{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.4398988232706478,
  "eval_steps": 500,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00021994941163532388,
      "grad_norm": 0.7960259914398193,
      "learning_rate": 8e-05,
      "loss": 2.2745,
      "step": 1
    },
    {
      "epoch": 0.00043989882327064776,
      "grad_norm": 0.3960127830505371,
      "learning_rate": 8e-05,
      "loss": 1.9155,
      "step": 2
    },
    {
      "epoch": 0.0006598482349059716,
      "grad_norm": 0.3869185745716095,
      "learning_rate": 8e-05,
      "loss": 1.8754,
      "step": 3
    },
    {
      "epoch": 0.0008797976465412955,
      "grad_norm": 0.33234134316444397,
      "learning_rate": 8e-05,
      "loss": 1.9126,
      "step": 4
    },
    {
      "epoch": 0.0010997470581766194,
      "grad_norm": 0.3670472502708435,
      "learning_rate": 8e-05,
      "loss": 1.9912,
      "step": 5
    },
    {
      "epoch": 0.0013196964698119432,
      "grad_norm": 0.32942938804626465,
      "learning_rate": 8e-05,
      "loss": 1.8746,
      "step": 6
    },
    {
      "epoch": 0.001539645881447267,
      "grad_norm": 0.3588086664676666,
      "learning_rate": 8e-05,
      "loss": 1.9545,
      "step": 7
    },
    {
      "epoch": 0.001759595293082591,
      "grad_norm": 0.33002492785453796,
      "learning_rate": 8e-05,
      "loss": 1.8781,
      "step": 8
    },
    {
      "epoch": 0.001979544704717915,
      "grad_norm": 0.3024381101131439,
      "learning_rate": 8e-05,
      "loss": 1.8859,
      "step": 9
    },
    {
      "epoch": 0.002199494116353239,
      "grad_norm": 0.3224199712276459,
      "learning_rate": 8e-05,
      "loss": 1.829,
      "step": 10
    },
    {
      "epoch": 0.0024194435279885624,
      "grad_norm": 0.31481102108955383,
      "learning_rate": 8e-05,
      "loss": 1.817,
      "step": 11
    },
    {
      "epoch": 0.0026393929396238865,
      "grad_norm": 0.3078259825706482,
      "learning_rate": 8e-05,
      "loss": 1.9631,
      "step": 12
    },
    {
      "epoch": 0.0028593423512592105,
      "grad_norm": 0.3141743540763855,
      "learning_rate": 8e-05,
      "loss": 1.8879,
      "step": 13
    },
    {
      "epoch": 0.003079291762894534,
      "grad_norm": 0.29574745893478394,
      "learning_rate": 8e-05,
      "loss": 1.799,
      "step": 14
    },
    {
      "epoch": 0.003299241174529858,
      "grad_norm": 0.3095031976699829,
      "learning_rate": 8e-05,
      "loss": 1.8741,
      "step": 15
    },
    {
      "epoch": 0.003519190586165182,
      "grad_norm": 0.28804337978363037,
      "learning_rate": 8e-05,
      "loss": 1.9931,
      "step": 16
    },
    {
      "epoch": 0.0037391399978005057,
      "grad_norm": 0.25137585401535034,
      "learning_rate": 8e-05,
      "loss": 1.6762,
      "step": 17
    },
    {
      "epoch": 0.00395908940943583,
      "grad_norm": 0.28565698862075806,
      "learning_rate": 8e-05,
      "loss": 1.8489,
      "step": 18
    },
    {
      "epoch": 0.004179038821071153,
      "grad_norm": 0.2877500355243683,
      "learning_rate": 8e-05,
      "loss": 1.8871,
      "step": 19
    },
    {
      "epoch": 0.004398988232706478,
      "grad_norm": 0.28803154826164246,
      "learning_rate": 8e-05,
      "loss": 1.6956,
      "step": 20
    },
    {
      "epoch": 0.004618937644341801,
      "grad_norm": 0.32161301374435425,
      "learning_rate": 8e-05,
      "loss": 1.8243,
      "step": 21
    },
    {
      "epoch": 0.004838887055977125,
      "grad_norm": 0.2959391176700592,
      "learning_rate": 8e-05,
      "loss": 1.8991,
      "step": 22
    },
    {
      "epoch": 0.005058836467612449,
      "grad_norm": 0.3021189868450165,
      "learning_rate": 8e-05,
      "loss": 1.9975,
      "step": 23
    },
    {
      "epoch": 0.005278785879247773,
      "grad_norm": 0.2793104946613312,
      "learning_rate": 8e-05,
      "loss": 1.8792,
      "step": 24
    },
    {
      "epoch": 0.0054987352908830965,
      "grad_norm": 0.2658381760120392,
      "learning_rate": 8e-05,
      "loss": 1.6467,
      "step": 25
    },
    {
      "epoch": 0.005718684702518421,
      "grad_norm": 0.2793010175228119,
      "learning_rate": 8e-05,
      "loss": 1.7479,
      "step": 26
    },
    {
      "epoch": 0.0059386341141537445,
      "grad_norm": 0.2800044119358063,
      "learning_rate": 8e-05,
      "loss": 1.7885,
      "step": 27
    },
    {
      "epoch": 0.006158583525789068,
      "grad_norm": 0.2864585220813751,
      "learning_rate": 8e-05,
      "loss": 1.9257,
      "step": 28
    },
    {
      "epoch": 0.006378532937424393,
      "grad_norm": 0.301496684551239,
      "learning_rate": 8e-05,
      "loss": 1.8586,
      "step": 29
    },
    {
      "epoch": 0.006598482349059716,
      "grad_norm": 0.2858293354511261,
      "learning_rate": 8e-05,
      "loss": 1.8541,
      "step": 30
    },
    {
      "epoch": 0.00681843176069504,
      "grad_norm": 0.31271278858184814,
      "learning_rate": 8e-05,
      "loss": 1.8774,
      "step": 31
    },
    {
      "epoch": 0.007038381172330364,
      "grad_norm": 0.30428266525268555,
      "learning_rate": 8e-05,
      "loss": 1.8464,
      "step": 32
    },
    {
      "epoch": 0.007258330583965688,
      "grad_norm": 0.26637139916419983,
      "learning_rate": 8e-05,
      "loss": 1.7896,
      "step": 33
    },
    {
      "epoch": 0.007478279995601011,
      "grad_norm": 0.2802716791629791,
      "learning_rate": 8e-05,
      "loss": 1.9534,
      "step": 34
    },
    {
      "epoch": 0.007698229407236336,
      "grad_norm": 0.35580113530158997,
      "learning_rate": 8e-05,
      "loss": 1.8236,
      "step": 35
    },
    {
      "epoch": 0.00791817881887166,
      "grad_norm": 0.2794848382472992,
      "learning_rate": 8e-05,
      "loss": 1.8089,
      "step": 36
    },
    {
      "epoch": 0.008138128230506983,
      "grad_norm": 0.27942711114883423,
      "learning_rate": 8e-05,
      "loss": 1.7725,
      "step": 37
    },
    {
      "epoch": 0.008358077642142307,
      "grad_norm": 0.2882610857486725,
      "learning_rate": 8e-05,
      "loss": 1.8279,
      "step": 38
    },
    {
      "epoch": 0.008578027053777632,
      "grad_norm": 0.29375842213630676,
      "learning_rate": 8e-05,
      "loss": 2.0123,
      "step": 39
    },
    {
      "epoch": 0.008797976465412955,
      "grad_norm": 0.26120567321777344,
      "learning_rate": 8e-05,
      "loss": 1.6731,
      "step": 40
    },
    {
      "epoch": 0.009017925877048279,
      "grad_norm": 0.25272971391677856,
      "learning_rate": 8e-05,
      "loss": 1.5723,
      "step": 41
    },
    {
      "epoch": 0.009237875288683603,
      "grad_norm": 0.30548569560050964,
      "learning_rate": 8e-05,
      "loss": 1.9269,
      "step": 42
    },
    {
      "epoch": 0.009457824700318926,
      "grad_norm": 0.2714739441871643,
      "learning_rate": 8e-05,
      "loss": 1.6715,
      "step": 43
    },
    {
      "epoch": 0.00967777411195425,
      "grad_norm": 0.3086313009262085,
      "learning_rate": 8e-05,
      "loss": 1.8903,
      "step": 44
    },
    {
      "epoch": 0.009897723523589575,
      "grad_norm": 0.28676554560661316,
      "learning_rate": 8e-05,
      "loss": 1.8257,
      "step": 45
    },
    {
      "epoch": 0.010117672935224899,
      "grad_norm": 0.2898331880569458,
      "learning_rate": 8e-05,
      "loss": 1.822,
      "step": 46
    },
    {
      "epoch": 0.010337622346860222,
      "grad_norm": 0.2887754440307617,
      "learning_rate": 8e-05,
      "loss": 1.7629,
      "step": 47
    },
    {
      "epoch": 0.010557571758495546,
      "grad_norm": 0.28026437759399414,
      "learning_rate": 8e-05,
      "loss": 1.8874,
      "step": 48
    },
    {
      "epoch": 0.01077752117013087,
      "grad_norm": 0.29256439208984375,
      "learning_rate": 8e-05,
      "loss": 1.9169,
      "step": 49
    },
    {
      "epoch": 0.010997470581766193,
      "grad_norm": 0.29388460516929626,
      "learning_rate": 8e-05,
      "loss": 1.8341,
      "step": 50
    },
    {
      "epoch": 0.011217419993401518,
      "grad_norm": 0.29456326365470886,
      "learning_rate": 8e-05,
      "loss": 1.7088,
      "step": 51
    },
    {
      "epoch": 0.011437369405036842,
      "grad_norm": 0.2810533046722412,
      "learning_rate": 8e-05,
      "loss": 1.8564,
      "step": 52
    },
    {
      "epoch": 0.011657318816672166,
      "grad_norm": 0.3049224019050598,
      "learning_rate": 8e-05,
      "loss": 2.0114,
      "step": 53
    },
    {
      "epoch": 0.011877268228307489,
      "grad_norm": 0.347817987203598,
      "learning_rate": 8e-05,
      "loss": 1.6936,
      "step": 54
    },
    {
      "epoch": 0.012097217639942813,
      "grad_norm": 0.28999242186546326,
      "learning_rate": 8e-05,
      "loss": 1.5852,
      "step": 55
    },
    {
      "epoch": 0.012317167051578136,
      "grad_norm": 0.32856103777885437,
      "learning_rate": 8e-05,
      "loss": 1.8249,
      "step": 56
    },
    {
      "epoch": 0.012537116463213462,
      "grad_norm": 0.3450610339641571,
      "learning_rate": 8e-05,
      "loss": 1.9309,
      "step": 57
    },
    {
      "epoch": 0.012757065874848785,
      "grad_norm": 0.27445971965789795,
      "learning_rate": 8e-05,
      "loss": 1.8153,
      "step": 58
    },
    {
      "epoch": 0.012977015286484109,
      "grad_norm": 0.28595077991485596,
      "learning_rate": 8e-05,
      "loss": 1.8061,
      "step": 59
    },
    {
      "epoch": 0.013196964698119432,
      "grad_norm": 0.2909082770347595,
      "learning_rate": 8e-05,
      "loss": 1.7793,
      "step": 60
    },
    {
      "epoch": 0.013416914109754756,
      "grad_norm": 0.28822049498558044,
      "learning_rate": 8e-05,
      "loss": 1.7218,
      "step": 61
    },
    {
      "epoch": 0.01363686352139008,
      "grad_norm": 0.29159948229789734,
      "learning_rate": 8e-05,
      "loss": 1.7985,
      "step": 62
    },
    {
      "epoch": 0.013856812933025405,
      "grad_norm": 0.29802417755126953,
      "learning_rate": 8e-05,
      "loss": 1.8903,
      "step": 63
    },
    {
      "epoch": 0.014076762344660728,
      "grad_norm": 0.29128944873809814,
      "learning_rate": 8e-05,
      "loss": 1.7928,
      "step": 64
    },
    {
      "epoch": 0.014296711756296052,
      "grad_norm": 0.3093227446079254,
      "learning_rate": 8e-05,
      "loss": 1.8409,
      "step": 65
    },
    {
      "epoch": 0.014516661167931376,
      "grad_norm": 0.2688956558704376,
      "learning_rate": 8e-05,
      "loss": 1.6136,
      "step": 66
    },
    {
      "epoch": 0.0147366105795667,
      "grad_norm": 0.316579133272171,
      "learning_rate": 8e-05,
      "loss": 1.8153,
      "step": 67
    },
    {
      "epoch": 0.014956559991202023,
      "grad_norm": 0.30234795808792114,
      "learning_rate": 8e-05,
      "loss": 1.7311,
      "step": 68
    },
    {
      "epoch": 0.015176509402837348,
      "grad_norm": 0.2790556848049164,
      "learning_rate": 8e-05,
      "loss": 1.7714,
      "step": 69
    },
    {
      "epoch": 0.015396458814472672,
      "grad_norm": 0.29012972116470337,
      "learning_rate": 8e-05,
      "loss": 1.7528,
      "step": 70
    },
    {
      "epoch": 0.015616408226107995,
      "grad_norm": 0.28507527709007263,
      "learning_rate": 8e-05,
      "loss": 1.6452,
      "step": 71
    },
    {
      "epoch": 0.01583635763774332,
      "grad_norm": 0.28862133622169495,
      "learning_rate": 8e-05,
      "loss": 1.7473,
      "step": 72
    },
    {
      "epoch": 0.016056307049378642,
      "grad_norm": 0.2726048231124878,
      "learning_rate": 8e-05,
      "loss": 1.7519,
      "step": 73
    },
    {
      "epoch": 0.016276256461013966,
      "grad_norm": 0.26808786392211914,
      "learning_rate": 8e-05,
      "loss": 1.6332,
      "step": 74
    },
    {
      "epoch": 0.01649620587264929,
      "grad_norm": 0.32144519686698914,
      "learning_rate": 8e-05,
      "loss": 1.7115,
      "step": 75
    },
    {
      "epoch": 0.016716155284284613,
      "grad_norm": 0.26930421590805054,
      "learning_rate": 8e-05,
      "loss": 1.7854,
      "step": 76
    },
    {
      "epoch": 0.016936104695919937,
      "grad_norm": 0.29462486505508423,
      "learning_rate": 8e-05,
      "loss": 1.6919,
      "step": 77
    },
    {
      "epoch": 0.017156054107555264,
      "grad_norm": 0.2780003249645233,
      "learning_rate": 8e-05,
      "loss": 1.6355,
      "step": 78
    },
    {
      "epoch": 0.017376003519190587,
      "grad_norm": 0.29219016432762146,
      "learning_rate": 8e-05,
      "loss": 1.883,
      "step": 79
    },
    {
      "epoch": 0.01759595293082591,
      "grad_norm": 0.2893241047859192,
      "learning_rate": 8e-05,
      "loss": 1.8548,
      "step": 80
    },
    {
      "epoch": 0.017815902342461234,
      "grad_norm": 0.283512145280838,
      "learning_rate": 8e-05,
      "loss": 1.79,
      "step": 81
    },
    {
      "epoch": 0.018035851754096558,
      "grad_norm": 0.2679024040699005,
      "learning_rate": 8e-05,
      "loss": 1.5866,
      "step": 82
    },
    {
      "epoch": 0.01825580116573188,
      "grad_norm": 0.2892123758792877,
      "learning_rate": 8e-05,
      "loss": 1.9033,
      "step": 83
    },
    {
      "epoch": 0.018475750577367205,
      "grad_norm": 0.2680201530456543,
      "learning_rate": 8e-05,
      "loss": 1.8557,
      "step": 84
    },
    {
      "epoch": 0.01869569998900253,
      "grad_norm": 0.30922645330429077,
      "learning_rate": 8e-05,
      "loss": 1.8885,
      "step": 85
    },
    {
      "epoch": 0.018915649400637852,
      "grad_norm": 0.2735271751880646,
      "learning_rate": 8e-05,
      "loss": 1.8765,
      "step": 86
    },
    {
      "epoch": 0.019135598812273176,
      "grad_norm": 0.28639712929725647,
      "learning_rate": 8e-05,
      "loss": 1.9429,
      "step": 87
    },
    {
      "epoch": 0.0193555482239085,
      "grad_norm": 0.28437235951423645,
      "learning_rate": 8e-05,
      "loss": 1.8405,
      "step": 88
    },
    {
      "epoch": 0.019575497635543827,
      "grad_norm": 0.276517778635025,
      "learning_rate": 8e-05,
      "loss": 1.7496,
      "step": 89
    },
    {
      "epoch": 0.01979544704717915,
      "grad_norm": 0.273404598236084,
      "learning_rate": 8e-05,
      "loss": 1.704,
      "step": 90
    },
    {
      "epoch": 0.020015396458814474,
      "grad_norm": 0.2707740366458893,
      "learning_rate": 8e-05,
      "loss": 1.8274,
      "step": 91
    },
    {
      "epoch": 0.020235345870449797,
      "grad_norm": 0.26880595088005066,
      "learning_rate": 8e-05,
      "loss": 1.7695,
      "step": 92
    },
    {
      "epoch": 0.02045529528208512,
      "grad_norm": 0.28712528944015503,
      "learning_rate": 8e-05,
      "loss": 1.9436,
      "step": 93
    },
    {
      "epoch": 0.020675244693720445,
      "grad_norm": 0.26633599400520325,
      "learning_rate": 8e-05,
      "loss": 1.7877,
      "step": 94
    },
    {
      "epoch": 0.020895194105355768,
      "grad_norm": 0.2843431532382965,
      "learning_rate": 8e-05,
      "loss": 1.8389,
      "step": 95
    },
    {
      "epoch": 0.02111514351699109,
      "grad_norm": 0.2597465515136719,
      "learning_rate": 8e-05,
      "loss": 1.7047,
      "step": 96
    },
    {
      "epoch": 0.021335092928626415,
      "grad_norm": 0.2804902493953705,
      "learning_rate": 8e-05,
      "loss": 1.9375,
      "step": 97
    },
    {
      "epoch": 0.02155504234026174,
      "grad_norm": 0.2825285792350769,
      "learning_rate": 8e-05,
      "loss": 1.8348,
      "step": 98
    },
    {
      "epoch": 0.021774991751897062,
      "grad_norm": 0.26459112763404846,
      "learning_rate": 8e-05,
      "loss": 1.7416,
      "step": 99
    },
    {
      "epoch": 0.021994941163532386,
      "grad_norm": 0.28523096442222595,
      "learning_rate": 8e-05,
      "loss": 1.9202,
      "step": 100
    },
    {
      "epoch": 0.022214890575167713,
      "grad_norm": 0.2679818570613861,
      "learning_rate": 8e-05,
      "loss": 1.6741,
      "step": 101
    },
    {
      "epoch": 0.022434839986803037,
      "grad_norm": 0.2798464894294739,
      "learning_rate": 8e-05,
      "loss": 1.6622,
      "step": 102
    },
    {
      "epoch": 0.02265478939843836,
      "grad_norm": 0.2826269567012787,
      "learning_rate": 8e-05,
      "loss": 1.7577,
      "step": 103
    },
    {
      "epoch": 0.022874738810073684,
      "grad_norm": 0.3859495222568512,
      "learning_rate": 8e-05,
      "loss": 1.9705,
      "step": 104
    },
    {
      "epoch": 0.023094688221709007,
      "grad_norm": 0.2766650319099426,
      "learning_rate": 8e-05,
      "loss": 1.7706,
      "step": 105
    },
    {
      "epoch": 0.02331463763334433,
      "grad_norm": 0.2804067134857178,
      "learning_rate": 8e-05,
      "loss": 1.8007,
      "step": 106
    },
    {
      "epoch": 0.023534587044979655,
      "grad_norm": 0.27818629145622253,
      "learning_rate": 8e-05,
      "loss": 1.7913,
      "step": 107
    },
    {
      "epoch": 0.023754536456614978,
      "grad_norm": 0.2697458267211914,
      "learning_rate": 8e-05,
      "loss": 1.8458,
      "step": 108
    },
    {
      "epoch": 0.023974485868250302,
      "grad_norm": 0.28805410861968994,
      "learning_rate": 8e-05,
      "loss": 1.7543,
      "step": 109
    },
    {
      "epoch": 0.024194435279885625,
      "grad_norm": 0.28452396392822266,
      "learning_rate": 8e-05,
      "loss": 1.8499,
      "step": 110
    },
    {
      "epoch": 0.02441438469152095,
      "grad_norm": 0.2837978005409241,
      "learning_rate": 8e-05,
      "loss": 1.797,
      "step": 111
    },
    {
      "epoch": 0.024634334103156273,
      "grad_norm": 0.2965853810310364,
      "learning_rate": 8e-05,
      "loss": 1.7988,
      "step": 112
    },
    {
      "epoch": 0.0248542835147916,
      "grad_norm": 0.28529393672943115,
      "learning_rate": 8e-05,
      "loss": 1.7886,
      "step": 113
    },
    {
      "epoch": 0.025074232926426923,
      "grad_norm": 0.285199910402298,
      "learning_rate": 8e-05,
      "loss": 1.9112,
      "step": 114
    },
    {
      "epoch": 0.025294182338062247,
      "grad_norm": 0.286316454410553,
      "learning_rate": 8e-05,
      "loss": 1.6735,
      "step": 115
    },
    {
      "epoch": 0.02551413174969757,
      "grad_norm": 0.2648874819278717,
      "learning_rate": 8e-05,
      "loss": 1.5333,
      "step": 116
    },
    {
      "epoch": 0.025734081161332894,
      "grad_norm": 0.2834017276763916,
      "learning_rate": 8e-05,
      "loss": 1.7524,
      "step": 117
    },
    {
      "epoch": 0.025954030572968217,
      "grad_norm": 0.27846938371658325,
      "learning_rate": 8e-05,
      "loss": 1.8448,
      "step": 118
    },
    {
      "epoch": 0.02617397998460354,
      "grad_norm": 0.3278025984764099,
      "learning_rate": 8e-05,
      "loss": 1.9158,
      "step": 119
    },
    {
      "epoch": 0.026393929396238865,
      "grad_norm": 0.30259498953819275,
      "learning_rate": 8e-05,
      "loss": 1.7897,
      "step": 120
    },
    {
      "epoch": 0.026613878807874188,
      "grad_norm": 0.27566099166870117,
      "learning_rate": 8e-05,
      "loss": 1.682,
      "step": 121
    },
    {
      "epoch": 0.026833828219509512,
      "grad_norm": 0.2959173321723938,
      "learning_rate": 8e-05,
      "loss": 1.9032,
      "step": 122
    },
    {
      "epoch": 0.027053777631144835,
      "grad_norm": 0.29449525475502014,
      "learning_rate": 8e-05,
      "loss": 1.6174,
      "step": 123
    },
    {
      "epoch": 0.02727372704278016,
      "grad_norm": 0.3012568950653076,
      "learning_rate": 8e-05,
      "loss": 1.6817,
      "step": 124
    },
    {
      "epoch": 0.027493676454415486,
      "grad_norm": 0.29086676239967346,
      "learning_rate": 8e-05,
      "loss": 1.833,
      "step": 125
    },
    {
      "epoch": 0.02771362586605081,
      "grad_norm": 0.2756067216396332,
      "learning_rate": 8e-05,
      "loss": 1.7807,
      "step": 126
    },
    {
      "epoch": 0.027933575277686133,
      "grad_norm": 0.3420695662498474,
      "learning_rate": 8e-05,
      "loss": 1.8652,
      "step": 127
    },
    {
      "epoch": 0.028153524689321457,
      "grad_norm": 0.2899749279022217,
      "learning_rate": 8e-05,
      "loss": 1.7199,
      "step": 128
    },
    {
      "epoch": 0.02837347410095678,
      "grad_norm": 0.274718701839447,
      "learning_rate": 8e-05,
      "loss": 1.7322,
      "step": 129
    },
    {
      "epoch": 0.028593423512592104,
      "grad_norm": 0.3784034848213196,
      "learning_rate": 8e-05,
      "loss": 1.8917,
      "step": 130
    },
    {
      "epoch": 0.028813372924227428,
      "grad_norm": 0.2814437448978424,
      "learning_rate": 8e-05,
      "loss": 1.726,
      "step": 131
    },
    {
      "epoch": 0.02903332233586275,
      "grad_norm": 0.287701815366745,
      "learning_rate": 8e-05,
      "loss": 1.8166,
      "step": 132
    },
    {
      "epoch": 0.029253271747498075,
      "grad_norm": 0.28487101197242737,
      "learning_rate": 8e-05,
      "loss": 1.7183,
      "step": 133
    },
    {
      "epoch": 0.0294732211591334,
      "grad_norm": 0.27141597867012024,
      "learning_rate": 8e-05,
      "loss": 1.7436,
      "step": 134
    },
    {
      "epoch": 0.029693170570768722,
      "grad_norm": 0.2708652913570404,
      "learning_rate": 8e-05,
      "loss": 1.8116,
      "step": 135
    },
    {
      "epoch": 0.029913119982404045,
      "grad_norm": 0.2789991796016693,
      "learning_rate": 8e-05,
      "loss": 1.7942,
      "step": 136
    },
    {
      "epoch": 0.030133069394039372,
      "grad_norm": 0.3053725063800812,
      "learning_rate": 8e-05,
      "loss": 1.8508,
      "step": 137
    },
    {
      "epoch": 0.030353018805674696,
      "grad_norm": 0.30432772636413574,
      "learning_rate": 8e-05,
      "loss": 1.8129,
      "step": 138
    },
    {
      "epoch": 0.03057296821731002,
      "grad_norm": 0.2873070240020752,
      "learning_rate": 8e-05,
      "loss": 1.8713,
      "step": 139
    },
    {
      "epoch": 0.030792917628945343,
      "grad_norm": 0.2777135968208313,
      "learning_rate": 8e-05,
      "loss": 1.7065,
      "step": 140
    },
    {
      "epoch": 0.031012867040580667,
      "grad_norm": 0.29774004220962524,
      "learning_rate": 8e-05,
      "loss": 1.6471,
      "step": 141
    },
    {
      "epoch": 0.03123281645221599,
      "grad_norm": 0.2803782522678375,
      "learning_rate": 8e-05,
      "loss": 1.6992,
      "step": 142
    },
    {
      "epoch": 0.03145276586385132,
      "grad_norm": 0.2777007818222046,
      "learning_rate": 8e-05,
      "loss": 1.8398,
      "step": 143
    },
    {
      "epoch": 0.03167271527548664,
      "grad_norm": 0.26938894391059875,
      "learning_rate": 8e-05,
      "loss": 1.6082,
      "step": 144
    },
    {
      "epoch": 0.031892664687121965,
      "grad_norm": 0.2934747338294983,
      "learning_rate": 8e-05,
      "loss": 1.6929,
      "step": 145
    },
    {
      "epoch": 0.032112614098757285,
      "grad_norm": 0.2687772214412689,
      "learning_rate": 8e-05,
      "loss": 1.6472,
      "step": 146
    },
    {
      "epoch": 0.03233256351039261,
      "grad_norm": 0.2758256793022156,
      "learning_rate": 8e-05,
      "loss": 1.7128,
      "step": 147
    },
    {
      "epoch": 0.03255251292202793,
      "grad_norm": 0.26065707206726074,
      "learning_rate": 8e-05,
      "loss": 1.7108,
      "step": 148
    },
    {
      "epoch": 0.03277246233366326,
      "grad_norm": 0.31668898463249207,
      "learning_rate": 8e-05,
      "loss": 1.9365,
      "step": 149
    },
    {
      "epoch": 0.03299241174529858,
      "grad_norm": 0.2915947437286377,
      "learning_rate": 8e-05,
      "loss": 1.855,
      "step": 150
    },
    {
      "epoch": 0.033212361156933906,
      "grad_norm": 0.2741534113883972,
      "learning_rate": 8e-05,
      "loss": 1.735,
      "step": 151
    },
    {
      "epoch": 0.033432310568569226,
      "grad_norm": 0.300800085067749,
      "learning_rate": 8e-05,
      "loss": 1.7161,
      "step": 152
    },
    {
      "epoch": 0.03365225998020455,
      "grad_norm": 0.26691076159477234,
      "learning_rate": 8e-05,
      "loss": 1.665,
      "step": 153
    },
    {
      "epoch": 0.03387220939183987,
      "grad_norm": 0.2605098485946655,
      "learning_rate": 8e-05,
      "loss": 1.7288,
      "step": 154
    },
    {
      "epoch": 0.0340921588034752,
      "grad_norm": 0.2728619873523712,
      "learning_rate": 8e-05,
      "loss": 1.7237,
      "step": 155
    },
    {
      "epoch": 0.03431210821511053,
      "grad_norm": 0.29627877473831177,
      "learning_rate": 8e-05,
      "loss": 1.8024,
      "step": 156
    },
    {
      "epoch": 0.03453205762674585,
      "grad_norm": 0.27106964588165283,
      "learning_rate": 8e-05,
      "loss": 1.8166,
      "step": 157
    },
    {
      "epoch": 0.034752007038381175,
      "grad_norm": 0.26806893944740295,
      "learning_rate": 8e-05,
      "loss": 1.7061,
      "step": 158
    },
    {
      "epoch": 0.034971956450016495,
      "grad_norm": 0.2509767413139343,
      "learning_rate": 8e-05,
      "loss": 1.6897,
      "step": 159
    },
    {
      "epoch": 0.03519190586165182,
      "grad_norm": 0.34342750906944275,
      "learning_rate": 8e-05,
      "loss": 1.7151,
      "step": 160
    },
    {
      "epoch": 0.03541185527328714,
      "grad_norm": 0.27948594093322754,
      "learning_rate": 8e-05,
      "loss": 1.6574,
      "step": 161
    },
    {
      "epoch": 0.03563180468492247,
      "grad_norm": 0.28651687502861023,
      "learning_rate": 8e-05,
      "loss": 1.839,
      "step": 162
    },
    {
      "epoch": 0.03585175409655779,
      "grad_norm": 0.2787701189517975,
      "learning_rate": 8e-05,
      "loss": 1.8146,
      "step": 163
    },
    {
      "epoch": 0.036071703508193116,
      "grad_norm": 0.2596721351146698,
      "learning_rate": 8e-05,
      "loss": 1.6088,
      "step": 164
    },
    {
      "epoch": 0.036291652919828436,
      "grad_norm": 0.2630285322666168,
      "learning_rate": 8e-05,
      "loss": 1.6941,
      "step": 165
    },
    {
      "epoch": 0.03651160233146376,
      "grad_norm": 0.30072465538978577,
      "learning_rate": 8e-05,
      "loss": 1.8684,
      "step": 166
    },
    {
      "epoch": 0.03673155174309909,
      "grad_norm": 0.2789234519004822,
      "learning_rate": 8e-05,
      "loss": 1.9136,
      "step": 167
    },
    {
      "epoch": 0.03695150115473441,
      "grad_norm": 0.25597283244132996,
      "learning_rate": 8e-05,
      "loss": 1.669,
      "step": 168
    },
    {
      "epoch": 0.03717145056636974,
      "grad_norm": 0.30354219675064087,
      "learning_rate": 8e-05,
      "loss": 1.7845,
      "step": 169
    },
    {
      "epoch": 0.03739139997800506,
      "grad_norm": 0.26998043060302734,
      "learning_rate": 8e-05,
      "loss": 1.6626,
      "step": 170
    },
    {
      "epoch": 0.037611349389640385,
      "grad_norm": 0.27418825030326843,
      "learning_rate": 8e-05,
      "loss": 1.6444,
      "step": 171
    },
    {
      "epoch": 0.037831298801275705,
      "grad_norm": 0.2858507037162781,
      "learning_rate": 8e-05,
      "loss": 1.8584,
      "step": 172
    },
    {
      "epoch": 0.03805124821291103,
      "grad_norm": 0.26513633131980896,
      "learning_rate": 8e-05,
      "loss": 1.7107,
      "step": 173
    },
    {
      "epoch": 0.03827119762454635,
      "grad_norm": 0.3162567913532257,
      "learning_rate": 8e-05,
      "loss": 1.7153,
      "step": 174
    },
    {
      "epoch": 0.03849114703618168,
      "grad_norm": 0.28961601853370667,
      "learning_rate": 8e-05,
      "loss": 1.8455,
      "step": 175
    },
    {
      "epoch": 0.038711096447817,
      "grad_norm": 0.29676249623298645,
      "learning_rate": 8e-05,
      "loss": 1.9303,
      "step": 176
    },
    {
      "epoch": 0.038931045859452326,
      "grad_norm": 0.2863664925098419,
      "learning_rate": 8e-05,
      "loss": 1.6975,
      "step": 177
    },
    {
      "epoch": 0.03915099527108765,
      "grad_norm": 0.2715422213077545,
      "learning_rate": 8e-05,
      "loss": 1.5472,
      "step": 178
    },
    {
      "epoch": 0.03937094468272297,
      "grad_norm": 0.2740415036678314,
      "learning_rate": 8e-05,
      "loss": 1.7113,
      "step": 179
    },
    {
      "epoch": 0.0395908940943583,
      "grad_norm": 0.29612302780151367,
      "learning_rate": 8e-05,
      "loss": 1.8689,
      "step": 180
    },
    {
      "epoch": 0.03981084350599362,
      "grad_norm": 0.26745903491973877,
      "learning_rate": 8e-05,
      "loss": 1.6076,
      "step": 181
    },
    {
      "epoch": 0.04003079291762895,
      "grad_norm": 0.296695739030838,
      "learning_rate": 8e-05,
      "loss": 1.846,
      "step": 182
    },
    {
      "epoch": 0.04025074232926427,
      "grad_norm": 0.27626705169677734,
      "learning_rate": 8e-05,
      "loss": 1.8103,
      "step": 183
    },
    {
      "epoch": 0.040470691740899595,
      "grad_norm": 0.2597677409648895,
      "learning_rate": 8e-05,
      "loss": 1.6432,
      "step": 184
    },
    {
      "epoch": 0.040690641152534915,
      "grad_norm": 0.2738899290561676,
      "learning_rate": 8e-05,
      "loss": 1.8351,
      "step": 185
    },
    {
      "epoch": 0.04091059056417024,
      "grad_norm": 0.2683742344379425,
      "learning_rate": 8e-05,
      "loss": 1.6453,
      "step": 186
    },
    {
      "epoch": 0.04113053997580556,
      "grad_norm": 0.28722816705703735,
      "learning_rate": 8e-05,
      "loss": 1.7685,
      "step": 187
    },
    {
      "epoch": 0.04135048938744089,
      "grad_norm": 0.2851015627384186,
      "learning_rate": 8e-05,
      "loss": 1.8464,
      "step": 188
    },
    {
      "epoch": 0.04157043879907621,
      "grad_norm": 0.2630920112133026,
      "learning_rate": 8e-05,
      "loss": 1.7176,
      "step": 189
    },
    {
      "epoch": 0.041790388210711536,
      "grad_norm": 0.2678779661655426,
      "learning_rate": 8e-05,
      "loss": 1.671,
      "step": 190
    },
    {
      "epoch": 0.04201033762234686,
      "grad_norm": 0.27810946106910706,
      "learning_rate": 8e-05,
      "loss": 1.6467,
      "step": 191
    },
    {
      "epoch": 0.04223028703398218,
      "grad_norm": 0.2831014394760132,
      "learning_rate": 8e-05,
      "loss": 1.8784,
      "step": 192
    },
    {
      "epoch": 0.04245023644561751,
      "grad_norm": 0.2643384635448456,
      "learning_rate": 8e-05,
      "loss": 1.6239,
      "step": 193
    },
    {
      "epoch": 0.04267018585725283,
      "grad_norm": 0.27143070101737976,
      "learning_rate": 8e-05,
      "loss": 1.8012,
      "step": 194
    },
    {
      "epoch": 0.04289013526888816,
      "grad_norm": 0.28524088859558105,
      "learning_rate": 8e-05,
      "loss": 1.7534,
      "step": 195
    },
    {
      "epoch": 0.04311008468052348,
      "grad_norm": 0.27226153016090393,
      "learning_rate": 8e-05,
      "loss": 1.847,
      "step": 196
    },
    {
      "epoch": 0.043330034092158805,
      "grad_norm": 0.27042534947395325,
      "learning_rate": 8e-05,
      "loss": 1.698,
      "step": 197
    },
    {
      "epoch": 0.043549983503794125,
      "grad_norm": 0.2673223912715912,
      "learning_rate": 8e-05,
      "loss": 1.7825,
      "step": 198
    },
    {
      "epoch": 0.04376993291542945,
      "grad_norm": 0.26485180854797363,
      "learning_rate": 8e-05,
      "loss": 1.7755,
      "step": 199
    },
    {
      "epoch": 0.04398988232706477,
      "grad_norm": 0.26945164799690247,
      "learning_rate": 8e-05,
      "loss": 1.8612,
      "step": 200
    },
    {
      "epoch": 0.0442098317387001,
      "grad_norm": 0.30337756872177124,
      "learning_rate": 8e-05,
      "loss": 1.8556,
      "step": 201
    },
    {
      "epoch": 0.044429781150335426,
      "grad_norm": 0.26593855023384094,
      "learning_rate": 8e-05,
      "loss": 1.7633,
      "step": 202
    },
    {
      "epoch": 0.044649730561970746,
      "grad_norm": 0.26703208684921265,
      "learning_rate": 8e-05,
      "loss": 1.7787,
      "step": 203
    },
    {
      "epoch": 0.04486967997360607,
      "grad_norm": 0.2799319922924042,
      "learning_rate": 8e-05,
      "loss": 1.8946,
      "step": 204
    },
    {
      "epoch": 0.04508962938524139,
      "grad_norm": 0.261406809091568,
      "learning_rate": 8e-05,
      "loss": 1.714,
      "step": 205
    },
    {
      "epoch": 0.04530957879687672,
      "grad_norm": 0.30923140048980713,
      "learning_rate": 8e-05,
      "loss": 1.9953,
      "step": 206
    },
    {
      "epoch": 0.04552952820851204,
      "grad_norm": 0.28189903497695923,
      "learning_rate": 8e-05,
      "loss": 1.8068,
      "step": 207
    },
    {
      "epoch": 0.04574947762014737,
      "grad_norm": 0.28659504652023315,
      "learning_rate": 8e-05,
      "loss": 1.7961,
      "step": 208
    },
    {
      "epoch": 0.04596942703178269,
      "grad_norm": 0.27828094363212585,
      "learning_rate": 8e-05,
      "loss": 1.6398,
      "step": 209
    },
    {
      "epoch": 0.046189376443418015,
      "grad_norm": 0.2826248109340668,
      "learning_rate": 8e-05,
      "loss": 1.8442,
      "step": 210
    },
    {
      "epoch": 0.046409325855053335,
      "grad_norm": 0.2596709430217743,
      "learning_rate": 8e-05,
      "loss": 1.7269,
      "step": 211
    },
    {
      "epoch": 0.04662927526668866,
      "grad_norm": 0.26883357763290405,
      "learning_rate": 8e-05,
      "loss": 1.7396,
      "step": 212
    },
    {
      "epoch": 0.04684922467832398,
      "grad_norm": 0.2834852933883667,
      "learning_rate": 8e-05,
      "loss": 1.6992,
      "step": 213
    },
    {
      "epoch": 0.04706917408995931,
      "grad_norm": 0.30232125520706177,
      "learning_rate": 8e-05,
      "loss": 1.8216,
      "step": 214
    },
    {
      "epoch": 0.047289123501594636,
      "grad_norm": 0.2887151539325714,
      "learning_rate": 8e-05,
      "loss": 1.5633,
      "step": 215
    },
    {
      "epoch": 0.047509072913229956,
      "grad_norm": 0.27171874046325684,
      "learning_rate": 8e-05,
      "loss": 1.8272,
      "step": 216
    },
    {
      "epoch": 0.04772902232486528,
      "grad_norm": 0.35441088676452637,
      "learning_rate": 8e-05,
      "loss": 1.8308,
      "step": 217
    },
    {
      "epoch": 0.047948971736500604,
      "grad_norm": 0.28351160883903503,
      "learning_rate": 8e-05,
      "loss": 1.8697,
      "step": 218
    },
    {
      "epoch": 0.04816892114813593,
      "grad_norm": 0.26361364126205444,
      "learning_rate": 8e-05,
      "loss": 1.7044,
      "step": 219
    },
    {
      "epoch": 0.04838887055977125,
      "grad_norm": 0.2720041871070862,
      "learning_rate": 8e-05,
      "loss": 1.7718,
      "step": 220
    },
    {
      "epoch": 0.04860881997140658,
      "grad_norm": 0.28131023049354553,
      "learning_rate": 8e-05,
      "loss": 1.8066,
      "step": 221
    },
    {
      "epoch": 0.0488287693830419,
      "grad_norm": 0.2640543580055237,
      "learning_rate": 8e-05,
      "loss": 1.69,
      "step": 222
    },
    {
      "epoch": 0.049048718794677225,
      "grad_norm": 0.26101046800613403,
      "learning_rate": 8e-05,
      "loss": 1.6372,
      "step": 223
    },
    {
      "epoch": 0.049268668206312545,
      "grad_norm": 0.3021651804447174,
      "learning_rate": 8e-05,
      "loss": 1.8528,
      "step": 224
    },
    {
      "epoch": 0.04948861761794787,
      "grad_norm": 0.2655261158943176,
      "learning_rate": 8e-05,
      "loss": 1.7406,
      "step": 225
    },
    {
      "epoch": 0.0497085670295832,
      "grad_norm": 0.2873914837837219,
      "learning_rate": 8e-05,
      "loss": 1.7643,
      "step": 226
    },
    {
      "epoch": 0.04992851644121852,
      "grad_norm": 0.31813880801200867,
      "learning_rate": 8e-05,
      "loss": 1.8645,
      "step": 227
    },
    {
      "epoch": 0.050148465852853846,
      "grad_norm": 0.2996014654636383,
      "learning_rate": 8e-05,
      "loss": 1.6685,
      "step": 228
    },
    {
      "epoch": 0.050368415264489166,
      "grad_norm": 0.2837509512901306,
      "learning_rate": 8e-05,
      "loss": 1.9227,
      "step": 229
    },
    {
      "epoch": 0.05058836467612449,
      "grad_norm": 0.29532885551452637,
      "learning_rate": 8e-05,
      "loss": 1.9073,
      "step": 230
    },
    {
      "epoch": 0.050808314087759814,
      "grad_norm": 0.285295307636261,
      "learning_rate": 8e-05,
      "loss": 1.8248,
      "step": 231
    },
    {
      "epoch": 0.05102826349939514,
      "grad_norm": 0.26331770420074463,
      "learning_rate": 8e-05,
      "loss": 1.7146,
      "step": 232
    },
    {
      "epoch": 0.05124821291103046,
      "grad_norm": 0.24956567585468292,
      "learning_rate": 8e-05,
      "loss": 1.5574,
      "step": 233
    },
    {
      "epoch": 0.05146816232266579,
      "grad_norm": 0.27515965700149536,
      "learning_rate": 8e-05,
      "loss": 1.7854,
      "step": 234
    },
    {
      "epoch": 0.05168811173430111,
      "grad_norm": 0.28268730640411377,
      "learning_rate": 8e-05,
      "loss": 1.8294,
      "step": 235
    },
    {
      "epoch": 0.051908061145936435,
      "grad_norm": 0.25420427322387695,
      "learning_rate": 8e-05,
      "loss": 1.6735,
      "step": 236
    },
    {
      "epoch": 0.052128010557571755,
      "grad_norm": 0.2869463860988617,
      "learning_rate": 8e-05,
      "loss": 1.808,
      "step": 237
    },
    {
      "epoch": 0.05234795996920708,
      "grad_norm": 0.2574792206287384,
      "learning_rate": 8e-05,
      "loss": 1.7563,
      "step": 238
    },
    {
      "epoch": 0.05256790938084241,
      "grad_norm": 0.26652273535728455,
      "learning_rate": 8e-05,
      "loss": 1.743,
      "step": 239
    },
    {
      "epoch": 0.05278785879247773,
      "grad_norm": 0.2956235408782959,
      "learning_rate": 8e-05,
      "loss": 1.9169,
      "step": 240
    },
    {
      "epoch": 0.053007808204113056,
      "grad_norm": 0.274142861366272,
      "learning_rate": 8e-05,
      "loss": 1.8321,
      "step": 241
    },
    {
      "epoch": 0.053227757615748376,
      "grad_norm": 0.27525436878204346,
      "learning_rate": 8e-05,
      "loss": 1.8206,
      "step": 242
    },
    {
      "epoch": 0.053447707027383703,
      "grad_norm": 0.26323091983795166,
      "learning_rate": 8e-05,
      "loss": 1.7574,
      "step": 243
    },
    {
      "epoch": 0.053667656439019024,
      "grad_norm": 0.28554126620292664,
      "learning_rate": 8e-05,
      "loss": 1.9293,
      "step": 244
    },
    {
      "epoch": 0.05388760585065435,
      "grad_norm": 0.2651476562023163,
      "learning_rate": 8e-05,
      "loss": 1.808,
      "step": 245
    },
    {
      "epoch": 0.05410755526228967,
      "grad_norm": 0.27941837906837463,
      "learning_rate": 8e-05,
      "loss": 1.7838,
      "step": 246
    },
    {
      "epoch": 0.054327504673925,
      "grad_norm": 0.26575711369514465,
      "learning_rate": 8e-05,
      "loss": 1.6117,
      "step": 247
    },
    {
      "epoch": 0.05454745408556032,
      "grad_norm": 0.2620556354522705,
      "learning_rate": 8e-05,
      "loss": 1.7703,
      "step": 248
    },
    {
      "epoch": 0.054767403497195645,
      "grad_norm": 0.2782936990261078,
      "learning_rate": 8e-05,
      "loss": 1.753,
      "step": 249
    },
    {
      "epoch": 0.05498735290883097,
      "grad_norm": 0.28347843885421753,
      "learning_rate": 8e-05,
      "loss": 1.8365,
      "step": 250
    },
    {
      "epoch": 0.05520730232046629,
      "grad_norm": 0.2740314304828644,
      "learning_rate": 8e-05,
      "loss": 1.7448,
      "step": 251
    },
    {
      "epoch": 0.05542725173210162,
      "grad_norm": 0.2779199779033661,
      "learning_rate": 8e-05,
      "loss": 1.8025,
      "step": 252
    },
    {
      "epoch": 0.05564720114373694,
      "grad_norm": 0.27700838446617126,
      "learning_rate": 8e-05,
      "loss": 1.6368,
      "step": 253
    },
    {
      "epoch": 0.055867150555372266,
      "grad_norm": 0.2753797173500061,
      "learning_rate": 8e-05,
      "loss": 1.7058,
      "step": 254
    },
    {
      "epoch": 0.056087099967007586,
      "grad_norm": 0.2677604556083679,
      "learning_rate": 8e-05,
      "loss": 1.772,
      "step": 255
    },
    {
      "epoch": 0.056307049378642914,
      "grad_norm": 0.291358083486557,
      "learning_rate": 8e-05,
      "loss": 1.7229,
      "step": 256
    },
    {
      "epoch": 0.056526998790278234,
      "grad_norm": 0.2605611979961395,
      "learning_rate": 8e-05,
      "loss": 1.6654,
      "step": 257
    },
    {
      "epoch": 0.05674694820191356,
      "grad_norm": 0.2726796865463257,
      "learning_rate": 8e-05,
      "loss": 1.8524,
      "step": 258
    },
    {
      "epoch": 0.05696689761354888,
      "grad_norm": 0.2769307494163513,
      "learning_rate": 8e-05,
      "loss": 1.913,
      "step": 259
    },
    {
      "epoch": 0.05718684702518421,
      "grad_norm": 0.27163514494895935,
      "learning_rate": 8e-05,
      "loss": 1.7076,
      "step": 260
    },
    {
      "epoch": 0.057406796436819535,
      "grad_norm": 0.27037522196769714,
      "learning_rate": 8e-05,
      "loss": 1.7461,
      "step": 261
    },
    {
      "epoch": 0.057626745848454855,
      "grad_norm": 0.2570153772830963,
      "learning_rate": 8e-05,
      "loss": 1.6714,
      "step": 262
    },
    {
      "epoch": 0.05784669526009018,
      "grad_norm": 0.2802227735519409,
      "learning_rate": 8e-05,
      "loss": 1.6782,
      "step": 263
    },
    {
      "epoch": 0.0580666446717255,
      "grad_norm": 0.293969064950943,
      "learning_rate": 8e-05,
      "loss": 1.6253,
      "step": 264
    },
    {
      "epoch": 0.05828659408336083,
      "grad_norm": 0.28199446201324463,
      "learning_rate": 8e-05,
      "loss": 1.791,
      "step": 265
    },
    {
      "epoch": 0.05850654349499615,
      "grad_norm": 0.3037835657596588,
      "learning_rate": 8e-05,
      "loss": 1.8553,
      "step": 266
    },
    {
      "epoch": 0.058726492906631476,
      "grad_norm": 0.2814860939979553,
      "learning_rate": 8e-05,
      "loss": 1.7237,
      "step": 267
    },
    {
      "epoch": 0.0589464423182668,
      "grad_norm": 0.29769864678382874,
      "learning_rate": 8e-05,
      "loss": 1.8635,
      "step": 268
    },
    {
      "epoch": 0.059166391729902124,
      "grad_norm": 0.26650169491767883,
      "learning_rate": 8e-05,
      "loss": 1.8173,
      "step": 269
    },
    {
      "epoch": 0.059386341141537444,
      "grad_norm": 0.29682958126068115,
      "learning_rate": 8e-05,
      "loss": 1.6548,
      "step": 270
    },
    {
      "epoch": 0.05960629055317277,
      "grad_norm": 0.2702498137950897,
      "learning_rate": 8e-05,
      "loss": 1.6022,
      "step": 271
    },
    {
      "epoch": 0.05982623996480809,
      "grad_norm": 0.2940424680709839,
      "learning_rate": 8e-05,
      "loss": 1.7955,
      "step": 272
    },
    {
      "epoch": 0.06004618937644342,
      "grad_norm": 0.2655317485332489,
      "learning_rate": 8e-05,
      "loss": 1.786,
      "step": 273
    },
    {
      "epoch": 0.060266138788078745,
      "grad_norm": 0.28093400597572327,
      "learning_rate": 8e-05,
      "loss": 1.9798,
      "step": 274
    },
    {
      "epoch": 0.060486088199714065,
      "grad_norm": 0.2635514736175537,
      "learning_rate": 8e-05,
      "loss": 1.6737,
      "step": 275
    },
    {
      "epoch": 0.06070603761134939,
      "grad_norm": 0.2648226320743561,
      "learning_rate": 8e-05,
      "loss": 1.8771,
      "step": 276
    },
    {
      "epoch": 0.06092598702298471,
      "grad_norm": 0.2934603691101074,
      "learning_rate": 8e-05,
      "loss": 1.4751,
      "step": 277
    },
    {
      "epoch": 0.06114593643462004,
      "grad_norm": 0.26369500160217285,
      "learning_rate": 8e-05,
      "loss": 1.7832,
      "step": 278
    },
    {
      "epoch": 0.06136588584625536,
      "grad_norm": 0.26159989833831787,
      "learning_rate": 8e-05,
      "loss": 1.7276,
      "step": 279
    },
    {
      "epoch": 0.061585835257890686,
      "grad_norm": 0.2826705873012543,
      "learning_rate": 8e-05,
      "loss": 1.8767,
      "step": 280
    },
    {
      "epoch": 0.06180578466952601,
      "grad_norm": 0.2911459505558014,
      "learning_rate": 8e-05,
      "loss": 1.7795,
      "step": 281
    },
    {
      "epoch": 0.062025734081161334,
      "grad_norm": 0.27846869826316833,
      "learning_rate": 8e-05,
      "loss": 1.838,
      "step": 282
    },
    {
      "epoch": 0.062245683492796654,
      "grad_norm": 0.33195585012435913,
      "learning_rate": 8e-05,
      "loss": 1.8576,
      "step": 283
    },
    {
      "epoch": 0.06246563290443198,
      "grad_norm": 0.26306337118148804,
      "learning_rate": 8e-05,
      "loss": 1.7202,
      "step": 284
    },
    {
      "epoch": 0.0626855823160673,
      "grad_norm": 0.2703022360801697,
      "learning_rate": 8e-05,
      "loss": 1.6962,
      "step": 285
    },
    {
      "epoch": 0.06290553172770263,
      "grad_norm": 0.2754605710506439,
      "learning_rate": 8e-05,
      "loss": 1.6468,
      "step": 286
    },
    {
      "epoch": 0.06312548113933796,
      "grad_norm": 0.2995694577693939,
      "learning_rate": 8e-05,
      "loss": 1.9298,
      "step": 287
    },
    {
      "epoch": 0.06334543055097328,
      "grad_norm": 0.27501800656318665,
      "learning_rate": 8e-05,
      "loss": 1.8152,
      "step": 288
    },
    {
      "epoch": 0.0635653799626086,
      "grad_norm": 0.2668202519416809,
      "learning_rate": 8e-05,
      "loss": 1.8809,
      "step": 289
    },
    {
      "epoch": 0.06378532937424393,
      "grad_norm": 0.26209571957588196,
      "learning_rate": 8e-05,
      "loss": 1.4927,
      "step": 290
    },
    {
      "epoch": 0.06400527878587925,
      "grad_norm": 0.35276591777801514,
      "learning_rate": 8e-05,
      "loss": 1.9654,
      "step": 291
    },
    {
      "epoch": 0.06422522819751457,
      "grad_norm": 0.26070040464401245,
      "learning_rate": 8e-05,
      "loss": 1.7332,
      "step": 292
    },
    {
      "epoch": 0.06444517760914989,
      "grad_norm": 0.26518604159355164,
      "learning_rate": 8e-05,
      "loss": 1.6867,
      "step": 293
    },
    {
      "epoch": 0.06466512702078522,
      "grad_norm": 0.28992095589637756,
      "learning_rate": 8e-05,
      "loss": 1.7498,
      "step": 294
    },
    {
      "epoch": 0.06488507643242054,
      "grad_norm": 0.27465108036994934,
      "learning_rate": 8e-05,
      "loss": 1.6095,
      "step": 295
    },
    {
      "epoch": 0.06510502584405586,
      "grad_norm": 0.2841359078884125,
      "learning_rate": 8e-05,
      "loss": 1.6869,
      "step": 296
    },
    {
      "epoch": 0.0653249752556912,
      "grad_norm": 0.28873759508132935,
      "learning_rate": 8e-05,
      "loss": 1.7954,
      "step": 297
    },
    {
      "epoch": 0.06554492466732652,
      "grad_norm": 0.2542605698108673,
      "learning_rate": 8e-05,
      "loss": 1.6075,
      "step": 298
    },
    {
      "epoch": 0.06576487407896184,
      "grad_norm": 0.270823210477829,
      "learning_rate": 8e-05,
      "loss": 1.7238,
      "step": 299
    },
    {
      "epoch": 0.06598482349059716,
      "grad_norm": 0.2610267102718353,
      "learning_rate": 8e-05,
      "loss": 1.697,
      "step": 300
    },
    {
      "epoch": 0.06620477290223249,
      "grad_norm": 0.28088685870170593,
      "learning_rate": 8e-05,
      "loss": 1.6806,
      "step": 301
    },
    {
      "epoch": 0.06642472231386781,
      "grad_norm": 0.2656930088996887,
      "learning_rate": 8e-05,
      "loss": 1.8744,
      "step": 302
    },
    {
      "epoch": 0.06664467172550313,
      "grad_norm": 0.2721637189388275,
      "learning_rate": 8e-05,
      "loss": 1.6903,
      "step": 303
    },
    {
      "epoch": 0.06686462113713845,
      "grad_norm": 0.2612883746623993,
      "learning_rate": 8e-05,
      "loss": 1.7444,
      "step": 304
    },
    {
      "epoch": 0.06708457054877379,
      "grad_norm": 0.2533530592918396,
      "learning_rate": 8e-05,
      "loss": 1.6427,
      "step": 305
    },
    {
      "epoch": 0.0673045199604091,
      "grad_norm": 0.27200043201446533,
      "learning_rate": 8e-05,
      "loss": 1.769,
      "step": 306
    },
    {
      "epoch": 0.06752446937204443,
      "grad_norm": 0.2626403272151947,
      "learning_rate": 8e-05,
      "loss": 1.64,
      "step": 307
    },
    {
      "epoch": 0.06774441878367975,
      "grad_norm": 0.3720408082008362,
      "learning_rate": 8e-05,
      "loss": 1.9055,
      "step": 308
    },
    {
      "epoch": 0.06796436819531508,
      "grad_norm": 0.2745527923107147,
      "learning_rate": 8e-05,
      "loss": 1.7844,
      "step": 309
    },
    {
      "epoch": 0.0681843176069504,
      "grad_norm": 0.2568323612213135,
      "learning_rate": 8e-05,
      "loss": 1.6728,
      "step": 310
    },
    {
      "epoch": 0.06840426701858572,
      "grad_norm": 0.2704140543937683,
      "learning_rate": 8e-05,
      "loss": 1.7685,
      "step": 311
    },
    {
      "epoch": 0.06862421643022105,
      "grad_norm": 0.27828502655029297,
      "learning_rate": 8e-05,
      "loss": 1.7957,
      "step": 312
    },
    {
      "epoch": 0.06884416584185638,
      "grad_norm": 0.2951858341693878,
      "learning_rate": 8e-05,
      "loss": 1.7709,
      "step": 313
    },
    {
      "epoch": 0.0690641152534917,
      "grad_norm": 0.2756475806236267,
      "learning_rate": 8e-05,
      "loss": 1.6348,
      "step": 314
    },
    {
      "epoch": 0.06928406466512702,
      "grad_norm": 0.2913607954978943,
      "learning_rate": 8e-05,
      "loss": 1.7888,
      "step": 315
    },
    {
      "epoch": 0.06950401407676235,
      "grad_norm": 0.2798636853694916,
      "learning_rate": 8e-05,
      "loss": 1.7806,
      "step": 316
    },
    {
      "epoch": 0.06972396348839767,
      "grad_norm": 0.27596554160118103,
      "learning_rate": 8e-05,
      "loss": 1.7458,
      "step": 317
    },
    {
      "epoch": 0.06994391290003299,
      "grad_norm": 0.26655322313308716,
      "learning_rate": 8e-05,
      "loss": 1.5985,
      "step": 318
    },
    {
      "epoch": 0.07016386231166831,
      "grad_norm": 0.2731332778930664,
      "learning_rate": 8e-05,
      "loss": 1.5995,
      "step": 319
    },
    {
      "epoch": 0.07038381172330364,
      "grad_norm": 0.2769210934638977,
      "learning_rate": 8e-05,
      "loss": 1.6748,
      "step": 320
    },
    {
      "epoch": 0.07060376113493896,
      "grad_norm": 0.290889173746109,
      "learning_rate": 8e-05,
      "loss": 1.9427,
      "step": 321
    },
    {
      "epoch": 0.07082371054657428,
      "grad_norm": 0.2911258339881897,
      "learning_rate": 8e-05,
      "loss": 1.7723,
      "step": 322
    },
    {
      "epoch": 0.07104365995820962,
      "grad_norm": 0.301992267370224,
      "learning_rate": 8e-05,
      "loss": 1.7772,
      "step": 323
    },
    {
      "epoch": 0.07126360936984494,
      "grad_norm": 0.3023516535758972,
      "learning_rate": 8e-05,
      "loss": 1.8363,
      "step": 324
    },
    {
      "epoch": 0.07148355878148026,
      "grad_norm": 0.3058542013168335,
      "learning_rate": 8e-05,
      "loss": 1.8762,
      "step": 325
    },
    {
      "epoch": 0.07170350819311558,
      "grad_norm": 0.3215092718601227,
      "learning_rate": 8e-05,
      "loss": 1.7265,
      "step": 326
    },
    {
      "epoch": 0.07192345760475091,
      "grad_norm": 0.2762998342514038,
      "learning_rate": 8e-05,
      "loss": 1.6361,
      "step": 327
    },
    {
      "epoch": 0.07214340701638623,
      "grad_norm": 0.258635014295578,
      "learning_rate": 8e-05,
      "loss": 1.7031,
      "step": 328
    },
    {
      "epoch": 0.07236335642802155,
      "grad_norm": 0.27160710096359253,
      "learning_rate": 8e-05,
      "loss": 1.6759,
      "step": 329
    },
    {
      "epoch": 0.07258330583965687,
      "grad_norm": 0.31089314818382263,
      "learning_rate": 8e-05,
      "loss": 1.8141,
      "step": 330
    },
    {
      "epoch": 0.0728032552512922,
      "grad_norm": 0.3026575744152069,
      "learning_rate": 8e-05,
      "loss": 1.9513,
      "step": 331
    },
    {
      "epoch": 0.07302320466292753,
      "grad_norm": 0.2692122161388397,
      "learning_rate": 8e-05,
      "loss": 1.8277,
      "step": 332
    },
    {
      "epoch": 0.07324315407456285,
      "grad_norm": 0.27460286021232605,
      "learning_rate": 8e-05,
      "loss": 1.6426,
      "step": 333
    },
    {
      "epoch": 0.07346310348619818,
      "grad_norm": 0.2557325065135956,
      "learning_rate": 8e-05,
      "loss": 1.6418,
      "step": 334
    },
    {
      "epoch": 0.0736830528978335,
      "grad_norm": 0.28074318170547485,
      "learning_rate": 8e-05,
      "loss": 1.79,
      "step": 335
    },
    {
      "epoch": 0.07390300230946882,
      "grad_norm": 0.28538671135902405,
      "learning_rate": 8e-05,
      "loss": 1.7363,
      "step": 336
    },
    {
      "epoch": 0.07412295172110414,
      "grad_norm": 0.27379995584487915,
      "learning_rate": 8e-05,
      "loss": 1.7881,
      "step": 337
    },
    {
      "epoch": 0.07434290113273948,
      "grad_norm": 0.2628316283226013,
      "learning_rate": 8e-05,
      "loss": 1.745,
      "step": 338
    },
    {
      "epoch": 0.0745628505443748,
      "grad_norm": 0.2573058009147644,
      "learning_rate": 8e-05,
      "loss": 1.7997,
      "step": 339
    },
    {
      "epoch": 0.07478279995601012,
      "grad_norm": 0.31905651092529297,
      "learning_rate": 8e-05,
      "loss": 1.8125,
      "step": 340
    },
    {
      "epoch": 0.07500274936764544,
      "grad_norm": 0.2501446604728699,
      "learning_rate": 8e-05,
      "loss": 1.557,
      "step": 341
    },
    {
      "epoch": 0.07522269877928077,
      "grad_norm": 0.26969289779663086,
      "learning_rate": 8e-05,
      "loss": 1.7819,
      "step": 342
    },
    {
      "epoch": 0.07544264819091609,
      "grad_norm": 0.28457415103912354,
      "learning_rate": 8e-05,
      "loss": 1.7682,
      "step": 343
    },
    {
      "epoch": 0.07566259760255141,
      "grad_norm": 0.27833452820777893,
      "learning_rate": 8e-05,
      "loss": 1.8436,
      "step": 344
    },
    {
      "epoch": 0.07588254701418674,
      "grad_norm": 0.2574867010116577,
      "learning_rate": 8e-05,
      "loss": 1.7196,
      "step": 345
    },
    {
      "epoch": 0.07610249642582206,
      "grad_norm": 0.30035245418548584,
      "learning_rate": 8e-05,
      "loss": 1.7159,
      "step": 346
    },
    {
      "epoch": 0.07632244583745738,
      "grad_norm": 0.284169465303421,
      "learning_rate": 8e-05,
      "loss": 1.7238,
      "step": 347
    },
    {
      "epoch": 0.0765423952490927,
      "grad_norm": 0.257168173789978,
      "learning_rate": 8e-05,
      "loss": 1.8531,
      "step": 348
    },
    {
      "epoch": 0.07676234466072804,
      "grad_norm": 0.2611413300037384,
      "learning_rate": 8e-05,
      "loss": 1.7753,
      "step": 349
    },
    {
      "epoch": 0.07698229407236336,
      "grad_norm": 0.26592132449150085,
      "learning_rate": 8e-05,
      "loss": 1.7557,
      "step": 350
    },
    {
      "epoch": 0.07720224348399868,
      "grad_norm": 0.27427396178245544,
      "learning_rate": 8e-05,
      "loss": 1.8699,
      "step": 351
    },
    {
      "epoch": 0.077422192895634,
      "grad_norm": 0.27014485001564026,
      "learning_rate": 8e-05,
      "loss": 1.816,
      "step": 352
    },
    {
      "epoch": 0.07764214230726933,
      "grad_norm": 0.27720019221305847,
      "learning_rate": 8e-05,
      "loss": 1.9601,
      "step": 353
    },
    {
      "epoch": 0.07786209171890465,
      "grad_norm": 0.3222314417362213,
      "learning_rate": 8e-05,
      "loss": 1.6726,
      "step": 354
    },
    {
      "epoch": 0.07808204113053997,
      "grad_norm": 0.2675410211086273,
      "learning_rate": 8e-05,
      "loss": 1.7113,
      "step": 355
    },
    {
      "epoch": 0.0783019905421753,
      "grad_norm": 0.2902251183986664,
      "learning_rate": 8e-05,
      "loss": 1.7734,
      "step": 356
    },
    {
      "epoch": 0.07852193995381063,
      "grad_norm": 0.2985514998435974,
      "learning_rate": 8e-05,
      "loss": 1.9182,
      "step": 357
    },
    {
      "epoch": 0.07874188936544595,
      "grad_norm": 0.30351343750953674,
      "learning_rate": 8e-05,
      "loss": 1.7795,
      "step": 358
    },
    {
      "epoch": 0.07896183877708127,
      "grad_norm": 0.2885829210281372,
      "learning_rate": 8e-05,
      "loss": 1.8054,
      "step": 359
    },
    {
      "epoch": 0.0791817881887166,
      "grad_norm": 0.273366242647171,
      "learning_rate": 8e-05,
      "loss": 1.7903,
      "step": 360
    },
    {
      "epoch": 0.07940173760035192,
      "grad_norm": 0.2959200441837311,
      "learning_rate": 8e-05,
      "loss": 1.9163,
      "step": 361
    },
    {
      "epoch": 0.07962168701198724,
      "grad_norm": 0.2587856948375702,
      "learning_rate": 8e-05,
      "loss": 1.5969,
      "step": 362
    },
    {
      "epoch": 0.07984163642362256,
      "grad_norm": 0.27777665853500366,
      "learning_rate": 8e-05,
      "loss": 1.8769,
      "step": 363
    },
    {
      "epoch": 0.0800615858352579,
      "grad_norm": 0.2635156512260437,
      "learning_rate": 8e-05,
      "loss": 1.8236,
      "step": 364
    },
    {
      "epoch": 0.08028153524689322,
      "grad_norm": 0.26534774899482727,
      "learning_rate": 8e-05,
      "loss": 1.6824,
      "step": 365
    },
    {
      "epoch": 0.08050148465852854,
      "grad_norm": 0.26372772455215454,
      "learning_rate": 8e-05,
      "loss": 1.517,
      "step": 366
    },
    {
      "epoch": 0.08072143407016386,
      "grad_norm": 0.2707895338535309,
      "learning_rate": 8e-05,
      "loss": 1.6757,
      "step": 367
    },
    {
      "epoch": 0.08094138348179919,
      "grad_norm": 0.2712070345878601,
      "learning_rate": 8e-05,
      "loss": 1.7261,
      "step": 368
    },
    {
      "epoch": 0.08116133289343451,
      "grad_norm": 0.2870525121688843,
      "learning_rate": 8e-05,
      "loss": 1.6337,
      "step": 369
    },
    {
      "epoch": 0.08138128230506983,
      "grad_norm": 0.30548396706581116,
      "learning_rate": 8e-05,
      "loss": 1.8733,
      "step": 370
    },
    {
      "epoch": 0.08160123171670516,
      "grad_norm": 0.2853962182998657,
      "learning_rate": 8e-05,
      "loss": 1.7938,
      "step": 371
    },
    {
      "epoch": 0.08182118112834048,
      "grad_norm": 0.2716579735279083,
      "learning_rate": 8e-05,
      "loss": 1.6733,
      "step": 372
    },
    {
      "epoch": 0.0820411305399758,
      "grad_norm": 0.3110131025314331,
      "learning_rate": 8e-05,
      "loss": 1.8554,
      "step": 373
    },
    {
      "epoch": 0.08226107995161112,
      "grad_norm": 0.28003835678100586,
      "learning_rate": 8e-05,
      "loss": 1.8032,
      "step": 374
    },
    {
      "epoch": 0.08248102936324646,
      "grad_norm": 0.28504347801208496,
      "learning_rate": 8e-05,
      "loss": 1.942,
      "step": 375
    },
    {
      "epoch": 0.08270097877488178,
      "grad_norm": 0.2593232989311218,
      "learning_rate": 8e-05,
      "loss": 1.4993,
      "step": 376
    },
    {
      "epoch": 0.0829209281865171,
      "grad_norm": 0.35680094361305237,
      "learning_rate": 8e-05,
      "loss": 1.8997,
      "step": 377
    },
    {
      "epoch": 0.08314087759815242,
      "grad_norm": 0.2747777998447418,
      "learning_rate": 8e-05,
      "loss": 1.7364,
      "step": 378
    },
    {
      "epoch": 0.08336082700978775,
      "grad_norm": 0.26816287636756897,
      "learning_rate": 8e-05,
      "loss": 1.7011,
      "step": 379
    },
    {
      "epoch": 0.08358077642142307,
      "grad_norm": 0.31877851486206055,
      "learning_rate": 8e-05,
      "loss": 1.6131,
      "step": 380
    },
    {
      "epoch": 0.08380072583305839,
      "grad_norm": 0.2845601737499237,
      "learning_rate": 8e-05,
      "loss": 1.6544,
      "step": 381
    },
    {
      "epoch": 0.08402067524469373,
      "grad_norm": 0.27758803963661194,
      "learning_rate": 8e-05,
      "loss": 1.8891,
      "step": 382
    },
    {
      "epoch": 0.08424062465632905,
      "grad_norm": 0.2832657992839813,
      "learning_rate": 8e-05,
      "loss": 1.7505,
      "step": 383
    },
    {
      "epoch": 0.08446057406796437,
      "grad_norm": 0.2901705801486969,
      "learning_rate": 8e-05,
      "loss": 1.7501,
      "step": 384
    },
    {
      "epoch": 0.08468052347959969,
      "grad_norm": 0.31189531087875366,
      "learning_rate": 8e-05,
      "loss": 1.8132,
      "step": 385
    },
    {
      "epoch": 0.08490047289123502,
      "grad_norm": 0.27582603693008423,
      "learning_rate": 8e-05,
      "loss": 1.7693,
      "step": 386
    },
    {
      "epoch": 0.08512042230287034,
      "grad_norm": 0.3030100464820862,
      "learning_rate": 8e-05,
      "loss": 1.7327,
      "step": 387
    },
    {
      "epoch": 0.08534037171450566,
      "grad_norm": 0.26879045367240906,
      "learning_rate": 8e-05,
      "loss": 1.6614,
      "step": 388
    },
    {
      "epoch": 0.08556032112614098,
      "grad_norm": 0.29507508873939514,
      "learning_rate": 8e-05,
      "loss": 1.9483,
      "step": 389
    },
    {
      "epoch": 0.08578027053777632,
      "grad_norm": 0.27386122941970825,
      "learning_rate": 8e-05,
      "loss": 1.8974,
      "step": 390
    },
    {
      "epoch": 0.08600021994941164,
      "grad_norm": 0.27103161811828613,
      "learning_rate": 8e-05,
      "loss": 1.7579,
      "step": 391
    },
    {
      "epoch": 0.08622016936104696,
      "grad_norm": 0.3045141100883484,
      "learning_rate": 8e-05,
      "loss": 1.8175,
      "step": 392
    },
    {
      "epoch": 0.08644011877268229,
      "grad_norm": 0.29032695293426514,
      "learning_rate": 8e-05,
      "loss": 1.7493,
      "step": 393
    },
    {
      "epoch": 0.08666006818431761,
      "grad_norm": 0.27853158116340637,
      "learning_rate": 8e-05,
      "loss": 1.7297,
      "step": 394
    },
    {
      "epoch": 0.08688001759595293,
      "grad_norm": 0.3007650375366211,
      "learning_rate": 8e-05,
      "loss": 1.6736,
      "step": 395
    },
    {
      "epoch": 0.08709996700758825,
      "grad_norm": 0.28009670972824097,
      "learning_rate": 8e-05,
      "loss": 1.9539,
      "step": 396
    },
    {
      "epoch": 0.08731991641922358,
      "grad_norm": 0.2512955665588379,
      "learning_rate": 8e-05,
      "loss": 1.6362,
      "step": 397
    },
    {
      "epoch": 0.0875398658308589,
      "grad_norm": 0.297489732503891,
      "learning_rate": 8e-05,
      "loss": 1.9097,
      "step": 398
    },
    {
      "epoch": 0.08775981524249422,
      "grad_norm": 0.2735532522201538,
      "learning_rate": 8e-05,
      "loss": 1.8348,
      "step": 399
    },
    {
      "epoch": 0.08797976465412954,
      "grad_norm": 0.2559053897857666,
      "learning_rate": 8e-05,
      "loss": 1.685,
      "step": 400
    },
    {
      "epoch": 0.08819971406576488,
      "grad_norm": 0.27982097864151,
      "learning_rate": 8e-05,
      "loss": 1.6801,
      "step": 401
    },
    {
      "epoch": 0.0884196634774002,
      "grad_norm": 0.26066988706588745,
      "learning_rate": 8e-05,
      "loss": 1.7732,
      "step": 402
    },
    {
      "epoch": 0.08863961288903552,
      "grad_norm": 0.26763463020324707,
      "learning_rate": 8e-05,
      "loss": 1.7214,
      "step": 403
    },
    {
      "epoch": 0.08885956230067085,
      "grad_norm": 0.2795925736427307,
      "learning_rate": 8e-05,
      "loss": 1.8387,
      "step": 404
    },
    {
      "epoch": 0.08907951171230617,
      "grad_norm": 0.266305148601532,
      "learning_rate": 8e-05,
      "loss": 1.6515,
      "step": 405
    },
    {
      "epoch": 0.08929946112394149,
      "grad_norm": 0.27049583196640015,
      "learning_rate": 8e-05,
      "loss": 1.7824,
      "step": 406
    },
    {
      "epoch": 0.08951941053557681,
      "grad_norm": 0.2959458529949188,
      "learning_rate": 8e-05,
      "loss": 1.8766,
      "step": 407
    },
    {
      "epoch": 0.08973935994721215,
      "grad_norm": 0.28563347458839417,
      "learning_rate": 8e-05,
      "loss": 1.8618,
      "step": 408
    },
    {
      "epoch": 0.08995930935884747,
      "grad_norm": 0.2840110659599304,
      "learning_rate": 8e-05,
      "loss": 1.6834,
      "step": 409
    },
    {
      "epoch": 0.09017925877048279,
      "grad_norm": 0.25303247570991516,
      "learning_rate": 8e-05,
      "loss": 1.6477,
      "step": 410
    },
    {
      "epoch": 0.09039920818211811,
      "grad_norm": 0.27236899733543396,
      "learning_rate": 8e-05,
      "loss": 1.7004,
      "step": 411
    },
    {
      "epoch": 0.09061915759375344,
      "grad_norm": 0.2795659899711609,
      "learning_rate": 8e-05,
      "loss": 1.7492,
      "step": 412
    },
    {
      "epoch": 0.09083910700538876,
      "grad_norm": 0.26019132137298584,
      "learning_rate": 8e-05,
      "loss": 1.691,
      "step": 413
    },
    {
      "epoch": 0.09105905641702408,
      "grad_norm": 0.26624274253845215,
      "learning_rate": 8e-05,
      "loss": 1.7001,
      "step": 414
    },
    {
      "epoch": 0.09127900582865942,
      "grad_norm": 0.2661585509777069,
      "learning_rate": 8e-05,
      "loss": 1.6762,
      "step": 415
    },
    {
      "epoch": 0.09149895524029474,
      "grad_norm": 0.2719002068042755,
      "learning_rate": 8e-05,
      "loss": 1.6915,
      "step": 416
    },
    {
      "epoch": 0.09171890465193006,
      "grad_norm": 0.24670244753360748,
      "learning_rate": 8e-05,
      "loss": 1.5598,
      "step": 417
    },
    {
      "epoch": 0.09193885406356538,
      "grad_norm": 0.2550405263900757,
      "learning_rate": 8e-05,
      "loss": 1.4817,
      "step": 418
    },
    {
      "epoch": 0.09215880347520071,
      "grad_norm": 0.26272761821746826,
      "learning_rate": 8e-05,
      "loss": 1.7016,
      "step": 419
    },
    {
      "epoch": 0.09237875288683603,
      "grad_norm": 0.2673632502555847,
      "learning_rate": 8e-05,
      "loss": 1.7626,
      "step": 420
    },
    {
      "epoch": 0.09259870229847135,
      "grad_norm": 0.25949448347091675,
      "learning_rate": 8e-05,
      "loss": 1.6273,
      "step": 421
    },
    {
      "epoch": 0.09281865171010667,
      "grad_norm": 0.27953028678894043,
      "learning_rate": 8e-05,
      "loss": 1.8843,
      "step": 422
    },
    {
      "epoch": 0.093038601121742,
      "grad_norm": 0.2534630298614502,
      "learning_rate": 8e-05,
      "loss": 1.7305,
      "step": 423
    },
    {
      "epoch": 0.09325855053337732,
      "grad_norm": 0.2573072910308838,
      "learning_rate": 8e-05,
      "loss": 1.6397,
      "step": 424
    },
    {
      "epoch": 0.09347849994501264,
      "grad_norm": 0.2604135572910309,
      "learning_rate": 8e-05,
      "loss": 1.6696,
      "step": 425
    },
    {
      "epoch": 0.09369844935664796,
      "grad_norm": 0.25805628299713135,
      "learning_rate": 8e-05,
      "loss": 1.6441,
      "step": 426
    },
    {
      "epoch": 0.0939183987682833,
      "grad_norm": 0.2935563027858734,
      "learning_rate": 8e-05,
      "loss": 1.6475,
      "step": 427
    },
    {
      "epoch": 0.09413834817991862,
      "grad_norm": 0.25222933292388916,
      "learning_rate": 8e-05,
      "loss": 1.727,
      "step": 428
    },
    {
      "epoch": 0.09435829759155394,
      "grad_norm": 0.2593076527118683,
      "learning_rate": 8e-05,
      "loss": 1.7066,
      "step": 429
    },
    {
      "epoch": 0.09457824700318927,
      "grad_norm": 0.25259336829185486,
      "learning_rate": 8e-05,
      "loss": 1.6821,
      "step": 430
    },
    {
      "epoch": 0.09479819641482459,
      "grad_norm": 0.2512541115283966,
      "learning_rate": 8e-05,
      "loss": 1.5923,
      "step": 431
    },
    {
      "epoch": 0.09501814582645991,
      "grad_norm": 0.2711183726787567,
      "learning_rate": 8e-05,
      "loss": 1.755,
      "step": 432
    },
    {
      "epoch": 0.09523809523809523,
      "grad_norm": 0.2782961130142212,
      "learning_rate": 8e-05,
      "loss": 1.8914,
      "step": 433
    },
    {
      "epoch": 0.09545804464973057,
      "grad_norm": 0.25964146852493286,
      "learning_rate": 8e-05,
      "loss": 1.6588,
      "step": 434
    },
    {
      "epoch": 0.09567799406136589,
      "grad_norm": 0.27077510952949524,
      "learning_rate": 8e-05,
      "loss": 1.753,
      "step": 435
    },
    {
      "epoch": 0.09589794347300121,
      "grad_norm": 0.2923937141895294,
      "learning_rate": 8e-05,
      "loss": 1.8218,
      "step": 436
    },
    {
      "epoch": 0.09611789288463653,
      "grad_norm": 0.2513190805912018,
      "learning_rate": 8e-05,
      "loss": 1.6232,
      "step": 437
    },
    {
      "epoch": 0.09633784229627186,
      "grad_norm": 0.28531181812286377,
      "learning_rate": 8e-05,
      "loss": 1.7199,
      "step": 438
    },
    {
      "epoch": 0.09655779170790718,
      "grad_norm": 0.302020400762558,
      "learning_rate": 8e-05,
      "loss": 1.8359,
      "step": 439
    },
    {
      "epoch": 0.0967777411195425,
      "grad_norm": 0.28001338243484497,
      "learning_rate": 8e-05,
      "loss": 1.8434,
      "step": 440
    },
    {
      "epoch": 0.09699769053117784,
      "grad_norm": 0.2990663945674896,
      "learning_rate": 8e-05,
      "loss": 1.6995,
      "step": 441
    },
    {
      "epoch": 0.09721763994281316,
      "grad_norm": 0.266197144985199,
      "learning_rate": 8e-05,
      "loss": 1.6195,
      "step": 442
    },
    {
      "epoch": 0.09743758935444848,
      "grad_norm": 0.28108519315719604,
      "learning_rate": 8e-05,
      "loss": 1.8108,
      "step": 443
    },
    {
      "epoch": 0.0976575387660838,
      "grad_norm": 0.26744788885116577,
      "learning_rate": 8e-05,
      "loss": 1.6497,
      "step": 444
    },
    {
      "epoch": 0.09787748817771913,
      "grad_norm": 0.28030574321746826,
      "learning_rate": 8e-05,
      "loss": 1.8143,
      "step": 445
    },
    {
      "epoch": 0.09809743758935445,
      "grad_norm": 0.27872079610824585,
      "learning_rate": 8e-05,
      "loss": 1.6319,
      "step": 446
    },
    {
      "epoch": 0.09831738700098977,
      "grad_norm": 0.2816067039966583,
      "learning_rate": 8e-05,
      "loss": 1.8385,
      "step": 447
    },
    {
      "epoch": 0.09853733641262509,
      "grad_norm": 0.25677627325057983,
      "learning_rate": 8e-05,
      "loss": 1.6885,
      "step": 448
    },
    {
      "epoch": 0.09875728582426042,
      "grad_norm": 0.276569128036499,
      "learning_rate": 8e-05,
      "loss": 1.7652,
      "step": 449
    },
    {
      "epoch": 0.09897723523589574,
      "grad_norm": 0.2765633463859558,
      "learning_rate": 8e-05,
      "loss": 1.7763,
      "step": 450
    },
    {
      "epoch": 0.09919718464753106,
      "grad_norm": 0.27050015330314636,
      "learning_rate": 8e-05,
      "loss": 1.6459,
      "step": 451
    },
    {
      "epoch": 0.0994171340591664,
      "grad_norm": 0.2552846372127533,
      "learning_rate": 8e-05,
      "loss": 1.6877,
      "step": 452
    },
    {
      "epoch": 0.09963708347080172,
      "grad_norm": 0.2653469741344452,
      "learning_rate": 8e-05,
      "loss": 1.6536,
      "step": 453
    },
    {
      "epoch": 0.09985703288243704,
      "grad_norm": 0.28801941871643066,
      "learning_rate": 8e-05,
      "loss": 1.7643,
      "step": 454
    },
    {
      "epoch": 0.10007698229407236,
      "grad_norm": 0.2930269241333008,
      "learning_rate": 8e-05,
      "loss": 1.7766,
      "step": 455
    },
    {
      "epoch": 0.10029693170570769,
      "grad_norm": 0.2718334496021271,
      "learning_rate": 8e-05,
      "loss": 1.7347,
      "step": 456
    },
    {
      "epoch": 0.10051688111734301,
      "grad_norm": 0.2807629704475403,
      "learning_rate": 8e-05,
      "loss": 1.7245,
      "step": 457
    },
    {
      "epoch": 0.10073683052897833,
      "grad_norm": 0.2801489531993866,
      "learning_rate": 8e-05,
      "loss": 1.7854,
      "step": 458
    },
    {
      "epoch": 0.10095677994061365,
      "grad_norm": 0.2616996765136719,
      "learning_rate": 8e-05,
      "loss": 1.6179,
      "step": 459
    },
    {
      "epoch": 0.10117672935224899,
      "grad_norm": 0.2626480758190155,
      "learning_rate": 8e-05,
      "loss": 1.7475,
      "step": 460
    },
    {
      "epoch": 0.10139667876388431,
      "grad_norm": 0.27338841557502747,
      "learning_rate": 8e-05,
      "loss": 1.8972,
      "step": 461
    },
    {
      "epoch": 0.10161662817551963,
      "grad_norm": 0.2695038616657257,
      "learning_rate": 8e-05,
      "loss": 1.7279,
      "step": 462
    },
    {
      "epoch": 0.10183657758715496,
      "grad_norm": 0.25614050030708313,
      "learning_rate": 8e-05,
      "loss": 1.6,
      "step": 463
    },
    {
      "epoch": 0.10205652699879028,
      "grad_norm": 0.2722180187702179,
      "learning_rate": 8e-05,
      "loss": 1.9241,
      "step": 464
    },
    {
      "epoch": 0.1022764764104256,
      "grad_norm": 0.2580203115940094,
      "learning_rate": 8e-05,
      "loss": 1.693,
      "step": 465
    },
    {
      "epoch": 0.10249642582206092,
      "grad_norm": 0.2848857641220093,
      "learning_rate": 8e-05,
      "loss": 1.9072,
      "step": 466
    },
    {
      "epoch": 0.10271637523369626,
      "grad_norm": 0.2783052325248718,
      "learning_rate": 8e-05,
      "loss": 1.9102,
      "step": 467
    },
    {
      "epoch": 0.10293632464533158,
      "grad_norm": 0.279695987701416,
      "learning_rate": 8e-05,
      "loss": 1.7491,
      "step": 468
    },
    {
      "epoch": 0.1031562740569669,
      "grad_norm": 0.2493034154176712,
      "learning_rate": 8e-05,
      "loss": 1.6789,
      "step": 469
    },
    {
      "epoch": 0.10337622346860222,
      "grad_norm": 0.2751196622848511,
      "learning_rate": 8e-05,
      "loss": 1.8132,
      "step": 470
    },
    {
      "epoch": 0.10359617288023755,
      "grad_norm": 0.2739677131175995,
      "learning_rate": 8e-05,
      "loss": 1.7945,
      "step": 471
    },
    {
      "epoch": 0.10381612229187287,
      "grad_norm": 0.30357351899147034,
      "learning_rate": 8e-05,
      "loss": 1.9113,
      "step": 472
    },
    {
      "epoch": 0.10403607170350819,
      "grad_norm": 0.2646970748901367,
      "learning_rate": 8e-05,
      "loss": 1.811,
      "step": 473
    },
    {
      "epoch": 0.10425602111514351,
      "grad_norm": 0.2626940608024597,
      "learning_rate": 8e-05,
      "loss": 1.6911,
      "step": 474
    },
    {
      "epoch": 0.10447597052677884,
      "grad_norm": 0.2613508701324463,
      "learning_rate": 8e-05,
      "loss": 1.7209,
      "step": 475
    },
    {
      "epoch": 0.10469591993841416,
      "grad_norm": 0.2609264552593231,
      "learning_rate": 8e-05,
      "loss": 1.6303,
      "step": 476
    },
    {
      "epoch": 0.10491586935004948,
      "grad_norm": 0.2549975514411926,
      "learning_rate": 8e-05,
      "loss": 1.7769,
      "step": 477
    },
    {
      "epoch": 0.10513581876168482,
      "grad_norm": 0.2742570638656616,
      "learning_rate": 8e-05,
      "loss": 1.8101,
      "step": 478
    },
    {
      "epoch": 0.10535576817332014,
      "grad_norm": 0.267070472240448,
      "learning_rate": 8e-05,
      "loss": 1.787,
      "step": 479
    },
    {
      "epoch": 0.10557571758495546,
      "grad_norm": 0.2735085189342499,
      "learning_rate": 8e-05,
      "loss": 1.8112,
      "step": 480
    },
    {
      "epoch": 0.10579566699659078,
      "grad_norm": 0.260111540555954,
      "learning_rate": 8e-05,
      "loss": 1.6926,
      "step": 481
    },
    {
      "epoch": 0.10601561640822611,
      "grad_norm": 0.26309284567832947,
      "learning_rate": 8e-05,
      "loss": 1.778,
      "step": 482
    },
    {
      "epoch": 0.10623556581986143,
      "grad_norm": 0.2658458948135376,
      "learning_rate": 8e-05,
      "loss": 1.7179,
      "step": 483
    },
    {
      "epoch": 0.10645551523149675,
      "grad_norm": 0.27498647570610046,
      "learning_rate": 8e-05,
      "loss": 1.6689,
      "step": 484
    },
    {
      "epoch": 0.10667546464313207,
      "grad_norm": 0.2658367156982422,
      "learning_rate": 8e-05,
      "loss": 1.6786,
      "step": 485
    },
    {
      "epoch": 0.10689541405476741,
      "grad_norm": 0.26023292541503906,
      "learning_rate": 8e-05,
      "loss": 1.6995,
      "step": 486
    },
    {
      "epoch": 0.10711536346640273,
      "grad_norm": 0.25749459862709045,
      "learning_rate": 8e-05,
      "loss": 1.6614,
      "step": 487
    },
    {
      "epoch": 0.10733531287803805,
      "grad_norm": 0.26305267214775085,
      "learning_rate": 8e-05,
      "loss": 1.6838,
      "step": 488
    },
    {
      "epoch": 0.10755526228967338,
      "grad_norm": 0.25277695059776306,
      "learning_rate": 8e-05,
      "loss": 1.6975,
      "step": 489
    },
    {
      "epoch": 0.1077752117013087,
      "grad_norm": 0.2584420144557953,
      "learning_rate": 8e-05,
      "loss": 1.7434,
      "step": 490
    },
    {
      "epoch": 0.10799516111294402,
      "grad_norm": 0.28107360005378723,
      "learning_rate": 8e-05,
      "loss": 1.8037,
      "step": 491
    },
    {
      "epoch": 0.10821511052457934,
      "grad_norm": 0.553341269493103,
      "learning_rate": 8e-05,
      "loss": 1.8896,
      "step": 492
    },
    {
      "epoch": 0.10843505993621468,
      "grad_norm": 0.2718677222728729,
      "learning_rate": 8e-05,
      "loss": 1.6646,
      "step": 493
    },
    {
      "epoch": 0.10865500934785,
      "grad_norm": 0.27301734685897827,
      "learning_rate": 8e-05,
      "loss": 1.6663,
      "step": 494
    },
    {
      "epoch": 0.10887495875948532,
      "grad_norm": 0.26952439546585083,
      "learning_rate": 8e-05,
      "loss": 1.7228,
      "step": 495
    },
    {
      "epoch": 0.10909490817112064,
      "grad_norm": 0.3017599582672119,
      "learning_rate": 8e-05,
      "loss": 1.7936,
      "step": 496
    },
    {
      "epoch": 0.10931485758275597,
      "grad_norm": 0.2676602303981781,
      "learning_rate": 8e-05,
      "loss": 1.7861,
      "step": 497
    },
    {
      "epoch": 0.10953480699439129,
      "grad_norm": 0.27192267775535583,
      "learning_rate": 8e-05,
      "loss": 1.8032,
      "step": 498
    },
    {
      "epoch": 0.10975475640602661,
      "grad_norm": 0.2807183861732483,
      "learning_rate": 8e-05,
      "loss": 1.6331,
      "step": 499
    },
    {
      "epoch": 0.10997470581766194,
      "grad_norm": 0.2652963399887085,
      "learning_rate": 8e-05,
      "loss": 1.6231,
      "step": 500
    },
    {
      "epoch": 0.11019465522929726,
      "grad_norm": 0.26010751724243164,
      "learning_rate": 8e-05,
      "loss": 1.729,
      "step": 501
    },
    {
      "epoch": 0.11041460464093258,
      "grad_norm": 0.29573148488998413,
      "learning_rate": 8e-05,
      "loss": 1.8082,
      "step": 502
    },
    {
      "epoch": 0.1106345540525679,
      "grad_norm": 0.28008025884628296,
      "learning_rate": 8e-05,
      "loss": 1.6829,
      "step": 503
    },
    {
      "epoch": 0.11085450346420324,
      "grad_norm": 0.3029135763645172,
      "learning_rate": 8e-05,
      "loss": 1.7699,
      "step": 504
    },
    {
      "epoch": 0.11107445287583856,
      "grad_norm": 0.2821674346923828,
      "learning_rate": 8e-05,
      "loss": 1.7337,
      "step": 505
    },
    {
      "epoch": 0.11129440228747388,
      "grad_norm": 0.274880975484848,
      "learning_rate": 8e-05,
      "loss": 1.7973,
      "step": 506
    },
    {
      "epoch": 0.1115143516991092,
      "grad_norm": 0.28885796666145325,
      "learning_rate": 8e-05,
      "loss": 1.7756,
      "step": 507
    },
    {
      "epoch": 0.11173430111074453,
      "grad_norm": 0.2744079530239105,
      "learning_rate": 8e-05,
      "loss": 1.7991,
      "step": 508
    },
    {
      "epoch": 0.11195425052237985,
      "grad_norm": 0.2645000219345093,
      "learning_rate": 8e-05,
      "loss": 1.6566,
      "step": 509
    },
    {
      "epoch": 0.11217419993401517,
      "grad_norm": 0.2640466094017029,
      "learning_rate": 8e-05,
      "loss": 1.6649,
      "step": 510
    },
    {
      "epoch": 0.11239414934565051,
      "grad_norm": 0.2965867817401886,
      "learning_rate": 8e-05,
      "loss": 1.7733,
      "step": 511
    },
    {
      "epoch": 0.11261409875728583,
      "grad_norm": 0.2533203661441803,
      "learning_rate": 8e-05,
      "loss": 1.7194,
      "step": 512
    },
    {
      "epoch": 0.11283404816892115,
      "grad_norm": 0.261994868516922,
      "learning_rate": 8e-05,
      "loss": 1.7387,
      "step": 513
    },
    {
      "epoch": 0.11305399758055647,
      "grad_norm": 0.2868165969848633,
      "learning_rate": 8e-05,
      "loss": 1.7444,
      "step": 514
    },
    {
      "epoch": 0.1132739469921918,
      "grad_norm": 0.2836281657218933,
      "learning_rate": 8e-05,
      "loss": 1.6507,
      "step": 515
    },
    {
      "epoch": 0.11349389640382712,
      "grad_norm": 0.28675276041030884,
      "learning_rate": 8e-05,
      "loss": 1.7054,
      "step": 516
    },
    {
      "epoch": 0.11371384581546244,
      "grad_norm": 0.2745465040206909,
      "learning_rate": 8e-05,
      "loss": 1.77,
      "step": 517
    },
    {
      "epoch": 0.11393379522709776,
      "grad_norm": 0.27250972390174866,
      "learning_rate": 8e-05,
      "loss": 1.9102,
      "step": 518
    },
    {
      "epoch": 0.1141537446387331,
      "grad_norm": 0.2781262695789337,
      "learning_rate": 8e-05,
      "loss": 1.8126,
      "step": 519
    },
    {
      "epoch": 0.11437369405036842,
      "grad_norm": 0.2691183388233185,
      "learning_rate": 8e-05,
      "loss": 1.5978,
      "step": 520
    },
    {
      "epoch": 0.11459364346200374,
      "grad_norm": 0.29496780037879944,
      "learning_rate": 8e-05,
      "loss": 1.9214,
      "step": 521
    },
    {
      "epoch": 0.11481359287363907,
      "grad_norm": 0.27725401520729065,
      "learning_rate": 8e-05,
      "loss": 1.8722,
      "step": 522
    },
    {
      "epoch": 0.11503354228527439,
      "grad_norm": 0.28819364309310913,
      "learning_rate": 8e-05,
      "loss": 1.6739,
      "step": 523
    },
    {
      "epoch": 0.11525349169690971,
      "grad_norm": 0.278857946395874,
      "learning_rate": 8e-05,
      "loss": 1.8137,
      "step": 524
    },
    {
      "epoch": 0.11547344110854503,
      "grad_norm": 0.26911258697509766,
      "learning_rate": 8e-05,
      "loss": 1.7123,
      "step": 525
    },
    {
      "epoch": 0.11569339052018036,
      "grad_norm": 0.2656850814819336,
      "learning_rate": 8e-05,
      "loss": 1.8124,
      "step": 526
    },
    {
      "epoch": 0.11591333993181568,
      "grad_norm": 0.26521819829940796,
      "learning_rate": 8e-05,
      "loss": 1.8188,
      "step": 527
    },
    {
      "epoch": 0.116133289343451,
      "grad_norm": 0.2821720540523529,
      "learning_rate": 8e-05,
      "loss": 1.7607,
      "step": 528
    },
    {
      "epoch": 0.11635323875508632,
      "grad_norm": 0.294612854719162,
      "learning_rate": 8e-05,
      "loss": 1.8142,
      "step": 529
    },
    {
      "epoch": 0.11657318816672166,
      "grad_norm": 0.29858094453811646,
      "learning_rate": 8e-05,
      "loss": 1.8795,
      "step": 530
    },
    {
      "epoch": 0.11679313757835698,
      "grad_norm": 0.2726878821849823,
      "learning_rate": 8e-05,
      "loss": 1.7988,
      "step": 531
    },
    {
      "epoch": 0.1170130869899923,
      "grad_norm": 0.2651258111000061,
      "learning_rate": 8e-05,
      "loss": 1.8106,
      "step": 532
    },
    {
      "epoch": 0.11723303640162762,
      "grad_norm": 0.2681291997432709,
      "learning_rate": 8e-05,
      "loss": 1.6692,
      "step": 533
    },
    {
      "epoch": 0.11745298581326295,
      "grad_norm": 0.2641060948371887,
      "learning_rate": 8e-05,
      "loss": 1.6479,
      "step": 534
    },
    {
      "epoch": 0.11767293522489827,
      "grad_norm": 0.2850191593170166,
      "learning_rate": 8e-05,
      "loss": 1.7337,
      "step": 535
    },
    {
      "epoch": 0.1178928846365336,
      "grad_norm": 0.2718667685985565,
      "learning_rate": 8e-05,
      "loss": 1.7069,
      "step": 536
    },
    {
      "epoch": 0.11811283404816893,
      "grad_norm": 0.27950581908226013,
      "learning_rate": 8e-05,
      "loss": 1.83,
      "step": 537
    },
    {
      "epoch": 0.11833278345980425,
      "grad_norm": 0.26720213890075684,
      "learning_rate": 8e-05,
      "loss": 1.6787,
      "step": 538
    },
    {
      "epoch": 0.11855273287143957,
      "grad_norm": 0.25440508127212524,
      "learning_rate": 8e-05,
      "loss": 1.5966,
      "step": 539
    },
    {
      "epoch": 0.11877268228307489,
      "grad_norm": 0.2716729938983917,
      "learning_rate": 8e-05,
      "loss": 1.793,
      "step": 540
    },
    {
      "epoch": 0.11899263169471022,
      "grad_norm": 0.26204821467399597,
      "learning_rate": 8e-05,
      "loss": 1.5882,
      "step": 541
    },
    {
      "epoch": 0.11921258110634554,
      "grad_norm": 0.2756775915622711,
      "learning_rate": 8e-05,
      "loss": 1.7529,
      "step": 542
    },
    {
      "epoch": 0.11943253051798086,
      "grad_norm": 0.27235740423202515,
      "learning_rate": 8e-05,
      "loss": 1.7607,
      "step": 543
    },
    {
      "epoch": 0.11965247992961618,
      "grad_norm": 0.27712538838386536,
      "learning_rate": 8e-05,
      "loss": 1.7504,
      "step": 544
    },
    {
      "epoch": 0.11987242934125152,
      "grad_norm": 0.27800193428993225,
      "learning_rate": 8e-05,
      "loss": 1.7421,
      "step": 545
    },
    {
      "epoch": 0.12009237875288684,
      "grad_norm": 0.27911701798439026,
      "learning_rate": 8e-05,
      "loss": 1.6683,
      "step": 546
    },
    {
      "epoch": 0.12031232816452216,
      "grad_norm": 0.27643364667892456,
      "learning_rate": 8e-05,
      "loss": 1.6393,
      "step": 547
    },
    {
      "epoch": 0.12053227757615749,
      "grad_norm": 0.25785166025161743,
      "learning_rate": 8e-05,
      "loss": 1.641,
      "step": 548
    },
    {
      "epoch": 0.12075222698779281,
      "grad_norm": 0.2791956067085266,
      "learning_rate": 8e-05,
      "loss": 1.7585,
      "step": 549
    },
    {
      "epoch": 0.12097217639942813,
      "grad_norm": 0.28245967626571655,
      "learning_rate": 8e-05,
      "loss": 1.8716,
      "step": 550
    },
    {
      "epoch": 0.12119212581106345,
      "grad_norm": 0.27160346508026123,
      "learning_rate": 8e-05,
      "loss": 1.7023,
      "step": 551
    },
    {
      "epoch": 0.12141207522269878,
      "grad_norm": 0.2670506536960602,
      "learning_rate": 8e-05,
      "loss": 1.5844,
      "step": 552
    },
    {
      "epoch": 0.1216320246343341,
      "grad_norm": 0.2762441337108612,
      "learning_rate": 8e-05,
      "loss": 1.7286,
      "step": 553
    },
    {
      "epoch": 0.12185197404596942,
      "grad_norm": 0.29608720541000366,
      "learning_rate": 8e-05,
      "loss": 1.7875,
      "step": 554
    },
    {
      "epoch": 0.12207192345760474,
      "grad_norm": 0.2847777307033539,
      "learning_rate": 8e-05,
      "loss": 1.7388,
      "step": 555
    },
    {
      "epoch": 0.12229187286924008,
      "grad_norm": 0.2769443988800049,
      "learning_rate": 8e-05,
      "loss": 1.8129,
      "step": 556
    },
    {
      "epoch": 0.1225118222808754,
      "grad_norm": 0.27490487694740295,
      "learning_rate": 8e-05,
      "loss": 1.678,
      "step": 557
    },
    {
      "epoch": 0.12273177169251072,
      "grad_norm": 0.2851822078227997,
      "learning_rate": 8e-05,
      "loss": 1.8268,
      "step": 558
    },
    {
      "epoch": 0.12295172110414605,
      "grad_norm": 0.31336653232574463,
      "learning_rate": 8e-05,
      "loss": 1.8247,
      "step": 559
    },
    {
      "epoch": 0.12317167051578137,
      "grad_norm": 0.26455923914909363,
      "learning_rate": 8e-05,
      "loss": 1.5548,
      "step": 560
    },
    {
      "epoch": 0.12339161992741669,
      "grad_norm": 0.2750054597854614,
      "learning_rate": 8e-05,
      "loss": 1.7912,
      "step": 561
    },
    {
      "epoch": 0.12361156933905201,
      "grad_norm": 0.28016433119773865,
      "learning_rate": 8e-05,
      "loss": 1.7367,
      "step": 562
    },
    {
      "epoch": 0.12383151875068735,
      "grad_norm": 0.30594533681869507,
      "learning_rate": 8e-05,
      "loss": 1.7959,
      "step": 563
    },
    {
      "epoch": 0.12405146816232267,
      "grad_norm": 0.2753421664237976,
      "learning_rate": 8e-05,
      "loss": 1.6714,
      "step": 564
    },
    {
      "epoch": 0.12427141757395799,
      "grad_norm": 0.3309609889984131,
      "learning_rate": 8e-05,
      "loss": 1.7632,
      "step": 565
    },
    {
      "epoch": 0.12449136698559331,
      "grad_norm": 0.3116569221019745,
      "learning_rate": 8e-05,
      "loss": 1.8312,
      "step": 566
    },
    {
      "epoch": 0.12471131639722864,
      "grad_norm": 0.27756184339523315,
      "learning_rate": 8e-05,
      "loss": 1.6622,
      "step": 567
    },
    {
      "epoch": 0.12493126580886396,
      "grad_norm": 0.2740349769592285,
      "learning_rate": 8e-05,
      "loss": 1.7015,
      "step": 568
    },
    {
      "epoch": 0.1251512152204993,
      "grad_norm": 0.2696126401424408,
      "learning_rate": 8e-05,
      "loss": 1.6063,
      "step": 569
    },
    {
      "epoch": 0.1253711646321346,
      "grad_norm": 0.29191461205482483,
      "learning_rate": 8e-05,
      "loss": 1.8429,
      "step": 570
    },
    {
      "epoch": 0.12559111404376994,
      "grad_norm": 0.2984013855457306,
      "learning_rate": 8e-05,
      "loss": 1.8194,
      "step": 571
    },
    {
      "epoch": 0.12581106345540527,
      "grad_norm": 0.27315613627433777,
      "learning_rate": 8e-05,
      "loss": 1.7027,
      "step": 572
    },
    {
      "epoch": 0.12603101286704058,
      "grad_norm": 0.28547149896621704,
      "learning_rate": 8e-05,
      "loss": 1.694,
      "step": 573
    },
    {
      "epoch": 0.1262509622786759,
      "grad_norm": 0.26458805799484253,
      "learning_rate": 8e-05,
      "loss": 1.7978,
      "step": 574
    },
    {
      "epoch": 0.12647091169031122,
      "grad_norm": 0.29676830768585205,
      "learning_rate": 8e-05,
      "loss": 1.8295,
      "step": 575
    },
    {
      "epoch": 0.12669086110194655,
      "grad_norm": 0.28077611327171326,
      "learning_rate": 8e-05,
      "loss": 1.7711,
      "step": 576
    },
    {
      "epoch": 0.12691081051358188,
      "grad_norm": 0.256736159324646,
      "learning_rate": 8e-05,
      "loss": 1.5371,
      "step": 577
    },
    {
      "epoch": 0.1271307599252172,
      "grad_norm": 0.2888578474521637,
      "learning_rate": 8e-05,
      "loss": 1.7532,
      "step": 578
    },
    {
      "epoch": 0.12735070933685252,
      "grad_norm": 0.29349133372306824,
      "learning_rate": 8e-05,
      "loss": 1.856,
      "step": 579
    },
    {
      "epoch": 0.12757065874848786,
      "grad_norm": 0.2626110911369324,
      "learning_rate": 8e-05,
      "loss": 1.5482,
      "step": 580
    },
    {
      "epoch": 0.12779060816012316,
      "grad_norm": 0.2715248167514801,
      "learning_rate": 8e-05,
      "loss": 1.7003,
      "step": 581
    },
    {
      "epoch": 0.1280105575717585,
      "grad_norm": 0.2800534963607788,
      "learning_rate": 8e-05,
      "loss": 1.7065,
      "step": 582
    },
    {
      "epoch": 0.12823050698339383,
      "grad_norm": 0.3190186619758606,
      "learning_rate": 8e-05,
      "loss": 1.8099,
      "step": 583
    },
    {
      "epoch": 0.12845045639502914,
      "grad_norm": 0.2689470648765564,
      "learning_rate": 8e-05,
      "loss": 1.7824,
      "step": 584
    },
    {
      "epoch": 0.12867040580666447,
      "grad_norm": 0.2715473473072052,
      "learning_rate": 8e-05,
      "loss": 1.7721,
      "step": 585
    },
    {
      "epoch": 0.12889035521829978,
      "grad_norm": 0.27956798672676086,
      "learning_rate": 8e-05,
      "loss": 1.7888,
      "step": 586
    },
    {
      "epoch": 0.1291103046299351,
      "grad_norm": 0.2842330038547516,
      "learning_rate": 8e-05,
      "loss": 1.7131,
      "step": 587
    },
    {
      "epoch": 0.12933025404157045,
      "grad_norm": 0.2888692021369934,
      "learning_rate": 8e-05,
      "loss": 1.7509,
      "step": 588
    },
    {
      "epoch": 0.12955020345320575,
      "grad_norm": 0.27673423290252686,
      "learning_rate": 8e-05,
      "loss": 1.7235,
      "step": 589
    },
    {
      "epoch": 0.1297701528648411,
      "grad_norm": 0.26007330417633057,
      "learning_rate": 8e-05,
      "loss": 1.7157,
      "step": 590
    },
    {
      "epoch": 0.12999010227647642,
      "grad_norm": 0.27521616220474243,
      "learning_rate": 8e-05,
      "loss": 1.7519,
      "step": 591
    },
    {
      "epoch": 0.13021005168811173,
      "grad_norm": 0.2753496766090393,
      "learning_rate": 8e-05,
      "loss": 1.6956,
      "step": 592
    },
    {
      "epoch": 0.13043000109974706,
      "grad_norm": 0.25559505820274353,
      "learning_rate": 8e-05,
      "loss": 1.522,
      "step": 593
    },
    {
      "epoch": 0.1306499505113824,
      "grad_norm": 0.26815375685691833,
      "learning_rate": 8e-05,
      "loss": 1.7658,
      "step": 594
    },
    {
      "epoch": 0.1308698999230177,
      "grad_norm": 0.26870042085647583,
      "learning_rate": 8e-05,
      "loss": 1.779,
      "step": 595
    },
    {
      "epoch": 0.13108984933465304,
      "grad_norm": 0.27346327900886536,
      "learning_rate": 8e-05,
      "loss": 1.7397,
      "step": 596
    },
    {
      "epoch": 0.13130979874628834,
      "grad_norm": 0.26674172282218933,
      "learning_rate": 8e-05,
      "loss": 1.906,
      "step": 597
    },
    {
      "epoch": 0.13152974815792368,
      "grad_norm": 0.266916960477829,
      "learning_rate": 8e-05,
      "loss": 1.6896,
      "step": 598
    },
    {
      "epoch": 0.131749697569559,
      "grad_norm": 0.2620035707950592,
      "learning_rate": 8e-05,
      "loss": 1.8032,
      "step": 599
    },
    {
      "epoch": 0.13196964698119432,
      "grad_norm": 0.2721168100833893,
      "learning_rate": 8e-05,
      "loss": 1.7992,
      "step": 600
    },
    {
      "epoch": 0.13218959639282965,
      "grad_norm": 0.2902929186820984,
      "learning_rate": 8e-05,
      "loss": 1.8392,
      "step": 601
    },
    {
      "epoch": 0.13240954580446498,
      "grad_norm": 0.267459899187088,
      "learning_rate": 8e-05,
      "loss": 1.8469,
      "step": 602
    },
    {
      "epoch": 0.1326294952161003,
      "grad_norm": 0.25643131136894226,
      "learning_rate": 8e-05,
      "loss": 1.5562,
      "step": 603
    },
    {
      "epoch": 0.13284944462773562,
      "grad_norm": 0.2919185757637024,
      "learning_rate": 8e-05,
      "loss": 1.7108,
      "step": 604
    },
    {
      "epoch": 0.13306939403937096,
      "grad_norm": 0.2631925046443939,
      "learning_rate": 8e-05,
      "loss": 1.4344,
      "step": 605
    },
    {
      "epoch": 0.13328934345100626,
      "grad_norm": 0.2710738182067871,
      "learning_rate": 8e-05,
      "loss": 1.6774,
      "step": 606
    },
    {
      "epoch": 0.1335092928626416,
      "grad_norm": 0.2641798257827759,
      "learning_rate": 8e-05,
      "loss": 1.8032,
      "step": 607
    },
    {
      "epoch": 0.1337292422742769,
      "grad_norm": 0.2571311891078949,
      "learning_rate": 8e-05,
      "loss": 1.63,
      "step": 608
    },
    {
      "epoch": 0.13394919168591224,
      "grad_norm": 0.24528057873249054,
      "learning_rate": 8e-05,
      "loss": 1.4576,
      "step": 609
    },
    {
      "epoch": 0.13416914109754757,
      "grad_norm": 0.270641028881073,
      "learning_rate": 8e-05,
      "loss": 1.7896,
      "step": 610
    },
    {
      "epoch": 0.13438909050918288,
      "grad_norm": 0.2723008990287781,
      "learning_rate": 8e-05,
      "loss": 1.7894,
      "step": 611
    },
    {
      "epoch": 0.1346090399208182,
      "grad_norm": 0.26487669348716736,
      "learning_rate": 8e-05,
      "loss": 1.7646,
      "step": 612
    },
    {
      "epoch": 0.13482898933245355,
      "grad_norm": 0.26771143078804016,
      "learning_rate": 8e-05,
      "loss": 1.8015,
      "step": 613
    },
    {
      "epoch": 0.13504893874408885,
      "grad_norm": 0.2585919499397278,
      "learning_rate": 8e-05,
      "loss": 1.6487,
      "step": 614
    },
    {
      "epoch": 0.1352688881557242,
      "grad_norm": 0.28161996603012085,
      "learning_rate": 8e-05,
      "loss": 1.7813,
      "step": 615
    },
    {
      "epoch": 0.1354888375673595,
      "grad_norm": 0.25246456265449524,
      "learning_rate": 8e-05,
      "loss": 1.5549,
      "step": 616
    },
    {
      "epoch": 0.13570878697899483,
      "grad_norm": 0.2803630530834198,
      "learning_rate": 8e-05,
      "loss": 1.7545,
      "step": 617
    },
    {
      "epoch": 0.13592873639063016,
      "grad_norm": 0.2587769031524658,
      "learning_rate": 8e-05,
      "loss": 1.6755,
      "step": 618
    },
    {
      "epoch": 0.13614868580226547,
      "grad_norm": 0.2890148162841797,
      "learning_rate": 8e-05,
      "loss": 1.9753,
      "step": 619
    },
    {
      "epoch": 0.1363686352139008,
      "grad_norm": 0.2924948036670685,
      "learning_rate": 8e-05,
      "loss": 1.7611,
      "step": 620
    },
    {
      "epoch": 0.13658858462553614,
      "grad_norm": 0.2594594359397888,
      "learning_rate": 8e-05,
      "loss": 1.6945,
      "step": 621
    },
    {
      "epoch": 0.13680853403717144,
      "grad_norm": 0.2853068709373474,
      "learning_rate": 8e-05,
      "loss": 1.8637,
      "step": 622
    },
    {
      "epoch": 0.13702848344880678,
      "grad_norm": 0.2696111798286438,
      "learning_rate": 8e-05,
      "loss": 1.7777,
      "step": 623
    },
    {
      "epoch": 0.1372484328604421,
      "grad_norm": 0.3137861490249634,
      "learning_rate": 8e-05,
      "loss": 1.8799,
      "step": 624
    },
    {
      "epoch": 0.13746838227207742,
      "grad_norm": 0.25645750761032104,
      "learning_rate": 8e-05,
      "loss": 1.5023,
      "step": 625
    },
    {
      "epoch": 0.13768833168371275,
      "grad_norm": 0.29853489995002747,
      "learning_rate": 8e-05,
      "loss": 1.9131,
      "step": 626
    },
    {
      "epoch": 0.13790828109534806,
      "grad_norm": 0.2653225362300873,
      "learning_rate": 8e-05,
      "loss": 1.6835,
      "step": 627
    },
    {
      "epoch": 0.1381282305069834,
      "grad_norm": 0.26686328649520874,
      "learning_rate": 8e-05,
      "loss": 1.7667,
      "step": 628
    },
    {
      "epoch": 0.13834817991861872,
      "grad_norm": 0.26114073395729065,
      "learning_rate": 8e-05,
      "loss": 1.6925,
      "step": 629
    },
    {
      "epoch": 0.13856812933025403,
      "grad_norm": 0.2520682215690613,
      "learning_rate": 8e-05,
      "loss": 1.6065,
      "step": 630
    },
    {
      "epoch": 0.13878807874188936,
      "grad_norm": 0.2676456868648529,
      "learning_rate": 8e-05,
      "loss": 1.7353,
      "step": 631
    },
    {
      "epoch": 0.1390080281535247,
      "grad_norm": 0.2525452673435211,
      "learning_rate": 8e-05,
      "loss": 1.5993,
      "step": 632
    },
    {
      "epoch": 0.13922797756516,
      "grad_norm": 0.25620371103286743,
      "learning_rate": 8e-05,
      "loss": 1.7188,
      "step": 633
    },
    {
      "epoch": 0.13944792697679534,
      "grad_norm": 0.4071904420852661,
      "learning_rate": 8e-05,
      "loss": 1.9348,
      "step": 634
    },
    {
      "epoch": 0.13966787638843067,
      "grad_norm": 0.2656376361846924,
      "learning_rate": 8e-05,
      "loss": 1.7833,
      "step": 635
    },
    {
      "epoch": 0.13988782580006598,
      "grad_norm": 0.25558993220329285,
      "learning_rate": 8e-05,
      "loss": 1.715,
      "step": 636
    },
    {
      "epoch": 0.1401077752117013,
      "grad_norm": 0.28318601846694946,
      "learning_rate": 8e-05,
      "loss": 1.8012,
      "step": 637
    },
    {
      "epoch": 0.14032772462333662,
      "grad_norm": 0.2558564245700836,
      "learning_rate": 8e-05,
      "loss": 1.5802,
      "step": 638
    },
    {
      "epoch": 0.14054767403497195,
      "grad_norm": 0.26874974370002747,
      "learning_rate": 8e-05,
      "loss": 1.7884,
      "step": 639
    },
    {
      "epoch": 0.1407676234466073,
      "grad_norm": 0.2960795760154724,
      "learning_rate": 8e-05,
      "loss": 1.7884,
      "step": 640
    },
    {
      "epoch": 0.1409875728582426,
      "grad_norm": 0.3098964989185333,
      "learning_rate": 8e-05,
      "loss": 1.8844,
      "step": 641
    },
    {
      "epoch": 0.14120752226987793,
      "grad_norm": 0.2819165885448456,
      "learning_rate": 8e-05,
      "loss": 1.7111,
      "step": 642
    },
    {
      "epoch": 0.14142747168151326,
      "grad_norm": 0.26352617144584656,
      "learning_rate": 8e-05,
      "loss": 1.7337,
      "step": 643
    },
    {
      "epoch": 0.14164742109314857,
      "grad_norm": 0.2622654139995575,
      "learning_rate": 8e-05,
      "loss": 1.7284,
      "step": 644
    },
    {
      "epoch": 0.1418673705047839,
      "grad_norm": 0.2793010473251343,
      "learning_rate": 8e-05,
      "loss": 1.8534,
      "step": 645
    },
    {
      "epoch": 0.14208731991641924,
      "grad_norm": 0.27972397208213806,
      "learning_rate": 8e-05,
      "loss": 1.7658,
      "step": 646
    },
    {
      "epoch": 0.14230726932805454,
      "grad_norm": 0.25940972566604614,
      "learning_rate": 8e-05,
      "loss": 1.6676,
      "step": 647
    },
    {
      "epoch": 0.14252721873968988,
      "grad_norm": 0.29578897356987,
      "learning_rate": 8e-05,
      "loss": 1.8002,
      "step": 648
    },
    {
      "epoch": 0.14274716815132518,
      "grad_norm": 0.2577681541442871,
      "learning_rate": 8e-05,
      "loss": 1.6154,
      "step": 649
    },
    {
      "epoch": 0.14296711756296052,
      "grad_norm": 0.2615002989768982,
      "learning_rate": 8e-05,
      "loss": 1.7539,
      "step": 650
    },
    {
      "epoch": 0.14318706697459585,
      "grad_norm": 0.26044437289237976,
      "learning_rate": 8e-05,
      "loss": 1.5284,
      "step": 651
    },
    {
      "epoch": 0.14340701638623116,
      "grad_norm": 0.28386443853378296,
      "learning_rate": 8e-05,
      "loss": 1.7188,
      "step": 652
    },
    {
      "epoch": 0.1436269657978665,
      "grad_norm": 0.2579086124897003,
      "learning_rate": 8e-05,
      "loss": 1.6758,
      "step": 653
    },
    {
      "epoch": 0.14384691520950182,
      "grad_norm": 0.263192743062973,
      "learning_rate": 8e-05,
      "loss": 1.7013,
      "step": 654
    },
    {
      "epoch": 0.14406686462113713,
      "grad_norm": 0.26551106572151184,
      "learning_rate": 8e-05,
      "loss": 1.7314,
      "step": 655
    },
    {
      "epoch": 0.14428681403277246,
      "grad_norm": 0.26143091917037964,
      "learning_rate": 8e-05,
      "loss": 1.7041,
      "step": 656
    },
    {
      "epoch": 0.1445067634444078,
      "grad_norm": 0.26432663202285767,
      "learning_rate": 8e-05,
      "loss": 1.601,
      "step": 657
    },
    {
      "epoch": 0.1447267128560431,
      "grad_norm": 0.2831920087337494,
      "learning_rate": 8e-05,
      "loss": 1.8573,
      "step": 658
    },
    {
      "epoch": 0.14494666226767844,
      "grad_norm": 0.3045855462551117,
      "learning_rate": 8e-05,
      "loss": 1.7853,
      "step": 659
    },
    {
      "epoch": 0.14516661167931375,
      "grad_norm": 0.28249257802963257,
      "learning_rate": 8e-05,
      "loss": 1.7525,
      "step": 660
    },
    {
      "epoch": 0.14538656109094908,
      "grad_norm": 0.27501189708709717,
      "learning_rate": 8e-05,
      "loss": 1.6939,
      "step": 661
    },
    {
      "epoch": 0.1456065105025844,
      "grad_norm": 0.28419750928878784,
      "learning_rate": 8e-05,
      "loss": 1.837,
      "step": 662
    },
    {
      "epoch": 0.14582645991421972,
      "grad_norm": 0.28872454166412354,
      "learning_rate": 8e-05,
      "loss": 1.623,
      "step": 663
    },
    {
      "epoch": 0.14604640932585505,
      "grad_norm": 0.2926316559314728,
      "learning_rate": 8e-05,
      "loss": 1.7438,
      "step": 664
    },
    {
      "epoch": 0.1462663587374904,
      "grad_norm": 0.2716543972492218,
      "learning_rate": 8e-05,
      "loss": 1.8925,
      "step": 665
    },
    {
      "epoch": 0.1464863081491257,
      "grad_norm": 0.2707289159297943,
      "learning_rate": 8e-05,
      "loss": 1.8218,
      "step": 666
    },
    {
      "epoch": 0.14670625756076103,
      "grad_norm": 0.2609579265117645,
      "learning_rate": 8e-05,
      "loss": 1.4612,
      "step": 667
    },
    {
      "epoch": 0.14692620697239636,
      "grad_norm": 0.2958548367023468,
      "learning_rate": 8e-05,
      "loss": 1.6191,
      "step": 668
    },
    {
      "epoch": 0.14714615638403167,
      "grad_norm": 0.2585492730140686,
      "learning_rate": 8e-05,
      "loss": 1.7161,
      "step": 669
    },
    {
      "epoch": 0.147366105795667,
      "grad_norm": 0.2637808322906494,
      "learning_rate": 8e-05,
      "loss": 1.6534,
      "step": 670
    },
    {
      "epoch": 0.1475860552073023,
      "grad_norm": 0.2885671854019165,
      "learning_rate": 8e-05,
      "loss": 1.7663,
      "step": 671
    },
    {
      "epoch": 0.14780600461893764,
      "grad_norm": 0.27028244733810425,
      "learning_rate": 8e-05,
      "loss": 1.7718,
      "step": 672
    },
    {
      "epoch": 0.14802595403057298,
      "grad_norm": 0.27723586559295654,
      "learning_rate": 8e-05,
      "loss": 1.7762,
      "step": 673
    },
    {
      "epoch": 0.14824590344220828,
      "grad_norm": 0.26336848735809326,
      "learning_rate": 8e-05,
      "loss": 1.6114,
      "step": 674
    },
    {
      "epoch": 0.14846585285384362,
      "grad_norm": 0.26031750440597534,
      "learning_rate": 8e-05,
      "loss": 1.7259,
      "step": 675
    },
    {
      "epoch": 0.14868580226547895,
      "grad_norm": 0.30176040530204773,
      "learning_rate": 8e-05,
      "loss": 1.7007,
      "step": 676
    },
    {
      "epoch": 0.14890575167711426,
      "grad_norm": 0.25952771306037903,
      "learning_rate": 8e-05,
      "loss": 1.6573,
      "step": 677
    },
    {
      "epoch": 0.1491257010887496,
      "grad_norm": 0.2727009356021881,
      "learning_rate": 8e-05,
      "loss": 1.7725,
      "step": 678
    },
    {
      "epoch": 0.14934565050038492,
      "grad_norm": 0.26398420333862305,
      "learning_rate": 8e-05,
      "loss": 1.7245,
      "step": 679
    },
    {
      "epoch": 0.14956559991202023,
      "grad_norm": 0.273967981338501,
      "learning_rate": 8e-05,
      "loss": 1.7231,
      "step": 680
    },
    {
      "epoch": 0.14978554932365556,
      "grad_norm": 0.27241724729537964,
      "learning_rate": 8e-05,
      "loss": 1.6896,
      "step": 681
    },
    {
      "epoch": 0.15000549873529087,
      "grad_norm": 0.26996085047721863,
      "learning_rate": 8e-05,
      "loss": 1.6767,
      "step": 682
    },
    {
      "epoch": 0.1502254481469262,
      "grad_norm": 0.27165672183036804,
      "learning_rate": 8e-05,
      "loss": 1.7747,
      "step": 683
    },
    {
      "epoch": 0.15044539755856154,
      "grad_norm": 0.26840028166770935,
      "learning_rate": 8e-05,
      "loss": 1.7616,
      "step": 684
    },
    {
      "epoch": 0.15066534697019685,
      "grad_norm": 0.27101555466651917,
      "learning_rate": 8e-05,
      "loss": 1.622,
      "step": 685
    },
    {
      "epoch": 0.15088529638183218,
      "grad_norm": 0.2691043019294739,
      "learning_rate": 8e-05,
      "loss": 1.7514,
      "step": 686
    },
    {
      "epoch": 0.1511052457934675,
      "grad_norm": 0.2926357090473175,
      "learning_rate": 8e-05,
      "loss": 1.6953,
      "step": 687
    },
    {
      "epoch": 0.15132519520510282,
      "grad_norm": 0.2730226516723633,
      "learning_rate": 8e-05,
      "loss": 1.6286,
      "step": 688
    },
    {
      "epoch": 0.15154514461673815,
      "grad_norm": 0.2618841826915741,
      "learning_rate": 8e-05,
      "loss": 1.7194,
      "step": 689
    },
    {
      "epoch": 0.1517650940283735,
      "grad_norm": 0.2584119737148285,
      "learning_rate": 8e-05,
      "loss": 1.6032,
      "step": 690
    },
    {
      "epoch": 0.1519850434400088,
      "grad_norm": 0.26063093543052673,
      "learning_rate": 8e-05,
      "loss": 1.63,
      "step": 691
    },
    {
      "epoch": 0.15220499285164413,
      "grad_norm": 0.267938494682312,
      "learning_rate": 8e-05,
      "loss": 1.7087,
      "step": 692
    },
    {
      "epoch": 0.15242494226327943,
      "grad_norm": 0.2709169089794159,
      "learning_rate": 8e-05,
      "loss": 1.6663,
      "step": 693
    },
    {
      "epoch": 0.15264489167491477,
      "grad_norm": 0.3015836775302887,
      "learning_rate": 8e-05,
      "loss": 1.6797,
      "step": 694
    },
    {
      "epoch": 0.1528648410865501,
      "grad_norm": 0.27824944257736206,
      "learning_rate": 8e-05,
      "loss": 1.7972,
      "step": 695
    },
    {
      "epoch": 0.1530847904981854,
      "grad_norm": 0.31089073419570923,
      "learning_rate": 8e-05,
      "loss": 1.7352,
      "step": 696
    },
    {
      "epoch": 0.15330473990982074,
      "grad_norm": 0.2804546654224396,
      "learning_rate": 8e-05,
      "loss": 1.6898,
      "step": 697
    },
    {
      "epoch": 0.15352468932145608,
      "grad_norm": 0.2804514765739441,
      "learning_rate": 8e-05,
      "loss": 1.8409,
      "step": 698
    },
    {
      "epoch": 0.15374463873309138,
      "grad_norm": 0.31666815280914307,
      "learning_rate": 8e-05,
      "loss": 1.6569,
      "step": 699
    },
    {
      "epoch": 0.15396458814472672,
      "grad_norm": 0.2846215069293976,
      "learning_rate": 8e-05,
      "loss": 1.8081,
      "step": 700
    },
    {
      "epoch": 0.15418453755636205,
      "grad_norm": 0.2656068801879883,
      "learning_rate": 8e-05,
      "loss": 1.5747,
      "step": 701
    },
    {
      "epoch": 0.15440448696799736,
      "grad_norm": 0.2633317708969116,
      "learning_rate": 8e-05,
      "loss": 1.6027,
      "step": 702
    },
    {
      "epoch": 0.1546244363796327,
      "grad_norm": 0.2669740319252014,
      "learning_rate": 8e-05,
      "loss": 1.6964,
      "step": 703
    },
    {
      "epoch": 0.154844385791268,
      "grad_norm": 0.2878497540950775,
      "learning_rate": 8e-05,
      "loss": 1.677,
      "step": 704
    },
    {
      "epoch": 0.15506433520290333,
      "grad_norm": 0.2624325156211853,
      "learning_rate": 8e-05,
      "loss": 1.6247,
      "step": 705
    },
    {
      "epoch": 0.15528428461453866,
      "grad_norm": 0.2894291579723358,
      "learning_rate": 8e-05,
      "loss": 1.7271,
      "step": 706
    },
    {
      "epoch": 0.15550423402617397,
      "grad_norm": 0.2924456298351288,
      "learning_rate": 8e-05,
      "loss": 1.7475,
      "step": 707
    },
    {
      "epoch": 0.1557241834378093,
      "grad_norm": 0.2519112229347229,
      "learning_rate": 8e-05,
      "loss": 1.6306,
      "step": 708
    },
    {
      "epoch": 0.15594413284944464,
      "grad_norm": 0.2831405699253082,
      "learning_rate": 8e-05,
      "loss": 1.7571,
      "step": 709
    },
    {
      "epoch": 0.15616408226107995,
      "grad_norm": 0.2804257273674011,
      "learning_rate": 8e-05,
      "loss": 1.6721,
      "step": 710
    },
    {
      "epoch": 0.15638403167271528,
      "grad_norm": 0.27130362391471863,
      "learning_rate": 8e-05,
      "loss": 1.7451,
      "step": 711
    },
    {
      "epoch": 0.1566039810843506,
      "grad_norm": 0.27843937277793884,
      "learning_rate": 8e-05,
      "loss": 1.7187,
      "step": 712
    },
    {
      "epoch": 0.15682393049598592,
      "grad_norm": 0.26205387711524963,
      "learning_rate": 8e-05,
      "loss": 1.7667,
      "step": 713
    },
    {
      "epoch": 0.15704387990762125,
      "grad_norm": 0.25978967547416687,
      "learning_rate": 8e-05,
      "loss": 1.6595,
      "step": 714
    },
    {
      "epoch": 0.15726382931925656,
      "grad_norm": 0.26331478357315063,
      "learning_rate": 8e-05,
      "loss": 1.8067,
      "step": 715
    },
    {
      "epoch": 0.1574837787308919,
      "grad_norm": 0.26023924350738525,
      "learning_rate": 8e-05,
      "loss": 1.8533,
      "step": 716
    },
    {
      "epoch": 0.15770372814252723,
      "grad_norm": 0.27147844433784485,
      "learning_rate": 8e-05,
      "loss": 1.6309,
      "step": 717
    },
    {
      "epoch": 0.15792367755416253,
      "grad_norm": 0.286035418510437,
      "learning_rate": 8e-05,
      "loss": 1.72,
      "step": 718
    },
    {
      "epoch": 0.15814362696579787,
      "grad_norm": 0.3167229890823364,
      "learning_rate": 8e-05,
      "loss": 1.9007,
      "step": 719
    },
    {
      "epoch": 0.1583635763774332,
      "grad_norm": 0.283975750207901,
      "learning_rate": 8e-05,
      "loss": 1.6662,
      "step": 720
    },
    {
      "epoch": 0.1585835257890685,
      "grad_norm": 0.2812137007713318,
      "learning_rate": 8e-05,
      "loss": 1.7651,
      "step": 721
    },
    {
      "epoch": 0.15880347520070384,
      "grad_norm": 0.2737642526626587,
      "learning_rate": 8e-05,
      "loss": 1.7679,
      "step": 722
    },
    {
      "epoch": 0.15902342461233915,
      "grad_norm": 0.30812978744506836,
      "learning_rate": 8e-05,
      "loss": 1.8408,
      "step": 723
    },
    {
      "epoch": 0.15924337402397448,
      "grad_norm": 0.27026352286338806,
      "learning_rate": 8e-05,
      "loss": 1.7362,
      "step": 724
    },
    {
      "epoch": 0.15946332343560982,
      "grad_norm": 0.2788861393928528,
      "learning_rate": 8e-05,
      "loss": 1.8371,
      "step": 725
    },
    {
      "epoch": 0.15968327284724512,
      "grad_norm": 0.2623996138572693,
      "learning_rate": 8e-05,
      "loss": 1.5855,
      "step": 726
    },
    {
      "epoch": 0.15990322225888046,
      "grad_norm": 0.2764820158481598,
      "learning_rate": 8e-05,
      "loss": 1.8185,
      "step": 727
    },
    {
      "epoch": 0.1601231716705158,
      "grad_norm": 0.27394816279411316,
      "learning_rate": 8e-05,
      "loss": 1.641,
      "step": 728
    },
    {
      "epoch": 0.1603431210821511,
      "grad_norm": 0.2726307511329651,
      "learning_rate": 8e-05,
      "loss": 1.6128,
      "step": 729
    },
    {
      "epoch": 0.16056307049378643,
      "grad_norm": 0.28221258521080017,
      "learning_rate": 8e-05,
      "loss": 1.8413,
      "step": 730
    },
    {
      "epoch": 0.16078301990542176,
      "grad_norm": 0.2649543881416321,
      "learning_rate": 8e-05,
      "loss": 1.5707,
      "step": 731
    },
    {
      "epoch": 0.16100296931705707,
      "grad_norm": 0.2659435570240021,
      "learning_rate": 8e-05,
      "loss": 1.6761,
      "step": 732
    },
    {
      "epoch": 0.1612229187286924,
      "grad_norm": 0.3131570518016815,
      "learning_rate": 8e-05,
      "loss": 1.9439,
      "step": 733
    },
    {
      "epoch": 0.1614428681403277,
      "grad_norm": 0.263069748878479,
      "learning_rate": 8e-05,
      "loss": 1.7069,
      "step": 734
    },
    {
      "epoch": 0.16166281755196305,
      "grad_norm": 0.2708505392074585,
      "learning_rate": 8e-05,
      "loss": 1.8031,
      "step": 735
    },
    {
      "epoch": 0.16188276696359838,
      "grad_norm": 0.26446613669395447,
      "learning_rate": 8e-05,
      "loss": 1.6419,
      "step": 736
    },
    {
      "epoch": 0.16210271637523369,
      "grad_norm": 0.27720367908477783,
      "learning_rate": 8e-05,
      "loss": 1.8291,
      "step": 737
    },
    {
      "epoch": 0.16232266578686902,
      "grad_norm": 0.25950226187705994,
      "learning_rate": 8e-05,
      "loss": 1.7498,
      "step": 738
    },
    {
      "epoch": 0.16254261519850435,
      "grad_norm": 0.25445327162742615,
      "learning_rate": 8e-05,
      "loss": 1.6804,
      "step": 739
    },
    {
      "epoch": 0.16276256461013966,
      "grad_norm": 0.2868766784667969,
      "learning_rate": 8e-05,
      "loss": 1.8058,
      "step": 740
    },
    {
      "epoch": 0.162982514021775,
      "grad_norm": 0.2775559425354004,
      "learning_rate": 8e-05,
      "loss": 1.7971,
      "step": 741
    },
    {
      "epoch": 0.16320246343341033,
      "grad_norm": 0.2822381556034088,
      "learning_rate": 8e-05,
      "loss": 1.7294,
      "step": 742
    },
    {
      "epoch": 0.16342241284504563,
      "grad_norm": 0.26617857813835144,
      "learning_rate": 8e-05,
      "loss": 1.8011,
      "step": 743
    },
    {
      "epoch": 0.16364236225668097,
      "grad_norm": 0.25615090131759644,
      "learning_rate": 8e-05,
      "loss": 1.6328,
      "step": 744
    },
    {
      "epoch": 0.16386231166831627,
      "grad_norm": 0.25831338763237,
      "learning_rate": 8e-05,
      "loss": 1.6174,
      "step": 745
    },
    {
      "epoch": 0.1640822610799516,
      "grad_norm": 0.2707291543483734,
      "learning_rate": 8e-05,
      "loss": 1.8217,
      "step": 746
    },
    {
      "epoch": 0.16430221049158694,
      "grad_norm": 0.3028862774372101,
      "learning_rate": 8e-05,
      "loss": 1.5852,
      "step": 747
    },
    {
      "epoch": 0.16452215990322225,
      "grad_norm": 0.26598575711250305,
      "learning_rate": 8e-05,
      "loss": 1.7213,
      "step": 748
    },
    {
      "epoch": 0.16474210931485758,
      "grad_norm": 0.27408871054649353,
      "learning_rate": 8e-05,
      "loss": 1.7109,
      "step": 749
    },
    {
      "epoch": 0.16496205872649292,
      "grad_norm": 0.27065837383270264,
      "learning_rate": 8e-05,
      "loss": 1.6696,
      "step": 750
    },
    {
      "epoch": 0.16518200813812822,
      "grad_norm": 0.2721879184246063,
      "learning_rate": 8e-05,
      "loss": 1.7055,
      "step": 751
    },
    {
      "epoch": 0.16540195754976356,
      "grad_norm": 0.29569125175476074,
      "learning_rate": 8e-05,
      "loss": 1.5921,
      "step": 752
    },
    {
      "epoch": 0.1656219069613989,
      "grad_norm": 0.28580978512763977,
      "learning_rate": 8e-05,
      "loss": 1.7518,
      "step": 753
    },
    {
      "epoch": 0.1658418563730342,
      "grad_norm": 0.2869469225406647,
      "learning_rate": 8e-05,
      "loss": 1.8164,
      "step": 754
    },
    {
      "epoch": 0.16606180578466953,
      "grad_norm": 0.2796071171760559,
      "learning_rate": 8e-05,
      "loss": 1.8325,
      "step": 755
    },
    {
      "epoch": 0.16628175519630484,
      "grad_norm": 0.27365031838417053,
      "learning_rate": 8e-05,
      "loss": 1.7287,
      "step": 756
    },
    {
      "epoch": 0.16650170460794017,
      "grad_norm": 0.2524491846561432,
      "learning_rate": 8e-05,
      "loss": 1.5379,
      "step": 757
    },
    {
      "epoch": 0.1667216540195755,
      "grad_norm": 0.259860634803772,
      "learning_rate": 8e-05,
      "loss": 1.5204,
      "step": 758
    },
    {
      "epoch": 0.1669416034312108,
      "grad_norm": 0.2714100182056427,
      "learning_rate": 8e-05,
      "loss": 1.7245,
      "step": 759
    },
    {
      "epoch": 0.16716155284284615,
      "grad_norm": 0.2729417383670807,
      "learning_rate": 8e-05,
      "loss": 1.6889,
      "step": 760
    },
    {
      "epoch": 0.16738150225448148,
      "grad_norm": 0.2753896415233612,
      "learning_rate": 8e-05,
      "loss": 1.7345,
      "step": 761
    },
    {
      "epoch": 0.16760145166611679,
      "grad_norm": 0.2830727994441986,
      "learning_rate": 8e-05,
      "loss": 1.6884,
      "step": 762
    },
    {
      "epoch": 0.16782140107775212,
      "grad_norm": 0.27818116545677185,
      "learning_rate": 8e-05,
      "loss": 1.7819,
      "step": 763
    },
    {
      "epoch": 0.16804135048938745,
      "grad_norm": 0.2601570785045624,
      "learning_rate": 8e-05,
      "loss": 1.6323,
      "step": 764
    },
    {
      "epoch": 0.16826129990102276,
      "grad_norm": 0.2638706564903259,
      "learning_rate": 8e-05,
      "loss": 1.5957,
      "step": 765
    },
    {
      "epoch": 0.1684812493126581,
      "grad_norm": 0.2798631489276886,
      "learning_rate": 8e-05,
      "loss": 1.7946,
      "step": 766
    },
    {
      "epoch": 0.1687011987242934,
      "grad_norm": 0.2975100874900818,
      "learning_rate": 8e-05,
      "loss": 1.871,
      "step": 767
    },
    {
      "epoch": 0.16892114813592873,
      "grad_norm": 0.28308364748954773,
      "learning_rate": 8e-05,
      "loss": 1.7184,
      "step": 768
    },
    {
      "epoch": 0.16914109754756407,
      "grad_norm": 0.2594911456108093,
      "learning_rate": 8e-05,
      "loss": 1.5867,
      "step": 769
    },
    {
      "epoch": 0.16936104695919937,
      "grad_norm": 0.27594470977783203,
      "learning_rate": 8e-05,
      "loss": 1.7722,
      "step": 770
    },
    {
      "epoch": 0.1695809963708347,
      "grad_norm": 0.2783298194408417,
      "learning_rate": 8e-05,
      "loss": 1.7891,
      "step": 771
    },
    {
      "epoch": 0.16980094578247004,
      "grad_norm": 0.2863733172416687,
      "learning_rate": 8e-05,
      "loss": 1.6274,
      "step": 772
    },
    {
      "epoch": 0.17002089519410535,
      "grad_norm": 0.27953147888183594,
      "learning_rate": 8e-05,
      "loss": 1.7287,
      "step": 773
    },
    {
      "epoch": 0.17024084460574068,
      "grad_norm": 0.2736772894859314,
      "learning_rate": 8e-05,
      "loss": 1.6802,
      "step": 774
    },
    {
      "epoch": 0.17046079401737602,
      "grad_norm": 0.27663713693618774,
      "learning_rate": 8e-05,
      "loss": 1.6607,
      "step": 775
    },
    {
      "epoch": 0.17068074342901132,
      "grad_norm": 0.3064086437225342,
      "learning_rate": 8e-05,
      "loss": 1.8244,
      "step": 776
    },
    {
      "epoch": 0.17090069284064666,
      "grad_norm": 0.29848581552505493,
      "learning_rate": 8e-05,
      "loss": 1.7702,
      "step": 777
    },
    {
      "epoch": 0.17112064225228196,
      "grad_norm": 0.3101220726966858,
      "learning_rate": 8e-05,
      "loss": 1.7714,
      "step": 778
    },
    {
      "epoch": 0.1713405916639173,
      "grad_norm": 0.2754581868648529,
      "learning_rate": 8e-05,
      "loss": 1.6367,
      "step": 779
    },
    {
      "epoch": 0.17156054107555263,
      "grad_norm": 0.2706362307071686,
      "learning_rate": 8e-05,
      "loss": 1.6236,
      "step": 780
    },
    {
      "epoch": 0.17178049048718794,
      "grad_norm": 0.29135438799858093,
      "learning_rate": 8e-05,
      "loss": 1.8478,
      "step": 781
    },
    {
      "epoch": 0.17200043989882327,
      "grad_norm": 0.2751868963241577,
      "learning_rate": 8e-05,
      "loss": 1.751,
      "step": 782
    },
    {
      "epoch": 0.1722203893104586,
      "grad_norm": 0.2871004045009613,
      "learning_rate": 8e-05,
      "loss": 1.6793,
      "step": 783
    },
    {
      "epoch": 0.1724403387220939,
      "grad_norm": 0.31024861335754395,
      "learning_rate": 8e-05,
      "loss": 1.7419,
      "step": 784
    },
    {
      "epoch": 0.17266028813372924,
      "grad_norm": 0.2917722165584564,
      "learning_rate": 8e-05,
      "loss": 1.8913,
      "step": 785
    },
    {
      "epoch": 0.17288023754536458,
      "grad_norm": 0.25443291664123535,
      "learning_rate": 8e-05,
      "loss": 1.6991,
      "step": 786
    },
    {
      "epoch": 0.17310018695699989,
      "grad_norm": 0.2827921211719513,
      "learning_rate": 8e-05,
      "loss": 1.8408,
      "step": 787
    },
    {
      "epoch": 0.17332013636863522,
      "grad_norm": 0.26190435886383057,
      "learning_rate": 8e-05,
      "loss": 1.6841,
      "step": 788
    },
    {
      "epoch": 0.17354008578027053,
      "grad_norm": 0.31557098031044006,
      "learning_rate": 8e-05,
      "loss": 1.8838,
      "step": 789
    },
    {
      "epoch": 0.17376003519190586,
      "grad_norm": 0.27622002363204956,
      "learning_rate": 8e-05,
      "loss": 1.6558,
      "step": 790
    },
    {
      "epoch": 0.1739799846035412,
      "grad_norm": 0.3161294758319855,
      "learning_rate": 8e-05,
      "loss": 1.5771,
      "step": 791
    },
    {
      "epoch": 0.1741999340151765,
      "grad_norm": 0.3014603555202484,
      "learning_rate": 8e-05,
      "loss": 1.7742,
      "step": 792
    },
    {
      "epoch": 0.17441988342681183,
      "grad_norm": 0.24996457993984222,
      "learning_rate": 8e-05,
      "loss": 1.5667,
      "step": 793
    },
    {
      "epoch": 0.17463983283844717,
      "grad_norm": 0.29180648922920227,
      "learning_rate": 8e-05,
      "loss": 1.7703,
      "step": 794
    },
    {
      "epoch": 0.17485978225008247,
      "grad_norm": 0.26707547903060913,
      "learning_rate": 8e-05,
      "loss": 1.7964,
      "step": 795
    },
    {
      "epoch": 0.1750797316617178,
      "grad_norm": 0.24924349784851074,
      "learning_rate": 8e-05,
      "loss": 1.6619,
      "step": 796
    },
    {
      "epoch": 0.17529968107335314,
      "grad_norm": 0.29872292280197144,
      "learning_rate": 8e-05,
      "loss": 1.8561,
      "step": 797
    },
    {
      "epoch": 0.17551963048498845,
      "grad_norm": 0.2770175337791443,
      "learning_rate": 8e-05,
      "loss": 1.6352,
      "step": 798
    },
    {
      "epoch": 0.17573957989662378,
      "grad_norm": 0.26890453696250916,
      "learning_rate": 8e-05,
      "loss": 1.885,
      "step": 799
    },
    {
      "epoch": 0.1759595293082591,
      "grad_norm": 0.2830483317375183,
      "learning_rate": 8e-05,
      "loss": 1.6029,
      "step": 800
    },
    {
      "epoch": 0.17617947871989442,
      "grad_norm": 0.27421921491622925,
      "learning_rate": 8e-05,
      "loss": 1.6845,
      "step": 801
    },
    {
      "epoch": 0.17639942813152976,
      "grad_norm": 0.29273220896720886,
      "learning_rate": 8e-05,
      "loss": 1.8135,
      "step": 802
    },
    {
      "epoch": 0.17661937754316506,
      "grad_norm": 0.2675575315952301,
      "learning_rate": 8e-05,
      "loss": 1.571,
      "step": 803
    },
    {
      "epoch": 0.1768393269548004,
      "grad_norm": 0.2821138799190521,
      "learning_rate": 8e-05,
      "loss": 1.9244,
      "step": 804
    },
    {
      "epoch": 0.17705927636643573,
      "grad_norm": 0.28082311153411865,
      "learning_rate": 8e-05,
      "loss": 1.7395,
      "step": 805
    },
    {
      "epoch": 0.17727922577807104,
      "grad_norm": 0.27897313237190247,
      "learning_rate": 8e-05,
      "loss": 1.6347,
      "step": 806
    },
    {
      "epoch": 0.17749917518970637,
      "grad_norm": 0.27358707785606384,
      "learning_rate": 8e-05,
      "loss": 1.7643,
      "step": 807
    },
    {
      "epoch": 0.1777191246013417,
      "grad_norm": 0.284059077501297,
      "learning_rate": 8e-05,
      "loss": 1.5789,
      "step": 808
    },
    {
      "epoch": 0.177939074012977,
      "grad_norm": 0.26125824451446533,
      "learning_rate": 8e-05,
      "loss": 1.7029,
      "step": 809
    },
    {
      "epoch": 0.17815902342461234,
      "grad_norm": 0.26438888907432556,
      "learning_rate": 8e-05,
      "loss": 1.6424,
      "step": 810
    },
    {
      "epoch": 0.17837897283624765,
      "grad_norm": 0.2746163010597229,
      "learning_rate": 8e-05,
      "loss": 1.7992,
      "step": 811
    },
    {
      "epoch": 0.17859892224788299,
      "grad_norm": 0.27717527747154236,
      "learning_rate": 8e-05,
      "loss": 1.7603,
      "step": 812
    },
    {
      "epoch": 0.17881887165951832,
      "grad_norm": 0.28336596488952637,
      "learning_rate": 8e-05,
      "loss": 1.7133,
      "step": 813
    },
    {
      "epoch": 0.17903882107115363,
      "grad_norm": 0.2701306939125061,
      "learning_rate": 8e-05,
      "loss": 1.7724,
      "step": 814
    },
    {
      "epoch": 0.17925877048278896,
      "grad_norm": 0.2807336449623108,
      "learning_rate": 8e-05,
      "loss": 1.7832,
      "step": 815
    },
    {
      "epoch": 0.1794787198944243,
      "grad_norm": 0.2847912907600403,
      "learning_rate": 8e-05,
      "loss": 1.7042,
      "step": 816
    },
    {
      "epoch": 0.1796986693060596,
      "grad_norm": 0.2836345434188843,
      "learning_rate": 8e-05,
      "loss": 1.8506,
      "step": 817
    },
    {
      "epoch": 0.17991861871769493,
      "grad_norm": 0.30620551109313965,
      "learning_rate": 8e-05,
      "loss": 1.7695,
      "step": 818
    },
    {
      "epoch": 0.18013856812933027,
      "grad_norm": 0.2698993980884552,
      "learning_rate": 8e-05,
      "loss": 1.6388,
      "step": 819
    },
    {
      "epoch": 0.18035851754096557,
      "grad_norm": 0.2937266528606415,
      "learning_rate": 8e-05,
      "loss": 1.8648,
      "step": 820
    },
    {
      "epoch": 0.1805784669526009,
      "grad_norm": 0.2661988139152527,
      "learning_rate": 8e-05,
      "loss": 1.7563,
      "step": 821
    },
    {
      "epoch": 0.18079841636423621,
      "grad_norm": 0.2944018840789795,
      "learning_rate": 8e-05,
      "loss": 1.882,
      "step": 822
    },
    {
      "epoch": 0.18101836577587155,
      "grad_norm": 0.2774435579776764,
      "learning_rate": 8e-05,
      "loss": 1.8117,
      "step": 823
    },
    {
      "epoch": 0.18123831518750688,
      "grad_norm": 0.27865204215049744,
      "learning_rate": 8e-05,
      "loss": 1.8815,
      "step": 824
    },
    {
      "epoch": 0.1814582645991422,
      "grad_norm": 0.26444011926651,
      "learning_rate": 8e-05,
      "loss": 1.5844,
      "step": 825
    },
    {
      "epoch": 0.18167821401077752,
      "grad_norm": 0.27044716477394104,
      "learning_rate": 8e-05,
      "loss": 1.7403,
      "step": 826
    },
    {
      "epoch": 0.18189816342241286,
      "grad_norm": 0.28727805614471436,
      "learning_rate": 8e-05,
      "loss": 1.8556,
      "step": 827
    },
    {
      "epoch": 0.18211811283404816,
      "grad_norm": 0.26131972670555115,
      "learning_rate": 8e-05,
      "loss": 1.7727,
      "step": 828
    },
    {
      "epoch": 0.1823380622456835,
      "grad_norm": 0.269638329744339,
      "learning_rate": 8e-05,
      "loss": 1.6795,
      "step": 829
    },
    {
      "epoch": 0.18255801165731883,
      "grad_norm": 0.2671653628349304,
      "learning_rate": 8e-05,
      "loss": 1.6811,
      "step": 830
    },
    {
      "epoch": 0.18277796106895414,
      "grad_norm": 0.2659014165401459,
      "learning_rate": 8e-05,
      "loss": 1.7166,
      "step": 831
    },
    {
      "epoch": 0.18299791048058947,
      "grad_norm": 0.2719801962375641,
      "learning_rate": 8e-05,
      "loss": 1.6938,
      "step": 832
    },
    {
      "epoch": 0.18321785989222478,
      "grad_norm": 0.3272366225719452,
      "learning_rate": 8e-05,
      "loss": 1.8213,
      "step": 833
    },
    {
      "epoch": 0.1834378093038601,
      "grad_norm": 0.2635113000869751,
      "learning_rate": 8e-05,
      "loss": 1.6291,
      "step": 834
    },
    {
      "epoch": 0.18365775871549544,
      "grad_norm": 0.29401281476020813,
      "learning_rate": 8e-05,
      "loss": 1.8234,
      "step": 835
    },
    {
      "epoch": 0.18387770812713075,
      "grad_norm": 0.29188451170921326,
      "learning_rate": 8e-05,
      "loss": 1.7359,
      "step": 836
    },
    {
      "epoch": 0.18409765753876609,
      "grad_norm": 0.2688080072402954,
      "learning_rate": 8e-05,
      "loss": 1.7088,
      "step": 837
    },
    {
      "epoch": 0.18431760695040142,
      "grad_norm": 0.27907344698905945,
      "learning_rate": 8e-05,
      "loss": 1.6762,
      "step": 838
    },
    {
      "epoch": 0.18453755636203673,
      "grad_norm": 0.2875908315181732,
      "learning_rate": 8e-05,
      "loss": 1.7612,
      "step": 839
    },
    {
      "epoch": 0.18475750577367206,
      "grad_norm": 0.2683177888393402,
      "learning_rate": 8e-05,
      "loss": 1.5965,
      "step": 840
    },
    {
      "epoch": 0.18497745518530737,
      "grad_norm": 0.29948660731315613,
      "learning_rate": 8e-05,
      "loss": 1.7358,
      "step": 841
    },
    {
      "epoch": 0.1851974045969427,
      "grad_norm": 0.28153204917907715,
      "learning_rate": 8e-05,
      "loss": 1.8089,
      "step": 842
    },
    {
      "epoch": 0.18541735400857803,
      "grad_norm": 0.29185283184051514,
      "learning_rate": 8e-05,
      "loss": 1.8357,
      "step": 843
    },
    {
      "epoch": 0.18563730342021334,
      "grad_norm": 0.27565860748291016,
      "learning_rate": 8e-05,
      "loss": 1.9565,
      "step": 844
    },
    {
      "epoch": 0.18585725283184867,
      "grad_norm": 0.2811479866504669,
      "learning_rate": 8e-05,
      "loss": 1.8493,
      "step": 845
    },
    {
      "epoch": 0.186077202243484,
      "grad_norm": 0.271893173456192,
      "learning_rate": 8e-05,
      "loss": 1.7622,
      "step": 846
    },
    {
      "epoch": 0.18629715165511931,
      "grad_norm": 0.26383113861083984,
      "learning_rate": 8e-05,
      "loss": 1.7392,
      "step": 847
    },
    {
      "epoch": 0.18651710106675465,
      "grad_norm": 0.2863881289958954,
      "learning_rate": 8e-05,
      "loss": 1.7367,
      "step": 848
    },
    {
      "epoch": 0.18673705047838998,
      "grad_norm": 0.28036433458328247,
      "learning_rate": 8e-05,
      "loss": 1.6587,
      "step": 849
    },
    {
      "epoch": 0.1869569998900253,
      "grad_norm": 0.2938581705093384,
      "learning_rate": 8e-05,
      "loss": 1.7411,
      "step": 850
    },
    {
      "epoch": 0.18717694930166062,
      "grad_norm": 0.27487799525260925,
      "learning_rate": 8e-05,
      "loss": 1.8054,
      "step": 851
    },
    {
      "epoch": 0.18739689871329593,
      "grad_norm": 0.2693670690059662,
      "learning_rate": 8e-05,
      "loss": 1.7361,
      "step": 852
    },
    {
      "epoch": 0.18761684812493126,
      "grad_norm": 0.2999705970287323,
      "learning_rate": 8e-05,
      "loss": 1.909,
      "step": 853
    },
    {
      "epoch": 0.1878367975365666,
      "grad_norm": 0.28235265612602234,
      "learning_rate": 8e-05,
      "loss": 1.8611,
      "step": 854
    },
    {
      "epoch": 0.1880567469482019,
      "grad_norm": 0.28417298197746277,
      "learning_rate": 8e-05,
      "loss": 1.683,
      "step": 855
    },
    {
      "epoch": 0.18827669635983724,
      "grad_norm": 0.2697356045246124,
      "learning_rate": 8e-05,
      "loss": 1.7138,
      "step": 856
    },
    {
      "epoch": 0.18849664577147257,
      "grad_norm": 0.26900357007980347,
      "learning_rate": 8e-05,
      "loss": 1.5579,
      "step": 857
    },
    {
      "epoch": 0.18871659518310788,
      "grad_norm": 0.259941041469574,
      "learning_rate": 8e-05,
      "loss": 1.7106,
      "step": 858
    },
    {
      "epoch": 0.1889365445947432,
      "grad_norm": 0.26958781480789185,
      "learning_rate": 8e-05,
      "loss": 1.6454,
      "step": 859
    },
    {
      "epoch": 0.18915649400637854,
      "grad_norm": 0.26425305008888245,
      "learning_rate": 8e-05,
      "loss": 1.6408,
      "step": 860
    },
    {
      "epoch": 0.18937644341801385,
      "grad_norm": 0.26996907591819763,
      "learning_rate": 8e-05,
      "loss": 1.6949,
      "step": 861
    },
    {
      "epoch": 0.18959639282964919,
      "grad_norm": 0.25882837176322937,
      "learning_rate": 8e-05,
      "loss": 1.6142,
      "step": 862
    },
    {
      "epoch": 0.1898163422412845,
      "grad_norm": 0.28000783920288086,
      "learning_rate": 8e-05,
      "loss": 1.8007,
      "step": 863
    },
    {
      "epoch": 0.19003629165291983,
      "grad_norm": 0.2744222581386566,
      "learning_rate": 8e-05,
      "loss": 1.6604,
      "step": 864
    },
    {
      "epoch": 0.19025624106455516,
      "grad_norm": 0.2791576683521271,
      "learning_rate": 8e-05,
      "loss": 1.7061,
      "step": 865
    },
    {
      "epoch": 0.19047619047619047,
      "grad_norm": 0.27878084778785706,
      "learning_rate": 8e-05,
      "loss": 1.8604,
      "step": 866
    },
    {
      "epoch": 0.1906961398878258,
      "grad_norm": 0.3818608820438385,
      "learning_rate": 8e-05,
      "loss": 1.8616,
      "step": 867
    },
    {
      "epoch": 0.19091608929946113,
      "grad_norm": 0.27952665090560913,
      "learning_rate": 8e-05,
      "loss": 1.7616,
      "step": 868
    },
    {
      "epoch": 0.19113603871109644,
      "grad_norm": 0.2711832523345947,
      "learning_rate": 8e-05,
      "loss": 1.7974,
      "step": 869
    },
    {
      "epoch": 0.19135598812273177,
      "grad_norm": 0.2572176456451416,
      "learning_rate": 8e-05,
      "loss": 1.584,
      "step": 870
    },
    {
      "epoch": 0.1915759375343671,
      "grad_norm": 0.2847760319709778,
      "learning_rate": 8e-05,
      "loss": 1.8598,
      "step": 871
    },
    {
      "epoch": 0.19179588694600241,
      "grad_norm": 0.29798731207847595,
      "learning_rate": 8e-05,
      "loss": 1.6689,
      "step": 872
    },
    {
      "epoch": 0.19201583635763775,
      "grad_norm": 0.2674097716808319,
      "learning_rate": 8e-05,
      "loss": 1.6694,
      "step": 873
    },
    {
      "epoch": 0.19223578576927305,
      "grad_norm": 0.27707335352897644,
      "learning_rate": 8e-05,
      "loss": 1.7251,
      "step": 874
    },
    {
      "epoch": 0.1924557351809084,
      "grad_norm": 0.2801666259765625,
      "learning_rate": 8e-05,
      "loss": 1.7251,
      "step": 875
    },
    {
      "epoch": 0.19267568459254372,
      "grad_norm": 0.2656191885471344,
      "learning_rate": 8e-05,
      "loss": 1.6232,
      "step": 876
    },
    {
      "epoch": 0.19289563400417903,
      "grad_norm": 0.2588733732700348,
      "learning_rate": 8e-05,
      "loss": 1.7595,
      "step": 877
    },
    {
      "epoch": 0.19311558341581436,
      "grad_norm": 0.2999958097934723,
      "learning_rate": 8e-05,
      "loss": 1.9095,
      "step": 878
    },
    {
      "epoch": 0.1933355328274497,
      "grad_norm": 0.27143120765686035,
      "learning_rate": 8e-05,
      "loss": 1.6698,
      "step": 879
    },
    {
      "epoch": 0.193555482239085,
      "grad_norm": 0.29155731201171875,
      "learning_rate": 8e-05,
      "loss": 1.6437,
      "step": 880
    },
    {
      "epoch": 0.19377543165072034,
      "grad_norm": 0.26307716965675354,
      "learning_rate": 8e-05,
      "loss": 1.6161,
      "step": 881
    },
    {
      "epoch": 0.19399538106235567,
      "grad_norm": 0.27041196823120117,
      "learning_rate": 8e-05,
      "loss": 1.5374,
      "step": 882
    },
    {
      "epoch": 0.19421533047399098,
      "grad_norm": 0.2752692699432373,
      "learning_rate": 8e-05,
      "loss": 1.6543,
      "step": 883
    },
    {
      "epoch": 0.1944352798856263,
      "grad_norm": 0.2883388102054596,
      "learning_rate": 8e-05,
      "loss": 1.7503,
      "step": 884
    },
    {
      "epoch": 0.19465522929726162,
      "grad_norm": 0.27332282066345215,
      "learning_rate": 8e-05,
      "loss": 1.8456,
      "step": 885
    },
    {
      "epoch": 0.19487517870889695,
      "grad_norm": 0.26226627826690674,
      "learning_rate": 8e-05,
      "loss": 1.2577,
      "step": 886
    },
    {
      "epoch": 0.19509512812053229,
      "grad_norm": 0.2709749639034271,
      "learning_rate": 8e-05,
      "loss": 1.7854,
      "step": 887
    },
    {
      "epoch": 0.1953150775321676,
      "grad_norm": 0.28380879759788513,
      "learning_rate": 8e-05,
      "loss": 1.7151,
      "step": 888
    },
    {
      "epoch": 0.19553502694380293,
      "grad_norm": 0.2702254354953766,
      "learning_rate": 8e-05,
      "loss": 1.6779,
      "step": 889
    },
    {
      "epoch": 0.19575497635543826,
      "grad_norm": 0.2620486617088318,
      "learning_rate": 8e-05,
      "loss": 1.7166,
      "step": 890
    },
    {
      "epoch": 0.19597492576707357,
      "grad_norm": 0.27195873856544495,
      "learning_rate": 8e-05,
      "loss": 1.7263,
      "step": 891
    },
    {
      "epoch": 0.1961948751787089,
      "grad_norm": 0.2719867527484894,
      "learning_rate": 8e-05,
      "loss": 1.6982,
      "step": 892
    },
    {
      "epoch": 0.19641482459034423,
      "grad_norm": 0.27889111638069153,
      "learning_rate": 8e-05,
      "loss": 1.7726,
      "step": 893
    },
    {
      "epoch": 0.19663477400197954,
      "grad_norm": 0.2745397686958313,
      "learning_rate": 8e-05,
      "loss": 1.7737,
      "step": 894
    },
    {
      "epoch": 0.19685472341361487,
      "grad_norm": 0.2698670029640198,
      "learning_rate": 8e-05,
      "loss": 1.7149,
      "step": 895
    },
    {
      "epoch": 0.19707467282525018,
      "grad_norm": 0.27113667130470276,
      "learning_rate": 8e-05,
      "loss": 1.7887,
      "step": 896
    },
    {
      "epoch": 0.19729462223688551,
      "grad_norm": 0.2772979140281677,
      "learning_rate": 8e-05,
      "loss": 1.8163,
      "step": 897
    },
    {
      "epoch": 0.19751457164852085,
      "grad_norm": 0.2757657766342163,
      "learning_rate": 8e-05,
      "loss": 1.636,
      "step": 898
    },
    {
      "epoch": 0.19773452106015615,
      "grad_norm": 0.26945242285728455,
      "learning_rate": 8e-05,
      "loss": 1.7639,
      "step": 899
    },
    {
      "epoch": 0.1979544704717915,
      "grad_norm": 0.27328991889953613,
      "learning_rate": 8e-05,
      "loss": 1.8421,
      "step": 900
    },
    {
      "epoch": 0.19817441988342682,
      "grad_norm": 0.2721468210220337,
      "learning_rate": 8e-05,
      "loss": 1.6926,
      "step": 901
    },
    {
      "epoch": 0.19839436929506213,
      "grad_norm": 0.2633766233921051,
      "learning_rate": 8e-05,
      "loss": 1.7629,
      "step": 902
    },
    {
      "epoch": 0.19861431870669746,
      "grad_norm": 0.26183879375457764,
      "learning_rate": 8e-05,
      "loss": 1.7149,
      "step": 903
    },
    {
      "epoch": 0.1988342681183328,
      "grad_norm": 0.2837960422039032,
      "learning_rate": 8e-05,
      "loss": 1.7923,
      "step": 904
    },
    {
      "epoch": 0.1990542175299681,
      "grad_norm": 0.30745571851730347,
      "learning_rate": 8e-05,
      "loss": 1.8315,
      "step": 905
    },
    {
      "epoch": 0.19927416694160344,
      "grad_norm": 0.2734341323375702,
      "learning_rate": 8e-05,
      "loss": 1.7424,
      "step": 906
    },
    {
      "epoch": 0.19949411635323874,
      "grad_norm": 0.2613460123538971,
      "learning_rate": 8e-05,
      "loss": 1.6445,
      "step": 907
    },
    {
      "epoch": 0.19971406576487408,
      "grad_norm": 0.27867522835731506,
      "learning_rate": 8e-05,
      "loss": 1.7075,
      "step": 908
    },
    {
      "epoch": 0.1999340151765094,
      "grad_norm": 0.269789457321167,
      "learning_rate": 8e-05,
      "loss": 1.6801,
      "step": 909
    },
    {
      "epoch": 0.20015396458814472,
      "grad_norm": 0.2684427797794342,
      "learning_rate": 8e-05,
      "loss": 1.6862,
      "step": 910
    },
    {
      "epoch": 0.20037391399978005,
      "grad_norm": 0.2929883897304535,
      "learning_rate": 8e-05,
      "loss": 1.7972,
      "step": 911
    },
    {
      "epoch": 0.20059386341141539,
      "grad_norm": 0.2757764756679535,
      "learning_rate": 8e-05,
      "loss": 1.6198,
      "step": 912
    },
    {
      "epoch": 0.2008138128230507,
      "grad_norm": 0.28071129322052,
      "learning_rate": 8e-05,
      "loss": 1.8268,
      "step": 913
    },
    {
      "epoch": 0.20103376223468603,
      "grad_norm": 0.2964448928833008,
      "learning_rate": 8e-05,
      "loss": 1.7806,
      "step": 914
    },
    {
      "epoch": 0.20125371164632136,
      "grad_norm": 0.2682490050792694,
      "learning_rate": 8e-05,
      "loss": 1.7016,
      "step": 915
    },
    {
      "epoch": 0.20147366105795667,
      "grad_norm": 0.2838338613510132,
      "learning_rate": 8e-05,
      "loss": 1.7402,
      "step": 916
    },
    {
      "epoch": 0.201693610469592,
      "grad_norm": 0.27621790766716003,
      "learning_rate": 8e-05,
      "loss": 1.7442,
      "step": 917
    },
    {
      "epoch": 0.2019135598812273,
      "grad_norm": 0.29265734553337097,
      "learning_rate": 8e-05,
      "loss": 1.6924,
      "step": 918
    },
    {
      "epoch": 0.20213350929286264,
      "grad_norm": 0.27404630184173584,
      "learning_rate": 8e-05,
      "loss": 1.7312,
      "step": 919
    },
    {
      "epoch": 0.20235345870449797,
      "grad_norm": 0.2742730975151062,
      "learning_rate": 8e-05,
      "loss": 1.8645,
      "step": 920
    },
    {
      "epoch": 0.20257340811613328,
      "grad_norm": 0.28536343574523926,
      "learning_rate": 8e-05,
      "loss": 1.717,
      "step": 921
    },
    {
      "epoch": 0.20279335752776861,
      "grad_norm": 0.28739288449287415,
      "learning_rate": 8e-05,
      "loss": 1.8356,
      "step": 922
    },
    {
      "epoch": 0.20301330693940395,
      "grad_norm": 0.2650564908981323,
      "learning_rate": 8e-05,
      "loss": 1.717,
      "step": 923
    },
    {
      "epoch": 0.20323325635103925,
      "grad_norm": 0.28638410568237305,
      "learning_rate": 8e-05,
      "loss": 1.8822,
      "step": 924
    },
    {
      "epoch": 0.2034532057626746,
      "grad_norm": 0.25474488735198975,
      "learning_rate": 8e-05,
      "loss": 1.5533,
      "step": 925
    },
    {
      "epoch": 0.20367315517430992,
      "grad_norm": 0.2719588279724121,
      "learning_rate": 8e-05,
      "loss": 1.7887,
      "step": 926
    },
    {
      "epoch": 0.20389310458594523,
      "grad_norm": 0.2572193741798401,
      "learning_rate": 8e-05,
      "loss": 1.5735,
      "step": 927
    },
    {
      "epoch": 0.20411305399758056,
      "grad_norm": 0.2975933253765106,
      "learning_rate": 8e-05,
      "loss": 1.7875,
      "step": 928
    },
    {
      "epoch": 0.20433300340921587,
      "grad_norm": 0.2562117874622345,
      "learning_rate": 8e-05,
      "loss": 1.5977,
      "step": 929
    },
    {
      "epoch": 0.2045529528208512,
      "grad_norm": 0.2524821162223816,
      "learning_rate": 8e-05,
      "loss": 1.6356,
      "step": 930
    },
    {
      "epoch": 0.20477290223248654,
      "grad_norm": 0.2621130347251892,
      "learning_rate": 8e-05,
      "loss": 1.6082,
      "step": 931
    },
    {
      "epoch": 0.20499285164412184,
      "grad_norm": 0.27930355072021484,
      "learning_rate": 8e-05,
      "loss": 1.7618,
      "step": 932
    },
    {
      "epoch": 0.20521280105575718,
      "grad_norm": 0.29147934913635254,
      "learning_rate": 8e-05,
      "loss": 1.8223,
      "step": 933
    },
    {
      "epoch": 0.2054327504673925,
      "grad_norm": 0.2584928870201111,
      "learning_rate": 8e-05,
      "loss": 1.6916,
      "step": 934
    },
    {
      "epoch": 0.20565269987902782,
      "grad_norm": 0.27299705147743225,
      "learning_rate": 8e-05,
      "loss": 1.5535,
      "step": 935
    },
    {
      "epoch": 0.20587264929066315,
      "grad_norm": 0.2682443857192993,
      "learning_rate": 8e-05,
      "loss": 1.7119,
      "step": 936
    },
    {
      "epoch": 0.20609259870229849,
      "grad_norm": 0.29716598987579346,
      "learning_rate": 8e-05,
      "loss": 2.0561,
      "step": 937
    },
    {
      "epoch": 0.2063125481139338,
      "grad_norm": 0.27801281213760376,
      "learning_rate": 8e-05,
      "loss": 1.7605,
      "step": 938
    },
    {
      "epoch": 0.20653249752556913,
      "grad_norm": 0.26767662167549133,
      "learning_rate": 8e-05,
      "loss": 1.6462,
      "step": 939
    },
    {
      "epoch": 0.20675244693720443,
      "grad_norm": 0.27354639768600464,
      "learning_rate": 8e-05,
      "loss": 1.7241,
      "step": 940
    },
    {
      "epoch": 0.20697239634883977,
      "grad_norm": 0.2684631049633026,
      "learning_rate": 8e-05,
      "loss": 1.8066,
      "step": 941
    },
    {
      "epoch": 0.2071923457604751,
      "grad_norm": 0.27846816182136536,
      "learning_rate": 8e-05,
      "loss": 1.7553,
      "step": 942
    },
    {
      "epoch": 0.2074122951721104,
      "grad_norm": 0.2820284366607666,
      "learning_rate": 8e-05,
      "loss": 1.8302,
      "step": 943
    },
    {
      "epoch": 0.20763224458374574,
      "grad_norm": 0.28080835938453674,
      "learning_rate": 8e-05,
      "loss": 1.7544,
      "step": 944
    },
    {
      "epoch": 0.20785219399538107,
      "grad_norm": 0.28095102310180664,
      "learning_rate": 8e-05,
      "loss": 1.7271,
      "step": 945
    },
    {
      "epoch": 0.20807214340701638,
      "grad_norm": 0.27856144309043884,
      "learning_rate": 8e-05,
      "loss": 1.8441,
      "step": 946
    },
    {
      "epoch": 0.20829209281865171,
      "grad_norm": 0.27816230058670044,
      "learning_rate": 8e-05,
      "loss": 1.981,
      "step": 947
    },
    {
      "epoch": 0.20851204223028702,
      "grad_norm": 0.2954215705394745,
      "learning_rate": 8e-05,
      "loss": 1.8001,
      "step": 948
    },
    {
      "epoch": 0.20873199164192235,
      "grad_norm": 0.24413350224494934,
      "learning_rate": 8e-05,
      "loss": 1.5436,
      "step": 949
    },
    {
      "epoch": 0.2089519410535577,
      "grad_norm": 0.2849874198436737,
      "learning_rate": 8e-05,
      "loss": 1.7027,
      "step": 950
    },
    {
      "epoch": 0.209171890465193,
      "grad_norm": 0.2710252106189728,
      "learning_rate": 8e-05,
      "loss": 1.7222,
      "step": 951
    },
    {
      "epoch": 0.20939183987682833,
      "grad_norm": 0.2557348608970642,
      "learning_rate": 8e-05,
      "loss": 1.6469,
      "step": 952
    },
    {
      "epoch": 0.20961178928846366,
      "grad_norm": 0.2688618004322052,
      "learning_rate": 8e-05,
      "loss": 1.6471,
      "step": 953
    },
    {
      "epoch": 0.20983173870009897,
      "grad_norm": 0.28641626238822937,
      "learning_rate": 8e-05,
      "loss": 1.8796,
      "step": 954
    },
    {
      "epoch": 0.2100516881117343,
      "grad_norm": 0.2582222521305084,
      "learning_rate": 8e-05,
      "loss": 1.5961,
      "step": 955
    },
    {
      "epoch": 0.21027163752336964,
      "grad_norm": 0.2615504562854767,
      "learning_rate": 8e-05,
      "loss": 1.6668,
      "step": 956
    },
    {
      "epoch": 0.21049158693500494,
      "grad_norm": 0.2669670879840851,
      "learning_rate": 8e-05,
      "loss": 1.663,
      "step": 957
    },
    {
      "epoch": 0.21071153634664028,
      "grad_norm": 0.2649092972278595,
      "learning_rate": 8e-05,
      "loss": 1.5377,
      "step": 958
    },
    {
      "epoch": 0.21093148575827558,
      "grad_norm": 0.2936461865901947,
      "learning_rate": 8e-05,
      "loss": 1.5659,
      "step": 959
    },
    {
      "epoch": 0.21115143516991092,
      "grad_norm": 0.2878846824169159,
      "learning_rate": 8e-05,
      "loss": 1.8567,
      "step": 960
    },
    {
      "epoch": 0.21137138458154625,
      "grad_norm": 0.2928799092769623,
      "learning_rate": 8e-05,
      "loss": 1.8423,
      "step": 961
    },
    {
      "epoch": 0.21159133399318156,
      "grad_norm": 0.2641200125217438,
      "learning_rate": 8e-05,
      "loss": 1.6403,
      "step": 962
    },
    {
      "epoch": 0.2118112834048169,
      "grad_norm": 0.26553985476493835,
      "learning_rate": 8e-05,
      "loss": 1.7436,
      "step": 963
    },
    {
      "epoch": 0.21203123281645223,
      "grad_norm": 0.25616276264190674,
      "learning_rate": 8e-05,
      "loss": 1.5959,
      "step": 964
    },
    {
      "epoch": 0.21225118222808753,
      "grad_norm": 0.29729175567626953,
      "learning_rate": 8e-05,
      "loss": 1.7164,
      "step": 965
    },
    {
      "epoch": 0.21247113163972287,
      "grad_norm": 0.2739759683609009,
      "learning_rate": 8e-05,
      "loss": 1.7797,
      "step": 966
    },
    {
      "epoch": 0.2126910810513582,
      "grad_norm": 0.2686353921890259,
      "learning_rate": 8e-05,
      "loss": 1.6974,
      "step": 967
    },
    {
      "epoch": 0.2129110304629935,
      "grad_norm": 0.261820912361145,
      "learning_rate": 8e-05,
      "loss": 1.6864,
      "step": 968
    },
    {
      "epoch": 0.21313097987462884,
      "grad_norm": 0.26877105236053467,
      "learning_rate": 8e-05,
      "loss": 1.6164,
      "step": 969
    },
    {
      "epoch": 0.21335092928626415,
      "grad_norm": 0.2555043399333954,
      "learning_rate": 8e-05,
      "loss": 1.6898,
      "step": 970
    },
    {
      "epoch": 0.21357087869789948,
      "grad_norm": 0.28584909439086914,
      "learning_rate": 8e-05,
      "loss": 1.9125,
      "step": 971
    },
    {
      "epoch": 0.21379082810953481,
      "grad_norm": 0.2830945551395416,
      "learning_rate": 8e-05,
      "loss": 1.6416,
      "step": 972
    },
    {
      "epoch": 0.21401077752117012,
      "grad_norm": 0.27979904413223267,
      "learning_rate": 8e-05,
      "loss": 1.8355,
      "step": 973
    },
    {
      "epoch": 0.21423072693280545,
      "grad_norm": 0.2672286033630371,
      "learning_rate": 8e-05,
      "loss": 1.685,
      "step": 974
    },
    {
      "epoch": 0.2144506763444408,
      "grad_norm": 0.26699069142341614,
      "learning_rate": 8e-05,
      "loss": 1.5951,
      "step": 975
    },
    {
      "epoch": 0.2146706257560761,
      "grad_norm": 0.2720418870449066,
      "learning_rate": 8e-05,
      "loss": 1.7558,
      "step": 976
    },
    {
      "epoch": 0.21489057516771143,
      "grad_norm": 0.26792463660240173,
      "learning_rate": 8e-05,
      "loss": 1.7407,
      "step": 977
    },
    {
      "epoch": 0.21511052457934676,
      "grad_norm": 0.2763652503490448,
      "learning_rate": 8e-05,
      "loss": 1.7525,
      "step": 978
    },
    {
      "epoch": 0.21533047399098207,
      "grad_norm": 0.2952554225921631,
      "learning_rate": 8e-05,
      "loss": 1.6535,
      "step": 979
    },
    {
      "epoch": 0.2155504234026174,
      "grad_norm": 0.24981874227523804,
      "learning_rate": 8e-05,
      "loss": 1.6055,
      "step": 980
    },
    {
      "epoch": 0.2157703728142527,
      "grad_norm": 0.29071807861328125,
      "learning_rate": 8e-05,
      "loss": 1.7461,
      "step": 981
    },
    {
      "epoch": 0.21599032222588804,
      "grad_norm": 0.26875782012939453,
      "learning_rate": 8e-05,
      "loss": 1.5809,
      "step": 982
    },
    {
      "epoch": 0.21621027163752338,
      "grad_norm": 0.2519072890281677,
      "learning_rate": 8e-05,
      "loss": 1.7001,
      "step": 983
    },
    {
      "epoch": 0.21643022104915868,
      "grad_norm": 0.2748781144618988,
      "learning_rate": 8e-05,
      "loss": 1.8367,
      "step": 984
    },
    {
      "epoch": 0.21665017046079402,
      "grad_norm": 0.274047315120697,
      "learning_rate": 8e-05,
      "loss": 1.7698,
      "step": 985
    },
    {
      "epoch": 0.21687011987242935,
      "grad_norm": 0.2614712119102478,
      "learning_rate": 8e-05,
      "loss": 1.5411,
      "step": 986
    },
    {
      "epoch": 0.21709006928406466,
      "grad_norm": 0.2714536190032959,
      "learning_rate": 8e-05,
      "loss": 1.6058,
      "step": 987
    },
    {
      "epoch": 0.2173100186957,
      "grad_norm": 0.28763729333877563,
      "learning_rate": 8e-05,
      "loss": 1.6711,
      "step": 988
    },
    {
      "epoch": 0.21752996810733533,
      "grad_norm": 0.26780402660369873,
      "learning_rate": 8e-05,
      "loss": 1.549,
      "step": 989
    },
    {
      "epoch": 0.21774991751897063,
      "grad_norm": 0.28782159090042114,
      "learning_rate": 8e-05,
      "loss": 1.8305,
      "step": 990
    },
    {
      "epoch": 0.21796986693060597,
      "grad_norm": 0.2859013080596924,
      "learning_rate": 8e-05,
      "loss": 1.7794,
      "step": 991
    },
    {
      "epoch": 0.21818981634224127,
      "grad_norm": 0.2893369197845459,
      "learning_rate": 8e-05,
      "loss": 1.6284,
      "step": 992
    },
    {
      "epoch": 0.2184097657538766,
      "grad_norm": 0.2809627652168274,
      "learning_rate": 8e-05,
      "loss": 1.6401,
      "step": 993
    },
    {
      "epoch": 0.21862971516551194,
      "grad_norm": 0.2700895667076111,
      "learning_rate": 8e-05,
      "loss": 1.7153,
      "step": 994
    },
    {
      "epoch": 0.21884966457714725,
      "grad_norm": 0.26506903767585754,
      "learning_rate": 8e-05,
      "loss": 1.4688,
      "step": 995
    },
    {
      "epoch": 0.21906961398878258,
      "grad_norm": 0.28202024102211,
      "learning_rate": 8e-05,
      "loss": 1.7009,
      "step": 996
    },
    {
      "epoch": 0.21928956340041791,
      "grad_norm": 0.2625409960746765,
      "learning_rate": 8e-05,
      "loss": 1.6491,
      "step": 997
    },
    {
      "epoch": 0.21950951281205322,
      "grad_norm": 0.29967787861824036,
      "learning_rate": 8e-05,
      "loss": 1.7231,
      "step": 998
    },
    {
      "epoch": 0.21972946222368855,
      "grad_norm": 0.2992357909679413,
      "learning_rate": 8e-05,
      "loss": 1.7028,
      "step": 999
    },
    {
      "epoch": 0.2199494116353239,
      "grad_norm": 0.28712475299835205,
      "learning_rate": 8e-05,
      "loss": 1.763,
      "step": 1000
    },
    {
      "epoch": 0.2201693610469592,
      "grad_norm": 0.26186901330947876,
      "learning_rate": 8e-05,
      "loss": 1.5695,
      "step": 1001
    },
    {
      "epoch": 0.22038931045859453,
      "grad_norm": 0.2897952198982239,
      "learning_rate": 8e-05,
      "loss": 1.6303,
      "step": 1002
    },
    {
      "epoch": 0.22060925987022983,
      "grad_norm": 0.2761494815349579,
      "learning_rate": 8e-05,
      "loss": 1.7448,
      "step": 1003
    },
    {
      "epoch": 0.22082920928186517,
      "grad_norm": 0.2604154944419861,
      "learning_rate": 8e-05,
      "loss": 1.53,
      "step": 1004
    },
    {
      "epoch": 0.2210491586935005,
      "grad_norm": 0.2897418737411499,
      "learning_rate": 8e-05,
      "loss": 1.7639,
      "step": 1005
    },
    {
      "epoch": 0.2212691081051358,
      "grad_norm": 0.28289687633514404,
      "learning_rate": 8e-05,
      "loss": 1.7202,
      "step": 1006
    },
    {
      "epoch": 0.22148905751677114,
      "grad_norm": 0.26917099952697754,
      "learning_rate": 8e-05,
      "loss": 1.7183,
      "step": 1007
    },
    {
      "epoch": 0.22170900692840648,
      "grad_norm": 0.26708024740219116,
      "learning_rate": 8e-05,
      "loss": 1.636,
      "step": 1008
    },
    {
      "epoch": 0.22192895634004178,
      "grad_norm": 0.2759459316730499,
      "learning_rate": 8e-05,
      "loss": 1.6537,
      "step": 1009
    },
    {
      "epoch": 0.22214890575167712,
      "grad_norm": 0.3040393590927124,
      "learning_rate": 8e-05,
      "loss": 1.7849,
      "step": 1010
    },
    {
      "epoch": 0.22236885516331245,
      "grad_norm": 0.2729750871658325,
      "learning_rate": 8e-05,
      "loss": 1.8199,
      "step": 1011
    },
    {
      "epoch": 0.22258880457494776,
      "grad_norm": 0.28002965450286865,
      "learning_rate": 8e-05,
      "loss": 1.8286,
      "step": 1012
    },
    {
      "epoch": 0.2228087539865831,
      "grad_norm": 0.27389100193977356,
      "learning_rate": 8e-05,
      "loss": 1.6472,
      "step": 1013
    },
    {
      "epoch": 0.2230287033982184,
      "grad_norm": 0.2610195279121399,
      "learning_rate": 8e-05,
      "loss": 1.6096,
      "step": 1014
    },
    {
      "epoch": 0.22324865280985373,
      "grad_norm": 0.2683162987232208,
      "learning_rate": 8e-05,
      "loss": 1.6477,
      "step": 1015
    },
    {
      "epoch": 0.22346860222148907,
      "grad_norm": 0.26524773240089417,
      "learning_rate": 8e-05,
      "loss": 1.6224,
      "step": 1016
    },
    {
      "epoch": 0.22368855163312437,
      "grad_norm": 0.26295366883277893,
      "learning_rate": 8e-05,
      "loss": 1.7316,
      "step": 1017
    },
    {
      "epoch": 0.2239085010447597,
      "grad_norm": 0.2837565243244171,
      "learning_rate": 8e-05,
      "loss": 1.9452,
      "step": 1018
    },
    {
      "epoch": 0.22412845045639504,
      "grad_norm": 0.28365132212638855,
      "learning_rate": 8e-05,
      "loss": 1.6577,
      "step": 1019
    },
    {
      "epoch": 0.22434839986803035,
      "grad_norm": 0.2736522853374481,
      "learning_rate": 8e-05,
      "loss": 1.6644,
      "step": 1020
    },
    {
      "epoch": 0.22456834927966568,
      "grad_norm": 0.2878374755382538,
      "learning_rate": 8e-05,
      "loss": 1.5844,
      "step": 1021
    },
    {
      "epoch": 0.22478829869130101,
      "grad_norm": 0.28223422169685364,
      "learning_rate": 8e-05,
      "loss": 1.881,
      "step": 1022
    },
    {
      "epoch": 0.22500824810293632,
      "grad_norm": 0.26408734917640686,
      "learning_rate": 8e-05,
      "loss": 1.6201,
      "step": 1023
    },
    {
      "epoch": 0.22522819751457165,
      "grad_norm": 0.28506824374198914,
      "learning_rate": 8e-05,
      "loss": 1.8146,
      "step": 1024
    },
    {
      "epoch": 0.22544814692620696,
      "grad_norm": 0.2808188796043396,
      "learning_rate": 8e-05,
      "loss": 1.8394,
      "step": 1025
    },
    {
      "epoch": 0.2256680963378423,
      "grad_norm": 0.2950645387172699,
      "learning_rate": 8e-05,
      "loss": 1.7993,
      "step": 1026
    },
    {
      "epoch": 0.22588804574947763,
      "grad_norm": 0.27935850620269775,
      "learning_rate": 8e-05,
      "loss": 1.6506,
      "step": 1027
    },
    {
      "epoch": 0.22610799516111293,
      "grad_norm": 0.2576957643032074,
      "learning_rate": 8e-05,
      "loss": 1.6987,
      "step": 1028
    },
    {
      "epoch": 0.22632794457274827,
      "grad_norm": 0.2719384729862213,
      "learning_rate": 8e-05,
      "loss": 1.6407,
      "step": 1029
    },
    {
      "epoch": 0.2265478939843836,
      "grad_norm": 0.25457167625427246,
      "learning_rate": 8e-05,
      "loss": 1.6877,
      "step": 1030
    },
    {
      "epoch": 0.2267678433960189,
      "grad_norm": 0.2758035659790039,
      "learning_rate": 8e-05,
      "loss": 1.6739,
      "step": 1031
    },
    {
      "epoch": 0.22698779280765424,
      "grad_norm": 0.27135321497917175,
      "learning_rate": 8e-05,
      "loss": 1.7124,
      "step": 1032
    },
    {
      "epoch": 0.22720774221928958,
      "grad_norm": 0.2675740420818329,
      "learning_rate": 8e-05,
      "loss": 1.7857,
      "step": 1033
    },
    {
      "epoch": 0.22742769163092488,
      "grad_norm": 0.28627943992614746,
      "learning_rate": 8e-05,
      "loss": 1.7012,
      "step": 1034
    },
    {
      "epoch": 0.22764764104256022,
      "grad_norm": 0.2710109353065491,
      "learning_rate": 8e-05,
      "loss": 1.6463,
      "step": 1035
    },
    {
      "epoch": 0.22786759045419552,
      "grad_norm": 0.27190473675727844,
      "learning_rate": 8e-05,
      "loss": 1.7288,
      "step": 1036
    },
    {
      "epoch": 0.22808753986583086,
      "grad_norm": 0.2503564953804016,
      "learning_rate": 8e-05,
      "loss": 1.566,
      "step": 1037
    },
    {
      "epoch": 0.2283074892774662,
      "grad_norm": 0.26503992080688477,
      "learning_rate": 8e-05,
      "loss": 1.7034,
      "step": 1038
    },
    {
      "epoch": 0.2285274386891015,
      "grad_norm": 0.29445260763168335,
      "learning_rate": 8e-05,
      "loss": 1.6739,
      "step": 1039
    },
    {
      "epoch": 0.22874738810073683,
      "grad_norm": 0.25705471634864807,
      "learning_rate": 8e-05,
      "loss": 1.6503,
      "step": 1040
    },
    {
      "epoch": 0.22896733751237217,
      "grad_norm": 0.27109014987945557,
      "learning_rate": 8e-05,
      "loss": 1.8045,
      "step": 1041
    },
    {
      "epoch": 0.22918728692400747,
      "grad_norm": 0.2972055673599243,
      "learning_rate": 8e-05,
      "loss": 1.6439,
      "step": 1042
    },
    {
      "epoch": 0.2294072363356428,
      "grad_norm": 0.27126485109329224,
      "learning_rate": 8e-05,
      "loss": 1.672,
      "step": 1043
    },
    {
      "epoch": 0.22962718574727814,
      "grad_norm": 0.2731145918369293,
      "learning_rate": 8e-05,
      "loss": 1.7795,
      "step": 1044
    },
    {
      "epoch": 0.22984713515891345,
      "grad_norm": 0.2768365442752838,
      "learning_rate": 8e-05,
      "loss": 1.7145,
      "step": 1045
    },
    {
      "epoch": 0.23006708457054878,
      "grad_norm": 0.2606940269470215,
      "learning_rate": 8e-05,
      "loss": 1.6169,
      "step": 1046
    },
    {
      "epoch": 0.2302870339821841,
      "grad_norm": 0.2898729741573334,
      "learning_rate": 8e-05,
      "loss": 1.7315,
      "step": 1047
    },
    {
      "epoch": 0.23050698339381942,
      "grad_norm": 0.2772413194179535,
      "learning_rate": 8e-05,
      "loss": 1.8632,
      "step": 1048
    },
    {
      "epoch": 0.23072693280545475,
      "grad_norm": 0.25808605551719666,
      "learning_rate": 8e-05,
      "loss": 1.6626,
      "step": 1049
    },
    {
      "epoch": 0.23094688221709006,
      "grad_norm": 0.2727161645889282,
      "learning_rate": 8e-05,
      "loss": 1.7848,
      "step": 1050
    },
    {
      "epoch": 0.2311668316287254,
      "grad_norm": 0.25677087903022766,
      "learning_rate": 8e-05,
      "loss": 1.6168,
      "step": 1051
    },
    {
      "epoch": 0.23138678104036073,
      "grad_norm": 0.2761050760746002,
      "learning_rate": 8e-05,
      "loss": 1.8615,
      "step": 1052
    },
    {
      "epoch": 0.23160673045199603,
      "grad_norm": 0.2862778604030609,
      "learning_rate": 8e-05,
      "loss": 1.8728,
      "step": 1053
    },
    {
      "epoch": 0.23182667986363137,
      "grad_norm": 0.27526941895484924,
      "learning_rate": 8e-05,
      "loss": 1.7627,
      "step": 1054
    },
    {
      "epoch": 0.23204662927526667,
      "grad_norm": 0.2932235896587372,
      "learning_rate": 8e-05,
      "loss": 1.8539,
      "step": 1055
    },
    {
      "epoch": 0.232266578686902,
      "grad_norm": 0.2770839035511017,
      "learning_rate": 8e-05,
      "loss": 1.7393,
      "step": 1056
    },
    {
      "epoch": 0.23248652809853734,
      "grad_norm": 0.2741580307483673,
      "learning_rate": 8e-05,
      "loss": 1.6076,
      "step": 1057
    },
    {
      "epoch": 0.23270647751017265,
      "grad_norm": 0.2788783311843872,
      "learning_rate": 8e-05,
      "loss": 1.7615,
      "step": 1058
    },
    {
      "epoch": 0.23292642692180798,
      "grad_norm": 0.28565406799316406,
      "learning_rate": 8e-05,
      "loss": 1.6266,
      "step": 1059
    },
    {
      "epoch": 0.23314637633344332,
      "grad_norm": 0.26543545722961426,
      "learning_rate": 8e-05,
      "loss": 1.7192,
      "step": 1060
    },
    {
      "epoch": 0.23336632574507862,
      "grad_norm": 0.2770478129386902,
      "learning_rate": 8e-05,
      "loss": 1.8056,
      "step": 1061
    },
    {
      "epoch": 0.23358627515671396,
      "grad_norm": 0.27805015444755554,
      "learning_rate": 8e-05,
      "loss": 1.6735,
      "step": 1062
    },
    {
      "epoch": 0.2338062245683493,
      "grad_norm": 0.309862345457077,
      "learning_rate": 8e-05,
      "loss": 1.7235,
      "step": 1063
    },
    {
      "epoch": 0.2340261739799846,
      "grad_norm": 0.27140697836875916,
      "learning_rate": 8e-05,
      "loss": 1.6883,
      "step": 1064
    },
    {
      "epoch": 0.23424612339161993,
      "grad_norm": 0.3052090108394623,
      "learning_rate": 8e-05,
      "loss": 1.8792,
      "step": 1065
    },
    {
      "epoch": 0.23446607280325524,
      "grad_norm": 0.2995065450668335,
      "learning_rate": 8e-05,
      "loss": 1.6632,
      "step": 1066
    },
    {
      "epoch": 0.23468602221489057,
      "grad_norm": 0.2782532870769501,
      "learning_rate": 8e-05,
      "loss": 1.7395,
      "step": 1067
    },
    {
      "epoch": 0.2349059716265259,
      "grad_norm": 0.28436902165412903,
      "learning_rate": 8e-05,
      "loss": 1.8416,
      "step": 1068
    },
    {
      "epoch": 0.2351259210381612,
      "grad_norm": 0.2740377187728882,
      "learning_rate": 8e-05,
      "loss": 1.9026,
      "step": 1069
    },
    {
      "epoch": 0.23534587044979655,
      "grad_norm": 0.2978285849094391,
      "learning_rate": 8e-05,
      "loss": 1.7277,
      "step": 1070
    },
    {
      "epoch": 0.23556581986143188,
      "grad_norm": 0.27265986800193787,
      "learning_rate": 8e-05,
      "loss": 1.7376,
      "step": 1071
    },
    {
      "epoch": 0.2357857692730672,
      "grad_norm": 0.24915599822998047,
      "learning_rate": 8e-05,
      "loss": 1.6151,
      "step": 1072
    },
    {
      "epoch": 0.23600571868470252,
      "grad_norm": 0.28203171491622925,
      "learning_rate": 8e-05,
      "loss": 1.7713,
      "step": 1073
    },
    {
      "epoch": 0.23622566809633785,
      "grad_norm": 0.278793066740036,
      "learning_rate": 8e-05,
      "loss": 1.6717,
      "step": 1074
    },
    {
      "epoch": 0.23644561750797316,
      "grad_norm": 0.2760609984397888,
      "learning_rate": 8e-05,
      "loss": 1.5866,
      "step": 1075
    },
    {
      "epoch": 0.2366655669196085,
      "grad_norm": 0.2726036012172699,
      "learning_rate": 8e-05,
      "loss": 1.6774,
      "step": 1076
    },
    {
      "epoch": 0.2368855163312438,
      "grad_norm": 0.27443891763687134,
      "learning_rate": 8e-05,
      "loss": 1.7615,
      "step": 1077
    },
    {
      "epoch": 0.23710546574287913,
      "grad_norm": 0.2818880081176758,
      "learning_rate": 8e-05,
      "loss": 1.7433,
      "step": 1078
    },
    {
      "epoch": 0.23732541515451447,
      "grad_norm": 0.2646252512931824,
      "learning_rate": 8e-05,
      "loss": 1.5498,
      "step": 1079
    },
    {
      "epoch": 0.23754536456614977,
      "grad_norm": 0.2964784502983093,
      "learning_rate": 8e-05,
      "loss": 1.6162,
      "step": 1080
    },
    {
      "epoch": 0.2377653139777851,
      "grad_norm": 0.3044411242008209,
      "learning_rate": 8e-05,
      "loss": 1.7395,
      "step": 1081
    },
    {
      "epoch": 0.23798526338942044,
      "grad_norm": 0.28679221868515015,
      "learning_rate": 8e-05,
      "loss": 1.8126,
      "step": 1082
    },
    {
      "epoch": 0.23820521280105575,
      "grad_norm": 0.26326417922973633,
      "learning_rate": 8e-05,
      "loss": 1.6451,
      "step": 1083
    },
    {
      "epoch": 0.23842516221269108,
      "grad_norm": 0.28527480363845825,
      "learning_rate": 8e-05,
      "loss": 1.8442,
      "step": 1084
    },
    {
      "epoch": 0.23864511162432642,
      "grad_norm": 0.28897759318351746,
      "learning_rate": 8e-05,
      "loss": 1.8224,
      "step": 1085
    },
    {
      "epoch": 0.23886506103596172,
      "grad_norm": 0.2955721616744995,
      "learning_rate": 8e-05,
      "loss": 1.7304,
      "step": 1086
    },
    {
      "epoch": 0.23908501044759706,
      "grad_norm": 0.26267075538635254,
      "learning_rate": 8e-05,
      "loss": 1.67,
      "step": 1087
    },
    {
      "epoch": 0.23930495985923236,
      "grad_norm": 0.27105912566185,
      "learning_rate": 8e-05,
      "loss": 1.7461,
      "step": 1088
    },
    {
      "epoch": 0.2395249092708677,
      "grad_norm": 0.26483941078186035,
      "learning_rate": 8e-05,
      "loss": 1.6215,
      "step": 1089
    },
    {
      "epoch": 0.23974485868250303,
      "grad_norm": 0.2804373800754547,
      "learning_rate": 8e-05,
      "loss": 1.6618,
      "step": 1090
    },
    {
      "epoch": 0.23996480809413834,
      "grad_norm": 0.26146185398101807,
      "learning_rate": 8e-05,
      "loss": 1.6641,
      "step": 1091
    },
    {
      "epoch": 0.24018475750577367,
      "grad_norm": 0.2839837372303009,
      "learning_rate": 8e-05,
      "loss": 1.5898,
      "step": 1092
    },
    {
      "epoch": 0.240404706917409,
      "grad_norm": 0.26833322644233704,
      "learning_rate": 8e-05,
      "loss": 1.8341,
      "step": 1093
    },
    {
      "epoch": 0.2406246563290443,
      "grad_norm": 0.2779574394226074,
      "learning_rate": 8e-05,
      "loss": 1.7142,
      "step": 1094
    },
    {
      "epoch": 0.24084460574067965,
      "grad_norm": 0.2821759879589081,
      "learning_rate": 8e-05,
      "loss": 1.7261,
      "step": 1095
    },
    {
      "epoch": 0.24106455515231498,
      "grad_norm": 0.2849150002002716,
      "learning_rate": 8e-05,
      "loss": 1.6834,
      "step": 1096
    },
    {
      "epoch": 0.24128450456395029,
      "grad_norm": 0.277148574590683,
      "learning_rate": 8e-05,
      "loss": 1.5617,
      "step": 1097
    },
    {
      "epoch": 0.24150445397558562,
      "grad_norm": 0.28307756781578064,
      "learning_rate": 8e-05,
      "loss": 1.8104,
      "step": 1098
    },
    {
      "epoch": 0.24172440338722093,
      "grad_norm": 0.28540289402008057,
      "learning_rate": 8e-05,
      "loss": 1.7331,
      "step": 1099
    },
    {
      "epoch": 0.24194435279885626,
      "grad_norm": 0.277544766664505,
      "learning_rate": 8e-05,
      "loss": 1.762,
      "step": 1100
    },
    {
      "epoch": 0.2421643022104916,
      "grad_norm": 0.259435772895813,
      "learning_rate": 8e-05,
      "loss": 1.5474,
      "step": 1101
    },
    {
      "epoch": 0.2423842516221269,
      "grad_norm": 0.2759372591972351,
      "learning_rate": 8e-05,
      "loss": 1.6535,
      "step": 1102
    },
    {
      "epoch": 0.24260420103376223,
      "grad_norm": 0.27163347601890564,
      "learning_rate": 8e-05,
      "loss": 1.6035,
      "step": 1103
    },
    {
      "epoch": 0.24282415044539757,
      "grad_norm": 0.26722922921180725,
      "learning_rate": 8e-05,
      "loss": 1.7607,
      "step": 1104
    },
    {
      "epoch": 0.24304409985703287,
      "grad_norm": 0.2925039529800415,
      "learning_rate": 8e-05,
      "loss": 1.6441,
      "step": 1105
    },
    {
      "epoch": 0.2432640492686682,
      "grad_norm": 0.271672785282135,
      "learning_rate": 8e-05,
      "loss": 1.658,
      "step": 1106
    },
    {
      "epoch": 0.24348399868030354,
      "grad_norm": 0.2827896773815155,
      "learning_rate": 8e-05,
      "loss": 1.6258,
      "step": 1107
    },
    {
      "epoch": 0.24370394809193885,
      "grad_norm": 0.2732497751712799,
      "learning_rate": 8e-05,
      "loss": 1.6379,
      "step": 1108
    },
    {
      "epoch": 0.24392389750357418,
      "grad_norm": 0.28081193566322327,
      "learning_rate": 8e-05,
      "loss": 1.7901,
      "step": 1109
    },
    {
      "epoch": 0.2441438469152095,
      "grad_norm": 0.2799675762653351,
      "learning_rate": 8e-05,
      "loss": 1.8323,
      "step": 1110
    },
    {
      "epoch": 0.24436379632684482,
      "grad_norm": 0.2677648961544037,
      "learning_rate": 8e-05,
      "loss": 1.7372,
      "step": 1111
    },
    {
      "epoch": 0.24458374573848016,
      "grad_norm": 0.2644648551940918,
      "learning_rate": 8e-05,
      "loss": 1.6594,
      "step": 1112
    },
    {
      "epoch": 0.24480369515011546,
      "grad_norm": 0.2704750895500183,
      "learning_rate": 8e-05,
      "loss": 1.706,
      "step": 1113
    },
    {
      "epoch": 0.2450236445617508,
      "grad_norm": 0.2762587368488312,
      "learning_rate": 8e-05,
      "loss": 1.7445,
      "step": 1114
    },
    {
      "epoch": 0.24524359397338613,
      "grad_norm": 0.2578018307685852,
      "learning_rate": 8e-05,
      "loss": 1.5707,
      "step": 1115
    },
    {
      "epoch": 0.24546354338502144,
      "grad_norm": 0.2892129719257355,
      "learning_rate": 8e-05,
      "loss": 1.805,
      "step": 1116
    },
    {
      "epoch": 0.24568349279665677,
      "grad_norm": 0.2868081033229828,
      "learning_rate": 8e-05,
      "loss": 1.7756,
      "step": 1117
    },
    {
      "epoch": 0.2459034422082921,
      "grad_norm": 0.2820534110069275,
      "learning_rate": 8e-05,
      "loss": 1.7826,
      "step": 1118
    },
    {
      "epoch": 0.2461233916199274,
      "grad_norm": 0.2824958264827728,
      "learning_rate": 8e-05,
      "loss": 1.6752,
      "step": 1119
    },
    {
      "epoch": 0.24634334103156275,
      "grad_norm": 0.2782610356807709,
      "learning_rate": 8e-05,
      "loss": 1.7536,
      "step": 1120
    },
    {
      "epoch": 0.24656329044319805,
      "grad_norm": 0.27147912979125977,
      "learning_rate": 8e-05,
      "loss": 1.6783,
      "step": 1121
    },
    {
      "epoch": 0.24678323985483339,
      "grad_norm": 0.2740795612335205,
      "learning_rate": 8e-05,
      "loss": 1.7702,
      "step": 1122
    },
    {
      "epoch": 0.24700318926646872,
      "grad_norm": 0.2922619879245758,
      "learning_rate": 8e-05,
      "loss": 1.8204,
      "step": 1123
    },
    {
      "epoch": 0.24722313867810403,
      "grad_norm": 0.2872619926929474,
      "learning_rate": 8e-05,
      "loss": 1.714,
      "step": 1124
    },
    {
      "epoch": 0.24744308808973936,
      "grad_norm": 0.27333369851112366,
      "learning_rate": 8e-05,
      "loss": 1.6575,
      "step": 1125
    },
    {
      "epoch": 0.2476630375013747,
      "grad_norm": 0.28192320466041565,
      "learning_rate": 8e-05,
      "loss": 1.7221,
      "step": 1126
    },
    {
      "epoch": 0.24788298691301,
      "grad_norm": 0.26607248187065125,
      "learning_rate": 8e-05,
      "loss": 1.7262,
      "step": 1127
    },
    {
      "epoch": 0.24810293632464533,
      "grad_norm": 0.279690682888031,
      "learning_rate": 8e-05,
      "loss": 1.7004,
      "step": 1128
    },
    {
      "epoch": 0.24832288573628067,
      "grad_norm": 0.27289190888404846,
      "learning_rate": 8e-05,
      "loss": 1.6916,
      "step": 1129
    },
    {
      "epoch": 0.24854283514791597,
      "grad_norm": 0.27388349175453186,
      "learning_rate": 8e-05,
      "loss": 1.6656,
      "step": 1130
    },
    {
      "epoch": 0.2487627845595513,
      "grad_norm": 0.2912501096725464,
      "learning_rate": 8e-05,
      "loss": 1.8086,
      "step": 1131
    },
    {
      "epoch": 0.24898273397118661,
      "grad_norm": 0.2999799847602844,
      "learning_rate": 8e-05,
      "loss": 1.7659,
      "step": 1132
    },
    {
      "epoch": 0.24920268338282195,
      "grad_norm": 0.262207955121994,
      "learning_rate": 8e-05,
      "loss": 1.6581,
      "step": 1133
    },
    {
      "epoch": 0.24942263279445728,
      "grad_norm": 0.2571624517440796,
      "learning_rate": 8e-05,
      "loss": 1.6509,
      "step": 1134
    },
    {
      "epoch": 0.2496425822060926,
      "grad_norm": 0.26213690638542175,
      "learning_rate": 8e-05,
      "loss": 1.6044,
      "step": 1135
    },
    {
      "epoch": 0.24986253161772792,
      "grad_norm": 0.2870398461818695,
      "learning_rate": 8e-05,
      "loss": 1.6678,
      "step": 1136
    },
    {
      "epoch": 0.25008248102936326,
      "grad_norm": 0.2672583758831024,
      "learning_rate": 8e-05,
      "loss": 1.6563,
      "step": 1137
    },
    {
      "epoch": 0.2503024304409986,
      "grad_norm": 0.29864680767059326,
      "learning_rate": 8e-05,
      "loss": 1.858,
      "step": 1138
    },
    {
      "epoch": 0.25052237985263387,
      "grad_norm": 0.3096907436847687,
      "learning_rate": 8e-05,
      "loss": 1.7731,
      "step": 1139
    },
    {
      "epoch": 0.2507423292642692,
      "grad_norm": 0.2668014466762543,
      "learning_rate": 8e-05,
      "loss": 1.6173,
      "step": 1140
    },
    {
      "epoch": 0.25096227867590454,
      "grad_norm": 0.275074303150177,
      "learning_rate": 8e-05,
      "loss": 1.704,
      "step": 1141
    },
    {
      "epoch": 0.25118222808753987,
      "grad_norm": 0.29657119512557983,
      "learning_rate": 8e-05,
      "loss": 1.9789,
      "step": 1142
    },
    {
      "epoch": 0.2514021774991752,
      "grad_norm": 0.26117807626724243,
      "learning_rate": 8e-05,
      "loss": 1.6815,
      "step": 1143
    },
    {
      "epoch": 0.25162212691081054,
      "grad_norm": 0.2738019824028015,
      "learning_rate": 8e-05,
      "loss": 1.7031,
      "step": 1144
    },
    {
      "epoch": 0.2518420763224458,
      "grad_norm": 0.27922967076301575,
      "learning_rate": 8e-05,
      "loss": 1.7914,
      "step": 1145
    },
    {
      "epoch": 0.25206202573408115,
      "grad_norm": 0.2876172661781311,
      "learning_rate": 8e-05,
      "loss": 1.721,
      "step": 1146
    },
    {
      "epoch": 0.2522819751457165,
      "grad_norm": 0.28017961978912354,
      "learning_rate": 8e-05,
      "loss": 1.6731,
      "step": 1147
    },
    {
      "epoch": 0.2525019245573518,
      "grad_norm": 0.2898389399051666,
      "learning_rate": 8e-05,
      "loss": 1.8749,
      "step": 1148
    },
    {
      "epoch": 0.25272187396898715,
      "grad_norm": 0.2742408812046051,
      "learning_rate": 8e-05,
      "loss": 1.6811,
      "step": 1149
    },
    {
      "epoch": 0.25294182338062243,
      "grad_norm": 0.2806207835674286,
      "learning_rate": 8e-05,
      "loss": 1.7082,
      "step": 1150
    },
    {
      "epoch": 0.25316177279225777,
      "grad_norm": 0.27871328592300415,
      "learning_rate": 8e-05,
      "loss": 1.7142,
      "step": 1151
    },
    {
      "epoch": 0.2533817222038931,
      "grad_norm": 0.2792799472808838,
      "learning_rate": 8e-05,
      "loss": 1.5703,
      "step": 1152
    },
    {
      "epoch": 0.25360167161552843,
      "grad_norm": 0.27358901500701904,
      "learning_rate": 8e-05,
      "loss": 1.7576,
      "step": 1153
    },
    {
      "epoch": 0.25382162102716377,
      "grad_norm": 0.26983192563056946,
      "learning_rate": 8e-05,
      "loss": 1.6646,
      "step": 1154
    },
    {
      "epoch": 0.2540415704387991,
      "grad_norm": 0.2711959183216095,
      "learning_rate": 8e-05,
      "loss": 1.7698,
      "step": 1155
    },
    {
      "epoch": 0.2542615198504344,
      "grad_norm": 0.28412333130836487,
      "learning_rate": 8e-05,
      "loss": 1.7446,
      "step": 1156
    },
    {
      "epoch": 0.2544814692620697,
      "grad_norm": 0.2698575258255005,
      "learning_rate": 8e-05,
      "loss": 1.6861,
      "step": 1157
    },
    {
      "epoch": 0.25470141867370505,
      "grad_norm": 0.2806732952594757,
      "learning_rate": 8e-05,
      "loss": 1.7308,
      "step": 1158
    },
    {
      "epoch": 0.2549213680853404,
      "grad_norm": 0.2715948522090912,
      "learning_rate": 8e-05,
      "loss": 1.852,
      "step": 1159
    },
    {
      "epoch": 0.2551413174969757,
      "grad_norm": 0.33048170804977417,
      "learning_rate": 8e-05,
      "loss": 1.881,
      "step": 1160
    },
    {
      "epoch": 0.255361266908611,
      "grad_norm": 0.27907994389533997,
      "learning_rate": 8e-05,
      "loss": 1.6501,
      "step": 1161
    },
    {
      "epoch": 0.25558121632024633,
      "grad_norm": 0.2747988998889923,
      "learning_rate": 8e-05,
      "loss": 1.7265,
      "step": 1162
    },
    {
      "epoch": 0.25580116573188166,
      "grad_norm": 0.28321677446365356,
      "learning_rate": 8e-05,
      "loss": 1.8602,
      "step": 1163
    },
    {
      "epoch": 0.256021115143517,
      "grad_norm": 0.2695465683937073,
      "learning_rate": 8e-05,
      "loss": 1.6091,
      "step": 1164
    },
    {
      "epoch": 0.25624106455515233,
      "grad_norm": 0.272135466337204,
      "learning_rate": 8e-05,
      "loss": 1.6236,
      "step": 1165
    },
    {
      "epoch": 0.25646101396678767,
      "grad_norm": 0.2715020775794983,
      "learning_rate": 8e-05,
      "loss": 1.674,
      "step": 1166
    },
    {
      "epoch": 0.25668096337842294,
      "grad_norm": 0.2879820764064789,
      "learning_rate": 8e-05,
      "loss": 1.8393,
      "step": 1167
    },
    {
      "epoch": 0.2569009127900583,
      "grad_norm": 0.2616657018661499,
      "learning_rate": 8e-05,
      "loss": 1.6391,
      "step": 1168
    },
    {
      "epoch": 0.2571208622016936,
      "grad_norm": 0.2558441460132599,
      "learning_rate": 8e-05,
      "loss": 1.606,
      "step": 1169
    },
    {
      "epoch": 0.25734081161332895,
      "grad_norm": 0.26944512128829956,
      "learning_rate": 8e-05,
      "loss": 1.7288,
      "step": 1170
    },
    {
      "epoch": 0.2575607610249643,
      "grad_norm": 0.26958367228507996,
      "learning_rate": 8e-05,
      "loss": 1.6233,
      "step": 1171
    },
    {
      "epoch": 0.25778071043659956,
      "grad_norm": 0.29003527760505676,
      "learning_rate": 8e-05,
      "loss": 1.777,
      "step": 1172
    },
    {
      "epoch": 0.2580006598482349,
      "grad_norm": 0.2677457630634308,
      "learning_rate": 8e-05,
      "loss": 1.6835,
      "step": 1173
    },
    {
      "epoch": 0.2582206092598702,
      "grad_norm": 0.28062689304351807,
      "learning_rate": 8e-05,
      "loss": 1.726,
      "step": 1174
    },
    {
      "epoch": 0.25844055867150556,
      "grad_norm": 0.26764920353889465,
      "learning_rate": 8e-05,
      "loss": 1.6575,
      "step": 1175
    },
    {
      "epoch": 0.2586605080831409,
      "grad_norm": 0.28183332085609436,
      "learning_rate": 8e-05,
      "loss": 1.784,
      "step": 1176
    },
    {
      "epoch": 0.25888045749477623,
      "grad_norm": 0.25718390941619873,
      "learning_rate": 8e-05,
      "loss": 1.6317,
      "step": 1177
    },
    {
      "epoch": 0.2591004069064115,
      "grad_norm": 0.25523149967193604,
      "learning_rate": 8e-05,
      "loss": 1.5634,
      "step": 1178
    },
    {
      "epoch": 0.25932035631804684,
      "grad_norm": 0.2539874315261841,
      "learning_rate": 8e-05,
      "loss": 1.5878,
      "step": 1179
    },
    {
      "epoch": 0.2595403057296822,
      "grad_norm": 0.2868393659591675,
      "learning_rate": 8e-05,
      "loss": 1.7301,
      "step": 1180
    },
    {
      "epoch": 0.2597602551413175,
      "grad_norm": 0.27819645404815674,
      "learning_rate": 8e-05,
      "loss": 1.6895,
      "step": 1181
    },
    {
      "epoch": 0.25998020455295284,
      "grad_norm": 0.27499255537986755,
      "learning_rate": 8e-05,
      "loss": 1.7462,
      "step": 1182
    },
    {
      "epoch": 0.2602001539645881,
      "grad_norm": 0.2858695685863495,
      "learning_rate": 8e-05,
      "loss": 1.8199,
      "step": 1183
    },
    {
      "epoch": 0.26042010337622346,
      "grad_norm": 0.2646760642528534,
      "learning_rate": 8e-05,
      "loss": 1.6597,
      "step": 1184
    },
    {
      "epoch": 0.2606400527878588,
      "grad_norm": 0.2831268310546875,
      "learning_rate": 8e-05,
      "loss": 1.8383,
      "step": 1185
    },
    {
      "epoch": 0.2608600021994941,
      "grad_norm": 0.2593746483325958,
      "learning_rate": 8e-05,
      "loss": 1.6115,
      "step": 1186
    },
    {
      "epoch": 0.26107995161112946,
      "grad_norm": 0.26519641280174255,
      "learning_rate": 8e-05,
      "loss": 1.5959,
      "step": 1187
    },
    {
      "epoch": 0.2612999010227648,
      "grad_norm": 0.2733252942562103,
      "learning_rate": 8e-05,
      "loss": 1.6318,
      "step": 1188
    },
    {
      "epoch": 0.26151985043440007,
      "grad_norm": 0.27299511432647705,
      "learning_rate": 8e-05,
      "loss": 1.7313,
      "step": 1189
    },
    {
      "epoch": 0.2617397998460354,
      "grad_norm": 0.2684955894947052,
      "learning_rate": 8e-05,
      "loss": 1.5826,
      "step": 1190
    },
    {
      "epoch": 0.26195974925767074,
      "grad_norm": 0.2747553586959839,
      "learning_rate": 8e-05,
      "loss": 1.7008,
      "step": 1191
    },
    {
      "epoch": 0.26217969866930607,
      "grad_norm": 0.26033639907836914,
      "learning_rate": 8e-05,
      "loss": 1.5571,
      "step": 1192
    },
    {
      "epoch": 0.2623996480809414,
      "grad_norm": 0.2640804350376129,
      "learning_rate": 8e-05,
      "loss": 1.5317,
      "step": 1193
    },
    {
      "epoch": 0.2626195974925767,
      "grad_norm": 0.27063700556755066,
      "learning_rate": 8e-05,
      "loss": 1.5501,
      "step": 1194
    },
    {
      "epoch": 0.262839546904212,
      "grad_norm": 0.2677111029624939,
      "learning_rate": 8e-05,
      "loss": 1.5894,
      "step": 1195
    },
    {
      "epoch": 0.26305949631584735,
      "grad_norm": 0.28144168853759766,
      "learning_rate": 8e-05,
      "loss": 1.7496,
      "step": 1196
    },
    {
      "epoch": 0.2632794457274827,
      "grad_norm": 0.2602388858795166,
      "learning_rate": 8e-05,
      "loss": 1.571,
      "step": 1197
    },
    {
      "epoch": 0.263499395139118,
      "grad_norm": 0.2941505014896393,
      "learning_rate": 8e-05,
      "loss": 1.6692,
      "step": 1198
    },
    {
      "epoch": 0.26371934455075335,
      "grad_norm": 0.264433354139328,
      "learning_rate": 8e-05,
      "loss": 1.6922,
      "step": 1199
    },
    {
      "epoch": 0.26393929396238863,
      "grad_norm": 0.25587090849876404,
      "learning_rate": 8e-05,
      "loss": 1.5599,
      "step": 1200
    },
    {
      "epoch": 0.26415924337402397,
      "grad_norm": 0.3012869358062744,
      "learning_rate": 8e-05,
      "loss": 1.9195,
      "step": 1201
    },
    {
      "epoch": 0.2643791927856593,
      "grad_norm": 0.2762719392776489,
      "learning_rate": 8e-05,
      "loss": 1.898,
      "step": 1202
    },
    {
      "epoch": 0.26459914219729463,
      "grad_norm": 0.2701188325881958,
      "learning_rate": 8e-05,
      "loss": 1.7312,
      "step": 1203
    },
    {
      "epoch": 0.26481909160892997,
      "grad_norm": 0.29665982723236084,
      "learning_rate": 8e-05,
      "loss": 1.8089,
      "step": 1204
    },
    {
      "epoch": 0.26503904102056525,
      "grad_norm": 0.26700517535209656,
      "learning_rate": 8e-05,
      "loss": 1.8401,
      "step": 1205
    },
    {
      "epoch": 0.2652589904322006,
      "grad_norm": 0.2828493118286133,
      "learning_rate": 8e-05,
      "loss": 1.8622,
      "step": 1206
    },
    {
      "epoch": 0.2654789398438359,
      "grad_norm": 0.2746271789073944,
      "learning_rate": 8e-05,
      "loss": 1.6521,
      "step": 1207
    },
    {
      "epoch": 0.26569888925547125,
      "grad_norm": 0.2882270812988281,
      "learning_rate": 8e-05,
      "loss": 1.7168,
      "step": 1208
    },
    {
      "epoch": 0.2659188386671066,
      "grad_norm": 0.29784512519836426,
      "learning_rate": 8e-05,
      "loss": 1.6968,
      "step": 1209
    },
    {
      "epoch": 0.2661387880787419,
      "grad_norm": 0.2807427942752838,
      "learning_rate": 8e-05,
      "loss": 1.6004,
      "step": 1210
    },
    {
      "epoch": 0.2663587374903772,
      "grad_norm": 0.2956424951553345,
      "learning_rate": 8e-05,
      "loss": 1.8325,
      "step": 1211
    },
    {
      "epoch": 0.26657868690201253,
      "grad_norm": 0.2647739350795746,
      "learning_rate": 8e-05,
      "loss": 1.6391,
      "step": 1212
    },
    {
      "epoch": 0.26679863631364786,
      "grad_norm": 0.2955171465873718,
      "learning_rate": 8e-05,
      "loss": 1.7893,
      "step": 1213
    },
    {
      "epoch": 0.2670185857252832,
      "grad_norm": 0.27241894602775574,
      "learning_rate": 8e-05,
      "loss": 1.781,
      "step": 1214
    },
    {
      "epoch": 0.26723853513691853,
      "grad_norm": 0.2841251492500305,
      "learning_rate": 8e-05,
      "loss": 1.8612,
      "step": 1215
    },
    {
      "epoch": 0.2674584845485538,
      "grad_norm": 0.327891081571579,
      "learning_rate": 8e-05,
      "loss": 1.844,
      "step": 1216
    },
    {
      "epoch": 0.26767843396018914,
      "grad_norm": 0.26434099674224854,
      "learning_rate": 8e-05,
      "loss": 1.6325,
      "step": 1217
    },
    {
      "epoch": 0.2678983833718245,
      "grad_norm": 0.2868417799472809,
      "learning_rate": 8e-05,
      "loss": 1.7087,
      "step": 1218
    },
    {
      "epoch": 0.2681183327834598,
      "grad_norm": 0.27408069372177124,
      "learning_rate": 8e-05,
      "loss": 1.6006,
      "step": 1219
    },
    {
      "epoch": 0.26833828219509515,
      "grad_norm": 0.2697390019893646,
      "learning_rate": 8e-05,
      "loss": 1.6833,
      "step": 1220
    },
    {
      "epoch": 0.2685582316067304,
      "grad_norm": 0.27598559856414795,
      "learning_rate": 8e-05,
      "loss": 1.7192,
      "step": 1221
    },
    {
      "epoch": 0.26877818101836576,
      "grad_norm": 0.26871007680892944,
      "learning_rate": 8e-05,
      "loss": 1.6301,
      "step": 1222
    },
    {
      "epoch": 0.2689981304300011,
      "grad_norm": 0.2739337980747223,
      "learning_rate": 8e-05,
      "loss": 1.6828,
      "step": 1223
    },
    {
      "epoch": 0.2692180798416364,
      "grad_norm": 0.286530464887619,
      "learning_rate": 8e-05,
      "loss": 1.6484,
      "step": 1224
    },
    {
      "epoch": 0.26943802925327176,
      "grad_norm": 0.27509886026382446,
      "learning_rate": 8e-05,
      "loss": 1.6647,
      "step": 1225
    },
    {
      "epoch": 0.2696579786649071,
      "grad_norm": 0.2916969358921051,
      "learning_rate": 8e-05,
      "loss": 1.7908,
      "step": 1226
    },
    {
      "epoch": 0.2698779280765424,
      "grad_norm": 0.26566174626350403,
      "learning_rate": 8e-05,
      "loss": 1.6075,
      "step": 1227
    },
    {
      "epoch": 0.2700978774881777,
      "grad_norm": 0.27648022770881653,
      "learning_rate": 8e-05,
      "loss": 1.7536,
      "step": 1228
    },
    {
      "epoch": 0.27031782689981304,
      "grad_norm": 0.27313023805618286,
      "learning_rate": 8e-05,
      "loss": 1.6978,
      "step": 1229
    },
    {
      "epoch": 0.2705377763114484,
      "grad_norm": 0.2755061388015747,
      "learning_rate": 8e-05,
      "loss": 1.7196,
      "step": 1230
    },
    {
      "epoch": 0.2707577257230837,
      "grad_norm": 0.25907769799232483,
      "learning_rate": 8e-05,
      "loss": 1.5518,
      "step": 1231
    },
    {
      "epoch": 0.270977675134719,
      "grad_norm": 0.26485681533813477,
      "learning_rate": 8e-05,
      "loss": 1.5053,
      "step": 1232
    },
    {
      "epoch": 0.2711976245463543,
      "grad_norm": 0.27980178594589233,
      "learning_rate": 8e-05,
      "loss": 1.7824,
      "step": 1233
    },
    {
      "epoch": 0.27141757395798966,
      "grad_norm": 0.2750954329967499,
      "learning_rate": 8e-05,
      "loss": 1.6973,
      "step": 1234
    },
    {
      "epoch": 0.271637523369625,
      "grad_norm": 0.27367594838142395,
      "learning_rate": 8e-05,
      "loss": 1.6691,
      "step": 1235
    },
    {
      "epoch": 0.2718574727812603,
      "grad_norm": 0.27089521288871765,
      "learning_rate": 8e-05,
      "loss": 1.7532,
      "step": 1236
    },
    {
      "epoch": 0.27207742219289566,
      "grad_norm": 0.30656641721725464,
      "learning_rate": 8e-05,
      "loss": 1.8411,
      "step": 1237
    },
    {
      "epoch": 0.27229737160453094,
      "grad_norm": 0.25732672214508057,
      "learning_rate": 8e-05,
      "loss": 1.5599,
      "step": 1238
    },
    {
      "epoch": 0.27251732101616627,
      "grad_norm": 0.2643807828426361,
      "learning_rate": 8e-05,
      "loss": 1.6654,
      "step": 1239
    },
    {
      "epoch": 0.2727372704278016,
      "grad_norm": 0.2703326344490051,
      "learning_rate": 8e-05,
      "loss": 1.594,
      "step": 1240
    },
    {
      "epoch": 0.27295721983943694,
      "grad_norm": 0.27907243371009827,
      "learning_rate": 8e-05,
      "loss": 1.7531,
      "step": 1241
    },
    {
      "epoch": 0.27317716925107227,
      "grad_norm": 0.2482902854681015,
      "learning_rate": 8e-05,
      "loss": 1.3586,
      "step": 1242
    },
    {
      "epoch": 0.27339711866270755,
      "grad_norm": 0.2879469394683838,
      "learning_rate": 8e-05,
      "loss": 1.76,
      "step": 1243
    },
    {
      "epoch": 0.2736170680743429,
      "grad_norm": 0.26334571838378906,
      "learning_rate": 8e-05,
      "loss": 1.536,
      "step": 1244
    },
    {
      "epoch": 0.2738370174859782,
      "grad_norm": 0.27328065037727356,
      "learning_rate": 8e-05,
      "loss": 1.7199,
      "step": 1245
    },
    {
      "epoch": 0.27405696689761355,
      "grad_norm": 0.27392926812171936,
      "learning_rate": 8e-05,
      "loss": 1.7731,
      "step": 1246
    },
    {
      "epoch": 0.2742769163092489,
      "grad_norm": 0.29755476117134094,
      "learning_rate": 8e-05,
      "loss": 1.7184,
      "step": 1247
    },
    {
      "epoch": 0.2744968657208842,
      "grad_norm": 0.29554107785224915,
      "learning_rate": 8e-05,
      "loss": 1.7442,
      "step": 1248
    },
    {
      "epoch": 0.2747168151325195,
      "grad_norm": 0.2562367618083954,
      "learning_rate": 8e-05,
      "loss": 1.63,
      "step": 1249
    },
    {
      "epoch": 0.27493676454415483,
      "grad_norm": 0.27746453881263733,
      "learning_rate": 8e-05,
      "loss": 1.7396,
      "step": 1250
    },
    {
      "epoch": 0.27515671395579017,
      "grad_norm": 0.2747843265533447,
      "learning_rate": 8e-05,
      "loss": 1.6628,
      "step": 1251
    },
    {
      "epoch": 0.2753766633674255,
      "grad_norm": 0.2650463581085205,
      "learning_rate": 8e-05,
      "loss": 1.6409,
      "step": 1252
    },
    {
      "epoch": 0.27559661277906083,
      "grad_norm": 0.30537328124046326,
      "learning_rate": 8e-05,
      "loss": 1.4927,
      "step": 1253
    },
    {
      "epoch": 0.2758165621906961,
      "grad_norm": 0.26015424728393555,
      "learning_rate": 8e-05,
      "loss": 1.718,
      "step": 1254
    },
    {
      "epoch": 0.27603651160233145,
      "grad_norm": 0.2512992322444916,
      "learning_rate": 8e-05,
      "loss": 1.4757,
      "step": 1255
    },
    {
      "epoch": 0.2762564610139668,
      "grad_norm": 0.28478461503982544,
      "learning_rate": 8e-05,
      "loss": 1.9081,
      "step": 1256
    },
    {
      "epoch": 0.2764764104256021,
      "grad_norm": 0.28490516543388367,
      "learning_rate": 8e-05,
      "loss": 1.8495,
      "step": 1257
    },
    {
      "epoch": 0.27669635983723745,
      "grad_norm": 0.2758481204509735,
      "learning_rate": 8e-05,
      "loss": 1.767,
      "step": 1258
    },
    {
      "epoch": 0.2769163092488728,
      "grad_norm": 0.28743213415145874,
      "learning_rate": 8e-05,
      "loss": 1.6548,
      "step": 1259
    },
    {
      "epoch": 0.27713625866050806,
      "grad_norm": 0.2738385796546936,
      "learning_rate": 8e-05,
      "loss": 1.5616,
      "step": 1260
    },
    {
      "epoch": 0.2773562080721434,
      "grad_norm": 0.27758583426475525,
      "learning_rate": 8e-05,
      "loss": 1.7793,
      "step": 1261
    },
    {
      "epoch": 0.27757615748377873,
      "grad_norm": 0.2830480635166168,
      "learning_rate": 8e-05,
      "loss": 1.8048,
      "step": 1262
    },
    {
      "epoch": 0.27779610689541406,
      "grad_norm": 0.296036034822464,
      "learning_rate": 8e-05,
      "loss": 1.7844,
      "step": 1263
    },
    {
      "epoch": 0.2780160563070494,
      "grad_norm": 0.28651297092437744,
      "learning_rate": 8e-05,
      "loss": 1.7239,
      "step": 1264
    },
    {
      "epoch": 0.2782360057186847,
      "grad_norm": 0.2826116979122162,
      "learning_rate": 8e-05,
      "loss": 1.8415,
      "step": 1265
    },
    {
      "epoch": 0.27845595513032,
      "grad_norm": 0.27445724606513977,
      "learning_rate": 8e-05,
      "loss": 1.6738,
      "step": 1266
    },
    {
      "epoch": 0.27867590454195534,
      "grad_norm": 0.28153640031814575,
      "learning_rate": 8e-05,
      "loss": 1.6519,
      "step": 1267
    },
    {
      "epoch": 0.2788958539535907,
      "grad_norm": 0.27389946579933167,
      "learning_rate": 8e-05,
      "loss": 1.681,
      "step": 1268
    },
    {
      "epoch": 0.279115803365226,
      "grad_norm": 0.2639203667640686,
      "learning_rate": 8e-05,
      "loss": 1.6398,
      "step": 1269
    },
    {
      "epoch": 0.27933575277686135,
      "grad_norm": 0.2787509560585022,
      "learning_rate": 8e-05,
      "loss": 1.7199,
      "step": 1270
    },
    {
      "epoch": 0.2795557021884966,
      "grad_norm": 0.28468430042266846,
      "learning_rate": 8e-05,
      "loss": 1.8668,
      "step": 1271
    },
    {
      "epoch": 0.27977565160013196,
      "grad_norm": 0.2907005250453949,
      "learning_rate": 8e-05,
      "loss": 1.9328,
      "step": 1272
    },
    {
      "epoch": 0.2799956010117673,
      "grad_norm": 0.2607463300228119,
      "learning_rate": 8e-05,
      "loss": 1.5958,
      "step": 1273
    },
    {
      "epoch": 0.2802155504234026,
      "grad_norm": 0.2695181965827942,
      "learning_rate": 8e-05,
      "loss": 1.6708,
      "step": 1274
    },
    {
      "epoch": 0.28043549983503796,
      "grad_norm": 0.28671538829803467,
      "learning_rate": 8e-05,
      "loss": 1.7736,
      "step": 1275
    },
    {
      "epoch": 0.28065544924667324,
      "grad_norm": 0.3246489465236664,
      "learning_rate": 8e-05,
      "loss": 1.8145,
      "step": 1276
    },
    {
      "epoch": 0.2808753986583086,
      "grad_norm": 0.2879314720630646,
      "learning_rate": 8e-05,
      "loss": 1.782,
      "step": 1277
    },
    {
      "epoch": 0.2810953480699439,
      "grad_norm": 0.27141574025154114,
      "learning_rate": 8e-05,
      "loss": 1.8069,
      "step": 1278
    },
    {
      "epoch": 0.28131529748157924,
      "grad_norm": 0.2893892228603363,
      "learning_rate": 8e-05,
      "loss": 1.7893,
      "step": 1279
    },
    {
      "epoch": 0.2815352468932146,
      "grad_norm": 0.2985538840293884,
      "learning_rate": 8e-05,
      "loss": 1.7804,
      "step": 1280
    },
    {
      "epoch": 0.2817551963048499,
      "grad_norm": 0.2664276957511902,
      "learning_rate": 8e-05,
      "loss": 1.6785,
      "step": 1281
    },
    {
      "epoch": 0.2819751457164852,
      "grad_norm": 0.3002198040485382,
      "learning_rate": 8e-05,
      "loss": 1.6109,
      "step": 1282
    },
    {
      "epoch": 0.2821950951281205,
      "grad_norm": 0.27687907218933105,
      "learning_rate": 8e-05,
      "loss": 1.6322,
      "step": 1283
    },
    {
      "epoch": 0.28241504453975586,
      "grad_norm": 0.28822144865989685,
      "learning_rate": 8e-05,
      "loss": 1.6785,
      "step": 1284
    },
    {
      "epoch": 0.2826349939513912,
      "grad_norm": 0.2801685333251953,
      "learning_rate": 8e-05,
      "loss": 1.69,
      "step": 1285
    },
    {
      "epoch": 0.2828549433630265,
      "grad_norm": 0.27876734733581543,
      "learning_rate": 8e-05,
      "loss": 1.6442,
      "step": 1286
    },
    {
      "epoch": 0.2830748927746618,
      "grad_norm": 0.2990095317363739,
      "learning_rate": 8e-05,
      "loss": 1.7439,
      "step": 1287
    },
    {
      "epoch": 0.28329484218629714,
      "grad_norm": 0.2710682451725006,
      "learning_rate": 8e-05,
      "loss": 1.6908,
      "step": 1288
    },
    {
      "epoch": 0.28351479159793247,
      "grad_norm": 0.2922731935977936,
      "learning_rate": 8e-05,
      "loss": 1.8361,
      "step": 1289
    },
    {
      "epoch": 0.2837347410095678,
      "grad_norm": 0.2638223171234131,
      "learning_rate": 8e-05,
      "loss": 1.6233,
      "step": 1290
    },
    {
      "epoch": 0.28395469042120314,
      "grad_norm": 0.27564552426338196,
      "learning_rate": 8e-05,
      "loss": 1.7624,
      "step": 1291
    },
    {
      "epoch": 0.28417463983283847,
      "grad_norm": 0.28238940238952637,
      "learning_rate": 8e-05,
      "loss": 1.8649,
      "step": 1292
    },
    {
      "epoch": 0.28439458924447375,
      "grad_norm": 0.27798035740852356,
      "learning_rate": 8e-05,
      "loss": 1.7877,
      "step": 1293
    },
    {
      "epoch": 0.2846145386561091,
      "grad_norm": 0.29618534445762634,
      "learning_rate": 8e-05,
      "loss": 1.816,
      "step": 1294
    },
    {
      "epoch": 0.2848344880677444,
      "grad_norm": 0.27669045329093933,
      "learning_rate": 8e-05,
      "loss": 1.7014,
      "step": 1295
    },
    {
      "epoch": 0.28505443747937975,
      "grad_norm": 0.27973508834838867,
      "learning_rate": 8e-05,
      "loss": 1.7491,
      "step": 1296
    },
    {
      "epoch": 0.2852743868910151,
      "grad_norm": 0.28833356499671936,
      "learning_rate": 8e-05,
      "loss": 1.6948,
      "step": 1297
    },
    {
      "epoch": 0.28549433630265036,
      "grad_norm": 0.2751030921936035,
      "learning_rate": 8e-05,
      "loss": 1.6846,
      "step": 1298
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 0.2766781449317932,
      "learning_rate": 8e-05,
      "loss": 1.5442,
      "step": 1299
    },
    {
      "epoch": 0.28593423512592103,
      "grad_norm": 0.29664894938468933,
      "learning_rate": 8e-05,
      "loss": 1.6884,
      "step": 1300
    },
    {
      "epoch": 0.28615418453755637,
      "grad_norm": 0.2771795392036438,
      "learning_rate": 8e-05,
      "loss": 1.6479,
      "step": 1301
    },
    {
      "epoch": 0.2863741339491917,
      "grad_norm": 0.2623322904109955,
      "learning_rate": 8e-05,
      "loss": 1.5803,
      "step": 1302
    },
    {
      "epoch": 0.28659408336082703,
      "grad_norm": 0.2821153998374939,
      "learning_rate": 8e-05,
      "loss": 1.7758,
      "step": 1303
    },
    {
      "epoch": 0.2868140327724623,
      "grad_norm": 0.29058384895324707,
      "learning_rate": 8e-05,
      "loss": 1.7244,
      "step": 1304
    },
    {
      "epoch": 0.28703398218409765,
      "grad_norm": 0.2811940312385559,
      "learning_rate": 8e-05,
      "loss": 1.6708,
      "step": 1305
    },
    {
      "epoch": 0.287253931595733,
      "grad_norm": 0.2773367762565613,
      "learning_rate": 8e-05,
      "loss": 1.7857,
      "step": 1306
    },
    {
      "epoch": 0.2874738810073683,
      "grad_norm": 0.2689999043941498,
      "learning_rate": 8e-05,
      "loss": 1.7432,
      "step": 1307
    },
    {
      "epoch": 0.28769383041900365,
      "grad_norm": 0.26896870136260986,
      "learning_rate": 8e-05,
      "loss": 1.6389,
      "step": 1308
    },
    {
      "epoch": 0.2879137798306389,
      "grad_norm": 0.2981964349746704,
      "learning_rate": 8e-05,
      "loss": 1.8771,
      "step": 1309
    },
    {
      "epoch": 0.28813372924227426,
      "grad_norm": 0.2872856855392456,
      "learning_rate": 8e-05,
      "loss": 1.785,
      "step": 1310
    },
    {
      "epoch": 0.2883536786539096,
      "grad_norm": 0.3186649680137634,
      "learning_rate": 8e-05,
      "loss": 1.9051,
      "step": 1311
    },
    {
      "epoch": 0.28857362806554493,
      "grad_norm": 0.2802119255065918,
      "learning_rate": 8e-05,
      "loss": 1.6532,
      "step": 1312
    },
    {
      "epoch": 0.28879357747718026,
      "grad_norm": 0.2864134907722473,
      "learning_rate": 8e-05,
      "loss": 1.7373,
      "step": 1313
    },
    {
      "epoch": 0.2890135268888156,
      "grad_norm": 0.2739737331867218,
      "learning_rate": 8e-05,
      "loss": 1.5365,
      "step": 1314
    },
    {
      "epoch": 0.2892334763004509,
      "grad_norm": 0.2707555294036865,
      "learning_rate": 8e-05,
      "loss": 1.6516,
      "step": 1315
    },
    {
      "epoch": 0.2894534257120862,
      "grad_norm": 0.2895212173461914,
      "learning_rate": 8e-05,
      "loss": 1.5634,
      "step": 1316
    },
    {
      "epoch": 0.28967337512372154,
      "grad_norm": 0.26424047350883484,
      "learning_rate": 8e-05,
      "loss": 1.6543,
      "step": 1317
    },
    {
      "epoch": 0.2898933245353569,
      "grad_norm": 0.26237159967422485,
      "learning_rate": 8e-05,
      "loss": 1.6814,
      "step": 1318
    },
    {
      "epoch": 0.2901132739469922,
      "grad_norm": 0.27964159846305847,
      "learning_rate": 8e-05,
      "loss": 1.7505,
      "step": 1319
    },
    {
      "epoch": 0.2903332233586275,
      "grad_norm": 0.27128270268440247,
      "learning_rate": 8e-05,
      "loss": 1.7229,
      "step": 1320
    },
    {
      "epoch": 0.2905531727702628,
      "grad_norm": 0.3012688159942627,
      "learning_rate": 8e-05,
      "loss": 1.6851,
      "step": 1321
    },
    {
      "epoch": 0.29077312218189816,
      "grad_norm": 0.2725695073604584,
      "learning_rate": 8e-05,
      "loss": 1.6552,
      "step": 1322
    },
    {
      "epoch": 0.2909930715935335,
      "grad_norm": 0.2855455279350281,
      "learning_rate": 8e-05,
      "loss": 1.7779,
      "step": 1323
    },
    {
      "epoch": 0.2912130210051688,
      "grad_norm": 0.2906174659729004,
      "learning_rate": 8e-05,
      "loss": 1.8209,
      "step": 1324
    },
    {
      "epoch": 0.29143297041680416,
      "grad_norm": 0.26015472412109375,
      "learning_rate": 8e-05,
      "loss": 1.5403,
      "step": 1325
    },
    {
      "epoch": 0.29165291982843944,
      "grad_norm": 0.29065820574760437,
      "learning_rate": 8e-05,
      "loss": 1.8499,
      "step": 1326
    },
    {
      "epoch": 0.2918728692400748,
      "grad_norm": 0.28715917468070984,
      "learning_rate": 8e-05,
      "loss": 1.854,
      "step": 1327
    },
    {
      "epoch": 0.2920928186517101,
      "grad_norm": 0.26932859420776367,
      "learning_rate": 8e-05,
      "loss": 1.6254,
      "step": 1328
    },
    {
      "epoch": 0.29231276806334544,
      "grad_norm": 0.2757404148578644,
      "learning_rate": 8e-05,
      "loss": 1.6077,
      "step": 1329
    },
    {
      "epoch": 0.2925327174749808,
      "grad_norm": 0.26532551646232605,
      "learning_rate": 8e-05,
      "loss": 1.6551,
      "step": 1330
    },
    {
      "epoch": 0.29275266688661605,
      "grad_norm": 0.2754289209842682,
      "learning_rate": 8e-05,
      "loss": 1.7276,
      "step": 1331
    },
    {
      "epoch": 0.2929726162982514,
      "grad_norm": 0.290568470954895,
      "learning_rate": 8e-05,
      "loss": 1.7622,
      "step": 1332
    },
    {
      "epoch": 0.2931925657098867,
      "grad_norm": 0.3045903742313385,
      "learning_rate": 8e-05,
      "loss": 1.7937,
      "step": 1333
    },
    {
      "epoch": 0.29341251512152206,
      "grad_norm": 0.2594483196735382,
      "learning_rate": 8e-05,
      "loss": 1.6163,
      "step": 1334
    },
    {
      "epoch": 0.2936324645331574,
      "grad_norm": 0.3054102957248688,
      "learning_rate": 8e-05,
      "loss": 1.7767,
      "step": 1335
    },
    {
      "epoch": 0.2938524139447927,
      "grad_norm": 0.27347666025161743,
      "learning_rate": 8e-05,
      "loss": 1.682,
      "step": 1336
    },
    {
      "epoch": 0.294072363356428,
      "grad_norm": 0.2639494836330414,
      "learning_rate": 8e-05,
      "loss": 1.4616,
      "step": 1337
    },
    {
      "epoch": 0.29429231276806334,
      "grad_norm": 0.2842942178249359,
      "learning_rate": 8e-05,
      "loss": 1.7625,
      "step": 1338
    },
    {
      "epoch": 0.29451226217969867,
      "grad_norm": 0.2895960509777069,
      "learning_rate": 8e-05,
      "loss": 1.7127,
      "step": 1339
    },
    {
      "epoch": 0.294732211591334,
      "grad_norm": 0.2836678624153137,
      "learning_rate": 8e-05,
      "loss": 1.7765,
      "step": 1340
    },
    {
      "epoch": 0.29495216100296934,
      "grad_norm": 0.26315444707870483,
      "learning_rate": 8e-05,
      "loss": 1.6592,
      "step": 1341
    },
    {
      "epoch": 0.2951721104146046,
      "grad_norm": 0.2601313591003418,
      "learning_rate": 8e-05,
      "loss": 1.5803,
      "step": 1342
    },
    {
      "epoch": 0.29539205982623995,
      "grad_norm": 0.28084784746170044,
      "learning_rate": 8e-05,
      "loss": 1.6172,
      "step": 1343
    },
    {
      "epoch": 0.2956120092378753,
      "grad_norm": 0.27707698941230774,
      "learning_rate": 8e-05,
      "loss": 1.6774,
      "step": 1344
    },
    {
      "epoch": 0.2958319586495106,
      "grad_norm": 0.28750407695770264,
      "learning_rate": 8e-05,
      "loss": 1.7775,
      "step": 1345
    },
    {
      "epoch": 0.29605190806114595,
      "grad_norm": 0.27315664291381836,
      "learning_rate": 8e-05,
      "loss": 1.6578,
      "step": 1346
    },
    {
      "epoch": 0.2962718574727813,
      "grad_norm": 0.26131486892700195,
      "learning_rate": 8e-05,
      "loss": 1.6429,
      "step": 1347
    },
    {
      "epoch": 0.29649180688441656,
      "grad_norm": 0.27198976278305054,
      "learning_rate": 8e-05,
      "loss": 1.6594,
      "step": 1348
    },
    {
      "epoch": 0.2967117562960519,
      "grad_norm": 0.2785218060016632,
      "learning_rate": 8e-05,
      "loss": 1.6959,
      "step": 1349
    },
    {
      "epoch": 0.29693170570768723,
      "grad_norm": 0.26987215876579285,
      "learning_rate": 8e-05,
      "loss": 1.6561,
      "step": 1350
    },
    {
      "epoch": 0.29715165511932257,
      "grad_norm": 0.2634013295173645,
      "learning_rate": 8e-05,
      "loss": 1.6817,
      "step": 1351
    },
    {
      "epoch": 0.2973716045309579,
      "grad_norm": 0.2584557831287384,
      "learning_rate": 8e-05,
      "loss": 1.5104,
      "step": 1352
    },
    {
      "epoch": 0.2975915539425932,
      "grad_norm": 0.28787991404533386,
      "learning_rate": 8e-05,
      "loss": 1.8217,
      "step": 1353
    },
    {
      "epoch": 0.2978115033542285,
      "grad_norm": 0.5047094225883484,
      "learning_rate": 8e-05,
      "loss": 1.7733,
      "step": 1354
    },
    {
      "epoch": 0.29803145276586385,
      "grad_norm": 0.26776471734046936,
      "learning_rate": 8e-05,
      "loss": 1.6961,
      "step": 1355
    },
    {
      "epoch": 0.2982514021774992,
      "grad_norm": 0.30351778864860535,
      "learning_rate": 8e-05,
      "loss": 1.7104,
      "step": 1356
    },
    {
      "epoch": 0.2984713515891345,
      "grad_norm": 0.27889010310173035,
      "learning_rate": 8e-05,
      "loss": 1.7276,
      "step": 1357
    },
    {
      "epoch": 0.29869130100076985,
      "grad_norm": 0.2656184136867523,
      "learning_rate": 8e-05,
      "loss": 1.7438,
      "step": 1358
    },
    {
      "epoch": 0.2989112504124051,
      "grad_norm": 0.27338340878486633,
      "learning_rate": 8e-05,
      "loss": 1.7526,
      "step": 1359
    },
    {
      "epoch": 0.29913119982404046,
      "grad_norm": 0.3266398310661316,
      "learning_rate": 8e-05,
      "loss": 1.8091,
      "step": 1360
    },
    {
      "epoch": 0.2993511492356758,
      "grad_norm": 0.309469997882843,
      "learning_rate": 8e-05,
      "loss": 2.0485,
      "step": 1361
    },
    {
      "epoch": 0.29957109864731113,
      "grad_norm": 0.2768929600715637,
      "learning_rate": 8e-05,
      "loss": 1.7977,
      "step": 1362
    },
    {
      "epoch": 0.29979104805894646,
      "grad_norm": 0.27685433626174927,
      "learning_rate": 8e-05,
      "loss": 1.5712,
      "step": 1363
    },
    {
      "epoch": 0.30001099747058174,
      "grad_norm": 0.26404622197151184,
      "learning_rate": 8e-05,
      "loss": 1.6639,
      "step": 1364
    },
    {
      "epoch": 0.3002309468822171,
      "grad_norm": 0.2719237208366394,
      "learning_rate": 8e-05,
      "loss": 1.788,
      "step": 1365
    },
    {
      "epoch": 0.3004508962938524,
      "grad_norm": 0.27983394265174866,
      "learning_rate": 8e-05,
      "loss": 1.7361,
      "step": 1366
    },
    {
      "epoch": 0.30067084570548774,
      "grad_norm": 0.2673875689506531,
      "learning_rate": 8e-05,
      "loss": 1.6288,
      "step": 1367
    },
    {
      "epoch": 0.3008907951171231,
      "grad_norm": 0.2850426435470581,
      "learning_rate": 8e-05,
      "loss": 1.8328,
      "step": 1368
    },
    {
      "epoch": 0.3011107445287584,
      "grad_norm": 0.2577967345714569,
      "learning_rate": 8e-05,
      "loss": 1.6267,
      "step": 1369
    },
    {
      "epoch": 0.3013306939403937,
      "grad_norm": 0.276094913482666,
      "learning_rate": 8e-05,
      "loss": 1.7673,
      "step": 1370
    },
    {
      "epoch": 0.301550643352029,
      "grad_norm": 0.2834344208240509,
      "learning_rate": 8e-05,
      "loss": 1.6692,
      "step": 1371
    },
    {
      "epoch": 0.30177059276366436,
      "grad_norm": 0.2617560029029846,
      "learning_rate": 8e-05,
      "loss": 1.7734,
      "step": 1372
    },
    {
      "epoch": 0.3019905421752997,
      "grad_norm": 0.27122870087623596,
      "learning_rate": 8e-05,
      "loss": 1.6988,
      "step": 1373
    },
    {
      "epoch": 0.302210491586935,
      "grad_norm": 0.26526594161987305,
      "learning_rate": 8e-05,
      "loss": 1.7459,
      "step": 1374
    },
    {
      "epoch": 0.3024304409985703,
      "grad_norm": 0.2893051207065582,
      "learning_rate": 8e-05,
      "loss": 1.8214,
      "step": 1375
    },
    {
      "epoch": 0.30265039041020564,
      "grad_norm": 0.2735356092453003,
      "learning_rate": 8e-05,
      "loss": 1.8437,
      "step": 1376
    },
    {
      "epoch": 0.302870339821841,
      "grad_norm": 0.2743459939956665,
      "learning_rate": 8e-05,
      "loss": 1.8365,
      "step": 1377
    },
    {
      "epoch": 0.3030902892334763,
      "grad_norm": 0.28047019243240356,
      "learning_rate": 8e-05,
      "loss": 1.6143,
      "step": 1378
    },
    {
      "epoch": 0.30331023864511164,
      "grad_norm": 0.268197238445282,
      "learning_rate": 8e-05,
      "loss": 1.591,
      "step": 1379
    },
    {
      "epoch": 0.303530188056747,
      "grad_norm": 0.2890843451023102,
      "learning_rate": 8e-05,
      "loss": 1.7757,
      "step": 1380
    },
    {
      "epoch": 0.30375013746838225,
      "grad_norm": 0.2765072286128998,
      "learning_rate": 8e-05,
      "loss": 1.6363,
      "step": 1381
    },
    {
      "epoch": 0.3039700868800176,
      "grad_norm": 0.290147602558136,
      "learning_rate": 8e-05,
      "loss": 1.7615,
      "step": 1382
    },
    {
      "epoch": 0.3041900362916529,
      "grad_norm": 0.2721220850944519,
      "learning_rate": 8e-05,
      "loss": 1.7101,
      "step": 1383
    },
    {
      "epoch": 0.30440998570328826,
      "grad_norm": 0.27125662565231323,
      "learning_rate": 8e-05,
      "loss": 1.7291,
      "step": 1384
    },
    {
      "epoch": 0.3046299351149236,
      "grad_norm": 0.2594304084777832,
      "learning_rate": 8e-05,
      "loss": 1.6754,
      "step": 1385
    },
    {
      "epoch": 0.30484988452655887,
      "grad_norm": 0.28582707047462463,
      "learning_rate": 8e-05,
      "loss": 1.6808,
      "step": 1386
    },
    {
      "epoch": 0.3050698339381942,
      "grad_norm": 0.2853895425796509,
      "learning_rate": 8e-05,
      "loss": 1.779,
      "step": 1387
    },
    {
      "epoch": 0.30528978334982954,
      "grad_norm": 0.2580530345439911,
      "learning_rate": 8e-05,
      "loss": 1.6316,
      "step": 1388
    },
    {
      "epoch": 0.30550973276146487,
      "grad_norm": 0.2793220281600952,
      "learning_rate": 8e-05,
      "loss": 1.7326,
      "step": 1389
    },
    {
      "epoch": 0.3057296821731002,
      "grad_norm": 0.2672085165977478,
      "learning_rate": 8e-05,
      "loss": 1.6544,
      "step": 1390
    },
    {
      "epoch": 0.30594963158473554,
      "grad_norm": 0.27718111872673035,
      "learning_rate": 8e-05,
      "loss": 1.6307,
      "step": 1391
    },
    {
      "epoch": 0.3061695809963708,
      "grad_norm": 0.29295554757118225,
      "learning_rate": 8e-05,
      "loss": 1.502,
      "step": 1392
    },
    {
      "epoch": 0.30638953040800615,
      "grad_norm": 0.2840512990951538,
      "learning_rate": 8e-05,
      "loss": 1.6326,
      "step": 1393
    },
    {
      "epoch": 0.3066094798196415,
      "grad_norm": 0.2897029519081116,
      "learning_rate": 8e-05,
      "loss": 1.7543,
      "step": 1394
    },
    {
      "epoch": 0.3068294292312768,
      "grad_norm": 0.28060710430145264,
      "learning_rate": 8e-05,
      "loss": 1.7227,
      "step": 1395
    },
    {
      "epoch": 0.30704937864291215,
      "grad_norm": 0.27874305844306946,
      "learning_rate": 8e-05,
      "loss": 1.6639,
      "step": 1396
    },
    {
      "epoch": 0.30726932805454743,
      "grad_norm": 0.2679193615913391,
      "learning_rate": 8e-05,
      "loss": 1.7226,
      "step": 1397
    },
    {
      "epoch": 0.30748927746618276,
      "grad_norm": 0.2769779562950134,
      "learning_rate": 8e-05,
      "loss": 1.6384,
      "step": 1398
    },
    {
      "epoch": 0.3077092268778181,
      "grad_norm": 0.26620879769325256,
      "learning_rate": 8e-05,
      "loss": 1.7134,
      "step": 1399
    },
    {
      "epoch": 0.30792917628945343,
      "grad_norm": 0.277423620223999,
      "learning_rate": 8e-05,
      "loss": 1.7376,
      "step": 1400
    },
    {
      "epoch": 0.30814912570108877,
      "grad_norm": 0.2629416882991791,
      "learning_rate": 8e-05,
      "loss": 1.598,
      "step": 1401
    },
    {
      "epoch": 0.3083690751127241,
      "grad_norm": 0.2844812572002411,
      "learning_rate": 8e-05,
      "loss": 1.7067,
      "step": 1402
    },
    {
      "epoch": 0.3085890245243594,
      "grad_norm": 0.2731526494026184,
      "learning_rate": 8e-05,
      "loss": 1.8571,
      "step": 1403
    },
    {
      "epoch": 0.3088089739359947,
      "grad_norm": 0.287438303232193,
      "learning_rate": 8e-05,
      "loss": 1.7612,
      "step": 1404
    },
    {
      "epoch": 0.30902892334763005,
      "grad_norm": 0.266718327999115,
      "learning_rate": 8e-05,
      "loss": 1.6106,
      "step": 1405
    },
    {
      "epoch": 0.3092488727592654,
      "grad_norm": 0.28080686926841736,
      "learning_rate": 8e-05,
      "loss": 1.8281,
      "step": 1406
    },
    {
      "epoch": 0.3094688221709007,
      "grad_norm": 0.27558308839797974,
      "learning_rate": 8e-05,
      "loss": 1.8677,
      "step": 1407
    },
    {
      "epoch": 0.309688771582536,
      "grad_norm": 0.2798183262348175,
      "learning_rate": 8e-05,
      "loss": 1.7867,
      "step": 1408
    },
    {
      "epoch": 0.3099087209941713,
      "grad_norm": 0.25823187828063965,
      "learning_rate": 8e-05,
      "loss": 1.6743,
      "step": 1409
    },
    {
      "epoch": 0.31012867040580666,
      "grad_norm": 0.27356335520744324,
      "learning_rate": 8e-05,
      "loss": 1.7039,
      "step": 1410
    },
    {
      "epoch": 0.310348619817442,
      "grad_norm": 0.2842661440372467,
      "learning_rate": 8e-05,
      "loss": 1.7046,
      "step": 1411
    },
    {
      "epoch": 0.31056856922907733,
      "grad_norm": 0.2561197876930237,
      "learning_rate": 8e-05,
      "loss": 1.4887,
      "step": 1412
    },
    {
      "epoch": 0.31078851864071266,
      "grad_norm": 0.2851184904575348,
      "learning_rate": 8e-05,
      "loss": 1.7074,
      "step": 1413
    },
    {
      "epoch": 0.31100846805234794,
      "grad_norm": 0.2655506432056427,
      "learning_rate": 8e-05,
      "loss": 1.6049,
      "step": 1414
    },
    {
      "epoch": 0.3112284174639833,
      "grad_norm": 0.26412099599838257,
      "learning_rate": 8e-05,
      "loss": 1.6052,
      "step": 1415
    },
    {
      "epoch": 0.3114483668756186,
      "grad_norm": 0.3026227056980133,
      "learning_rate": 8e-05,
      "loss": 1.7085,
      "step": 1416
    },
    {
      "epoch": 0.31166831628725394,
      "grad_norm": 0.28821703791618347,
      "learning_rate": 8e-05,
      "loss": 1.7573,
      "step": 1417
    },
    {
      "epoch": 0.3118882656988893,
      "grad_norm": 0.26806455850601196,
      "learning_rate": 8e-05,
      "loss": 1.7136,
      "step": 1418
    },
    {
      "epoch": 0.31210821511052456,
      "grad_norm": 0.28336799144744873,
      "learning_rate": 8e-05,
      "loss": 1.8445,
      "step": 1419
    },
    {
      "epoch": 0.3123281645221599,
      "grad_norm": 0.2772139012813568,
      "learning_rate": 8e-05,
      "loss": 1.692,
      "step": 1420
    },
    {
      "epoch": 0.3125481139337952,
      "grad_norm": 0.2815256714820862,
      "learning_rate": 8e-05,
      "loss": 1.77,
      "step": 1421
    },
    {
      "epoch": 0.31276806334543056,
      "grad_norm": 0.4029920697212219,
      "learning_rate": 8e-05,
      "loss": 1.8103,
      "step": 1422
    },
    {
      "epoch": 0.3129880127570659,
      "grad_norm": 0.2677610218524933,
      "learning_rate": 8e-05,
      "loss": 1.5898,
      "step": 1423
    },
    {
      "epoch": 0.3132079621687012,
      "grad_norm": 0.2605397701263428,
      "learning_rate": 8e-05,
      "loss": 1.5735,
      "step": 1424
    },
    {
      "epoch": 0.3134279115803365,
      "grad_norm": 0.2831586003303528,
      "learning_rate": 8e-05,
      "loss": 1.6641,
      "step": 1425
    },
    {
      "epoch": 0.31364786099197184,
      "grad_norm": 0.2746485471725464,
      "learning_rate": 8e-05,
      "loss": 1.618,
      "step": 1426
    },
    {
      "epoch": 0.3138678104036072,
      "grad_norm": 0.283342182636261,
      "learning_rate": 8e-05,
      "loss": 1.6963,
      "step": 1427
    },
    {
      "epoch": 0.3140877598152425,
      "grad_norm": 0.27635300159454346,
      "learning_rate": 8e-05,
      "loss": 1.6911,
      "step": 1428
    },
    {
      "epoch": 0.31430770922687784,
      "grad_norm": 0.2719132900238037,
      "learning_rate": 8e-05,
      "loss": 1.7063,
      "step": 1429
    },
    {
      "epoch": 0.3145276586385131,
      "grad_norm": 0.27162256836891174,
      "learning_rate": 8e-05,
      "loss": 1.6397,
      "step": 1430
    },
    {
      "epoch": 0.31474760805014845,
      "grad_norm": 0.2934938073158264,
      "learning_rate": 8e-05,
      "loss": 1.7555,
      "step": 1431
    },
    {
      "epoch": 0.3149675574617838,
      "grad_norm": 0.3060123920440674,
      "learning_rate": 8e-05,
      "loss": 1.642,
      "step": 1432
    },
    {
      "epoch": 0.3151875068734191,
      "grad_norm": 0.280846506357193,
      "learning_rate": 8e-05,
      "loss": 1.6805,
      "step": 1433
    },
    {
      "epoch": 0.31540745628505445,
      "grad_norm": 0.2768997550010681,
      "learning_rate": 8e-05,
      "loss": 1.7359,
      "step": 1434
    },
    {
      "epoch": 0.3156274056966898,
      "grad_norm": 0.29172810912132263,
      "learning_rate": 8e-05,
      "loss": 1.821,
      "step": 1435
    },
    {
      "epoch": 0.31584735510832507,
      "grad_norm": 0.30742648243904114,
      "learning_rate": 8e-05,
      "loss": 1.8198,
      "step": 1436
    },
    {
      "epoch": 0.3160673045199604,
      "grad_norm": 0.2889997065067291,
      "learning_rate": 8e-05,
      "loss": 1.6733,
      "step": 1437
    },
    {
      "epoch": 0.31628725393159574,
      "grad_norm": 0.2859675884246826,
      "learning_rate": 8e-05,
      "loss": 1.7655,
      "step": 1438
    },
    {
      "epoch": 0.31650720334323107,
      "grad_norm": 0.2926831543445587,
      "learning_rate": 8e-05,
      "loss": 1.7871,
      "step": 1439
    },
    {
      "epoch": 0.3167271527548664,
      "grad_norm": 0.28924524784088135,
      "learning_rate": 8e-05,
      "loss": 1.665,
      "step": 1440
    },
    {
      "epoch": 0.3169471021665017,
      "grad_norm": 0.2940097749233246,
      "learning_rate": 8e-05,
      "loss": 1.8364,
      "step": 1441
    },
    {
      "epoch": 0.317167051578137,
      "grad_norm": 0.2923974096775055,
      "learning_rate": 8e-05,
      "loss": 1.8071,
      "step": 1442
    },
    {
      "epoch": 0.31738700098977235,
      "grad_norm": 0.28991878032684326,
      "learning_rate": 8e-05,
      "loss": 1.7445,
      "step": 1443
    },
    {
      "epoch": 0.3176069504014077,
      "grad_norm": 0.283600777387619,
      "learning_rate": 8e-05,
      "loss": 1.8043,
      "step": 1444
    },
    {
      "epoch": 0.317826899813043,
      "grad_norm": 0.3082323372364044,
      "learning_rate": 8e-05,
      "loss": 1.7858,
      "step": 1445
    },
    {
      "epoch": 0.3180468492246783,
      "grad_norm": 0.28433462977409363,
      "learning_rate": 8e-05,
      "loss": 1.6911,
      "step": 1446
    },
    {
      "epoch": 0.31826679863631363,
      "grad_norm": 0.27776578068733215,
      "learning_rate": 8e-05,
      "loss": 1.7212,
      "step": 1447
    },
    {
      "epoch": 0.31848674804794896,
      "grad_norm": 0.29395151138305664,
      "learning_rate": 8e-05,
      "loss": 1.7221,
      "step": 1448
    },
    {
      "epoch": 0.3187066974595843,
      "grad_norm": 0.27507245540618896,
      "learning_rate": 8e-05,
      "loss": 1.7358,
      "step": 1449
    },
    {
      "epoch": 0.31892664687121963,
      "grad_norm": 0.25614190101623535,
      "learning_rate": 8e-05,
      "loss": 1.5138,
      "step": 1450
    },
    {
      "epoch": 0.31914659628285497,
      "grad_norm": 0.2908024489879608,
      "learning_rate": 8e-05,
      "loss": 1.756,
      "step": 1451
    },
    {
      "epoch": 0.31936654569449024,
      "grad_norm": 0.2729463577270508,
      "learning_rate": 8e-05,
      "loss": 1.5542,
      "step": 1452
    },
    {
      "epoch": 0.3195864951061256,
      "grad_norm": 0.27094194293022156,
      "learning_rate": 8e-05,
      "loss": 1.5917,
      "step": 1453
    },
    {
      "epoch": 0.3198064445177609,
      "grad_norm": 0.28125494718551636,
      "learning_rate": 8e-05,
      "loss": 1.6584,
      "step": 1454
    },
    {
      "epoch": 0.32002639392939625,
      "grad_norm": 0.29033198952674866,
      "learning_rate": 8e-05,
      "loss": 1.7332,
      "step": 1455
    },
    {
      "epoch": 0.3202463433410316,
      "grad_norm": 0.26570284366607666,
      "learning_rate": 8e-05,
      "loss": 1.6159,
      "step": 1456
    },
    {
      "epoch": 0.32046629275266686,
      "grad_norm": 0.307412713766098,
      "learning_rate": 8e-05,
      "loss": 1.7351,
      "step": 1457
    },
    {
      "epoch": 0.3206862421643022,
      "grad_norm": 0.29387474060058594,
      "learning_rate": 8e-05,
      "loss": 1.9386,
      "step": 1458
    },
    {
      "epoch": 0.3209061915759375,
      "grad_norm": 0.26545315980911255,
      "learning_rate": 8e-05,
      "loss": 1.6343,
      "step": 1459
    },
    {
      "epoch": 0.32112614098757286,
      "grad_norm": 0.279238224029541,
      "learning_rate": 8e-05,
      "loss": 1.6245,
      "step": 1460
    },
    {
      "epoch": 0.3213460903992082,
      "grad_norm": 0.2766862213611603,
      "learning_rate": 8e-05,
      "loss": 1.7135,
      "step": 1461
    },
    {
      "epoch": 0.32156603981084353,
      "grad_norm": 0.2705351412296295,
      "learning_rate": 8e-05,
      "loss": 1.6526,
      "step": 1462
    },
    {
      "epoch": 0.3217859892224788,
      "grad_norm": 0.27870967984199524,
      "learning_rate": 8e-05,
      "loss": 1.6512,
      "step": 1463
    },
    {
      "epoch": 0.32200593863411414,
      "grad_norm": 0.284407377243042,
      "learning_rate": 8e-05,
      "loss": 1.755,
      "step": 1464
    },
    {
      "epoch": 0.3222258880457495,
      "grad_norm": 0.2897641062736511,
      "learning_rate": 8e-05,
      "loss": 1.8383,
      "step": 1465
    },
    {
      "epoch": 0.3224458374573848,
      "grad_norm": 0.2667568624019623,
      "learning_rate": 8e-05,
      "loss": 1.6989,
      "step": 1466
    },
    {
      "epoch": 0.32266578686902014,
      "grad_norm": 0.26580294966697693,
      "learning_rate": 8e-05,
      "loss": 1.5895,
      "step": 1467
    },
    {
      "epoch": 0.3228857362806554,
      "grad_norm": 0.26188549399375916,
      "learning_rate": 8e-05,
      "loss": 1.5799,
      "step": 1468
    },
    {
      "epoch": 0.32310568569229076,
      "grad_norm": 0.27703747153282166,
      "learning_rate": 8e-05,
      "loss": 1.8306,
      "step": 1469
    },
    {
      "epoch": 0.3233256351039261,
      "grad_norm": 0.27643802762031555,
      "learning_rate": 8e-05,
      "loss": 1.6864,
      "step": 1470
    },
    {
      "epoch": 0.3235455845155614,
      "grad_norm": 0.27216553688049316,
      "learning_rate": 8e-05,
      "loss": 1.6006,
      "step": 1471
    },
    {
      "epoch": 0.32376553392719676,
      "grad_norm": 0.2984940707683563,
      "learning_rate": 8e-05,
      "loss": 1.7548,
      "step": 1472
    },
    {
      "epoch": 0.3239854833388321,
      "grad_norm": 0.30579298734664917,
      "learning_rate": 8e-05,
      "loss": 1.8307,
      "step": 1473
    },
    {
      "epoch": 0.32420543275046737,
      "grad_norm": 0.27524709701538086,
      "learning_rate": 8e-05,
      "loss": 1.6134,
      "step": 1474
    },
    {
      "epoch": 0.3244253821621027,
      "grad_norm": 0.2788650393486023,
      "learning_rate": 8e-05,
      "loss": 1.8194,
      "step": 1475
    },
    {
      "epoch": 0.32464533157373804,
      "grad_norm": 0.28263744711875916,
      "learning_rate": 8e-05,
      "loss": 1.7633,
      "step": 1476
    },
    {
      "epoch": 0.3248652809853734,
      "grad_norm": 0.30234408378601074,
      "learning_rate": 8e-05,
      "loss": 1.6057,
      "step": 1477
    },
    {
      "epoch": 0.3250852303970087,
      "grad_norm": 0.2820134162902832,
      "learning_rate": 8e-05,
      "loss": 1.6913,
      "step": 1478
    },
    {
      "epoch": 0.325305179808644,
      "grad_norm": 0.28929245471954346,
      "learning_rate": 8e-05,
      "loss": 1.9538,
      "step": 1479
    },
    {
      "epoch": 0.3255251292202793,
      "grad_norm": 0.26399463415145874,
      "learning_rate": 8e-05,
      "loss": 1.7309,
      "step": 1480
    },
    {
      "epoch": 0.32574507863191465,
      "grad_norm": 0.2722630202770233,
      "learning_rate": 8e-05,
      "loss": 1.5595,
      "step": 1481
    },
    {
      "epoch": 0.32596502804355,
      "grad_norm": 0.2759261727333069,
      "learning_rate": 8e-05,
      "loss": 1.6272,
      "step": 1482
    },
    {
      "epoch": 0.3261849774551853,
      "grad_norm": 0.28047022223472595,
      "learning_rate": 8e-05,
      "loss": 1.6933,
      "step": 1483
    },
    {
      "epoch": 0.32640492686682065,
      "grad_norm": 0.2835995554924011,
      "learning_rate": 8e-05,
      "loss": 1.7165,
      "step": 1484
    },
    {
      "epoch": 0.32662487627845593,
      "grad_norm": 0.28965097665786743,
      "learning_rate": 8e-05,
      "loss": 1.6399,
      "step": 1485
    },
    {
      "epoch": 0.32684482569009127,
      "grad_norm": 0.2729817032814026,
      "learning_rate": 8e-05,
      "loss": 1.7397,
      "step": 1486
    },
    {
      "epoch": 0.3270647751017266,
      "grad_norm": 0.26809874176979065,
      "learning_rate": 8e-05,
      "loss": 1.6954,
      "step": 1487
    },
    {
      "epoch": 0.32728472451336194,
      "grad_norm": 0.29766684770584106,
      "learning_rate": 8e-05,
      "loss": 1.6947,
      "step": 1488
    },
    {
      "epoch": 0.32750467392499727,
      "grad_norm": 0.27032670378685,
      "learning_rate": 8e-05,
      "loss": 1.8036,
      "step": 1489
    },
    {
      "epoch": 0.32772462333663255,
      "grad_norm": 0.2694716453552246,
      "learning_rate": 8e-05,
      "loss": 1.6856,
      "step": 1490
    },
    {
      "epoch": 0.3279445727482679,
      "grad_norm": 0.27968841791152954,
      "learning_rate": 8e-05,
      "loss": 1.7466,
      "step": 1491
    },
    {
      "epoch": 0.3281645221599032,
      "grad_norm": 0.2956348955631256,
      "learning_rate": 8e-05,
      "loss": 1.833,
      "step": 1492
    },
    {
      "epoch": 0.32838447157153855,
      "grad_norm": 0.27069491147994995,
      "learning_rate": 8e-05,
      "loss": 1.715,
      "step": 1493
    },
    {
      "epoch": 0.3286044209831739,
      "grad_norm": 0.26747795939445496,
      "learning_rate": 8e-05,
      "loss": 1.6663,
      "step": 1494
    },
    {
      "epoch": 0.3288243703948092,
      "grad_norm": 0.2619915008544922,
      "learning_rate": 8e-05,
      "loss": 1.6503,
      "step": 1495
    },
    {
      "epoch": 0.3290443198064445,
      "grad_norm": 0.2720276117324829,
      "learning_rate": 8e-05,
      "loss": 1.7174,
      "step": 1496
    },
    {
      "epoch": 0.32926426921807983,
      "grad_norm": 0.26874253153800964,
      "learning_rate": 8e-05,
      "loss": 1.6503,
      "step": 1497
    },
    {
      "epoch": 0.32948421862971516,
      "grad_norm": 0.28397336602211,
      "learning_rate": 8e-05,
      "loss": 1.67,
      "step": 1498
    },
    {
      "epoch": 0.3297041680413505,
      "grad_norm": 0.2544403076171875,
      "learning_rate": 8e-05,
      "loss": 1.5153,
      "step": 1499
    },
    {
      "epoch": 0.32992411745298583,
      "grad_norm": 0.2819180488586426,
      "learning_rate": 8e-05,
      "loss": 1.6704,
      "step": 1500
    },
    {
      "epoch": 0.3301440668646211,
      "grad_norm": 0.28150951862335205,
      "learning_rate": 8e-05,
      "loss": 1.8451,
      "step": 1501
    },
    {
      "epoch": 0.33036401627625644,
      "grad_norm": 0.27396339178085327,
      "learning_rate": 8e-05,
      "loss": 1.7631,
      "step": 1502
    },
    {
      "epoch": 0.3305839656878918,
      "grad_norm": 0.2954351007938385,
      "learning_rate": 8e-05,
      "loss": 1.8101,
      "step": 1503
    },
    {
      "epoch": 0.3308039150995271,
      "grad_norm": 0.27129319310188293,
      "learning_rate": 8e-05,
      "loss": 1.6484,
      "step": 1504
    },
    {
      "epoch": 0.33102386451116245,
      "grad_norm": 0.27612754702568054,
      "learning_rate": 8e-05,
      "loss": 1.6178,
      "step": 1505
    },
    {
      "epoch": 0.3312438139227978,
      "grad_norm": 0.26097655296325684,
      "learning_rate": 8e-05,
      "loss": 1.5781,
      "step": 1506
    },
    {
      "epoch": 0.33146376333443306,
      "grad_norm": 0.2704753577709198,
      "learning_rate": 8e-05,
      "loss": 1.6919,
      "step": 1507
    },
    {
      "epoch": 0.3316837127460684,
      "grad_norm": 0.26866593956947327,
      "learning_rate": 8e-05,
      "loss": 1.6795,
      "step": 1508
    },
    {
      "epoch": 0.3319036621577037,
      "grad_norm": 0.31797948479652405,
      "learning_rate": 8e-05,
      "loss": 1.7511,
      "step": 1509
    },
    {
      "epoch": 0.33212361156933906,
      "grad_norm": 0.29456841945648193,
      "learning_rate": 8e-05,
      "loss": 1.7041,
      "step": 1510
    },
    {
      "epoch": 0.3323435609809744,
      "grad_norm": 0.28345033526420593,
      "learning_rate": 8e-05,
      "loss": 1.7499,
      "step": 1511
    },
    {
      "epoch": 0.3325635103926097,
      "grad_norm": 0.28679129481315613,
      "learning_rate": 8e-05,
      "loss": 1.8304,
      "step": 1512
    },
    {
      "epoch": 0.332783459804245,
      "grad_norm": 0.2799399793148041,
      "learning_rate": 8e-05,
      "loss": 1.6461,
      "step": 1513
    },
    {
      "epoch": 0.33300340921588034,
      "grad_norm": 0.3234422206878662,
      "learning_rate": 8e-05,
      "loss": 1.5622,
      "step": 1514
    },
    {
      "epoch": 0.3332233586275157,
      "grad_norm": 0.27786344289779663,
      "learning_rate": 8e-05,
      "loss": 1.6718,
      "step": 1515
    },
    {
      "epoch": 0.333443308039151,
      "grad_norm": 0.27040839195251465,
      "learning_rate": 8e-05,
      "loss": 1.7428,
      "step": 1516
    },
    {
      "epoch": 0.33366325745078634,
      "grad_norm": 0.2837252616882324,
      "learning_rate": 8e-05,
      "loss": 1.6929,
      "step": 1517
    },
    {
      "epoch": 0.3338832068624216,
      "grad_norm": 0.27352792024612427,
      "learning_rate": 8e-05,
      "loss": 1.7804,
      "step": 1518
    },
    {
      "epoch": 0.33410315627405696,
      "grad_norm": 0.27237218618392944,
      "learning_rate": 8e-05,
      "loss": 1.7652,
      "step": 1519
    },
    {
      "epoch": 0.3343231056856923,
      "grad_norm": 0.3166270852088928,
      "learning_rate": 8e-05,
      "loss": 1.6363,
      "step": 1520
    },
    {
      "epoch": 0.3345430550973276,
      "grad_norm": 0.2650817930698395,
      "learning_rate": 8e-05,
      "loss": 1.6954,
      "step": 1521
    },
    {
      "epoch": 0.33476300450896296,
      "grad_norm": 0.2907481789588928,
      "learning_rate": 8e-05,
      "loss": 1.809,
      "step": 1522
    },
    {
      "epoch": 0.33498295392059824,
      "grad_norm": 0.2754502296447754,
      "learning_rate": 8e-05,
      "loss": 1.8143,
      "step": 1523
    },
    {
      "epoch": 0.33520290333223357,
      "grad_norm": 0.2890012264251709,
      "learning_rate": 8e-05,
      "loss": 1.6603,
      "step": 1524
    },
    {
      "epoch": 0.3354228527438689,
      "grad_norm": 0.271720826625824,
      "learning_rate": 8e-05,
      "loss": 1.7186,
      "step": 1525
    },
    {
      "epoch": 0.33564280215550424,
      "grad_norm": 0.2845331132411957,
      "learning_rate": 8e-05,
      "loss": 1.7739,
      "step": 1526
    },
    {
      "epoch": 0.3358627515671396,
      "grad_norm": 0.2787776291370392,
      "learning_rate": 8e-05,
      "loss": 1.6146,
      "step": 1527
    },
    {
      "epoch": 0.3360827009787749,
      "grad_norm": 0.2612919211387634,
      "learning_rate": 8e-05,
      "loss": 1.5575,
      "step": 1528
    },
    {
      "epoch": 0.3363026503904102,
      "grad_norm": 0.279220849275589,
      "learning_rate": 8e-05,
      "loss": 1.7661,
      "step": 1529
    },
    {
      "epoch": 0.3365225998020455,
      "grad_norm": 0.2812168300151825,
      "learning_rate": 8e-05,
      "loss": 1.7011,
      "step": 1530
    },
    {
      "epoch": 0.33674254921368085,
      "grad_norm": 0.28216826915740967,
      "learning_rate": 8e-05,
      "loss": 1.7856,
      "step": 1531
    },
    {
      "epoch": 0.3369624986253162,
      "grad_norm": 0.279895156621933,
      "learning_rate": 8e-05,
      "loss": 1.6793,
      "step": 1532
    },
    {
      "epoch": 0.3371824480369515,
      "grad_norm": 0.2694056034088135,
      "learning_rate": 8e-05,
      "loss": 1.6289,
      "step": 1533
    },
    {
      "epoch": 0.3374023974485868,
      "grad_norm": 0.2692592740058899,
      "learning_rate": 8e-05,
      "loss": 1.5595,
      "step": 1534
    },
    {
      "epoch": 0.33762234686022213,
      "grad_norm": 0.32149383425712585,
      "learning_rate": 8e-05,
      "loss": 1.6667,
      "step": 1535
    },
    {
      "epoch": 0.33784229627185747,
      "grad_norm": 0.28884437680244446,
      "learning_rate": 8e-05,
      "loss": 1.7836,
      "step": 1536
    },
    {
      "epoch": 0.3380622456834928,
      "grad_norm": 0.276017963886261,
      "learning_rate": 8e-05,
      "loss": 1.712,
      "step": 1537
    },
    {
      "epoch": 0.33828219509512814,
      "grad_norm": 0.26901450753211975,
      "learning_rate": 8e-05,
      "loss": 1.6442,
      "step": 1538
    },
    {
      "epoch": 0.33850214450676347,
      "grad_norm": 0.29827412962913513,
      "learning_rate": 8e-05,
      "loss": 1.7619,
      "step": 1539
    },
    {
      "epoch": 0.33872209391839875,
      "grad_norm": 0.2763231098651886,
      "learning_rate": 8e-05,
      "loss": 1.6344,
      "step": 1540
    },
    {
      "epoch": 0.3389420433300341,
      "grad_norm": 0.26493677496910095,
      "learning_rate": 8e-05,
      "loss": 1.6964,
      "step": 1541
    },
    {
      "epoch": 0.3391619927416694,
      "grad_norm": 0.2956371605396271,
      "learning_rate": 8e-05,
      "loss": 1.7328,
      "step": 1542
    },
    {
      "epoch": 0.33938194215330475,
      "grad_norm": 0.2845339775085449,
      "learning_rate": 8e-05,
      "loss": 1.6477,
      "step": 1543
    },
    {
      "epoch": 0.3396018915649401,
      "grad_norm": 0.29501214623451233,
      "learning_rate": 8e-05,
      "loss": 1.7951,
      "step": 1544
    },
    {
      "epoch": 0.33982184097657536,
      "grad_norm": 0.2859644591808319,
      "learning_rate": 8e-05,
      "loss": 1.6607,
      "step": 1545
    },
    {
      "epoch": 0.3400417903882107,
      "grad_norm": 0.2733168303966522,
      "learning_rate": 8e-05,
      "loss": 1.6397,
      "step": 1546
    },
    {
      "epoch": 0.34026173979984603,
      "grad_norm": 0.2580598294734955,
      "learning_rate": 8e-05,
      "loss": 1.5692,
      "step": 1547
    },
    {
      "epoch": 0.34048168921148136,
      "grad_norm": 0.3042803406715393,
      "learning_rate": 8e-05,
      "loss": 1.7063,
      "step": 1548
    },
    {
      "epoch": 0.3407016386231167,
      "grad_norm": 0.2833859324455261,
      "learning_rate": 8e-05,
      "loss": 1.7531,
      "step": 1549
    },
    {
      "epoch": 0.34092158803475203,
      "grad_norm": 0.259620726108551,
      "learning_rate": 8e-05,
      "loss": 1.6179,
      "step": 1550
    },
    {
      "epoch": 0.3411415374463873,
      "grad_norm": 0.268355131149292,
      "learning_rate": 8e-05,
      "loss": 1.6009,
      "step": 1551
    },
    {
      "epoch": 0.34136148685802264,
      "grad_norm": 0.2858780324459076,
      "learning_rate": 8e-05,
      "loss": 1.7033,
      "step": 1552
    },
    {
      "epoch": 0.341581436269658,
      "grad_norm": 0.2777354121208191,
      "learning_rate": 8e-05,
      "loss": 1.7615,
      "step": 1553
    },
    {
      "epoch": 0.3418013856812933,
      "grad_norm": 0.27899524569511414,
      "learning_rate": 8e-05,
      "loss": 1.6684,
      "step": 1554
    },
    {
      "epoch": 0.34202133509292865,
      "grad_norm": 0.3156200349330902,
      "learning_rate": 8e-05,
      "loss": 1.5658,
      "step": 1555
    },
    {
      "epoch": 0.3422412845045639,
      "grad_norm": 0.27549582719802856,
      "learning_rate": 8e-05,
      "loss": 1.692,
      "step": 1556
    },
    {
      "epoch": 0.34246123391619926,
      "grad_norm": 0.27770310640335083,
      "learning_rate": 8e-05,
      "loss": 1.5891,
      "step": 1557
    },
    {
      "epoch": 0.3426811833278346,
      "grad_norm": 0.28138646483421326,
      "learning_rate": 8e-05,
      "loss": 1.5949,
      "step": 1558
    },
    {
      "epoch": 0.3429011327394699,
      "grad_norm": 0.2790684998035431,
      "learning_rate": 8e-05,
      "loss": 1.6371,
      "step": 1559
    },
    {
      "epoch": 0.34312108215110526,
      "grad_norm": 0.303230345249176,
      "learning_rate": 8e-05,
      "loss": 1.7416,
      "step": 1560
    },
    {
      "epoch": 0.3433410315627406,
      "grad_norm": 0.26891767978668213,
      "learning_rate": 8e-05,
      "loss": 1.8044,
      "step": 1561
    },
    {
      "epoch": 0.3435609809743759,
      "grad_norm": 0.2734631896018982,
      "learning_rate": 8e-05,
      "loss": 1.7171,
      "step": 1562
    },
    {
      "epoch": 0.3437809303860112,
      "grad_norm": 0.29556018114089966,
      "learning_rate": 8e-05,
      "loss": 1.9085,
      "step": 1563
    },
    {
      "epoch": 0.34400087979764654,
      "grad_norm": 0.26478004455566406,
      "learning_rate": 8e-05,
      "loss": 1.6153,
      "step": 1564
    },
    {
      "epoch": 0.3442208292092819,
      "grad_norm": 0.27655404806137085,
      "learning_rate": 8e-05,
      "loss": 1.7384,
      "step": 1565
    },
    {
      "epoch": 0.3444407786209172,
      "grad_norm": 0.2902698218822479,
      "learning_rate": 8e-05,
      "loss": 1.6589,
      "step": 1566
    },
    {
      "epoch": 0.3446607280325525,
      "grad_norm": 0.2857147455215454,
      "learning_rate": 8e-05,
      "loss": 1.6598,
      "step": 1567
    },
    {
      "epoch": 0.3448806774441878,
      "grad_norm": 0.28339943289756775,
      "learning_rate": 8e-05,
      "loss": 1.7356,
      "step": 1568
    },
    {
      "epoch": 0.34510062685582316,
      "grad_norm": 0.29340776801109314,
      "learning_rate": 8e-05,
      "loss": 1.8316,
      "step": 1569
    },
    {
      "epoch": 0.3453205762674585,
      "grad_norm": 0.26669397950172424,
      "learning_rate": 8e-05,
      "loss": 1.6803,
      "step": 1570
    },
    {
      "epoch": 0.3455405256790938,
      "grad_norm": 0.28508248925209045,
      "learning_rate": 8e-05,
      "loss": 1.7702,
      "step": 1571
    },
    {
      "epoch": 0.34576047509072916,
      "grad_norm": 0.25610047578811646,
      "learning_rate": 8e-05,
      "loss": 1.6343,
      "step": 1572
    },
    {
      "epoch": 0.34598042450236444,
      "grad_norm": 0.2758273482322693,
      "learning_rate": 8e-05,
      "loss": 1.7875,
      "step": 1573
    },
    {
      "epoch": 0.34620037391399977,
      "grad_norm": 0.2674688398838043,
      "learning_rate": 8e-05,
      "loss": 1.6804,
      "step": 1574
    },
    {
      "epoch": 0.3464203233256351,
      "grad_norm": 0.2796163558959961,
      "learning_rate": 8e-05,
      "loss": 1.6135,
      "step": 1575
    },
    {
      "epoch": 0.34664027273727044,
      "grad_norm": 0.26260775327682495,
      "learning_rate": 8e-05,
      "loss": 1.6752,
      "step": 1576
    },
    {
      "epoch": 0.3468602221489058,
      "grad_norm": 0.2897137403488159,
      "learning_rate": 8e-05,
      "loss": 1.6743,
      "step": 1577
    },
    {
      "epoch": 0.34708017156054105,
      "grad_norm": 0.27681732177734375,
      "learning_rate": 8e-05,
      "loss": 1.6436,
      "step": 1578
    },
    {
      "epoch": 0.3473001209721764,
      "grad_norm": 0.2694265842437744,
      "learning_rate": 8e-05,
      "loss": 1.6343,
      "step": 1579
    },
    {
      "epoch": 0.3475200703838117,
      "grad_norm": 0.28179508447647095,
      "learning_rate": 8e-05,
      "loss": 1.676,
      "step": 1580
    },
    {
      "epoch": 0.34774001979544705,
      "grad_norm": 0.29600057005882263,
      "learning_rate": 8e-05,
      "loss": 1.786,
      "step": 1581
    },
    {
      "epoch": 0.3479599692070824,
      "grad_norm": 0.28932616114616394,
      "learning_rate": 8e-05,
      "loss": 1.7151,
      "step": 1582
    },
    {
      "epoch": 0.3481799186187177,
      "grad_norm": 0.2912417948246002,
      "learning_rate": 8e-05,
      "loss": 1.7788,
      "step": 1583
    },
    {
      "epoch": 0.348399868030353,
      "grad_norm": 0.2844431698322296,
      "learning_rate": 8e-05,
      "loss": 1.5585,
      "step": 1584
    },
    {
      "epoch": 0.34861981744198833,
      "grad_norm": 0.2916630804538727,
      "learning_rate": 8e-05,
      "loss": 1.7484,
      "step": 1585
    },
    {
      "epoch": 0.34883976685362367,
      "grad_norm": 0.2785089612007141,
      "learning_rate": 8e-05,
      "loss": 1.61,
      "step": 1586
    },
    {
      "epoch": 0.349059716265259,
      "grad_norm": 0.2777422368526459,
      "learning_rate": 8e-05,
      "loss": 1.7183,
      "step": 1587
    },
    {
      "epoch": 0.34927966567689434,
      "grad_norm": 0.28772565722465515,
      "learning_rate": 8e-05,
      "loss": 1.7161,
      "step": 1588
    },
    {
      "epoch": 0.3494996150885296,
      "grad_norm": 0.28452831506729126,
      "learning_rate": 8e-05,
      "loss": 1.8004,
      "step": 1589
    },
    {
      "epoch": 0.34971956450016495,
      "grad_norm": 0.2837449014186859,
      "learning_rate": 8e-05,
      "loss": 1.7992,
      "step": 1590
    },
    {
      "epoch": 0.3499395139118003,
      "grad_norm": 0.2874920666217804,
      "learning_rate": 8e-05,
      "loss": 1.6408,
      "step": 1591
    },
    {
      "epoch": 0.3501594633234356,
      "grad_norm": 0.26615065336227417,
      "learning_rate": 8e-05,
      "loss": 1.658,
      "step": 1592
    },
    {
      "epoch": 0.35037941273507095,
      "grad_norm": 0.27493569254875183,
      "learning_rate": 8e-05,
      "loss": 1.6843,
      "step": 1593
    },
    {
      "epoch": 0.3505993621467063,
      "grad_norm": 0.291886568069458,
      "learning_rate": 8e-05,
      "loss": 1.7683,
      "step": 1594
    },
    {
      "epoch": 0.35081931155834156,
      "grad_norm": 0.2868814468383789,
      "learning_rate": 8e-05,
      "loss": 1.7825,
      "step": 1595
    },
    {
      "epoch": 0.3510392609699769,
      "grad_norm": 0.30988067388534546,
      "learning_rate": 8e-05,
      "loss": 1.7511,
      "step": 1596
    },
    {
      "epoch": 0.35125921038161223,
      "grad_norm": 0.2746553122997284,
      "learning_rate": 8e-05,
      "loss": 1.6298,
      "step": 1597
    },
    {
      "epoch": 0.35147915979324756,
      "grad_norm": 0.3013536036014557,
      "learning_rate": 8e-05,
      "loss": 1.7883,
      "step": 1598
    },
    {
      "epoch": 0.3516991092048829,
      "grad_norm": 0.2906748056411743,
      "learning_rate": 8e-05,
      "loss": 1.5819,
      "step": 1599
    },
    {
      "epoch": 0.3519190586165182,
      "grad_norm": 0.28082364797592163,
      "learning_rate": 8e-05,
      "loss": 1.7525,
      "step": 1600
    },
    {
      "epoch": 0.3521390080281535,
      "grad_norm": 0.28713324666023254,
      "learning_rate": 8e-05,
      "loss": 1.79,
      "step": 1601
    },
    {
      "epoch": 0.35235895743978884,
      "grad_norm": 0.2819896638393402,
      "learning_rate": 8e-05,
      "loss": 1.6514,
      "step": 1602
    },
    {
      "epoch": 0.3525789068514242,
      "grad_norm": 0.27669310569763184,
      "learning_rate": 8e-05,
      "loss": 1.5888,
      "step": 1603
    },
    {
      "epoch": 0.3527988562630595,
      "grad_norm": 0.2873641848564148,
      "learning_rate": 8e-05,
      "loss": 1.8206,
      "step": 1604
    },
    {
      "epoch": 0.35301880567469485,
      "grad_norm": 0.28426647186279297,
      "learning_rate": 8e-05,
      "loss": 1.7736,
      "step": 1605
    },
    {
      "epoch": 0.3532387550863301,
      "grad_norm": 0.2733590602874756,
      "learning_rate": 8e-05,
      "loss": 1.6653,
      "step": 1606
    },
    {
      "epoch": 0.35345870449796546,
      "grad_norm": 0.26751479506492615,
      "learning_rate": 8e-05,
      "loss": 1.5841,
      "step": 1607
    },
    {
      "epoch": 0.3536786539096008,
      "grad_norm": 0.2767663598060608,
      "learning_rate": 8e-05,
      "loss": 1.6859,
      "step": 1608
    },
    {
      "epoch": 0.3538986033212361,
      "grad_norm": 0.28359255194664,
      "learning_rate": 8e-05,
      "loss": 1.8799,
      "step": 1609
    },
    {
      "epoch": 0.35411855273287146,
      "grad_norm": 0.27551594376564026,
      "learning_rate": 8e-05,
      "loss": 1.6429,
      "step": 1610
    },
    {
      "epoch": 0.35433850214450674,
      "grad_norm": 0.26260972023010254,
      "learning_rate": 8e-05,
      "loss": 1.6068,
      "step": 1611
    },
    {
      "epoch": 0.3545584515561421,
      "grad_norm": 0.2778937518596649,
      "learning_rate": 8e-05,
      "loss": 1.8057,
      "step": 1612
    },
    {
      "epoch": 0.3547784009677774,
      "grad_norm": 0.27607765793800354,
      "learning_rate": 8e-05,
      "loss": 1.7439,
      "step": 1613
    },
    {
      "epoch": 0.35499835037941274,
      "grad_norm": 0.2628287076950073,
      "learning_rate": 8e-05,
      "loss": 1.6916,
      "step": 1614
    },
    {
      "epoch": 0.3552182997910481,
      "grad_norm": 0.2767592966556549,
      "learning_rate": 8e-05,
      "loss": 1.7185,
      "step": 1615
    },
    {
      "epoch": 0.3554382492026834,
      "grad_norm": 0.2666943669319153,
      "learning_rate": 8e-05,
      "loss": 1.7351,
      "step": 1616
    },
    {
      "epoch": 0.3556581986143187,
      "grad_norm": 0.28780093789100647,
      "learning_rate": 8e-05,
      "loss": 1.757,
      "step": 1617
    },
    {
      "epoch": 0.355878148025954,
      "grad_norm": 0.30761584639549255,
      "learning_rate": 8e-05,
      "loss": 1.8096,
      "step": 1618
    },
    {
      "epoch": 0.35609809743758936,
      "grad_norm": 0.2926090359687805,
      "learning_rate": 8e-05,
      "loss": 1.8609,
      "step": 1619
    },
    {
      "epoch": 0.3563180468492247,
      "grad_norm": 0.27546852827072144,
      "learning_rate": 8e-05,
      "loss": 1.6422,
      "step": 1620
    },
    {
      "epoch": 0.35653799626086,
      "grad_norm": 0.28559309244155884,
      "learning_rate": 8e-05,
      "loss": 1.8225,
      "step": 1621
    },
    {
      "epoch": 0.3567579456724953,
      "grad_norm": 0.2804494798183441,
      "learning_rate": 8e-05,
      "loss": 1.9108,
      "step": 1622
    },
    {
      "epoch": 0.35697789508413064,
      "grad_norm": 0.2643645703792572,
      "learning_rate": 8e-05,
      "loss": 1.5462,
      "step": 1623
    },
    {
      "epoch": 0.35719784449576597,
      "grad_norm": 0.2888531982898712,
      "learning_rate": 8e-05,
      "loss": 1.701,
      "step": 1624
    },
    {
      "epoch": 0.3574177939074013,
      "grad_norm": 0.28601035475730896,
      "learning_rate": 8e-05,
      "loss": 1.628,
      "step": 1625
    },
    {
      "epoch": 0.35763774331903664,
      "grad_norm": 0.2877524197101593,
      "learning_rate": 8e-05,
      "loss": 1.8403,
      "step": 1626
    },
    {
      "epoch": 0.357857692730672,
      "grad_norm": 0.2658945918083191,
      "learning_rate": 8e-05,
      "loss": 1.4552,
      "step": 1627
    },
    {
      "epoch": 0.35807764214230725,
      "grad_norm": 0.2911885976791382,
      "learning_rate": 8e-05,
      "loss": 1.7753,
      "step": 1628
    },
    {
      "epoch": 0.3582975915539426,
      "grad_norm": 0.29072439670562744,
      "learning_rate": 8e-05,
      "loss": 1.7229,
      "step": 1629
    },
    {
      "epoch": 0.3585175409655779,
      "grad_norm": 0.29961150884628296,
      "learning_rate": 8e-05,
      "loss": 1.7694,
      "step": 1630
    },
    {
      "epoch": 0.35873749037721325,
      "grad_norm": 0.2760653793811798,
      "learning_rate": 8e-05,
      "loss": 1.719,
      "step": 1631
    },
    {
      "epoch": 0.3589574397888486,
      "grad_norm": 0.2739832103252411,
      "learning_rate": 8e-05,
      "loss": 1.7367,
      "step": 1632
    },
    {
      "epoch": 0.35917738920048387,
      "grad_norm": 0.2669771611690521,
      "learning_rate": 8e-05,
      "loss": 1.5306,
      "step": 1633
    },
    {
      "epoch": 0.3593973386121192,
      "grad_norm": 0.2744583189487457,
      "learning_rate": 8e-05,
      "loss": 1.5713,
      "step": 1634
    },
    {
      "epoch": 0.35961728802375453,
      "grad_norm": 0.2943086326122284,
      "learning_rate": 8e-05,
      "loss": 1.6569,
      "step": 1635
    },
    {
      "epoch": 0.35983723743538987,
      "grad_norm": 0.2873243987560272,
      "learning_rate": 8e-05,
      "loss": 1.6864,
      "step": 1636
    },
    {
      "epoch": 0.3600571868470252,
      "grad_norm": 0.27217867970466614,
      "learning_rate": 8e-05,
      "loss": 1.7519,
      "step": 1637
    },
    {
      "epoch": 0.36027713625866054,
      "grad_norm": 0.28656938672065735,
      "learning_rate": 8e-05,
      "loss": 1.7892,
      "step": 1638
    },
    {
      "epoch": 0.3604970856702958,
      "grad_norm": 0.2876884937286377,
      "learning_rate": 8e-05,
      "loss": 1.6709,
      "step": 1639
    },
    {
      "epoch": 0.36071703508193115,
      "grad_norm": 0.2873481512069702,
      "learning_rate": 8e-05,
      "loss": 1.8336,
      "step": 1640
    },
    {
      "epoch": 0.3609369844935665,
      "grad_norm": 0.28285419940948486,
      "learning_rate": 8e-05,
      "loss": 1.5887,
      "step": 1641
    },
    {
      "epoch": 0.3611569339052018,
      "grad_norm": 0.2624582052230835,
      "learning_rate": 8e-05,
      "loss": 1.6248,
      "step": 1642
    },
    {
      "epoch": 0.36137688331683715,
      "grad_norm": 0.2794424891471863,
      "learning_rate": 8e-05,
      "loss": 1.7191,
      "step": 1643
    },
    {
      "epoch": 0.36159683272847243,
      "grad_norm": 0.2890479862689972,
      "learning_rate": 8e-05,
      "loss": 1.905,
      "step": 1644
    },
    {
      "epoch": 0.36181678214010776,
      "grad_norm": 0.28444570302963257,
      "learning_rate": 8e-05,
      "loss": 1.6948,
      "step": 1645
    },
    {
      "epoch": 0.3620367315517431,
      "grad_norm": 0.27037203311920166,
      "learning_rate": 8e-05,
      "loss": 1.6245,
      "step": 1646
    },
    {
      "epoch": 0.36225668096337843,
      "grad_norm": 0.2864437699317932,
      "learning_rate": 8e-05,
      "loss": 1.688,
      "step": 1647
    },
    {
      "epoch": 0.36247663037501376,
      "grad_norm": 0.27912065386772156,
      "learning_rate": 8e-05,
      "loss": 1.6056,
      "step": 1648
    },
    {
      "epoch": 0.3626965797866491,
      "grad_norm": 0.26467230916023254,
      "learning_rate": 8e-05,
      "loss": 1.5786,
      "step": 1649
    },
    {
      "epoch": 0.3629165291982844,
      "grad_norm": 0.2793690264225006,
      "learning_rate": 8e-05,
      "loss": 1.6003,
      "step": 1650
    },
    {
      "epoch": 0.3631364786099197,
      "grad_norm": 0.288629949092865,
      "learning_rate": 8e-05,
      "loss": 1.6752,
      "step": 1651
    },
    {
      "epoch": 0.36335642802155504,
      "grad_norm": 0.283195823431015,
      "learning_rate": 8e-05,
      "loss": 1.6854,
      "step": 1652
    },
    {
      "epoch": 0.3635763774331904,
      "grad_norm": 0.2929665446281433,
      "learning_rate": 8e-05,
      "loss": 1.7191,
      "step": 1653
    },
    {
      "epoch": 0.3637963268448257,
      "grad_norm": 0.28676289319992065,
      "learning_rate": 8e-05,
      "loss": 1.6959,
      "step": 1654
    },
    {
      "epoch": 0.364016276256461,
      "grad_norm": 0.264635294675827,
      "learning_rate": 8e-05,
      "loss": 1.6232,
      "step": 1655
    },
    {
      "epoch": 0.3642362256680963,
      "grad_norm": 0.2763380706310272,
      "learning_rate": 8e-05,
      "loss": 1.7631,
      "step": 1656
    },
    {
      "epoch": 0.36445617507973166,
      "grad_norm": 0.2624233365058899,
      "learning_rate": 8e-05,
      "loss": 1.6635,
      "step": 1657
    },
    {
      "epoch": 0.364676124491367,
      "grad_norm": 0.2564058303833008,
      "learning_rate": 8e-05,
      "loss": 1.4745,
      "step": 1658
    },
    {
      "epoch": 0.3648960739030023,
      "grad_norm": 0.2966236174106598,
      "learning_rate": 8e-05,
      "loss": 1.6892,
      "step": 1659
    },
    {
      "epoch": 0.36511602331463766,
      "grad_norm": 0.30588555335998535,
      "learning_rate": 8e-05,
      "loss": 1.6884,
      "step": 1660
    },
    {
      "epoch": 0.36533597272627294,
      "grad_norm": 0.2692076861858368,
      "learning_rate": 8e-05,
      "loss": 1.7158,
      "step": 1661
    },
    {
      "epoch": 0.3655559221379083,
      "grad_norm": 0.29388558864593506,
      "learning_rate": 8e-05,
      "loss": 1.7133,
      "step": 1662
    },
    {
      "epoch": 0.3657758715495436,
      "grad_norm": 0.28685635328292847,
      "learning_rate": 8e-05,
      "loss": 1.7444,
      "step": 1663
    },
    {
      "epoch": 0.36599582096117894,
      "grad_norm": 0.2885795831680298,
      "learning_rate": 8e-05,
      "loss": 1.7537,
      "step": 1664
    },
    {
      "epoch": 0.3662157703728143,
      "grad_norm": 0.3066631853580475,
      "learning_rate": 8e-05,
      "loss": 1.7843,
      "step": 1665
    },
    {
      "epoch": 0.36643571978444955,
      "grad_norm": 0.31112298369407654,
      "learning_rate": 8e-05,
      "loss": 1.4934,
      "step": 1666
    },
    {
      "epoch": 0.3666556691960849,
      "grad_norm": 0.2751656472682953,
      "learning_rate": 8e-05,
      "loss": 1.7463,
      "step": 1667
    },
    {
      "epoch": 0.3668756186077202,
      "grad_norm": 0.2834889590740204,
      "learning_rate": 8e-05,
      "loss": 1.7757,
      "step": 1668
    },
    {
      "epoch": 0.36709556801935556,
      "grad_norm": 0.2778145968914032,
      "learning_rate": 8e-05,
      "loss": 1.7423,
      "step": 1669
    },
    {
      "epoch": 0.3673155174309909,
      "grad_norm": 0.32161521911621094,
      "learning_rate": 8e-05,
      "loss": 1.8312,
      "step": 1670
    },
    {
      "epoch": 0.36753546684262617,
      "grad_norm": 0.27995115518569946,
      "learning_rate": 8e-05,
      "loss": 1.7694,
      "step": 1671
    },
    {
      "epoch": 0.3677554162542615,
      "grad_norm": 0.27701541781425476,
      "learning_rate": 8e-05,
      "loss": 1.8054,
      "step": 1672
    },
    {
      "epoch": 0.36797536566589684,
      "grad_norm": 0.2757355570793152,
      "learning_rate": 8e-05,
      "loss": 1.7036,
      "step": 1673
    },
    {
      "epoch": 0.36819531507753217,
      "grad_norm": 0.27305907011032104,
      "learning_rate": 8e-05,
      "loss": 1.6627,
      "step": 1674
    },
    {
      "epoch": 0.3684152644891675,
      "grad_norm": 0.299679696559906,
      "learning_rate": 8e-05,
      "loss": 1.7552,
      "step": 1675
    },
    {
      "epoch": 0.36863521390080284,
      "grad_norm": 0.2728777825832367,
      "learning_rate": 8e-05,
      "loss": 1.7649,
      "step": 1676
    },
    {
      "epoch": 0.3688551633124381,
      "grad_norm": 0.26330089569091797,
      "learning_rate": 8e-05,
      "loss": 1.5887,
      "step": 1677
    },
    {
      "epoch": 0.36907511272407345,
      "grad_norm": 0.2850317060947418,
      "learning_rate": 8e-05,
      "loss": 1.6255,
      "step": 1678
    },
    {
      "epoch": 0.3692950621357088,
      "grad_norm": 0.2784862220287323,
      "learning_rate": 8e-05,
      "loss": 1.7123,
      "step": 1679
    },
    {
      "epoch": 0.3695150115473441,
      "grad_norm": 0.284298300743103,
      "learning_rate": 8e-05,
      "loss": 1.5809,
      "step": 1680
    },
    {
      "epoch": 0.36973496095897945,
      "grad_norm": 0.2725334167480469,
      "learning_rate": 8e-05,
      "loss": 1.7037,
      "step": 1681
    },
    {
      "epoch": 0.36995491037061473,
      "grad_norm": 0.2760758399963379,
      "learning_rate": 8e-05,
      "loss": 1.6827,
      "step": 1682
    },
    {
      "epoch": 0.37017485978225007,
      "grad_norm": 0.2661541700363159,
      "learning_rate": 8e-05,
      "loss": 1.7042,
      "step": 1683
    },
    {
      "epoch": 0.3703948091938854,
      "grad_norm": 0.27737516164779663,
      "learning_rate": 8e-05,
      "loss": 1.7689,
      "step": 1684
    },
    {
      "epoch": 0.37061475860552073,
      "grad_norm": 0.2607424259185791,
      "learning_rate": 8e-05,
      "loss": 1.6356,
      "step": 1685
    },
    {
      "epoch": 0.37083470801715607,
      "grad_norm": 0.2802969217300415,
      "learning_rate": 8e-05,
      "loss": 1.7004,
      "step": 1686
    },
    {
      "epoch": 0.3710546574287914,
      "grad_norm": 0.2660817801952362,
      "learning_rate": 8e-05,
      "loss": 1.5539,
      "step": 1687
    },
    {
      "epoch": 0.3712746068404267,
      "grad_norm": 0.27867192029953003,
      "learning_rate": 8e-05,
      "loss": 1.6531,
      "step": 1688
    },
    {
      "epoch": 0.371494556252062,
      "grad_norm": 0.27857083082199097,
      "learning_rate": 8e-05,
      "loss": 1.8023,
      "step": 1689
    },
    {
      "epoch": 0.37171450566369735,
      "grad_norm": 0.2689161002635956,
      "learning_rate": 8e-05,
      "loss": 1.7601,
      "step": 1690
    },
    {
      "epoch": 0.3719344550753327,
      "grad_norm": 0.297826886177063,
      "learning_rate": 8e-05,
      "loss": 1.7627,
      "step": 1691
    },
    {
      "epoch": 0.372154404486968,
      "grad_norm": 0.2592705190181732,
      "learning_rate": 8e-05,
      "loss": 1.7132,
      "step": 1692
    },
    {
      "epoch": 0.3723743538986033,
      "grad_norm": 0.28288522362709045,
      "learning_rate": 8e-05,
      "loss": 1.7604,
      "step": 1693
    },
    {
      "epoch": 0.37259430331023863,
      "grad_norm": 0.30823859572410583,
      "learning_rate": 8e-05,
      "loss": 1.8563,
      "step": 1694
    },
    {
      "epoch": 0.37281425272187396,
      "grad_norm": 0.27835527062416077,
      "learning_rate": 8e-05,
      "loss": 1.6816,
      "step": 1695
    },
    {
      "epoch": 0.3730342021335093,
      "grad_norm": 0.2626672089099884,
      "learning_rate": 8e-05,
      "loss": 1.6185,
      "step": 1696
    },
    {
      "epoch": 0.37325415154514463,
      "grad_norm": 0.2489227056503296,
      "learning_rate": 8e-05,
      "loss": 1.6119,
      "step": 1697
    },
    {
      "epoch": 0.37347410095677996,
      "grad_norm": 0.28637897968292236,
      "learning_rate": 8e-05,
      "loss": 1.6695,
      "step": 1698
    },
    {
      "epoch": 0.37369405036841524,
      "grad_norm": 0.27077022194862366,
      "learning_rate": 8e-05,
      "loss": 1.6095,
      "step": 1699
    },
    {
      "epoch": 0.3739139997800506,
      "grad_norm": 0.32049357891082764,
      "learning_rate": 8e-05,
      "loss": 1.875,
      "step": 1700
    },
    {
      "epoch": 0.3741339491916859,
      "grad_norm": 0.2890382707118988,
      "learning_rate": 8e-05,
      "loss": 1.7129,
      "step": 1701
    },
    {
      "epoch": 0.37435389860332124,
      "grad_norm": 0.2785224914550781,
      "learning_rate": 8e-05,
      "loss": 1.7162,
      "step": 1702
    },
    {
      "epoch": 0.3745738480149566,
      "grad_norm": 0.2685299217700958,
      "learning_rate": 8e-05,
      "loss": 1.7358,
      "step": 1703
    },
    {
      "epoch": 0.37479379742659186,
      "grad_norm": 0.2840120494365692,
      "learning_rate": 8e-05,
      "loss": 1.9123,
      "step": 1704
    },
    {
      "epoch": 0.3750137468382272,
      "grad_norm": 0.27426856756210327,
      "learning_rate": 8e-05,
      "loss": 1.7144,
      "step": 1705
    },
    {
      "epoch": 0.3752336962498625,
      "grad_norm": 0.2707318663597107,
      "learning_rate": 8e-05,
      "loss": 1.6961,
      "step": 1706
    },
    {
      "epoch": 0.37545364566149786,
      "grad_norm": 0.3059745728969574,
      "learning_rate": 8e-05,
      "loss": 1.7491,
      "step": 1707
    },
    {
      "epoch": 0.3756735950731332,
      "grad_norm": 0.27109962701797485,
      "learning_rate": 8e-05,
      "loss": 1.6515,
      "step": 1708
    },
    {
      "epoch": 0.3758935444847685,
      "grad_norm": 0.26874709129333496,
      "learning_rate": 8e-05,
      "loss": 1.7119,
      "step": 1709
    },
    {
      "epoch": 0.3761134938964038,
      "grad_norm": 0.27959340810775757,
      "learning_rate": 8e-05,
      "loss": 1.5449,
      "step": 1710
    },
    {
      "epoch": 0.37633344330803914,
      "grad_norm": 0.284386545419693,
      "learning_rate": 8e-05,
      "loss": 1.8336,
      "step": 1711
    },
    {
      "epoch": 0.3765533927196745,
      "grad_norm": 0.27861231565475464,
      "learning_rate": 8e-05,
      "loss": 1.7547,
      "step": 1712
    },
    {
      "epoch": 0.3767733421313098,
      "grad_norm": 0.26845625042915344,
      "learning_rate": 8e-05,
      "loss": 1.6838,
      "step": 1713
    },
    {
      "epoch": 0.37699329154294514,
      "grad_norm": 0.31240981817245483,
      "learning_rate": 8e-05,
      "loss": 1.7489,
      "step": 1714
    },
    {
      "epoch": 0.3772132409545804,
      "grad_norm": 0.2878013253211975,
      "learning_rate": 8e-05,
      "loss": 1.7533,
      "step": 1715
    },
    {
      "epoch": 0.37743319036621575,
      "grad_norm": 0.27676892280578613,
      "learning_rate": 8e-05,
      "loss": 1.6218,
      "step": 1716
    },
    {
      "epoch": 0.3776531397778511,
      "grad_norm": 0.2782065272331238,
      "learning_rate": 8e-05,
      "loss": 1.6311,
      "step": 1717
    },
    {
      "epoch": 0.3778730891894864,
      "grad_norm": 0.2829797863960266,
      "learning_rate": 8e-05,
      "loss": 1.5863,
      "step": 1718
    },
    {
      "epoch": 0.37809303860112176,
      "grad_norm": 0.2851261794567108,
      "learning_rate": 8e-05,
      "loss": 1.8365,
      "step": 1719
    },
    {
      "epoch": 0.3783129880127571,
      "grad_norm": 0.2844488322734833,
      "learning_rate": 8e-05,
      "loss": 1.7765,
      "step": 1720
    },
    {
      "epoch": 0.37853293742439237,
      "grad_norm": 0.2976120412349701,
      "learning_rate": 8e-05,
      "loss": 1.7334,
      "step": 1721
    },
    {
      "epoch": 0.3787528868360277,
      "grad_norm": 0.27947840094566345,
      "learning_rate": 8e-05,
      "loss": 1.6641,
      "step": 1722
    },
    {
      "epoch": 0.37897283624766304,
      "grad_norm": 0.2986278831958771,
      "learning_rate": 8e-05,
      "loss": 1.8201,
      "step": 1723
    },
    {
      "epoch": 0.37919278565929837,
      "grad_norm": 0.26200374960899353,
      "learning_rate": 8e-05,
      "loss": 1.5835,
      "step": 1724
    },
    {
      "epoch": 0.3794127350709337,
      "grad_norm": 0.2846388816833496,
      "learning_rate": 8e-05,
      "loss": 1.7863,
      "step": 1725
    },
    {
      "epoch": 0.379632684482569,
      "grad_norm": 0.2809320390224457,
      "learning_rate": 8e-05,
      "loss": 1.6667,
      "step": 1726
    },
    {
      "epoch": 0.3798526338942043,
      "grad_norm": 0.28523099422454834,
      "learning_rate": 8e-05,
      "loss": 1.6647,
      "step": 1727
    },
    {
      "epoch": 0.38007258330583965,
      "grad_norm": 0.2719436287879944,
      "learning_rate": 8e-05,
      "loss": 1.6,
      "step": 1728
    },
    {
      "epoch": 0.380292532717475,
      "grad_norm": 0.2762429118156433,
      "learning_rate": 8e-05,
      "loss": 1.6888,
      "step": 1729
    },
    {
      "epoch": 0.3805124821291103,
      "grad_norm": 0.30161863565444946,
      "learning_rate": 8e-05,
      "loss": 1.6659,
      "step": 1730
    },
    {
      "epoch": 0.38073243154074565,
      "grad_norm": 0.27962687611579895,
      "learning_rate": 8e-05,
      "loss": 1.629,
      "step": 1731
    },
    {
      "epoch": 0.38095238095238093,
      "grad_norm": 0.27580323815345764,
      "learning_rate": 8e-05,
      "loss": 1.689,
      "step": 1732
    },
    {
      "epoch": 0.38117233036401627,
      "grad_norm": 0.2676113545894623,
      "learning_rate": 8e-05,
      "loss": 1.7195,
      "step": 1733
    },
    {
      "epoch": 0.3813922797756516,
      "grad_norm": 0.27840152382850647,
      "learning_rate": 8e-05,
      "loss": 1.6433,
      "step": 1734
    },
    {
      "epoch": 0.38161222918728693,
      "grad_norm": 0.27100005745887756,
      "learning_rate": 8e-05,
      "loss": 1.6517,
      "step": 1735
    },
    {
      "epoch": 0.38183217859892227,
      "grad_norm": 0.2874828577041626,
      "learning_rate": 8e-05,
      "loss": 1.9139,
      "step": 1736
    },
    {
      "epoch": 0.38205212801055755,
      "grad_norm": 0.2685931324958801,
      "learning_rate": 8e-05,
      "loss": 1.7373,
      "step": 1737
    },
    {
      "epoch": 0.3822720774221929,
      "grad_norm": 0.2895548641681671,
      "learning_rate": 8e-05,
      "loss": 1.7828,
      "step": 1738
    },
    {
      "epoch": 0.3824920268338282,
      "grad_norm": 0.29109206795692444,
      "learning_rate": 8e-05,
      "loss": 1.6347,
      "step": 1739
    },
    {
      "epoch": 0.38271197624546355,
      "grad_norm": 0.2804923951625824,
      "learning_rate": 8e-05,
      "loss": 1.5978,
      "step": 1740
    },
    {
      "epoch": 0.3829319256570989,
      "grad_norm": 0.2829732894897461,
      "learning_rate": 8e-05,
      "loss": 1.6271,
      "step": 1741
    },
    {
      "epoch": 0.3831518750687342,
      "grad_norm": 0.28979840874671936,
      "learning_rate": 8e-05,
      "loss": 1.7244,
      "step": 1742
    },
    {
      "epoch": 0.3833718244803695,
      "grad_norm": 0.30159792304039,
      "learning_rate": 8e-05,
      "loss": 1.8074,
      "step": 1743
    },
    {
      "epoch": 0.38359177389200483,
      "grad_norm": 0.28228580951690674,
      "learning_rate": 8e-05,
      "loss": 1.6669,
      "step": 1744
    },
    {
      "epoch": 0.38381172330364016,
      "grad_norm": 0.27950945496559143,
      "learning_rate": 8e-05,
      "loss": 1.6583,
      "step": 1745
    },
    {
      "epoch": 0.3840316727152755,
      "grad_norm": 0.2708896994590759,
      "learning_rate": 8e-05,
      "loss": 1.5793,
      "step": 1746
    },
    {
      "epoch": 0.38425162212691083,
      "grad_norm": 0.27368029952049255,
      "learning_rate": 8e-05,
      "loss": 1.6371,
      "step": 1747
    },
    {
      "epoch": 0.3844715715385461,
      "grad_norm": 0.27621379494667053,
      "learning_rate": 8e-05,
      "loss": 1.5737,
      "step": 1748
    },
    {
      "epoch": 0.38469152095018144,
      "grad_norm": 0.27143922448158264,
      "learning_rate": 8e-05,
      "loss": 1.7289,
      "step": 1749
    },
    {
      "epoch": 0.3849114703618168,
      "grad_norm": 0.28887274861335754,
      "learning_rate": 8e-05,
      "loss": 1.7262,
      "step": 1750
    },
    {
      "epoch": 0.3851314197734521,
      "grad_norm": 0.26516541838645935,
      "learning_rate": 8e-05,
      "loss": 1.6358,
      "step": 1751
    },
    {
      "epoch": 0.38535136918508744,
      "grad_norm": 0.31475701928138733,
      "learning_rate": 8e-05,
      "loss": 1.8599,
      "step": 1752
    },
    {
      "epoch": 0.3855713185967228,
      "grad_norm": 0.27711552381515503,
      "learning_rate": 8e-05,
      "loss": 1.632,
      "step": 1753
    },
    {
      "epoch": 0.38579126800835806,
      "grad_norm": 0.27542901039123535,
      "learning_rate": 8e-05,
      "loss": 1.6741,
      "step": 1754
    },
    {
      "epoch": 0.3860112174199934,
      "grad_norm": 0.2941054701805115,
      "learning_rate": 8e-05,
      "loss": 1.6396,
      "step": 1755
    },
    {
      "epoch": 0.3862311668316287,
      "grad_norm": 0.27836698293685913,
      "learning_rate": 8e-05,
      "loss": 1.5689,
      "step": 1756
    },
    {
      "epoch": 0.38645111624326406,
      "grad_norm": 0.29147645831108093,
      "learning_rate": 8e-05,
      "loss": 1.7523,
      "step": 1757
    },
    {
      "epoch": 0.3866710656548994,
      "grad_norm": 0.30084285140037537,
      "learning_rate": 8e-05,
      "loss": 1.6932,
      "step": 1758
    },
    {
      "epoch": 0.38689101506653467,
      "grad_norm": 0.2850727140903473,
      "learning_rate": 8e-05,
      "loss": 1.7276,
      "step": 1759
    },
    {
      "epoch": 0.38711096447817,
      "grad_norm": 0.27011391520500183,
      "learning_rate": 8e-05,
      "loss": 1.6635,
      "step": 1760
    },
    {
      "epoch": 0.38733091388980534,
      "grad_norm": 0.28682348132133484,
      "learning_rate": 8e-05,
      "loss": 1.7642,
      "step": 1761
    },
    {
      "epoch": 0.3875508633014407,
      "grad_norm": 0.27676117420196533,
      "learning_rate": 8e-05,
      "loss": 1.7406,
      "step": 1762
    },
    {
      "epoch": 0.387770812713076,
      "grad_norm": 0.2654523551464081,
      "learning_rate": 8e-05,
      "loss": 1.6484,
      "step": 1763
    },
    {
      "epoch": 0.38799076212471134,
      "grad_norm": 0.28026026487350464,
      "learning_rate": 8e-05,
      "loss": 1.6714,
      "step": 1764
    },
    {
      "epoch": 0.3882107115363466,
      "grad_norm": 0.3003789782524109,
      "learning_rate": 8e-05,
      "loss": 1.8121,
      "step": 1765
    },
    {
      "epoch": 0.38843066094798195,
      "grad_norm": 0.35523107647895813,
      "learning_rate": 8e-05,
      "loss": 1.9299,
      "step": 1766
    },
    {
      "epoch": 0.3886506103596173,
      "grad_norm": 0.26844245195388794,
      "learning_rate": 8e-05,
      "loss": 1.6358,
      "step": 1767
    },
    {
      "epoch": 0.3888705597712526,
      "grad_norm": 0.27308356761932373,
      "learning_rate": 8e-05,
      "loss": 1.6104,
      "step": 1768
    },
    {
      "epoch": 0.38909050918288796,
      "grad_norm": 0.2775373160839081,
      "learning_rate": 8e-05,
      "loss": 1.5679,
      "step": 1769
    },
    {
      "epoch": 0.38931045859452323,
      "grad_norm": 0.29753705859184265,
      "learning_rate": 8e-05,
      "loss": 1.7678,
      "step": 1770
    },
    {
      "epoch": 0.38953040800615857,
      "grad_norm": 0.2798722982406616,
      "learning_rate": 8e-05,
      "loss": 1.7034,
      "step": 1771
    },
    {
      "epoch": 0.3897503574177939,
      "grad_norm": 0.2842818796634674,
      "learning_rate": 8e-05,
      "loss": 1.727,
      "step": 1772
    },
    {
      "epoch": 0.38997030682942924,
      "grad_norm": 0.27555832266807556,
      "learning_rate": 8e-05,
      "loss": 1.7438,
      "step": 1773
    },
    {
      "epoch": 0.39019025624106457,
      "grad_norm": 0.2824547588825226,
      "learning_rate": 8e-05,
      "loss": 1.5733,
      "step": 1774
    },
    {
      "epoch": 0.3904102056526999,
      "grad_norm": 0.2658035159111023,
      "learning_rate": 8e-05,
      "loss": 1.5997,
      "step": 1775
    },
    {
      "epoch": 0.3906301550643352,
      "grad_norm": 0.27601394057273865,
      "learning_rate": 8e-05,
      "loss": 1.7025,
      "step": 1776
    },
    {
      "epoch": 0.3908501044759705,
      "grad_norm": 0.2990022897720337,
      "learning_rate": 8e-05,
      "loss": 1.8278,
      "step": 1777
    },
    {
      "epoch": 0.39107005388760585,
      "grad_norm": 0.29378873109817505,
      "learning_rate": 8e-05,
      "loss": 1.7549,
      "step": 1778
    },
    {
      "epoch": 0.3912900032992412,
      "grad_norm": 0.29202136397361755,
      "learning_rate": 8e-05,
      "loss": 1.6602,
      "step": 1779
    },
    {
      "epoch": 0.3915099527108765,
      "grad_norm": 0.28191903233528137,
      "learning_rate": 8e-05,
      "loss": 1.6618,
      "step": 1780
    },
    {
      "epoch": 0.3917299021225118,
      "grad_norm": 0.26916682720184326,
      "learning_rate": 8e-05,
      "loss": 1.6523,
      "step": 1781
    },
    {
      "epoch": 0.39194985153414713,
      "grad_norm": 0.2886850833892822,
      "learning_rate": 8e-05,
      "loss": 1.5752,
      "step": 1782
    },
    {
      "epoch": 0.39216980094578247,
      "grad_norm": 0.2749246656894684,
      "learning_rate": 8e-05,
      "loss": 1.6731,
      "step": 1783
    },
    {
      "epoch": 0.3923897503574178,
      "grad_norm": 0.28945374488830566,
      "learning_rate": 8e-05,
      "loss": 1.7404,
      "step": 1784
    },
    {
      "epoch": 0.39260969976905313,
      "grad_norm": 0.27297836542129517,
      "learning_rate": 8e-05,
      "loss": 1.7176,
      "step": 1785
    },
    {
      "epoch": 0.39282964918068847,
      "grad_norm": 0.2738782465457916,
      "learning_rate": 8e-05,
      "loss": 1.6844,
      "step": 1786
    },
    {
      "epoch": 0.39304959859232375,
      "grad_norm": 0.2897050082683563,
      "learning_rate": 8e-05,
      "loss": 1.6522,
      "step": 1787
    },
    {
      "epoch": 0.3932695480039591,
      "grad_norm": 0.31031668186187744,
      "learning_rate": 8e-05,
      "loss": 1.8216,
      "step": 1788
    },
    {
      "epoch": 0.3934894974155944,
      "grad_norm": 0.2869516909122467,
      "learning_rate": 8e-05,
      "loss": 1.7598,
      "step": 1789
    },
    {
      "epoch": 0.39370944682722975,
      "grad_norm": 0.3080596625804901,
      "learning_rate": 8e-05,
      "loss": 1.7645,
      "step": 1790
    },
    {
      "epoch": 0.3939293962388651,
      "grad_norm": 0.27992716431617737,
      "learning_rate": 8e-05,
      "loss": 1.8373,
      "step": 1791
    },
    {
      "epoch": 0.39414934565050036,
      "grad_norm": 0.2761777341365814,
      "learning_rate": 8e-05,
      "loss": 1.6767,
      "step": 1792
    },
    {
      "epoch": 0.3943692950621357,
      "grad_norm": 0.30193084478378296,
      "learning_rate": 8e-05,
      "loss": 1.7394,
      "step": 1793
    },
    {
      "epoch": 0.39458924447377103,
      "grad_norm": 0.29375529289245605,
      "learning_rate": 8e-05,
      "loss": 1.761,
      "step": 1794
    },
    {
      "epoch": 0.39480919388540636,
      "grad_norm": 0.32190364599227905,
      "learning_rate": 8e-05,
      "loss": 1.7045,
      "step": 1795
    },
    {
      "epoch": 0.3950291432970417,
      "grad_norm": 0.27505311369895935,
      "learning_rate": 8e-05,
      "loss": 1.6678,
      "step": 1796
    },
    {
      "epoch": 0.39524909270867703,
      "grad_norm": 0.28678107261657715,
      "learning_rate": 8e-05,
      "loss": 1.7155,
      "step": 1797
    },
    {
      "epoch": 0.3954690421203123,
      "grad_norm": 0.28372088074684143,
      "learning_rate": 8e-05,
      "loss": 1.8833,
      "step": 1798
    },
    {
      "epoch": 0.39568899153194764,
      "grad_norm": 0.27803388237953186,
      "learning_rate": 8e-05,
      "loss": 1.7125,
      "step": 1799
    },
    {
      "epoch": 0.395908940943583,
      "grad_norm": 0.278728187084198,
      "learning_rate": 8e-05,
      "loss": 1.7019,
      "step": 1800
    },
    {
      "epoch": 0.3961288903552183,
      "grad_norm": 0.29563671350479126,
      "learning_rate": 8e-05,
      "loss": 1.6588,
      "step": 1801
    },
    {
      "epoch": 0.39634883976685364,
      "grad_norm": 0.35585105419158936,
      "learning_rate": 8e-05,
      "loss": 1.8864,
      "step": 1802
    },
    {
      "epoch": 0.3965687891784889,
      "grad_norm": 0.27399691939353943,
      "learning_rate": 8e-05,
      "loss": 1.6222,
      "step": 1803
    },
    {
      "epoch": 0.39678873859012426,
      "grad_norm": 0.2557234764099121,
      "learning_rate": 8e-05,
      "loss": 1.4835,
      "step": 1804
    },
    {
      "epoch": 0.3970086880017596,
      "grad_norm": 0.2929818332195282,
      "learning_rate": 8e-05,
      "loss": 1.7737,
      "step": 1805
    },
    {
      "epoch": 0.3972286374133949,
      "grad_norm": 0.279729425907135,
      "learning_rate": 8e-05,
      "loss": 1.6073,
      "step": 1806
    },
    {
      "epoch": 0.39744858682503026,
      "grad_norm": 0.2622847259044647,
      "learning_rate": 8e-05,
      "loss": 1.5368,
      "step": 1807
    },
    {
      "epoch": 0.3976685362366656,
      "grad_norm": 0.2704038619995117,
      "learning_rate": 8e-05,
      "loss": 1.5809,
      "step": 1808
    },
    {
      "epoch": 0.39788848564830087,
      "grad_norm": 0.2785516679286957,
      "learning_rate": 8e-05,
      "loss": 1.6416,
      "step": 1809
    },
    {
      "epoch": 0.3981084350599362,
      "grad_norm": 0.29611852765083313,
      "learning_rate": 8e-05,
      "loss": 1.7186,
      "step": 1810
    },
    {
      "epoch": 0.39832838447157154,
      "grad_norm": 0.28127896785736084,
      "learning_rate": 8e-05,
      "loss": 1.6248,
      "step": 1811
    },
    {
      "epoch": 0.3985483338832069,
      "grad_norm": 0.2746615707874298,
      "learning_rate": 8e-05,
      "loss": 1.7029,
      "step": 1812
    },
    {
      "epoch": 0.3987682832948422,
      "grad_norm": 0.2650880515575409,
      "learning_rate": 8e-05,
      "loss": 1.7828,
      "step": 1813
    },
    {
      "epoch": 0.3989882327064775,
      "grad_norm": 0.28278401494026184,
      "learning_rate": 8e-05,
      "loss": 1.6383,
      "step": 1814
    },
    {
      "epoch": 0.3992081821181128,
      "grad_norm": 0.2749755382537842,
      "learning_rate": 8e-05,
      "loss": 1.7551,
      "step": 1815
    },
    {
      "epoch": 0.39942813152974815,
      "grad_norm": 0.26788878440856934,
      "learning_rate": 8e-05,
      "loss": 1.5206,
      "step": 1816
    },
    {
      "epoch": 0.3996480809413835,
      "grad_norm": 0.28166842460632324,
      "learning_rate": 8e-05,
      "loss": 1.7134,
      "step": 1817
    },
    {
      "epoch": 0.3998680303530188,
      "grad_norm": 0.2781674563884735,
      "learning_rate": 8e-05,
      "loss": 1.7864,
      "step": 1818
    },
    {
      "epoch": 0.40008797976465416,
      "grad_norm": 0.2810186445713043,
      "learning_rate": 8e-05,
      "loss": 1.7017,
      "step": 1819
    },
    {
      "epoch": 0.40030792917628943,
      "grad_norm": 0.2872167229652405,
      "learning_rate": 8e-05,
      "loss": 1.563,
      "step": 1820
    },
    {
      "epoch": 0.40052787858792477,
      "grad_norm": 0.2864447832107544,
      "learning_rate": 8e-05,
      "loss": 1.8261,
      "step": 1821
    },
    {
      "epoch": 0.4007478279995601,
      "grad_norm": 0.2633639872074127,
      "learning_rate": 8e-05,
      "loss": 1.5864,
      "step": 1822
    },
    {
      "epoch": 0.40096777741119544,
      "grad_norm": 0.3556326925754547,
      "learning_rate": 8e-05,
      "loss": 1.5781,
      "step": 1823
    },
    {
      "epoch": 0.40118772682283077,
      "grad_norm": 0.2832813560962677,
      "learning_rate": 8e-05,
      "loss": 1.7145,
      "step": 1824
    },
    {
      "epoch": 0.40140767623446605,
      "grad_norm": 0.2862699627876282,
      "learning_rate": 8e-05,
      "loss": 1.7753,
      "step": 1825
    },
    {
      "epoch": 0.4016276256461014,
      "grad_norm": 0.3274460732936859,
      "learning_rate": 8e-05,
      "loss": 1.7979,
      "step": 1826
    },
    {
      "epoch": 0.4018475750577367,
      "grad_norm": 0.277118444442749,
      "learning_rate": 8e-05,
      "loss": 1.6089,
      "step": 1827
    },
    {
      "epoch": 0.40206752446937205,
      "grad_norm": 0.278337687253952,
      "learning_rate": 8e-05,
      "loss": 1.5788,
      "step": 1828
    },
    {
      "epoch": 0.4022874738810074,
      "grad_norm": 0.28072914481163025,
      "learning_rate": 8e-05,
      "loss": 1.6735,
      "step": 1829
    },
    {
      "epoch": 0.4025074232926427,
      "grad_norm": 0.2815505564212799,
      "learning_rate": 8e-05,
      "loss": 1.7593,
      "step": 1830
    },
    {
      "epoch": 0.402727372704278,
      "grad_norm": 0.2957006096839905,
      "learning_rate": 8e-05,
      "loss": 1.865,
      "step": 1831
    },
    {
      "epoch": 0.40294732211591333,
      "grad_norm": 0.3079582452774048,
      "learning_rate": 8e-05,
      "loss": 1.7421,
      "step": 1832
    },
    {
      "epoch": 0.40316727152754867,
      "grad_norm": 0.2924387454986572,
      "learning_rate": 8e-05,
      "loss": 1.7462,
      "step": 1833
    },
    {
      "epoch": 0.403387220939184,
      "grad_norm": 0.28879454731941223,
      "learning_rate": 8e-05,
      "loss": 1.7433,
      "step": 1834
    },
    {
      "epoch": 0.40360717035081933,
      "grad_norm": 0.27446237206459045,
      "learning_rate": 8e-05,
      "loss": 1.5869,
      "step": 1835
    },
    {
      "epoch": 0.4038271197624546,
      "grad_norm": 0.3164878487586975,
      "learning_rate": 8e-05,
      "loss": 1.7505,
      "step": 1836
    },
    {
      "epoch": 0.40404706917408995,
      "grad_norm": 0.25979530811309814,
      "learning_rate": 8e-05,
      "loss": 1.6001,
      "step": 1837
    },
    {
      "epoch": 0.4042670185857253,
      "grad_norm": 0.30625709891319275,
      "learning_rate": 8e-05,
      "loss": 1.7907,
      "step": 1838
    },
    {
      "epoch": 0.4044869679973606,
      "grad_norm": 0.27351540327072144,
      "learning_rate": 8e-05,
      "loss": 1.5835,
      "step": 1839
    },
    {
      "epoch": 0.40470691740899595,
      "grad_norm": 0.302372545003891,
      "learning_rate": 8e-05,
      "loss": 1.7821,
      "step": 1840
    },
    {
      "epoch": 0.4049268668206313,
      "grad_norm": 0.2910183370113373,
      "learning_rate": 8e-05,
      "loss": 1.7993,
      "step": 1841
    },
    {
      "epoch": 0.40514681623226656,
      "grad_norm": 0.2934883236885071,
      "learning_rate": 8e-05,
      "loss": 1.5928,
      "step": 1842
    },
    {
      "epoch": 0.4053667656439019,
      "grad_norm": 0.2586327791213989,
      "learning_rate": 8e-05,
      "loss": 1.5714,
      "step": 1843
    },
    {
      "epoch": 0.40558671505553723,
      "grad_norm": 0.27952027320861816,
      "learning_rate": 8e-05,
      "loss": 1.8168,
      "step": 1844
    },
    {
      "epoch": 0.40580666446717256,
      "grad_norm": 0.2987437844276428,
      "learning_rate": 8e-05,
      "loss": 1.555,
      "step": 1845
    },
    {
      "epoch": 0.4060266138788079,
      "grad_norm": 0.29165002703666687,
      "learning_rate": 8e-05,
      "loss": 1.6865,
      "step": 1846
    },
    {
      "epoch": 0.4062465632904432,
      "grad_norm": 0.2825503945350647,
      "learning_rate": 8e-05,
      "loss": 1.8814,
      "step": 1847
    },
    {
      "epoch": 0.4064665127020785,
      "grad_norm": 0.27995482087135315,
      "learning_rate": 8e-05,
      "loss": 1.6897,
      "step": 1848
    },
    {
      "epoch": 0.40668646211371384,
      "grad_norm": 0.2735064923763275,
      "learning_rate": 8e-05,
      "loss": 1.7279,
      "step": 1849
    },
    {
      "epoch": 0.4069064115253492,
      "grad_norm": 0.2850511074066162,
      "learning_rate": 8e-05,
      "loss": 1.6459,
      "step": 1850
    },
    {
      "epoch": 0.4071263609369845,
      "grad_norm": 0.3000599145889282,
      "learning_rate": 8e-05,
      "loss": 1.7301,
      "step": 1851
    },
    {
      "epoch": 0.40734631034861984,
      "grad_norm": 0.2768002152442932,
      "learning_rate": 8e-05,
      "loss": 1.5748,
      "step": 1852
    },
    {
      "epoch": 0.4075662597602551,
      "grad_norm": 0.26737141609191895,
      "learning_rate": 8e-05,
      "loss": 1.5895,
      "step": 1853
    },
    {
      "epoch": 0.40778620917189046,
      "grad_norm": 0.26408424973487854,
      "learning_rate": 8e-05,
      "loss": 1.5611,
      "step": 1854
    },
    {
      "epoch": 0.4080061585835258,
      "grad_norm": 0.2646276354789734,
      "learning_rate": 8e-05,
      "loss": 1.5865,
      "step": 1855
    },
    {
      "epoch": 0.4082261079951611,
      "grad_norm": 0.27871212363243103,
      "learning_rate": 8e-05,
      "loss": 1.8202,
      "step": 1856
    },
    {
      "epoch": 0.40844605740679646,
      "grad_norm": 0.3234533965587616,
      "learning_rate": 8e-05,
      "loss": 1.8213,
      "step": 1857
    },
    {
      "epoch": 0.40866600681843174,
      "grad_norm": 0.2705099284648895,
      "learning_rate": 8e-05,
      "loss": 1.6637,
      "step": 1858
    },
    {
      "epoch": 0.40888595623006707,
      "grad_norm": 0.28647711873054504,
      "learning_rate": 8e-05,
      "loss": 1.7396,
      "step": 1859
    },
    {
      "epoch": 0.4091059056417024,
      "grad_norm": 0.2812083959579468,
      "learning_rate": 8e-05,
      "loss": 1.5663,
      "step": 1860
    },
    {
      "epoch": 0.40932585505333774,
      "grad_norm": 0.2818193733692169,
      "learning_rate": 8e-05,
      "loss": 1.6073,
      "step": 1861
    },
    {
      "epoch": 0.4095458044649731,
      "grad_norm": 0.29906994104385376,
      "learning_rate": 8e-05,
      "loss": 1.7061,
      "step": 1862
    },
    {
      "epoch": 0.4097657538766084,
      "grad_norm": 0.27941465377807617,
      "learning_rate": 8e-05,
      "loss": 1.7282,
      "step": 1863
    },
    {
      "epoch": 0.4099857032882437,
      "grad_norm": 0.27629899978637695,
      "learning_rate": 8e-05,
      "loss": 1.6879,
      "step": 1864
    },
    {
      "epoch": 0.410205652699879,
      "grad_norm": 0.2792319059371948,
      "learning_rate": 8e-05,
      "loss": 1.7196,
      "step": 1865
    },
    {
      "epoch": 0.41042560211151435,
      "grad_norm": 0.2763090431690216,
      "learning_rate": 8e-05,
      "loss": 1.673,
      "step": 1866
    },
    {
      "epoch": 0.4106455515231497,
      "grad_norm": 0.2930999994277954,
      "learning_rate": 8e-05,
      "loss": 1.6919,
      "step": 1867
    },
    {
      "epoch": 0.410865500934785,
      "grad_norm": 0.2748461365699768,
      "learning_rate": 8e-05,
      "loss": 1.7553,
      "step": 1868
    },
    {
      "epoch": 0.4110854503464203,
      "grad_norm": 0.2742187976837158,
      "learning_rate": 8e-05,
      "loss": 1.6786,
      "step": 1869
    },
    {
      "epoch": 0.41130539975805563,
      "grad_norm": 0.3050731420516968,
      "learning_rate": 8e-05,
      "loss": 1.4902,
      "step": 1870
    },
    {
      "epoch": 0.41152534916969097,
      "grad_norm": 0.29456627368927,
      "learning_rate": 8e-05,
      "loss": 1.758,
      "step": 1871
    },
    {
      "epoch": 0.4117452985813263,
      "grad_norm": 0.2844219505786896,
      "learning_rate": 8e-05,
      "loss": 1.6206,
      "step": 1872
    },
    {
      "epoch": 0.41196524799296164,
      "grad_norm": 0.28889915347099304,
      "learning_rate": 8e-05,
      "loss": 1.6907,
      "step": 1873
    },
    {
      "epoch": 0.41218519740459697,
      "grad_norm": 0.27245181798934937,
      "learning_rate": 8e-05,
      "loss": 1.6749,
      "step": 1874
    },
    {
      "epoch": 0.41240514681623225,
      "grad_norm": 0.2927252948284149,
      "learning_rate": 8e-05,
      "loss": 1.6382,
      "step": 1875
    },
    {
      "epoch": 0.4126250962278676,
      "grad_norm": 0.27153030037879944,
      "learning_rate": 8e-05,
      "loss": 1.6011,
      "step": 1876
    },
    {
      "epoch": 0.4128450456395029,
      "grad_norm": 0.2807110846042633,
      "learning_rate": 8e-05,
      "loss": 1.7126,
      "step": 1877
    },
    {
      "epoch": 0.41306499505113825,
      "grad_norm": 0.27375784516334534,
      "learning_rate": 8e-05,
      "loss": 1.7443,
      "step": 1878
    },
    {
      "epoch": 0.4132849444627736,
      "grad_norm": 0.27330929040908813,
      "learning_rate": 8e-05,
      "loss": 1.6305,
      "step": 1879
    },
    {
      "epoch": 0.41350489387440886,
      "grad_norm": 0.27126336097717285,
      "learning_rate": 8e-05,
      "loss": 1.6688,
      "step": 1880
    },
    {
      "epoch": 0.4137248432860442,
      "grad_norm": 0.2768147885799408,
      "learning_rate": 8e-05,
      "loss": 1.7274,
      "step": 1881
    },
    {
      "epoch": 0.41394479269767953,
      "grad_norm": 0.2686031460762024,
      "learning_rate": 8e-05,
      "loss": 1.6445,
      "step": 1882
    },
    {
      "epoch": 0.41416474210931487,
      "grad_norm": 0.27737778425216675,
      "learning_rate": 8e-05,
      "loss": 1.5226,
      "step": 1883
    },
    {
      "epoch": 0.4143846915209502,
      "grad_norm": 0.2761901319026947,
      "learning_rate": 8e-05,
      "loss": 1.6884,
      "step": 1884
    },
    {
      "epoch": 0.41460464093258553,
      "grad_norm": 0.28609856963157654,
      "learning_rate": 8e-05,
      "loss": 1.7719,
      "step": 1885
    },
    {
      "epoch": 0.4148245903442208,
      "grad_norm": 0.2904943525791168,
      "learning_rate": 8e-05,
      "loss": 1.6979,
      "step": 1886
    },
    {
      "epoch": 0.41504453975585615,
      "grad_norm": 0.3016435503959656,
      "learning_rate": 8e-05,
      "loss": 1.7912,
      "step": 1887
    },
    {
      "epoch": 0.4152644891674915,
      "grad_norm": 0.27562782168388367,
      "learning_rate": 8e-05,
      "loss": 1.5822,
      "step": 1888
    },
    {
      "epoch": 0.4154844385791268,
      "grad_norm": 0.2841348648071289,
      "learning_rate": 8e-05,
      "loss": 1.7524,
      "step": 1889
    },
    {
      "epoch": 0.41570438799076215,
      "grad_norm": 0.26393935084342957,
      "learning_rate": 8e-05,
      "loss": 1.6219,
      "step": 1890
    },
    {
      "epoch": 0.4159243374023974,
      "grad_norm": 0.2792678773403168,
      "learning_rate": 8e-05,
      "loss": 1.7243,
      "step": 1891
    },
    {
      "epoch": 0.41614428681403276,
      "grad_norm": 0.291425496339798,
      "learning_rate": 8e-05,
      "loss": 1.7499,
      "step": 1892
    },
    {
      "epoch": 0.4163642362256681,
      "grad_norm": 0.2737634778022766,
      "learning_rate": 8e-05,
      "loss": 1.6565,
      "step": 1893
    },
    {
      "epoch": 0.41658418563730343,
      "grad_norm": 0.26807767152786255,
      "learning_rate": 8e-05,
      "loss": 1.6149,
      "step": 1894
    },
    {
      "epoch": 0.41680413504893876,
      "grad_norm": 0.28826507925987244,
      "learning_rate": 8e-05,
      "loss": 1.6857,
      "step": 1895
    },
    {
      "epoch": 0.41702408446057404,
      "grad_norm": 0.27604466676712036,
      "learning_rate": 8e-05,
      "loss": 1.7689,
      "step": 1896
    },
    {
      "epoch": 0.4172440338722094,
      "grad_norm": 0.27355703711509705,
      "learning_rate": 8e-05,
      "loss": 1.7254,
      "step": 1897
    },
    {
      "epoch": 0.4174639832838447,
      "grad_norm": 0.26692044734954834,
      "learning_rate": 8e-05,
      "loss": 1.6372,
      "step": 1898
    },
    {
      "epoch": 0.41768393269548004,
      "grad_norm": 0.27527916431427,
      "learning_rate": 8e-05,
      "loss": 1.6913,
      "step": 1899
    },
    {
      "epoch": 0.4179038821071154,
      "grad_norm": 0.26881837844848633,
      "learning_rate": 8e-05,
      "loss": 1.6663,
      "step": 1900
    },
    {
      "epoch": 0.4181238315187507,
      "grad_norm": 0.27977946400642395,
      "learning_rate": 8e-05,
      "loss": 1.8117,
      "step": 1901
    },
    {
      "epoch": 0.418343780930386,
      "grad_norm": 0.2958911955356598,
      "learning_rate": 8e-05,
      "loss": 1.6603,
      "step": 1902
    },
    {
      "epoch": 0.4185637303420213,
      "grad_norm": 0.2845151424407959,
      "learning_rate": 8e-05,
      "loss": 1.7517,
      "step": 1903
    },
    {
      "epoch": 0.41878367975365666,
      "grad_norm": 0.2804581820964813,
      "learning_rate": 8e-05,
      "loss": 1.765,
      "step": 1904
    },
    {
      "epoch": 0.419003629165292,
      "grad_norm": 0.29568520188331604,
      "learning_rate": 8e-05,
      "loss": 1.6995,
      "step": 1905
    },
    {
      "epoch": 0.4192235785769273,
      "grad_norm": 0.303100049495697,
      "learning_rate": 8e-05,
      "loss": 1.77,
      "step": 1906
    },
    {
      "epoch": 0.4194435279885626,
      "grad_norm": 0.26847636699676514,
      "learning_rate": 8e-05,
      "loss": 1.6964,
      "step": 1907
    },
    {
      "epoch": 0.41966347740019794,
      "grad_norm": 0.2791590094566345,
      "learning_rate": 8e-05,
      "loss": 1.5912,
      "step": 1908
    },
    {
      "epoch": 0.41988342681183327,
      "grad_norm": 0.2687268555164337,
      "learning_rate": 8e-05,
      "loss": 1.6163,
      "step": 1909
    },
    {
      "epoch": 0.4201033762234686,
      "grad_norm": 0.29087433218955994,
      "learning_rate": 8e-05,
      "loss": 1.6588,
      "step": 1910
    },
    {
      "epoch": 0.42032332563510394,
      "grad_norm": 0.29639971256256104,
      "learning_rate": 8e-05,
      "loss": 1.6911,
      "step": 1911
    },
    {
      "epoch": 0.4205432750467393,
      "grad_norm": 0.27669841051101685,
      "learning_rate": 8e-05,
      "loss": 1.7078,
      "step": 1912
    },
    {
      "epoch": 0.42076322445837455,
      "grad_norm": 0.2851327955722809,
      "learning_rate": 8e-05,
      "loss": 1.7217,
      "step": 1913
    },
    {
      "epoch": 0.4209831738700099,
      "grad_norm": 0.27069011330604553,
      "learning_rate": 8e-05,
      "loss": 1.7076,
      "step": 1914
    },
    {
      "epoch": 0.4212031232816452,
      "grad_norm": 0.26195240020751953,
      "learning_rate": 8e-05,
      "loss": 1.6647,
      "step": 1915
    },
    {
      "epoch": 0.42142307269328055,
      "grad_norm": 0.3046209216117859,
      "learning_rate": 8e-05,
      "loss": 1.5303,
      "step": 1916
    },
    {
      "epoch": 0.4216430221049159,
      "grad_norm": 0.29437899589538574,
      "learning_rate": 8e-05,
      "loss": 1.6589,
      "step": 1917
    },
    {
      "epoch": 0.42186297151655117,
      "grad_norm": 0.2954728603363037,
      "learning_rate": 8e-05,
      "loss": 1.777,
      "step": 1918
    },
    {
      "epoch": 0.4220829209281865,
      "grad_norm": 0.2612738609313965,
      "learning_rate": 8e-05,
      "loss": 1.5668,
      "step": 1919
    },
    {
      "epoch": 0.42230287033982183,
      "grad_norm": 0.3015122413635254,
      "learning_rate": 8e-05,
      "loss": 1.7861,
      "step": 1920
    },
    {
      "epoch": 0.42252281975145717,
      "grad_norm": 0.3785838484764099,
      "learning_rate": 8e-05,
      "loss": 1.8979,
      "step": 1921
    },
    {
      "epoch": 0.4227427691630925,
      "grad_norm": 0.2849038541316986,
      "learning_rate": 8e-05,
      "loss": 1.6891,
      "step": 1922
    },
    {
      "epoch": 0.42296271857472784,
      "grad_norm": 0.278728723526001,
      "learning_rate": 8e-05,
      "loss": 1.7891,
      "step": 1923
    },
    {
      "epoch": 0.4231826679863631,
      "grad_norm": 0.27032172679901123,
      "learning_rate": 8e-05,
      "loss": 1.6963,
      "step": 1924
    },
    {
      "epoch": 0.42340261739799845,
      "grad_norm": 0.2731832265853882,
      "learning_rate": 8e-05,
      "loss": 1.632,
      "step": 1925
    },
    {
      "epoch": 0.4236225668096338,
      "grad_norm": 0.30378425121307373,
      "learning_rate": 8e-05,
      "loss": 1.7823,
      "step": 1926
    },
    {
      "epoch": 0.4238425162212691,
      "grad_norm": 0.27693971991539,
      "learning_rate": 8e-05,
      "loss": 1.483,
      "step": 1927
    },
    {
      "epoch": 0.42406246563290445,
      "grad_norm": 0.2719477415084839,
      "learning_rate": 8e-05,
      "loss": 1.6708,
      "step": 1928
    },
    {
      "epoch": 0.42428241504453973,
      "grad_norm": 0.26625335216522217,
      "learning_rate": 8e-05,
      "loss": 1.4946,
      "step": 1929
    },
    {
      "epoch": 0.42450236445617506,
      "grad_norm": 0.2843473553657532,
      "learning_rate": 8e-05,
      "loss": 1.722,
      "step": 1930
    },
    {
      "epoch": 0.4247223138678104,
      "grad_norm": 0.3453083336353302,
      "learning_rate": 8e-05,
      "loss": 1.7238,
      "step": 1931
    },
    {
      "epoch": 0.42494226327944573,
      "grad_norm": 0.25626078248023987,
      "learning_rate": 8e-05,
      "loss": 1.4706,
      "step": 1932
    },
    {
      "epoch": 0.42516221269108107,
      "grad_norm": 0.2908123731613159,
      "learning_rate": 8e-05,
      "loss": 1.7105,
      "step": 1933
    },
    {
      "epoch": 0.4253821621027164,
      "grad_norm": 0.33517104387283325,
      "learning_rate": 8e-05,
      "loss": 1.8023,
      "step": 1934
    },
    {
      "epoch": 0.4256021115143517,
      "grad_norm": 0.28047069907188416,
      "learning_rate": 8e-05,
      "loss": 1.6266,
      "step": 1935
    },
    {
      "epoch": 0.425822060925987,
      "grad_norm": 0.2778942584991455,
      "learning_rate": 8e-05,
      "loss": 1.6866,
      "step": 1936
    },
    {
      "epoch": 0.42604201033762235,
      "grad_norm": 0.3038877248764038,
      "learning_rate": 8e-05,
      "loss": 1.6075,
      "step": 1937
    },
    {
      "epoch": 0.4262619597492577,
      "grad_norm": 0.2814297378063202,
      "learning_rate": 8e-05,
      "loss": 1.5939,
      "step": 1938
    },
    {
      "epoch": 0.426481909160893,
      "grad_norm": 0.27854403853416443,
      "learning_rate": 8e-05,
      "loss": 1.5943,
      "step": 1939
    },
    {
      "epoch": 0.4267018585725283,
      "grad_norm": 0.2924019694328308,
      "learning_rate": 8e-05,
      "loss": 1.8193,
      "step": 1940
    },
    {
      "epoch": 0.4269218079841636,
      "grad_norm": 0.2862766683101654,
      "learning_rate": 8e-05,
      "loss": 1.6065,
      "step": 1941
    },
    {
      "epoch": 0.42714175739579896,
      "grad_norm": 0.2696346342563629,
      "learning_rate": 8e-05,
      "loss": 1.5343,
      "step": 1942
    },
    {
      "epoch": 0.4273617068074343,
      "grad_norm": 0.2578338384628296,
      "learning_rate": 8e-05,
      "loss": 1.5786,
      "step": 1943
    },
    {
      "epoch": 0.42758165621906963,
      "grad_norm": 0.28594937920570374,
      "learning_rate": 8e-05,
      "loss": 1.725,
      "step": 1944
    },
    {
      "epoch": 0.42780160563070496,
      "grad_norm": 0.2808282971382141,
      "learning_rate": 8e-05,
      "loss": 1.7951,
      "step": 1945
    },
    {
      "epoch": 0.42802155504234024,
      "grad_norm": 0.32533401250839233,
      "learning_rate": 8e-05,
      "loss": 1.9645,
      "step": 1946
    },
    {
      "epoch": 0.4282415044539756,
      "grad_norm": 0.2737642228603363,
      "learning_rate": 8e-05,
      "loss": 1.6243,
      "step": 1947
    },
    {
      "epoch": 0.4284614538656109,
      "grad_norm": 0.2885657250881195,
      "learning_rate": 8e-05,
      "loss": 1.7338,
      "step": 1948
    },
    {
      "epoch": 0.42868140327724624,
      "grad_norm": 0.2788100242614746,
      "learning_rate": 8e-05,
      "loss": 1.76,
      "step": 1949
    },
    {
      "epoch": 0.4289013526888816,
      "grad_norm": 0.2899073362350464,
      "learning_rate": 8e-05,
      "loss": 1.7739,
      "step": 1950
    },
    {
      "epoch": 0.42912130210051685,
      "grad_norm": 0.2874782681465149,
      "learning_rate": 8e-05,
      "loss": 1.8283,
      "step": 1951
    },
    {
      "epoch": 0.4293412515121522,
      "grad_norm": 0.2757413685321808,
      "learning_rate": 8e-05,
      "loss": 1.641,
      "step": 1952
    },
    {
      "epoch": 0.4295612009237875,
      "grad_norm": 0.2811121940612793,
      "learning_rate": 8e-05,
      "loss": 1.7231,
      "step": 1953
    },
    {
      "epoch": 0.42978115033542286,
      "grad_norm": 0.3400493860244751,
      "learning_rate": 8e-05,
      "loss": 1.8431,
      "step": 1954
    },
    {
      "epoch": 0.4300010997470582,
      "grad_norm": 0.29006627202033997,
      "learning_rate": 8e-05,
      "loss": 1.7438,
      "step": 1955
    },
    {
      "epoch": 0.4302210491586935,
      "grad_norm": 0.30233392119407654,
      "learning_rate": 8e-05,
      "loss": 1.6603,
      "step": 1956
    },
    {
      "epoch": 0.4304409985703288,
      "grad_norm": 0.2921263873577118,
      "learning_rate": 8e-05,
      "loss": 1.6604,
      "step": 1957
    },
    {
      "epoch": 0.43066094798196414,
      "grad_norm": 0.27695250511169434,
      "learning_rate": 8e-05,
      "loss": 1.7536,
      "step": 1958
    },
    {
      "epoch": 0.43088089739359947,
      "grad_norm": 0.2827337980270386,
      "learning_rate": 8e-05,
      "loss": 1.6324,
      "step": 1959
    },
    {
      "epoch": 0.4311008468052348,
      "grad_norm": 0.27993375062942505,
      "learning_rate": 8e-05,
      "loss": 1.7168,
      "step": 1960
    },
    {
      "epoch": 0.43132079621687014,
      "grad_norm": 0.2801220417022705,
      "learning_rate": 8e-05,
      "loss": 1.705,
      "step": 1961
    },
    {
      "epoch": 0.4315407456285054,
      "grad_norm": 0.27520567178726196,
      "learning_rate": 8e-05,
      "loss": 1.664,
      "step": 1962
    },
    {
      "epoch": 0.43176069504014075,
      "grad_norm": 0.26910632848739624,
      "learning_rate": 8e-05,
      "loss": 1.3616,
      "step": 1963
    },
    {
      "epoch": 0.4319806444517761,
      "grad_norm": 0.27770352363586426,
      "learning_rate": 8e-05,
      "loss": 1.6689,
      "step": 1964
    },
    {
      "epoch": 0.4322005938634114,
      "grad_norm": 0.27606719732284546,
      "learning_rate": 8e-05,
      "loss": 1.6644,
      "step": 1965
    },
    {
      "epoch": 0.43242054327504675,
      "grad_norm": 0.27787330746650696,
      "learning_rate": 8e-05,
      "loss": 1.8854,
      "step": 1966
    },
    {
      "epoch": 0.4326404926866821,
      "grad_norm": 0.26479870080947876,
      "learning_rate": 8e-05,
      "loss": 1.5904,
      "step": 1967
    },
    {
      "epoch": 0.43286044209831737,
      "grad_norm": 0.27598053216934204,
      "learning_rate": 8e-05,
      "loss": 1.5666,
      "step": 1968
    },
    {
      "epoch": 0.4330803915099527,
      "grad_norm": 0.27461937069892883,
      "learning_rate": 8e-05,
      "loss": 1.5487,
      "step": 1969
    },
    {
      "epoch": 0.43330034092158803,
      "grad_norm": 0.2928270399570465,
      "learning_rate": 8e-05,
      "loss": 1.8173,
      "step": 1970
    },
    {
      "epoch": 0.43352029033322337,
      "grad_norm": 0.30754199624061584,
      "learning_rate": 8e-05,
      "loss": 1.6762,
      "step": 1971
    },
    {
      "epoch": 0.4337402397448587,
      "grad_norm": 0.2676936089992523,
      "learning_rate": 8e-05,
      "loss": 1.7314,
      "step": 1972
    },
    {
      "epoch": 0.433960189156494,
      "grad_norm": 0.2919710576534271,
      "learning_rate": 8e-05,
      "loss": 1.8586,
      "step": 1973
    },
    {
      "epoch": 0.4341801385681293,
      "grad_norm": 0.28165963292121887,
      "learning_rate": 8e-05,
      "loss": 1.7943,
      "step": 1974
    },
    {
      "epoch": 0.43440008797976465,
      "grad_norm": 0.2700537443161011,
      "learning_rate": 8e-05,
      "loss": 1.613,
      "step": 1975
    },
    {
      "epoch": 0.4346200373914,
      "grad_norm": 0.26830658316612244,
      "learning_rate": 8e-05,
      "loss": 1.5854,
      "step": 1976
    },
    {
      "epoch": 0.4348399868030353,
      "grad_norm": 0.28799256682395935,
      "learning_rate": 8e-05,
      "loss": 1.8246,
      "step": 1977
    },
    {
      "epoch": 0.43505993621467065,
      "grad_norm": 0.27226150035858154,
      "learning_rate": 8e-05,
      "loss": 1.6252,
      "step": 1978
    },
    {
      "epoch": 0.43527988562630593,
      "grad_norm": 0.2646162807941437,
      "learning_rate": 8e-05,
      "loss": 1.5699,
      "step": 1979
    },
    {
      "epoch": 0.43549983503794126,
      "grad_norm": 0.27331140637397766,
      "learning_rate": 8e-05,
      "loss": 1.6893,
      "step": 1980
    },
    {
      "epoch": 0.4357197844495766,
      "grad_norm": 0.26996269822120667,
      "learning_rate": 8e-05,
      "loss": 1.6004,
      "step": 1981
    },
    {
      "epoch": 0.43593973386121193,
      "grad_norm": 0.29484307765960693,
      "learning_rate": 8e-05,
      "loss": 1.551,
      "step": 1982
    },
    {
      "epoch": 0.43615968327284727,
      "grad_norm": 0.28224268555641174,
      "learning_rate": 8e-05,
      "loss": 1.6898,
      "step": 1983
    },
    {
      "epoch": 0.43637963268448254,
      "grad_norm": 0.26172178983688354,
      "learning_rate": 8e-05,
      "loss": 1.4375,
      "step": 1984
    },
    {
      "epoch": 0.4365995820961179,
      "grad_norm": 0.2603735029697418,
      "learning_rate": 8e-05,
      "loss": 1.4528,
      "step": 1985
    },
    {
      "epoch": 0.4368195315077532,
      "grad_norm": 0.30643707513809204,
      "learning_rate": 8e-05,
      "loss": 1.688,
      "step": 1986
    },
    {
      "epoch": 0.43703948091938855,
      "grad_norm": 0.2951216995716095,
      "learning_rate": 8e-05,
      "loss": 1.7769,
      "step": 1987
    },
    {
      "epoch": 0.4372594303310239,
      "grad_norm": 0.2939329445362091,
      "learning_rate": 8e-05,
      "loss": 1.8161,
      "step": 1988
    },
    {
      "epoch": 0.4374793797426592,
      "grad_norm": 0.27539846301078796,
      "learning_rate": 8e-05,
      "loss": 1.6019,
      "step": 1989
    },
    {
      "epoch": 0.4376993291542945,
      "grad_norm": 0.2770693898200989,
      "learning_rate": 8e-05,
      "loss": 1.5972,
      "step": 1990
    },
    {
      "epoch": 0.4379192785659298,
      "grad_norm": 0.2832552492618561,
      "learning_rate": 8e-05,
      "loss": 1.7467,
      "step": 1991
    },
    {
      "epoch": 0.43813922797756516,
      "grad_norm": 0.2983148992061615,
      "learning_rate": 8e-05,
      "loss": 1.7181,
      "step": 1992
    },
    {
      "epoch": 0.4383591773892005,
      "grad_norm": 0.2829340994358063,
      "learning_rate": 8e-05,
      "loss": 1.5984,
      "step": 1993
    },
    {
      "epoch": 0.43857912680083583,
      "grad_norm": 0.2857687473297119,
      "learning_rate": 8e-05,
      "loss": 1.6471,
      "step": 1994
    },
    {
      "epoch": 0.4387990762124711,
      "grad_norm": 0.2669824957847595,
      "learning_rate": 8e-05,
      "loss": 1.6215,
      "step": 1995
    },
    {
      "epoch": 0.43901902562410644,
      "grad_norm": 0.28832894563674927,
      "learning_rate": 8e-05,
      "loss": 1.6884,
      "step": 1996
    },
    {
      "epoch": 0.4392389750357418,
      "grad_norm": 0.2919970154762268,
      "learning_rate": 8e-05,
      "loss": 1.7462,
      "step": 1997
    },
    {
      "epoch": 0.4394589244473771,
      "grad_norm": 0.2998509109020233,
      "learning_rate": 8e-05,
      "loss": 1.7219,
      "step": 1998
    },
    {
      "epoch": 0.43967887385901244,
      "grad_norm": 0.2780647575855255,
      "learning_rate": 8e-05,
      "loss": 1.8219,
      "step": 1999
    },
    {
      "epoch": 0.4398988232706478,
      "grad_norm": 0.2833268940448761,
      "learning_rate": 8e-05,
      "loss": 1.6873,
      "step": 2000
    }
  ],
  "logging_steps": 1,
  "max_steps": 4546,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.77128394686464e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}