352112676056,
"grad_norm": 0.6876509805065439,
"learning_rate": 1.0777333536249873e-06,
"loss": 0.4901,
"step": 861
},
{
"epoch": 2.428169014084507,
"grad_norm": 0.6927349527894001,
"learning_rate": 1.067585476380023e-06,
"loss": 0.4973,
"step": 862
},
{
"epoch": 2.4309859154929576,
"grad_norm": 0.6939033155329007,
"learning_rate": 1.0574798881558834e-06,
"loss": 0.5595,
"step": 863
},
{
"epoch": 2.4338028169014083,
"grad_norm": 0.7110587459331594,
"learning_rate": 1.0474166976276396e-06,
"loss": 0.5999,
"step": 864
},
{
"epoch": 2.436619718309859,
"grad_norm": 0.7552175181737919,
"learning_rate": 1.0373960130144206e-06,
"loss": 0.5635,
"step": 865
},
{
"epoch": 2.4394366197183097,
"grad_norm": 0.7322000886422406,
"learning_rate": 1.0274179420782487e-06,
"loss": 0.5159,
"step": 866
},
{
"epoch": 2.4422535211267604,
"grad_norm": 0.6978339620980003,
"learning_rate": 1.0174825921228803e-06,
"loss": 0.5035,
"step": 867
},
{
"epoch": 2.445070422535211,
"grad_norm": 0.7233687516678328,
"learning_rate": 1.0075900699926523e-06,
"loss": 0.4615,
"step": 868
},
{
"epoch": 2.447887323943662,
"grad_norm": 0.7029331443183017,
"learning_rate": 9.977404820713315e-07,
"loss": 0.5125,
"step": 869
},
{
"epoch": 2.4507042253521125,
"grad_norm": 0.795131750879923,
"learning_rate": 9.879339342809741e-07,
"loss": 0.5948,
"step": 870
},
{
"epoch": 2.453521126760563,
"grad_norm": 0.6927831704928772,
"learning_rate": 9.781705320807833e-07,
"loss": 0.5056,
"step": 871
},
{
"epoch": 2.456338028169014,
"grad_norm": 0.7217137637353825,
"learning_rate": 9.684503804659773e-07,
"loss": 0.5544,
"step": 872
},
{
"epoch": 2.4591549295774646,
"grad_norm": 0.7363521243814246,
"learning_rate": 9.587735839666573e-07,
"loss": 0.5393,
"step": 873
},
{
"epoch": 2.4619718309859153,
"grad_norm": 0.7313271519524395,
"learning_rate": 9.491402466466893e-07,
"loss": 0.5885,
"step": 874
},
{
"epoch": 2.464788732394366,
"grad_norm": 0.7118534580991188,
"learning_rate": 9.395504721025773e-07,
"loss": 0.5589,
"step": 875
},
{
"epoch": 2.4676056338028167,
"grad_norm": 0.7610936167591283,
"learning_rate": 9.300043634623546e-07,
"loss": 0.5192,
"step": 876
},
{
"epoch": 2.4704225352112674,
"grad_norm": 0.7178990480634524,
"learning_rate": 9.205020233844736e-07,
"loss": 0.4867,
"step": 877
},
{
"epoch": 2.473239436619718,
"grad_norm": 0.7145917047867616,
"learning_rate": 9.110435540567003e-07,
"loss": 0.5257,
"step": 878
},
{
"epoch": 2.476056338028169,
"grad_norm": 0.6846734741974113,
"learning_rate": 9.016290571950171e-07,
"loss": 0.5265,
"step": 879
},
{
"epoch": 2.4788732394366195,
"grad_norm": 0.7431090137610343,
"learning_rate": 8.922586340425288e-07,
"loss": 0.5516,
"step": 880
},
{
"epoch": 2.4816901408450702,
"grad_norm": 0.770495805259206,
"learning_rate": 8.829323853683719e-07,
"loss": 0.6051,
"step": 881
},
{
"epoch": 2.4845070422535214,
"grad_norm": 0.6952836508138835,
"learning_rate": 8.736504114666345e-07,
"loss": 0.511,
"step": 882
},
{
"epoch": 2.4873239436619716,
"grad_norm": 0.7402309170159516,
"learning_rate": 8.644128121552742e-07,
"loss": 0.5602,
"step": 883
},
{
"epoch": 2.4901408450704228,
"grad_norm": 0.7102436346119819,
"learning_rate": 8.552196867750462e-07,
"loss": 0.5326,
"step": 884
},
{
"epoch": 2.492957746478873,
"grad_norm": 0.7904150584495637,
"learning_rate": 8.460711341884353e-07,
"loss": 0.5704,
"step": 885
},
{
"epoch": 2.495774647887324,
"grad_norm": 0.7701966990313119,
"learning_rate": 8.36967252778591e-07,
"loss": 0.5924,
"step": 886
},
{
"epoch": 2.4985915492957744,
"grad_norm": 0.7369703311898811,
"learning_rate": 8.279081404482748e-07,
"loss": 0.5543,
"step": 887
},
{
"epoch": 2.5014084507042256,
"grad_norm": 0.7235281553818372,
"learning_rate": 8.188938946187991e-07,
"loss": 0.5459,
"step": 888
},
{
"epoch": 2.504225352112676,
"grad_norm": 0.7571994304517072,
"learning_rate": 8.099246122289861e-07,
"loss": 0.5572,
"step": 889
},
{
"epoch": 2.507042253521127,
"grad_norm": 0.6593905074365984,
"learning_rate": 8.010003897341212e-07,
"loss": 0.4757,
"step": 890
},
{
"epoch": 2.5098591549295772,
"grad_norm": 0.7809776499441126,
"learning_rate": 7.921213231049196e-07,
"loss": 0.583,
"step": 891
},
{
"epoch": 2.5126760563380284,
"grad_norm": 0.7763572290071068,
"learning_rate": 7.832875078264912e-07,
"loss": 0.5567,
"step": 892
},
{
"epoch": 2.5154929577464786,
"grad_norm": 0.8164057637936308,
"learning_rate": 7.74499038897315e-07,
"loss": 0.6174,
"step": 893
},
{
"epoch": 2.5183098591549298,
"grad_norm": 0.7487499523494067,
"learning_rate": 7.65756010828217e-07,
"loss": 0.5256,
"step": 894
},
{
"epoch": 2.52112676056338,
"grad_norm": 0.692947524253345,
"learning_rate": 7.570585176413547e-07,
"loss": 0.527,
"step": 895
},
{
"epoch": 2.523943661971831,
"grad_norm": 0.7745072471556728,
"learning_rate": 7.484066528692041e-07,
"loss": 0.5582,
"step": 896
},
{
"epoch": 2.5267605633802814,
"grad_norm": 0.759370710618239,
"learning_rate": 7.398005095535565e-07,
"loss": 0.5745,
"step": 897
},
{
"epoch": 2.5295774647887326,
"grad_norm": 0.7408984808818695,
"learning_rate": 7.312401802445169e-07,
"loss": 0.5308,
"step": 898
},
{
"epoch": 2.532394366197183,
"grad_norm": 0.7722385173607145,
"learning_rate": 7.227257569995061e-07,
"loss": 0.5938,
"step": 899
},
{
"epoch": 2.535211267605634,
"grad_norm": 0.7358053652279244,
"learning_rate": 7.142573313822754e-07,
"loss": 0.5992,
"step": 900
},
{
"epoch": 2.5380281690140842,
"grad_norm": 0.7192406498904013,
"learning_rate": 7.058349944619186e-07,
"loss": 0.5353,
"step": 901
},
{
"epoch": 2.5408450704225354,
"grad_norm": 0.7158007177960015,
"learning_rate": 6.974588368118934e-07,
"loss": 0.5445,
"step": 902
},
{
"epoch": 2.543661971830986,
"grad_norm": 0.7826333376030878,
"learning_rate": 6.89128948509048e-07,
"loss": 0.6257,
"step": 903
},
{
"epoch": 2.546478873239437,
"grad_norm": 0.7252757135255095,
"learning_rate": 6.808454191326519e-07,
"loss": 0.523,
"step": 904
},
{
"epoch": 2.5492957746478875,
"grad_norm": 0.7296052273393889,
"learning_rate": 6.726083377634318e-07,
"loss": 0.4892,
"step": 905
},
{
"epoch": 2.552112676056338,
"grad_norm": 0.6998049271823165,
"learning_rate": 6.644177929826162e-07,
"loss": 0.5785,
"step": 906
},
{
"epoch": 2.554929577464789,
"grad_norm": 0.6866771635498244,
"learning_rate": 6.562738728709795e-07,
"loss": 0.5141,
"step": 907
},
{
"epoch": 2.5577464788732396,
"grad_norm": 0.762851841283065,
"learning_rate": 6.481766650078969e-07,
"loss": 0.5784,
"step": 908
},
{
"epoch": 2.5605633802816903,
"grad_norm": 0.745798205319872,
"learning_rate": 6.401262564704019e-07,
"loss": 0.5615,
"step": 909
},
{
"epoch": 2.563380281690141,
"grad_norm": 0.7267076581981277,
"learning_rate": 6.321227338322511e-07,
"loss": 0.588,
"step": 910
},
{
"epoch": 2.5661971830985917,
"grad_norm": 0.683833936306517,
"learning_rate": 6.241661831629902e-07,
"loss": 0.5261,
"step": 911
},
{
"epoch": 2.5690140845070424,
"grad_norm": 0.785279539380278,
"learning_rate": 6.162566900270311e-07,
"loss": 0.5273,
"step": 912
},
{
"epoch": 2.571830985915493,
"grad_norm": 0.6821061292671917,
"learning_rate": 6.083943394827329e-07,
"loss": 0.4928,
"step": 913
},
{
"epoch": 2.574647887323944,
"grad_norm": 0.7156545834103997,
"learning_rate": 6.005792160814821e-07,
"loss": 0.5511,
"step": 914
},
{
"epoch": 2.5774647887323945,
"grad_norm": 0.6867901055418119,
"learning_rate": 5.928114038667888e-07,
"loss": 0.5047,
"step": 915
},
{
"epoch": 2.580281690140845,
"grad_norm": 0.6960326420245707,
"learning_rate": 5.850909863733784e-07,
"loss": 0.5591,
"step": 916
},
{
"epoch": 2.583098591549296,
"grad_norm": 0.7774414120953896,
"learning_rate": 5.774180466262985e-07,
"loss": 0.6052,
"step": 917
},
{
"epoch": 2.5859154929577466,
"grad_norm": 0.73166882890799,
"learning_rate": 5.697926671400194e-07,
"loss": 0.5807,
"step": 918
},
{
"epoch": 2.5887323943661973,
"grad_norm": 0.754328972659702,
"learning_rate": 5.622149299175539e-07,
"loss": 0.5593,
"step": 919
},
{
"epoch": 2.591549295774648,
"grad_norm": 0.7172975394869928,
"learning_rate": 5.546849164495688e-07,
"loss": 0.5434,
"step": 920
},
{
"epoch": 2.5943661971830987,
"grad_norm": 0.7653703517730494,
"learning_rate": 5.472027077135145e-07,
"loss": 0.5656,
"step": 921
},
{
"epoch": 2.5971830985915494,
"grad_norm": 0.6981722575591085,
"learning_rate": 5.397683841727485e-07,
"loss": 0.4886,
"step": 922
},
{
"epoch": 2.6,
"grad_norm": 0.7444541339360583,
"learning_rate": 5.323820257756745e-07,
"loss": 0.5395,
"step": 923
},
{
"epoch": 2.602816901408451,
"grad_norm": 0.6835507442292554,
"learning_rate": 5.250437119548817e-07,
"loss": 0.518,
"step": 924
},
{
"epoch": 2.6056338028169015,
"grad_norm": 0.6763495669522698,
"learning_rate": 5.177535216262885e-07,
"loss": 0.4711,
"step": 925
},
{
"epoch": 2.608450704225352,
"grad_norm": 0.6976570030316313,
"learning_rate": 5.105115331882954e-07,
"loss": 0.5282,
"step": 926
},
{
"epoch": 2.611267605633803,
"grad_norm": 0.7230013623985598,
"learning_rate": 5.033178245209436e-07,
"loss": 0.5319,
"step": 927
},
{
"epoch": 2.6140845070422536,
"grad_norm": 0.701113900863958,
"learning_rate": 4.961724729850731e-07,
"loss": 0.4762,
"step": 928
},
{
"epoch": 2.6169014084507043,
"grad_norm": 0.7593924318858984,
"learning_rate": 4.890755554214954e-07,
"loss": 0.5236,
"step": 929
},
{
"epoch": 2.619718309859155,
"grad_norm": 0.7245468462578131,
"learning_rate": 4.820271481501642e-07,
"loss": 0.5012,
"step": 930
},
{
"epoch": 2.6225352112676057,
"grad_norm": 0.742178416513282,
"learning_rate": 4.7502732696935507e-07,
"loss": 0.5618,
"step": 931
},
{
"epoch": 2.6253521126760564,
"grad_norm": 0.7757926136824119,
"learning_rate": 4.680761671548517e-07,
"loss": 0.5689,
"step": 932
},
{
"epoch": 2.628169014084507,
"grad_norm": 0.757413214376559,
"learning_rate": 4.6117374345913454e-07,
"loss": 0.6093,
"step": 933
},
{
"epoch": 2.630985915492958,
"grad_norm": 0.7808878960276914,
"learning_rate": 4.5432013011057984e-07,
"loss": 0.6574,
"step": 934
},
{
"epoch": 2.6338028169014085,
"grad_norm": 0.7432341814513662,
"learning_rate": 4.4751540081265645e-07,
"loss": 0.5579,
"step": 935
},
{
"epoch": 2.636619718309859,
"grad_norm": 0.7432292701218901,
"learning_rate": 4.407596287431387e-07,
"loss": 0.464,
"step": 936
},
{
"epoch": 2.63943661971831,
"grad_norm": 0.7142037612001625,
"learning_rate": 4.340528865533161e-07,
"loss": 0.5436,
"step": 937
},
{
"epoch": 2.6422535211267606,
"grad_norm": 0.7587147906576946,
"learning_rate": 4.2739524636721207e-07,
"loss": 0.5685,
"step": 938
},
{
"epoch": 2.6450704225352113,
"grad_norm": 0.7250510864530743,
"learning_rate": 4.207867797808102e-07,
"loss": 0.5706,
"step": 939
},
{
"epoch": 2.647887323943662,
"grad_norm": 0.7444089098403069,
"learning_rate": 4.1422755786128364e-07,
"loss": 0.5828,
"step": 940
},
{
"epoch": 2.6507042253521127,
"grad_norm": 0.7111869917066532,
"learning_rate": 4.0771765114622886e-07,
"loss": 0.5301,
"step": 941
},
{
"epoch": 2.6535211267605634,
"grad_norm": 0.6726007711358601,
"learning_rate": 4.012571296429102e-07,
"loss": 0.4944,
"step": 942
},
{
"epoch": 2.656338028169014,
"grad_norm": 0.7311381996970749,
"learning_rate": 3.948460628275047e-07,
"loss": 0.5329,
"step": 943
},
{
"epoch": 2.659154929577465,
"grad_norm": 0.7535498452656507,
"learning_rate": 3.8848451964435594e-07,
"loss": 0.5611,
"step": 944
},
{
"epoch": 2.6619718309859155,
"grad_norm": 0.7368613397695026,
"learning_rate": 3.8217256850523243e-07,
"loss": 0.575,
"step": 945
},
{
"epoch": 2.664788732394366,
"grad_norm": 0.8283481948228351,
"learning_rate": 3.759102772885925e-07,
"loss": 0.6088,
"step": 946
},
{
"epoch": 2.667605633802817,
"grad_norm": 0.6831237388540137,
"learning_rate": 3.696977133388524e-07,
"loss": 0.5037,
"step": 947
},
{
"epoch": 2.6704225352112676,
"grad_norm": 0.654320245207509,
"learning_rate": 3.635349434656638e-07,
"loss": 0.4809,
"step": 948
},
{
"epoch": 2.6732394366197183,
"grad_norm": 0.6823803621068965,
"learning_rate": 3.5742203394319606e-07,
"loss": 0.4552,
"step": 949
},
{
"epoch": 2.676056338028169,
"grad_norm": 0.7047964358595283,
"learning_rate": 3.513590505094222e-07,
"loss": 0.5088,
"step": 950
},
{
"epoch": 2.6788732394366197,
"grad_norm": 0.7489158544565429,
"learning_rate": 3.453460583654106e-07,
"loss": 0.52,
"step": 951
},
{
"epoch": 2.6816901408450704,
"grad_norm": 0.7482041266928414,
"learning_rate": 3.3938312217462686e-07,
"loss": 0.4875,
"step": 952
},
{
"epoch": 2.684507042253521,
"grad_norm": 0.7319451096212228,
"learning_rate": 3.334703060622374e-07,
"loss": 0.546,
"step": 953
},
{
"epoch": 2.687323943661972,
"grad_norm": 0.793605571383183,
"learning_rate": 3.2760767361441847e-07,
"loss": 0.5887,
"step": 954
},
{
"epoch": 2.6901408450704225,
"grad_norm": 0.695997307638334,
"learning_rate": 3.2179528787767313e-07,
"loss": 0.5182,
"step": 955
},
{
"epoch": 2.692957746478873,
"grad_norm": 0.7360062330805416,
"learning_rate": 3.1603321135815434e-07,
"loss": 0.6046,
"step": 956
},
{
"epoch": 2.695774647887324,
"grad_norm": 0.7164054436806284,
"learning_rate": 3.103215060209902e-07,
"loss": 0.5753,
"step": 957
},
{
"epoch": 2.6985915492957746,
"grad_norm": 0.7148121631621239,
"learning_rate": 3.046602332896209e-07,
"loss": 0.5756,
"step": 958
},
{
"epoch": 2.7014084507042253,
"grad_norm": 0.6847968930046422,
"learning_rate": 2.9904945404513606e-07,
"loss": 0.5609,
"step": 959
},
{
"epoch": 2.704225352112676,
"grad_norm": 0.7182111847241899,
"learning_rate": 2.9348922862562025e-07,
"loss": 0.5507,
"step": 960
},
{
"epoch": 2.7070422535211267,
"grad_norm": 0.782842512987059,
"learning_rate": 2.879796168255039e-07,
"loss": 0.5537,
"step": 961
},
{
"epoch": 2.7098591549295774,
"grad_norm": 0.7642184480760954,
"learning_rate": 2.825206778949219e-07,
"loss": 0.5525,
"step": 962
},
{
"epoch": 2.712676056338028,
"grad_norm": 0.7731085239034594,
"learning_rate": 2.771124705390743e-07,
"loss": 0.6081,
"step": 963
},
{
"epoch": 2.715492957746479,
"grad_norm": 0.7025162728788912,
"learning_rate": 2.717550529175955e-07,
"loss": 0.4821,
"step": 964
},
{
"epoch": 2.7183098591549295,
"grad_norm": 0.7198454479509169,
"learning_rate": 2.664484826439312e-07,
"loss": 0.5213,
"step": 965
},
{
"epoch": 2.72112676056338,
"grad_norm": 0.7576950704979507,
"learning_rate": 2.611928167847133e-07,
"loss": 0.5119,
"step": 966
},
{
"epoch": 2.723943661971831,
"grad_norm": 0.7263675362698118,
"learning_rate": 2.5598811185915397e-07,
"loss": 0.5309,
"step": 967
},
{
"epoch": 2.7267605633802816,
"grad_norm": 0.8071498912093302,
"learning_rate": 2.5083442383843136e-07,
"loss": 0.6026,
"step": 968
},
{
"epoch": 2.7295774647887323,
"grad_norm": 0.716757593430612,
"learning_rate": 2.457318081450899e-07,
"loss": 0.4898,
"step": 969
},
{
"epoch": 2.732394366197183,
"grad_norm": 0.7652310058804378,
"learning_rate": 2.4068031965244486e-07,
"loss": 0.5468,
"step": 970
},
{
"epoch": 2.7352112676056337,
"grad_norm": 0.7250545500999873,
"learning_rate": 2.3568001268399088e-07,
"loss": 0.5776,
"step": 971
},
{
"epoch": 2.7380281690140844,
"grad_norm": 0.728736775512018,
"learning_rate": 2.3073094101282056e-07,
"loss": 0.5576,
"step": 972
},
{
"epoch": 2.740845070422535,
"grad_norm": 0.6672516406987512,
"learning_rate": 2.258331578610423e-07,
"loss": 0.4974,
"step": 973
},
{
"epoch": 2.743661971830986,
"grad_norm": 0.706845139654572,
"learning_rate": 2.2098671589921017e-07,
"loss": 0.5409,
"step": 974
},
{
"epoch": 2.7464788732394365,
"grad_norm": 0.6557210086436945,
"learning_rate": 2.1619166724575924e-07,
"loss": 0.5145,
"step": 975
},
{
"epoch": 2.749295774647887,
"grad_norm": 0.6968785117284902,
"learning_rate": 2.114480634664401e-07,
"loss": 0.525,
"step": 976
},
{
"epoch": 2.752112676056338,
"grad_norm": 0.7560876394254348,
"learning_rate": 2.0675595557376916e-07,
"loss": 0.5601,
"step": 977
},
{
"epoch": 2.7549295774647886,
"grad_norm": 0.7526488510684558,
"learning_rate": 2.0211539402647807e-07,
"loss": 0.5691,
"step": 978
},
{
"epoch": 2.7577464788732393,
"grad_norm": 0.6816358178509648,
"learning_rate": 1.9752642872897078e-07,
"loss": 0.4796,
"step": 979
},
{
"epoch": 2.76056338028169,
"grad_norm": 0.7374455810319034,
"learning_rate": 1.9298910903078838e-07,
"loss": 0.5415,
"step": 980
},
{
"epoch": 2.7633802816901407,
"grad_norm": 0.6844607412321583,
"learning_rate": 1.8850348372607575e-07,
"loss": 0.5275,
"step": 981
},
{
"epoch": 2.7661971830985914,
"grad_norm": 0.6959147100628197,
"learning_rate": 1.840696010530596e-07,
"loss": 0.5354,
"step": 982
},
{
"epoch": 2.769014084507042,
"grad_norm": 0.7374698652571952,
"learning_rate": 1.7968750869352792e-07,
"loss": 0.5119,
"step": 983
},
{
"epoch": 2.771830985915493,
"grad_norm": 0.717663869275445,
"learning_rate": 1.753572537723186e-07,
"loss": 0.5049,
"step": 984
},
{
"epoch": 2.7746478873239435,
"grad_norm": 0.6950965076865067,
"learning_rate": 1.7107888285681106e-07,
"loss": 0.55,
"step": 985
},
{
"epoch": 2.777464788732394,
"grad_norm": 0.7644851193031408,
"learning_rate": 1.6685244195642715e-07,
"loss": 0.5415,
"step": 986
},
{
"epoch": 2.780281690140845,
"grad_norm": 0.7470884993973315,
"learning_rate": 1.6267797652213368e-07,
"loss": 0.5456,
"step": 987
},
{
"epoch": 2.7830985915492956,
"grad_norm": 0.7605411571407109,
"learning_rate": 1.5855553144595858e-07,
"loss": 0.5865,
"step": 988
},
{
"epoch": 2.7859154929577463,
"grad_norm": 0.7366936546933648,
"learning_rate": 1.5448515106050165e-07,
"loss": 0.4865,
"step": 989
},
{
"epoch": 2.788732394366197,
"grad_norm": 0.7539626563855862,
"learning_rate": 1.5046687913846392e-07,
"loss": 0.5467,
"step": 990
},
{
"epoch": 2.7915492957746477,
"grad_norm": 0.7014547717515902,
"learning_rate": 1.4650075889217297e-07,
"loss": 0.5575,
"step": 991
},
{
"epoch": 2.7943661971830984,
"grad_norm": 0.683981370181626,
"learning_rate": 1.4258683297311891e-07,
"loss": 0.5005,
"step": 992
},
{
"epoch": 2.7971830985915496,
"grad_norm": 0.7191901291359036,
"learning_rate": 1.3872514347149756e-07,
"loss": 0.5829,
"step": 993
},
{
"epoch": 2.8,
"grad_norm": 0.6988468027896046,
"learning_rate": 1.3491573191575513e-07,
"loss": 0.4915,
"step": 994
},
{
"epoch": 2.802816901408451,
"grad_norm": 0.7458383149541131,
"learning_rate": 1.3115863927214423e-07,
"loss": 0.5119,
"step": 995
},
{
"epoch": 2.8056338028169012,
"grad_norm": 0.7792116484568382,
"learning_rate": 1.2745390594428143e-07,
"loss": 0.5684,
"step": 996
},
{
"epoch": 2.8084507042253524,
"grad_norm": 0.7680837077144288,
"learning_rate": 1.2380157177271369e-07,
"loss": 0.5421,
"step": 997
},
{
"epoch": 2.8112676056338026,
"grad_norm": 0.7354184387370749,
"learning_rate": 1.202016760344893e-07,
"loss": 0.5652,
"step": 998
},
{
"epoch": 2.8140845070422538,
"grad_norm": 0.7345418602331815,
"learning_rate": 1.166542574427354e-07,
"loss": 0.5414,
"step": 999
},
{
"epoch": 2.816901408450704,
"grad_norm": 0.7264331561764557,
"learning_rate": 1.131593541462439e-07,
"loss": 0.5053,
"step": 1000
},
{
"epoch": 2.819718309859155,
"grad_norm": 0.7215698428704066,
"learning_rate": 1.0971700372905736e-07,
"loss": 0.5694,
"step": 1001
},
{
"epoch": 2.8225352112676054,
"grad_norm": 0.6924431328276708,
"learning_rate": 1.0632724321006816e-07,
"loss": 0.4931,
"step": 1002
},
{
"epoch": 2.8253521126760566,
"grad_norm": 0.7433999913262525,
"learning_rate": 1.0299010904261886e-07,
"loss": 0.5269,
"step": 1003
},
{
"epoch": 2.828169014084507,
"grad_norm": 0.7591641686077665,
"learning_rate": 9.970563711410974e-08,
"loss": 0.5789,
"step": 1004
},
{
"epoch": 2.830985915492958,
"grad_norm": 0.7206709788993783,
"learning_rate": 9.647386274561466e-08,
"loss": 0.5783,
"step": 1005
},
{
"epoch": 2.8338028169014082,
"grad_norm": 0.763303001346764,
"learning_rate": 9.329482069149853e-08,
"loss": 0.5639,
"step": 1006
},
{
"epoch": 2.8366197183098594,
"grad_norm": 0.7171907005458857,
"learning_rate": 9.016854513904828e-08,
"loss": 0.5147,
"step": 1007
},
{
"epoch": 2.8394366197183096,
"grad_norm": 0.7333285024549918,
"learning_rate": 8.709506970809855e-08,
"loss": 0.5431,
"step": 1008
},
{
"epoch": 2.8422535211267608,
"grad_norm": 0.7283810889890997,
"learning_rate": 8.407442745067552e-08,
"loss": 0.5362,
"step": 1009
},
{
"epoch": 2.845070422535211,
"grad_norm": 0.7149664274563876,
"learning_rate": 8.110665085063918e-08,
"loss": 0.5162,
"step": 1010
},
{
"epoch": 2.847887323943662,
"grad_norm": 0.7302249666055635,
"learning_rate": 7.819177182333493e-08,
"loss": 0.5649,
"step": 1011
},
{
"epoch": 2.8507042253521124,
"grad_norm": 0.6819672875462226,
"learning_rate": 7.532982171524927e-08,
"loss": 0.5188,
"step": 1012
},
{
"epoch": 2.8535211267605636,
"grad_norm": 0.7429108794162049,
"learning_rate": 7.252083130367349e-08,
"loss": 0.5547,
"step": 1013
},
{
"epoch": 2.856338028169014,
"grad_norm": 0.6925223924918105,
"learning_rate": 6.97648307963733e-08,
"loss": 0.5092,
"step": 1014
},
{
"epoch": 2.859154929577465,
"grad_norm": 0.731757804946538,
"learning_rate": 6.706184983126196e-08,
"loss": 0.5549,
"step": 1015
},
{
"epoch": 2.8619718309859152,
"grad_norm": 0.7104692499216267,
"learning_rate": 6.441191747608322e-08,
"loss": 0.5388,
"step": 1016
},
{
"epoch": 2.8647887323943664,
"grad_norm": 0.6701176981914706,
"learning_rate": 6.181506222809885e-08,
"loss": 0.5234,
"step": 1017
},
{
"epoch": 2.8676056338028166,
"grad_norm": 0.7709788734633877,
"learning_rate": 5.9271312013781665e-08,
"loss": 0.593,
"step": 1018
},
{
"epoch": 2.870422535211268,
"grad_norm": 0.7829907431052303,
"learning_rate": 5.678069418851351e-08,
"loss": 0.6111,
"step": 1019
},
{
"epoch": 2.873239436619718,
"grad_norm": 0.7037842335096338,
"learning_rate": 5.4343235536294944e-08,
"loss": 0.4803,
"step": 1020
},
{
"epoch": 2.876056338028169,
"grad_norm": 0.667841942594328,
"learning_rate": 5.195896226945385e-08,
"loss": 0.4703,
"step": 1021
},
{
"epoch": 2.87887323943662,
"grad_norm": 0.7093517015710021,
"learning_rate": 4.9627900028365036e-08,
"loss": 0.5247,
"step": 1022
},
{
"epoch": 2.8816901408450706,
"grad_norm": 0.7804094006727855,
"learning_rate": 4.735007388117441e-08,
"loss": 0.5622,
"step": 1023
},
{
"epoch": 2.8845070422535213,
"grad_norm": 0.7281503997189487,
"learning_rate": 4.5125508323528025e-08,
"loss": 0.5361,
"step": 1024
},
{
"epoch": 2.887323943661972,
"grad_norm": 0.7459062923541441,
"learning_rate": 4.2954227278310666e-08,
"loss": 0.5618,
"step": 1025
},
{
"epoch": 2.8901408450704227,
"grad_norm": 0.7441252088100604,
"learning_rate": 4.083625409538772e-08,
"loss": 0.6007,
"step": 1026
},
{
"epoch": 2.8929577464788734,
"grad_norm": 0.717554049706811,
"learning_rate": 3.877161155135423e-08,
"loss": 0.5736,
"step": 1027
},
{
"epoch": 2.895774647887324,
"grad_norm": 0.696592390563458,
"learning_rate": 3.676032184928957e-08,
"loss": 0.5046,
"step": 1028
},
{
"epoch": 2.898591549295775,
"grad_norm": 0.6542018044465501,
"learning_rate": 3.4802406618518195e-08,
"loss": 0.4593,
"step": 1029
},
{
"epoch": 2.9014084507042255,
"grad_norm": 0.6815251687238045,
"learning_rate": 3.289788691437923e-08,
"loss": 0.4588,
"step": 1030
},
{
"epoch": 2.904225352112676,
"grad_norm": 0.7027235666996269,
"learning_rate": 3.10467832179967e-08,
"loss": 0.5113,
"step": 1031
},
{
"epoch": 2.907042253521127,
"grad_norm": 0.7607538460194083,
"learning_rate": 2.9249115436063013e-08,
"loss": 0.5941,
"step": 1032
},
{
"epoch": 2.9098591549295776,
"grad_norm": 0.7220736725918624,
"learning_rate": 2.7504902900621356e-08,
"loss": 0.4933,
"step": 1033
},
{
"epoch": 2.9126760563380283,
"grad_norm": 0.7166615863712383,
"learning_rate": 2.581416436886086e-08,
"loss": 0.5728,
"step": 1034
},
{
"epoch": 2.915492957746479,
"grad_norm": 0.7971855756389676,
"learning_rate": 2.4176918022912333e-08,
"loss": 0.6093,
"step": 1035
},
{
"epoch": 2.9183098591549297,
"grad_norm": 0.708380543577938,
"learning_rate": 2.2593181469653945e-08,
"loss": 0.5801,
"step": 1036
},
{
"epoch": 2.9211267605633804,
"grad_norm": 0.7432289103223266,
"learning_rate": 2.1062971740523076e-08,
"loss": 0.5307,
"step": 1037
},
{
"epoch": 2.923943661971831,
"grad_norm": 0.7786874168213611,
"learning_rate": 1.958630529133032e-08,
"loss": 0.5472,
"step": 1038
},
{
"epoch": 2.926760563380282,
"grad_norm": 0.7737112697875131,
"learning_rate": 1.816319800208466e-08,
"loss": 0.6477,
"step": 1039
},
{
"epoch": 2.9295774647887325,
"grad_norm": 0.755421734849608,
"learning_rate": 1.679366517682246e-08,
"loss": 0.4996,
"step": 1040
},
{
"epoch": 2.932394366197183,
"grad_norm": 0.7344390660036586,
"learning_rate": 1.547772154344207e-08,
"loss": 0.5684,
"step": 1041
},
{
"epoch": 2.935211267605634,
"grad_norm": 0.7455018669291504,
"learning_rate": 1.4215381253546712e-08,
"loss": 0.5352,
"step": 1042
},
{
"epoch": 2.9380281690140846,
"grad_norm": 0.6500928946288003,
"learning_rate": 1.3006657882290163e-08,
"loss": 0.4517,
"step": 1043
},
{
"epoch": 2.9408450704225353,
"grad_norm": 0.7398268482770803,
"learning_rate": 1.1851564428232987e-08,
"loss": 0.4815,
"step": 1044
},
{
"epoch": 2.943661971830986,
"grad_norm": 0.6758826237123673,
"learning_rate": 1.0750113313202082e-08,
"loss": 0.4791,
"step": 1045
},
{
"epoch": 2.9464788732394367,
"grad_norm": 0.7585979778426581,
"learning_rate": 9.70231638215635e-09,
"loss": 0.6102,
"step": 1046
},
{
"epoch": 2.9492957746478874,
"grad_norm": 0.7103789084083153,
"learning_rate": 8.708184903060135e-09,
"loss": 0.4698,
"step": 1047
},
{
"epoch": 2.952112676056338,
"grad_norm": 0.7183494091050963,
"learning_rate": 7.76772956676164e-09,
"loss": 0.5356,
"step": 1048
},
{
"epoch": 2.954929577464789,
"grad_norm": 0.7193080568255721,
"learning_rate": 6.880960486877475e-09,
"loss": 0.5064,
"step": 1049
},
{
"epoch": 2.9577464788732395,
"grad_norm": 0.7797853793145919,
"learning_rate": 6.047887199686075e-09,
"loss": 0.5628,
"step": 1050
},
{
"epoch": 2.96056338028169,
"grad_norm": 0.7388278933529127,
"learning_rate": 5.268518664022226e-09,
"loss": 0.5486,
"step": 1051
},
{
"epoch": 2.963380281690141,
"grad_norm": 0.7159582661188039,
"learning_rate": 4.54286326118214e-09,
"loss": 0.5335,
"step": 1052
},
{
"epoch": 2.9661971830985916,
"grad_norm": 0.7133437665306763,
"learning_rate": 3.870928794834083e-09,
"loss": 0.4978,
"step": 1053
},
{
"epoch": 2.9690140845070423,
"grad_norm": 0.732528489265048,
"learning_rate": 3.2527224909328914e-09,
"loss": 0.561,
"step": 1054
},
{
"epoch": 2.971830985915493,
"grad_norm": 0.7353062443834698,
"learning_rate": 2.6882509976433604e-09,
"loss": 0.5916,
"step": 1055
},
{
"epoch": 2.9746478873239437,
"grad_norm": 0.7331761189776203,
"learning_rate": 2.1775203852680836e-09,
"loss": 0.4967,
"step": 1056
},
{
"epoch": 2.9774647887323944,
"grad_norm": 0.7162322949974765,
"learning_rate": 1.7205361461825054e-09,
"loss": 0.5463,
"step": 1057
},
{
"epoch": 2.980281690140845,
"grad_norm": 0.7187767220372333,
"learning_rate": 1.317303194776076e-09,
"loss": 0.4954,
"step": 1058
},
{
"epoch": 2.983098591549296,
"grad_norm": 0.7051041793653454,
"learning_rate": 9.678258673995189e-10,
"loss": 0.5135,
"step": 1059
},
{
"epoch": 2.9859154929577465,
"grad_norm": 0.7268714889415322,
"learning_rate": 6.721079223165339e-10,
"loss": 0.5272,
"step": 1060
},
{
"epoch": 2.988732394366197,
"grad_norm": 1.1890425193935092,
"learning_rate": 4.301525396666062e-10,
"loss": 0.6821,
"step": 1061
},
{
"epoch": 2.991549295774648,
"grad_norm": 0.7030030929326179,
"learning_rate": 2.4196232142614794e-10,
"loss": 0.5686,
"step": 1062
},
{
"epoch": 2.9943661971830986,
"grad_norm": 0.6908075641092671,
"learning_rate": 1.0753929138629382e-10,
"loss": 0.5257,
"step": 1063
},
{
"epoch": 2.9971830985915493,
"grad_norm": 0.7428862677010641,
"learning_rate": 2.6884895126255695e-11,
"loss": 0.4927,
"step": 1064
},
{
"epoch": 3.0,
"grad_norm": 0.7428387204524438,
"learning_rate": 0.0,
"loss": 0.5818,
"step": 1065
},
{
"epoch": 3.0,
"step": 1065,
"total_flos": 171917700349952.0,
"train_loss": 0.6223782367549592,
"train_runtime": 3879.5995,
"train_samples_per_second": 26.353,
"train_steps_per_second": 0.275
}
],
"logging_steps": 1,
"max_steps": 1065,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 171917700349952.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}