{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 884,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004527447651386531,
      "grad_norm": 9.912898659097554,
      "learning_rate": 0.0,
      "loss": 0.704662561416626,
      "step": 1
    },
    {
      "epoch": 0.009054895302773062,
      "grad_norm": 11.570635762100624,
      "learning_rate": 1.1235955056179775e-06,
      "loss": 0.7413114905357361,
      "step": 2
    },
    {
      "epoch": 0.013582342954159592,
      "grad_norm": 11.534539965780745,
      "learning_rate": 2.247191011235955e-06,
      "loss": 0.7407010197639465,
      "step": 3
    },
    {
      "epoch": 0.018109790605546124,
      "grad_norm": 11.197515354711427,
      "learning_rate": 3.3707865168539327e-06,
      "loss": 0.763998806476593,
      "step": 4
    },
    {
      "epoch": 0.022637238256932653,
      "grad_norm": 8.440914434182256,
      "learning_rate": 4.49438202247191e-06,
      "loss": 0.7183060646057129,
      "step": 5
    },
    {
      "epoch": 0.027164685908319185,
      "grad_norm": 5.699585146399442,
      "learning_rate": 5.617977528089888e-06,
      "loss": 0.638984203338623,
      "step": 6
    },
    {
      "epoch": 0.03169213355970572,
      "grad_norm": 1.8448435550812394,
      "learning_rate": 6.741573033707865e-06,
      "loss": 0.48158273100852966,
      "step": 7
    },
    {
      "epoch": 0.03621958121109225,
      "grad_norm": 1.6420510885571051,
      "learning_rate": 7.865168539325843e-06,
      "loss": 0.5159077048301697,
      "step": 8
    },
    {
      "epoch": 0.04074702886247878,
      "grad_norm": 1.0582363331837332,
      "learning_rate": 8.98876404494382e-06,
      "loss": 0.4606141448020935,
      "step": 9
    },
    {
      "epoch": 0.045274476513865305,
      "grad_norm": 1.3855426484013416,
      "learning_rate": 1.0112359550561798e-05,
      "loss": 0.4743913412094116,
      "step": 10
    },
    {
      "epoch": 0.04980192416525184,
      "grad_norm": 0.9317004579840986,
      "learning_rate": 1.1235955056179776e-05,
      "loss": 0.4422619342803955,
      "step": 11
    },
    {
      "epoch": 0.05432937181663837,
      "grad_norm": 0.7778790634981517,
      "learning_rate": 1.2359550561797752e-05,
      "loss": 0.4157835841178894,
      "step": 12
    },
    {
      "epoch": 0.0588568194680249,
      "grad_norm": 0.7973555323801547,
      "learning_rate": 1.348314606741573e-05,
      "loss": 0.38240519165992737,
      "step": 13
    },
    {
      "epoch": 0.06338426711941143,
      "grad_norm": 0.6808969539392029,
      "learning_rate": 1.4606741573033709e-05,
      "loss": 0.3956570029258728,
      "step": 14
    },
    {
      "epoch": 0.06791171477079797,
      "grad_norm": 0.6476893513084165,
      "learning_rate": 1.5730337078651687e-05,
      "loss": 0.38245508074760437,
      "step": 15
    },
    {
      "epoch": 0.0724391624221845,
      "grad_norm": 0.5574203855090996,
      "learning_rate": 1.6853932584269665e-05,
      "loss": 0.3645451068878174,
      "step": 16
    },
    {
      "epoch": 0.07696661007357103,
      "grad_norm": 0.556142105956913,
      "learning_rate": 1.797752808988764e-05,
      "loss": 0.37906813621520996,
      "step": 17
    },
    {
      "epoch": 0.08149405772495756,
      "grad_norm": 0.5390987389446441,
      "learning_rate": 1.9101123595505618e-05,
      "loss": 0.35216137766838074,
      "step": 18
    },
    {
      "epoch": 0.08602150537634409,
      "grad_norm": 0.42363300643952606,
      "learning_rate": 2.0224719101123596e-05,
      "loss": 0.32894521951675415,
      "step": 19
    },
    {
      "epoch": 0.09054895302773061,
      "grad_norm": 0.4504410413461609,
      "learning_rate": 2.1348314606741574e-05,
      "loss": 0.3588491380214691,
      "step": 20
    },
    {
      "epoch": 0.09507640067911714,
      "grad_norm": 0.4248016680645426,
      "learning_rate": 2.2471910112359552e-05,
      "loss": 0.35492587089538574,
      "step": 21
    },
    {
      "epoch": 0.09960384833050367,
      "grad_norm": 0.45292204165027866,
      "learning_rate": 2.359550561797753e-05,
      "loss": 0.3395863175392151,
      "step": 22
    },
    {
      "epoch": 0.1041312959818902,
      "grad_norm": 0.4127018468390898,
      "learning_rate": 2.4719101123595505e-05,
      "loss": 0.3350675702095032,
      "step": 23
    },
    {
      "epoch": 0.10865874363327674,
      "grad_norm": 0.41087243703557136,
      "learning_rate": 2.5842696629213486e-05,
      "loss": 0.32198527455329895,
      "step": 24
    },
    {
      "epoch": 0.11318619128466327,
      "grad_norm": 0.4141054386106995,
      "learning_rate": 2.696629213483146e-05,
      "loss": 0.31030064821243286,
      "step": 25
    },
    {
      "epoch": 0.1177136389360498,
      "grad_norm": 0.4007754482311067,
      "learning_rate": 2.8089887640449443e-05,
      "loss": 0.3337652087211609,
      "step": 26
    },
    {
      "epoch": 0.12224108658743633,
      "grad_norm": 0.3445113938247397,
      "learning_rate": 2.9213483146067417e-05,
      "loss": 0.3164421319961548,
      "step": 27
    },
    {
      "epoch": 0.12676853423882287,
      "grad_norm": 0.4165236729788368,
      "learning_rate": 3.0337078651685396e-05,
      "loss": 0.3188047409057617,
      "step": 28
    },
    {
      "epoch": 0.13129598189020938,
      "grad_norm": 0.35793264806538255,
      "learning_rate": 3.1460674157303374e-05,
      "loss": 0.3220544457435608,
      "step": 29
    },
    {
      "epoch": 0.13582342954159593,
      "grad_norm": 0.3693389074810848,
      "learning_rate": 3.258426966292135e-05,
      "loss": 0.2884308099746704,
      "step": 30
    },
    {
      "epoch": 0.14035087719298245,
      "grad_norm": 0.3693829230677698,
      "learning_rate": 3.370786516853933e-05,
      "loss": 0.28860390186309814,
      "step": 31
    },
    {
      "epoch": 0.144878324844369,
      "grad_norm": 0.2871781736218002,
      "learning_rate": 3.483146067415731e-05,
      "loss": 0.29432493448257446,
      "step": 32
    },
    {
      "epoch": 0.1494057724957555,
      "grad_norm": 0.34659072254483303,
      "learning_rate": 3.595505617977528e-05,
      "loss": 0.29653051495552063,
      "step": 33
    },
    {
      "epoch": 0.15393322014714206,
      "grad_norm": 0.3353984969007762,
      "learning_rate": 3.7078651685393264e-05,
      "loss": 0.31739920377731323,
      "step": 34
    },
    {
      "epoch": 0.15846066779852858,
      "grad_norm": 0.30470773813342095,
      "learning_rate": 3.8202247191011236e-05,
      "loss": 0.2955687344074249,
      "step": 35
    },
    {
      "epoch": 0.16298811544991512,
      "grad_norm": 0.3365610498312007,
      "learning_rate": 3.9325842696629214e-05,
      "loss": 0.28539973497390747,
      "step": 36
    },
    {
      "epoch": 0.16751556310130164,
      "grad_norm": 0.3630734236817004,
      "learning_rate": 4.044943820224719e-05,
      "loss": 0.26939618587493896,
      "step": 37
    },
    {
      "epoch": 0.17204301075268819,
      "grad_norm": 0.30664574697744273,
      "learning_rate": 4.157303370786517e-05,
      "loss": 0.2930394113063812,
      "step": 38
    },
    {
      "epoch": 0.1765704584040747,
      "grad_norm": 0.3172929701625085,
      "learning_rate": 4.269662921348315e-05,
      "loss": 0.2929205000400543,
      "step": 39
    },
    {
      "epoch": 0.18109790605546122,
      "grad_norm": 0.30282779948812566,
      "learning_rate": 4.3820224719101126e-05,
      "loss": 0.28868985176086426,
      "step": 40
    },
    {
      "epoch": 0.18562535370684777,
      "grad_norm": 0.36387145692382533,
      "learning_rate": 4.4943820224719104e-05,
      "loss": 0.2901768386363983,
      "step": 41
    },
    {
      "epoch": 0.19015280135823429,
      "grad_norm": 0.32552743560482894,
      "learning_rate": 4.606741573033708e-05,
      "loss": 0.2603440284729004,
      "step": 42
    },
    {
      "epoch": 0.19468024900962083,
      "grad_norm": 0.32257099007122275,
      "learning_rate": 4.719101123595506e-05,
      "loss": 0.2538576126098633,
      "step": 43
    },
    {
      "epoch": 0.19920769666100735,
      "grad_norm": 0.30260575607567747,
      "learning_rate": 4.831460674157304e-05,
      "loss": 0.270669549703598,
      "step": 44
    },
    {
      "epoch": 0.2037351443123939,
      "grad_norm": 0.29630104759035064,
      "learning_rate": 4.943820224719101e-05,
      "loss": 0.299394428730011,
      "step": 45
    },
    {
      "epoch": 0.2082625919637804,
      "grad_norm": 0.3143547122724143,
      "learning_rate": 5.0561797752808995e-05,
      "loss": 0.27803701162338257,
      "step": 46
    },
    {
      "epoch": 0.21279003961516696,
      "grad_norm": 0.3069473445693767,
      "learning_rate": 5.168539325842697e-05,
      "loss": 0.2667008936405182,
      "step": 47
    },
    {
      "epoch": 0.21731748726655348,
      "grad_norm": 0.32098032594841497,
      "learning_rate": 5.2808988764044944e-05,
      "loss": 0.25864067673683167,
      "step": 48
    },
    {
      "epoch": 0.22184493491794002,
      "grad_norm": 0.3555508737165522,
      "learning_rate": 5.393258426966292e-05,
      "loss": 0.25849252939224243,
      "step": 49
    },
    {
      "epoch": 0.22637238256932654,
      "grad_norm": 0.3557992878598279,
      "learning_rate": 5.50561797752809e-05,
      "loss": 0.25682830810546875,
      "step": 50
    },
    {
      "epoch": 0.23089983022071306,
      "grad_norm": 0.3302326973115044,
      "learning_rate": 5.6179775280898885e-05,
      "loss": 0.2713906168937683,
      "step": 51
    },
    {
      "epoch": 0.2354272778720996,
      "grad_norm": 0.3203937826524574,
      "learning_rate": 5.730337078651685e-05,
      "loss": 0.2719656527042389,
      "step": 52
    },
    {
      "epoch": 0.23995472552348612,
      "grad_norm": 0.34684564119613337,
      "learning_rate": 5.8426966292134835e-05,
      "loss": 0.26817768812179565,
      "step": 53
    },
    {
      "epoch": 0.24448217317487267,
      "grad_norm": 0.43126929639643324,
      "learning_rate": 5.955056179775281e-05,
      "loss": 0.26956790685653687,
      "step": 54
    },
    {
      "epoch": 0.2490096208262592,
      "grad_norm": 0.3867030193280389,
      "learning_rate": 6.067415730337079e-05,
      "loss": 0.2570400834083557,
      "step": 55
    },
    {
      "epoch": 0.25353706847764573,
      "grad_norm": 0.4120938739335183,
      "learning_rate": 6.179775280898876e-05,
      "loss": 0.2433938980102539,
      "step": 56
    },
    {
      "epoch": 0.25806451612903225,
      "grad_norm": 0.3586156961196402,
      "learning_rate": 6.292134831460675e-05,
      "loss": 0.28030940890312195,
      "step": 57
    },
    {
      "epoch": 0.26259196378041877,
      "grad_norm": 0.3666232259035239,
      "learning_rate": 6.404494382022472e-05,
      "loss": 0.2750133275985718,
      "step": 58
    },
    {
      "epoch": 0.26711941143180534,
      "grad_norm": 0.3742199995843273,
      "learning_rate": 6.51685393258427e-05,
      "loss": 0.26811373233795166,
      "step": 59
    },
    {
      "epoch": 0.27164685908319186,
      "grad_norm": 0.340295805859951,
      "learning_rate": 6.629213483146067e-05,
      "loss": 0.26556718349456787,
      "step": 60
    },
    {
      "epoch": 0.2761743067345784,
      "grad_norm": 0.3910827471892104,
      "learning_rate": 6.741573033707866e-05,
      "loss": 0.26536938548088074,
      "step": 61
    },
    {
      "epoch": 0.2807017543859649,
      "grad_norm": 0.4013017063957163,
      "learning_rate": 6.853932584269663e-05,
      "loss": 0.24445080757141113,
      "step": 62
    },
    {
      "epoch": 0.28522920203735147,
      "grad_norm": 0.37310303240899334,
      "learning_rate": 6.966292134831462e-05,
      "loss": 0.2769097685813904,
      "step": 63
    },
    {
      "epoch": 0.289756649688738,
      "grad_norm": 0.41547291584982365,
      "learning_rate": 7.078651685393259e-05,
      "loss": 0.27888572216033936,
      "step": 64
    },
    {
      "epoch": 0.2942840973401245,
      "grad_norm": 0.3978298563806116,
      "learning_rate": 7.191011235955056e-05,
      "loss": 0.25978708267211914,
      "step": 65
    },
    {
      "epoch": 0.298811544991511,
      "grad_norm": 0.34304353624930783,
      "learning_rate": 7.303370786516854e-05,
      "loss": 0.2618892788887024,
      "step": 66
    },
    {
      "epoch": 0.30333899264289754,
      "grad_norm": 0.3568235886758767,
      "learning_rate": 7.415730337078653e-05,
      "loss": 0.27624228596687317,
      "step": 67
    },
    {
      "epoch": 0.3078664402942841,
      "grad_norm": 0.4068129432872794,
      "learning_rate": 7.52808988764045e-05,
      "loss": 0.26165643334388733,
      "step": 68
    },
    {
      "epoch": 0.31239388794567063,
      "grad_norm": 0.35618727820338963,
      "learning_rate": 7.640449438202247e-05,
      "loss": 0.23117734491825104,
      "step": 69
    },
    {
      "epoch": 0.31692133559705715,
      "grad_norm": 0.29522791383335695,
      "learning_rate": 7.752808988764046e-05,
      "loss": 0.26893192529678345,
      "step": 70
    },
    {
      "epoch": 0.32144878324844367,
      "grad_norm": 0.34184780027859696,
      "learning_rate": 7.865168539325843e-05,
      "loss": 0.265208899974823,
      "step": 71
    },
    {
      "epoch": 0.32597623089983024,
      "grad_norm": 0.36453052433385097,
      "learning_rate": 7.97752808988764e-05,
      "loss": 0.26777154207229614,
      "step": 72
    },
    {
      "epoch": 0.33050367855121676,
      "grad_norm": 0.31626063151863026,
      "learning_rate": 8.089887640449438e-05,
      "loss": 0.25969862937927246,
      "step": 73
    },
    {
      "epoch": 0.3350311262026033,
      "grad_norm": 0.36078791784625247,
      "learning_rate": 8.202247191011237e-05,
      "loss": 0.252025842666626,
      "step": 74
    },
    {
      "epoch": 0.3395585738539898,
      "grad_norm": 0.36592820199700676,
      "learning_rate": 8.314606741573034e-05,
      "loss": 0.23486942052841187,
      "step": 75
    },
    {
      "epoch": 0.34408602150537637,
      "grad_norm": 0.37291138339143015,
      "learning_rate": 8.426966292134831e-05,
      "loss": 0.2732143998146057,
      "step": 76
    },
    {
      "epoch": 0.3486134691567629,
      "grad_norm": 0.40292620977238186,
      "learning_rate": 8.53932584269663e-05,
      "loss": 0.27056077122688293,
      "step": 77
    },
    {
      "epoch": 0.3531409168081494,
      "grad_norm": 0.3178919083590533,
      "learning_rate": 8.651685393258427e-05,
      "loss": 0.24804702401161194,
      "step": 78
    },
    {
      "epoch": 0.3576683644595359,
      "grad_norm": 0.3824213813807792,
      "learning_rate": 8.764044943820225e-05,
      "loss": 0.2524891197681427,
      "step": 79
    },
    {
      "epoch": 0.36219581211092244,
      "grad_norm": 0.39732669158813094,
      "learning_rate": 8.876404494382022e-05,
      "loss": 0.25263312458992004,
      "step": 80
    },
    {
      "epoch": 0.366723259762309,
      "grad_norm": 0.40167655171657085,
      "learning_rate": 8.988764044943821e-05,
      "loss": 0.2401430904865265,
      "step": 81
    },
    {
      "epoch": 0.37125070741369554,
      "grad_norm": 0.4156895140870112,
      "learning_rate": 9.101123595505618e-05,
      "loss": 0.2767913043498993,
      "step": 82
    },
    {
      "epoch": 0.37577815506508205,
      "grad_norm": 0.36470302185591663,
      "learning_rate": 9.213483146067416e-05,
      "loss": 0.27237945795059204,
      "step": 83
    },
    {
      "epoch": 0.38030560271646857,
      "grad_norm": 0.33621491342371873,
      "learning_rate": 9.325842696629214e-05,
      "loss": 0.26383328437805176,
      "step": 84
    },
    {
      "epoch": 0.38483305036785515,
      "grad_norm": 0.41752007008704656,
      "learning_rate": 9.438202247191012e-05,
      "loss": 0.2715482711791992,
      "step": 85
    },
    {
      "epoch": 0.38936049801924166,
      "grad_norm": 0.3810992799422323,
      "learning_rate": 9.550561797752809e-05,
      "loss": 0.2509955167770386,
      "step": 86
    },
    {
      "epoch": 0.3938879456706282,
      "grad_norm": 0.37395173174434054,
      "learning_rate": 9.662921348314608e-05,
      "loss": 0.2497263103723526,
      "step": 87
    },
    {
      "epoch": 0.3984153933220147,
      "grad_norm": 0.41548687649152877,
      "learning_rate": 9.775280898876405e-05,
      "loss": 0.2719443738460541,
      "step": 88
    },
    {
      "epoch": 0.4029428409734012,
      "grad_norm": 0.3589498353469027,
      "learning_rate": 9.887640449438202e-05,
      "loss": 0.26369816064834595,
      "step": 89
    },
    {
      "epoch": 0.4074702886247878,
      "grad_norm": 0.33451587322094073,
      "learning_rate": 0.0001,
      "loss": 0.2591651976108551,
      "step": 90
    },
    {
      "epoch": 0.4119977362761743,
      "grad_norm": 0.3629069875769648,
      "learning_rate": 9.999960960438442e-05,
      "loss": 0.257312148809433,
      "step": 91
    },
    {
      "epoch": 0.4165251839275608,
      "grad_norm": 0.3840755219827805,
      "learning_rate": 9.999843842363401e-05,
      "loss": 0.2674788236618042,
      "step": 92
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 0.3303447013720405,
      "learning_rate": 9.999648647603774e-05,
      "loss": 0.24443332850933075,
      "step": 93
    },
    {
      "epoch": 0.4255800792303339,
      "grad_norm": 0.31673300660852427,
      "learning_rate": 9.99937537920769e-05,
      "loss": 0.2551422417163849,
      "step": 94
    },
    {
      "epoch": 0.43010752688172044,
      "grad_norm": 0.30992880019470387,
      "learning_rate": 9.999024041442456e-05,
      "loss": 0.2655659317970276,
      "step": 95
    },
    {
      "epoch": 0.43463497453310695,
      "grad_norm": 0.3361276340451737,
      "learning_rate": 9.998594639794501e-05,
      "loss": 0.2640272378921509,
      "step": 96
    },
    {
      "epoch": 0.4391624221844935,
      "grad_norm": 0.31861601679153306,
      "learning_rate": 9.998087180969289e-05,
      "loss": 0.2508934438228607,
      "step": 97
    },
    {
      "epoch": 0.44368986983588005,
      "grad_norm": 0.3141874772255754,
      "learning_rate": 9.997501672891207e-05,
      "loss": 0.2459731549024582,
      "step": 98
    },
    {
      "epoch": 0.44821731748726656,
      "grad_norm": 0.3417995555895257,
      "learning_rate": 9.996838124703447e-05,
      "loss": 0.2545221447944641,
      "step": 99
    },
    {
      "epoch": 0.4527447651386531,
      "grad_norm": 0.32772554154297434,
      "learning_rate": 9.99609654676786e-05,
      "loss": 0.23231828212738037,
      "step": 100
    },
    {
      "epoch": 0.4572722127900396,
      "grad_norm": 0.31390097725170435,
      "learning_rate": 9.995276950664796e-05,
      "loss": 0.26731887459754944,
      "step": 101
    },
    {
      "epoch": 0.4617996604414261,
      "grad_norm": 0.36839050365963416,
      "learning_rate": 9.994379349192926e-05,
      "loss": 0.2589717507362366,
      "step": 102
    },
    {
      "epoch": 0.4663271080928127,
      "grad_norm": 0.3230085260059877,
      "learning_rate": 9.993403756369038e-05,
      "loss": 0.2471804916858673,
      "step": 103
    },
    {
      "epoch": 0.4708545557441992,
      "grad_norm": 0.3527635871153749,
      "learning_rate": 9.992350187427815e-05,
      "loss": 0.2583211660385132,
      "step": 104
    },
    {
      "epoch": 0.47538200339558573,
      "grad_norm": 0.3257945006410995,
      "learning_rate": 9.991218658821608e-05,
      "loss": 0.23961997032165527,
      "step": 105
    },
    {
      "epoch": 0.47990945104697225,
      "grad_norm": 0.320334711261071,
      "learning_rate": 9.990009188220167e-05,
      "loss": 0.2344263792037964,
      "step": 106
    },
    {
      "epoch": 0.4844368986983588,
      "grad_norm": 0.2826225136540489,
      "learning_rate": 9.988721794510374e-05,
      "loss": 0.24988439679145813,
      "step": 107
    },
    {
      "epoch": 0.48896434634974534,
      "grad_norm": 0.28247379517350096,
      "learning_rate": 9.987356497795943e-05,
      "loss": 0.26465803384780884,
      "step": 108
    },
    {
      "epoch": 0.49349179400113186,
      "grad_norm": 0.30751957231400473,
      "learning_rate": 9.985913319397109e-05,
      "loss": 0.26271963119506836,
      "step": 109
    },
    {
      "epoch": 0.4980192416525184,
      "grad_norm": 0.29507363950644233,
      "learning_rate": 9.984392281850293e-05,
      "loss": 0.25772029161453247,
      "step": 110
    },
    {
      "epoch": 0.5025466893039049,
      "grad_norm": 0.3177993118197023,
      "learning_rate": 9.982793408907747e-05,
      "loss": 0.24420583248138428,
      "step": 111
    },
    {
      "epoch": 0.5070741369552915,
      "grad_norm": 0.30002396119316954,
      "learning_rate": 9.981116725537194e-05,
      "loss": 0.23018613457679749,
      "step": 112
    },
    {
      "epoch": 0.511601584606678,
      "grad_norm": 0.30591462803235053,
      "learning_rate": 9.979362257921427e-05,
      "loss": 0.25676852464675903,
      "step": 113
    },
    {
      "epoch": 0.5161290322580645,
      "grad_norm": 0.2771867071105219,
      "learning_rate": 9.977530033457905e-05,
      "loss": 0.26490455865859985,
      "step": 114
    },
    {
      "epoch": 0.5206564799094511,
      "grad_norm": 0.2695074487782023,
      "learning_rate": 9.97562008075832e-05,
      "loss": 0.26009300351142883,
      "step": 115
    },
    {
      "epoch": 0.5251839275608375,
      "grad_norm": 0.29418055761111117,
      "learning_rate": 9.973632429648165e-05,
      "loss": 0.2581276297569275,
      "step": 116
    },
    {
      "epoch": 0.5297113752122241,
      "grad_norm": 0.2797820785269771,
      "learning_rate": 9.971567111166246e-05,
      "loss": 0.24163666367530823,
      "step": 117
    },
    {
      "epoch": 0.5342388228636107,
      "grad_norm": 0.3325911505915364,
      "learning_rate": 9.969424157564215e-05,
      "loss": 0.24011248350143433,
      "step": 118
    },
    {
      "epoch": 0.5387662705149971,
      "grad_norm": 0.2844676569188647,
      "learning_rate": 9.967203602306061e-05,
      "loss": 0.23980462551116943,
      "step": 119
    },
    {
      "epoch": 0.5432937181663837,
      "grad_norm": 0.2862344840570254,
      "learning_rate": 9.964905480067586e-05,
      "loss": 0.2630372643470764,
      "step": 120
    },
    {
      "epoch": 0.5478211658177702,
      "grad_norm": 0.28341803695593515,
      "learning_rate": 9.96252982673586e-05,
      "loss": 0.25794416666030884,
      "step": 121
    },
    {
      "epoch": 0.5523486134691568,
      "grad_norm": 0.2823295105070198,
      "learning_rate": 9.960076679408674e-05,
      "loss": 0.24132010340690613,
      "step": 122
    },
    {
      "epoch": 0.5568760611205433,
      "grad_norm": 0.27570977649505757,
      "learning_rate": 9.957546076393943e-05,
      "loss": 0.23130124807357788,
      "step": 123
    },
    {
      "epoch": 0.5614035087719298,
      "grad_norm": 0.2790707776447979,
      "learning_rate": 9.954938057209121e-05,
      "loss": 0.22781060636043549,
      "step": 124
    },
    {
      "epoch": 0.5659309564233164,
      "grad_norm": 0.2939545202863887,
      "learning_rate": 9.952252662580579e-05,
      "loss": 0.2287454903125763,
      "step": 125
    },
    {
      "epoch": 0.5704584040747029,
      "grad_norm": 0.3370977301578632,
      "learning_rate": 9.949489934442966e-05,
      "loss": 0.2620600759983063,
      "step": 126
    },
    {
      "epoch": 0.5749858517260894,
      "grad_norm": 0.3152020396356069,
      "learning_rate": 9.946649915938562e-05,
      "loss": 0.2481548935174942,
      "step": 127
    },
    {
      "epoch": 0.579513299377476,
      "grad_norm": 0.2839964481543971,
      "learning_rate": 9.943732651416597e-05,
      "loss": 0.24060115218162537,
      "step": 128
    },
    {
      "epoch": 0.5840407470288624,
      "grad_norm": 0.3082572196378021,
      "learning_rate": 9.940738186432565e-05,
      "loss": 0.24669115245342255,
      "step": 129
    },
    {
      "epoch": 0.588568194680249,
      "grad_norm": 0.3175326451046873,
      "learning_rate": 9.937666567747501e-05,
      "loss": 0.232182115316391,
      "step": 130
    },
    {
      "epoch": 0.5930956423316356,
      "grad_norm": 0.2946456608292691,
      "learning_rate": 9.934517843327269e-05,
      "loss": 0.22415456175804138,
      "step": 131
    },
    {
      "epoch": 0.597623089983022,
      "grad_norm": 0.3029418466381808,
      "learning_rate": 9.931292062341793e-05,
      "loss": 0.25596803426742554,
      "step": 132
    },
    {
      "epoch": 0.6021505376344086,
      "grad_norm": 0.26570149470696636,
      "learning_rate": 9.927989275164305e-05,
      "loss": 0.25962620973587036,
      "step": 133
    },
    {
      "epoch": 0.6066779852857951,
      "grad_norm": 0.26184979344110004,
      "learning_rate": 9.924609533370551e-05,
      "loss": 0.2516033351421356,
      "step": 134
    },
    {
      "epoch": 0.6112054329371817,
      "grad_norm": 0.29194357336431614,
      "learning_rate": 9.921152889737984e-05,
      "loss": 0.2449331432580948,
      "step": 135
    },
    {
      "epoch": 0.6157328805885682,
      "grad_norm": 0.2817364203286285,
      "learning_rate": 9.917619398244949e-05,
      "loss": 0.23890982568264008,
      "step": 136
    },
    {
      "epoch": 0.6202603282399547,
      "grad_norm": 0.28221458993411364,
      "learning_rate": 9.914009114069824e-05,
      "loss": 0.2346743941307068,
      "step": 137
    },
    {
      "epoch": 0.6247877758913413,
      "grad_norm": 0.2890703412087473,
      "learning_rate": 9.910322093590177e-05,
      "loss": 0.26403021812438965,
      "step": 138
    },
    {
      "epoch": 0.6293152235427278,
      "grad_norm": 0.2775688212501171,
      "learning_rate": 9.90655839438187e-05,
      "loss": 0.2611105442047119,
      "step": 139
    },
    {
      "epoch": 0.6338426711941143,
      "grad_norm": 0.28175236393188907,
      "learning_rate": 9.902718075218176e-05,
      "loss": 0.2559444308280945,
      "step": 140
    },
    {
      "epoch": 0.6383701188455009,
      "grad_norm": 0.2808421807548938,
      "learning_rate": 9.898801196068839e-05,
      "loss": 0.24988169968128204,
      "step": 141
    },
    {
      "epoch": 0.6428975664968873,
      "grad_norm": 0.28565156352811083,
      "learning_rate": 9.89480781809916e-05,
      "loss": 0.23623761534690857,
      "step": 142
    },
    {
      "epoch": 0.6474250141482739,
      "grad_norm": 0.29055336575982305,
      "learning_rate": 9.890738003669029e-05,
      "loss": 0.23513942956924438,
      "step": 143
    },
    {
      "epoch": 0.6519524617996605,
      "grad_norm": 0.2609119776160892,
      "learning_rate": 9.886591816331954e-05,
      "loss": 0.2382441908121109,
      "step": 144
    },
    {
      "epoch": 0.656479909451047,
      "grad_norm": 0.2625283116555289,
      "learning_rate": 9.882369320834069e-05,
      "loss": 0.25793686509132385,
      "step": 145
    },
    {
      "epoch": 0.6610073571024335,
      "grad_norm": 0.2618134108702404,
      "learning_rate": 9.878070583113123e-05,
      "loss": 0.25634902715682983,
      "step": 146
    },
    {
      "epoch": 0.66553480475382,
      "grad_norm": 0.2618703653278327,
      "learning_rate": 9.87369567029745e-05,
      "loss": 0.23825976252555847,
      "step": 147
    },
    {
      "epoch": 0.6700622524052066,
      "grad_norm": 0.24549141424367749,
      "learning_rate": 9.869244650704923e-05,
      "loss": 0.23139911890029907,
      "step": 148
    },
    {
      "epoch": 0.6745897000565931,
      "grad_norm": 0.2919179784703248,
      "learning_rate": 9.864717593841883e-05,
      "loss": 0.2364065945148468,
      "step": 149
    },
    {
      "epoch": 0.6791171477079796,
      "grad_norm": 0.3068776756186207,
      "learning_rate": 9.860114570402054e-05,
      "loss": 0.22805152833461761,
      "step": 150
    },
    {
      "epoch": 0.6836445953593662,
      "grad_norm": 0.24926557146946038,
      "learning_rate": 9.855435652265446e-05,
      "loss": 0.25066664814949036,
      "step": 151
    },
    {
      "epoch": 0.6881720430107527,
      "grad_norm": 0.31210984138209535,
      "learning_rate": 9.85068091249722e-05,
      "loss": 0.2638513743877411,
      "step": 152
    },
    {
      "epoch": 0.6926994906621392,
      "grad_norm": 0.29266749441113704,
      "learning_rate": 9.845850425346563e-05,
      "loss": 0.24220243096351624,
      "step": 153
    },
    {
      "epoch": 0.6972269383135258,
      "grad_norm": 0.2700294950759208,
      "learning_rate": 9.840944266245511e-05,
      "loss": 0.22989992797374725,
      "step": 154
    },
    {
      "epoch": 0.7017543859649122,
      "grad_norm": 0.281010988179117,
      "learning_rate": 9.835962511807786e-05,
      "loss": 0.2375098466873169,
      "step": 155
    },
    {
      "epoch": 0.7062818336162988,
      "grad_norm": 0.30729269930306746,
      "learning_rate": 9.830905239827593e-05,
      "loss": 0.2361062467098236,
      "step": 156
    },
    {
      "epoch": 0.7108092812676854,
      "grad_norm": 0.25590332210065875,
      "learning_rate": 9.825772529278401e-05,
      "loss": 0.23588624596595764,
      "step": 157
    },
    {
      "epoch": 0.7153367289190719,
      "grad_norm": 0.31448280989647853,
      "learning_rate": 9.820564460311718e-05,
      "loss": 0.26085129380226135,
      "step": 158
    },
    {
      "epoch": 0.7198641765704584,
      "grad_norm": 0.26424643377935236,
      "learning_rate": 9.815281114255841e-05,
      "loss": 0.24308937788009644,
      "step": 159
    },
    {
      "epoch": 0.7243916242218449,
      "grad_norm": 0.2861457501124968,
      "learning_rate": 9.809922573614569e-05,
      "loss": 0.23594915866851807,
      "step": 160
    },
    {
      "epoch": 0.7289190718732315,
      "grad_norm": 0.27887102169366423,
      "learning_rate": 9.804488922065937e-05,
      "loss": 0.22787603735923767,
      "step": 161
    },
    {
      "epoch": 0.733446519524618,
      "grad_norm": 0.25822092011332204,
      "learning_rate": 9.798980244460893e-05,
      "loss": 0.2265864610671997,
      "step": 162
    },
    {
      "epoch": 0.7379739671760045,
      "grad_norm": 0.2671374759174116,
      "learning_rate": 9.79339662682198e-05,
      "loss": 0.24127787351608276,
      "step": 163
    },
    {
      "epoch": 0.7425014148273911,
      "grad_norm": 0.2548892719346753,
      "learning_rate": 9.787738156341992e-05,
      "loss": 0.2710321247577667,
      "step": 164
    },
    {
      "epoch": 0.7470288624787776,
      "grad_norm": 0.275831569381297,
      "learning_rate": 9.782004921382612e-05,
      "loss": 0.2556942403316498,
      "step": 165
    },
    {
      "epoch": 0.7515563101301641,
      "grad_norm": 0.269064177544399,
      "learning_rate": 9.776197011473033e-05,
      "loss": 0.235366553068161,
      "step": 166
    },
    {
      "epoch": 0.7560837577815507,
      "grad_norm": 0.26612381338020386,
      "learning_rate": 9.770314517308554e-05,
      "loss": 0.22850793600082397,
      "step": 167
    },
    {
      "epoch": 0.7606112054329371,
      "grad_norm": 0.2950948787933902,
      "learning_rate": 9.764357530749178e-05,
      "loss": 0.21261030435562134,
      "step": 168
    },
    {
      "epoch": 0.7651386530843237,
      "grad_norm": 0.25123182860694,
      "learning_rate": 9.758326144818155e-05,
      "loss": 0.2538388967514038,
      "step": 169
    },
    {
      "epoch": 0.7696661007357103,
      "grad_norm": 0.2732466231923803,
      "learning_rate": 9.752220453700556e-05,
      "loss": 0.26699525117874146,
      "step": 170
    },
    {
      "epoch": 0.7741935483870968,
      "grad_norm": 0.2782731752817353,
      "learning_rate": 9.74604055274178e-05,
      "loss": 0.24641463160514832,
      "step": 171
    },
    {
      "epoch": 0.7787209960384833,
      "grad_norm": 0.247210966114068,
      "learning_rate": 9.739786538446076e-05,
      "loss": 0.2501038908958435,
      "step": 172
    },
    {
      "epoch": 0.7832484436898698,
      "grad_norm": 0.27530294426655916,
      "learning_rate": 9.733458508475037e-05,
      "loss": 0.2417881190776825,
      "step": 173
    },
    {
      "epoch": 0.7877758913412564,
      "grad_norm": 0.2523811344572268,
      "learning_rate": 9.727056561646066e-05,
      "loss": 0.22153747081756592,
      "step": 174
    },
    {
      "epoch": 0.7923033389926429,
      "grad_norm": 0.31380982209356834,
      "learning_rate": 9.720580797930845e-05,
      "loss": 0.21761545538902283,
      "step": 175
    },
    {
      "epoch": 0.7968307866440294,
      "grad_norm": 0.2608751562910921,
      "learning_rate": 9.714031318453764e-05,
      "loss": 0.26552385091781616,
      "step": 176
    },
    {
      "epoch": 0.801358234295416,
      "grad_norm": 0.26240711982162157,
      "learning_rate": 9.707408225490344e-05,
      "loss": 0.24315059185028076,
      "step": 177
    },
    {
      "epoch": 0.8058856819468024,
      "grad_norm": 0.2492667246400354,
      "learning_rate": 9.700711622465643e-05,
      "loss": 0.23814553022384644,
      "step": 178
    },
    {
      "epoch": 0.810413129598189,
      "grad_norm": 0.2798579758688783,
      "learning_rate": 9.693941613952642e-05,
      "loss": 0.2307881861925125,
      "step": 179
    },
    {
      "epoch": 0.8149405772495756,
      "grad_norm": 0.2526797413748999,
      "learning_rate": 9.687098305670605e-05,
      "loss": 0.2314993441104889,
      "step": 180
    },
    {
      "epoch": 0.819468024900962,
      "grad_norm": 0.27342696621039475,
      "learning_rate": 9.680181804483434e-05,
      "loss": 0.21843808889389038,
      "step": 181
    },
    {
      "epoch": 0.8239954725523486,
      "grad_norm": 0.27599884141353176,
      "learning_rate": 9.673192218398e-05,
      "loss": 0.25147268176078796,
      "step": 182
    },
    {
      "epoch": 0.8285229202037352,
      "grad_norm": 0.2379839159731896,
      "learning_rate": 9.66612965656245e-05,
      "loss": 0.23470696806907654,
      "step": 183
    },
    {
      "epoch": 0.8330503678551217,
      "grad_norm": 0.2377607729961128,
      "learning_rate": 9.658994229264514e-05,
      "loss": 0.23892493546009064,
      "step": 184
    },
    {
      "epoch": 0.8375778155065082,
      "grad_norm": 0.2765508729309311,
      "learning_rate": 9.651786047929773e-05,
      "loss": 0.22539551556110382,
      "step": 185
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 0.2519593343985967,
      "learning_rate": 9.644505225119922e-05,
      "loss": 0.22245247662067413,
      "step": 186
    },
    {
      "epoch": 0.8466327108092813,
      "grad_norm": 0.2628128409473277,
      "learning_rate": 9.637151874531014e-05,
      "loss": 0.22229069471359253,
      "step": 187
    },
    {
      "epoch": 0.8511601584606678,
      "grad_norm": 0.24756613696937937,
      "learning_rate": 9.62972611099168e-05,
      "loss": 0.23303383588790894,
      "step": 188
    },
    {
      "epoch": 0.8556876061120543,
      "grad_norm": 0.21745096846866624,
      "learning_rate": 9.622228050461343e-05,
      "loss": 0.2513193190097809,
      "step": 189
    },
    {
      "epoch": 0.8602150537634409,
      "grad_norm": 0.25236639225485624,
      "learning_rate": 9.614657810028402e-05,
      "loss": 0.24287942051887512,
      "step": 190
    },
    {
      "epoch": 0.8647425014148273,
      "grad_norm": 0.22852889983816463,
      "learning_rate": 9.607015507908401e-05,
      "loss": 0.2401147484779358,
      "step": 191
    },
    {
      "epoch": 0.8692699490662139,
      "grad_norm": 0.2674473200163257,
      "learning_rate": 9.599301263442192e-05,
      "loss": 0.22889262437820435,
      "step": 192
    },
    {
      "epoch": 0.8737973967176005,
      "grad_norm": 0.24774642015877638,
      "learning_rate": 9.591515197094064e-05,
      "loss": 0.22174108028411865,
      "step": 193
    },
    {
      "epoch": 0.878324844368987,
      "grad_norm": 0.2251412757821093,
      "learning_rate": 9.583657430449862e-05,
      "loss": 0.22398972511291504,
      "step": 194
    },
    {
      "epoch": 0.8828522920203735,
      "grad_norm": 0.240165320306096,
      "learning_rate": 9.575728086215092e-05,
      "loss": 0.24194234609603882,
      "step": 195
    },
    {
      "epoch": 0.8873797396717601,
      "grad_norm": 0.3695367603182094,
      "learning_rate": 9.567727288213005e-05,
      "loss": 0.26238495111465454,
      "step": 196
    },
    {
      "epoch": 0.8919071873231466,
      "grad_norm": 0.23280374098825996,
      "learning_rate": 9.559655161382657e-05,
      "loss": 0.22860734164714813,
      "step": 197
    },
    {
      "epoch": 0.8964346349745331,
      "grad_norm": 0.25973865710646277,
      "learning_rate": 9.551511831776965e-05,
      "loss": 0.23258307576179504,
      "step": 198
    },
    {
      "epoch": 0.9009620826259196,
      "grad_norm": 0.24401070938951658,
      "learning_rate": 9.543297426560739e-05,
      "loss": 0.22125867009162903,
      "step": 199
    },
    {
      "epoch": 0.9054895302773062,
      "grad_norm": 0.2822334392942023,
      "learning_rate": 9.535012074008687e-05,
      "loss": 0.21612870693206787,
      "step": 200
    },
    {
      "epoch": 0.9100169779286927,
      "grad_norm": 0.25403982737539815,
      "learning_rate": 9.526655903503423e-05,
      "loss": 0.2510525584220886,
      "step": 201
    },
    {
      "epoch": 0.9145444255800792,
      "grad_norm": 0.2673253097092775,
      "learning_rate": 9.518229045533438e-05,
      "loss": 0.22703340649604797,
      "step": 202
    },
    {
      "epoch": 0.9190718732314658,
      "grad_norm": 0.2544786736455726,
      "learning_rate": 9.50973163169107e-05,
      "loss": 0.24126993119716644,
      "step": 203
    },
    {
      "epoch": 0.9235993208828522,
      "grad_norm": 0.25793075974148677,
      "learning_rate": 9.501163794670444e-05,
      "loss": 0.22873973846435547,
      "step": 204
    },
    {
      "epoch": 0.9281267685342388,
      "grad_norm": 0.2752285689352442,
      "learning_rate": 9.492525668265399e-05,
      "loss": 0.21138757467269897,
      "step": 205
    },
    {
      "epoch": 0.9326542161856254,
      "grad_norm": 0.24474589938416696,
      "learning_rate": 9.483817387367403e-05,
      "loss": 0.2193620204925537,
      "step": 206
    },
    {
      "epoch": 0.9371816638370118,
      "grad_norm": 0.2528310377042809,
      "learning_rate": 9.475039087963442e-05,
      "loss": 0.2507357895374298,
      "step": 207
    },
    {
      "epoch": 0.9417091114883984,
      "grad_norm": 0.23129738265349994,
      "learning_rate": 9.4661909071339e-05,
      "loss": 0.2444463074207306,
      "step": 208
    },
    {
      "epoch": 0.946236559139785,
      "grad_norm": 0.2573095558750083,
      "learning_rate": 9.45727298305042e-05,
      "loss": 0.23638638854026794,
      "step": 209
    },
    {
      "epoch": 0.9507640067911715,
      "grad_norm": 0.2515375484392366,
      "learning_rate": 9.448285454973738e-05,
      "loss": 0.2397586703300476,
      "step": 210
    },
    {
      "epoch": 0.955291454442558,
      "grad_norm": 0.2372193469399196,
      "learning_rate": 9.439228463251515e-05,
      "loss": 0.2245749980211258,
      "step": 211
    },
    {
      "epoch": 0.9598189020939445,
      "grad_norm": 0.2714440784648232,
      "learning_rate": 9.430102149316146e-05,
      "loss": 0.23028632998466492,
      "step": 212
    },
    {
      "epoch": 0.9643463497453311,
      "grad_norm": 0.23968073230614428,
      "learning_rate": 9.420906655682553e-05,
      "loss": 0.2380228042602539,
      "step": 213
    },
    {
      "epoch": 0.9688737973967176,
      "grad_norm": 0.2212932861961108,
      "learning_rate": 9.411642125945945e-05,
      "loss": 0.24913515150547028,
      "step": 214
    },
    {
      "epoch": 0.9734012450481041,
      "grad_norm": 0.2454853876146069,
      "learning_rate": 9.402308704779599e-05,
      "loss": 0.23333734273910522,
      "step": 215
    },
    {
      "epoch": 0.9779286926994907,
      "grad_norm": 0.22069122865008636,
      "learning_rate": 9.392906537932582e-05,
      "loss": 0.2362390011548996,
      "step": 216
    },
    {
      "epoch": 0.9824561403508771,
      "grad_norm": 0.2615361622462132,
      "learning_rate": 9.38343577222748e-05,
      "loss": 0.22081774473190308,
      "step": 217
    },
    {
      "epoch": 0.9869835880022637,
      "grad_norm": 0.24607026565299575,
      "learning_rate": 9.373896555558113e-05,
      "loss": 0.22281277179718018,
      "step": 218
    },
    {
      "epoch": 0.9915110356536503,
      "grad_norm": 0.3213662621119955,
      "learning_rate": 9.364289036887213e-05,
      "loss": 0.2349088191986084,
      "step": 219
    },
    {
      "epoch": 0.9960384833050367,
      "grad_norm": 0.23379740443138045,
      "learning_rate": 9.354613366244108e-05,
      "loss": 0.23271436989307404,
      "step": 220
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.26499035972613255,
      "learning_rate": 9.344869694722372e-05,
      "loss": 0.22342568635940552,
      "step": 221
    },
    {
      "epoch": 1.0045274476513866,
      "grad_norm": 0.2752850648594146,
      "learning_rate": 9.335058174477471e-05,
      "loss": 0.20739434659481049,
      "step": 222
    },
    {
      "epoch": 1.0090548953027731,
      "grad_norm": 0.24098221436784062,
      "learning_rate": 9.325178958724386e-05,
      "loss": 0.2148832529783249,
      "step": 223
    },
    {
      "epoch": 1.0135823429541595,
      "grad_norm": 0.2879563972448631,
      "learning_rate": 9.315232201735217e-05,
      "loss": 0.19599905610084534,
      "step": 224
    },
    {
      "epoch": 1.018109790605546,
      "grad_norm": 0.2922679962376902,
      "learning_rate": 9.305218058836778e-05,
      "loss": 0.18351894617080688,
      "step": 225
    },
    {
      "epoch": 1.0226372382569326,
      "grad_norm": 0.2638706498950157,
      "learning_rate": 9.295136686408166e-05,
      "loss": 0.17019593715667725,
      "step": 226
    },
    {
      "epoch": 1.0271646859083192,
      "grad_norm": 0.28406751291904697,
      "learning_rate": 9.284988241878326e-05,
      "loss": 0.1615096628665924,
      "step": 227
    },
    {
      "epoch": 1.0316921335597058,
      "grad_norm": 0.28488039655510994,
      "learning_rate": 9.274772883723587e-05,
      "loss": 0.18971635401248932,
      "step": 228
    },
    {
      "epoch": 1.0362195812110921,
      "grad_norm": 0.22879868791939567,
      "learning_rate": 9.264490771465191e-05,
      "loss": 0.20442956686019897,
      "step": 229
    },
    {
      "epoch": 1.0407470288624787,
      "grad_norm": 0.25817967842281037,
      "learning_rate": 9.254142065666801e-05,
      "loss": 0.18615120649337769,
      "step": 230
    },
    {
      "epoch": 1.0452744765138653,
      "grad_norm": 0.2579986697138899,
      "learning_rate": 9.243726927931991e-05,
      "loss": 0.17054522037506104,
      "step": 231
    },
    {
      "epoch": 1.0498019241652519,
      "grad_norm": 0.2530770386261921,
      "learning_rate": 9.233245520901723e-05,
      "loss": 0.1674395054578781,
      "step": 232
    },
    {
      "epoch": 1.0543293718166384,
      "grad_norm": 0.25560438874254243,
      "learning_rate": 9.222698008251813e-05,
      "loss": 0.166593998670578,
      "step": 233
    },
    {
      "epoch": 1.058856819468025,
      "grad_norm": 0.32317543774295115,
      "learning_rate": 9.21208455469037e-05,
      "loss": 0.199741929769516,
      "step": 234
    },
    {
      "epoch": 1.0633842671194114,
      "grad_norm": 0.24884612149048577,
      "learning_rate": 9.201405325955221e-05,
      "loss": 0.19953754544258118,
      "step": 235
    },
    {
      "epoch": 1.067911714770798,
      "grad_norm": 0.23481904213991484,
      "learning_rate": 9.190660488811331e-05,
      "loss": 0.1850803941488266,
      "step": 236
    },
    {
      "epoch": 1.0724391624221845,
      "grad_norm": 0.26830396060800593,
      "learning_rate": 9.179850211048193e-05,
      "loss": 0.19193042814731598,
      "step": 237
    },
    {
      "epoch": 1.076966610073571,
      "grad_norm": 0.25425212272986186,
      "learning_rate": 9.168974661477205e-05,
      "loss": 0.17169542610645294,
      "step": 238
    },
    {
      "epoch": 1.0814940577249577,
      "grad_norm": 0.24164852163929176,
      "learning_rate": 9.158034009929046e-05,
      "loss": 0.16295838356018066,
      "step": 239
    },
    {
      "epoch": 1.086021505376344,
      "grad_norm": 0.2960035170475587,
      "learning_rate": 9.14702842725101e-05,
      "loss": 0.19935685396194458,
      "step": 240
    },
    {
      "epoch": 1.0905489530277306,
      "grad_norm": 0.29022028829516544,
      "learning_rate": 9.135958085304344e-05,
      "loss": 0.20939302444458008,
      "step": 241
    },
    {
      "epoch": 1.0950764006791172,
      "grad_norm": 0.2464121104543295,
      "learning_rate": 9.12482315696157e-05,
      "loss": 0.1964278519153595,
      "step": 242
    },
    {
      "epoch": 1.0996038483305037,
      "grad_norm": 0.2780392524554886,
      "learning_rate": 9.113623816103773e-05,
      "loss": 0.18398499488830566,
      "step": 243
    },
    {
      "epoch": 1.1041312959818903,
      "grad_norm": 0.24843742573980285,
      "learning_rate": 9.102360237617899e-05,
      "loss": 0.1648891270160675,
      "step": 244
    },
    {
      "epoch": 1.1086587436332767,
      "grad_norm": 0.2719985733182779,
      "learning_rate": 9.091032597394012e-05,
      "loss": 0.16693006455898285,
      "step": 245
    },
    {
      "epoch": 1.1131861912846632,
      "grad_norm": 0.2922019324194098,
      "learning_rate": 9.079641072322556e-05,
      "loss": 0.17014384269714355,
      "step": 246
    },
    {
      "epoch": 1.1177136389360498,
      "grad_norm": 0.2694898285990861,
      "learning_rate": 9.068185840291588e-05,
      "loss": 0.20611386001110077,
      "step": 247
    },
    {
      "epoch": 1.1222410865874364,
      "grad_norm": 0.25913955722976884,
      "learning_rate": 9.056667080184003e-05,
      "loss": 0.20523002743721008,
      "step": 248
    },
    {
      "epoch": 1.126768534238823,
      "grad_norm": 0.23805155408702844,
      "learning_rate": 9.045084971874738e-05,
      "loss": 0.1908320188522339,
      "step": 249
    },
    {
      "epoch": 1.1312959818902093,
      "grad_norm": 0.2470384177768955,
      "learning_rate": 9.033439696227965e-05,
      "loss": 0.17448925971984863,
      "step": 250
    },
    {
      "epoch": 1.1358234295415959,
      "grad_norm": 0.24421211016429645,
      "learning_rate": 9.021731435094268e-05,
      "loss": 0.17978568375110626,
      "step": 251
    },
    {
      "epoch": 1.1403508771929824,
      "grad_norm": 0.25947666303083733,
      "learning_rate": 9.009960371307798e-05,
      "loss": 0.15618249773979187,
      "step": 252
    },
    {
      "epoch": 1.144878324844369,
      "grad_norm": 0.27569138136904253,
      "learning_rate": 8.998126688683422e-05,
      "loss": 0.20588532090187073,
      "step": 253
    },
    {
      "epoch": 1.1494057724957556,
      "grad_norm": 0.27422459568616203,
      "learning_rate": 8.986230572013855e-05,
      "loss": 0.21392375230789185,
      "step": 254
    },
    {
      "epoch": 1.1539332201471422,
      "grad_norm": 0.24852824385492137,
      "learning_rate": 8.974272207066767e-05,
      "loss": 0.18582111597061157,
      "step": 255
    },
    {
      "epoch": 1.1584606677985285,
      "grad_norm": 0.24375124438036957,
      "learning_rate": 8.962251780581887e-05,
      "loss": 0.18185541033744812,
      "step": 256
    },
    {
      "epoch": 1.162988115449915,
      "grad_norm": 0.2547521793599632,
      "learning_rate": 8.95016948026809e-05,
      "loss": 0.1750154048204422,
      "step": 257
    },
    {
      "epoch": 1.1675155631013017,
      "grad_norm": 0.28224762417970106,
      "learning_rate": 8.938025494800454e-05,
      "loss": 0.16390518844127655,
      "step": 258
    },
    {
      "epoch": 1.1720430107526882,
      "grad_norm": 0.24033758475859487,
      "learning_rate": 8.925820013817329e-05,
      "loss": 0.19912275671958923,
      "step": 259
    },
    {
      "epoch": 1.1765704584040746,
      "grad_norm": 0.24817487616754882,
      "learning_rate": 8.913553227917367e-05,
      "loss": 0.20072248578071594,
      "step": 260
    },
    {
      "epoch": 1.1810979060554612,
      "grad_norm": 0.24205577998995131,
      "learning_rate": 8.901225328656542e-05,
      "loss": 0.19208306074142456,
      "step": 261
    },
    {
      "epoch": 1.1856253537068477,
      "grad_norm": 0.22991487472466549,
      "learning_rate": 8.88883650854517e-05,
      "loss": 0.18774566054344177,
      "step": 262
    },
    {
      "epoch": 1.1901528013582343,
      "grad_norm": 0.2578167724932888,
      "learning_rate": 8.876386961044891e-05,
      "loss": 0.17961907386779785,
      "step": 263
    },
    {
      "epoch": 1.1946802490096209,
      "grad_norm": 0.25461586663077734,
      "learning_rate": 8.863876880565656e-05,
      "loss": 0.1687210202217102,
      "step": 264
    },
    {
      "epoch": 1.1992076966610075,
      "grad_norm": 0.23594236397959223,
      "learning_rate": 8.851306462462688e-05,
      "loss": 0.18248504400253296,
      "step": 265
    },
    {
      "epoch": 1.2037351443123938,
      "grad_norm": 0.23685664063354384,
      "learning_rate": 8.83867590303343e-05,
      "loss": 0.199457585811615,
      "step": 266
    },
    {
      "epoch": 1.2082625919637804,
      "grad_norm": 0.23288586801493874,
      "learning_rate": 8.825985399514487e-05,
      "loss": 0.19683831930160522,
      "step": 267
    },
    {
      "epoch": 1.212790039615167,
      "grad_norm": 0.22652733194791236,
      "learning_rate": 8.813235150078531e-05,
      "loss": 0.1852969229221344,
      "step": 268
    },
    {
      "epoch": 1.2173174872665535,
      "grad_norm": 0.2228879720848376,
      "learning_rate": 8.800425353831226e-05,
      "loss": 0.17336446046829224,
      "step": 269
    },
    {
      "epoch": 1.22184493491794,
      "grad_norm": 0.22594774756703834,
      "learning_rate": 8.787556210808101e-05,
      "loss": 0.16956844925880432,
      "step": 270
    },
    {
      "epoch": 1.2263723825693265,
      "grad_norm": 0.24916486672724852,
      "learning_rate": 8.774627921971436e-05,
      "loss": 0.15064643323421478,
      "step": 271
    },
    {
      "epoch": 1.230899830220713,
      "grad_norm": 0.22725656650553938,
      "learning_rate": 8.761640689207123e-05,
      "loss": 0.22100770473480225,
      "step": 272
    },
    {
      "epoch": 1.2354272778720996,
      "grad_norm": 0.22601605546228898,
      "learning_rate": 8.748594715321512e-05,
      "loss": 0.1962413638830185,
      "step": 273
    },
    {
      "epoch": 1.2399547255234862,
      "grad_norm": 0.2332067860133986,
      "learning_rate": 8.735490204038243e-05,
      "loss": 0.19212886691093445,
      "step": 274
    },
    {
      "epoch": 1.2444821731748728,
      "grad_norm": 0.22983107806955186,
      "learning_rate": 8.722327359995064e-05,
      "loss": 0.1842222809791565,
      "step": 275
    },
    {
      "epoch": 1.249009620826259,
      "grad_norm": 0.23211304709258715,
      "learning_rate": 8.709106388740642e-05,
      "loss": 0.166344553232193,
      "step": 276
    },
    {
      "epoch": 1.2535370684776457,
      "grad_norm": 0.24571103366679994,
      "learning_rate": 8.695827496731342e-05,
      "loss": 0.15337689220905304,
      "step": 277
    },
    {
      "epoch": 1.2580645161290323,
      "grad_norm": 0.25541992708482025,
      "learning_rate": 8.682490891328016e-05,
      "loss": 0.19148650765419006,
      "step": 278
    },
    {
      "epoch": 1.2625919637804188,
      "grad_norm": 0.22861549262407976,
      "learning_rate": 8.669096780792753e-05,
      "loss": 0.2060171663761139,
      "step": 279
    },
    {
      "epoch": 1.2671194114318054,
      "grad_norm": 0.2584206185400821,
      "learning_rate": 8.655645374285637e-05,
      "loss": 0.2004309892654419,
      "step": 280
    },
    {
      "epoch": 1.2716468590831917,
      "grad_norm": 0.23603568932463323,
      "learning_rate": 8.64213688186147e-05,
      "loss": 0.19571299850940704,
      "step": 281
    },
    {
      "epoch": 1.2761743067345783,
      "grad_norm": 0.26642701179023764,
      "learning_rate": 8.628571514466501e-05,
      "loss": 0.1712074726819992,
      "step": 282
    },
    {
      "epoch": 1.280701754385965,
      "grad_norm": 0.2328806692093194,
      "learning_rate": 8.61494948393513e-05,
      "loss": 0.16103923320770264,
      "step": 283
    },
    {
      "epoch": 1.2852292020373515,
      "grad_norm": 0.2904277029583309,
      "learning_rate": 8.601271002986595e-05,
      "loss": 0.20578786730766296,
      "step": 284
    },
    {
      "epoch": 1.289756649688738,
      "grad_norm": 0.3006894192800676,
      "learning_rate": 8.587536285221656e-05,
      "loss": 0.20829525589942932,
      "step": 285
    },
    {
      "epoch": 1.2942840973401246,
      "grad_norm": 0.2501043326869867,
      "learning_rate": 8.573745545119257e-05,
      "loss": 0.20398181676864624,
      "step": 286
    },
    {
      "epoch": 1.298811544991511,
      "grad_norm": 0.24264916410989754,
      "learning_rate": 8.559898998033178e-05,
      "loss": 0.18764348328113556,
      "step": 287
    },
    {
      "epoch": 1.3033389926428975,
      "grad_norm": 0.2562744146657136,
      "learning_rate": 8.545996860188668e-05,
      "loss": 0.18195241689682007,
      "step": 288
    },
    {
      "epoch": 1.3078664402942841,
      "grad_norm": 0.24730795883244333,
      "learning_rate": 8.532039348679073e-05,
      "loss": 0.1701197326183319,
      "step": 289
    },
    {
      "epoch": 1.3123938879456707,
      "grad_norm": 0.23718454397392824,
      "learning_rate": 8.518026681462448e-05,
      "loss": 0.17796757817268372,
      "step": 290
    },
    {
      "epoch": 1.316921335597057,
      "grad_norm": 0.26644257424403134,
      "learning_rate": 8.503959077358143e-05,
      "loss": 0.204569011926651,
      "step": 291
    },
    {
      "epoch": 1.3214487832484436,
      "grad_norm": 0.2381331025166414,
      "learning_rate": 8.4898367560434e-05,
      "loss": 0.18774446845054626,
      "step": 292
    },
    {
      "epoch": 1.3259762308998302,
      "grad_norm": 0.2618993159968482,
      "learning_rate": 8.475659938049912e-05,
      "loss": 0.18409812450408936,
      "step": 293
    },
    {
      "epoch": 1.3305036785512168,
      "grad_norm": 0.2302993540643751,
      "learning_rate": 8.46142884476038e-05,
      "loss": 0.18080274760723114,
      "step": 294
    },
    {
      "epoch": 1.3350311262026033,
      "grad_norm": 0.24254296224043767,
      "learning_rate": 8.44714369840506e-05,
      "loss": 0.1700076162815094,
      "step": 295
    },
    {
      "epoch": 1.33955857385399,
      "grad_norm": 0.2434792462353294,
      "learning_rate": 8.432804722058296e-05,
      "loss": 0.16379863023757935,
      "step": 296
    },
    {
      "epoch": 1.3440860215053765,
      "grad_norm": 0.2292742396093162,
      "learning_rate": 8.418412139635025e-05,
      "loss": 0.20876599848270416,
      "step": 297
    },
    {
      "epoch": 1.3486134691567628,
      "grad_norm": 0.24015507761248567,
      "learning_rate": 8.403966175887292e-05,
      "loss": 0.19511157274246216,
      "step": 298
    },
    {
      "epoch": 1.3531409168081494,
      "grad_norm": 0.23835404601487503,
      "learning_rate": 8.389467056400732e-05,
      "loss": 0.18732516467571259,
      "step": 299
    },
    {
      "epoch": 1.357668364459536,
      "grad_norm": 0.23394631208167474,
      "learning_rate": 8.374915007591053e-05,
      "loss": 0.17540541291236877,
      "step": 300
    },
    {
      "epoch": 1.3621958121109223,
      "grad_norm": 0.24058113236241577,
      "learning_rate": 8.360310256700497e-05,
      "loss": 0.16894260048866272,
      "step": 301
    },
    {
      "epoch": 1.366723259762309,
      "grad_norm": 0.22971701577193754,
      "learning_rate": 8.345653031794292e-05,
      "loss": 0.15845853090286255,
      "step": 302
    },
    {
      "epoch": 1.3712507074136955,
      "grad_norm": 0.22497275896865934,
      "learning_rate": 8.330943561757092e-05,
      "loss": 0.19490782916545868,
      "step": 303
    },
    {
      "epoch": 1.375778155065082,
      "grad_norm": 0.2397201329461442,
      "learning_rate": 8.316182076289401e-05,
      "loss": 0.20050162076950073,
      "step": 304
    },
    {
      "epoch": 1.3803056027164686,
      "grad_norm": 0.23131154556846992,
      "learning_rate": 8.301368805903988e-05,
      "loss": 0.18538133800029755,
      "step": 305
    },
    {
      "epoch": 1.3848330503678552,
      "grad_norm": 0.22438363660363442,
      "learning_rate": 8.286503981922283e-05,
      "loss": 0.17794030904769897,
      "step": 306
    },
    {
      "epoch": 1.3893604980192418,
      "grad_norm": 0.22915377851132254,
      "learning_rate": 8.271587836470775e-05,
      "loss": 0.17239966988563538,
      "step": 307
    },
    {
      "epoch": 1.3938879456706281,
      "grad_norm": 0.22301765516003927,
      "learning_rate": 8.256620602477372e-05,
      "loss": 0.15613433718681335,
      "step": 308
    },
    {
      "epoch": 1.3984153933220147,
      "grad_norm": 0.2556039458430123,
      "learning_rate": 8.241602513667774e-05,
      "loss": 0.19233815371990204,
      "step": 309
    },
    {
      "epoch": 1.4029428409734013,
      "grad_norm": 0.22161397456653995,
      "learning_rate": 8.226533804561827e-05,
      "loss": 0.20076030492782593,
      "step": 310
    },
    {
      "epoch": 1.4074702886247878,
      "grad_norm": 0.2158517522160699,
      "learning_rate": 8.211414710469845e-05,
      "loss": 0.19023515284061432,
      "step": 311
    },
    {
      "epoch": 1.4119977362761742,
      "grad_norm": 0.23301307869370144,
      "learning_rate": 8.19624546748895e-05,
      "loss": 0.18671788275241852,
      "step": 312
    },
    {
      "epoch": 1.4165251839275608,
      "grad_norm": 0.2350909348433444,
      "learning_rate": 8.181026312499383e-05,
      "loss": 0.19153177738189697,
      "step": 313
    },
    {
      "epoch": 1.4210526315789473,
      "grad_norm": 0.25490025232781116,
      "learning_rate": 8.165757483160798e-05,
      "loss": 0.1734253168106079,
      "step": 314
    },
    {
      "epoch": 1.425580079230334,
      "grad_norm": 0.2186984511700432,
      "learning_rate": 8.150439217908556e-05,
      "loss": 0.17426717281341553,
      "step": 315
    },
    {
      "epoch": 1.4301075268817205,
      "grad_norm": 0.24210463849625682,
      "learning_rate": 8.13507175595e-05,
      "loss": 0.2044614553451538,
      "step": 316
    },
    {
      "epoch": 1.434634974533107,
      "grad_norm": 0.2291072741742975,
      "learning_rate": 8.11965533726072e-05,
      "loss": 0.1858120560646057,
      "step": 317
    },
    {
      "epoch": 1.4391624221844934,
      "grad_norm": 0.2271627598789281,
      "learning_rate": 8.104190202580812e-05,
      "loss": 0.18891949951648712,
      "step": 318
    },
    {
      "epoch": 1.44368986983588,
      "grad_norm": 0.2258569909298315,
      "learning_rate": 8.0886765934111e-05,
      "loss": 0.16802790760993958,
      "step": 319
    },
    {
      "epoch": 1.4482173174872666,
      "grad_norm": 0.23439655207595847,
      "learning_rate": 8.073114752009387e-05,
      "loss": 0.16781140863895416,
      "step": 320
    },
    {
      "epoch": 1.4527447651386531,
      "grad_norm": 0.2792513119608853,
      "learning_rate": 8.05750492138666e-05,
      "loss": 0.1735672950744629,
      "step": 321
    },
    {
      "epoch": 1.4572722127900395,
      "grad_norm": 0.20612097968855694,
      "learning_rate": 8.041847345303297e-05,
      "loss": 0.20380868017673492,
      "step": 322
    },
    {
      "epoch": 1.461799660441426,
      "grad_norm": 0.2313183526770185,
      "learning_rate": 8.026142268265256e-05,
      "loss": 0.19325749576091766,
      "step": 323
    },
    {
      "epoch": 1.4663271080928126,
      "grad_norm": 0.23353137321186834,
      "learning_rate": 8.01038993552027e-05,
      "loss": 0.19468310475349426,
      "step": 324
    },
    {
      "epoch": 1.4708545557441992,
      "grad_norm": 0.23764416321851503,
      "learning_rate": 7.994590593054001e-05,
      "loss": 0.18415693938732147,
      "step": 325
    },
    {
      "epoch": 1.4753820033955858,
      "grad_norm": 0.2235625506267048,
      "learning_rate": 7.978744487586214e-05,
      "loss": 0.15884950757026672,
      "step": 326
    },
    {
      "epoch": 1.4799094510469724,
      "grad_norm": 0.2664821870999596,
      "learning_rate": 7.962851866566912e-05,
      "loss": 0.16488111019134521,
      "step": 327
    },
    {
      "epoch": 1.484436898698359,
      "grad_norm": 0.24207702126869313,
      "learning_rate": 7.946912978172474e-05,
      "loss": 0.208705335855484,
      "step": 328
    },
    {
      "epoch": 1.4889643463497453,
      "grad_norm": 0.23499117845400197,
      "learning_rate": 7.93092807130179e-05,
      "loss": 0.2106453776359558,
      "step": 329
    },
    {
      "epoch": 1.4934917940011319,
      "grad_norm": 0.22987971104295032,
      "learning_rate": 7.91489739557236e-05,
      "loss": 0.20180118083953857,
      "step": 330
    },
    {
      "epoch": 1.4980192416525184,
      "grad_norm": 0.2394203019376777,
      "learning_rate": 7.898821201316407e-05,
      "loss": 0.18206928670406342,
      "step": 331
    },
    {
      "epoch": 1.5025466893039048,
      "grad_norm": 0.23449731981443797,
      "learning_rate": 7.882699739576959e-05,
      "loss": 0.18937627971172333,
      "step": 332
    },
    {
      "epoch": 1.5070741369552914,
      "grad_norm": 0.23370924946757207,
      "learning_rate": 7.866533262103936e-05,
      "loss": 0.15429037809371948,
      "step": 333
    },
    {
      "epoch": 1.511601584606678,
      "grad_norm": 0.2348912406022975,
      "learning_rate": 7.850322021350215e-05,
      "loss": 0.18988651037216187,
      "step": 334
    },
    {
      "epoch": 1.5161290322580645,
      "grad_norm": 0.2432086507465226,
      "learning_rate": 7.83406627046769e-05,
      "loss": 0.2089129537343979,
      "step": 335
    },
    {
      "epoch": 1.520656479909451,
      "grad_norm": 0.22816092626220283,
      "learning_rate": 7.817766263303313e-05,
      "loss": 0.1933603584766388,
      "step": 336
    },
    {
      "epoch": 1.5251839275608376,
      "grad_norm": 0.22858247706166318,
      "learning_rate": 7.801422254395138e-05,
      "loss": 0.1819375455379486,
      "step": 337
    },
    {
      "epoch": 1.5297113752122242,
      "grad_norm": 0.22790539922779043,
      "learning_rate": 7.785034498968344e-05,
      "loss": 0.17599374055862427,
      "step": 338
    },
    {
      "epoch": 1.5342388228636108,
      "grad_norm": 0.2441118933134464,
      "learning_rate": 7.768603252931243e-05,
      "loss": 0.17194731533527374,
      "step": 339
    },
    {
      "epoch": 1.5387662705149971,
      "grad_norm": 0.22048564382617603,
      "learning_rate": 7.752128772871292e-05,
      "loss": 0.17797675728797913,
      "step": 340
    },
    {
      "epoch": 1.5432937181663837,
      "grad_norm": 0.2330896122968261,
      "learning_rate": 7.735611316051084e-05,
      "loss": 0.21639850735664368,
      "step": 341
    },
    {
      "epoch": 1.54782116581777,
      "grad_norm": 0.2389971369177504,
      "learning_rate": 7.719051140404327e-05,
      "loss": 0.19379103183746338,
      "step": 342
    },
    {
      "epoch": 1.5523486134691566,
      "grad_norm": 0.23465330342048735,
      "learning_rate": 7.702448504531819e-05,
      "loss": 0.18040522933006287,
      "step": 343
    },
    {
      "epoch": 1.5568760611205432,
      "grad_norm": 0.2356363676068827,
      "learning_rate": 7.685803667697411e-05,
      "loss": 0.17366576194763184,
      "step": 344
    },
    {
      "epoch": 1.5614035087719298,
      "grad_norm": 0.23082440956515077,
      "learning_rate": 7.669116889823955e-05,
      "loss": 0.16583794355392456,
      "step": 345
    },
    {
      "epoch": 1.5659309564233164,
      "grad_norm": 0.25911599556300013,
      "learning_rate": 7.652388431489248e-05,
      "loss": 0.1619320809841156,
      "step": 346
    },
    {
      "epoch": 1.570458404074703,
      "grad_norm": 0.2097933254162827,
      "learning_rate": 7.635618553921962e-05,
      "loss": 0.20942410826683044,
      "step": 347
    },
    {
      "epoch": 1.5749858517260895,
      "grad_norm": 0.2233737518899933,
      "learning_rate": 7.618807518997563e-05,
      "loss": 0.20726366341114044,
      "step": 348
    },
    {
      "epoch": 1.579513299377476,
      "grad_norm": 0.23295362591375984,
      "learning_rate": 7.601955589234227e-05,
      "loss": 0.19247621297836304,
      "step": 349
    },
    {
      "epoch": 1.5840407470288624,
      "grad_norm": 0.23071211084375837,
      "learning_rate": 7.585063027788731e-05,
      "loss": 0.182029128074646,
      "step": 350
    },
    {
      "epoch": 1.588568194680249,
      "grad_norm": 0.22351989125506522,
      "learning_rate": 7.568130098452351e-05,
      "loss": 0.16755558550357819,
      "step": 351
    },
    {
      "epoch": 1.5930956423316356,
      "grad_norm": 0.25599011363609997,
      "learning_rate": 7.551157065646746e-05,
      "loss": 0.16772229969501495,
      "step": 352
    },
    {
      "epoch": 1.597623089983022,
      "grad_norm": 0.23397623453567362,
      "learning_rate": 7.534144194419817e-05,
      "loss": 0.20039445161819458,
      "step": 353
    },
    {
      "epoch": 1.6021505376344085,
      "grad_norm": 0.20716371187078184,
      "learning_rate": 7.517091750441576e-05,
      "loss": 0.1948099136352539,
      "step": 354
    },
    {
      "epoch": 1.606677985285795,
      "grad_norm": 0.22909614590863156,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.18500792980194092,
      "step": 355
    },
    {
      "epoch": 1.6112054329371817,
      "grad_norm": 0.24806486473796108,
      "learning_rate": 7.482869209996867e-05,
      "loss": 0.1827983558177948,
      "step": 356
    },
    {
      "epoch": 1.6157328805885682,
      "grad_norm": 0.20995458240393158,
      "learning_rate": 7.465699647943586e-05,
      "loss": 0.1701834797859192,
      "step": 357
    },
    {
      "epoch": 1.6202603282399548,
      "grad_norm": 0.2435204448171681,
      "learning_rate": 7.44849158195703e-05,
      "loss": 0.16099783778190613,
      "step": 358
    },
    {
      "epoch": 1.6247877758913414,
      "grad_norm": 0.22196230609403023,
      "learning_rate": 7.431245280755336e-05,
      "loss": 0.18706944584846497,
      "step": 359
    },
    {
      "epoch": 1.629315223542728,
      "grad_norm": 0.2177395907118696,
      "learning_rate": 7.413961013653726e-05,
      "loss": 0.1985095590353012,
      "step": 360
    },
    {
      "epoch": 1.6338426711941143,
      "grad_norm": 0.23420218333689144,
      "learning_rate": 7.396639050560275e-05,
      "loss": 0.1964409351348877,
      "step": 361
    },
    {
      "epoch": 1.6383701188455009,
      "grad_norm": 0.24252526568770688,
      "learning_rate": 7.379279661971727e-05,
      "loss": 0.1915147602558136,
      "step": 362
    },
    {
      "epoch": 1.6428975664968872,
      "grad_norm": 0.21486825839557844,
      "learning_rate": 7.361883118969247e-05,
      "loss": 0.17337551712989807,
      "step": 363
    },
    {
      "epoch": 1.6474250141482738,
      "grad_norm": 0.24153242168834782,
      "learning_rate": 7.3444496932142e-05,
      "loss": 0.1554442048072815,
      "step": 364
    },
    {
      "epoch": 1.6519524617996604,
      "grad_norm": 0.2304945640598229,
      "learning_rate": 7.326979656943906e-05,
      "loss": 0.17819085717201233,
      "step": 365
    },
    {
      "epoch": 1.656479909451047,
      "grad_norm": 0.21864772860961737,
      "learning_rate": 7.309473282967387e-05,
      "loss": 0.19845563173294067,
      "step": 366
    },
    {
      "epoch": 1.6610073571024335,
      "grad_norm": 0.21212791611920695,
      "learning_rate": 7.291930844661109e-05,
      "loss": 0.19407084584236145,
      "step": 367
    },
    {
      "epoch": 1.66553480475382,
      "grad_norm": 0.23000132762868372,
      "learning_rate": 7.274352615964712e-05,
      "loss": 0.19369813799858093,
      "step": 368
    },
    {
      "epoch": 1.6700622524052067,
      "grad_norm": 0.26386625451467977,
      "learning_rate": 7.256738871376732e-05,
      "loss": 0.1805361658334732,
      "step": 369
    },
    {
      "epoch": 1.6745897000565932,
      "grad_norm": 0.2250077167287235,
      "learning_rate": 7.239089885950316e-05,
      "loss": 0.16192588210105896,
      "step": 370
    },
    {
      "epoch": 1.6791171477079796,
      "grad_norm": 0.25514592594562496,
      "learning_rate": 7.221405935288925e-05,
      "loss": 0.1664281040430069,
      "step": 371
    },
    {
      "epoch": 1.6836445953593662,
      "grad_norm": 0.22878780832134368,
      "learning_rate": 7.203687295542032e-05,
      "loss": 0.20167440176010132,
      "step": 372
    },
    {
      "epoch": 1.6881720430107527,
      "grad_norm": 0.23966301175008684,
      "learning_rate": 7.185934243400806e-05,
      "loss": 0.1990775465965271,
      "step": 373
    },
    {
      "epoch": 1.692699490662139,
      "grad_norm": 0.21754089132557736,
      "learning_rate": 7.168147056093797e-05,
      "loss": 0.18479225039482117,
      "step": 374
    },
    {
      "epoch": 1.6972269383135257,
      "grad_norm": 0.28801731760009536,
      "learning_rate": 7.150326011382604e-05,
      "loss": 0.1853107511997223,
      "step": 375
    },
    {
      "epoch": 1.7017543859649122,
      "grad_norm": 0.2303035094624934,
      "learning_rate": 7.132471387557532e-05,
      "loss": 0.17413756251335144,
      "step": 376
    },
    {
      "epoch": 1.7062818336162988,
      "grad_norm": 0.21566297620020786,
      "learning_rate": 7.114583463433259e-05,
      "loss": 0.15298089385032654,
      "step": 377
    },
    {
      "epoch": 1.7108092812676854,
      "grad_norm": 0.22573410481597223,
      "learning_rate": 7.096662518344468e-05,
      "loss": 0.19422583281993866,
      "step": 378
    },
    {
      "epoch": 1.715336728919072,
      "grad_norm": 0.21405199236051928,
      "learning_rate": 7.078708832141497e-05,
      "loss": 0.18652589619159698,
      "step": 379
    },
    {
      "epoch": 1.7198641765704585,
      "grad_norm": 0.21441763042624978,
      "learning_rate": 7.060722685185961e-05,
      "loss": 0.19040265679359436,
      "step": 380
    },
    {
      "epoch": 1.7243916242218449,
      "grad_norm": 0.2287276042650557,
      "learning_rate": 7.042704358346375e-05,
      "loss": 0.18727976083755493,
      "step": 381
    },
    {
      "epoch": 1.7289190718732315,
      "grad_norm": 0.21306676953811082,
      "learning_rate": 7.024654132993772e-05,
      "loss": 0.1772461086511612,
      "step": 382
    },
    {
      "epoch": 1.733446519524618,
      "grad_norm": 0.22641914389290818,
      "learning_rate": 7.006572290997304e-05,
      "loss": 0.16065247356891632,
      "step": 383
    },
    {
      "epoch": 1.7379739671760044,
      "grad_norm": 0.2163723775253618,
      "learning_rate": 6.988459114719849e-05,
      "loss": 0.17697076499462128,
      "step": 384
    },
    {
      "epoch": 1.742501414827391,
      "grad_norm": 0.20537607404160005,
      "learning_rate": 6.970314887013584e-05,
      "loss": 0.21230991184711456,
      "step": 385
    },
    {
      "epoch": 1.7470288624787775,
      "grad_norm": 0.23281708772592882,
      "learning_rate": 6.952139891215593e-05,
      "loss": 0.1916704922914505,
      "step": 386
    },
    {
      "epoch": 1.751556310130164,
      "grad_norm": 0.20208193625524307,
      "learning_rate": 6.93393441114342e-05,
      "loss": 0.17192980647087097,
      "step": 387
    },
    {
      "epoch": 1.7560837577815507,
      "grad_norm": 0.23905156882195805,
      "learning_rate": 6.915698731090648e-05,
      "loss": 0.17362508177757263,
      "step": 388
    },
    {
      "epoch": 1.7606112054329373,
      "grad_norm": 0.20726985781015517,
      "learning_rate": 6.897433135822461e-05,
      "loss": 0.16086629033088684,
      "step": 389
    },
    {
      "epoch": 1.7651386530843238,
      "grad_norm": 0.21143137753141877,
      "learning_rate": 6.879137910571191e-05,
      "loss": 0.17281287908554077,
      "step": 390
    },
    {
      "epoch": 1.7696661007357104,
      "grad_norm": 0.1988056358395259,
      "learning_rate": 6.860813341031866e-05,
      "loss": 0.20971840620040894,
      "step": 391
    },
    {
      "epoch": 1.7741935483870968,
      "grad_norm": 0.20345665643593203,
      "learning_rate": 6.842459713357752e-05,
      "loss": 0.18777552247047424,
      "step": 392
    },
    {
      "epoch": 1.7787209960384833,
      "grad_norm": 0.21784178411369015,
      "learning_rate": 6.824077314155877e-05,
      "loss": 0.18649156391620636,
      "step": 393
    },
    {
      "epoch": 1.7832484436898697,
      "grad_norm": 0.22281633265480258,
      "learning_rate": 6.805666430482564e-05,
      "loss": 0.18137586116790771,
      "step": 394
    },
    {
      "epoch": 1.7877758913412563,
      "grad_norm": 0.21739305127964964,
      "learning_rate": 6.787227349838947e-05,
      "loss": 0.166051983833313,
      "step": 395
    },
    {
      "epoch": 1.7923033389926428,
      "grad_norm": 0.22627848415767696,
      "learning_rate": 6.768760360166471e-05,
      "loss": 0.15685385465621948,
      "step": 396
    },
    {
      "epoch": 1.7968307866440294,
      "grad_norm": 0.19512762189959926,
      "learning_rate": 6.750265749842409e-05,
      "loss": 0.18851293623447418,
      "step": 397
    },
    {
      "epoch": 1.801358234295416,
      "grad_norm": 0.19851096227569107,
      "learning_rate": 6.731743807675355e-05,
      "loss": 0.20142939686775208,
      "step": 398
    },
    {
      "epoch": 1.8058856819468025,
      "grad_norm": 0.1987549455469965,
      "learning_rate": 6.713194822900706e-05,
      "loss": 0.1752074658870697,
      "step": 399
    },
    {
      "epoch": 1.8104131295981891,
      "grad_norm": 0.24469634191868353,
      "learning_rate": 6.694619085176159e-05,
      "loss": 0.1878441721200943,
      "step": 400
    },
    {
      "epoch": 1.8149405772495757,
      "grad_norm": 0.23840098103990479,
      "learning_rate": 6.676016884577173e-05,
      "loss": 0.1690133512020111,
      "step": 401
    },
    {
      "epoch": 1.819468024900962,
      "grad_norm": 0.23227971846950474,
      "learning_rate": 6.657388511592452e-05,
      "loss": 0.15280716121196747,
      "step": 402
    },
    {
      "epoch": 1.8239954725523486,
      "grad_norm": 0.19879257030565525,
      "learning_rate": 6.638734257119401e-05,
      "loss": 0.18653661012649536,
      "step": 403
    },
    {
      "epoch": 1.8285229202037352,
      "grad_norm": 0.20695685518794696,
      "learning_rate": 6.620054412459588e-05,
      "loss": 0.2069815844297409,
      "step": 404
    },
    {
      "epoch": 1.8330503678551215,
      "grad_norm": 0.20256246903129402,
      "learning_rate": 6.601349269314188e-05,
      "loss": 0.19434502720832825,
      "step": 405
    },
    {
      "epoch": 1.8375778155065081,
      "grad_norm": 0.22025258519215757,
      "learning_rate": 6.582619119779439e-05,
      "loss": 0.18385529518127441,
      "step": 406
    },
    {
      "epoch": 1.8421052631578947,
      "grad_norm": 0.2137361777451771,
      "learning_rate": 6.56386425634207e-05,
      "loss": 0.17453709244728088,
      "step": 407
    },
    {
      "epoch": 1.8466327108092813,
      "grad_norm": 0.21800698801636118,
      "learning_rate": 6.545084971874738e-05,
      "loss": 0.16962510347366333,
      "step": 408
    },
    {
      "epoch": 1.8511601584606678,
      "grad_norm": 0.21773841184668813,
      "learning_rate": 6.526281559631458e-05,
      "loss": 0.18594738841056824,
      "step": 409
    },
    {
      "epoch": 1.8556876061120544,
      "grad_norm": 0.20543663680928975,
      "learning_rate": 6.507454313243015e-05,
      "loss": 0.19622986018657684,
      "step": 410
    },
    {
      "epoch": 1.860215053763441,
      "grad_norm": 0.21878558101742387,
      "learning_rate": 6.48860352671239e-05,
      "loss": 0.1954154521226883,
      "step": 411
    },
    {
      "epoch": 1.8647425014148273,
      "grad_norm": 0.21524365537025683,
      "learning_rate": 6.469729494410158e-05,
      "loss": 0.17159169912338257,
      "step": 412
    },
    {
      "epoch": 1.869269949066214,
      "grad_norm": 0.20616359246250593,
      "learning_rate": 6.450832511069897e-05,
      "loss": 0.1622459441423416,
      "step": 413
    },
    {
      "epoch": 1.8737973967176005,
      "grad_norm": 0.2234602536197417,
      "learning_rate": 6.431912871783586e-05,
      "loss": 0.1627485156059265,
      "step": 414
    },
    {
      "epoch": 1.8783248443689868,
      "grad_norm": 0.20253941588453686,
      "learning_rate": 6.412970871996995e-05,
      "loss": 0.16856414079666138,
      "step": 415
    },
    {
      "epoch": 1.8828522920203734,
      "grad_norm": 0.21881514866997123,
      "learning_rate": 6.394006807505067e-05,
      "loss": 0.21652480959892273,
      "step": 416
    },
    {
      "epoch": 1.88737973967176,
      "grad_norm": 0.21735463374069466,
      "learning_rate": 6.37502097444731e-05,
      "loss": 0.18492794036865234,
      "step": 417
    },
    {
      "epoch": 1.8919071873231466,
      "grad_norm": 0.21855523766409582,
      "learning_rate": 6.356013669303162e-05,
      "loss": 0.18506664037704468,
      "step": 418
    },
    {
      "epoch": 1.8964346349745331,
      "grad_norm": 0.20599360139804854,
      "learning_rate": 6.336985188887366e-05,
      "loss": 0.16712680459022522,
      "step": 419
    },
    {
      "epoch": 1.9009620826259197,
      "grad_norm": 0.212516416997445,
      "learning_rate": 6.317935830345338e-05,
      "loss": 0.16304892301559448,
      "step": 420
    },
    {
      "epoch": 1.9054895302773063,
      "grad_norm": 0.23929914649032236,
      "learning_rate": 6.298865891148518e-05,
      "loss": 0.15581056475639343,
      "step": 421
    },
    {
      "epoch": 1.9100169779286928,
      "grad_norm": 0.20335819449154932,
      "learning_rate": 6.279775669089733e-05,
      "loss": 0.21459367871284485,
      "step": 422
    },
    {
      "epoch": 1.9145444255800792,
      "grad_norm": 0.21257734416198903,
      "learning_rate": 6.260665462278544e-05,
      "loss": 0.1929590404033661,
      "step": 423
    },
    {
      "epoch": 1.9190718732314658,
      "grad_norm": 0.23415671857217688,
      "learning_rate": 6.241535569136584e-05,
      "loss": 0.17095708847045898,
      "step": 424
    },
    {
      "epoch": 1.9235993208828521,
      "grad_norm": 0.21520631254678835,
      "learning_rate": 6.222386288392913e-05,
      "loss": 0.17006826400756836,
      "step": 425
    },
    {
      "epoch": 1.9281267685342387,
      "grad_norm": 0.2200610745944257,
      "learning_rate": 6.203217919079342e-05,
      "loss": 0.155431866645813,
      "step": 426
    },
    {
      "epoch": 1.9326542161856253,
      "grad_norm": 0.2223264165520048,
      "learning_rate": 6.184030760525764e-05,
      "loss": 0.1626005470752716,
      "step": 427
    },
    {
      "epoch": 1.9371816638370118,
      "grad_norm": 0.20514982553421754,
      "learning_rate": 6.164825112355477e-05,
      "loss": 0.1874426305294037,
      "step": 428
    },
    {
      "epoch": 1.9417091114883984,
      "grad_norm": 0.2040379032179284,
      "learning_rate": 6.145601274480521e-05,
      "loss": 0.18967613577842712,
      "step": 429
    },
    {
      "epoch": 1.946236559139785,
      "grad_norm": 0.21479986147172914,
      "learning_rate": 6.126359547096975e-05,
      "loss": 0.178785502910614,
      "step": 430
    },
    {
      "epoch": 1.9507640067911716,
      "grad_norm": 0.21286903742116525,
      "learning_rate": 6.107100230680279e-05,
      "loss": 0.16879679262638092,
      "step": 431
    },
    {
      "epoch": 1.9552914544425581,
      "grad_norm": 0.22034910660483908,
      "learning_rate": 6.0878236259805396e-05,
      "loss": 0.16959422826766968,
      "step": 432
    },
    {
      "epoch": 1.9598189020939445,
      "grad_norm": 0.24349120753831688,
      "learning_rate": 6.068530034017835e-05,
      "loss": 0.16129666566848755,
      "step": 433
    },
    {
      "epoch": 1.964346349745331,
      "grad_norm": 0.18827473527386088,
      "learning_rate": 6.049219756077514e-05,
      "loss": 0.17796547710895538,
      "step": 434
    },
    {
      "epoch": 1.9688737973967176,
      "grad_norm": 0.18918405467160337,
      "learning_rate": 6.029893093705492e-05,
      "loss": 0.18614676594734192,
      "step": 435
    },
    {
      "epoch": 1.973401245048104,
      "grad_norm": 0.2246979279481908,
      "learning_rate": 6.010550348703538e-05,
      "loss": 0.18553641438484192,
      "step": 436
    },
    {
      "epoch": 1.9779286926994906,
      "grad_norm": 0.21206168874010198,
      "learning_rate": 5.991191823124565e-05,
      "loss": 0.1744600236415863,
      "step": 437
    },
    {
      "epoch": 1.9824561403508771,
      "grad_norm": 0.21999951783024038,
      "learning_rate": 5.971817819267913e-05,
      "loss": 0.16687238216400146,
      "step": 438
    },
    {
      "epoch": 1.9869835880022637,
      "grad_norm": 0.23967428162189694,
      "learning_rate": 5.952428639674632e-05,
      "loss": 0.1604543924331665,
      "step": 439
    },
    {
      "epoch": 1.9915110356536503,
      "grad_norm": 0.21086931201900394,
      "learning_rate": 5.9330245871227454e-05,
      "loss": 0.17042537033557892,
      "step": 440
    },
    {
      "epoch": 1.9960384833050369,
      "grad_norm": 0.21783306620521006,
      "learning_rate": 5.9136059646225375e-05,
      "loss": 0.19700202345848083,
      "step": 441
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.2840852610301388,
      "learning_rate": 5.8941730754118116e-05,
      "loss": 0.16136552393436432,
      "step": 442
    },
    {
      "epoch": 2.0045274476513866,
      "grad_norm": 0.34043201285761693,
      "learning_rate": 5.874726222951157e-05,
      "loss": 0.1516227424144745,
      "step": 443
    },
    {
      "epoch": 2.009054895302773,
      "grad_norm": 0.24500131072597228,
      "learning_rate": 5.855265710919211e-05,
      "loss": 0.13573628664016724,
      "step": 444
    },
    {
      "epoch": 2.0135823429541597,
      "grad_norm": 0.5473269358411129,
      "learning_rate": 5.835791843207916e-05,
      "loss": 0.12611983716487885,
      "step": 445
    },
    {
      "epoch": 2.0181097906055463,
      "grad_norm": 0.29649620899571866,
      "learning_rate": 5.8163049239177784e-05,
      "loss": 0.11450044810771942,
      "step": 446
    },
    {
      "epoch": 2.022637238256933,
      "grad_norm": 0.29036589789947115,
      "learning_rate": 5.7968052573531084e-05,
      "loss": 0.10637161135673523,
      "step": 447
    },
    {
      "epoch": 2.027164685908319,
      "grad_norm": 0.2771109717391068,
      "learning_rate": 5.7772931480172795e-05,
      "loss": 0.09776574373245239,
      "step": 448
    },
    {
      "epoch": 2.0316921335597056,
      "grad_norm": 0.323631395065372,
      "learning_rate": 5.757768900607972e-05,
      "loss": 0.14583757519721985,
      "step": 449
    },
    {
      "epoch": 2.036219581211092,
      "grad_norm": 0.3061023195343294,
      "learning_rate": 5.738232820012407e-05,
      "loss": 0.1459420919418335,
      "step": 450
    },
    {
      "epoch": 2.0407470288624787,
      "grad_norm": 0.2342789652760014,
      "learning_rate": 5.718685211302592e-05,
      "loss": 0.130174919962883,
      "step": 451
    },
    {
      "epoch": 2.0452744765138653,
      "grad_norm": 0.2572683144609063,
      "learning_rate": 5.699126379730559e-05,
      "loss": 0.11415386199951172,
      "step": 452
    },
    {
      "epoch": 2.049801924165252,
      "grad_norm": 0.26093673764142417,
      "learning_rate": 5.6795566307235915e-05,
      "loss": 0.10571029037237167,
      "step": 453
    },
    {
      "epoch": 2.0543293718166384,
      "grad_norm": 0.23275157616142492,
      "learning_rate": 5.6599762698794554e-05,
      "loss": 0.09423750638961792,
      "step": 454
    },
    {
      "epoch": 2.058856819468025,
      "grad_norm": 0.21870818736956663,
      "learning_rate": 5.640385602961634e-05,
      "loss": 0.12411186099052429,
      "step": 455
    },
    {
      "epoch": 2.0633842671194116,
      "grad_norm": 0.2595869834999154,
      "learning_rate": 5.620784935894547e-05,
      "loss": 0.14633986353874207,
      "step": 456
    },
    {
      "epoch": 2.067911714770798,
      "grad_norm": 0.2311176864459013,
      "learning_rate": 5.601174574758771e-05,
      "loss": 0.12183500826358795,
      "step": 457
    },
    {
      "epoch": 2.0724391624221843,
      "grad_norm": 0.22784673976437347,
      "learning_rate": 5.581554825786267e-05,
      "loss": 0.11389859020709991,
      "step": 458
    },
    {
      "epoch": 2.076966610073571,
      "grad_norm": 0.23797601078837152,
      "learning_rate": 5.5619259953555945e-05,
      "loss": 0.10225850343704224,
      "step": 459
    },
    {
      "epoch": 2.0814940577249574,
      "grad_norm": 0.2352587872410654,
      "learning_rate": 5.5422883899871284e-05,
      "loss": 0.0940392017364502,
      "step": 460
    },
    {
      "epoch": 2.086021505376344,
      "grad_norm": 0.22058429661210505,
      "learning_rate": 5.522642316338268e-05,
      "loss": 0.12159469723701477,
      "step": 461
    },
    {
      "epoch": 2.0905489530277306,
      "grad_norm": 0.21887132931344502,
      "learning_rate": 5.5029880811986544e-05,
      "loss": 0.1463402509689331,
      "step": 462
    },
    {
      "epoch": 2.095076400679117,
      "grad_norm": 0.2164365588361623,
      "learning_rate": 5.483325991485379e-05,
      "loss": 0.13759233057498932,
      "step": 463
    },
    {
      "epoch": 2.0996038483305037,
      "grad_norm": 0.21488628541756813,
      "learning_rate": 5.463656354238184e-05,
      "loss": 0.11654265224933624,
      "step": 464
    },
    {
      "epoch": 2.1041312959818903,
      "grad_norm": 0.22603763343851777,
      "learning_rate": 5.4439794766146746e-05,
      "loss": 0.10867158323526382,
      "step": 465
    },
    {
      "epoch": 2.108658743633277,
      "grad_norm": 0.21410526450166345,
      "learning_rate": 5.424295665885523e-05,
      "loss": 0.10072112083435059,
      "step": 466
    },
    {
      "epoch": 2.1131861912846635,
      "grad_norm": 0.26519511640876037,
      "learning_rate": 5.404605229429664e-05,
      "loss": 0.09951277077198029,
      "step": 467
    },
    {
      "epoch": 2.11771363893605,
      "grad_norm": 0.2152943228441027,
      "learning_rate": 5.384908474729501e-05,
      "loss": 0.15281015634536743,
      "step": 468
    },
    {
      "epoch": 2.122241086587436,
      "grad_norm": 0.22369944720082735,
      "learning_rate": 5.365205709366099e-05,
      "loss": 0.14525587856769562,
      "step": 469
    },
    {
      "epoch": 2.1267685342388227,
      "grad_norm": 0.24122685007261,
      "learning_rate": 5.34549724101439e-05,
      "loss": 0.1253020465373993,
      "step": 470
    },
    {
      "epoch": 2.1312959818902093,
      "grad_norm": 0.2260455077583836,
      "learning_rate": 5.325783377438357e-05,
      "loss": 0.11360302567481995,
      "step": 471
    },
    {
      "epoch": 2.135823429541596,
      "grad_norm": 0.22335618437779742,
      "learning_rate": 5.306064426486237e-05,
      "loss": 0.1024991124868393,
      "step": 472
    },
    {
      "epoch": 2.1403508771929824,
      "grad_norm": 0.25230876999441276,
      "learning_rate": 5.286340696085709e-05,
      "loss": 0.09407860040664673,
      "step": 473
    },
    {
      "epoch": 2.144878324844369,
      "grad_norm": 0.22751802306615446,
      "learning_rate": 5.266612494239088e-05,
      "loss": 0.13089381158351898,
      "step": 474
    },
    {
      "epoch": 2.1494057724957556,
      "grad_norm": 0.2320769608199613,
      "learning_rate": 5.246880129018516e-05,
      "loss": 0.14164891839027405,
      "step": 475
    },
    {
      "epoch": 2.153933220147142,
      "grad_norm": 0.2438198816777789,
      "learning_rate": 5.227143908561145e-05,
      "loss": 0.13016922771930695,
      "step": 476
    },
    {
      "epoch": 2.1584606677985287,
      "grad_norm": 0.24106291314947187,
      "learning_rate": 5.207404141064334e-05,
      "loss": 0.11469609290361404,
      "step": 477
    },
    {
      "epoch": 2.1629881154499153,
      "grad_norm": 0.2417691338448816,
      "learning_rate": 5.187661134780829e-05,
      "loss": 0.10605259984731674,
      "step": 478
    },
    {
      "epoch": 2.1675155631013014,
      "grad_norm": 0.23120471522513697,
      "learning_rate": 5.1679151980139564e-05,
      "loss": 0.0900178849697113,
      "step": 479
    },
    {
      "epoch": 2.172043010752688,
      "grad_norm": 0.23818899603412247,
      "learning_rate": 5.148166639112799e-05,
      "loss": 0.12464714050292969,
      "step": 480
    },
    {
      "epoch": 2.1765704584040746,
      "grad_norm": 0.2435869489375957,
      "learning_rate": 5.128415766467392e-05,
      "loss": 0.14309638738632202,
      "step": 481
    },
    {
      "epoch": 2.181097906055461,
      "grad_norm": 0.2550388153795191,
      "learning_rate": 5.1086628885038946e-05,
      "loss": 0.1305043250322342,
      "step": 482
    },
    {
      "epoch": 2.1856253537068477,
      "grad_norm": 0.2748501673226725,
      "learning_rate": 5.0889083136797875e-05,
      "loss": 0.1189241111278534,
      "step": 483
    },
    {
      "epoch": 2.1901528013582343,
      "grad_norm": 0.23791461412724066,
      "learning_rate": 5.0691523504790474e-05,
      "loss": 0.11003842949867249,
      "step": 484
    },
    {
      "epoch": 2.194680249009621,
      "grad_norm": 0.2571332410212008,
      "learning_rate": 5.049395307407329e-05,
      "loss": 0.09716764092445374,
      "step": 485
    },
    {
      "epoch": 2.1992076966610075,
      "grad_norm": 0.2384425630442099,
      "learning_rate": 5.029637492987153e-05,
      "loss": 0.12285800278186798,
      "step": 486
    },
    {
      "epoch": 2.203735144312394,
      "grad_norm": 0.23327436477533126,
      "learning_rate": 5.009879215753085e-05,
      "loss": 0.14675894379615784,
      "step": 487
    },
    {
      "epoch": 2.2082625919637806,
      "grad_norm": 0.23660300722487457,
      "learning_rate": 4.990120784246917e-05,
      "loss": 0.12845291197299957,
      "step": 488
    },
    {
      "epoch": 2.212790039615167,
      "grad_norm": 0.2529987599525421,
      "learning_rate": 4.970362507012848e-05,
      "loss": 0.11380189657211304,
      "step": 489
    },
    {
      "epoch": 2.2173174872665533,
      "grad_norm": 0.2257026219548231,
      "learning_rate": 4.950604692592672e-05,
      "loss": 0.10727782547473907,
      "step": 490
    },
    {
      "epoch": 2.22184493491794,
      "grad_norm": 0.24938286972845045,
      "learning_rate": 4.9308476495209544e-05,
      "loss": 0.09760070592164993,
      "step": 491
    },
    {
      "epoch": 2.2263723825693265,
      "grad_norm": 0.256045129385456,
      "learning_rate": 4.911091686320213e-05,
      "loss": 0.09533251821994781,
      "step": 492
    },
    {
      "epoch": 2.230899830220713,
      "grad_norm": 0.20731473595370586,
      "learning_rate": 4.891337111496107e-05,
      "loss": 0.1371329128742218,
      "step": 493
    },
    {
      "epoch": 2.2354272778720996,
      "grad_norm": 0.2342242439023739,
      "learning_rate": 4.87158423353261e-05,
      "loss": 0.13537269830703735,
      "step": 494
    },
    {
      "epoch": 2.239954725523486,
      "grad_norm": 0.26444134837772815,
      "learning_rate": 4.851833360887201e-05,
      "loss": 0.12884870171546936,
      "step": 495
    },
    {
      "epoch": 2.2444821731748728,
      "grad_norm": 0.23576112791818454,
      "learning_rate": 4.8320848019860454e-05,
      "loss": 0.11199313402175903,
      "step": 496
    },
    {
      "epoch": 2.2490096208262593,
      "grad_norm": 0.24720880815470767,
      "learning_rate": 4.8123388652191715e-05,
      "loss": 0.10179640352725983,
      "step": 497
    },
    {
      "epoch": 2.253537068477646,
      "grad_norm": 0.25523862330199726,
      "learning_rate": 4.7925958589356675e-05,
      "loss": 0.09644719958305359,
      "step": 498
    },
    {
      "epoch": 2.258064516129032,
      "grad_norm": 0.24865695878197794,
      "learning_rate": 4.7728560914388566e-05,
      "loss": 0.14680303633213043,
      "step": 499
    },
    {
      "epoch": 2.2625919637804186,
      "grad_norm": 0.23004145198330597,
      "learning_rate": 4.7531198709814854e-05,
      "loss": 0.14524027705192566,
      "step": 500
    },
    {
      "epoch": 2.267119411431805,
      "grad_norm": 0.279854048150932,
      "learning_rate": 4.7333875057609126e-05,
      "loss": 0.12221886962652206,
      "step": 501
    },
    {
      "epoch": 2.2716468590831917,
      "grad_norm": 0.2535516632210991,
      "learning_rate": 4.713659303914292e-05,
      "loss": 0.11302715539932251,
      "step": 502
    },
    {
      "epoch": 2.2761743067345783,
      "grad_norm": 0.24110440429763527,
      "learning_rate": 4.693935573513764e-05,
      "loss": 0.10129803419113159,
      "step": 503
    },
    {
      "epoch": 2.280701754385965,
      "grad_norm": 0.27220854726214855,
      "learning_rate": 4.674216622561644e-05,
      "loss": 0.09534797072410583,
      "step": 504
    },
    {
      "epoch": 2.2852292020373515,
      "grad_norm": 0.2512747146047165,
      "learning_rate": 4.654502758985611e-05,
      "loss": 0.13642863929271698,
      "step": 505
    },
    {
      "epoch": 2.289756649688738,
      "grad_norm": 0.2418677972815777,
      "learning_rate": 4.6347942906339015e-05,
      "loss": 0.14808616042137146,
      "step": 506
    },
    {
      "epoch": 2.2942840973401246,
      "grad_norm": 0.2567606605942382,
      "learning_rate": 4.615091525270501e-05,
      "loss": 0.12847869098186493,
      "step": 507
    },
    {
      "epoch": 2.298811544991511,
      "grad_norm": 0.2831968733809652,
      "learning_rate": 4.595394770570337e-05,
      "loss": 0.12522968649864197,
      "step": 508
    },
    {
      "epoch": 2.3033389926428978,
      "grad_norm": 0.2512488836803008,
      "learning_rate": 4.575704334114478e-05,
      "loss": 0.11159878224134445,
      "step": 509
    },
    {
      "epoch": 2.3078664402942843,
      "grad_norm": 0.2613370817591552,
      "learning_rate": 4.5560205233853266e-05,
      "loss": 0.09394368529319763,
      "step": 510
    },
    {
      "epoch": 2.3123938879456705,
      "grad_norm": 0.2967943489180437,
      "learning_rate": 4.5363436457618174e-05,
      "loss": 0.12963798642158508,
      "step": 511
    },
    {
      "epoch": 2.316921335597057,
      "grad_norm": 0.26304342645636347,
      "learning_rate": 4.516674008514623e-05,
      "loss": 0.15677320957183838,
      "step": 512
    },
    {
      "epoch": 2.3214487832484436,
      "grad_norm": 0.256715655421078,
      "learning_rate": 4.497011918801347e-05,
      "loss": 0.137124702334404,
      "step": 513
    },
    {
      "epoch": 2.32597623089983,
      "grad_norm": 0.3140837931142449,
      "learning_rate": 4.477357683661734e-05,
      "loss": 0.12650829553604126,
      "step": 514
    },
    {
      "epoch": 2.3305036785512168,
      "grad_norm": 0.28103324867288104,
      "learning_rate": 4.4577116100128735e-05,
      "loss": 0.1170065626502037,
      "step": 515
    },
    {
      "epoch": 2.3350311262026033,
      "grad_norm": 0.26198003618585153,
      "learning_rate": 4.4380740046444066e-05,
      "loss": 0.09977163374423981,
      "step": 516
    },
    {
      "epoch": 2.33955857385399,
      "grad_norm": 0.29320026049771986,
      "learning_rate": 4.418445174213734e-05,
      "loss": 0.09925702959299088,
      "step": 517
    },
    {
      "epoch": 2.3440860215053765,
      "grad_norm": 0.2585866257314453,
      "learning_rate": 4.39882542524123e-05,
      "loss": 0.13797470927238464,
      "step": 518
    },
    {
      "epoch": 2.348613469156763,
      "grad_norm": 0.2393657065001258,
      "learning_rate": 4.379215064105454e-05,
      "loss": 0.14010655879974365,
      "step": 519
    },
    {
      "epoch": 2.353140916808149,
      "grad_norm": 0.29148402803167184,
      "learning_rate": 4.3596143970383664e-05,
      "loss": 0.12116549909114838,
      "step": 520
    },
    {
      "epoch": 2.3576683644595358,
      "grad_norm": 0.27445223823357867,
      "learning_rate": 4.340023730120545e-05,
      "loss": 0.10586535930633545,
      "step": 521
    },
    {
      "epoch": 2.3621958121109223,
      "grad_norm": 0.25900953697788853,
      "learning_rate": 4.3204433692764096e-05,
      "loss": 0.09879538416862488,
      "step": 522
    },
    {
      "epoch": 2.366723259762309,
      "grad_norm": 0.2728929783602796,
      "learning_rate": 4.3008736202694414e-05,
      "loss": 0.093920037150383,
      "step": 523
    },
    {
      "epoch": 2.3712507074136955,
      "grad_norm": 0.24512095999370934,
      "learning_rate": 4.281314788697408e-05,
      "loss": 0.1439329981803894,
      "step": 524
    },
    {
      "epoch": 2.375778155065082,
      "grad_norm": 0.2361995166682201,
      "learning_rate": 4.2617671799875944e-05,
      "loss": 0.14218634366989136,
      "step": 525
    },
    {
      "epoch": 2.3803056027164686,
      "grad_norm": 0.24406215264374898,
      "learning_rate": 4.242231099392029e-05,
      "loss": 0.12147480249404907,
      "step": 526
    },
    {
      "epoch": 2.384833050367855,
      "grad_norm": 0.2520795253864757,
      "learning_rate": 4.222706851982721e-05,
      "loss": 0.10550867021083832,
      "step": 527
    },
    {
      "epoch": 2.3893604980192418,
      "grad_norm": 0.27284713189389487,
      "learning_rate": 4.203194742646893e-05,
      "loss": 0.11013522744178772,
      "step": 528
    },
    {
      "epoch": 2.3938879456706283,
      "grad_norm": 0.23626266673987192,
      "learning_rate": 4.1836950760822235e-05,
      "loss": 0.09375832974910736,
      "step": 529
    },
    {
      "epoch": 2.398415393322015,
      "grad_norm": 0.28795086578857015,
      "learning_rate": 4.1642081567920846e-05,
      "loss": 0.14555881917476654,
      "step": 530
    },
    {
      "epoch": 2.402942840973401,
      "grad_norm": 0.28861629308942055,
      "learning_rate": 4.144734289080791e-05,
      "loss": 0.15418826043605804,
      "step": 531
    },
    {
      "epoch": 2.4074702886247876,
      "grad_norm": 0.23754273031391096,
      "learning_rate": 4.125273777048845e-05,
      "loss": 0.13065479695796967,
      "step": 532
    },
    {
      "epoch": 2.411997736276174,
      "grad_norm": 0.254132024369517,
      "learning_rate": 4.1058269245881896e-05,
      "loss": 0.12865102291107178,
      "step": 533
    },
    {
      "epoch": 2.4165251839275608,
      "grad_norm": 0.27102840717897925,
      "learning_rate": 4.086394035377463e-05,
      "loss": 0.10483617335557938,
      "step": 534
    },
    {
      "epoch": 2.4210526315789473,
      "grad_norm": 0.24916755528181858,
      "learning_rate": 4.066975412877255e-05,
      "loss": 0.09119851142168045,
      "step": 535
    },
    {
      "epoch": 2.425580079230334,
      "grad_norm": 0.2331903011344317,
      "learning_rate": 4.04757136032537e-05,
      "loss": 0.10368061810731888,
      "step": 536
    },
    {
      "epoch": 2.4301075268817205,
      "grad_norm": 0.2507431798769252,
      "learning_rate": 4.028182180732088e-05,
      "loss": 0.14746174216270447,
      "step": 537
    },
    {
      "epoch": 2.434634974533107,
      "grad_norm": 0.2956217123784412,
      "learning_rate": 4.0088081768754365e-05,
      "loss": 0.12845945358276367,
      "step": 538
    },
    {
      "epoch": 2.4391624221844936,
      "grad_norm": 0.25429438138428373,
      "learning_rate": 3.9894496512964635e-05,
      "loss": 0.11777739971876144,
      "step": 539
    },
    {
      "epoch": 2.44368986983588,
      "grad_norm": 0.24156815633641798,
      "learning_rate": 3.970106906294509e-05,
      "loss": 0.1058419942855835,
      "step": 540
    },
    {
      "epoch": 2.4482173174872663,
      "grad_norm": 0.27220037685566956,
      "learning_rate": 3.950780243922486e-05,
      "loss": 0.0915253609418869,
      "step": 541
    },
    {
      "epoch": 2.452744765138653,
      "grad_norm": 0.24280319564635572,
      "learning_rate": 3.9314699659821666e-05,
      "loss": 0.09519818425178528,
      "step": 542
    },
    {
      "epoch": 2.4572722127900395,
      "grad_norm": 0.21923038682955656,
      "learning_rate": 3.9121763740194616e-05,
      "loss": 0.13519631326198578,
      "step": 543
    },
    {
      "epoch": 2.461799660441426,
      "grad_norm": 0.2488082494626004,
      "learning_rate": 3.892899769319722e-05,
      "loss": 0.13617262244224548,
      "step": 544
    },
    {
      "epoch": 2.4663271080928126,
      "grad_norm": 0.305635107885523,
      "learning_rate": 3.873640452903026e-05,
      "loss": 0.1264575570821762,
      "step": 545
    },
    {
      "epoch": 2.470854555744199,
      "grad_norm": 0.23570314768564468,
      "learning_rate": 3.854398725519479e-05,
      "loss": 0.11620922386646271,
      "step": 546
    },
    {
      "epoch": 2.475382003395586,
      "grad_norm": 0.24778946270413518,
      "learning_rate": 3.835174887644523e-05,
      "loss": 0.0968777984380722,
      "step": 547
    },
    {
      "epoch": 2.4799094510469724,
      "grad_norm": 0.296885121447023,
      "learning_rate": 3.815969239474238e-05,
      "loss": 0.09505044668912888,
      "step": 548
    },
    {
      "epoch": 2.484436898698359,
      "grad_norm": 0.25294822929702704,
      "learning_rate": 3.796782080920659e-05,
      "loss": 0.14779812097549438,
      "step": 549
    },
    {
      "epoch": 2.4889643463497455,
      "grad_norm": 0.21256636916880545,
      "learning_rate": 3.777613711607087e-05,
      "loss": 0.14114922285079956,
      "step": 550
    },
    {
      "epoch": 2.493491794001132,
      "grad_norm": 0.29488223629139126,
      "learning_rate": 3.758464430863417e-05,
      "loss": 0.12959524989128113,
      "step": 551
    },
    {
      "epoch": 2.498019241652518,
      "grad_norm": 0.27289699128683415,
      "learning_rate": 3.7393345377214586e-05,
      "loss": 0.11985202878713608,
      "step": 552
    },
    {
      "epoch": 2.502546689303905,
      "grad_norm": 0.22406212900795036,
      "learning_rate": 3.720224330910268e-05,
      "loss": 0.1026943251490593,
      "step": 553
    },
    {
      "epoch": 2.5070741369552914,
      "grad_norm": 0.2859349746609221,
      "learning_rate": 3.701134108851483e-05,
      "loss": 0.08981700241565704,
      "step": 554
    },
    {
      "epoch": 2.511601584606678,
      "grad_norm": 0.28214778358297454,
      "learning_rate": 3.682064169654663e-05,
      "loss": 0.13117776811122894,
      "step": 555
    },
    {
      "epoch": 2.5161290322580645,
      "grad_norm": 0.21994347959459834,
      "learning_rate": 3.663014811112634e-05,
      "loss": 0.14276379346847534,
      "step": 556
    },
    {
      "epoch": 2.520656479909451,
      "grad_norm": 0.2594577747985432,
      "learning_rate": 3.6439863306968395e-05,
      "loss": 0.12485526502132416,
      "step": 557
    },
    {
      "epoch": 2.5251839275608376,
      "grad_norm": 0.2961625882311709,
      "learning_rate": 3.6249790255526915e-05,
      "loss": 0.11515647172927856,
      "step": 558
    },
    {
      "epoch": 2.5297113752122242,
      "grad_norm": 0.24683329504623078,
      "learning_rate": 3.605993192494934e-05,
      "loss": 0.10419435799121857,
      "step": 559
    },
    {
      "epoch": 2.534238822863611,
      "grad_norm": 0.25364083068174226,
      "learning_rate": 3.587029128003006e-05,
      "loss": 0.09284784644842148,
      "step": 560
    },
    {
      "epoch": 2.538766270514997,
      "grad_norm": 0.2585483974392137,
      "learning_rate": 3.5680871282164144e-05,
      "loss": 0.11332575976848602,
      "step": 561
    },
    {
      "epoch": 2.5432937181663835,
      "grad_norm": 0.24237379986215338,
      "learning_rate": 3.549167488930103e-05,
      "loss": 0.14688560366630554,
      "step": 562
    },
    {
      "epoch": 2.54782116581777,
      "grad_norm": 0.22638908812977684,
      "learning_rate": 3.5302705055898425e-05,
      "loss": 0.13564583659172058,
      "step": 563
    },
    {
      "epoch": 2.5523486134691566,
      "grad_norm": 0.2596181991576638,
      "learning_rate": 3.5113964732876106e-05,
      "loss": 0.115118108689785,
      "step": 564
    },
    {
      "epoch": 2.556876061120543,
      "grad_norm": 0.26415733368451977,
      "learning_rate": 3.492545686756986e-05,
      "loss": 0.12011849135160446,
      "step": 565
    },
    {
      "epoch": 2.56140350877193,
      "grad_norm": 0.23714192087858377,
      "learning_rate": 3.473718440368544e-05,
      "loss": 0.10161202400922775,
      "step": 566
    },
    {
      "epoch": 2.5659309564233164,
      "grad_norm": 0.27025275979195007,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 0.0882958248257637,
      "step": 567
    },
    {
      "epoch": 2.570458404074703,
      "grad_norm": 0.23879461948112587,
      "learning_rate": 3.4361357436579316e-05,
      "loss": 0.14357197284698486,
      "step": 568
    },
    {
      "epoch": 2.5749858517260895,
      "grad_norm": 0.23316769041494662,
      "learning_rate": 3.417380880220563e-05,
      "loss": 0.13415873050689697,
      "step": 569
    },
    {
      "epoch": 2.579513299377476,
      "grad_norm": 0.27654361242855297,
      "learning_rate": 3.3986507306858125e-05,
      "loss": 0.12432755529880524,
      "step": 570
    },
    {
      "epoch": 2.5840407470288627,
      "grad_norm": 0.2567188323738886,
      "learning_rate": 3.379945587540414e-05,
      "loss": 0.10620371252298355,
      "step": 571
    },
    {
      "epoch": 2.5885681946802492,
      "grad_norm": 0.23508221202812904,
      "learning_rate": 3.361265742880599e-05,
      "loss": 0.09938931465148926,
      "step": 572
    },
    {
      "epoch": 2.593095642331636,
      "grad_norm": 0.26833749975488186,
      "learning_rate": 3.342611488407549e-05,
      "loss": 0.09262629598379135,
      "step": 573
    },
    {
      "epoch": 2.597623089983022,
      "grad_norm": 0.24976917871377982,
      "learning_rate": 3.323983115422827e-05,
      "loss": 0.12932151556015015,
      "step": 574
    },
    {
      "epoch": 2.6021505376344085,
      "grad_norm": 0.20685234324195712,
      "learning_rate": 3.3053809148238426e-05,
      "loss": 0.13880762457847595,
      "step": 575
    },
    {
      "epoch": 2.606677985285795,
      "grad_norm": 0.24401216152089034,
      "learning_rate": 3.286805177099293e-05,
      "loss": 0.12213125824928284,
      "step": 576
    },
    {
      "epoch": 2.6112054329371817,
      "grad_norm": 0.26388750626149415,
      "learning_rate": 3.268256192324647e-05,
      "loss": 0.11493176221847534,
      "step": 577
    },
    {
      "epoch": 2.6157328805885682,
      "grad_norm": 0.22428771518696666,
      "learning_rate": 3.249734250157592e-05,
      "loss": 0.09479294717311859,
      "step": 578
    },
    {
      "epoch": 2.620260328239955,
      "grad_norm": 0.24057632420007685,
      "learning_rate": 3.231239639833531e-05,
      "loss": 0.08547629415988922,
      "step": 579
    },
    {
      "epoch": 2.6247877758913414,
      "grad_norm": 0.2336601965665524,
      "learning_rate": 3.212772650161056e-05,
      "loss": 0.1252753734588623,
      "step": 580
    },
    {
      "epoch": 2.629315223542728,
      "grad_norm": 0.24165240309289301,
      "learning_rate": 3.1943335695174365e-05,
      "loss": 0.1408216655254364,
      "step": 581
    },
    {
      "epoch": 2.633842671194114,
      "grad_norm": 0.24033546379107754,
      "learning_rate": 3.175922685844125e-05,
      "loss": 0.12671208381652832,
      "step": 582
    },
    {
      "epoch": 2.6383701188455007,
      "grad_norm": 0.2356329228609367,
      "learning_rate": 3.15754028664225e-05,
      "loss": 0.11594489961862564,
      "step": 583
    },
    {
      "epoch": 2.6428975664968872,
      "grad_norm": 0.2644332823548845,
      "learning_rate": 3.1391866589681346e-05,
      "loss": 0.10527190566062927,
      "step": 584
    },
    {
      "epoch": 2.647425014148274,
      "grad_norm": 0.24214421981778728,
      "learning_rate": 3.12086208942881e-05,
      "loss": 0.0941970944404602,
      "step": 585
    },
    {
      "epoch": 2.6519524617996604,
      "grad_norm": 0.24680728714282923,
      "learning_rate": 3.10256686417754e-05,
      "loss": 0.12475133687257767,
      "step": 586
    },
    {
      "epoch": 2.656479909451047,
      "grad_norm": 0.25121624638306916,
      "learning_rate": 3.084301268909353e-05,
      "loss": 0.15046575665473938,
      "step": 587
    },
    {
      "epoch": 2.6610073571024335,
      "grad_norm": 0.2418561721157581,
      "learning_rate": 3.0660655888565825e-05,
      "loss": 0.13163983821868896,
      "step": 588
    },
    {
      "epoch": 2.66553480475382,
      "grad_norm": 0.24347981203163183,
      "learning_rate": 3.0478601087844096e-05,
      "loss": 0.11440954357385635,
      "step": 589
    },
    {
      "epoch": 2.6700622524052067,
      "grad_norm": 0.2649101408245476,
      "learning_rate": 3.0296851129864168e-05,
      "loss": 0.11191365122795105,
      "step": 590
    },
    {
      "epoch": 2.6745897000565932,
      "grad_norm": 0.2585452012810636,
      "learning_rate": 3.0115408852801535e-05,
      "loss": 0.09637647867202759,
      "step": 591
    },
    {
      "epoch": 2.67911714770798,
      "grad_norm": 0.2518425045559207,
      "learning_rate": 2.9934277090026964e-05,
      "loss": 0.09529514610767365,
      "step": 592
    },
    {
      "epoch": 2.6836445953593664,
      "grad_norm": 0.22520047038788044,
      "learning_rate": 2.97534586700623e-05,
      "loss": 0.13834667205810547,
      "step": 593
    },
    {
      "epoch": 2.688172043010753,
      "grad_norm": 0.25173496296604714,
      "learning_rate": 2.9572956416536267e-05,
      "loss": 0.13480937480926514,
      "step": 594
    },
    {
      "epoch": 2.692699490662139,
      "grad_norm": 0.2410585900956496,
      "learning_rate": 2.9392773148140408e-05,
      "loss": 0.11652663350105286,
      "step": 595
    },
    {
      "epoch": 2.6972269383135257,
      "grad_norm": 0.23487256508276858,
      "learning_rate": 2.9212911678585043e-05,
      "loss": 0.11234890669584274,
      "step": 596
    },
    {
      "epoch": 2.7017543859649122,
      "grad_norm": 0.2270191951541322,
      "learning_rate": 2.9033374816555338e-05,
      "loss": 0.09663751721382141,
      "step": 597
    },
    {
      "epoch": 2.706281833616299,
      "grad_norm": 0.25197285773571976,
      "learning_rate": 2.885416536566744e-05,
      "loss": 0.08802421391010284,
      "step": 598
    },
    {
      "epoch": 2.7108092812676854,
      "grad_norm": 0.26157029463529385,
      "learning_rate": 2.8675286124424693e-05,
      "loss": 0.13833507895469666,
      "step": 599
    },
    {
      "epoch": 2.715336728919072,
      "grad_norm": 0.21057652362655196,
      "learning_rate": 2.8496739886173995e-05,
      "loss": 0.13772332668304443,
      "step": 600
    },
    {
      "epoch": 2.7198641765704585,
      "grad_norm": 0.2630087930460932,
      "learning_rate": 2.8318529439062035e-05,
      "loss": 0.12039318680763245,
      "step": 601
    },
    {
      "epoch": 2.7243916242218447,
      "grad_norm": 0.26268925166743773,
      "learning_rate": 2.8140657565991958e-05,
      "loss": 0.1143353208899498,
      "step": 602
    },
    {
      "epoch": 2.7289190718732312,
      "grad_norm": 0.21784795757319134,
      "learning_rate": 2.7963127044579697e-05,
      "loss": 0.10443739593029022,
      "step": 603
    },
    {
      "epoch": 2.733446519524618,
      "grad_norm": 0.24171040540762112,
      "learning_rate": 2.7785940647110763e-05,
      "loss": 0.08580689132213593,
      "step": 604
    },
    {
      "epoch": 2.7379739671760044,
      "grad_norm": 0.25017672282012504,
      "learning_rate": 2.7609101140496863e-05,
      "loss": 0.1266556680202484,
      "step": 605
    },
    {
      "epoch": 2.742501414827391,
      "grad_norm": 0.22426404482783405,
      "learning_rate": 2.743261128623269e-05,
      "loss": 0.13873793184757233,
      "step": 606
    },
    {
      "epoch": 2.7470288624787775,
      "grad_norm": 0.21232226494818524,
      "learning_rate": 2.72564738403529e-05,
      "loss": 0.12767687439918518,
      "step": 607
    },
    {
      "epoch": 2.751556310130164,
      "grad_norm": 0.2495474667171734,
      "learning_rate": 2.708069155338892e-05,
      "loss": 0.11374086141586304,
      "step": 608
    },
    {
      "epoch": 2.7560837577815507,
      "grad_norm": 0.23903618675797153,
      "learning_rate": 2.6905267170326143e-05,
      "loss": 0.10475228726863861,
      "step": 609
    },
    {
      "epoch": 2.7606112054329373,
      "grad_norm": 0.22385101714825797,
      "learning_rate": 2.6730203430560947e-05,
      "loss": 0.09183211624622345,
      "step": 610
    },
    {
      "epoch": 2.765138653084324,
      "grad_norm": 0.2255108306449036,
      "learning_rate": 2.6555503067858013e-05,
      "loss": 0.120794378221035,
      "step": 611
    },
    {
      "epoch": 2.7696661007357104,
      "grad_norm": 0.21818161061861707,
      "learning_rate": 2.6381168810307533e-05,
      "loss": 0.13886579871177673,
      "step": 612
    },
    {
      "epoch": 2.774193548387097,
      "grad_norm": 0.2162769870443805,
      "learning_rate": 2.6207203380282747e-05,
      "loss": 0.1283240020275116,
      "step": 613
    },
    {
      "epoch": 2.7787209960384835,
      "grad_norm": 0.2358741304529575,
      "learning_rate": 2.603360949439727e-05,
      "loss": 0.121439628303051,
      "step": 614
    },
    {
      "epoch": 2.7832484436898697,
      "grad_norm": 0.2276407496979808,
      "learning_rate": 2.5860389863462765e-05,
      "loss": 0.10243870317935944,
      "step": 615
    },
    {
      "epoch": 2.7877758913412563,
      "grad_norm": 0.22912652601551836,
      "learning_rate": 2.5687547192446647e-05,
      "loss": 0.09506306052207947,
      "step": 616
    },
    {
      "epoch": 2.792303338992643,
      "grad_norm": 0.22281328036060236,
      "learning_rate": 2.5515084180429716e-05,
      "loss": 0.09046522527933121,
      "step": 617
    },
    {
      "epoch": 2.7968307866440294,
      "grad_norm": 0.1976431350430884,
      "learning_rate": 2.5343003520564158e-05,
      "loss": 0.13864973187446594,
      "step": 618
    },
    {
      "epoch": 2.801358234295416,
      "grad_norm": 0.20883640126867217,
      "learning_rate": 2.5171307900031345e-05,
      "loss": 0.13139835000038147,
      "step": 619
    },
    {
      "epoch": 2.8058856819468025,
      "grad_norm": 0.2125190189472025,
      "learning_rate": 2.500000000000001e-05,
      "loss": 0.1208663210272789,
      "step": 620
    },
    {
      "epoch": 2.810413129598189,
      "grad_norm": 0.21228317639875047,
      "learning_rate": 2.4829082495584242e-05,
      "loss": 0.10401903092861176,
      "step": 621
    },
    {
      "epoch": 2.8149405772495757,
      "grad_norm": 0.21279150723283488,
      "learning_rate": 2.465855805580185e-05,
      "loss": 0.09000042080879211,
      "step": 622
    },
    {
      "epoch": 2.819468024900962,
      "grad_norm": 0.2154703660458778,
      "learning_rate": 2.448842934353256e-05,
      "loss": 0.09436316788196564,
      "step": 623
    },
    {
      "epoch": 2.8239954725523484,
      "grad_norm": 0.18468130064262714,
      "learning_rate": 2.4318699015476493e-05,
      "loss": 0.12485232949256897,
      "step": 624
    },
    {
      "epoch": 2.828522920203735,
      "grad_norm": 0.2118189050847499,
      "learning_rate": 2.414936972211272e-05,
      "loss": 0.13756975531578064,
      "step": 625
    },
    {
      "epoch": 2.8330503678551215,
      "grad_norm": 0.2008637254946331,
      "learning_rate": 2.3980444107657747e-05,
      "loss": 0.1262628734111786,
      "step": 626
    },
    {
      "epoch": 2.837577815506508,
      "grad_norm": 0.19507177543957294,
      "learning_rate": 2.3811924810024384e-05,
      "loss": 0.11262847483158112,
      "step": 627
    },
    {
      "epoch": 2.8421052631578947,
      "grad_norm": 0.2160150070627766,
      "learning_rate": 2.3643814460780394e-05,
      "loss": 0.09914860129356384,
      "step": 628
    },
    {
      "epoch": 2.8466327108092813,
      "grad_norm": 0.22322489364690976,
      "learning_rate": 2.347611568510754e-05,
      "loss": 0.08965691924095154,
      "step": 629
    },
    {
      "epoch": 2.851160158460668,
      "grad_norm": 0.18405705268568368,
      "learning_rate": 2.3308831101760486e-05,
      "loss": 0.12905165553092957,
      "step": 630
    },
    {
      "epoch": 2.8556876061120544,
      "grad_norm": 0.1954145806659638,
      "learning_rate": 2.3141963323025916e-05,
      "loss": 0.14251145720481873,
      "step": 631
    },
    {
      "epoch": 2.860215053763441,
      "grad_norm": 0.2550894717469498,
      "learning_rate": 2.2975514954681838e-05,
      "loss": 0.1315610110759735,
      "step": 632
    },
    {
      "epoch": 2.8647425014148276,
      "grad_norm": 0.19866802480987966,
      "learning_rate": 2.2809488595956745e-05,
      "loss": 0.10726577043533325,
      "step": 633
    },
    {
      "epoch": 2.869269949066214,
      "grad_norm": 0.2059757913242525,
      "learning_rate": 2.264388683948918e-05,
      "loss": 0.09670988470315933,
      "step": 634
    },
    {
      "epoch": 2.8737973967176007,
      "grad_norm": 0.2209737989570976,
      "learning_rate": 2.247871227128709e-05,
      "loss": 0.09051385521888733,
      "step": 635
    },
    {
      "epoch": 2.878324844368987,
      "grad_norm": 0.19723744969911808,
      "learning_rate": 2.2313967470687593e-05,
      "loss": 0.11008135974407196,
      "step": 636
    },
    {
      "epoch": 2.8828522920203734,
      "grad_norm": 0.18491495577394462,
      "learning_rate": 2.2149655010316573e-05,
      "loss": 0.14236138761043549,
      "step": 637
    },
    {
      "epoch": 2.88737973967176,
      "grad_norm": 0.20555974856745457,
      "learning_rate": 2.1985777456048633e-05,
      "loss": 0.12351356446743011,
      "step": 638
    },
    {
      "epoch": 2.8919071873231466,
      "grad_norm": 0.20962920142081726,
      "learning_rate": 2.1822337366966898e-05,
      "loss": 0.11412464082241058,
      "step": 639
    },
    {
      "epoch": 2.896434634974533,
      "grad_norm": 0.18352752004787823,
      "learning_rate": 2.1659337295323118e-05,
      "loss": 0.10024420917034149,
      "step": 640
    },
    {
      "epoch": 2.9009620826259197,
      "grad_norm": 0.1961851446672569,
      "learning_rate": 2.149677978649786e-05,
      "loss": 0.08953743427991867,
      "step": 641
    },
    {
      "epoch": 2.9054895302773063,
      "grad_norm": 0.22264124964898163,
      "learning_rate": 2.1334667378960644e-05,
      "loss": 0.08613376319408417,
      "step": 642
    },
    {
      "epoch": 2.910016977928693,
      "grad_norm": 0.1841841809171048,
      "learning_rate": 2.1173002604230425e-05,
      "loss": 0.12837599217891693,
      "step": 643
    },
    {
      "epoch": 2.914544425580079,
      "grad_norm": 0.17570198473311538,
      "learning_rate": 2.1011787986835934e-05,
      "loss": 0.13058477640151978,
      "step": 644
    },
    {
      "epoch": 2.9190718732314656,
      "grad_norm": 0.1890424121279899,
      "learning_rate": 2.0851026044276406e-05,
      "loss": 0.11193414777517319,
      "step": 645
    },
    {
      "epoch": 2.923599320882852,
      "grad_norm": 0.2037958407485097,
      "learning_rate": 2.0690719286982125e-05,
      "loss": 0.10173594951629639,
      "step": 646
    },
    {
      "epoch": 2.9281267685342387,
      "grad_norm": 0.19365613909836157,
      "learning_rate": 2.0530870218275273e-05,
      "loss": 0.10059279203414917,
      "step": 647
    },
    {
      "epoch": 2.9326542161856253,
      "grad_norm": 0.18313628206502447,
      "learning_rate": 2.0371481334330912e-05,
      "loss": 0.08752919733524323,
      "step": 648
    },
    {
      "epoch": 2.937181663837012,
      "grad_norm": 0.18163662587745502,
      "learning_rate": 2.0212555124137866e-05,
      "loss": 0.1352725774049759,
      "step": 649
    },
    {
      "epoch": 2.9417091114883984,
      "grad_norm": 0.18305358707796393,
      "learning_rate": 2.005409406946e-05,
      "loss": 0.1280105859041214,
      "step": 650
    },
    {
      "epoch": 2.946236559139785,
      "grad_norm": 0.1695870301850634,
      "learning_rate": 1.9896100644797317e-05,
      "loss": 0.11544548720121384,
      "step": 651
    },
    {
      "epoch": 2.9507640067911716,
      "grad_norm": 0.178212092154041,
      "learning_rate": 1.973857731734746e-05,
      "loss": 0.10372395813465118,
      "step": 652
    },
    {
      "epoch": 2.955291454442558,
      "grad_norm": 0.20017714089350316,
      "learning_rate": 1.958152654696705e-05,
      "loss": 0.09953290224075317,
      "step": 653
    },
    {
      "epoch": 2.9598189020939447,
      "grad_norm": 0.19414570777122794,
      "learning_rate": 1.942495078613341e-05,
      "loss": 0.08335710316896439,
      "step": 654
    },
    {
      "epoch": 2.9643463497453313,
      "grad_norm": 0.16077247309449752,
      "learning_rate": 1.9268852479906147e-05,
      "loss": 0.11613577604293823,
      "step": 655
    },
    {
      "epoch": 2.968873797396718,
      "grad_norm": 0.16858390725017094,
      "learning_rate": 1.9113234065889014e-05,
      "loss": 0.14055851101875305,
      "step": 656
    },
    {
      "epoch": 2.973401245048104,
      "grad_norm": 0.179867973624298,
      "learning_rate": 1.8958097974191907e-05,
      "loss": 0.11342918127775192,
      "step": 657
    },
    {
      "epoch": 2.9779286926994906,
      "grad_norm": 0.1747443274776837,
      "learning_rate": 1.8803446627392797e-05,
      "loss": 0.11532742530107498,
      "step": 658
    },
    {
      "epoch": 2.982456140350877,
      "grad_norm": 0.1840967023486453,
      "learning_rate": 1.8649282440500015e-05,
      "loss": 0.09636199474334717,
      "step": 659
    },
    {
      "epoch": 2.9869835880022637,
      "grad_norm": 0.19968514929505088,
      "learning_rate": 1.849560782091445e-05,
      "loss": 0.08977682888507843,
      "step": 660
    },
    {
      "epoch": 2.9915110356536503,
      "grad_norm": 0.18660833054019316,
      "learning_rate": 1.834242516839203e-05,
      "loss": 0.11185477674007416,
      "step": 661
    },
    {
      "epoch": 2.996038483305037,
      "grad_norm": 0.17623246563776465,
      "learning_rate": 1.8189736875006185e-05,
      "loss": 0.1226733922958374,
      "step": 662
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.21161133741890636,
      "learning_rate": 1.8037545325110504e-05,
      "loss": 0.08953028917312622,
      "step": 663
    },
    {
      "epoch": 3.0045274476513866,
      "grad_norm": 0.3066364156494326,
      "learning_rate": 1.788585289530158e-05,
      "loss": 0.09976173937320709,
      "step": 664
    },
    {
      "epoch": 3.009054895302773,
      "grad_norm": 0.2618992530713181,
      "learning_rate": 1.7734661954381754e-05,
      "loss": 0.08678863942623138,
      "step": 665
    },
    {
      "epoch": 3.0135823429541597,
      "grad_norm": 0.19907300376185252,
      "learning_rate": 1.7583974863322274e-05,
      "loss": 0.06767494976520538,
      "step": 666
    },
    {
      "epoch": 3.0181097906055463,
      "grad_norm": 0.18817368766730283,
      "learning_rate": 1.7433793975226298e-05,
      "loss": 0.0600329264998436,
      "step": 667
    },
    {
      "epoch": 3.022637238256933,
      "grad_norm": 0.2061340158138551,
      "learning_rate": 1.728412163529227e-05,
      "loss": 0.0469946414232254,
      "step": 668
    },
    {
      "epoch": 3.027164685908319,
      "grad_norm": 0.23783508890294386,
      "learning_rate": 1.713496018077717e-05,
      "loss": 0.04690144211053848,
      "step": 669
    },
    {
      "epoch": 3.0316921335597056,
      "grad_norm": 0.43698570813754956,
      "learning_rate": 1.6986311940960147e-05,
      "loss": 0.09764593839645386,
      "step": 670
    },
    {
      "epoch": 3.036219581211092,
      "grad_norm": 0.32154495167200836,
      "learning_rate": 1.6838179237106016e-05,
      "loss": 0.08996790647506714,
      "step": 671
    },
    {
      "epoch": 3.0407470288624787,
      "grad_norm": 0.19989528083823166,
      "learning_rate": 1.66905643824291e-05,
      "loss": 0.0713300108909607,
      "step": 672
    },
    {
      "epoch": 3.0452744765138653,
      "grad_norm": 0.2103161752515197,
      "learning_rate": 1.6543469682057106e-05,
      "loss": 0.062314074486494064,
      "step": 673
    },
    {
      "epoch": 3.049801924165252,
      "grad_norm": 0.21310080332783007,
      "learning_rate": 1.6396897432995044e-05,
      "loss": 0.05069063603878021,
      "step": 674
    },
    {
      "epoch": 3.0543293718166384,
      "grad_norm": 0.2374684195733413,
      "learning_rate": 1.6250849924089484e-05,
      "loss": 0.043403737246990204,
      "step": 675
    },
    {
      "epoch": 3.058856819468025,
      "grad_norm": 0.21904849526165582,
      "learning_rate": 1.6105329435992682e-05,
      "loss": 0.0834536924958229,
      "step": 676
    },
    {
      "epoch": 3.0633842671194116,
      "grad_norm": 0.19383275443870573,
      "learning_rate": 1.5960338241127093e-05,
      "loss": 0.08995246887207031,
      "step": 677
    },
    {
      "epoch": 3.067911714770798,
      "grad_norm": 0.18879861173936469,
      "learning_rate": 1.581587860364977e-05,
      "loss": 0.07954228669404984,
      "step": 678
    },
    {
      "epoch": 3.0724391624221843,
      "grad_norm": 0.19092381008877815,
      "learning_rate": 1.567195277941706e-05,
      "loss": 0.06363485753536224,
      "step": 679
    },
    {
      "epoch": 3.076966610073571,
      "grad_norm": 0.19424073278225265,
      "learning_rate": 1.552856301594942e-05,
      "loss": 0.05385269224643707,
      "step": 680
    },
    {
      "epoch": 3.0814940577249574,
      "grad_norm": 0.18062190907689793,
      "learning_rate": 1.5385711552396227e-05,
      "loss": 0.04273997247219086,
      "step": 681
    },
    {
      "epoch": 3.086021505376344,
      "grad_norm": 0.19216670245288398,
      "learning_rate": 1.5243400619500903e-05,
      "loss": 0.06583520025014877,
      "step": 682
    },
    {
      "epoch": 3.0905489530277306,
      "grad_norm": 0.25510602354092327,
      "learning_rate": 1.5101632439565998e-05,
      "loss": 0.09310319274663925,
      "step": 683
    },
    {
      "epoch": 3.095076400679117,
      "grad_norm": 0.21622644779417574,
      "learning_rate": 1.4960409226418576e-05,
      "loss": 0.07124336808919907,
      "step": 684
    },
    {
      "epoch": 3.0996038483305037,
      "grad_norm": 0.20132594716382185,
      "learning_rate": 1.4819733185375534e-05,
      "loss": 0.06657753884792328,
      "step": 685
    },
    {
      "epoch": 3.1041312959818903,
      "grad_norm": 0.17173972561069642,
      "learning_rate": 1.4679606513209283e-05,
      "loss": 0.050332531332969666,
      "step": 686
    },
    {
      "epoch": 3.108658743633277,
      "grad_norm": 0.17209769554298748,
      "learning_rate": 1.4540031398113335e-05,
      "loss": 0.04407206177711487,
      "step": 687
    },
    {
      "epoch": 3.1131861912846635,
      "grad_norm": 0.1727555707541188,
      "learning_rate": 1.4401010019668226e-05,
      "loss": 0.0437740758061409,
      "step": 688
    },
    {
      "epoch": 3.11771363893605,
      "grad_norm": 0.17590697354246457,
      "learning_rate": 1.4262544548807432e-05,
      "loss": 0.08785104006528854,
      "step": 689
    },
    {
      "epoch": 3.122241086587436,
      "grad_norm": 0.17646701046970592,
      "learning_rate": 1.4124637147783432e-05,
      "loss": 0.08701614290475845,
      "step": 690
    },
    {
      "epoch": 3.1267685342388227,
      "grad_norm": 0.16910237298120426,
      "learning_rate": 1.3987289970134049e-05,
      "loss": 0.07015383988618851,
      "step": 691
    },
    {
      "epoch": 3.1312959818902093,
      "grad_norm": 0.16674344676254058,
      "learning_rate": 1.3850505160648709e-05,
      "loss": 0.06920076161623001,
      "step": 692
    },
    {
      "epoch": 3.135823429541596,
      "grad_norm": 0.1720701281820272,
      "learning_rate": 1.371428485533498e-05,
      "loss": 0.04850110039114952,
      "step": 693
    },
    {
      "epoch": 3.1403508771929824,
      "grad_norm": 0.17777914669888842,
      "learning_rate": 1.3578631181385305e-05,
      "loss": 0.04546038806438446,
      "step": 694
    },
    {
      "epoch": 3.144878324844369,
      "grad_norm": 0.17040333469142502,
      "learning_rate": 1.3443546257143624e-05,
      "loss": 0.0866042971611023,
      "step": 695
    },
    {
      "epoch": 3.1494057724957556,
      "grad_norm": 0.17121685270160847,
      "learning_rate": 1.3309032192072463e-05,
      "loss": 0.07414685934782028,
      "step": 696
    },
    {
      "epoch": 3.153933220147142,
      "grad_norm": 0.17655076877248685,
      "learning_rate": 1.3175091086719832e-05,
      "loss": 0.07138119637966156,
      "step": 697
    },
    {
      "epoch": 3.1584606677985287,
      "grad_norm": 0.16472616782119096,
      "learning_rate": 1.304172503268658e-05,
      "loss": 0.0625188946723938,
      "step": 698
    },
    {
      "epoch": 3.1629881154499153,
      "grad_norm": 0.16364050214385314,
      "learning_rate": 1.29089361125936e-05,
      "loss": 0.04717274382710457,
      "step": 699
    },
    {
      "epoch": 3.1675155631013014,
      "grad_norm": 0.17462613946871813,
      "learning_rate": 1.277672640004936e-05,
      "loss": 0.04338686913251877,
      "step": 700
    },
    {
      "epoch": 3.172043010752688,
      "grad_norm": 0.1712762009121397,
      "learning_rate": 1.2645097959617585e-05,
      "loss": 0.08021088689565659,
      "step": 701
    },
    {
      "epoch": 3.1765704584040746,
      "grad_norm": 0.17306948863643268,
      "learning_rate": 1.251405284678488e-05,
      "loss": 0.0890338122844696,
      "step": 702
    },
    {
      "epoch": 3.181097906055461,
      "grad_norm": 0.16734877496385878,
      "learning_rate": 1.238359310792877e-05,
      "loss": 0.07373081892728806,
      "step": 703
    },
    {
      "epoch": 3.1856253537068477,
      "grad_norm": 0.16681965977504648,
      "learning_rate": 1.2253720780285639e-05,
      "loss": 0.05940074473619461,
      "step": 704
    },
    {
      "epoch": 3.1901528013582343,
      "grad_norm": 0.1677395709964833,
      "learning_rate": 1.2124437891918993e-05,
      "loss": 0.049046434462070465,
      "step": 705
    },
    {
      "epoch": 3.194680249009621,
      "grad_norm": 0.16375245763793814,
      "learning_rate": 1.1995746461687734e-05,
      "loss": 0.04000747576355934,
      "step": 706
    },
    {
      "epoch": 3.1992076966610075,
      "grad_norm": 0.15889086433849803,
      "learning_rate": 1.186764849921468e-05,
      "loss": 0.06086156889796257,
      "step": 707
    },
    {
      "epoch": 3.203735144312394,
      "grad_norm": 0.15657853266586888,
      "learning_rate": 1.174014600485514e-05,
      "loss": 0.09235703200101852,
      "step": 708
    },
    {
      "epoch": 3.2082625919637806,
      "grad_norm": 0.16523424483152438,
      "learning_rate": 1.1613240969665685e-05,
      "loss": 0.0760895386338234,
      "step": 709
    },
    {
      "epoch": 3.212790039615167,
      "grad_norm": 0.17067373227429014,
      "learning_rate": 1.1486935375373126e-05,
      "loss": 0.06844381242990494,
      "step": 710
    },
    {
      "epoch": 3.2173174872665533,
      "grad_norm": 0.17285529090137522,
      "learning_rate": 1.1361231194343436e-05,
      "loss": 0.0570899173617363,
      "step": 711
    },
    {
      "epoch": 3.22184493491794,
      "grad_norm": 0.16098809089084598,
      "learning_rate": 1.1236130389551092e-05,
      "loss": 0.044195231050252914,
      "step": 712
    },
    {
      "epoch": 3.2263723825693265,
      "grad_norm": 0.17431764407498884,
      "learning_rate": 1.1111634914548297e-05,
      "loss": 0.04926883801817894,
      "step": 713
    },
    {
      "epoch": 3.230899830220713,
      "grad_norm": 0.15953958352407424,
      "learning_rate": 1.0987746713434576e-05,
      "loss": 0.08690941333770752,
      "step": 714
    },
    {
      "epoch": 3.2354272778720996,
      "grad_norm": 0.16045560003870812,
      "learning_rate": 1.0864467720826343e-05,
      "loss": 0.07696666568517685,
      "step": 715
    },
    {
      "epoch": 3.239954725523486,
      "grad_norm": 0.1617192142283404,
      "learning_rate": 1.0741799861826706e-05,
      "loss": 0.07071413844823837,
      "step": 716
    },
    {
      "epoch": 3.2444821731748728,
      "grad_norm": 0.16026922346768308,
      "learning_rate": 1.0619745051995472e-05,
      "loss": 0.05099868029356003,
      "step": 717
    },
    {
      "epoch": 3.2490096208262593,
      "grad_norm": 0.16355950842816344,
      "learning_rate": 1.0498305197319115e-05,
      "loss": 0.04914161562919617,
      "step": 718
    },
    {
      "epoch": 3.253537068477646,
      "grad_norm": 0.17319912138054358,
      "learning_rate": 1.0377482194181132e-05,
      "loss": 0.044711459428071976,
      "step": 719
    },
    {
      "epoch": 3.258064516129032,
      "grad_norm": 0.15120807983443754,
      "learning_rate": 1.0257277929332332e-05,
      "loss": 0.0844351053237915,
      "step": 720
    },
    {
      "epoch": 3.2625919637804186,
      "grad_norm": 0.15764732640093013,
      "learning_rate": 1.0137694279861454e-05,
      "loss": 0.08362387120723724,
      "step": 721
    },
    {
      "epoch": 3.267119411431805,
      "grad_norm": 0.1652434905849552,
      "learning_rate": 1.0018733113165773e-05,
      "loss": 0.07020123302936554,
      "step": 722
    },
    {
      "epoch": 3.2716468590831917,
      "grad_norm": 0.1615727759269588,
      "learning_rate": 9.900396286922026e-06,
      "loss": 0.05829951912164688,
      "step": 723
    },
    {
      "epoch": 3.2761743067345783,
      "grad_norm": 0.15540133620600818,
      "learning_rate": 9.782685649057333e-06,
      "loss": 0.0489511638879776,
      "step": 724
    },
    {
      "epoch": 3.280701754385965,
      "grad_norm": 0.17125472358784352,
      "learning_rate": 9.66560303772035e-06,
      "loss": 0.04217950999736786,
      "step": 725
    },
    {
      "epoch": 3.2852292020373515,
      "grad_norm": 0.16738285041744344,
      "learning_rate": 9.549150281252633e-06,
      "loss": 0.07959597557783127,
      "step": 726
    },
    {
      "epoch": 3.289756649688738,
      "grad_norm": 0.16850168868394563,
      "learning_rate": 9.433329198159974e-06,
      "loss": 0.09891261160373688,
      "step": 727
    },
    {
      "epoch": 3.2942840973401246,
      "grad_norm": 0.15597726589104297,
      "learning_rate": 9.31814159708413e-06,
      "loss": 0.07539305835962296,
      "step": 728
    },
    {
      "epoch": 3.298811544991511,
      "grad_norm": 0.16702347188447222,
      "learning_rate": 9.203589276774439e-06,
      "loss": 0.06475088000297546,
      "step": 729
    },
    {
      "epoch": 3.3033389926428978,
      "grad_norm": 0.15682792673122475,
      "learning_rate": 9.08967402605988e-06,
      "loss": 0.0529288724064827,
      "step": 730
    },
    {
      "epoch": 3.3078664402942843,
      "grad_norm": 0.16526227036485225,
      "learning_rate": 8.976397623821003e-06,
      "loss": 0.04603153094649315,
      "step": 731
    },
    {
      "epoch": 3.3123938879456705,
      "grad_norm": 0.16531667842750028,
      "learning_rate": 8.86376183896226e-06,
      "loss": 0.06262427568435669,
      "step": 732
    },
    {
      "epoch": 3.316921335597057,
      "grad_norm": 0.16283974682025906,
      "learning_rate": 8.751768430384305e-06,
      "loss": 0.0915934145450592,
      "step": 733
    },
    {
      "epoch": 3.3214487832484436,
      "grad_norm": 0.15764077900401163,
      "learning_rate": 8.640419146956557e-06,
      "loss": 0.0832981988787651,
      "step": 734
    },
    {
      "epoch": 3.32597623089983,
      "grad_norm": 0.1573195696429796,
      "learning_rate": 8.529715727489912e-06,
      "loss": 0.06754305958747864,
      "step": 735
    },
    {
      "epoch": 3.3305036785512168,
      "grad_norm": 0.1642300634758806,
      "learning_rate": 8.419659900709536e-06,
      "loss": 0.057708628475666046,
      "step": 736
    },
    {
      "epoch": 3.3350311262026033,
      "grad_norm": 0.15929872915842908,
      "learning_rate": 8.310253385227946e-06,
      "loss": 0.04438793659210205,
      "step": 737
    },
    {
      "epoch": 3.33955857385399,
      "grad_norm": 0.16519478212333719,
      "learning_rate": 8.201497889518073e-06,
      "loss": 0.042902104556560516,
      "step": 738
    },
    {
      "epoch": 3.3440860215053765,
      "grad_norm": 0.14737630137329194,
      "learning_rate": 8.093395111886687e-06,
      "loss": 0.08900769054889679,
      "step": 739
    },
    {
      "epoch": 3.348613469156763,
      "grad_norm": 0.1618938058753914,
      "learning_rate": 7.985946740447791e-06,
      "loss": 0.08507311344146729,
      "step": 740
    },
    {
      "epoch": 3.353140916808149,
      "grad_norm": 0.1596359027814366,
      "learning_rate": 7.879154453096304e-06,
      "loss": 0.06864621490240097,
      "step": 741
    },
    {
      "epoch": 3.3576683644595358,
      "grad_norm": 0.15567522378906307,
      "learning_rate": 7.773019917481872e-06,
      "loss": 0.054425135254859924,
      "step": 742
    },
    {
      "epoch": 3.3621958121109223,
      "grad_norm": 0.15758649025936708,
      "learning_rate": 7.667544790982778e-06,
      "loss": 0.047320909798145294,
      "step": 743
    },
    {
      "epoch": 3.366723259762309,
      "grad_norm": 0.16327258638928266,
      "learning_rate": 7.562730720680112e-06,
      "loss": 0.041624799370765686,
      "step": 744
    },
    {
      "epoch": 3.3712507074136955,
      "grad_norm": 0.15361939363105245,
      "learning_rate": 7.458579343331995e-06,
      "loss": 0.08947330713272095,
      "step": 745
    },
    {
      "epoch": 3.375778155065082,
      "grad_norm": 0.15597413816852934,
      "learning_rate": 7.3550922853480915e-06,
      "loss": 0.08114541321992874,
      "step": 746
    },
    {
      "epoch": 3.3803056027164686,
      "grad_norm": 0.16059378071093583,
      "learning_rate": 7.252271162764129e-06,
      "loss": 0.06897275149822235,
      "step": 747
    },
    {
      "epoch": 3.384833050367855,
      "grad_norm": 0.152879973115427,
      "learning_rate": 7.150117581216748e-06,
      "loss": 0.05720801651477814,
      "step": 748
    },
    {
      "epoch": 3.3893604980192418,
      "grad_norm": 0.15896141396588348,
      "learning_rate": 7.048633135918347e-06,
      "loss": 0.04943118989467621,
      "step": 749
    },
    {
      "epoch": 3.3938879456706283,
      "grad_norm": 0.1602689070908953,
      "learning_rate": 6.947819411632223e-06,
      "loss": 0.039460405707359314,
      "step": 750
    },
    {
      "epoch": 3.398415393322015,
      "grad_norm": 0.1669617663639801,
      "learning_rate": 6.8476779826478265e-06,
      "loss": 0.08177199959754944,
      "step": 751
    },
    {
      "epoch": 3.402942840973401,
      "grad_norm": 0.15975689129614332,
      "learning_rate": 6.748210412756134e-06,
      "loss": 0.08858094364404678,
      "step": 752
    },
    {
      "epoch": 3.4074702886247876,
      "grad_norm": 0.16135662158013386,
      "learning_rate": 6.649418255225298e-06,
      "loss": 0.07011917233467102,
      "step": 753
    },
    {
      "epoch": 3.411997736276174,
      "grad_norm": 0.15990295998471452,
      "learning_rate": 6.551303052776292e-06,
      "loss": 0.058572862297296524,
      "step": 754
    },
    {
      "epoch": 3.4165251839275608,
      "grad_norm": 0.15952088982362758,
      "learning_rate": 6.45386633755894e-06,
      "loss": 0.047646377235651016,
      "step": 755
    },
    {
      "epoch": 3.4210526315789473,
      "grad_norm": 0.1613594717934019,
      "learning_rate": 6.357109631127889e-06,
      "loss": 0.03935559466481209,
      "step": 756
    },
    {
      "epoch": 3.425580079230334,
      "grad_norm": 0.15912287807624553,
      "learning_rate": 6.261034444418879e-06,
      "loss": 0.06060168892145157,
      "step": 757
    },
    {
      "epoch": 3.4301075268817205,
      "grad_norm": 0.16111441855805086,
      "learning_rate": 6.165642277725203e-06,
      "loss": 0.09181486070156097,
      "step": 758
    },
    {
      "epoch": 3.434634974533107,
      "grad_norm": 0.16519976628545915,
      "learning_rate": 6.07093462067419e-06,
      "loss": 0.0789683610200882,
      "step": 759
    },
    {
      "epoch": 3.4391624221844936,
      "grad_norm": 0.1559387343193889,
      "learning_rate": 5.976912952204017e-06,
      "loss": 0.0634407326579094,
      "step": 760
    },
    {
      "epoch": 3.44368986983588,
      "grad_norm": 0.15643118236775796,
      "learning_rate": 5.883578740540546e-06,
      "loss": 0.05550439655780792,
      "step": 761
    },
    {
      "epoch": 3.4482173174872663,
      "grad_norm": 0.15747386201299873,
      "learning_rate": 5.79093344317449e-06,
      "loss": 0.04290204495191574,
      "step": 762
    },
    {
      "epoch": 3.452744765138653,
      "grad_norm": 0.1646471000344093,
      "learning_rate": 5.698978506838532e-06,
      "loss": 0.04180324822664261,
      "step": 763
    },
    {
      "epoch": 3.4572722127900395,
      "grad_norm": 0.14229847674612914,
      "learning_rate": 5.607715367484861e-06,
      "loss": 0.08869330585002899,
      "step": 764
    },
    {
      "epoch": 3.461799660441426,
      "grad_norm": 0.1471590674363114,
      "learning_rate": 5.51714545026264e-06,
      "loss": 0.08063746988773346,
      "step": 765
    },
    {
      "epoch": 3.4663271080928126,
      "grad_norm": 0.15814353243743898,
      "learning_rate": 5.4272701694958076e-06,
      "loss": 0.06893438845872879,
      "step": 766
    },
    {
      "epoch": 3.470854555744199,
      "grad_norm": 0.1556645213851928,
      "learning_rate": 5.338090928660999e-06,
      "loss": 0.05451403185725212,
      "step": 767
    },
    {
      "epoch": 3.475382003395586,
      "grad_norm": 0.15569393724140868,
      "learning_rate": 5.249609120365578e-06,
      "loss": 0.04844459146261215,
      "step": 768
    },
    {
      "epoch": 3.4799094510469724,
      "grad_norm": 0.16611731683384365,
      "learning_rate": 5.16182612632598e-06,
      "loss": 0.04286676645278931,
      "step": 769
    },
    {
      "epoch": 3.484436898698359,
      "grad_norm": 0.15293615219864046,
      "learning_rate": 5.074743317346009e-06,
      "loss": 0.08715027570724487,
      "step": 770
    },
    {
      "epoch": 3.4889643463497455,
      "grad_norm": 0.15074314633235772,
      "learning_rate": 4.988362053295564e-06,
      "loss": 0.08602124452590942,
      "step": 771
    },
    {
      "epoch": 3.493491794001132,
      "grad_norm": 0.15513477299232484,
      "learning_rate": 4.902683683089304e-06,
      "loss": 0.07156828045845032,
      "step": 772
    },
    {
      "epoch": 3.498019241652518,
      "grad_norm": 0.16174335126430972,
      "learning_rate": 4.817709544665628e-06,
      "loss": 0.0598549023270607,
      "step": 773
    },
    {
      "epoch": 3.502546689303905,
      "grad_norm": 0.15795200396841924,
      "learning_rate": 4.733440964965791e-06,
      "loss": 0.04977518320083618,
      "step": 774
    },
    {
      "epoch": 3.5070741369552914,
      "grad_norm": 0.15513788247676116,
      "learning_rate": 4.649879259913137e-06,
      "loss": 0.039833471179008484,
      "step": 775
    },
    {
      "epoch": 3.511601584606678,
      "grad_norm": 0.15698424696956995,
      "learning_rate": 4.567025734392622e-06,
      "loss": 0.08777929842472076,
      "step": 776
    },
    {
      "epoch": 3.5161290322580645,
      "grad_norm": 0.15265023394351482,
      "learning_rate": 4.484881682230341e-06,
      "loss": 0.08583962172269821,
      "step": 777
    },
    {
      "epoch": 3.520656479909451,
      "grad_norm": 0.15272885627714522,
      "learning_rate": 4.403448386173437e-06,
      "loss": 0.06798358261585236,
      "step": 778
    },
    {
      "epoch": 3.5251839275608376,
      "grad_norm": 0.15569180522696666,
      "learning_rate": 4.322727117869951e-06,
      "loss": 0.06092444807291031,
      "step": 779
    },
    {
      "epoch": 3.5297113752122242,
      "grad_norm": 0.1589263536844735,
      "learning_rate": 4.242719137849077e-06,
      "loss": 0.052914176136255264,
      "step": 780
    },
    {
      "epoch": 3.534238822863611,
      "grad_norm": 0.15055087585343038,
      "learning_rate": 4.163425695501388e-06,
      "loss": 0.038688283413648605,
      "step": 781
    },
    {
      "epoch": 3.538766270514997,
      "grad_norm": 0.15450216645361534,
      "learning_rate": 4.0848480290593625e-06,
      "loss": 0.060667701065540314,
      "step": 782
    },
    {
      "epoch": 3.5432937181663835,
      "grad_norm": 0.15175236104478088,
      "learning_rate": 4.00698736557808e-06,
      "loss": 0.09457433223724365,
      "step": 783
    },
    {
      "epoch": 3.54782116581777,
      "grad_norm": 0.15330083415530651,
      "learning_rate": 3.929844920915987e-06,
      "loss": 0.0753507912158966,
      "step": 784
    },
    {
      "epoch": 3.5523486134691566,
      "grad_norm": 0.15460276506056247,
      "learning_rate": 3.853421899715992e-06,
      "loss": 0.06055988371372223,
      "step": 785
    },
    {
      "epoch": 3.556876061120543,
      "grad_norm": 0.15268718665932088,
      "learning_rate": 3.7777194953865667e-06,
      "loss": 0.05165576934814453,
      "step": 786
    },
    {
      "epoch": 3.56140350877193,
      "grad_norm": 0.15764631090076642,
      "learning_rate": 3.702738890083207e-06,
      "loss": 0.04397202283143997,
      "step": 787
    },
    {
      "epoch": 3.5659309564233164,
      "grad_norm": 0.1756581403035511,
      "learning_rate": 3.628481254689875e-06,
      "loss": 0.046176545321941376,
      "step": 788
    },
    {
      "epoch": 3.570458404074703,
      "grad_norm": 0.13903295845121066,
      "learning_rate": 3.5549477488007854e-06,
      "loss": 0.08546517044305801,
      "step": 789
    },
    {
      "epoch": 3.5749858517260895,
      "grad_norm": 0.14625086414117394,
      "learning_rate": 3.4821395207022766e-06,
      "loss": 0.07490271329879761,
      "step": 790
    },
    {
      "epoch": 3.579513299377476,
      "grad_norm": 0.1529370303120228,
      "learning_rate": 3.4100577073548634e-06,
      "loss": 0.06298314034938812,
      "step": 791
    },
    {
      "epoch": 3.5840407470288627,
      "grad_norm": 0.1535335759252131,
      "learning_rate": 3.3387034343755065e-06,
      "loss": 0.05735611543059349,
      "step": 792
    },
    {
      "epoch": 3.5885681946802492,
      "grad_norm": 0.1541760869868739,
      "learning_rate": 3.2680778160200155e-06,
      "loss": 0.04231870174407959,
      "step": 793
    },
    {
      "epoch": 3.593095642331636,
      "grad_norm": 0.1678727786788288,
      "learning_rate": 3.198181955165669e-06,
      "loss": 0.04027624428272247,
      "step": 794
    },
    {
      "epoch": 3.597623089983022,
      "grad_norm": 0.14943363872847212,
      "learning_rate": 3.1290169432939553e-06,
      "loss": 0.07936602830886841,
      "step": 795
    },
    {
      "epoch": 3.6021505376344085,
      "grad_norm": 0.15394087286690059,
      "learning_rate": 3.060583860473587e-06,
      "loss": 0.08372323960065842,
      "step": 796
    },
    {
      "epoch": 3.606677985285795,
      "grad_norm": 0.15124484091267687,
      "learning_rate": 2.9928837753435746e-06,
      "loss": 0.06588123738765717,
      "step": 797
    },
    {
      "epoch": 3.6112054329371817,
      "grad_norm": 0.15914915572966265,
      "learning_rate": 2.9259177450965682e-06,
      "loss": 0.05927068740129471,
      "step": 798
    },
    {
      "epoch": 3.6157328805885682,
      "grad_norm": 0.15507298661949537,
      "learning_rate": 2.8596868154623703e-06,
      "loss": 0.047394055873155594,
      "step": 799
    },
    {
      "epoch": 3.620260328239955,
      "grad_norm": 0.15414879271979146,
      "learning_rate": 2.794192020691544e-06,
      "loss": 0.040023073554039,
      "step": 800
    },
    {
      "epoch": 3.6247877758913414,
      "grad_norm": 0.15063072941137776,
      "learning_rate": 2.7294343835393368e-06,
      "loss": 0.07365724444389343,
      "step": 801
    },
    {
      "epoch": 3.629315223542728,
      "grad_norm": 0.16206292605361364,
      "learning_rate": 2.665414915249631e-06,
      "loss": 0.09670303761959076,
      "step": 802
    },
    {
      "epoch": 3.633842671194114,
      "grad_norm": 0.15756959946827895,
      "learning_rate": 2.6021346155392423e-06,
      "loss": 0.0773511528968811,
      "step": 803
    },
    {
      "epoch": 3.6383701188455007,
      "grad_norm": 0.159147027221656,
      "learning_rate": 2.539594472582213e-06,
      "loss": 0.062302928417921066,
      "step": 804
    },
    {
      "epoch": 3.6428975664968872,
      "grad_norm": 0.1527324547096345,
      "learning_rate": 2.4777954629944477e-06,
      "loss": 0.05168925225734711,
      "step": 805
    },
    {
      "epoch": 3.647425014148274,
      "grad_norm": 0.15196338938892234,
      "learning_rate": 2.416738551818454e-06,
      "loss": 0.04184819757938385,
      "step": 806
    },
    {
      "epoch": 3.6519524617996604,
      "grad_norm": 0.16196166090266254,
      "learning_rate": 2.3564246925082357e-06,
      "loss": 0.06736510992050171,
      "step": 807
    },
    {
      "epoch": 3.656479909451047,
      "grad_norm": 0.14890893645296682,
      "learning_rate": 2.2968548269144574e-06,
      "loss": 0.09164802730083466,
      "step": 808
    },
    {
      "epoch": 3.6610073571024335,
      "grad_norm": 0.15408211922238227,
      "learning_rate": 2.238029885269677e-06,
      "loss": 0.07316052913665771,
      "step": 809
    },
    {
      "epoch": 3.66553480475382,
      "grad_norm": 0.15701808807503811,
      "learning_rate": 2.179950786173879e-06,
      "loss": 0.06162027269601822,
      "step": 810
    },
    {
      "epoch": 3.6700622524052067,
      "grad_norm": 0.15613269534431912,
      "learning_rate": 2.122618436580082e-06,
      "loss": 0.05187002941966057,
      "step": 811
    },
    {
      "epoch": 3.6745897000565932,
      "grad_norm": 0.1496676889935242,
      "learning_rate": 2.066033731780209e-06,
      "loss": 0.04026584327220917,
      "step": 812
    },
    {
      "epoch": 3.67911714770798,
      "grad_norm": 0.16572654137997284,
      "learning_rate": 2.01019755539108e-06,
      "loss": 0.041823290288448334,
      "step": 813
    },
    {
      "epoch": 3.6836445953593664,
      "grad_norm": 0.14034470741750407,
      "learning_rate": 1.9551107793406355e-06,
      "loss": 0.08595189452171326,
      "step": 814
    },
    {
      "epoch": 3.688172043010753,
      "grad_norm": 0.1527973789499426,
      "learning_rate": 1.9007742638543102e-06,
      "loss": 0.07732295989990234,
      "step": 815
    },
    {
      "epoch": 3.692699490662139,
      "grad_norm": 0.14730166164260208,
      "learning_rate": 1.8471888574415951e-06,
      "loss": 0.06500217318534851,
      "step": 816
    },
    {
      "epoch": 3.6972269383135257,
      "grad_norm": 0.1565767548547146,
      "learning_rate": 1.794355396882813e-06,
      "loss": 0.052856337279081345,
      "step": 817
    },
    {
      "epoch": 3.7017543859649122,
      "grad_norm": 0.15183313998733808,
      "learning_rate": 1.7422747072160017e-06,
      "loss": 0.0447005033493042,
      "step": 818
    },
    {
      "epoch": 3.706281833616299,
      "grad_norm": 0.1604975684005926,
      "learning_rate": 1.6909476017240912e-06,
      "loss": 0.04208570718765259,
      "step": 819
    },
    {
      "epoch": 3.7108092812676854,
      "grad_norm": 0.14703983351715297,
      "learning_rate": 1.6403748819221466e-06,
      "loss": 0.0831032320857048,
      "step": 820
    },
    {
      "epoch": 3.715336728919072,
      "grad_norm": 0.16135082507857954,
      "learning_rate": 1.5905573375449012e-06,
      "loss": 0.08684415370225906,
      "step": 821
    },
    {
      "epoch": 3.7198641765704585,
      "grad_norm": 0.1570784237060374,
      "learning_rate": 1.5414957465343882e-06,
      "loss": 0.06628584861755371,
      "step": 822
    },
    {
      "epoch": 3.7243916242218447,
      "grad_norm": 0.1510596052074201,
      "learning_rate": 1.4931908750278e-06,
      "loss": 0.057143792510032654,
      "step": 823
    },
    {
      "epoch": 3.7289190718732312,
      "grad_norm": 0.151662078759489,
      "learning_rate": 1.4456434773455541e-06,
      "loss": 0.05711142718791962,
      "step": 824
    },
    {
      "epoch": 3.733446519524618,
      "grad_norm": 0.1539657036570409,
      "learning_rate": 1.3988542959794627e-06,
      "loss": 0.03840760886669159,
      "step": 825
    },
    {
      "epoch": 3.7379739671760044,
      "grad_norm": 0.15620650235667358,
      "learning_rate": 1.3528240615811816e-06,
      "loss": 0.08128702640533447,
      "step": 826
    },
    {
      "epoch": 3.742501414827391,
      "grad_norm": 0.14950219877477688,
      "learning_rate": 1.3075534929507693e-06,
      "loss": 0.09000637382268906,
      "step": 827
    },
    {
      "epoch": 3.7470288624787775,
      "grad_norm": 0.15369244061554468,
      "learning_rate": 1.2630432970255013e-06,
      "loss": 0.07430066913366318,
      "step": 828
    },
    {
      "epoch": 3.751556310130164,
      "grad_norm": 0.16175879968517679,
      "learning_rate": 1.2192941688687843e-06,
      "loss": 0.05985051020979881,
      "step": 829
    },
    {
      "epoch": 3.7560837577815507,
      "grad_norm": 0.14747817211522232,
      "learning_rate": 1.1763067916593262e-06,
      "loss": 0.04820965975522995,
      "step": 830
    },
    {
      "epoch": 3.7606112054329373,
      "grad_norm": 0.15356666434221958,
      "learning_rate": 1.1340818366804729e-06,
      "loss": 0.04318070411682129,
      "step": 831
    },
    {
      "epoch": 3.765138653084324,
      "grad_norm": 0.16528905989080245,
      "learning_rate": 1.0926199633097157e-06,
      "loss": 0.06352683901786804,
      "step": 832
    },
    {
      "epoch": 3.7696661007357104,
      "grad_norm": 0.15731659177179078,
      "learning_rate": 1.0519218190084056e-06,
      "loss": 0.1001967266201973,
      "step": 833
    },
    {
      "epoch": 3.774193548387097,
      "grad_norm": 0.1541312314650151,
      "learning_rate": 1.0119880393116176e-06,
      "loss": 0.0764806866645813,
      "step": 834
    },
    {
      "epoch": 3.7787209960384835,
      "grad_norm": 0.14845792539339586,
      "learning_rate": 9.728192478182574e-07,
      "loss": 0.06443999707698822,
      "step": 835
    },
    {
      "epoch": 3.7832484436898697,
      "grad_norm": 0.1565016153733591,
      "learning_rate": 9.344160561812921e-07,
      "loss": 0.05295903980731964,
      "step": 836
    },
    {
      "epoch": 3.7877758913412563,
      "grad_norm": 0.1471871869309214,
      "learning_rate": 8.967790640982465e-07,
      "loss": 0.041768576949834824,
      "step": 837
    },
    {
      "epoch": 3.792303338992643,
      "grad_norm": 0.15978438538353343,
      "learning_rate": 8.599088593017723e-07,
      "loss": 0.04355769604444504,
      "step": 838
    },
    {
      "epoch": 3.7968307866440294,
      "grad_norm": 0.13835662528062842,
      "learning_rate": 8.238060175505269e-07,
      "loss": 0.08960022777318954,
      "step": 839
    },
    {
      "epoch": 3.801358234295416,
      "grad_norm": 0.15270774966828765,
      "learning_rate": 7.884711026201585e-07,
      "loss": 0.07562049478292465,
      "step": 840
    },
    {
      "epoch": 3.8058856819468025,
      "grad_norm": 0.15345558176555804,
      "learning_rate": 7.53904666294497e-07,
      "loss": 0.06913469731807709,
      "step": 841
    },
    {
      "epoch": 3.810413129598189,
      "grad_norm": 0.15096906039665564,
      "learning_rate": 7.201072483569549e-07,
      "loss": 0.053817637264728546,
      "step": 842
    },
    {
      "epoch": 3.8149405772495757,
      "grad_norm": 0.14873095528976368,
      "learning_rate": 6.870793765820782e-07,
      "loss": 0.04244130477309227,
      "step": 843
    },
    {
      "epoch": 3.819468024900962,
      "grad_norm": 0.15592437232678136,
      "learning_rate": 6.548215667273206e-07,
      "loss": 0.04185333102941513,
      "step": 844
    },
    {
      "epoch": 3.8239954725523484,
      "grad_norm": 0.1430641212527451,
      "learning_rate": 6.233343225249933e-07,
      "loss": 0.08310285210609436,
      "step": 845
    },
    {
      "epoch": 3.828522920203735,
      "grad_norm": 0.14868161257084533,
      "learning_rate": 5.92618135674361e-07,
      "loss": 0.08037766814231873,
      "step": 846
    },
    {
      "epoch": 3.8330503678551215,
      "grad_norm": 0.1519358880326252,
      "learning_rate": 5.626734858340255e-07,
      "loss": 0.06963315606117249,
      "step": 847
    },
    {
      "epoch": 3.837577815506508,
      "grad_norm": 0.15626148257749853,
      "learning_rate": 5.335008406143815e-07,
      "loss": 0.058459773659706116,
      "step": 848
    },
    {
      "epoch": 3.8421052631578947,
      "grad_norm": 0.14849571274572423,
      "learning_rate": 5.051006555703453e-07,
      "loss": 0.04667978733778,
      "step": 849
    },
    {
      "epoch": 3.8466327108092813,
      "grad_norm": 0.15380370744996802,
      "learning_rate": 4.774733741942206e-07,
      "loss": 0.0407584086060524,
      "step": 850
    },
    {
      "epoch": 3.851160158460668,
      "grad_norm": 0.142605966337961,
      "learning_rate": 4.5061942790879386e-07,
      "loss": 0.07849274575710297,
      "step": 851
    },
    {
      "epoch": 3.8556876061120544,
      "grad_norm": 0.15317537834760572,
      "learning_rate": 4.2453923606057265e-07,
      "loss": 0.09511102735996246,
      "step": 852
    },
    {
      "epoch": 3.860215053763441,
      "grad_norm": 0.15036969162974312,
      "learning_rate": 3.992332059132631e-07,
      "loss": 0.07533329725265503,
      "step": 853
    },
    {
      "epoch": 3.8647425014148276,
      "grad_norm": 0.15076748076530955,
      "learning_rate": 3.747017326413971e-07,
      "loss": 0.056159425526857376,
      "step": 854
    },
    {
      "epoch": 3.869269949066214,
      "grad_norm": 0.1515430338118417,
      "learning_rate": 3.5094519932415417e-07,
      "loss": 0.05289001017808914,
      "step": 855
    },
    {
      "epoch": 3.8737973967176007,
      "grad_norm": 0.15500832383952134,
      "learning_rate": 3.2796397693939385e-07,
      "loss": 0.039535753428936005,
      "step": 856
    },
    {
      "epoch": 3.878324844368987,
      "grad_norm": 0.14646644392684122,
      "learning_rate": 3.0575842435785486e-07,
      "loss": 0.0575360544025898,
      "step": 857
    },
    {
      "epoch": 3.8828522920203734,
      "grad_norm": 0.15652446928494548,
      "learning_rate": 2.843288883375539e-07,
      "loss": 0.0908580794930458,
      "step": 858
    },
    {
      "epoch": 3.88737973967176,
      "grad_norm": 0.15100612580392556,
      "learning_rate": 2.6367570351836237e-07,
      "loss": 0.07020405679941177,
      "step": 859
    },
    {
      "epoch": 3.8919071873231466,
      "grad_norm": 0.1567994785843258,
      "learning_rate": 2.437991924167937e-07,
      "loss": 0.06751954555511475,
      "step": 860
    },
    {
      "epoch": 3.896434634974533,
      "grad_norm": 0.15056768351261735,
      "learning_rate": 2.2469966542096322e-07,
      "loss": 0.05812934786081314,
      "step": 861
    },
    {
      "epoch": 3.9009620826259197,
      "grad_norm": 0.15300797430847607,
      "learning_rate": 2.0637742078573607e-07,
      "loss": 0.04170238599181175,
      "step": 862
    },
    {
      "epoch": 3.9054895302773063,
      "grad_norm": 0.1577533773184765,
      "learning_rate": 1.8883274462806467e-07,
      "loss": 0.04246639460325241,
      "step": 863
    },
    {
      "epoch": 3.910016977928693,
      "grad_norm": 0.1399590377761555,
      "learning_rate": 1.7206591092253642e-07,
      "loss": 0.09115473926067352,
      "step": 864
    },
    {
      "epoch": 3.914544425580079,
      "grad_norm": 0.15185387966957223,
      "learning_rate": 1.560771814970885e-07,
      "loss": 0.07839730381965637,
      "step": 865
    },
    {
      "epoch": 3.9190718732314656,
      "grad_norm": 0.1504348726008439,
      "learning_rate": 1.4086680602891643e-07,
      "loss": 0.06564494967460632,
      "step": 866
    },
    {
      "epoch": 3.923599320882852,
      "grad_norm": 0.15208995216401133,
      "learning_rate": 1.264350220405719e-07,
      "loss": 0.05529960244894028,
      "step": 867
    },
    {
      "epoch": 3.9281267685342387,
      "grad_norm": 0.14903132343569683,
      "learning_rate": 1.1278205489626547e-07,
      "loss": 0.04081864655017853,
      "step": 868
    },
    {
      "epoch": 3.9326542161856253,
      "grad_norm": 0.15881363321169775,
      "learning_rate": 9.99081177983363e-08,
      "loss": 0.04252634942531586,
      "step": 869
    },
    {
      "epoch": 3.937181663837012,
      "grad_norm": 0.14364090157421816,
      "learning_rate": 8.781341178393244e-08,
      "loss": 0.0835392028093338,
      "step": 870
    },
    {
      "epoch": 3.9417091114883984,
      "grad_norm": 0.15328856903966662,
      "learning_rate": 7.649812572185222e-08,
      "loss": 0.09088088572025299,
      "step": 871
    },
    {
      "epoch": 3.946236559139785,
      "grad_norm": 0.1461174844367331,
      "learning_rate": 6.596243630963006e-08,
      "loss": 0.07099941372871399,
      "step": 872
    },
    {
      "epoch": 3.9507640067911716,
      "grad_norm": 0.16113221779664158,
      "learning_rate": 5.620650807073857e-08,
      "loss": 0.05559823662042618,
      "step": 873
    },
    {
      "epoch": 3.955291454442558,
      "grad_norm": 0.15131008698282955,
      "learning_rate": 4.723049335204066e-08,
      "loss": 0.04776914045214653,
      "step": 874
    },
    {
      "epoch": 3.9598189020939447,
      "grad_norm": 0.15304548694602363,
      "learning_rate": 3.9034532321408076e-08,
      "loss": 0.037714093923568726,
      "step": 875
    },
    {
      "epoch": 3.9643463497453313,
      "grad_norm": 0.15407251192345708,
      "learning_rate": 3.161875296553429e-08,
      "loss": 0.07897656410932541,
      "step": 876
    },
    {
      "epoch": 3.968873797396718,
      "grad_norm": 0.1505951685106998,
      "learning_rate": 2.4983271087924974e-08,
      "loss": 0.08945634961128235,
      "step": 877
    },
    {
      "epoch": 3.973401245048104,
      "grad_norm": 0.1562160450088478,
      "learning_rate": 1.9128190307105e-08,
      "loss": 0.06681520491838455,
      "step": 878
    },
    {
      "epoch": 3.9779286926994906,
      "grad_norm": 0.14480850196719536,
      "learning_rate": 1.4053602054991955e-08,
      "loss": 0.056111957877874374,
      "step": 879
    },
    {
      "epoch": 3.982456140350877,
      "grad_norm": 0.15057235606341254,
      "learning_rate": 9.75958557545842e-09,
      "loss": 0.04651391878724098,
      "step": 880
    },
    {
      "epoch": 3.9869835880022637,
      "grad_norm": 0.148782705610793,
      "learning_rate": 6.246207923116254e-09,
      "loss": 0.03932764381170273,
      "step": 881
    },
    {
      "epoch": 3.9915110356536503,
      "grad_norm": 0.16613848323659036,
      "learning_rate": 3.513523962256349e-09,
      "loss": 0.06670388579368591,
      "step": 882
    },
    {
      "epoch": 3.996038483305037,
      "grad_norm": 0.1520550990812053,
      "learning_rate": 1.5615763659881933e-09,
      "loss": 0.07301482558250427,
      "step": 883
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.16232352886585782,
      "learning_rate": 3.903956155848487e-10,
      "loss": 0.04399577155709267,
      "step": 884
    },
    {
      "epoch": 4.0,
      "step": 884,
      "total_flos": 2826281377071104.0,
      "train_loss": 0.1602747890744274,
      "train_runtime": 158332.2671,
      "train_samples_per_second": 0.357,
      "train_steps_per_second": 0.006
    }
  ],
  "logging_steps": 1,
  "max_steps": 884,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2826281377071104.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}