{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 33504,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0002984718242597899,
      "grad_norm": 2.5109126567840576,
      "learning_rate": 1.1933174224343676e-07,
      "loss": 0.849,
      "mean_token_accuracy": 0.7716509580612183,
      "num_tokens": 167680.0,
      "step": 5
    },
    {
      "epoch": 0.0005969436485195798,
      "grad_norm": 2.5402281284332275,
      "learning_rate": 2.684964200477327e-07,
      "loss": 0.8594,
      "mean_token_accuracy": 0.7681856036186219,
      "num_tokens": 335360.0,
      "step": 10
    },
    {
      "epoch": 0.0008954154727793697,
      "grad_norm": 2.270322799682617,
      "learning_rate": 4.1766109785202865e-07,
      "loss": 0.8854,
      "mean_token_accuracy": 0.7613682389259339,
      "num_tokens": 503040.0,
      "step": 15
    },
    {
      "epoch": 0.0011938872970391596,
      "grad_norm": 2.388043165206909,
      "learning_rate": 5.668257756563246e-07,
      "loss": 0.8881,
      "mean_token_accuracy": 0.7604795455932617,
      "num_tokens": 670720.0,
      "step": 20
    },
    {
      "epoch": 0.0014923591212989494,
      "grad_norm": 2.682539224624634,
      "learning_rate": 7.159904534606206e-07,
      "loss": 0.8831,
      "mean_token_accuracy": 0.7617738366127014,
      "num_tokens": 838400.0,
      "step": 25
    },
    {
      "epoch": 0.0017908309455587394,
      "grad_norm": 1.9359753131866455,
      "learning_rate": 8.651551312649164e-07,
      "loss": 0.8246,
      "mean_token_accuracy": 0.7766312837600708,
      "num_tokens": 1006080.0,
      "step": 30
    },
    {
      "epoch": 0.002089302769818529,
      "grad_norm": 1.921677589416504,
      "learning_rate": 1.0143198090692125e-06,
      "loss": 0.8854,
      "mean_token_accuracy": 0.7587021350860595,
      "num_tokens": 1173760.0,
      "step": 35
    },
    {
      "epoch": 0.002387774594078319,
      "grad_norm": 1.6539034843444824,
      "learning_rate": 1.1634844868735085e-06,
      "loss": 0.881,
      "mean_token_accuracy": 0.7584575891494751,
      "num_tokens": 1341440.0,
      "step": 40
    },
    {
      "epoch": 0.0026862464183381087,
      "grad_norm": 1.6672691106796265,
      "learning_rate": 1.3126491646778044e-06,
      "loss": 0.8263,
      "mean_token_accuracy": 0.7719491720199585,
      "num_tokens": 1509120.0,
      "step": 45
    },
    {
      "epoch": 0.0029847182425978987,
      "grad_norm": 1.7613340616226196,
      "learning_rate": 1.4618138424821004e-06,
      "loss": 0.8306,
      "mean_token_accuracy": 0.7692890286445617,
      "num_tokens": 1676800.0,
      "step": 50
    },
    {
      "epoch": 0.0032831900668576887,
      "grad_norm": 1.4340014457702637,
      "learning_rate": 1.6109785202863964e-06,
      "loss": 0.7541,
      "mean_token_accuracy": 0.7874746680259704,
      "num_tokens": 1844480.0,
      "step": 55
    },
    {
      "epoch": 0.0035816618911174787,
      "grad_norm": 1.2414655685424805,
      "learning_rate": 1.7601431980906923e-06,
      "loss": 0.8135,
      "mean_token_accuracy": 0.7701061725616455,
      "num_tokens": 2012160.0,
      "step": 60
    },
    {
      "epoch": 0.0038801337153772683,
      "grad_norm": 1.2400470972061157,
      "learning_rate": 1.909307875894988e-06,
      "loss": 0.7695,
      "mean_token_accuracy": 0.7819217443466187,
      "num_tokens": 2179840.0,
      "step": 65
    },
    {
      "epoch": 0.004178605539637058,
      "grad_norm": 1.1827894449234009,
      "learning_rate": 2.0584725536992843e-06,
      "loss": 0.767,
      "mean_token_accuracy": 0.7819277167320251,
      "num_tokens": 2347520.0,
      "step": 70
    },
    {
      "epoch": 0.004477077363896848,
      "grad_norm": 1.2285350561141968,
      "learning_rate": 2.20763723150358e-06,
      "loss": 0.7641,
      "mean_token_accuracy": 0.7814624905586243,
      "num_tokens": 2515200.0,
      "step": 75
    },
    {
      "epoch": 0.004775549188156638,
      "grad_norm": 1.2129645347595215,
      "learning_rate": 2.3568019093078762e-06,
      "loss": 0.7559,
      "mean_token_accuracy": 0.7816056251525879,
      "num_tokens": 2682880.0,
      "step": 80
    },
    {
      "epoch": 0.005074021012416428,
      "grad_norm": 1.2340443134307861,
      "learning_rate": 2.505966587112172e-06,
      "loss": 0.759,
      "mean_token_accuracy": 0.7816712260246277,
      "num_tokens": 2850560.0,
      "step": 85
    },
    {
      "epoch": 0.0053724928366762174,
      "grad_norm": 1.1642875671386719,
      "learning_rate": 2.6551312649164677e-06,
      "loss": 0.7329,
      "mean_token_accuracy": 0.7863831520080566,
      "num_tokens": 3018240.0,
      "step": 90
    },
    {
      "epoch": 0.005670964660936008,
      "grad_norm": 1.1461609601974487,
      "learning_rate": 2.804295942720764e-06,
      "loss": 0.7503,
      "mean_token_accuracy": 0.7809912919998169,
      "num_tokens": 3185920.0,
      "step": 95
    },
    {
      "epoch": 0.0059694364851957974,
      "grad_norm": 1.0481009483337402,
      "learning_rate": 2.9534606205250597e-06,
      "loss": 0.7112,
      "mean_token_accuracy": 0.7925742626190185,
      "num_tokens": 3353600.0,
      "step": 100
    },
    {
      "epoch": 0.006267908309455587,
      "grad_norm": 1.0805094242095947,
      "learning_rate": 3.1026252983293554e-06,
      "loss": 0.7714,
      "mean_token_accuracy": 0.7766133904457092,
      "num_tokens": 3521280.0,
      "step": 105
    },
    {
      "epoch": 0.0065663801337153774,
      "grad_norm": 0.9588642716407776,
      "learning_rate": 3.251789976133652e-06,
      "loss": 0.7113,
      "mean_token_accuracy": 0.7901944398880005,
      "num_tokens": 3688960.0,
      "step": 110
    },
    {
      "epoch": 0.006864851957975167,
      "grad_norm": 1.172989845275879,
      "learning_rate": 3.4009546539379474e-06,
      "loss": 0.749,
      "mean_token_accuracy": 0.78075270652771,
      "num_tokens": 3856640.0,
      "step": 115
    },
    {
      "epoch": 0.0071633237822349575,
      "grad_norm": 1.03142511844635,
      "learning_rate": 3.5501193317422436e-06,
      "loss": 0.6881,
      "mean_token_accuracy": 0.7986937999725342,
      "num_tokens": 4024320.0,
      "step": 120
    },
    {
      "epoch": 0.007461795606494747,
      "grad_norm": 1.1663661003112793,
      "learning_rate": 3.6992840095465393e-06,
      "loss": 0.7456,
      "mean_token_accuracy": 0.781778609752655,
      "num_tokens": 4192000.0,
      "step": 125
    },
    {
      "epoch": 0.007760267430754537,
      "grad_norm": 1.00513756275177,
      "learning_rate": 3.8484486873508355e-06,
      "loss": 0.6995,
      "mean_token_accuracy": 0.7933019280433655,
      "num_tokens": 4359680.0,
      "step": 130
    },
    {
      "epoch": 0.008058739255014326,
      "grad_norm": 1.2561825513839722,
      "learning_rate": 3.997613365155131e-06,
      "loss": 0.686,
      "mean_token_accuracy": 0.797393536567688,
      "num_tokens": 4527360.0,
      "step": 135
    },
    {
      "epoch": 0.008357211079274117,
      "grad_norm": 1.193806529045105,
      "learning_rate": 4.146778042959427e-06,
      "loss": 0.7801,
      "mean_token_accuracy": 0.7724979162216187,
      "num_tokens": 4695040.0,
      "step": 140
    },
    {
      "epoch": 0.008655682903533907,
      "grad_norm": 1.072721242904663,
      "learning_rate": 4.295942720763723e-06,
      "loss": 0.7001,
      "mean_token_accuracy": 0.7940415143966675,
      "num_tokens": 4862720.0,
      "step": 145
    },
    {
      "epoch": 0.008954154727793696,
      "grad_norm": 1.1867351531982422,
      "learning_rate": 4.445107398568019e-06,
      "loss": 0.7206,
      "mean_token_accuracy": 0.7882440567016602,
      "num_tokens": 5030400.0,
      "step": 150
    },
    {
      "epoch": 0.009252626552053486,
      "grad_norm": 1.0713008642196655,
      "learning_rate": 4.594272076372315e-06,
      "loss": 0.6724,
      "mean_token_accuracy": 0.8011153697967529,
      "num_tokens": 5198080.0,
      "step": 155
    },
    {
      "epoch": 0.009551098376313277,
      "grad_norm": 1.1685636043548584,
      "learning_rate": 4.743436754176611e-06,
      "loss": 0.6886,
      "mean_token_accuracy": 0.7966300845146179,
      "num_tokens": 5365760.0,
      "step": 160
    },
    {
      "epoch": 0.009849570200573065,
      "grad_norm": 1.1419373750686646,
      "learning_rate": 4.8926014319809075e-06,
      "loss": 0.7177,
      "mean_token_accuracy": 0.7853572726249695,
      "num_tokens": 5533440.0,
      "step": 165
    },
    {
      "epoch": 0.010148042024832856,
      "grad_norm": 1.017261028289795,
      "learning_rate": 5.041766109785203e-06,
      "loss": 0.6972,
      "mean_token_accuracy": 0.7945604085922241,
      "num_tokens": 5701120.0,
      "step": 170
    },
    {
      "epoch": 0.010446513849092646,
      "grad_norm": 1.1668633222579956,
      "learning_rate": 5.190930787589499e-06,
      "loss": 0.6491,
      "mean_token_accuracy": 0.8055350184440613,
      "num_tokens": 5868800.0,
      "step": 175
    },
    {
      "epoch": 0.010744985673352435,
      "grad_norm": 0.9799959659576416,
      "learning_rate": 5.340095465393795e-06,
      "loss": 0.6254,
      "mean_token_accuracy": 0.8119766235351562,
      "num_tokens": 6036480.0,
      "step": 180
    },
    {
      "epoch": 0.011043457497612225,
      "grad_norm": 1.1462510824203491,
      "learning_rate": 5.489260143198091e-06,
      "loss": 0.6946,
      "mean_token_accuracy": 0.7927233815193176,
      "num_tokens": 6204160.0,
      "step": 185
    },
    {
      "epoch": 0.011341929321872016,
      "grad_norm": 1.156132459640503,
      "learning_rate": 5.638424821002386e-06,
      "loss": 0.7361,
      "mean_token_accuracy": 0.7793510675430297,
      "num_tokens": 6371840.0,
      "step": 190
    },
    {
      "epoch": 0.011640401146131804,
      "grad_norm": 1.0716009140014648,
      "learning_rate": 5.787589498806683e-06,
      "loss": 0.7308,
      "mean_token_accuracy": 0.7823571681976318,
      "num_tokens": 6539520.0,
      "step": 195
    },
    {
      "epoch": 0.011938872970391595,
      "grad_norm": 1.0945734977722168,
      "learning_rate": 5.936754176610979e-06,
      "loss": 0.6918,
      "mean_token_accuracy": 0.7942442893981934,
      "num_tokens": 6707200.0,
      "step": 200
    },
    {
      "epoch": 0.012237344794651385,
      "grad_norm": 1.0889852046966553,
      "learning_rate": 6.0859188544152745e-06,
      "loss": 0.6849,
      "mean_token_accuracy": 0.795276153087616,
      "num_tokens": 6874880.0,
      "step": 205
    },
    {
      "epoch": 0.012535816618911174,
      "grad_norm": 1.1265602111816406,
      "learning_rate": 6.23508353221957e-06,
      "loss": 0.6678,
      "mean_token_accuracy": 0.8014195442199707,
      "num_tokens": 7042560.0,
      "step": 210
    },
    {
      "epoch": 0.012834288443170964,
      "grad_norm": 1.0774636268615723,
      "learning_rate": 6.384248210023866e-06,
      "loss": 0.6611,
      "mean_token_accuracy": 0.8024692893028259,
      "num_tokens": 7210240.0,
      "step": 215
    },
    {
      "epoch": 0.013132760267430755,
      "grad_norm": 1.0906325578689575,
      "learning_rate": 6.533412887828163e-06,
      "loss": 0.6313,
      "mean_token_accuracy": 0.807944655418396,
      "num_tokens": 7377920.0,
      "step": 220
    },
    {
      "epoch": 0.013431232091690544,
      "grad_norm": 1.0445871353149414,
      "learning_rate": 6.682577565632458e-06,
      "loss": 0.6426,
      "mean_token_accuracy": 0.8060420036315918,
      "num_tokens": 7545600.0,
      "step": 225
    },
    {
      "epoch": 0.013729703915950334,
      "grad_norm": 1.1006596088409424,
      "learning_rate": 6.831742243436755e-06,
      "loss": 0.6281,
      "mean_token_accuracy": 0.8109686255455018,
      "num_tokens": 7713280.0,
      "step": 230
    },
    {
      "epoch": 0.014028175740210124,
      "grad_norm": 1.0418542623519897,
      "learning_rate": 6.980906921241051e-06,
      "loss": 0.6862,
      "mean_token_accuracy": 0.7943993926048278,
      "num_tokens": 7880960.0,
      "step": 235
    },
    {
      "epoch": 0.014326647564469915,
      "grad_norm": 1.0069177150726318,
      "learning_rate": 7.130071599045346e-06,
      "loss": 0.644,
      "mean_token_accuracy": 0.8059823513031006,
      "num_tokens": 8048640.0,
      "step": 240
    },
    {
      "epoch": 0.014625119388729704,
      "grad_norm": 1.1208637952804565,
      "learning_rate": 7.279236276849642e-06,
      "loss": 0.7192,
      "mean_token_accuracy": 0.7862698316574097,
      "num_tokens": 8216320.0,
      "step": 245
    },
    {
      "epoch": 0.014923591212989494,
      "grad_norm": 1.064270257949829,
      "learning_rate": 7.428400954653939e-06,
      "loss": 0.6886,
      "mean_token_accuracy": 0.7932064771652222,
      "num_tokens": 8384000.0,
      "step": 250
    },
    {
      "epoch": 0.015222063037249284,
      "grad_norm": 1.0383998155593872,
      "learning_rate": 7.577565632458235e-06,
      "loss": 0.6209,
      "mean_token_accuracy": 0.8120899319648742,
      "num_tokens": 8551680.0,
      "step": 255
    },
    {
      "epoch": 0.015520534861509073,
      "grad_norm": 1.0964311361312866,
      "learning_rate": 7.72673031026253e-06,
      "loss": 0.634,
      "mean_token_accuracy": 0.8092150688171387,
      "num_tokens": 8719360.0,
      "step": 260
    },
    {
      "epoch": 0.015819006685768862,
      "grad_norm": 1.1319653987884521,
      "learning_rate": 7.875894988066825e-06,
      "loss": 0.6299,
      "mean_token_accuracy": 0.8099904656410217,
      "num_tokens": 8887040.0,
      "step": 265
    },
    {
      "epoch": 0.016117478510028652,
      "grad_norm": 1.131912350654602,
      "learning_rate": 8.025059665871123e-06,
      "loss": 0.6335,
      "mean_token_accuracy": 0.8081832528114319,
      "num_tokens": 9054720.0,
      "step": 270
    },
    {
      "epoch": 0.016415950334288443,
      "grad_norm": 1.0844351053237915,
      "learning_rate": 8.174224343675418e-06,
      "loss": 0.6501,
      "mean_token_accuracy": 0.8047655940055847,
      "num_tokens": 9222400.0,
      "step": 275
    },
    {
      "epoch": 0.016714422158548233,
      "grad_norm": 1.0595623254776,
      "learning_rate": 8.323389021479713e-06,
      "loss": 0.6802,
      "mean_token_accuracy": 0.7956459403038025,
      "num_tokens": 9390080.0,
      "step": 280
    },
    {
      "epoch": 0.017012893982808024,
      "grad_norm": 1.1240721940994263,
      "learning_rate": 8.47255369928401e-06,
      "loss": 0.6342,
      "mean_token_accuracy": 0.8102648258209229,
      "num_tokens": 9557760.0,
      "step": 285
    },
    {
      "epoch": 0.017311365807067814,
      "grad_norm": 1.173274278640747,
      "learning_rate": 8.621718377088306e-06,
      "loss": 0.7119,
      "mean_token_accuracy": 0.7863771915435791,
      "num_tokens": 9725440.0,
      "step": 290
    },
    {
      "epoch": 0.0176098376313276,
      "grad_norm": 1.2010884284973145,
      "learning_rate": 8.770883054892602e-06,
      "loss": 0.6609,
      "mean_token_accuracy": 0.8019265174865723,
      "num_tokens": 9893120.0,
      "step": 295
    },
    {
      "epoch": 0.01790830945558739,
      "grad_norm": 1.0363003015518188,
      "learning_rate": 8.920047732696899e-06,
      "loss": 0.667,
      "mean_token_accuracy": 0.7988429069519043,
      "num_tokens": 10060800.0,
      "step": 300
    },
    {
      "epoch": 0.018206781279847182,
      "grad_norm": 1.1245827674865723,
      "learning_rate": 9.069212410501193e-06,
      "loss": 0.6813,
      "mean_token_accuracy": 0.7939043283462525,
      "num_tokens": 10228480.0,
      "step": 305
    },
    {
      "epoch": 0.018505253104106972,
      "grad_norm": 1.1633771657943726,
      "learning_rate": 9.218377088305489e-06,
      "loss": 0.6452,
      "mean_token_accuracy": 0.8054395914077759,
      "num_tokens": 10396160.0,
      "step": 310
    },
    {
      "epoch": 0.018803724928366763,
      "grad_norm": 1.150593638420105,
      "learning_rate": 9.367541766109786e-06,
      "loss": 0.7082,
      "mean_token_accuracy": 0.7869139909744263,
      "num_tokens": 10563840.0,
      "step": 315
    },
    {
      "epoch": 0.019102196752626553,
      "grad_norm": 1.0160619020462036,
      "learning_rate": 9.516706443914082e-06,
      "loss": 0.6665,
      "mean_token_accuracy": 0.7976798415184021,
      "num_tokens": 10731520.0,
      "step": 320
    },
    {
      "epoch": 0.019400668576886344,
      "grad_norm": 1.0531463623046875,
      "learning_rate": 9.665871121718378e-06,
      "loss": 0.6464,
      "mean_token_accuracy": 0.8056542992591857,
      "num_tokens": 10899200.0,
      "step": 325
    },
    {
      "epoch": 0.01969914040114613,
      "grad_norm": 1.1304893493652344,
      "learning_rate": 9.815035799522674e-06,
      "loss": 0.6846,
      "mean_token_accuracy": 0.793683660030365,
      "num_tokens": 11066880.0,
      "step": 330
    },
    {
      "epoch": 0.01999761222540592,
      "grad_norm": 1.0940200090408325,
      "learning_rate": 9.96420047732697e-06,
      "loss": 0.6519,
      "mean_token_accuracy": 0.802367889881134,
      "num_tokens": 11234560.0,
      "step": 335
    },
    {
      "epoch": 0.02029608404966571,
      "grad_norm": 1.0811877250671387,
      "learning_rate": 1.0113365155131265e-05,
      "loss": 0.631,
      "mean_token_accuracy": 0.8103244781494141,
      "num_tokens": 11402240.0,
      "step": 340
    },
    {
      "epoch": 0.020594555873925502,
      "grad_norm": 1.041429877281189,
      "learning_rate": 1.026252983293556e-05,
      "loss": 0.6405,
      "mean_token_accuracy": 0.8054097652435303,
      "num_tokens": 11569920.0,
      "step": 345
    },
    {
      "epoch": 0.020893027698185292,
      "grad_norm": 1.0615298748016357,
      "learning_rate": 1.0411694510739857e-05,
      "loss": 0.6304,
      "mean_token_accuracy": 0.809048080444336,
      "num_tokens": 11737600.0,
      "step": 350
    },
    {
      "epoch": 0.021191499522445083,
      "grad_norm": 1.08322274684906,
      "learning_rate": 1.0560859188544154e-05,
      "loss": 0.681,
      "mean_token_accuracy": 0.7950972080230713,
      "num_tokens": 11905280.0,
      "step": 355
    },
    {
      "epoch": 0.02148997134670487,
      "grad_norm": 1.1768213510513306,
      "learning_rate": 1.071002386634845e-05,
      "loss": 0.6674,
      "mean_token_accuracy": 0.8011630654335022,
      "num_tokens": 12072960.0,
      "step": 360
    },
    {
      "epoch": 0.02178844317096466,
      "grad_norm": 1.1367334127426147,
      "learning_rate": 1.0859188544152746e-05,
      "loss": 0.6181,
      "mean_token_accuracy": 0.8141178607940673,
      "num_tokens": 12240640.0,
      "step": 365
    },
    {
      "epoch": 0.02208691499522445,
      "grad_norm": 1.0722662210464478,
      "learning_rate": 1.1008353221957041e-05,
      "loss": 0.6741,
      "mean_token_accuracy": 0.7966122031211853,
      "num_tokens": 12408320.0,
      "step": 370
    },
    {
      "epoch": 0.02238538681948424,
      "grad_norm": 1.0842481851577759,
      "learning_rate": 1.1157517899761337e-05,
      "loss": 0.6307,
      "mean_token_accuracy": 0.8092866659164428,
      "num_tokens": 12576000.0,
      "step": 375
    },
    {
      "epoch": 0.02268385864374403,
      "grad_norm": 1.1080683469772339,
      "learning_rate": 1.1306682577565633e-05,
      "loss": 0.6275,
      "mean_token_accuracy": 0.8088691473007202,
      "num_tokens": 12743680.0,
      "step": 380
    },
    {
      "epoch": 0.022982330468003822,
      "grad_norm": 1.1539459228515625,
      "learning_rate": 1.145584725536993e-05,
      "loss": 0.6832,
      "mean_token_accuracy": 0.7943695425987244,
      "num_tokens": 12911360.0,
      "step": 385
    },
    {
      "epoch": 0.02328080229226361,
      "grad_norm": 1.051101565361023,
      "learning_rate": 1.1605011933174224e-05,
      "loss": 0.695,
      "mean_token_accuracy": 0.7907431840896606,
      "num_tokens": 13079040.0,
      "step": 390
    },
    {
      "epoch": 0.0235792741165234,
      "grad_norm": 1.1492129564285278,
      "learning_rate": 1.175417661097852e-05,
      "loss": 0.6588,
      "mean_token_accuracy": 0.8029404759407044,
      "num_tokens": 13246720.0,
      "step": 395
    },
    {
      "epoch": 0.02387774594078319,
      "grad_norm": 1.0017367601394653,
      "learning_rate": 1.1903341288782818e-05,
      "loss": 0.6531,
      "mean_token_accuracy": 0.8003161191940308,
      "num_tokens": 13414400.0,
      "step": 400
    },
    {
      "epoch": 0.02417621776504298,
      "grad_norm": 1.1234192848205566,
      "learning_rate": 1.2052505966587113e-05,
      "loss": 0.6429,
      "mean_token_accuracy": 0.805439579486847,
      "num_tokens": 13582080.0,
      "step": 405
    },
    {
      "epoch": 0.02447468958930277,
      "grad_norm": 1.051469326019287,
      "learning_rate": 1.2201670644391407e-05,
      "loss": 0.6853,
      "mean_token_accuracy": 0.792443037033081,
      "num_tokens": 13749760.0,
      "step": 410
    },
    {
      "epoch": 0.02477316141356256,
      "grad_norm": 1.0647557973861694,
      "learning_rate": 1.2350835322195705e-05,
      "loss": 0.6633,
      "mean_token_accuracy": 0.7992425084114074,
      "num_tokens": 13917440.0,
      "step": 415
    },
    {
      "epoch": 0.025071633237822348,
      "grad_norm": 1.1210533380508423,
      "learning_rate": 1.25e-05,
      "loss": 0.6623,
      "mean_token_accuracy": 0.799260425567627,
      "num_tokens": 14085120.0,
      "step": 420
    },
    {
      "epoch": 0.02537010506208214,
      "grad_norm": 1.1232095956802368,
      "learning_rate": 1.2649164677804295e-05,
      "loss": 0.6428,
      "mean_token_accuracy": 0.803471314907074,
      "num_tokens": 14252800.0,
      "step": 425
    },
    {
      "epoch": 0.02566857688634193,
      "grad_norm": 1.2449058294296265,
      "learning_rate": 1.2798329355608594e-05,
      "loss": 0.627,
      "mean_token_accuracy": 0.8090003490447998,
      "num_tokens": 14420480.0,
      "step": 430
    },
    {
      "epoch": 0.02596704871060172,
      "grad_norm": 1.0878791809082031,
      "learning_rate": 1.2947494033412888e-05,
      "loss": 0.6265,
      "mean_token_accuracy": 0.8103662133216858,
      "num_tokens": 14588160.0,
      "step": 435
    },
    {
      "epoch": 0.02626552053486151,
      "grad_norm": 1.0709093809127808,
      "learning_rate": 1.3096658711217185e-05,
      "loss": 0.6187,
      "mean_token_accuracy": 0.8117499709129333,
      "num_tokens": 14755840.0,
      "step": 440
    },
    {
      "epoch": 0.0265639923591213,
      "grad_norm": 1.0762311220169067,
      "learning_rate": 1.3245823389021481e-05,
      "loss": 0.6245,
      "mean_token_accuracy": 0.8105570912361145,
      "num_tokens": 14923520.0,
      "step": 445
    },
    {
      "epoch": 0.026862464183381087,
      "grad_norm": 1.1213608980178833,
      "learning_rate": 1.3394988066825775e-05,
      "loss": 0.658,
      "mean_token_accuracy": 0.8015686511993408,
      "num_tokens": 15091200.0,
      "step": 450
    },
    {
      "epoch": 0.027160936007640878,
      "grad_norm": 1.147695779800415,
      "learning_rate": 1.3544152744630073e-05,
      "loss": 0.6844,
      "mean_token_accuracy": 0.7944948196411132,
      "num_tokens": 15258880.0,
      "step": 455
    },
    {
      "epoch": 0.027459407831900668,
      "grad_norm": 1.0717793703079224,
      "learning_rate": 1.3693317422434368e-05,
      "loss": 0.6508,
      "mean_token_accuracy": 0.802367889881134,
      "num_tokens": 15426560.0,
      "step": 460
    },
    {
      "epoch": 0.02775787965616046,
      "grad_norm": 1.0311919450759888,
      "learning_rate": 1.3842482100238662e-05,
      "loss": 0.5861,
      "mean_token_accuracy": 0.8208457589149475,
      "num_tokens": 15594240.0,
      "step": 465
    },
    {
      "epoch": 0.02805635148042025,
      "grad_norm": 1.1428905725479126,
      "learning_rate": 1.399164677804296e-05,
      "loss": 0.5845,
      "mean_token_accuracy": 0.8203686118125916,
      "num_tokens": 15761920.0,
      "step": 470
    },
    {
      "epoch": 0.02835482330468004,
      "grad_norm": 1.1685853004455566,
      "learning_rate": 1.4140811455847256e-05,
      "loss": 0.6442,
      "mean_token_accuracy": 0.8049325942993164,
      "num_tokens": 15929600.0,
      "step": 475
    },
    {
      "epoch": 0.02865329512893983,
      "grad_norm": 1.172850251197815,
      "learning_rate": 1.4289976133651553e-05,
      "loss": 0.632,
      "mean_token_accuracy": 0.8073124289512634,
      "num_tokens": 16097280.0,
      "step": 480
    },
    {
      "epoch": 0.028951766953199617,
      "grad_norm": 0.9494922757148743,
      "learning_rate": 1.4439140811455847e-05,
      "loss": 0.619,
      "mean_token_accuracy": 0.8106346249580383,
      "num_tokens": 16264960.0,
      "step": 485
    },
    {
      "epoch": 0.029250238777459407,
      "grad_norm": 1.1580462455749512,
      "learning_rate": 1.4588305489260143e-05,
      "loss": 0.7286,
      "mean_token_accuracy": 0.7803709864616394,
      "num_tokens": 16432640.0,
      "step": 490
    },
    {
      "epoch": 0.029548710601719198,
      "grad_norm": 1.0796221494674683,
      "learning_rate": 1.473747016706444e-05,
      "loss": 0.6096,
      "mean_token_accuracy": 0.8138792872428894,
      "num_tokens": 16600320.0,
      "step": 495
    },
    {
      "epoch": 0.029847182425978988,
      "grad_norm": 1.0365972518920898,
      "learning_rate": 1.4886634844868736e-05,
      "loss": 0.6267,
      "mean_token_accuracy": 0.8090361475944519,
      "num_tokens": 16768000.0,
      "step": 500
    },
    {
      "epoch": 0.03014565425023878,
      "grad_norm": 1.0467289686203003,
      "learning_rate": 1.5035799522673034e-05,
      "loss": 0.6976,
      "mean_token_accuracy": 0.7893355488777161,
      "num_tokens": 16935680.0,
      "step": 505
    },
    {
      "epoch": 0.03044412607449857,
      "grad_norm": 1.037341833114624,
      "learning_rate": 1.5184964200477328e-05,
      "loss": 0.6205,
      "mean_token_accuracy": 0.8109686255455018,
      "num_tokens": 17103360.0,
      "step": 510
    },
    {
      "epoch": 0.030742597898758356,
      "grad_norm": 1.044974446296692,
      "learning_rate": 1.5334128878281622e-05,
      "loss": 0.6095,
      "mean_token_accuracy": 0.8160264730453491,
      "num_tokens": 17271040.0,
      "step": 515
    },
    {
      "epoch": 0.031041069723018146,
      "grad_norm": 1.0821094512939453,
      "learning_rate": 1.548329355608592e-05,
      "loss": 0.6583,
      "mean_token_accuracy": 0.8012883186340332,
      "num_tokens": 17438720.0,
      "step": 520
    },
    {
      "epoch": 0.03133954154727794,
      "grad_norm": 1.1046359539031982,
      "learning_rate": 1.5632458233890217e-05,
      "loss": 0.637,
      "mean_token_accuracy": 0.8047178745269775,
      "num_tokens": 17606400.0,
      "step": 525
    },
    {
      "epoch": 0.031638013371537724,
      "grad_norm": 0.9659026265144348,
      "learning_rate": 1.578162291169451e-05,
      "loss": 0.6691,
      "mean_token_accuracy": 0.7965585112571716,
      "num_tokens": 17774080.0,
      "step": 530
    },
    {
      "epoch": 0.03193648519579752,
      "grad_norm": 1.0919208526611328,
      "learning_rate": 1.5930787589498808e-05,
      "loss": 0.7246,
      "mean_token_accuracy": 0.7839854478836059,
      "num_tokens": 17941760.0,
      "step": 535
    },
    {
      "epoch": 0.032234957020057305,
      "grad_norm": 1.0334947109222412,
      "learning_rate": 1.6079952267303104e-05,
      "loss": 0.5953,
      "mean_token_accuracy": 0.8185196161270142,
      "num_tokens": 18109440.0,
      "step": 540
    },
    {
      "epoch": 0.0325334288443171,
      "grad_norm": 1.0578919649124146,
      "learning_rate": 1.62291169451074e-05,
      "loss": 0.6199,
      "mean_token_accuracy": 0.8102051854133606,
      "num_tokens": 18277120.0,
      "step": 545
    },
    {
      "epoch": 0.032831900668576885,
      "grad_norm": 1.219687581062317,
      "learning_rate": 1.6378281622911695e-05,
      "loss": 0.6447,
      "mean_token_accuracy": 0.8032029032707214,
      "num_tokens": 18444800.0,
      "step": 550
    },
    {
      "epoch": 0.03313037249283668,
      "grad_norm": 1.0638045072555542,
      "learning_rate": 1.652744630071599e-05,
      "loss": 0.6361,
      "mean_token_accuracy": 0.8058511257171631,
      "num_tokens": 18612480.0,
      "step": 555
    },
    {
      "epoch": 0.033428844317096466,
      "grad_norm": 1.2430840730667114,
      "learning_rate": 1.6676610978520287e-05,
      "loss": 0.6715,
      "mean_token_accuracy": 0.7967221140861511,
      "num_tokens": 18773299.0,
      "step": 560
    },
    {
      "epoch": 0.03372731614135625,
      "grad_norm": 1.077887773513794,
      "learning_rate": 1.6825775656324583e-05,
      "loss": 0.6505,
      "mean_token_accuracy": 0.8032744884490967,
      "num_tokens": 18940979.0,
      "step": 565
    },
    {
      "epoch": 0.03402578796561605,
      "grad_norm": 1.2823841571807861,
      "learning_rate": 1.6974940334128882e-05,
      "loss": 0.6655,
      "mean_token_accuracy": 0.8005189180374146,
      "num_tokens": 19108659.0,
      "step": 570
    },
    {
      "epoch": 0.034324259789875834,
      "grad_norm": 1.1820347309112549,
      "learning_rate": 1.7124105011933174e-05,
      "loss": 0.6662,
      "mean_token_accuracy": 0.7994632124900818,
      "num_tokens": 19276339.0,
      "step": 575
    },
    {
      "epoch": 0.03462273161413563,
      "grad_norm": 1.2058839797973633,
      "learning_rate": 1.727326968973747e-05,
      "loss": 0.6852,
      "mean_token_accuracy": 0.7928009033203125,
      "num_tokens": 19444019.0,
      "step": 580
    },
    {
      "epoch": 0.034921203438395415,
      "grad_norm": 1.1294993162155151,
      "learning_rate": 1.742243436754177e-05,
      "loss": 0.6431,
      "mean_token_accuracy": 0.8039544343948364,
      "num_tokens": 19611699.0,
      "step": 585
    },
    {
      "epoch": 0.0352196752626552,
      "grad_norm": 1.0808647871017456,
      "learning_rate": 1.757159904534606e-05,
      "loss": 0.6496,
      "mean_token_accuracy": 0.8030716896057128,
      "num_tokens": 19779379.0,
      "step": 590
    },
    {
      "epoch": 0.035518147086914996,
      "grad_norm": 1.1090660095214844,
      "learning_rate": 1.7720763723150357e-05,
      "loss": 0.6256,
      "mean_token_accuracy": 0.8078014969825744,
      "num_tokens": 19947059.0,
      "step": 595
    },
    {
      "epoch": 0.03581661891117478,
      "grad_norm": 1.2074198722839355,
      "learning_rate": 1.7869928400954656e-05,
      "loss": 0.6983,
      "mean_token_accuracy": 0.7884349346160888,
      "num_tokens": 20114739.0,
      "step": 600
    },
    {
      "epoch": 0.03611509073543458,
      "grad_norm": 1.1048452854156494,
      "learning_rate": 1.801909307875895e-05,
      "loss": 0.5936,
      "mean_token_accuracy": 0.8177740693092346,
      "num_tokens": 20282419.0,
      "step": 605
    },
    {
      "epoch": 0.036413562559694364,
      "grad_norm": 1.1176884174346924,
      "learning_rate": 1.8168257756563248e-05,
      "loss": 0.6529,
      "mean_token_accuracy": 0.8003340005874634,
      "num_tokens": 20450099.0,
      "step": 610
    },
    {
      "epoch": 0.03671203438395416,
      "grad_norm": 0.9916646480560303,
      "learning_rate": 1.8317422434367544e-05,
      "loss": 0.6221,
      "mean_token_accuracy": 0.8111475586891175,
      "num_tokens": 20617779.0,
      "step": 615
    },
    {
      "epoch": 0.037010506208213945,
      "grad_norm": 0.9652239680290222,
      "learning_rate": 1.8466587112171836e-05,
      "loss": 0.6334,
      "mean_token_accuracy": 0.8067756175994873,
      "num_tokens": 20785459.0,
      "step": 620
    },
    {
      "epoch": 0.03730897803247373,
      "grad_norm": 0.9723144173622131,
      "learning_rate": 1.8615751789976135e-05,
      "loss": 0.6706,
      "mean_token_accuracy": 0.7959560990333557,
      "num_tokens": 20953139.0,
      "step": 625
    },
    {
      "epoch": 0.037607449856733526,
      "grad_norm": 1.0488835573196411,
      "learning_rate": 1.876491646778043e-05,
      "loss": 0.6377,
      "mean_token_accuracy": 0.8059704065322876,
      "num_tokens": 21120819.0,
      "step": 630
    },
    {
      "epoch": 0.03790592168099331,
      "grad_norm": 0.9998219609260559,
      "learning_rate": 1.8914081145584727e-05,
      "loss": 0.647,
      "mean_token_accuracy": 0.800990104675293,
      "num_tokens": 21288499.0,
      "step": 635
    },
    {
      "epoch": 0.038204393505253106,
      "grad_norm": 0.9908907413482666,
      "learning_rate": 1.9063245823389023e-05,
      "loss": 0.6108,
      "mean_token_accuracy": 0.8143624067306519,
      "num_tokens": 21456179.0,
      "step": 640
    },
    {
      "epoch": 0.03850286532951289,
      "grad_norm": 1.0818791389465332,
      "learning_rate": 1.9212410501193318e-05,
      "loss": 0.683,
      "mean_token_accuracy": 0.7940295815467835,
      "num_tokens": 21623859.0,
      "step": 645
    },
    {
      "epoch": 0.03880133715377269,
      "grad_norm": 0.9645847082138062,
      "learning_rate": 1.9361575178997614e-05,
      "loss": 0.6214,
      "mean_token_accuracy": 0.812030303478241,
      "num_tokens": 21791539.0,
      "step": 650
    },
    {
      "epoch": 0.039099808978032474,
      "grad_norm": 1.0143963098526,
      "learning_rate": 1.951073985680191e-05,
      "loss": 0.6465,
      "mean_token_accuracy": 0.8018310785293579,
      "num_tokens": 21959219.0,
      "step": 655
    },
    {
      "epoch": 0.03939828080229226,
      "grad_norm": 0.9881317615509033,
      "learning_rate": 1.9659904534606206e-05,
      "loss": 0.6641,
      "mean_token_accuracy": 0.7982643365859985,
      "num_tokens": 22126899.0,
      "step": 660
    },
    {
      "epoch": 0.039696752626552055,
      "grad_norm": 1.0634610652923584,
      "learning_rate": 1.98090692124105e-05,
      "loss": 0.678,
      "mean_token_accuracy": 0.794172739982605,
      "num_tokens": 22294579.0,
      "step": 665
    },
    {
      "epoch": 0.03999522445081184,
      "grad_norm": 1.3175532817840576,
      "learning_rate": 1.9958233890214797e-05,
      "loss": 0.6285,
      "mean_token_accuracy": 0.8092270016670227,
      "num_tokens": 22462259.0,
      "step": 670
    },
    {
      "epoch": 0.040293696275071636,
      "grad_norm": 1.6304832696914673,
      "learning_rate": 2.0107398568019096e-05,
      "loss": 0.6206,
      "mean_token_accuracy": 0.8117857575416565,
      "num_tokens": 22629939.0,
      "step": 675
    },
    {
      "epoch": 0.04059216809933142,
      "grad_norm": 1.0529224872589111,
      "learning_rate": 2.025656324582339e-05,
      "loss": 0.6384,
      "mean_token_accuracy": 0.8057914733886719,
      "num_tokens": 22797619.0,
      "step": 680
    },
    {
      "epoch": 0.04089063992359121,
      "grad_norm": 0.9856318831443787,
      "learning_rate": 2.0405727923627684e-05,
      "loss": 0.6364,
      "mean_token_accuracy": 0.8046761274337768,
      "num_tokens": 22965299.0,
      "step": 685
    },
    {
      "epoch": 0.041189111747851004,
      "grad_norm": 1.0770418643951416,
      "learning_rate": 2.0554892601431983e-05,
      "loss": 0.6511,
      "mean_token_accuracy": 0.80174161195755,
      "num_tokens": 23132979.0,
      "step": 690
    },
    {
      "epoch": 0.04148758357211079,
      "grad_norm": 1.185211181640625,
      "learning_rate": 2.0704057279236276e-05,
      "loss": 0.6757,
      "mean_token_accuracy": 0.7944113135337829,
      "num_tokens": 23300659.0,
      "step": 695
    },
    {
      "epoch": 0.041786055396370585,
      "grad_norm": 1.0646095275878906,
      "learning_rate": 2.0853221957040575e-05,
      "loss": 0.71,
      "mean_token_accuracy": 0.7840629816055298,
      "num_tokens": 23468339.0,
      "step": 700
    },
    {
      "epoch": 0.04208452722063037,
      "grad_norm": 1.0418790578842163,
      "learning_rate": 2.100238663484487e-05,
      "loss": 0.6645,
      "mean_token_accuracy": 0.7989860415458679,
      "num_tokens": 23636019.0,
      "step": 705
    },
    {
      "epoch": 0.042382999044890166,
      "grad_norm": 1.008773684501648,
      "learning_rate": 2.1151551312649167e-05,
      "loss": 0.6334,
      "mean_token_accuracy": 0.8068233370780945,
      "num_tokens": 23803699.0,
      "step": 710
    },
    {
      "epoch": 0.04268147086914995,
      "grad_norm": 1.1428587436676025,
      "learning_rate": 2.1300715990453462e-05,
      "loss": 0.6831,
      "mean_token_accuracy": 0.7940176486968994,
      "num_tokens": 23971379.0,
      "step": 715
    },
    {
      "epoch": 0.04297994269340974,
      "grad_norm": 1.1082713603973389,
      "learning_rate": 2.1449880668257758e-05,
      "loss": 0.6637,
      "mean_token_accuracy": 0.7991053223609924,
      "num_tokens": 24139059.0,
      "step": 720
    },
    {
      "epoch": 0.04327841451766953,
      "grad_norm": 1.0348970890045166,
      "learning_rate": 2.1599045346062054e-05,
      "loss": 0.6329,
      "mean_token_accuracy": 0.807855200767517,
      "num_tokens": 24306739.0,
      "step": 725
    },
    {
      "epoch": 0.04357688634192932,
      "grad_norm": 0.9988929033279419,
      "learning_rate": 2.174821002386635e-05,
      "loss": 0.6476,
      "mean_token_accuracy": 0.8021412372589112,
      "num_tokens": 24474419.0,
      "step": 730
    },
    {
      "epoch": 0.043875358166189114,
      "grad_norm": 1.1111482381820679,
      "learning_rate": 2.1897374701670645e-05,
      "loss": 0.648,
      "mean_token_accuracy": 0.8020696640014648,
      "num_tokens": 24642099.0,
      "step": 735
    },
    {
      "epoch": 0.0441738299904489,
      "grad_norm": 1.0283012390136719,
      "learning_rate": 2.204653937947494e-05,
      "loss": 0.6115,
      "mean_token_accuracy": 0.8134319424629212,
      "num_tokens": 24809779.0,
      "step": 740
    },
    {
      "epoch": 0.04447230181470869,
      "grad_norm": 1.0280516147613525,
      "learning_rate": 2.2195704057279237e-05,
      "loss": 0.647,
      "mean_token_accuracy": 0.8046164870262146,
      "num_tokens": 24977459.0,
      "step": 745
    },
    {
      "epoch": 0.04477077363896848,
      "grad_norm": 0.9584552049636841,
      "learning_rate": 2.2344868735083533e-05,
      "loss": 0.6557,
      "mean_token_accuracy": 0.8007097721099854,
      "num_tokens": 25145139.0,
      "step": 750
    },
    {
      "epoch": 0.04506924546322827,
      "grad_norm": 1.1966413259506226,
      "learning_rate": 2.249403341288783e-05,
      "loss": 0.7648,
      "mean_token_accuracy": 0.7722533583641052,
      "num_tokens": 25312819.0,
      "step": 755
    },
    {
      "epoch": 0.04536771728748806,
      "grad_norm": 1.034635066986084,
      "learning_rate": 2.2643198090692124e-05,
      "loss": 0.6048,
      "mean_token_accuracy": 0.8161875128746032,
      "num_tokens": 25480499.0,
      "step": 760
    },
    {
      "epoch": 0.04566618911174785,
      "grad_norm": 0.9553530812263489,
      "learning_rate": 2.2792362768496423e-05,
      "loss": 0.7142,
      "mean_token_accuracy": 0.7846653819084167,
      "num_tokens": 25648179.0,
      "step": 765
    },
    {
      "epoch": 0.045964660936007644,
      "grad_norm": 0.9542186856269836,
      "learning_rate": 2.294152744630072e-05,
      "loss": 0.6467,
      "mean_token_accuracy": 0.8023380637168884,
      "num_tokens": 25815859.0,
      "step": 770
    },
    {
      "epoch": 0.04626313276026743,
      "grad_norm": 1.106665849685669,
      "learning_rate": 2.309069212410501e-05,
      "loss": 0.6754,
      "mean_token_accuracy": 0.7965465784072876,
      "num_tokens": 25983539.0,
      "step": 775
    },
    {
      "epoch": 0.04656160458452722,
      "grad_norm": 1.0282317399978638,
      "learning_rate": 2.323985680190931e-05,
      "loss": 0.7013,
      "mean_token_accuracy": 0.7864428043365479,
      "num_tokens": 26151219.0,
      "step": 780
    },
    {
      "epoch": 0.04686007640878701,
      "grad_norm": 1.0247565507888794,
      "learning_rate": 2.3389021479713606e-05,
      "loss": 0.6184,
      "mean_token_accuracy": 0.8110700130462647,
      "num_tokens": 26318899.0,
      "step": 785
    },
    {
      "epoch": 0.0471585482330468,
      "grad_norm": 0.9614169597625732,
      "learning_rate": 2.35381861575179e-05,
      "loss": 0.6165,
      "mean_token_accuracy": 0.8113205313682557,
      "num_tokens": 26486579.0,
      "step": 790
    },
    {
      "epoch": 0.04745702005730659,
      "grad_norm": 1.0571271181106567,
      "learning_rate": 2.3687350835322198e-05,
      "loss": 0.6383,
      "mean_token_accuracy": 0.8051592469215393,
      "num_tokens": 26654259.0,
      "step": 795
    },
    {
      "epoch": 0.04775549188156638,
      "grad_norm": 1.024620532989502,
      "learning_rate": 2.3836515513126494e-05,
      "loss": 0.6759,
      "mean_token_accuracy": 0.7939759016036987,
      "num_tokens": 26821939.0,
      "step": 800
    },
    {
      "epoch": 0.04805396370582617,
      "grad_norm": 1.002708077430725,
      "learning_rate": 2.398568019093079e-05,
      "loss": 0.6626,
      "mean_token_accuracy": 0.7992604017257691,
      "num_tokens": 26989619.0,
      "step": 805
    },
    {
      "epoch": 0.04835243553008596,
      "grad_norm": 1.1294392347335815,
      "learning_rate": 2.4134844868735085e-05,
      "loss": 0.6854,
      "mean_token_accuracy": 0.7950435400009155,
      "num_tokens": 27157299.0,
      "step": 810
    },
    {
      "epoch": 0.04865090735434575,
      "grad_norm": 1.0987032651901245,
      "learning_rate": 2.428400954653938e-05,
      "loss": 0.6595,
      "mean_token_accuracy": 0.7997435331344604,
      "num_tokens": 27324979.0,
      "step": 815
    },
    {
      "epoch": 0.04894937917860554,
      "grad_norm": 1.0007015466690063,
      "learning_rate": 2.4433174224343677e-05,
      "loss": 0.6433,
      "mean_token_accuracy": 0.8039246082305909,
      "num_tokens": 27492659.0,
      "step": 820
    },
    {
      "epoch": 0.04924785100286533,
      "grad_norm": 1.1488895416259766,
      "learning_rate": 2.4582338902147972e-05,
      "loss": 0.6779,
      "mean_token_accuracy": 0.7940593957901001,
      "num_tokens": 27660339.0,
      "step": 825
    },
    {
      "epoch": 0.04954632282712512,
      "grad_norm": 1.0630240440368652,
      "learning_rate": 2.473150357995227e-05,
      "loss": 0.6117,
      "mean_token_accuracy": 0.81239413022995,
      "num_tokens": 27828019.0,
      "step": 830
    },
    {
      "epoch": 0.04984479465138491,
      "grad_norm": 1.0150984525680542,
      "learning_rate": 2.4880668257756564e-05,
      "loss": 0.6497,
      "mean_token_accuracy": 0.8021114230155945,
      "num_tokens": 27995699.0,
      "step": 835
    },
    {
      "epoch": 0.050143266475644696,
      "grad_norm": 0.9663558006286621,
      "learning_rate": 2.502983293556086e-05,
      "loss": 0.5899,
      "mean_token_accuracy": 0.8176234364509583,
      "num_tokens": 28158354.0,
      "step": 840
    },
    {
      "epoch": 0.05044173829990449,
      "grad_norm": 0.8831707239151001,
      "learning_rate": 2.517899761336516e-05,
      "loss": 0.6278,
      "mean_token_accuracy": 0.8061493515968323,
      "num_tokens": 28326034.0,
      "step": 845
    },
    {
      "epoch": 0.05074021012416428,
      "grad_norm": 1.0511727333068848,
      "learning_rate": 2.5328162291169455e-05,
      "loss": 0.6259,
      "mean_token_accuracy": 0.8092270135879517,
      "num_tokens": 28493714.0,
      "step": 850
    },
    {
      "epoch": 0.05103868194842407,
      "grad_norm": 1.0011943578720093,
      "learning_rate": 2.5477326968973747e-05,
      "loss": 0.6973,
      "mean_token_accuracy": 0.7880174160003662,
      "num_tokens": 28661394.0,
      "step": 855
    },
    {
      "epoch": 0.05133715377268386,
      "grad_norm": 1.098341941833496,
      "learning_rate": 2.5626491646778046e-05,
      "loss": 0.7285,
      "mean_token_accuracy": 0.781629490852356,
      "num_tokens": 28829074.0,
      "step": 860
    },
    {
      "epoch": 0.05163562559694365,
      "grad_norm": 1.0473344326019287,
      "learning_rate": 2.5775656324582342e-05,
      "loss": 0.6242,
      "mean_token_accuracy": 0.8083561778068542,
      "num_tokens": 28996754.0,
      "step": 865
    },
    {
      "epoch": 0.05193409742120344,
      "grad_norm": 1.2244330644607544,
      "learning_rate": 2.5924821002386634e-05,
      "loss": 0.639,
      "mean_token_accuracy": 0.8053978323936463,
      "num_tokens": 29164434.0,
      "step": 870
    },
    {
      "epoch": 0.052232569245463226,
      "grad_norm": 1.067139983177185,
      "learning_rate": 2.6073985680190933e-05,
      "loss": 0.7022,
      "mean_token_accuracy": 0.7878802418708801,
      "num_tokens": 29332114.0,
      "step": 875
    },
    {
      "epoch": 0.05253104106972302,
      "grad_norm": 0.8787959814071655,
      "learning_rate": 2.622315035799523e-05,
      "loss": 0.6115,
      "mean_token_accuracy": 0.8129667282104492,
      "num_tokens": 29499794.0,
      "step": 880
    },
    {
      "epoch": 0.05282951289398281,
      "grad_norm": 0.9647467136383057,
      "learning_rate": 2.637231503579952e-05,
      "loss": 0.6358,
      "mean_token_accuracy": 0.8050101399421692,
      "num_tokens": 29667474.0,
      "step": 885
    },
    {
      "epoch": 0.0531279847182426,
      "grad_norm": 1.1024103164672852,
      "learning_rate": 2.652147971360382e-05,
      "loss": 0.6797,
      "mean_token_accuracy": 0.7921806097030639,
      "num_tokens": 29835154.0,
      "step": 890
    },
    {
      "epoch": 0.05342645654250239,
      "grad_norm": 0.9410717487335205,
      "learning_rate": 2.6670644391408116e-05,
      "loss": 0.6672,
      "mean_token_accuracy": 0.7984790682792664,
      "num_tokens": 30002834.0,
      "step": 895
    },
    {
      "epoch": 0.053724928366762174,
      "grad_norm": 0.9704106450080872,
      "learning_rate": 2.681980906921241e-05,
      "loss": 0.6639,
      "mean_token_accuracy": 0.7983418941497803,
      "num_tokens": 30170514.0,
      "step": 900
    },
    {
      "epoch": 0.05402340019102197,
      "grad_norm": 1.0035582780838013,
      "learning_rate": 2.6968973747016708e-05,
      "loss": 0.6865,
      "mean_token_accuracy": 0.7923416495323181,
      "num_tokens": 30338194.0,
      "step": 905
    },
    {
      "epoch": 0.054321872015281755,
      "grad_norm": 0.8491612672805786,
      "learning_rate": 2.7118138424821004e-05,
      "loss": 0.5953,
      "mean_token_accuracy": 0.8169628977775574,
      "num_tokens": 30505874.0,
      "step": 910
    },
    {
      "epoch": 0.05462034383954155,
      "grad_norm": 1.0254725217819214,
      "learning_rate": 2.7267303102625303e-05,
      "loss": 0.6688,
      "mean_token_accuracy": 0.796892523765564,
      "num_tokens": 30673554.0,
      "step": 915
    },
    {
      "epoch": 0.054918815663801336,
      "grad_norm": 0.8602766394615173,
      "learning_rate": 2.7416467780429595e-05,
      "loss": 0.6036,
      "mean_token_accuracy": 0.8152809262275695,
      "num_tokens": 30841234.0,
      "step": 920
    },
    {
      "epoch": 0.05521728748806113,
      "grad_norm": 1.065596103668213,
      "learning_rate": 2.756563245823389e-05,
      "loss": 0.6769,
      "mean_token_accuracy": 0.7956161379814148,
      "num_tokens": 31008914.0,
      "step": 925
    },
    {
      "epoch": 0.05551575931232092,
      "grad_norm": 0.9849673509597778,
      "learning_rate": 2.771479713603819e-05,
      "loss": 0.6844,
      "mean_token_accuracy": 0.7949421405792236,
      "num_tokens": 31176594.0,
      "step": 930
    },
    {
      "epoch": 0.055814231136580704,
      "grad_norm": 0.9741759896278381,
      "learning_rate": 2.7863961813842482e-05,
      "loss": 0.7201,
      "mean_token_accuracy": 0.7840331554412842,
      "num_tokens": 31344274.0,
      "step": 935
    },
    {
      "epoch": 0.0561127029608405,
      "grad_norm": 1.0898329019546509,
      "learning_rate": 2.8013126491646778e-05,
      "loss": 0.6348,
      "mean_token_accuracy": 0.8053441405296325,
      "num_tokens": 31511954.0,
      "step": 940
    },
    {
      "epoch": 0.056411174785100285,
      "grad_norm": 0.9574422836303711,
      "learning_rate": 2.8162291169451077e-05,
      "loss": 0.6288,
      "mean_token_accuracy": 0.809197187423706,
      "num_tokens": 31679634.0,
      "step": 945
    },
    {
      "epoch": 0.05670964660936008,
      "grad_norm": 1.055638313293457,
      "learning_rate": 2.831145584725537e-05,
      "loss": 0.6808,
      "mean_token_accuracy": 0.7933794617652893,
      "num_tokens": 31847314.0,
      "step": 950
    },
    {
      "epoch": 0.057008118433619866,
      "grad_norm": 0.9862667322158813,
      "learning_rate": 2.8460620525059666e-05,
      "loss": 0.6321,
      "mean_token_accuracy": 0.807079803943634,
      "num_tokens": 32014994.0,
      "step": 955
    },
    {
      "epoch": 0.05730659025787966,
      "grad_norm": 1.005582332611084,
      "learning_rate": 2.8609785202863965e-05,
      "loss": 0.658,
      "mean_token_accuracy": 0.7998329877853394,
      "num_tokens": 32182674.0,
      "step": 960
    },
    {
      "epoch": 0.05760506208213945,
      "grad_norm": 0.9280803203582764,
      "learning_rate": 2.8758949880668257e-05,
      "loss": 0.5842,
      "mean_token_accuracy": 0.8203745722770691,
      "num_tokens": 32350354.0,
      "step": 965
    },
    {
      "epoch": 0.057903533906399234,
      "grad_norm": 0.9932334423065186,
      "learning_rate": 2.8908114558472553e-05,
      "loss": 0.6319,
      "mean_token_accuracy": 0.8079804301261901,
      "num_tokens": 32518034.0,
      "step": 970
    },
    {
      "epoch": 0.05820200573065903,
      "grad_norm": 0.9722811579704285,
      "learning_rate": 2.9057279236276852e-05,
      "loss": 0.6078,
      "mean_token_accuracy": 0.8157580852508545,
      "num_tokens": 32685714.0,
      "step": 975
    },
    {
      "epoch": 0.058500477554918814,
      "grad_norm": 0.9054681658744812,
      "learning_rate": 2.920644391408115e-05,
      "loss": 0.626,
      "mean_token_accuracy": 0.8077597498893738,
      "num_tokens": 32853394.0,
      "step": 980
    },
    {
      "epoch": 0.05879894937917861,
      "grad_norm": 1.0661559104919434,
      "learning_rate": 2.935560859188544e-05,
      "loss": 0.6347,
      "mean_token_accuracy": 0.8061911106109619,
      "num_tokens": 33021074.0,
      "step": 985
    },
    {
      "epoch": 0.059097421203438395,
      "grad_norm": 1.0102673768997192,
      "learning_rate": 2.950477326968974e-05,
      "loss": 0.6031,
      "mean_token_accuracy": 0.8141417145729065,
      "num_tokens": 33188754.0,
      "step": 990
    },
    {
      "epoch": 0.05939589302769818,
      "grad_norm": 1.2798126935958862,
      "learning_rate": 2.965393794749404e-05,
      "loss": 0.6373,
      "mean_token_accuracy": 0.8037635564804078,
      "num_tokens": 33356434.0,
      "step": 995
    },
    {
      "epoch": 0.059694364851957976,
      "grad_norm": 0.8796353936195374,
      "learning_rate": 2.980310262529833e-05,
      "loss": 0.6399,
      "mean_token_accuracy": 0.8031969428062439,
      "num_tokens": 33524114.0,
      "step": 1000
    },
    {
      "epoch": 0.05999283667621776,
      "grad_norm": 1.0169320106506348,
      "learning_rate": 2.9952267303102627e-05,
      "loss": 0.6342,
      "mean_token_accuracy": 0.8077120304107666,
      "num_tokens": 33691794.0,
      "step": 1005
    },
    {
      "epoch": 0.06029130850047756,
      "grad_norm": 0.855898380279541,
      "learning_rate": 3.0101431980906926e-05,
      "loss": 0.5901,
      "mean_token_accuracy": 0.81674222946167,
      "num_tokens": 33859474.0,
      "step": 1010
    },
    {
      "epoch": 0.060589780324737344,
      "grad_norm": 0.8868576884269714,
      "learning_rate": 3.0250596658711218e-05,
      "loss": 0.6453,
      "mean_token_accuracy": 0.8030895829200745,
      "num_tokens": 34027154.0,
      "step": 1015
    },
    {
      "epoch": 0.06088825214899714,
      "grad_norm": 0.9104098677635193,
      "learning_rate": 3.0399761336515514e-05,
      "loss": 0.6973,
      "mean_token_accuracy": 0.7892520546913147,
      "num_tokens": 34194834.0,
      "step": 1020
    },
    {
      "epoch": 0.061186723973256925,
      "grad_norm": 0.9279430508613586,
      "learning_rate": 3.054892601431981e-05,
      "loss": 0.6034,
      "mean_token_accuracy": 0.8136287808418274,
      "num_tokens": 34362514.0,
      "step": 1025
    },
    {
      "epoch": 0.06148519579751671,
      "grad_norm": 0.972676694393158,
      "learning_rate": 3.06980906921241e-05,
      "loss": 0.6801,
      "mean_token_accuracy": 0.7960455656051636,
      "num_tokens": 34530194.0,
      "step": 1030
    },
    {
      "epoch": 0.061783667621776506,
      "grad_norm": 1.1245499849319458,
      "learning_rate": 3.08472553699284e-05,
      "loss": 0.6116,
      "mean_token_accuracy": 0.8137480616569519,
      "num_tokens": 34697874.0,
      "step": 1035
    },
    {
      "epoch": 0.06208213944603629,
      "grad_norm": 0.9801743030548096,
      "learning_rate": 3.09964200477327e-05,
      "loss": 0.615,
      "mean_token_accuracy": 0.8098950266838074,
      "num_tokens": 34865554.0,
      "step": 1040
    },
    {
      "epoch": 0.06238061127029609,
      "grad_norm": 0.9435410499572754,
      "learning_rate": 3.1145584725537e-05,
      "loss": 0.6644,
      "mean_token_accuracy": 0.7998210787773132,
      "num_tokens": 35033234.0,
      "step": 1045
    },
    {
      "epoch": 0.06267908309455587,
      "grad_norm": 0.9696136713027954,
      "learning_rate": 3.129474940334129e-05,
      "loss": 0.6557,
      "mean_token_accuracy": 0.8016775250434875,
      "num_tokens": 35193806.0,
      "step": 1050
    },
    {
      "epoch": 0.06297755491881567,
      "grad_norm": 0.8354589939117432,
      "learning_rate": 3.1443914081145584e-05,
      "loss": 0.6388,
      "mean_token_accuracy": 0.8039425134658813,
      "num_tokens": 35361486.0,
      "step": 1055
    },
    {
      "epoch": 0.06327602674307545,
      "grad_norm": 0.959191620349884,
      "learning_rate": 3.159307875894988e-05,
      "loss": 0.6353,
      "mean_token_accuracy": 0.8077001094818115,
      "num_tokens": 35529166.0,
      "step": 1060
    },
    {
      "epoch": 0.06357449856733524,
      "grad_norm": 0.9490755200386047,
      "learning_rate": 3.1742243436754176e-05,
      "loss": 0.6416,
      "mean_token_accuracy": 0.8047655940055847,
      "num_tokens": 35696846.0,
      "step": 1065
    },
    {
      "epoch": 0.06387297039159504,
      "grad_norm": 0.9717972278594971,
      "learning_rate": 3.1891408114558475e-05,
      "loss": 0.6697,
      "mean_token_accuracy": 0.7956936836242676,
      "num_tokens": 35864526.0,
      "step": 1070
    },
    {
      "epoch": 0.06417144221585483,
      "grad_norm": 0.8323968648910522,
      "learning_rate": 3.2040572792362774e-05,
      "loss": 0.6138,
      "mean_token_accuracy": 0.8139389395713806,
      "num_tokens": 36032206.0,
      "step": 1075
    },
    {
      "epoch": 0.06446991404011461,
      "grad_norm": 0.9752675294876099,
      "learning_rate": 3.2189737470167066e-05,
      "loss": 0.673,
      "mean_token_accuracy": 0.7970953106880188,
      "num_tokens": 36199886.0,
      "step": 1080
    },
    {
      "epoch": 0.0647683858643744,
      "grad_norm": 1.1442911624908447,
      "learning_rate": 3.233890214797136e-05,
      "loss": 0.7082,
      "mean_token_accuracy": 0.7877609491348266,
      "num_tokens": 36367566.0,
      "step": 1085
    },
    {
      "epoch": 0.0650668576886342,
      "grad_norm": 1.037018060684204,
      "learning_rate": 3.248806682577566e-05,
      "loss": 0.6543,
      "mean_token_accuracy": 0.8013300776481629,
      "num_tokens": 36535246.0,
      "step": 1090
    },
    {
      "epoch": 0.06536532951289398,
      "grad_norm": 1.0444648265838623,
      "learning_rate": 3.263723150357995e-05,
      "loss": 0.6876,
      "mean_token_accuracy": 0.7919957041740417,
      "num_tokens": 36702926.0,
      "step": 1095
    },
    {
      "epoch": 0.06566380133715377,
      "grad_norm": 1.0123920440673828,
      "learning_rate": 3.278639618138425e-05,
      "loss": 0.6297,
      "mean_token_accuracy": 0.8090182542800903,
      "num_tokens": 36870606.0,
      "step": 1100
    },
    {
      "epoch": 0.06596227316141356,
      "grad_norm": 0.96045982837677,
      "learning_rate": 3.293556085918855e-05,
      "loss": 0.6734,
      "mean_token_accuracy": 0.7954491376876831,
      "num_tokens": 37038286.0,
      "step": 1105
    },
    {
      "epoch": 0.06626074498567336,
      "grad_norm": 0.9951871633529663,
      "learning_rate": 3.308472553699285e-05,
      "loss": 0.6348,
      "mean_token_accuracy": 0.8046761274337768,
      "num_tokens": 37205966.0,
      "step": 1110
    },
    {
      "epoch": 0.06655921680993314,
      "grad_norm": 1.0533862113952637,
      "learning_rate": 3.323389021479713e-05,
      "loss": 0.6228,
      "mean_token_accuracy": 0.810300612449646,
      "num_tokens": 37373646.0,
      "step": 1115
    },
    {
      "epoch": 0.06685768863419293,
      "grad_norm": 0.9938288927078247,
      "learning_rate": 3.338305489260143e-05,
      "loss": 0.6618,
      "mean_token_accuracy": 0.7987295746803283,
      "num_tokens": 37541326.0,
      "step": 1120
    },
    {
      "epoch": 0.06715616045845273,
      "grad_norm": 0.910653293132782,
      "learning_rate": 3.353221957040573e-05,
      "loss": 0.6518,
      "mean_token_accuracy": 0.802618408203125,
      "num_tokens": 37709006.0,
      "step": 1125
    },
    {
      "epoch": 0.0674546322827125,
      "grad_norm": 0.8506009578704834,
      "learning_rate": 3.3681384248210024e-05,
      "loss": 0.636,
      "mean_token_accuracy": 0.8066980719566346,
      "num_tokens": 37876686.0,
      "step": 1130
    },
    {
      "epoch": 0.0677531041069723,
      "grad_norm": 0.8333922028541565,
      "learning_rate": 3.383054892601432e-05,
      "loss": 0.6315,
      "mean_token_accuracy": 0.8075211763381958,
      "num_tokens": 38044366.0,
      "step": 1135
    },
    {
      "epoch": 0.0680515759312321,
      "grad_norm": 0.8634788393974304,
      "learning_rate": 3.397971360381862e-05,
      "loss": 0.6177,
      "mean_token_accuracy": 0.813157594203949,
      "num_tokens": 38212046.0,
      "step": 1140
    },
    {
      "epoch": 0.06835004775549189,
      "grad_norm": 1.0894731283187866,
      "learning_rate": 3.4128878281622915e-05,
      "loss": 0.6542,
      "mean_token_accuracy": 0.8000238537788391,
      "num_tokens": 38379726.0,
      "step": 1145
    },
    {
      "epoch": 0.06864851957975167,
      "grad_norm": 0.877989649772644,
      "learning_rate": 3.427804295942721e-05,
      "loss": 0.6432,
      "mean_token_accuracy": 0.8051831126213074,
      "num_tokens": 38547406.0,
      "step": 1150
    },
    {
      "epoch": 0.06894699140401146,
      "grad_norm": 1.0099589824676514,
      "learning_rate": 3.4427207637231506e-05,
      "loss": 0.6757,
      "mean_token_accuracy": 0.7953238725662232,
      "num_tokens": 38715086.0,
      "step": 1155
    },
    {
      "epoch": 0.06924546322827126,
      "grad_norm": 0.991763710975647,
      "learning_rate": 3.45763723150358e-05,
      "loss": 0.6467,
      "mean_token_accuracy": 0.8043301939964295,
      "num_tokens": 38882766.0,
      "step": 1160
    },
    {
      "epoch": 0.06954393505253104,
      "grad_norm": 0.9633587598800659,
      "learning_rate": 3.47255369928401e-05,
      "loss": 0.6407,
      "mean_token_accuracy": 0.8049006700515747,
      "num_tokens": 39050412.0,
      "step": 1165
    },
    {
      "epoch": 0.06984240687679083,
      "grad_norm": 0.8255019187927246,
      "learning_rate": 3.48747016706444e-05,
      "loss": 0.6417,
      "mean_token_accuracy": 0.8049206852912902,
      "num_tokens": 39218092.0,
      "step": 1170
    },
    {
      "epoch": 0.07014087870105062,
      "grad_norm": 0.8424438238143921,
      "learning_rate": 3.502386634844869e-05,
      "loss": 0.6699,
      "mean_token_accuracy": 0.7992604017257691,
      "num_tokens": 39385772.0,
      "step": 1175
    },
    {
      "epoch": 0.0704393505253104,
      "grad_norm": 0.9406603574752808,
      "learning_rate": 3.517303102625298e-05,
      "loss": 0.6198,
      "mean_token_accuracy": 0.8089884281158447,
      "num_tokens": 39553452.0,
      "step": 1180
    },
    {
      "epoch": 0.0707378223495702,
      "grad_norm": 0.9525007605552673,
      "learning_rate": 3.532219570405728e-05,
      "loss": 0.6857,
      "mean_token_accuracy": 0.7931349277496338,
      "num_tokens": 39721132.0,
      "step": 1185
    },
    {
      "epoch": 0.07103629417382999,
      "grad_norm": 0.9634766578674316,
      "learning_rate": 3.547136038186158e-05,
      "loss": 0.6574,
      "mean_token_accuracy": 0.7991649746894837,
      "num_tokens": 39888812.0,
      "step": 1190
    },
    {
      "epoch": 0.07133476599808979,
      "grad_norm": 1.0032249689102173,
      "learning_rate": 3.562052505966587e-05,
      "loss": 0.7442,
      "mean_token_accuracy": 0.7792198657989502,
      "num_tokens": 40056492.0,
      "step": 1195
    },
    {
      "epoch": 0.07163323782234957,
      "grad_norm": 0.8983254432678223,
      "learning_rate": 3.576968973747017e-05,
      "loss": 0.6583,
      "mean_token_accuracy": 0.8014374256134034,
      "num_tokens": 40224172.0,
      "step": 1200
    },
    {
      "epoch": 0.07193170964660936,
      "grad_norm": 0.9061633348464966,
      "learning_rate": 3.5918854415274464e-05,
      "loss": 0.6116,
      "mean_token_accuracy": 0.8129309177398681,
      "num_tokens": 40391852.0,
      "step": 1205
    },
    {
      "epoch": 0.07223018147086915,
      "grad_norm": 0.8976230621337891,
      "learning_rate": 3.6068019093078756e-05,
      "loss": 0.6341,
      "mean_token_accuracy": 0.8084218025207519,
      "num_tokens": 40559532.0,
      "step": 1210
    },
    {
      "epoch": 0.07252865329512893,
      "grad_norm": 0.8993825316429138,
      "learning_rate": 3.6217183770883055e-05,
      "loss": 0.6783,
      "mean_token_accuracy": 0.7957413792610168,
      "num_tokens": 40727212.0,
      "step": 1215
    },
    {
      "epoch": 0.07282712511938873,
      "grad_norm": 0.9418155550956726,
      "learning_rate": 3.6366348448687354e-05,
      "loss": 0.5908,
      "mean_token_accuracy": 0.8176905632019043,
      "num_tokens": 40894892.0,
      "step": 1220
    },
    {
      "epoch": 0.07312559694364852,
      "grad_norm": 0.9506223201751709,
      "learning_rate": 3.651551312649165e-05,
      "loss": 0.6605,
      "mean_token_accuracy": 0.7986281871795654,
      "num_tokens": 41062572.0,
      "step": 1225
    },
    {
      "epoch": 0.07342406876790832,
      "grad_norm": 0.9626055955886841,
      "learning_rate": 3.6664677804295946e-05,
      "loss": 0.6432,
      "mean_token_accuracy": 0.8047894597053528,
      "num_tokens": 41230252.0,
      "step": 1230
    },
    {
      "epoch": 0.0737225405921681,
      "grad_norm": 0.9560362696647644,
      "learning_rate": 3.681384248210024e-05,
      "loss": 0.6858,
      "mean_token_accuracy": 0.7895235776901245,
      "num_tokens": 41390132.0,
      "step": 1235
    },
    {
      "epoch": 0.07402101241642789,
      "grad_norm": 0.8743177652359009,
      "learning_rate": 3.696300715990454e-05,
      "loss": 0.6634,
      "mean_token_accuracy": 0.7970774173736572,
      "num_tokens": 41557812.0,
      "step": 1240
    },
    {
      "epoch": 0.07431948424068768,
      "grad_norm": 0.831082284450531,
      "learning_rate": 3.711217183770883e-05,
      "loss": 0.6038,
      "mean_token_accuracy": 0.8142013549804688,
      "num_tokens": 41725492.0,
      "step": 1245
    },
    {
      "epoch": 0.07461795606494746,
      "grad_norm": 0.9832303524017334,
      "learning_rate": 3.726133651551313e-05,
      "loss": 0.6486,
      "mean_token_accuracy": 0.802982223033905,
      "num_tokens": 41893172.0,
      "step": 1250
    },
    {
      "epoch": 0.07491642788920726,
      "grad_norm": 0.9175562262535095,
      "learning_rate": 3.741050119331743e-05,
      "loss": 0.6443,
      "mean_token_accuracy": 0.8030120491981506,
      "num_tokens": 42060852.0,
      "step": 1255
    },
    {
      "epoch": 0.07521489971346705,
      "grad_norm": 0.9284515976905823,
      "learning_rate": 3.755966587112172e-05,
      "loss": 0.6633,
      "mean_token_accuracy": 0.7998687863349915,
      "num_tokens": 42228532.0,
      "step": 1260
    },
    {
      "epoch": 0.07551337153772684,
      "grad_norm": 1.1153919696807861,
      "learning_rate": 3.770883054892602e-05,
      "loss": 0.6454,
      "mean_token_accuracy": 0.8018296599388123,
      "num_tokens": 42389862.0,
      "step": 1265
    },
    {
      "epoch": 0.07581184336198662,
      "grad_norm": 0.9243282079696655,
      "learning_rate": 3.785799522673031e-05,
      "loss": 0.675,
      "mean_token_accuracy": 0.7935524225234986,
      "num_tokens": 42557542.0,
      "step": 1270
    },
    {
      "epoch": 0.07611031518624642,
      "grad_norm": 0.8942394256591797,
      "learning_rate": 3.8007159904534604e-05,
      "loss": 0.6568,
      "mean_token_accuracy": 0.800501024723053,
      "num_tokens": 42725222.0,
      "step": 1275
    },
    {
      "epoch": 0.07640878701050621,
      "grad_norm": 0.8383049964904785,
      "learning_rate": 3.8156324582338903e-05,
      "loss": 0.6224,
      "mean_token_accuracy": 0.8080102682113648,
      "num_tokens": 42892902.0,
      "step": 1280
    },
    {
      "epoch": 0.07670725883476599,
      "grad_norm": 0.926938533782959,
      "learning_rate": 3.83054892601432e-05,
      "loss": 0.6166,
      "mean_token_accuracy": 0.8101336002349854,
      "num_tokens": 43060582.0,
      "step": 1285
    },
    {
      "epoch": 0.07700573065902579,
      "grad_norm": 0.8695472478866577,
      "learning_rate": 3.8454653937947495e-05,
      "loss": 0.6384,
      "mean_token_accuracy": 0.8027496099472046,
      "num_tokens": 43228262.0,
      "step": 1290
    },
    {
      "epoch": 0.07730420248328558,
      "grad_norm": 0.8851755261421204,
      "learning_rate": 3.8603818615751794e-05,
      "loss": 0.6554,
      "mean_token_accuracy": 0.8046105146408081,
      "num_tokens": 43395942.0,
      "step": 1295
    },
    {
      "epoch": 0.07760267430754537,
      "grad_norm": 0.9108031988143921,
      "learning_rate": 3.8752983293556087e-05,
      "loss": 0.6103,
      "mean_token_accuracy": 0.8136168360710144,
      "num_tokens": 43563622.0,
      "step": 1300
    },
    {
      "epoch": 0.07790114613180515,
      "grad_norm": 0.9888355135917664,
      "learning_rate": 3.8902147971360386e-05,
      "loss": 0.6626,
      "mean_token_accuracy": 0.7978468298912048,
      "num_tokens": 43731302.0,
      "step": 1305
    },
    {
      "epoch": 0.07819961795606495,
      "grad_norm": 0.9139824509620667,
      "learning_rate": 3.905131264916468e-05,
      "loss": 0.6671,
      "mean_token_accuracy": 0.7976440548896789,
      "num_tokens": 43898982.0,
      "step": 1310
    },
    {
      "epoch": 0.07849808978032474,
      "grad_norm": 0.8039135336875916,
      "learning_rate": 3.920047732696898e-05,
      "loss": 0.616,
      "mean_token_accuracy": 0.8116008639335632,
      "num_tokens": 44066662.0,
      "step": 1315
    },
    {
      "epoch": 0.07879656160458452,
      "grad_norm": 0.8862246870994568,
      "learning_rate": 3.9349642004773276e-05,
      "loss": 0.645,
      "mean_token_accuracy": 0.8045687794685363,
      "num_tokens": 44234342.0,
      "step": 1320
    },
    {
      "epoch": 0.07909503342884432,
      "grad_norm": 0.8758423328399658,
      "learning_rate": 3.949880668257757e-05,
      "loss": 0.6708,
      "mean_token_accuracy": 0.7965107917785644,
      "num_tokens": 44402022.0,
      "step": 1325
    },
    {
      "epoch": 0.07939350525310411,
      "grad_norm": 0.8791078925132751,
      "learning_rate": 3.964797136038186e-05,
      "loss": 0.6026,
      "mean_token_accuracy": 0.8155731916427612,
      "num_tokens": 44569702.0,
      "step": 1330
    },
    {
      "epoch": 0.07969197707736389,
      "grad_norm": 0.8547496199607849,
      "learning_rate": 3.979713603818616e-05,
      "loss": 0.6969,
      "mean_token_accuracy": 0.7891864538192749,
      "num_tokens": 44737382.0,
      "step": 1335
    },
    {
      "epoch": 0.07999044890162368,
      "grad_norm": 1.9958553314208984,
      "learning_rate": 3.994630071599045e-05,
      "loss": 0.6901,
      "mean_token_accuracy": 0.7919897437095642,
      "num_tokens": 44905062.0,
      "step": 1340
    },
    {
      "epoch": 0.08028892072588348,
      "grad_norm": 0.9849021434783936,
      "learning_rate": 4.009546539379475e-05,
      "loss": 0.6632,
      "mean_token_accuracy": 0.798646068572998,
      "num_tokens": 45072742.0,
      "step": 1345
    },
    {
      "epoch": 0.08058739255014327,
      "grad_norm": 0.9506638646125793,
      "learning_rate": 4.024463007159905e-05,
      "loss": 0.6586,
      "mean_token_accuracy": 0.8004473328590394,
      "num_tokens": 45240422.0,
      "step": 1350
    },
    {
      "epoch": 0.08088586437440305,
      "grad_norm": 0.8287100791931152,
      "learning_rate": 4.039379474940334e-05,
      "loss": 0.6131,
      "mean_token_accuracy": 0.812656569480896,
      "num_tokens": 45408102.0,
      "step": 1355
    },
    {
      "epoch": 0.08118433619866285,
      "grad_norm": 0.850650429725647,
      "learning_rate": 4.0542959427207636e-05,
      "loss": 0.6021,
      "mean_token_accuracy": 0.8153942584991455,
      "num_tokens": 45575782.0,
      "step": 1360
    },
    {
      "epoch": 0.08148280802292264,
      "grad_norm": 0.8990254402160645,
      "learning_rate": 4.0692124105011935e-05,
      "loss": 0.6228,
      "mean_token_accuracy": 0.8096624255180359,
      "num_tokens": 45743462.0,
      "step": 1365
    },
    {
      "epoch": 0.08178127984718242,
      "grad_norm": 0.9089404940605164,
      "learning_rate": 4.0841288782816234e-05,
      "loss": 0.6486,
      "mean_token_accuracy": 0.802856981754303,
      "num_tokens": 45911142.0,
      "step": 1370
    },
    {
      "epoch": 0.08207975167144221,
      "grad_norm": 0.8626326322555542,
      "learning_rate": 4.0990453460620526e-05,
      "loss": 0.6474,
      "mean_token_accuracy": 0.8042645812034607,
      "num_tokens": 46078822.0,
      "step": 1375
    },
    {
      "epoch": 0.08237822349570201,
      "grad_norm": 0.969448983669281,
      "learning_rate": 4.1139618138424825e-05,
      "loss": 0.6653,
      "mean_token_accuracy": 0.7979422569274902,
      "num_tokens": 46246502.0,
      "step": 1380
    },
    {
      "epoch": 0.0826766953199618,
      "grad_norm": 0.9010327458381653,
      "learning_rate": 4.1288782816229125e-05,
      "loss": 0.66,
      "mean_token_accuracy": 0.7994512677192688,
      "num_tokens": 46414182.0,
      "step": 1385
    },
    {
      "epoch": 0.08297516714422158,
      "grad_norm": 0.9436922073364258,
      "learning_rate": 4.143794749403341e-05,
      "loss": 0.6776,
      "mean_token_accuracy": 0.7937075138092041,
      "num_tokens": 46581862.0,
      "step": 1390
    },
    {
      "epoch": 0.08327363896848138,
      "grad_norm": 0.7983666658401489,
      "learning_rate": 4.158711217183771e-05,
      "loss": 0.6541,
      "mean_token_accuracy": 0.8005606651306152,
      "num_tokens": 46749542.0,
      "step": 1395
    },
    {
      "epoch": 0.08357211079274117,
      "grad_norm": 0.8562968373298645,
      "learning_rate": 4.173627684964201e-05,
      "loss": 0.6757,
      "mean_token_accuracy": 0.7943874716758728,
      "num_tokens": 46917222.0,
      "step": 1400
    },
    {
      "epoch": 0.08387058261700095,
      "grad_norm": 0.9290655255317688,
      "learning_rate": 4.18854415274463e-05,
      "loss": 0.6422,
      "mean_token_accuracy": 0.8043600082397461,
      "num_tokens": 47084902.0,
      "step": 1405
    },
    {
      "epoch": 0.08416905444126074,
      "grad_norm": 0.8598992228507996,
      "learning_rate": 4.20346062052506e-05,
      "loss": 0.5864,
      "mean_token_accuracy": 0.8187403321266175,
      "num_tokens": 47252582.0,
      "step": 1410
    },
    {
      "epoch": 0.08446752626552054,
      "grad_norm": 0.8854230046272278,
      "learning_rate": 4.21837708830549e-05,
      "loss": 0.6655,
      "mean_token_accuracy": 0.7973040819168091,
      "num_tokens": 47420262.0,
      "step": 1415
    },
    {
      "epoch": 0.08476599808978033,
      "grad_norm": 0.8565343618392944,
      "learning_rate": 4.233293556085919e-05,
      "loss": 0.669,
      "mean_token_accuracy": 0.797894561290741,
      "num_tokens": 47587942.0,
      "step": 1420
    },
    {
      "epoch": 0.08506446991404011,
      "grad_norm": 0.8702725768089294,
      "learning_rate": 4.2482100238663484e-05,
      "loss": 0.633,
      "mean_token_accuracy": 0.8055409669876099,
      "num_tokens": 47755622.0,
      "step": 1425
    },
    {
      "epoch": 0.0853629417382999,
      "grad_norm": 0.8215453028678894,
      "learning_rate": 4.263126491646778e-05,
      "loss": 0.6068,
      "mean_token_accuracy": 0.8145174741744995,
      "num_tokens": 47923302.0,
      "step": 1430
    },
    {
      "epoch": 0.0856614135625597,
      "grad_norm": 0.8121721148490906,
      "learning_rate": 4.278042959427208e-05,
      "loss": 0.6683,
      "mean_token_accuracy": 0.7968984842300415,
      "num_tokens": 48090982.0,
      "step": 1435
    },
    {
      "epoch": 0.08595988538681948,
      "grad_norm": 0.8051170706748962,
      "learning_rate": 4.2929594272076375e-05,
      "loss": 0.654,
      "mean_token_accuracy": 0.8024573564529419,
      "num_tokens": 48258662.0,
      "step": 1440
    },
    {
      "epoch": 0.08625835721107927,
      "grad_norm": 0.7850134372711182,
      "learning_rate": 4.3078758949880674e-05,
      "loss": 0.6628,
      "mean_token_accuracy": 0.8003817200660706,
      "num_tokens": 48426342.0,
      "step": 1445
    },
    {
      "epoch": 0.08655682903533907,
      "grad_norm": 0.8776177763938904,
      "learning_rate": 4.3227923627684966e-05,
      "loss": 0.6251,
      "mean_token_accuracy": 0.808494484424591,
      "num_tokens": 48588230.0,
      "step": 1450
    },
    {
      "epoch": 0.08685530085959886,
      "grad_norm": 0.781536340713501,
      "learning_rate": 4.337708830548926e-05,
      "loss": 0.6573,
      "mean_token_accuracy": 0.7996501803398133,
      "num_tokens": 48752708.0,
      "step": 1455
    },
    {
      "epoch": 0.08715377268385864,
      "grad_norm": 0.8784968852996826,
      "learning_rate": 4.352625298329356e-05,
      "loss": 0.7195,
      "mean_token_accuracy": 0.7821126103401184,
      "num_tokens": 48920388.0,
      "step": 1460
    },
    {
      "epoch": 0.08745224450811843,
      "grad_norm": 0.8595447540283203,
      "learning_rate": 4.367541766109786e-05,
      "loss": 0.6471,
      "mean_token_accuracy": 0.8021114230155945,
      "num_tokens": 49088068.0,
      "step": 1465
    },
    {
      "epoch": 0.08775071633237823,
      "grad_norm": 0.8538692593574524,
      "learning_rate": 4.382458233890215e-05,
      "loss": 0.6269,
      "mean_token_accuracy": 0.8087796688079834,
      "num_tokens": 49255748.0,
      "step": 1470
    },
    {
      "epoch": 0.08804918815663801,
      "grad_norm": 0.8793528079986572,
      "learning_rate": 4.397374701670645e-05,
      "loss": 0.6285,
      "mean_token_accuracy": 0.8089884281158447,
      "num_tokens": 49423428.0,
      "step": 1475
    },
    {
      "epoch": 0.0883476599808978,
      "grad_norm": 0.8972448706626892,
      "learning_rate": 4.412291169451074e-05,
      "loss": 0.6615,
      "mean_token_accuracy": 0.8002743601799012,
      "num_tokens": 49591108.0,
      "step": 1480
    },
    {
      "epoch": 0.0886461318051576,
      "grad_norm": 0.8688410520553589,
      "learning_rate": 4.427207637231503e-05,
      "loss": 0.6493,
      "mean_token_accuracy": 0.8021889448165893,
      "num_tokens": 49758788.0,
      "step": 1485
    },
    {
      "epoch": 0.08894460362941738,
      "grad_norm": 0.8287907242774963,
      "learning_rate": 4.442124105011933e-05,
      "loss": 0.6451,
      "mean_token_accuracy": 0.8048848867416382,
      "num_tokens": 49926468.0,
      "step": 1490
    },
    {
      "epoch": 0.08924307545367717,
      "grad_norm": 0.8825296759605408,
      "learning_rate": 4.457040572792363e-05,
      "loss": 0.6704,
      "mean_token_accuracy": 0.7964332580566407,
      "num_tokens": 50094148.0,
      "step": 1495
    },
    {
      "epoch": 0.08954154727793696,
      "grad_norm": 0.8199141621589661,
      "learning_rate": 4.4719570405727924e-05,
      "loss": 0.6339,
      "mean_token_accuracy": 0.8072169899940491,
      "num_tokens": 50261828.0,
      "step": 1500
    },
    {
      "epoch": 0.08984001910219676,
      "grad_norm": 0.949461817741394,
      "learning_rate": 4.486873508353222e-05,
      "loss": 0.6471,
      "mean_token_accuracy": 0.802469277381897,
      "num_tokens": 50429508.0,
      "step": 1505
    },
    {
      "epoch": 0.09013849092645654,
      "grad_norm": 0.8446555137634277,
      "learning_rate": 4.5017899761336515e-05,
      "loss": 0.6793,
      "mean_token_accuracy": 0.792818808555603,
      "num_tokens": 50597188.0,
      "step": 1510
    },
    {
      "epoch": 0.09043696275071633,
      "grad_norm": 0.8719688653945923,
      "learning_rate": 4.5167064439140814e-05,
      "loss": 0.6364,
      "mean_token_accuracy": 0.8059286594390869,
      "num_tokens": 50764868.0,
      "step": 1515
    },
    {
      "epoch": 0.09073543457497613,
      "grad_norm": 0.8748178482055664,
      "learning_rate": 4.531622911694511e-05,
      "loss": 0.6535,
      "mean_token_accuracy": 0.8021829843521118,
      "num_tokens": 50932548.0,
      "step": 1520
    },
    {
      "epoch": 0.0910339063992359,
      "grad_norm": 0.7743371725082397,
      "learning_rate": 4.5465393794749406e-05,
      "loss": 0.6319,
      "mean_token_accuracy": 0.8074018836021424,
      "num_tokens": 51100228.0,
      "step": 1525
    },
    {
      "epoch": 0.0913323782234957,
      "grad_norm": 0.7892270684242249,
      "learning_rate": 4.5614558472553705e-05,
      "loss": 0.6156,
      "mean_token_accuracy": 0.8116068243980408,
      "num_tokens": 51267908.0,
      "step": 1530
    },
    {
      "epoch": 0.0916308500477555,
      "grad_norm": 0.8279834389686584,
      "learning_rate": 4.5763723150358e-05,
      "loss": 0.6131,
      "mean_token_accuracy": 0.8119527697563171,
      "num_tokens": 51435588.0,
      "step": 1535
    },
    {
      "epoch": 0.09192932187201529,
      "grad_norm": 0.8303422331809998,
      "learning_rate": 4.5912887828162297e-05,
      "loss": 0.6335,
      "mean_token_accuracy": 0.8053501129150391,
      "num_tokens": 51603268.0,
      "step": 1540
    },
    {
      "epoch": 0.09222779369627507,
      "grad_norm": 0.8752642869949341,
      "learning_rate": 4.606205250596659e-05,
      "loss": 0.6315,
      "mean_token_accuracy": 0.8056185126304627,
      "num_tokens": 51770948.0,
      "step": 1545
    },
    {
      "epoch": 0.09252626552053486,
      "grad_norm": 0.8765949010848999,
      "learning_rate": 4.621121718377088e-05,
      "loss": 0.6409,
      "mean_token_accuracy": 0.8050578594207763,
      "num_tokens": 51938628.0,
      "step": 1550
    },
    {
      "epoch": 0.09282473734479466,
      "grad_norm": 0.8335886001586914,
      "learning_rate": 4.636038186157518e-05,
      "loss": 0.604,
      "mean_token_accuracy": 0.8138911962509155,
      "num_tokens": 52106308.0,
      "step": 1555
    },
    {
      "epoch": 0.09312320916905444,
      "grad_norm": 0.9149099588394165,
      "learning_rate": 4.650954653937948e-05,
      "loss": 0.6456,
      "mean_token_accuracy": 0.8014195322990417,
      "num_tokens": 52273988.0,
      "step": 1560
    },
    {
      "epoch": 0.09342168099331423,
      "grad_norm": 0.7945675253868103,
      "learning_rate": 4.665871121718377e-05,
      "loss": 0.6676,
      "mean_token_accuracy": 0.7962125658988952,
      "num_tokens": 52441668.0,
      "step": 1565
    },
    {
      "epoch": 0.09372015281757402,
      "grad_norm": 0.7990531325340271,
      "learning_rate": 4.680787589498807e-05,
      "loss": 0.6695,
      "mean_token_accuracy": 0.7956877112388611,
      "num_tokens": 52609348.0,
      "step": 1570
    },
    {
      "epoch": 0.09401862464183382,
      "grad_norm": 0.9765162467956543,
      "learning_rate": 4.6957040572792363e-05,
      "loss": 0.7025,
      "mean_token_accuracy": 0.7895502805709839,
      "num_tokens": 52777028.0,
      "step": 1575
    },
    {
      "epoch": 0.0943170964660936,
      "grad_norm": 0.8342398405075073,
      "learning_rate": 4.710620525059666e-05,
      "loss": 0.636,
      "mean_token_accuracy": 0.8052308201789856,
      "num_tokens": 52944708.0,
      "step": 1580
    },
    {
      "epoch": 0.09461556829035339,
      "grad_norm": 0.9157741665840149,
      "learning_rate": 4.7255369928400955e-05,
      "loss": 0.6463,
      "mean_token_accuracy": 0.8030001282691955,
      "num_tokens": 53112388.0,
      "step": 1585
    },
    {
      "epoch": 0.09491404011461319,
      "grad_norm": 0.8565534353256226,
      "learning_rate": 4.7404534606205254e-05,
      "loss": 0.6883,
      "mean_token_accuracy": 0.790916132926941,
      "num_tokens": 53280068.0,
      "step": 1590
    },
    {
      "epoch": 0.09521251193887297,
      "grad_norm": 0.9609672427177429,
      "learning_rate": 4.755369928400955e-05,
      "loss": 0.686,
      "mean_token_accuracy": 0.7923416495323181,
      "num_tokens": 53447748.0,
      "step": 1595
    },
    {
      "epoch": 0.09551098376313276,
      "grad_norm": 0.8258156776428223,
      "learning_rate": 4.7702863961813846e-05,
      "loss": 0.6341,
      "mean_token_accuracy": 0.8055588722229003,
      "num_tokens": 53615428.0,
      "step": 1600
    },
    {
      "epoch": 0.09580945558739255,
      "grad_norm": 0.9373846650123596,
      "learning_rate": 4.785202863961814e-05,
      "loss": 0.6391,
      "mean_token_accuracy": 0.8042407274246216,
      "num_tokens": 53783108.0,
      "step": 1605
    },
    {
      "epoch": 0.09610792741165235,
      "grad_norm": 0.8284741044044495,
      "learning_rate": 4.800119331742244e-05,
      "loss": 0.6916,
      "mean_token_accuracy": 0.790802824497223,
      "num_tokens": 53950788.0,
      "step": 1610
    },
    {
      "epoch": 0.09640639923591213,
      "grad_norm": 0.8239485025405884,
      "learning_rate": 4.815035799522673e-05,
      "loss": 0.6504,
      "mean_token_accuracy": 0.8026899695396423,
      "num_tokens": 54118468.0,
      "step": 1615
    },
    {
      "epoch": 0.09670487106017192,
      "grad_norm": 0.8749241232872009,
      "learning_rate": 4.829952267303103e-05,
      "loss": 0.6443,
      "mean_token_accuracy": 0.8018966794013977,
      "num_tokens": 54286148.0,
      "step": 1620
    },
    {
      "epoch": 0.09700334288443171,
      "grad_norm": 0.7732098698616028,
      "learning_rate": 4.844868735083533e-05,
      "loss": 0.6376,
      "mean_token_accuracy": 0.8057556986808777,
      "num_tokens": 54453828.0,
      "step": 1625
    },
    {
      "epoch": 0.0973018147086915,
      "grad_norm": 0.8085273504257202,
      "learning_rate": 4.859785202863962e-05,
      "loss": 0.6205,
      "mean_token_accuracy": 0.8094891905784607,
      "num_tokens": 54613284.0,
      "step": 1630
    },
    {
      "epoch": 0.09760028653295129,
      "grad_norm": 0.8560582399368286,
      "learning_rate": 4.874701670644391e-05,
      "loss": 0.6885,
      "mean_token_accuracy": 0.7920732498168945,
      "num_tokens": 54780964.0,
      "step": 1635
    },
    {
      "epoch": 0.09789875835721108,
      "grad_norm": 0.7586135268211365,
      "learning_rate": 4.889618138424821e-05,
      "loss": 0.6188,
      "mean_token_accuracy": 0.8088154554367065,
      "num_tokens": 54948644.0,
      "step": 1640
    },
    {
      "epoch": 0.09819723018147086,
      "grad_norm": 0.9643146395683289,
      "learning_rate": 4.904534606205251e-05,
      "loss": 0.6594,
      "mean_token_accuracy": 0.8002862930297852,
      "num_tokens": 55116324.0,
      "step": 1645
    },
    {
      "epoch": 0.09849570200573066,
      "grad_norm": 0.8251019716262817,
      "learning_rate": 4.91945107398568e-05,
      "loss": 0.5937,
      "mean_token_accuracy": 0.8172074437141419,
      "num_tokens": 55284004.0,
      "step": 1650
    },
    {
      "epoch": 0.09879417382999045,
      "grad_norm": 0.8162339329719543,
      "learning_rate": 4.93436754176611e-05,
      "loss": 0.6679,
      "mean_token_accuracy": 0.7967016577720643,
      "num_tokens": 55451684.0,
      "step": 1655
    },
    {
      "epoch": 0.09909264565425024,
      "grad_norm": 0.7879742383956909,
      "learning_rate": 4.9492840095465395e-05,
      "loss": 0.6224,
      "mean_token_accuracy": 0.8104497194290161,
      "num_tokens": 55619364.0,
      "step": 1660
    },
    {
      "epoch": 0.09939111747851002,
      "grad_norm": 0.8442058563232422,
      "learning_rate": 4.964200477326969e-05,
      "loss": 0.6383,
      "mean_token_accuracy": 0.8046403408050538,
      "num_tokens": 55787044.0,
      "step": 1665
    },
    {
      "epoch": 0.09968958930276982,
      "grad_norm": 0.8841798305511475,
      "learning_rate": 4.9791169451073986e-05,
      "loss": 0.6689,
      "mean_token_accuracy": 0.798007869720459,
      "num_tokens": 55954724.0,
      "step": 1670
    },
    {
      "epoch": 0.09998806112702961,
      "grad_norm": 0.8683707118034363,
      "learning_rate": 4.9940334128878285e-05,
      "loss": 0.677,
      "mean_token_accuracy": 0.7960396051406861,
      "num_tokens": 56122404.0,
      "step": 1675
    },
    {
      "epoch": 0.10028653295128939,
      "grad_norm": 0.7513850927352905,
      "learning_rate": 4.999999901354773e-05,
      "loss": 0.6123,
      "mean_token_accuracy": 0.8120779991149902,
      "num_tokens": 56290084.0,
      "step": 1680
    },
    {
      "epoch": 0.10058500477554919,
      "grad_norm": 0.8978567719459534,
      "learning_rate": 4.9999992985228636e-05,
      "loss": 0.6895,
      "mean_token_accuracy": 0.7923416495323181,
      "num_tokens": 56457764.0,
      "step": 1685
    },
    {
      "epoch": 0.10088347659980898,
      "grad_norm": 0.820931613445282,
      "learning_rate": 4.9999981476620944e-05,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8212751984596253,
      "num_tokens": 56625444.0,
      "step": 1690
    },
    {
      "epoch": 0.10118194842406877,
      "grad_norm": 0.7980391383171082,
      "learning_rate": 4.9999964487727464e-05,
      "loss": 0.5925,
      "mean_token_accuracy": 0.8184659481048584,
      "num_tokens": 56793124.0,
      "step": 1695
    },
    {
      "epoch": 0.10148042024832855,
      "grad_norm": 0.7825395464897156,
      "learning_rate": 4.999994201855233e-05,
      "loss": 0.6576,
      "mean_token_accuracy": 0.8001013994216919,
      "num_tokens": 56960804.0,
      "step": 1700
    },
    {
      "epoch": 0.10177889207258835,
      "grad_norm": 0.8230543732643127,
      "learning_rate": 4.9999914069101016e-05,
      "loss": 0.623,
      "mean_token_accuracy": 0.8087916135787964,
      "num_tokens": 57128484.0,
      "step": 1705
    },
    {
      "epoch": 0.10207736389684814,
      "grad_norm": 0.7690154314041138,
      "learning_rate": 4.999988063938033e-05,
      "loss": 0.6198,
      "mean_token_accuracy": 0.8091315865516663,
      "num_tokens": 57296164.0,
      "step": 1710
    },
    {
      "epoch": 0.10237583572110792,
      "grad_norm": 0.8304437398910522,
      "learning_rate": 4.999984172939841e-05,
      "loss": 0.6461,
      "mean_token_accuracy": 0.8028092741966247,
      "num_tokens": 57463844.0,
      "step": 1715
    },
    {
      "epoch": 0.10267430754536772,
      "grad_norm": 0.8990681171417236,
      "learning_rate": 4.999979733916474e-05,
      "loss": 0.6983,
      "mean_token_accuracy": 0.785589873790741,
      "num_tokens": 57631524.0,
      "step": 1720
    },
    {
      "epoch": 0.10297277936962751,
      "grad_norm": 0.7804028391838074,
      "learning_rate": 4.999974746869013e-05,
      "loss": 0.6916,
      "mean_token_accuracy": 0.7889180541038513,
      "num_tokens": 57799204.0,
      "step": 1725
    },
    {
      "epoch": 0.1032712511938873,
      "grad_norm": 0.8725268840789795,
      "learning_rate": 4.999969211798672e-05,
      "loss": 0.646,
      "mean_token_accuracy": 0.8023380756378173,
      "num_tokens": 57966884.0,
      "step": 1730
    },
    {
      "epoch": 0.10356972301814708,
      "grad_norm": 0.8713625073432922,
      "learning_rate": 4.9999631287068004e-05,
      "loss": 0.7015,
      "mean_token_accuracy": 0.7879935503005981,
      "num_tokens": 58134564.0,
      "step": 1735
    },
    {
      "epoch": 0.10386819484240688,
      "grad_norm": 0.808779239654541,
      "learning_rate": 4.999956497594879e-05,
      "loss": 0.6943,
      "mean_token_accuracy": 0.7902600526809692,
      "num_tokens": 58302244.0,
      "step": 1740
    },
    {
      "epoch": 0.10416666666666667,
      "grad_norm": 0.7280786633491516,
      "learning_rate": 4.999949318464524e-05,
      "loss": 0.6228,
      "mean_token_accuracy": 0.8087319493293762,
      "num_tokens": 58469924.0,
      "step": 1745
    },
    {
      "epoch": 0.10446513849092645,
      "grad_norm": 0.8153372406959534,
      "learning_rate": 4.9999415913174825e-05,
      "loss": 0.6974,
      "mean_token_accuracy": 0.7893594145774842,
      "num_tokens": 58637604.0,
      "step": 1750
    },
    {
      "epoch": 0.10476361031518625,
      "grad_norm": 0.7588002681732178,
      "learning_rate": 4.9999333161556374e-05,
      "loss": 0.6347,
      "mean_token_accuracy": 0.8061724901199341,
      "num_tokens": 58804558.0,
      "step": 1755
    },
    {
      "epoch": 0.10506208213944604,
      "grad_norm": 0.7470396161079407,
      "learning_rate": 4.999924492981004e-05,
      "loss": 0.641,
      "mean_token_accuracy": 0.8059107780456543,
      "num_tokens": 58972238.0,
      "step": 1760
    },
    {
      "epoch": 0.10536055396370583,
      "grad_norm": 0.8653382062911987,
      "learning_rate": 4.999915121795732e-05,
      "loss": 0.6801,
      "mean_token_accuracy": 0.7929261684417724,
      "num_tokens": 59139918.0,
      "step": 1765
    },
    {
      "epoch": 0.10565902578796561,
      "grad_norm": 0.9816977977752686,
      "learning_rate": 4.9999052026021024e-05,
      "loss": 0.6644,
      "mean_token_accuracy": 0.7990134358406067,
      "num_tokens": 59305447.0,
      "step": 1770
    },
    {
      "epoch": 0.10595749761222541,
      "grad_norm": 0.8733177185058594,
      "learning_rate": 4.999894735402533e-05,
      "loss": 0.6167,
      "mean_token_accuracy": 0.8126625418663025,
      "num_tokens": 59473127.0,
      "step": 1775
    },
    {
      "epoch": 0.1062559694364852,
      "grad_norm": 0.7908738255500793,
      "learning_rate": 4.999883720199572e-05,
      "loss": 0.6168,
      "mean_token_accuracy": 0.8099069476127625,
      "num_tokens": 59640807.0,
      "step": 1780
    },
    {
      "epoch": 0.10655444126074498,
      "grad_norm": 0.8156537413597107,
      "learning_rate": 4.9998721569959035e-05,
      "loss": 0.6945,
      "mean_token_accuracy": 0.7903017997741699,
      "num_tokens": 59808487.0,
      "step": 1785
    },
    {
      "epoch": 0.10685291308500477,
      "grad_norm": 0.8189343214035034,
      "learning_rate": 4.999860045794343e-05,
      "loss": 0.6788,
      "mean_token_accuracy": 0.7944590210914612,
      "num_tokens": 59976167.0,
      "step": 1790
    },
    {
      "epoch": 0.10715138490926457,
      "grad_norm": 0.9364219307899475,
      "learning_rate": 4.999847386597841e-05,
      "loss": 0.6401,
      "mean_token_accuracy": 0.8055946588516235,
      "num_tokens": 60143847.0,
      "step": 1795
    },
    {
      "epoch": 0.10744985673352435,
      "grad_norm": 0.8389230966567993,
      "learning_rate": 4.9998341794094795e-05,
      "loss": 0.6314,
      "mean_token_accuracy": 0.8070320963859559,
      "num_tokens": 60311527.0,
      "step": 1800
    },
    {
      "epoch": 0.10774832855778414,
      "grad_norm": 0.8747956156730652,
      "learning_rate": 4.999820424232477e-05,
      "loss": 0.7154,
      "mean_token_accuracy": 0.7825838088989258,
      "num_tokens": 60479207.0,
      "step": 1805
    },
    {
      "epoch": 0.10804680038204394,
      "grad_norm": 4.572124004364014,
      "learning_rate": 4.9998061210701836e-05,
      "loss": 0.6549,
      "mean_token_accuracy": 0.8010378003120422,
      "num_tokens": 60646887.0,
      "step": 1810
    },
    {
      "epoch": 0.10834527220630373,
      "grad_norm": 0.7478983998298645,
      "learning_rate": 4.999791269926082e-05,
      "loss": 0.5921,
      "mean_token_accuracy": 0.8204623460769653,
      "num_tokens": 60810262.0,
      "step": 1815
    },
    {
      "epoch": 0.10864374403056351,
      "grad_norm": 1.0214107036590576,
      "learning_rate": 4.9997758708037906e-05,
      "loss": 0.671,
      "mean_token_accuracy": 0.7963855385780334,
      "num_tokens": 60977942.0,
      "step": 1820
    },
    {
      "epoch": 0.1089422158548233,
      "grad_norm": 0.9769614934921265,
      "learning_rate": 4.999759923707061e-05,
      "loss": 0.6606,
      "mean_token_accuracy": 0.7990695357322692,
      "num_tokens": 61145622.0,
      "step": 1825
    },
    {
      "epoch": 0.1092406876790831,
      "grad_norm": 0.7749030590057373,
      "learning_rate": 4.9997434286397746e-05,
      "loss": 0.6264,
      "mean_token_accuracy": 0.808421790599823,
      "num_tokens": 61313302.0,
      "step": 1830
    },
    {
      "epoch": 0.10953915950334288,
      "grad_norm": 1.0878808498382568,
      "learning_rate": 4.999726385605951e-05,
      "loss": 0.7611,
      "mean_token_accuracy": 0.7697423338890076,
      "num_tokens": 61480982.0,
      "step": 1835
    },
    {
      "epoch": 0.10983763132760267,
      "grad_norm": 0.7937517166137695,
      "learning_rate": 4.999708794609741e-05,
      "loss": 0.6456,
      "mean_token_accuracy": 0.8015149712562561,
      "num_tokens": 61648662.0,
      "step": 1840
    },
    {
      "epoch": 0.11013610315186247,
      "grad_norm": 0.8066228032112122,
      "learning_rate": 4.999690655655429e-05,
      "loss": 0.6776,
      "mean_token_accuracy": 0.792681610584259,
      "num_tokens": 61816342.0,
      "step": 1845
    },
    {
      "epoch": 0.11043457497612226,
      "grad_norm": 0.8846402764320374,
      "learning_rate": 4.999671968747434e-05,
      "loss": 0.6363,
      "mean_token_accuracy": 0.8043779015541077,
      "num_tokens": 61984022.0,
      "step": 1850
    },
    {
      "epoch": 0.11073304680038204,
      "grad_norm": 0.7564229369163513,
      "learning_rate": 4.999652733890306e-05,
      "loss": 0.6465,
      "mean_token_accuracy": 0.8019324660301208,
      "num_tokens": 62151702.0,
      "step": 1855
    },
    {
      "epoch": 0.11103151862464183,
      "grad_norm": 0.7707486152648926,
      "learning_rate": 4.999632951088732e-05,
      "loss": 0.6413,
      "mean_token_accuracy": 0.8040021538734436,
      "num_tokens": 62319382.0,
      "step": 1860
    },
    {
      "epoch": 0.11132999044890163,
      "grad_norm": 0.8063837885856628,
      "learning_rate": 4.999612620347528e-05,
      "loss": 0.6425,
      "mean_token_accuracy": 0.8032983303070068,
      "num_tokens": 62487062.0,
      "step": 1865
    },
    {
      "epoch": 0.11162846227316141,
      "grad_norm": 0.842122495174408,
      "learning_rate": 4.999591741671648e-05,
      "loss": 0.6961,
      "mean_token_accuracy": 0.789222252368927,
      "num_tokens": 62654742.0,
      "step": 1870
    },
    {
      "epoch": 0.1119269340974212,
      "grad_norm": 0.8580625057220459,
      "learning_rate": 4.9995703150661765e-05,
      "loss": 0.6505,
      "mean_token_accuracy": 0.8024335026741027,
      "num_tokens": 62822422.0,
      "step": 1875
    },
    {
      "epoch": 0.112225405921681,
      "grad_norm": 0.7586648464202881,
      "learning_rate": 4.999548340536332e-05,
      "loss": 0.6036,
      "mean_token_accuracy": 0.8125909566879272,
      "num_tokens": 62990102.0,
      "step": 1880
    },
    {
      "epoch": 0.11252387774594079,
      "grad_norm": 0.7322880029678345,
      "learning_rate": 4.9995258180874674e-05,
      "loss": 0.5983,
      "mean_token_accuracy": 0.8164380311965942,
      "num_tokens": 63157782.0,
      "step": 1885
    },
    {
      "epoch": 0.11282234957020057,
      "grad_norm": 0.75962233543396,
      "learning_rate": 4.9995027477250684e-05,
      "loss": 0.5895,
      "mean_token_accuracy": 0.8182989478111267,
      "num_tokens": 63325462.0,
      "step": 1890
    },
    {
      "epoch": 0.11312082139446036,
      "grad_norm": 0.7311409115791321,
      "learning_rate": 4.999479129454755e-05,
      "loss": 0.6697,
      "mean_token_accuracy": 0.796403431892395,
      "num_tokens": 63493142.0,
      "step": 1895
    },
    {
      "epoch": 0.11341929321872016,
      "grad_norm": 0.8264614343643188,
      "learning_rate": 4.999454963282278e-05,
      "loss": 0.6556,
      "mean_token_accuracy": 0.8048133254051208,
      "num_tokens": 63660822.0,
      "step": 1900
    },
    {
      "epoch": 0.11371776504297994,
      "grad_norm": 0.7244938015937805,
      "learning_rate": 4.9994302492135245e-05,
      "loss": 0.5858,
      "mean_token_accuracy": 0.8215078234672546,
      "num_tokens": 63828502.0,
      "step": 1905
    },
    {
      "epoch": 0.11401623686723973,
      "grad_norm": 0.825389564037323,
      "learning_rate": 4.999404987254514e-05,
      "loss": 0.6417,
      "mean_token_accuracy": 0.8050161004066467,
      "num_tokens": 63996182.0,
      "step": 1910
    },
    {
      "epoch": 0.11431470869149953,
      "grad_norm": 0.8485185503959656,
      "learning_rate": 4.9993791774114e-05,
      "loss": 0.6604,
      "mean_token_accuracy": 0.7995884537696838,
      "num_tokens": 64163862.0,
      "step": 1915
    },
    {
      "epoch": 0.11461318051575932,
      "grad_norm": 0.8129671812057495,
      "learning_rate": 4.999352819690469e-05,
      "loss": 0.6404,
      "mean_token_accuracy": 0.8028509974479675,
      "num_tokens": 64331542.0,
      "step": 1920
    },
    {
      "epoch": 0.1149116523400191,
      "grad_norm": 0.735074520111084,
      "learning_rate": 4.99932591409814e-05,
      "loss": 0.6448,
      "mean_token_accuracy": 0.8026720762252808,
      "num_tokens": 64499222.0,
      "step": 1925
    },
    {
      "epoch": 0.1152101241642789,
      "grad_norm": 0.8927582502365112,
      "learning_rate": 4.999298460640967e-05,
      "loss": 0.6991,
      "mean_token_accuracy": 0.7883752703666687,
      "num_tokens": 64666902.0,
      "step": 1930
    },
    {
      "epoch": 0.11550859598853869,
      "grad_norm": 0.761263906955719,
      "learning_rate": 4.9992704593256365e-05,
      "loss": 0.6513,
      "mean_token_accuracy": 0.7997017741203308,
      "num_tokens": 64834582.0,
      "step": 1935
    },
    {
      "epoch": 0.11580706781279847,
      "grad_norm": 0.7414693236351013,
      "learning_rate": 4.999241910158968e-05,
      "loss": 0.6103,
      "mean_token_accuracy": 0.8118871569633483,
      "num_tokens": 65002262.0,
      "step": 1940
    },
    {
      "epoch": 0.11610553963705826,
      "grad_norm": 0.7668074369430542,
      "learning_rate": 4.9992128131479167e-05,
      "loss": 0.6318,
      "mean_token_accuracy": 0.8079303145408631,
      "num_tokens": 65165118.0,
      "step": 1945
    },
    {
      "epoch": 0.11640401146131805,
      "grad_norm": 0.6964201331138611,
      "learning_rate": 4.9991831682995696e-05,
      "loss": 0.6713,
      "mean_token_accuracy": 0.7963974714279175,
      "num_tokens": 65332798.0,
      "step": 1950
    },
    {
      "epoch": 0.11670248328557783,
      "grad_norm": 0.7373561859130859,
      "learning_rate": 4.999152975621146e-05,
      "loss": 0.6682,
      "mean_token_accuracy": 0.7958904862403869,
      "num_tokens": 65500478.0,
      "step": 1955
    },
    {
      "epoch": 0.11700095510983763,
      "grad_norm": 0.91972815990448,
      "learning_rate": 4.999122235120001e-05,
      "loss": 0.6689,
      "mean_token_accuracy": 0.7971668839454651,
      "num_tokens": 65668158.0,
      "step": 1960
    },
    {
      "epoch": 0.11729942693409742,
      "grad_norm": 0.8223519325256348,
      "learning_rate": 4.999090946803621e-05,
      "loss": 0.6544,
      "mean_token_accuracy": 0.8011093854904174,
      "num_tokens": 65835838.0,
      "step": 1965
    },
    {
      "epoch": 0.11759789875835722,
      "grad_norm": 0.8795048594474792,
      "learning_rate": 4.9990591106796276e-05,
      "loss": 0.6549,
      "mean_token_accuracy": 0.7998031735420227,
      "num_tokens": 66003518.0,
      "step": 1970
    },
    {
      "epoch": 0.117896370582617,
      "grad_norm": 0.715836763381958,
      "learning_rate": 4.999026726755775e-05,
      "loss": 0.6149,
      "mean_token_accuracy": 0.8110819458961487,
      "num_tokens": 66171198.0,
      "step": 1975
    },
    {
      "epoch": 0.11819484240687679,
      "grad_norm": 0.7795296311378479,
      "learning_rate": 4.998993795039951e-05,
      "loss": 0.6072,
      "mean_token_accuracy": 0.8121853709220886,
      "num_tokens": 66338878.0,
      "step": 1980
    },
    {
      "epoch": 0.11849331423113658,
      "grad_norm": 0.8171116709709167,
      "learning_rate": 4.998960315540177e-05,
      "loss": 0.6071,
      "mean_token_accuracy": 0.8135213971138,
      "num_tokens": 66506558.0,
      "step": 1985
    },
    {
      "epoch": 0.11879178605539636,
      "grad_norm": 0.8781904578208923,
      "learning_rate": 4.998926288264606e-05,
      "loss": 0.5904,
      "mean_token_accuracy": 0.8168376445770263,
      "num_tokens": 66674238.0,
      "step": 1990
    },
    {
      "epoch": 0.11909025787965616,
      "grad_norm": 0.8734174370765686,
      "learning_rate": 4.998891713221528e-05,
      "loss": 0.6321,
      "mean_token_accuracy": 0.8071036696434021,
      "num_tokens": 66841918.0,
      "step": 1995
    },
    {
      "epoch": 0.11938872970391595,
      "grad_norm": 0.7292535901069641,
      "learning_rate": 4.998856590419363e-05,
      "loss": 0.6519,
      "mean_token_accuracy": 0.8018191576004028,
      "num_tokens": 67009598.0,
      "step": 2000
    },
    {
      "epoch": 0.11968720152817575,
      "grad_norm": 0.7165786027908325,
      "learning_rate": 4.9988209198666667e-05,
      "loss": 0.6151,
      "mean_token_accuracy": 0.8122390508651733,
      "num_tokens": 67177278.0,
      "step": 2005
    },
    {
      "epoch": 0.11998567335243553,
      "grad_norm": 0.694393515586853,
      "learning_rate": 4.998784701572127e-05,
      "loss": 0.6438,
      "mean_token_accuracy": 0.8043719410896302,
      "num_tokens": 67344958.0,
      "step": 2010
    },
    {
      "epoch": 0.12028414517669532,
      "grad_norm": 0.7211229205131531,
      "learning_rate": 4.9987479355445645e-05,
      "loss": 0.6312,
      "mean_token_accuracy": 0.8063461899757385,
      "num_tokens": 67512638.0,
      "step": 2015
    },
    {
      "epoch": 0.12058261700095511,
      "grad_norm": 0.8628489375114441,
      "learning_rate": 4.998710621792936e-05,
      "loss": 0.6962,
      "mean_token_accuracy": 0.788106894493103,
      "num_tokens": 67680318.0,
      "step": 2020
    },
    {
      "epoch": 0.1208810888252149,
      "grad_norm": 0.7633472681045532,
      "learning_rate": 4.998672760326329e-05,
      "loss": 0.6548,
      "mean_token_accuracy": 0.7996182680130005,
      "num_tokens": 67847998.0,
      "step": 2025
    },
    {
      "epoch": 0.12117956064947469,
      "grad_norm": 0.8130160570144653,
      "learning_rate": 4.9986343511539653e-05,
      "loss": 0.6258,
      "mean_token_accuracy": 0.8070142030715942,
      "num_tokens": 68015678.0,
      "step": 2030
    },
    {
      "epoch": 0.12147803247373448,
      "grad_norm": 0.6947306394577026,
      "learning_rate": 4.9985953942852e-05,
      "loss": 0.6495,
      "mean_token_accuracy": 0.801479172706604,
      "num_tokens": 68183358.0,
      "step": 2035
    },
    {
      "epoch": 0.12177650429799428,
      "grad_norm": 0.7634063363075256,
      "learning_rate": 4.998555889729523e-05,
      "loss": 0.6555,
      "mean_token_accuracy": 0.8009125709533691,
      "num_tokens": 68351038.0,
      "step": 2040
    },
    {
      "epoch": 0.12207497612225406,
      "grad_norm": 0.778538167476654,
      "learning_rate": 4.9985158374965543e-05,
      "loss": 0.5638,
      "mean_token_accuracy": 0.8260646462440491,
      "num_tokens": 68518718.0,
      "step": 2045
    },
    {
      "epoch": 0.12237344794651385,
      "grad_norm": 0.7564401030540466,
      "learning_rate": 4.998475237596051e-05,
      "loss": 0.6481,
      "mean_token_accuracy": 0.8013062119483948,
      "num_tokens": 68686398.0,
      "step": 2050
    },
    {
      "epoch": 0.12267191977077364,
      "grad_norm": 0.6965293288230896,
      "learning_rate": 4.9984340900379026e-05,
      "loss": 0.6503,
      "mean_token_accuracy": 0.8002326130867005,
      "num_tokens": 68854078.0,
      "step": 2055
    },
    {
      "epoch": 0.12297039159503342,
      "grad_norm": 0.7579342722892761,
      "learning_rate": 4.9983923948321287e-05,
      "loss": 0.639,
      "mean_token_accuracy": 0.804849112033844,
      "num_tokens": 69021758.0,
      "step": 2060
    },
    {
      "epoch": 0.12326886341929322,
      "grad_norm": 0.74488765001297,
      "learning_rate": 4.998350151988887e-05,
      "loss": 0.6642,
      "mean_token_accuracy": 0.7954431653022767,
      "num_tokens": 69189438.0,
      "step": 2065
    },
    {
      "epoch": 0.12356733524355301,
      "grad_norm": 0.8010881543159485,
      "learning_rate": 4.998307361518466e-05,
      "loss": 0.6219,
      "mean_token_accuracy": 0.8085888147354126,
      "num_tokens": 69357118.0,
      "step": 2070
    },
    {
      "epoch": 0.1238658070678128,
      "grad_norm": 0.7767967581748962,
      "learning_rate": 4.9982640234312875e-05,
      "loss": 0.6268,
      "mean_token_accuracy": 0.8063640713691711,
      "num_tokens": 69524798.0,
      "step": 2075
    },
    {
      "epoch": 0.12416427889207259,
      "grad_norm": 0.8799981474876404,
      "learning_rate": 4.998220137737909e-05,
      "loss": 0.6772,
      "mean_token_accuracy": 0.7931885957717896,
      "num_tokens": 69692478.0,
      "step": 2080
    },
    {
      "epoch": 0.12446275071633238,
      "grad_norm": 0.7806699275970459,
      "learning_rate": 4.998175704449019e-05,
      "loss": 0.6151,
      "mean_token_accuracy": 0.8097220540046692,
      "num_tokens": 69860158.0,
      "step": 2085
    },
    {
      "epoch": 0.12476122254059217,
      "grad_norm": 0.7636421918869019,
      "learning_rate": 4.998130723575438e-05,
      "loss": 0.6421,
      "mean_token_accuracy": 0.8033281683921814,
      "num_tokens": 70027838.0,
      "step": 2090
    },
    {
      "epoch": 0.12505969436485195,
      "grad_norm": 0.8024376630783081,
      "learning_rate": 4.998085195128124e-05,
      "loss": 0.6787,
      "mean_token_accuracy": 0.7936836361885071,
      "num_tokens": 70195518.0,
      "step": 2095
    },
    {
      "epoch": 0.12535816618911175,
      "grad_norm": 0.7960630059242249,
      "learning_rate": 4.998039119118166e-05,
      "loss": 0.6845,
      "mean_token_accuracy": 0.791202437877655,
      "num_tokens": 70363198.0,
      "step": 2100
    },
    {
      "epoch": 0.12565663801337154,
      "grad_norm": 0.8377149701118469,
      "learning_rate": 4.997992495556788e-05,
      "loss": 0.6462,
      "mean_token_accuracy": 0.8019623160362244,
      "num_tokens": 70530878.0,
      "step": 2105
    },
    {
      "epoch": 0.12595510983763133,
      "grad_norm": 0.8119000792503357,
      "learning_rate": 4.9979453244553424e-05,
      "loss": 0.6686,
      "mean_token_accuracy": 0.7952165126800537,
      "num_tokens": 70698558.0,
      "step": 2110
    },
    {
      "epoch": 0.12625358166189113,
      "grad_norm": 0.8080458045005798,
      "learning_rate": 4.9978976058253205e-05,
      "loss": 0.6991,
      "mean_token_accuracy": 0.789210295677185,
      "num_tokens": 70866238.0,
      "step": 2115
    },
    {
      "epoch": 0.1265520534861509,
      "grad_norm": 0.7718464732170105,
      "learning_rate": 4.997849339678347e-05,
      "loss": 0.6106,
      "mean_token_accuracy": 0.8126744508743287,
      "num_tokens": 71033918.0,
      "step": 2120
    },
    {
      "epoch": 0.1268505253104107,
      "grad_norm": 0.7288238406181335,
      "learning_rate": 4.997800526026176e-05,
      "loss": 0.649,
      "mean_token_accuracy": 0.8037635564804078,
      "num_tokens": 71201598.0,
      "step": 2125
    },
    {
      "epoch": 0.12714899713467048,
      "grad_norm": 0.7069281935691833,
      "learning_rate": 4.997751164880696e-05,
      "loss": 0.6353,
      "mean_token_accuracy": 0.8050817131996155,
      "num_tokens": 71369278.0,
      "step": 2130
    },
    {
      "epoch": 0.12744746895893028,
      "grad_norm": 0.7088195085525513,
      "learning_rate": 4.997701256253932e-05,
      "loss": 0.629,
      "mean_token_accuracy": 0.807306456565857,
      "num_tokens": 71536958.0,
      "step": 2135
    },
    {
      "epoch": 0.12774594078319007,
      "grad_norm": 0.7803517580032349,
      "learning_rate": 4.997650800158039e-05,
      "loss": 0.625,
      "mean_token_accuracy": 0.809537160396576,
      "num_tokens": 71704638.0,
      "step": 2140
    },
    {
      "epoch": 0.12804441260744986,
      "grad_norm": 0.8309139609336853,
      "learning_rate": 4.997599796605306e-05,
      "loss": 0.6059,
      "mean_token_accuracy": 0.8137361168861389,
      "num_tokens": 71872318.0,
      "step": 2145
    },
    {
      "epoch": 0.12834288443170966,
      "grad_norm": 0.754838228225708,
      "learning_rate": 4.997548245608158e-05,
      "loss": 0.6338,
      "mean_token_accuracy": 0.8056423664093018,
      "num_tokens": 72039998.0,
      "step": 2150
    },
    {
      "epoch": 0.12864135625596942,
      "grad_norm": 0.8363931179046631,
      "learning_rate": 4.9974961471791485e-05,
      "loss": 0.7028,
      "mean_token_accuracy": 0.787719190120697,
      "num_tokens": 72207678.0,
      "step": 2155
    },
    {
      "epoch": 0.12893982808022922,
      "grad_norm": 0.738895058631897,
      "learning_rate": 4.997443501330969e-05,
      "loss": 0.6208,
      "mean_token_accuracy": 0.8090898275375367,
      "num_tokens": 72375358.0,
      "step": 2160
    },
    {
      "epoch": 0.129238299904489,
      "grad_norm": 0.757001519203186,
      "learning_rate": 4.9973903080764416e-05,
      "loss": 0.6735,
      "mean_token_accuracy": 0.7951807141304016,
      "num_tokens": 72543038.0,
      "step": 2165
    },
    {
      "epoch": 0.1295367717287488,
      "grad_norm": 0.7282401919364929,
      "learning_rate": 4.997336567428523e-05,
      "loss": 0.6415,
      "mean_token_accuracy": 0.8038709282875061,
      "num_tokens": 72710718.0,
      "step": 2170
    },
    {
      "epoch": 0.1298352435530086,
      "grad_norm": 0.7109334468841553,
      "learning_rate": 4.9972822794003024e-05,
      "loss": 0.5997,
      "mean_token_accuracy": 0.8166050314903259,
      "num_tokens": 72878398.0,
      "step": 2175
    },
    {
      "epoch": 0.1301337153772684,
      "grad_norm": 0.7727491855621338,
      "learning_rate": 4.997227444005001e-05,
      "loss": 0.6291,
      "mean_token_accuracy": 0.807592761516571,
      "num_tokens": 73046078.0,
      "step": 2180
    },
    {
      "epoch": 0.1304321872015282,
      "grad_norm": 0.7181538939476013,
      "learning_rate": 4.997172061255978e-05,
      "loss": 0.6396,
      "mean_token_accuracy": 0.8033043026924134,
      "num_tokens": 73213758.0,
      "step": 2185
    },
    {
      "epoch": 0.13073065902578795,
      "grad_norm": 0.7119318842887878,
      "learning_rate": 4.9971161311667217e-05,
      "loss": 0.6073,
      "mean_token_accuracy": 0.812895143032074,
      "num_tokens": 73381438.0,
      "step": 2190
    },
    {
      "epoch": 0.13102913085004775,
      "grad_norm": 0.7757542133331299,
      "learning_rate": 4.9970596537508536e-05,
      "loss": 0.6207,
      "mean_token_accuracy": 0.8105809330940247,
      "num_tokens": 73549118.0,
      "step": 2195
    },
    {
      "epoch": 0.13132760267430754,
      "grad_norm": 0.7838243842124939,
      "learning_rate": 4.997002629022132e-05,
      "loss": 0.6521,
      "mean_token_accuracy": 0.8015567302703858,
      "num_tokens": 73716798.0,
      "step": 2200
    },
    {
      "epoch": 0.13162607449856734,
      "grad_norm": 0.7373713254928589,
      "learning_rate": 4.9969450569944435e-05,
      "loss": 0.6685,
      "mean_token_accuracy": 0.7966241240501404,
      "num_tokens": 73884478.0,
      "step": 2205
    },
    {
      "epoch": 0.13192454632282713,
      "grad_norm": 0.7715480327606201,
      "learning_rate": 4.996886937681814e-05,
      "loss": 0.6697,
      "mean_token_accuracy": 0.7969402313232422,
      "num_tokens": 74052158.0,
      "step": 2210
    },
    {
      "epoch": 0.13222301814708692,
      "grad_norm": 0.9639972448348999,
      "learning_rate": 4.996828271098398e-05,
      "loss": 0.6559,
      "mean_token_accuracy": 0.8007336139678956,
      "num_tokens": 74219838.0,
      "step": 2215
    },
    {
      "epoch": 0.13252148997134672,
      "grad_norm": 0.8003249764442444,
      "learning_rate": 4.996769057258484e-05,
      "loss": 0.6856,
      "mean_token_accuracy": 0.7893832802772522,
      "num_tokens": 74387518.0,
      "step": 2220
    },
    {
      "epoch": 0.13281996179560648,
      "grad_norm": 0.7151038646697998,
      "learning_rate": 4.996709296176496e-05,
      "loss": 0.6566,
      "mean_token_accuracy": 0.7986639738082886,
      "num_tokens": 74555198.0,
      "step": 2225
    },
    {
      "epoch": 0.13311843361986628,
      "grad_norm": 0.8242704272270203,
      "learning_rate": 4.9966489878669885e-05,
      "loss": 0.6345,
      "mean_token_accuracy": 0.8052427530288696,
      "num_tokens": 74722878.0,
      "step": 2230
    },
    {
      "epoch": 0.13341690544412607,
      "grad_norm": 0.7722378969192505,
      "learning_rate": 4.9965881323446534e-05,
      "loss": 0.6528,
      "mean_token_accuracy": 0.8037099003791809,
      "num_tokens": 74890558.0,
      "step": 2235
    },
    {
      "epoch": 0.13371537726838587,
      "grad_norm": 0.7419425845146179,
      "learning_rate": 4.99652672962431e-05,
      "loss": 0.6253,
      "mean_token_accuracy": 0.8088989734649659,
      "num_tokens": 75058238.0,
      "step": 2240
    },
    {
      "epoch": 0.13401384909264566,
      "grad_norm": 0.8640178442001343,
      "learning_rate": 4.996464779720916e-05,
      "loss": 0.6495,
      "mean_token_accuracy": 0.8026601433753967,
      "num_tokens": 75225918.0,
      "step": 2245
    },
    {
      "epoch": 0.13431232091690545,
      "grad_norm": 0.821284830570221,
      "learning_rate": 4.99640228264956e-05,
      "loss": 0.6683,
      "mean_token_accuracy": 0.7976380825042725,
      "num_tokens": 75393598.0,
      "step": 2250
    },
    {
      "epoch": 0.13461079274116525,
      "grad_norm": 0.6626023054122925,
      "learning_rate": 4.9963392384254635e-05,
      "loss": 0.6217,
      "mean_token_accuracy": 0.8092627882957458,
      "num_tokens": 75561278.0,
      "step": 2255
    },
    {
      "epoch": 0.134909264565425,
      "grad_norm": 0.7161763906478882,
      "learning_rate": 4.9962756470639846e-05,
      "loss": 0.6096,
      "mean_token_accuracy": 0.8110342264175415,
      "num_tokens": 75728958.0,
      "step": 2260
    },
    {
      "epoch": 0.1352077363896848,
      "grad_norm": 0.7436273097991943,
      "learning_rate": 4.9962115085806086e-05,
      "loss": 0.619,
      "mean_token_accuracy": 0.8112489461898804,
      "num_tokens": 75896638.0,
      "step": 2265
    },
    {
      "epoch": 0.1355062082139446,
      "grad_norm": 0.7392136454582214,
      "learning_rate": 4.99614682299096e-05,
      "loss": 0.6556,
      "mean_token_accuracy": 0.8005606532096863,
      "num_tokens": 76064318.0,
      "step": 2270
    },
    {
      "epoch": 0.1358046800382044,
      "grad_norm": 0.7133291959762573,
      "learning_rate": 4.996081590310794e-05,
      "loss": 0.6379,
      "mean_token_accuracy": 0.8034057021141052,
      "num_tokens": 76231998.0,
      "step": 2275
    },
    {
      "epoch": 0.1361031518624642,
      "grad_norm": 0.6754491925239563,
      "learning_rate": 4.996015810555997e-05,
      "loss": 0.5799,
      "mean_token_accuracy": 0.822205638885498,
      "num_tokens": 76399678.0,
      "step": 2280
    },
    {
      "epoch": 0.13640162368672398,
      "grad_norm": 0.7440774440765381,
      "learning_rate": 4.995949483742594e-05,
      "loss": 0.6735,
      "mean_token_accuracy": 0.7971609234809875,
      "num_tokens": 76567358.0,
      "step": 2285
    },
    {
      "epoch": 0.13670009551098378,
      "grad_norm": 0.7381521463394165,
      "learning_rate": 4.995882609886739e-05,
      "loss": 0.6281,
      "mean_token_accuracy": 0.8080042839050293,
      "num_tokens": 76735038.0,
      "step": 2290
    },
    {
      "epoch": 0.13699856733524354,
      "grad_norm": 0.8954285979270935,
      "learning_rate": 4.9958151890047196e-05,
      "loss": 0.6434,
      "mean_token_accuracy": 0.803107476234436,
      "num_tokens": 76902718.0,
      "step": 2295
    },
    {
      "epoch": 0.13729703915950334,
      "grad_norm": 0.7358289361000061,
      "learning_rate": 4.995747221112958e-05,
      "loss": 0.6329,
      "mean_token_accuracy": 0.8068173646926879,
      "num_tokens": 77070398.0,
      "step": 2300
    },
    {
      "epoch": 0.13759551098376313,
      "grad_norm": 0.7513579726219177,
      "learning_rate": 4.995678706228009e-05,
      "loss": 0.6164,
      "mean_token_accuracy": 0.8093999862670899,
      "num_tokens": 77238078.0,
      "step": 2305
    },
    {
      "epoch": 0.13789398280802292,
      "grad_norm": 0.7520550489425659,
      "learning_rate": 4.995609644366561e-05,
      "loss": 0.6325,
      "mean_token_accuracy": 0.8071692705154419,
      "num_tokens": 77405758.0,
      "step": 2310
    },
    {
      "epoch": 0.13819245463228272,
      "grad_norm": 0.7596616148948669,
      "learning_rate": 4.995540035545435e-05,
      "loss": 0.6234,
      "mean_token_accuracy": 0.8089466691017151,
      "num_tokens": 77573438.0,
      "step": 2315
    },
    {
      "epoch": 0.1384909264565425,
      "grad_norm": 0.7286480069160461,
      "learning_rate": 4.995469879781584e-05,
      "loss": 0.6152,
      "mean_token_accuracy": 0.8094775080680847,
      "num_tokens": 77741118.0,
      "step": 2320
    },
    {
      "epoch": 0.13878939828080228,
      "grad_norm": 0.7319132089614868,
      "learning_rate": 4.9953991770920984e-05,
      "loss": 0.7041,
      "mean_token_accuracy": 0.7839556336402893,
      "num_tokens": 77908798.0,
      "step": 2325
    },
    {
      "epoch": 0.13908787010506207,
      "grad_norm": 0.7207380533218384,
      "learning_rate": 4.995327927494198e-05,
      "loss": 0.6218,
      "mean_token_accuracy": 0.8090957880020142,
      "num_tokens": 78076478.0,
      "step": 2330
    },
    {
      "epoch": 0.13938634192932187,
      "grad_norm": 0.728185772895813,
      "learning_rate": 4.9952561310052365e-05,
      "loss": 0.5835,
      "mean_token_accuracy": 0.819336760044098,
      "num_tokens": 78244158.0,
      "step": 2335
    },
    {
      "epoch": 0.13968481375358166,
      "grad_norm": 0.7341372966766357,
      "learning_rate": 4.995183787642701e-05,
      "loss": 0.65,
      "mean_token_accuracy": 0.8011988639831543,
      "num_tokens": 78411838.0,
      "step": 2340
    },
    {
      "epoch": 0.13998328557784145,
      "grad_norm": 0.8130307793617249,
      "learning_rate": 4.9951108974242136e-05,
      "loss": 0.5887,
      "mean_token_accuracy": 0.816390311717987,
      "num_tokens": 78579518.0,
      "step": 2345
    },
    {
      "epoch": 0.14028175740210125,
      "grad_norm": 0.7904885411262512,
      "learning_rate": 4.995037460367527e-05,
      "loss": 0.6605,
      "mean_token_accuracy": 0.7984969615936279,
      "num_tokens": 78747198.0,
      "step": 2350
    },
    {
      "epoch": 0.14058022922636104,
      "grad_norm": 0.7222793698310852,
      "learning_rate": 4.9949634764905276e-05,
      "loss": 0.5814,
      "mean_token_accuracy": 0.8193964004516602,
      "num_tokens": 78914878.0,
      "step": 2355
    },
    {
      "epoch": 0.1408787010506208,
      "grad_norm": 0.6837400197982788,
      "learning_rate": 4.994888945811237e-05,
      "loss": 0.636,
      "mean_token_accuracy": 0.8061791658401489,
      "num_tokens": 79082558.0,
      "step": 2360
    },
    {
      "epoch": 0.1411771728748806,
      "grad_norm": 0.7385311722755432,
      "learning_rate": 4.9948138683478076e-05,
      "loss": 0.5857,
      "mean_token_accuracy": 0.8195335865020752,
      "num_tokens": 79250238.0,
      "step": 2365
    },
    {
      "epoch": 0.1414756446991404,
      "grad_norm": 0.7428717613220215,
      "learning_rate": 4.9947382441185255e-05,
      "loss": 0.6286,
      "mean_token_accuracy": 0.8067935109138489,
      "num_tokens": 79417918.0,
      "step": 2370
    },
    {
      "epoch": 0.1417741165234002,
      "grad_norm": 0.6756755113601685,
      "learning_rate": 4.9946620731418115e-05,
      "loss": 0.624,
      "mean_token_accuracy": 0.807807457447052,
      "num_tokens": 79585598.0,
      "step": 2375
    },
    {
      "epoch": 0.14207258834765998,
      "grad_norm": 0.6760066151618958,
      "learning_rate": 4.9945853554362177e-05,
      "loss": 0.6122,
      "mean_token_accuracy": 0.8115948915481568,
      "num_tokens": 79753278.0,
      "step": 2380
    },
    {
      "epoch": 0.14237106017191978,
      "grad_norm": 0.7346377968788147,
      "learning_rate": 4.99450809102043e-05,
      "loss": 0.6596,
      "mean_token_accuracy": 0.798007869720459,
      "num_tokens": 79920958.0,
      "step": 2385
    },
    {
      "epoch": 0.14266953199617957,
      "grad_norm": 0.717136561870575,
      "learning_rate": 4.994430279913268e-05,
      "loss": 0.6748,
      "mean_token_accuracy": 0.7956221103668213,
      "num_tokens": 80088638.0,
      "step": 2390
    },
    {
      "epoch": 0.14296800382043934,
      "grad_norm": 0.7092469930648804,
      "learning_rate": 4.9943519221336844e-05,
      "loss": 0.6876,
      "mean_token_accuracy": 0.7916497588157654,
      "num_tokens": 80256318.0,
      "step": 2395
    },
    {
      "epoch": 0.14326647564469913,
      "grad_norm": 0.7810541987419128,
      "learning_rate": 4.994273017700764e-05,
      "loss": 0.6212,
      "mean_token_accuracy": 0.8079028964042664,
      "num_tokens": 80423998.0,
      "step": 2400
    },
    {
      "epoch": 0.14356494746895893,
      "grad_norm": 0.7001931667327881,
      "learning_rate": 4.994193566633725e-05,
      "loss": 0.6196,
      "mean_token_accuracy": 0.8095968127250671,
      "num_tokens": 80591678.0,
      "step": 2405
    },
    {
      "epoch": 0.14386341929321872,
      "grad_norm": 0.7972292900085449,
      "learning_rate": 4.9941135689519206e-05,
      "loss": 0.6592,
      "mean_token_accuracy": 0.7987116694450378,
      "num_tokens": 80759358.0,
      "step": 2410
    },
    {
      "epoch": 0.1441618911174785,
      "grad_norm": 0.6630339622497559,
      "learning_rate": 4.9940330246748343e-05,
      "loss": 0.6221,
      "mean_token_accuracy": 0.8100262522697449,
      "num_tokens": 80927038.0,
      "step": 2415
    },
    {
      "epoch": 0.1444603629417383,
      "grad_norm": 0.8520267009735107,
      "learning_rate": 4.993951933822085e-05,
      "loss": 0.6998,
      "mean_token_accuracy": 0.7876178026199341,
      "num_tokens": 81094718.0,
      "step": 2420
    },
    {
      "epoch": 0.1447588347659981,
      "grad_norm": 0.7509779334068298,
      "learning_rate": 4.993870296413424e-05,
      "loss": 0.6515,
      "mean_token_accuracy": 0.802505087852478,
      "num_tokens": 81262398.0,
      "step": 2425
    },
    {
      "epoch": 0.14505730659025787,
      "grad_norm": 0.8092514276504517,
      "learning_rate": 4.993788112468734e-05,
      "loss": 0.6536,
      "mean_token_accuracy": 0.8006859183311462,
      "num_tokens": 81430078.0,
      "step": 2430
    },
    {
      "epoch": 0.14535577841451766,
      "grad_norm": 0.7288460731506348,
      "learning_rate": 4.993705382008035e-05,
      "loss": 0.6407,
      "mean_token_accuracy": 0.8037635803222656,
      "num_tokens": 81597758.0,
      "step": 2435
    },
    {
      "epoch": 0.14565425023877746,
      "grad_norm": 0.9207749962806702,
      "learning_rate": 4.993622105051475e-05,
      "loss": 0.6298,
      "mean_token_accuracy": 0.8070499777793885,
      "num_tokens": 81765438.0,
      "step": 2440
    },
    {
      "epoch": 0.14595272206303725,
      "grad_norm": 0.6609447598457336,
      "learning_rate": 4.993538281619339e-05,
      "loss": 0.5507,
      "mean_token_accuracy": 0.8290647864341736,
      "num_tokens": 81933118.0,
      "step": 2445
    },
    {
      "epoch": 0.14625119388729704,
      "grad_norm": 0.7385303974151611,
      "learning_rate": 4.993453911732045e-05,
      "loss": 0.6671,
      "mean_token_accuracy": 0.7958666324615479,
      "num_tokens": 82100798.0,
      "step": 2450
    },
    {
      "epoch": 0.14654966571155684,
      "grad_norm": 0.7430198192596436,
      "learning_rate": 4.993368995410139e-05,
      "loss": 0.6335,
      "mean_token_accuracy": 0.8055111527442932,
      "num_tokens": 82268478.0,
      "step": 2455
    },
    {
      "epoch": 0.14684813753581663,
      "grad_norm": 0.6414621472358704,
      "learning_rate": 4.993283532674308e-05,
      "loss": 0.6154,
      "mean_token_accuracy": 0.8096445202827454,
      "num_tokens": 82436158.0,
      "step": 2460
    },
    {
      "epoch": 0.1471466093600764,
      "grad_norm": 0.7025129199028015,
      "learning_rate": 4.9931975235453666e-05,
      "loss": 0.6248,
      "mean_token_accuracy": 0.8093164682388305,
      "num_tokens": 82603838.0,
      "step": 2465
    },
    {
      "epoch": 0.1474450811843362,
      "grad_norm": 0.690983235836029,
      "learning_rate": 4.9931109680442625e-05,
      "loss": 0.6328,
      "mean_token_accuracy": 0.8053441405296325,
      "num_tokens": 82771518.0,
      "step": 2470
    },
    {
      "epoch": 0.14774355300859598,
      "grad_norm": 0.7963226437568665,
      "learning_rate": 4.993023866192079e-05,
      "loss": 0.6131,
      "mean_token_accuracy": 0.81239413022995,
      "num_tokens": 82939198.0,
      "step": 2475
    },
    {
      "epoch": 0.14804202483285578,
      "grad_norm": 0.823704183101654,
      "learning_rate": 4.9929362180100326e-05,
      "loss": 0.6226,
      "mean_token_accuracy": 0.811153519153595,
      "num_tokens": 83106878.0,
      "step": 2480
    },
    {
      "epoch": 0.14834049665711557,
      "grad_norm": 0.6541733741760254,
      "learning_rate": 4.99284802351947e-05,
      "loss": 0.6085,
      "mean_token_accuracy": 0.8119646906852722,
      "num_tokens": 83274558.0,
      "step": 2485
    },
    {
      "epoch": 0.14863896848137537,
      "grad_norm": 0.7201209664344788,
      "learning_rate": 4.9927592827418725e-05,
      "loss": 0.6959,
      "mean_token_accuracy": 0.7875044703483581,
      "num_tokens": 83442238.0,
      "step": 2490
    },
    {
      "epoch": 0.14893744030563516,
      "grad_norm": 0.7420956492424011,
      "learning_rate": 4.992669995698855e-05,
      "loss": 0.6343,
      "mean_token_accuracy": 0.8059763789176941,
      "num_tokens": 83609918.0,
      "step": 2495
    },
    {
      "epoch": 0.14923591212989493,
      "grad_norm": 0.7263451218605042,
      "learning_rate": 4.9925801624121657e-05,
      "loss": 0.6304,
      "mean_token_accuracy": 0.8056423664093018,
      "num_tokens": 83777598.0,
      "step": 2500
    },
    {
      "epoch": 0.14953438395415472,
      "grad_norm": 0.7237770557403564,
      "learning_rate": 4.992489782903684e-05,
      "loss": 0.6519,
      "mean_token_accuracy": 0.8010855317115784,
      "num_tokens": 83945278.0,
      "step": 2505
    },
    {
      "epoch": 0.14983285577841451,
      "grad_norm": 0.6878868937492371,
      "learning_rate": 4.992398857195425e-05,
      "loss": 0.601,
      "mean_token_accuracy": 0.8140462875366211,
      "num_tokens": 84112958.0,
      "step": 2510
    },
    {
      "epoch": 0.1501313276026743,
      "grad_norm": 0.698435366153717,
      "learning_rate": 4.9923073853095334e-05,
      "loss": 0.6264,
      "mean_token_accuracy": 0.8081713080406189,
      "num_tokens": 84280638.0,
      "step": 2515
    },
    {
      "epoch": 0.1504297994269341,
      "grad_norm": 0.7715688943862915,
      "learning_rate": 4.992215367268291e-05,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8235536098480225,
      "num_tokens": 84448318.0,
      "step": 2520
    },
    {
      "epoch": 0.1507282712511939,
      "grad_norm": 0.718703031539917,
      "learning_rate": 4.9921228030941086e-05,
      "loss": 0.6642,
      "mean_token_accuracy": 0.7946916341781616,
      "num_tokens": 84615998.0,
      "step": 2525
    },
    {
      "epoch": 0.1510267430754537,
      "grad_norm": 0.696151077747345,
      "learning_rate": 4.992029692809532e-05,
      "loss": 0.6378,
      "mean_token_accuracy": 0.8031790614128113,
      "num_tokens": 84783678.0,
      "step": 2530
    },
    {
      "epoch": 0.15132521489971346,
      "grad_norm": 0.6987526416778564,
      "learning_rate": 4.991936036437241e-05,
      "loss": 0.6225,
      "mean_token_accuracy": 0.8104570865631103,
      "num_tokens": 84945647.0,
      "step": 2535
    },
    {
      "epoch": 0.15162368672397325,
      "grad_norm": 0.7180252075195312,
      "learning_rate": 4.991841834000048e-05,
      "loss": 0.6288,
      "mean_token_accuracy": 0.8090063214302063,
      "num_tokens": 85113327.0,
      "step": 2540
    },
    {
      "epoch": 0.15192215854823304,
      "grad_norm": 0.7295783162117004,
      "learning_rate": 4.991747085520895e-05,
      "loss": 0.6599,
      "mean_token_accuracy": 0.7979124426841736,
      "num_tokens": 85281007.0,
      "step": 2545
    },
    {
      "epoch": 0.15222063037249284,
      "grad_norm": 0.7051554322242737,
      "learning_rate": 4.991651791022863e-05,
      "loss": 0.6556,
      "mean_token_accuracy": 0.7968507647514343,
      "num_tokens": 85448687.0,
      "step": 2550
    },
    {
      "epoch": 0.15251910219675263,
      "grad_norm": 0.7344266176223755,
      "learning_rate": 4.9915559505291604e-05,
      "loss": 0.611,
      "mean_token_accuracy": 0.8117201447486877,
      "num_tokens": 85616367.0,
      "step": 2555
    },
    {
      "epoch": 0.15281757402101243,
      "grad_norm": 0.6918289661407471,
      "learning_rate": 4.991459564063132e-05,
      "loss": 0.6469,
      "mean_token_accuracy": 0.8017654657363892,
      "num_tokens": 85784047.0,
      "step": 2560
    },
    {
      "epoch": 0.15311604584527222,
      "grad_norm": 0.650274932384491,
      "learning_rate": 4.991362631648254e-05,
      "loss": 0.6114,
      "mean_token_accuracy": 0.8120899438858032,
      "num_tokens": 85951727.0,
      "step": 2565
    },
    {
      "epoch": 0.15341451766953199,
      "grad_norm": 0.7291713953018188,
      "learning_rate": 4.991265153308136e-05,
      "loss": 0.6401,
      "mean_token_accuracy": 0.8030537962913513,
      "num_tokens": 86119407.0,
      "step": 2570
    },
    {
      "epoch": 0.15371298949379178,
      "grad_norm": 0.66012042760849,
      "learning_rate": 4.99116712906652e-05,
      "loss": 0.6298,
      "mean_token_accuracy": 0.8075092554092407,
      "num_tokens": 86287087.0,
      "step": 2575
    },
    {
      "epoch": 0.15401146131805157,
      "grad_norm": 0.7682079672813416,
      "learning_rate": 4.991068558947284e-05,
      "loss": 0.6605,
      "mean_token_accuracy": 0.7970118045806884,
      "num_tokens": 86454767.0,
      "step": 2580
    },
    {
      "epoch": 0.15430993314231137,
      "grad_norm": 0.637602686882019,
      "learning_rate": 4.9909694429744346e-05,
      "loss": 0.6247,
      "mean_token_accuracy": 0.8069128036499024,
      "num_tokens": 86622447.0,
      "step": 2585
    },
    {
      "epoch": 0.15460840496657116,
      "grad_norm": 0.7632467150688171,
      "learning_rate": 4.990869781172114e-05,
      "loss": 0.6424,
      "mean_token_accuracy": 0.800876772403717,
      "num_tokens": 86790127.0,
      "step": 2590
    },
    {
      "epoch": 0.15490687679083096,
      "grad_norm": 0.67705237865448,
      "learning_rate": 4.990769573564596e-05,
      "loss": 0.6103,
      "mean_token_accuracy": 0.8119527578353882,
      "num_tokens": 86957807.0,
      "step": 2595
    },
    {
      "epoch": 0.15520534861509075,
      "grad_norm": 0.6863157153129578,
      "learning_rate": 4.990668820176289e-05,
      "loss": 0.6296,
      "mean_token_accuracy": 0.8071275234222413,
      "num_tokens": 87125487.0,
      "step": 2600
    },
    {
      "epoch": 0.15550382043935052,
      "grad_norm": 0.7088929414749146,
      "learning_rate": 4.990567521031733e-05,
      "loss": 0.6324,
      "mean_token_accuracy": 0.808038866519928,
      "num_tokens": 87286740.0,
      "step": 2605
    },
    {
      "epoch": 0.1558022922636103,
      "grad_norm": 0.7194373607635498,
      "learning_rate": 4.9904656761556e-05,
      "loss": 0.5995,
      "mean_token_accuracy": 0.8147023677825928,
      "num_tokens": 87454420.0,
      "step": 2610
    },
    {
      "epoch": 0.1561007640878701,
      "grad_norm": 0.6910497546195984,
      "learning_rate": 4.9903632855726983e-05,
      "loss": 0.606,
      "mean_token_accuracy": 0.8130621552467346,
      "num_tokens": 87622100.0,
      "step": 2615
    },
    {
      "epoch": 0.1563992359121299,
      "grad_norm": 0.794522225856781,
      "learning_rate": 4.990260349307966e-05,
      "loss": 0.6257,
      "mean_token_accuracy": 0.8078551769256592,
      "num_tokens": 87789780.0,
      "step": 2620
    },
    {
      "epoch": 0.1566977077363897,
      "grad_norm": 0.7447571754455566,
      "learning_rate": 4.990156867386475e-05,
      "loss": 0.6338,
      "mean_token_accuracy": 0.8046463012695313,
      "num_tokens": 87957460.0,
      "step": 2625
    },
    {
      "epoch": 0.15699617956064948,
      "grad_norm": 0.6099318265914917,
      "learning_rate": 4.9900528398334304e-05,
      "loss": 0.6216,
      "mean_token_accuracy": 0.8102111458778382,
      "num_tokens": 88125140.0,
      "step": 2630
    },
    {
      "epoch": 0.15729465138490925,
      "grad_norm": 0.7270940542221069,
      "learning_rate": 4.989948266674169e-05,
      "loss": 0.6803,
      "mean_token_accuracy": 0.7926100611686706,
      "num_tokens": 88292820.0,
      "step": 2635
    },
    {
      "epoch": 0.15759312320916904,
      "grad_norm": 0.8718281984329224,
      "learning_rate": 4.989843147934164e-05,
      "loss": 0.6381,
      "mean_token_accuracy": 0.8051175117492676,
      "num_tokens": 88460500.0,
      "step": 2640
    },
    {
      "epoch": 0.15789159503342884,
      "grad_norm": 0.7283470630645752,
      "learning_rate": 4.989737483639018e-05,
      "loss": 0.6376,
      "mean_token_accuracy": 0.8044673800468445,
      "num_tokens": 88628180.0,
      "step": 2645
    },
    {
      "epoch": 0.15819006685768863,
      "grad_norm": 0.7733397483825684,
      "learning_rate": 4.989631273814465e-05,
      "loss": 0.6539,
      "mean_token_accuracy": 0.7983180046081543,
      "num_tokens": 88795860.0,
      "step": 2650
    },
    {
      "epoch": 0.15848853868194843,
      "grad_norm": 0.7224664092063904,
      "learning_rate": 4.9895245184863775e-05,
      "loss": 0.6389,
      "mean_token_accuracy": 0.8033221960067749,
      "num_tokens": 88963540.0,
      "step": 2655
    },
    {
      "epoch": 0.15878701050620822,
      "grad_norm": 0.8396055102348328,
      "learning_rate": 4.989417217680757e-05,
      "loss": 0.6487,
      "mean_token_accuracy": 0.803220796585083,
      "num_tokens": 89131220.0,
      "step": 2660
    },
    {
      "epoch": 0.15908548233046801,
      "grad_norm": 0.7238900065422058,
      "learning_rate": 4.9893093714237384e-05,
      "loss": 0.6222,
      "mean_token_accuracy": 0.8101693868637085,
      "num_tokens": 89298900.0,
      "step": 2665
    },
    {
      "epoch": 0.15938395415472778,
      "grad_norm": 0.7223188877105713,
      "learning_rate": 4.9892009797415886e-05,
      "loss": 0.6257,
      "mean_token_accuracy": 0.8076583623886109,
      "num_tokens": 89466580.0,
      "step": 2670
    },
    {
      "epoch": 0.15968242597898757,
      "grad_norm": 0.7806656360626221,
      "learning_rate": 4.989092042660709e-05,
      "loss": 0.619,
      "mean_token_accuracy": 0.8088154673576355,
      "num_tokens": 89634260.0,
      "step": 2675
    },
    {
      "epoch": 0.15998089780324737,
      "grad_norm": 0.7205828428268433,
      "learning_rate": 4.988982560207635e-05,
      "loss": 0.6064,
      "mean_token_accuracy": 0.8118513584136963,
      "num_tokens": 89801940.0,
      "step": 2680
    },
    {
      "epoch": 0.16027936962750716,
      "grad_norm": 0.8668790459632874,
      "learning_rate": 4.98887253240903e-05,
      "loss": 0.6109,
      "mean_token_accuracy": 0.8123344779014587,
      "num_tokens": 89969620.0,
      "step": 2685
    },
    {
      "epoch": 0.16057784145176696,
      "grad_norm": 0.8031144738197327,
      "learning_rate": 4.9887619592916956e-05,
      "loss": 0.6615,
      "mean_token_accuracy": 0.7974710822105407,
      "num_tokens": 90137300.0,
      "step": 2690
    },
    {
      "epoch": 0.16087631327602675,
      "grad_norm": 0.7694268226623535,
      "learning_rate": 4.9886508408825635e-05,
      "loss": 0.5922,
      "mean_token_accuracy": 0.8175295233726502,
      "num_tokens": 90304980.0,
      "step": 2695
    },
    {
      "epoch": 0.16117478510028654,
      "grad_norm": 0.659078061580658,
      "learning_rate": 4.988539177208698e-05,
      "loss": 0.5937,
      "mean_token_accuracy": 0.8154300451278687,
      "num_tokens": 90472660.0,
      "step": 2700
    },
    {
      "epoch": 0.1614732569245463,
      "grad_norm": 0.7620792984962463,
      "learning_rate": 4.988426968297297e-05,
      "loss": 0.6588,
      "mean_token_accuracy": 0.7991291880607605,
      "num_tokens": 90640340.0,
      "step": 2705
    },
    {
      "epoch": 0.1617717287488061,
      "grad_norm": 0.7885122895240784,
      "learning_rate": 4.9883142141756925e-05,
      "loss": 0.6346,
      "mean_token_accuracy": 0.8048073530197144,
      "num_tokens": 90808020.0,
      "step": 2710
    },
    {
      "epoch": 0.1620702005730659,
      "grad_norm": 0.8176653385162354,
      "learning_rate": 4.9882009148713445e-05,
      "loss": 0.6411,
      "mean_token_accuracy": 0.8032267689704895,
      "num_tokens": 90975700.0,
      "step": 2715
    },
    {
      "epoch": 0.1623686723973257,
      "grad_norm": 0.7681753039360046,
      "learning_rate": 4.988087070411853e-05,
      "loss": 0.6167,
      "mean_token_accuracy": 0.8108433604240417,
      "num_tokens": 91143380.0,
      "step": 2720
    },
    {
      "epoch": 0.16266714422158549,
      "grad_norm": 0.9150328636169434,
      "learning_rate": 4.987972680824944e-05,
      "loss": 0.5778,
      "mean_token_accuracy": 0.822557556629181,
      "num_tokens": 91311060.0,
      "step": 2725
    },
    {
      "epoch": 0.16296561604584528,
      "grad_norm": 0.6954326033592224,
      "learning_rate": 4.987857746138481e-05,
      "loss": 0.5959,
      "mean_token_accuracy": 0.8174579501152038,
      "num_tokens": 91478740.0,
      "step": 2730
    },
    {
      "epoch": 0.16326408787010507,
      "grad_norm": 0.6668336391448975,
      "learning_rate": 4.9877422663804574e-05,
      "loss": 0.6131,
      "mean_token_accuracy": 0.8127937555313111,
      "num_tokens": 91646420.0,
      "step": 2735
    },
    {
      "epoch": 0.16356255969436484,
      "grad_norm": 0.7585785388946533,
      "learning_rate": 4.987626241579001e-05,
      "loss": 0.6386,
      "mean_token_accuracy": 0.8051234602928161,
      "num_tokens": 91814100.0,
      "step": 2740
    },
    {
      "epoch": 0.16386103151862463,
      "grad_norm": 0.7316513061523438,
      "learning_rate": 4.9875096717623716e-05,
      "loss": 0.6573,
      "mean_token_accuracy": 0.7992961883544922,
      "num_tokens": 91981780.0,
      "step": 2745
    },
    {
      "epoch": 0.16415950334288443,
      "grad_norm": 0.7090036869049072,
      "learning_rate": 4.987392556958962e-05,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8201240777969361,
      "num_tokens": 92149460.0,
      "step": 2750
    },
    {
      "epoch": 0.16445797516714422,
      "grad_norm": 0.6841456294059753,
      "learning_rate": 4.9872748971972976e-05,
      "loss": 0.6074,
      "mean_token_accuracy": 0.8100679874420166,
      "num_tokens": 92317140.0,
      "step": 2755
    },
    {
      "epoch": 0.16475644699140402,
      "grad_norm": 0.6879178285598755,
      "learning_rate": 4.987156692506037e-05,
      "loss": 0.6216,
      "mean_token_accuracy": 0.8083144426345825,
      "num_tokens": 92484820.0,
      "step": 2760
    },
    {
      "epoch": 0.1650549188156638,
      "grad_norm": 0.7845852375030518,
      "learning_rate": 4.9870379429139694e-05,
      "loss": 0.6308,
      "mean_token_accuracy": 0.8065370440483093,
      "num_tokens": 92652500.0,
      "step": 2765
    },
    {
      "epoch": 0.1653533906399236,
      "grad_norm": 0.7859224677085876,
      "learning_rate": 4.986918648450021e-05,
      "loss": 0.6197,
      "mean_token_accuracy": 0.8071454048156739,
      "num_tokens": 92820180.0,
      "step": 2770
    },
    {
      "epoch": 0.16565186246418337,
      "grad_norm": 0.7830347418785095,
      "learning_rate": 4.9867988091432475e-05,
      "loss": 0.6003,
      "mean_token_accuracy": 0.8133305430412292,
      "num_tokens": 92987860.0,
      "step": 2775
    },
    {
      "epoch": 0.16595033428844316,
      "grad_norm": 0.733372688293457,
      "learning_rate": 4.986678425022836e-05,
      "loss": 0.5964,
      "mean_token_accuracy": 0.816903269290924,
      "num_tokens": 93155540.0,
      "step": 2780
    },
    {
      "epoch": 0.16624880611270296,
      "grad_norm": 0.8022065758705139,
      "learning_rate": 4.986557496118111e-05,
      "loss": 0.7015,
      "mean_token_accuracy": 0.785088860988617,
      "num_tokens": 93323220.0,
      "step": 2785
    },
    {
      "epoch": 0.16654727793696275,
      "grad_norm": 0.7595603466033936,
      "learning_rate": 4.986436022458525e-05,
      "loss": 0.6271,
      "mean_token_accuracy": 0.8079506158828735,
      "num_tokens": 93490900.0,
      "step": 2790
    },
    {
      "epoch": 0.16684574976122254,
      "grad_norm": 0.7025103569030762,
      "learning_rate": 4.9863140040736664e-05,
      "loss": 0.626,
      "mean_token_accuracy": 0.8075271368026733,
      "num_tokens": 93658580.0,
      "step": 2795
    },
    {
      "epoch": 0.16714422158548234,
      "grad_norm": 0.7312909364700317,
      "learning_rate": 4.9861914409932534e-05,
      "loss": 0.5975,
      "mean_token_accuracy": 0.815901231765747,
      "num_tokens": 93826260.0,
      "step": 2800
    },
    {
      "epoch": 0.16744269340974213,
      "grad_norm": 0.6187756061553955,
      "learning_rate": 4.9860683332471405e-05,
      "loss": 0.602,
      "mean_token_accuracy": 0.8137599945068359,
      "num_tokens": 93993940.0,
      "step": 2805
    },
    {
      "epoch": 0.1677411652340019,
      "grad_norm": 0.6624660491943359,
      "learning_rate": 4.985944680865312e-05,
      "loss": 0.6563,
      "mean_token_accuracy": 0.7977514147758484,
      "num_tokens": 94161620.0,
      "step": 2810
    },
    {
      "epoch": 0.1680396370582617,
      "grad_norm": 0.7377040386199951,
      "learning_rate": 4.985820483877885e-05,
      "loss": 0.6423,
      "mean_token_accuracy": 0.8036800622940063,
      "num_tokens": 94329300.0,
      "step": 2815
    },
    {
      "epoch": 0.1683381088825215,
      "grad_norm": 0.6923050880432129,
      "learning_rate": 4.985695742315111e-05,
      "loss": 0.6382,
      "mean_token_accuracy": 0.8044733285903931,
      "num_tokens": 94496980.0,
      "step": 2820
    },
    {
      "epoch": 0.16863658070678128,
      "grad_norm": 0.6681461334228516,
      "learning_rate": 4.9855704562073725e-05,
      "loss": 0.6192,
      "mean_token_accuracy": 0.8077597379684448,
      "num_tokens": 94664660.0,
      "step": 2825
    },
    {
      "epoch": 0.16893505253104107,
      "grad_norm": 0.6664122939109802,
      "learning_rate": 4.985444625585186e-05,
      "loss": 0.6101,
      "mean_token_accuracy": 0.8125134229660034,
      "num_tokens": 94832340.0,
      "step": 2830
    },
    {
      "epoch": 0.16923352435530087,
      "grad_norm": 0.7340320944786072,
      "learning_rate": 4.985318250479199e-05,
      "loss": 0.6455,
      "mean_token_accuracy": 0.8028092503547668,
      "num_tokens": 95000020.0,
      "step": 2835
    },
    {
      "epoch": 0.16953199617956066,
      "grad_norm": 0.7466009259223938,
      "learning_rate": 4.985191330920193e-05,
      "loss": 0.6375,
      "mean_token_accuracy": 0.8048819541931153,
      "num_tokens": 95166202.0,
      "step": 2840
    },
    {
      "epoch": 0.16983046800382043,
      "grad_norm": 0.678203284740448,
      "learning_rate": 4.9850638669390816e-05,
      "loss": 0.5868,
      "mean_token_accuracy": 0.8190087080001831,
      "num_tokens": 95333882.0,
      "step": 2845
    },
    {
      "epoch": 0.17012893982808022,
      "grad_norm": 0.6863088607788086,
      "learning_rate": 4.98493585856691e-05,
      "loss": 0.6066,
      "mean_token_accuracy": 0.8119885325431824,
      "num_tokens": 95501562.0,
      "step": 2850
    },
    {
      "epoch": 0.17042741165234002,
      "grad_norm": 0.7498096227645874,
      "learning_rate": 4.984807305834859e-05,
      "loss": 0.6694,
      "mean_token_accuracy": 0.7951270341873169,
      "num_tokens": 95669242.0,
      "step": 2855
    },
    {
      "epoch": 0.1707258834765998,
      "grad_norm": 0.6950475573539734,
      "learning_rate": 4.9846782087742384e-05,
      "loss": 0.6398,
      "mean_token_accuracy": 0.8023857831954956,
      "num_tokens": 95836922.0,
      "step": 2860
    },
    {
      "epoch": 0.1710243553008596,
      "grad_norm": 0.7693982124328613,
      "learning_rate": 4.984548567416493e-05,
      "loss": 0.62,
      "mean_token_accuracy": 0.8073959231376648,
      "num_tokens": 96004602.0,
      "step": 2865
    },
    {
      "epoch": 0.1713228271251194,
      "grad_norm": 0.6745275855064392,
      "learning_rate": 4.9844183817931985e-05,
      "loss": 0.6382,
      "mean_token_accuracy": 0.80477055311203,
      "num_tokens": 96166897.0,
      "step": 2870
    },
    {
      "epoch": 0.1716212989493792,
      "grad_norm": 0.6974903345108032,
      "learning_rate": 4.9842876519360635e-05,
      "loss": 0.5975,
      "mean_token_accuracy": 0.8148157000541687,
      "num_tokens": 96334577.0,
      "step": 2875
    },
    {
      "epoch": 0.17191977077363896,
      "grad_norm": 0.723353385925293,
      "learning_rate": 4.984156377876932e-05,
      "loss": 0.6336,
      "mean_token_accuracy": 0.8037397146224976,
      "num_tokens": 96502257.0,
      "step": 2880
    },
    {
      "epoch": 0.17221824259789875,
      "grad_norm": 0.6630491018295288,
      "learning_rate": 4.984024559647777e-05,
      "loss": 0.6047,
      "mean_token_accuracy": 0.813408088684082,
      "num_tokens": 96669937.0,
      "step": 2885
    },
    {
      "epoch": 0.17251671442215855,
      "grad_norm": 0.6247820258140564,
      "learning_rate": 4.983892197280705e-05,
      "loss": 0.6116,
      "mean_token_accuracy": 0.8108195066452026,
      "num_tokens": 96837617.0,
      "step": 2890
    },
    {
      "epoch": 0.17281518624641834,
      "grad_norm": 0.6820305585861206,
      "learning_rate": 4.9837592908079544e-05,
      "loss": 0.5859,
      "mean_token_accuracy": 0.8177740693092346,
      "num_tokens": 97005297.0,
      "step": 2895
    },
    {
      "epoch": 0.17311365807067813,
      "grad_norm": 0.6570559740066528,
      "learning_rate": 4.983625840261899e-05,
      "loss": 0.6351,
      "mean_token_accuracy": 0.8049326062202453,
      "num_tokens": 97172977.0,
      "step": 2900
    },
    {
      "epoch": 0.17341212989493793,
      "grad_norm": 0.7028390169143677,
      "learning_rate": 4.9834918456750417e-05,
      "loss": 0.6003,
      "mean_token_accuracy": 0.8141059160232544,
      "num_tokens": 97340657.0,
      "step": 2905
    },
    {
      "epoch": 0.17371060171919772,
      "grad_norm": 0.7124725580215454,
      "learning_rate": 4.98335730708002e-05,
      "loss": 0.6173,
      "mean_token_accuracy": 0.8082548022270203,
      "num_tokens": 97508337.0,
      "step": 2910
    },
    {
      "epoch": 0.1740090735434575,
      "grad_norm": 0.6809436082839966,
      "learning_rate": 4.983222224509604e-05,
      "loss": 0.6272,
      "mean_token_accuracy": 0.8079923629760742,
      "num_tokens": 97676017.0,
      "step": 2915
    },
    {
      "epoch": 0.17430754536771728,
      "grad_norm": 0.7032147645950317,
      "learning_rate": 4.983086597996694e-05,
      "loss": 0.6416,
      "mean_token_accuracy": 0.8048073410987854,
      "num_tokens": 97843697.0,
      "step": 2920
    },
    {
      "epoch": 0.17460601719197708,
      "grad_norm": 0.7811126708984375,
      "learning_rate": 4.982950427574325e-05,
      "loss": 0.6362,
      "mean_token_accuracy": 0.8042168617248535,
      "num_tokens": 98011377.0,
      "step": 2925
    },
    {
      "epoch": 0.17490448901623687,
      "grad_norm": 0.8064723014831543,
      "learning_rate": 4.982813713275664e-05,
      "loss": 0.593,
      "mean_token_accuracy": 0.8176905512809753,
      "num_tokens": 98179057.0,
      "step": 2930
    },
    {
      "epoch": 0.17520296084049666,
      "grad_norm": 0.7125324606895447,
      "learning_rate": 4.982676455134009e-05,
      "loss": 0.5986,
      "mean_token_accuracy": 0.8163604974746704,
      "num_tokens": 98346737.0,
      "step": 2935
    },
    {
      "epoch": 0.17550143266475646,
      "grad_norm": 0.6731593012809753,
      "learning_rate": 4.982538653182794e-05,
      "loss": 0.6518,
      "mean_token_accuracy": 0.8014255046844483,
      "num_tokens": 98514417.0,
      "step": 2940
    },
    {
      "epoch": 0.17579990448901622,
      "grad_norm": 0.6540849208831787,
      "learning_rate": 4.982400307455582e-05,
      "loss": 0.5967,
      "mean_token_accuracy": 0.8152034044265747,
      "num_tokens": 98682097.0,
      "step": 2945
    },
    {
      "epoch": 0.17609837631327602,
      "grad_norm": 0.6249152421951294,
      "learning_rate": 4.982261417986068e-05,
      "loss": 0.5809,
      "mean_token_accuracy": 0.8181856155395508,
      "num_tokens": 98849777.0,
      "step": 2950
    },
    {
      "epoch": 0.1763968481375358,
      "grad_norm": 0.6893585324287415,
      "learning_rate": 4.9821219848080844e-05,
      "loss": 0.584,
      "mean_token_accuracy": 0.8165036320686341,
      "num_tokens": 99017457.0,
      "step": 2955
    },
    {
      "epoch": 0.1766953199617956,
      "grad_norm": 0.6397733688354492,
      "learning_rate": 4.981982007955592e-05,
      "loss": 0.5348,
      "mean_token_accuracy": 0.8326851844787597,
      "num_tokens": 99185137.0,
      "step": 2960
    },
    {
      "epoch": 0.1769937917860554,
      "grad_norm": 0.6592968106269836,
      "learning_rate": 4.981841487462683e-05,
      "loss": 0.6095,
      "mean_token_accuracy": 0.8119587302207947,
      "num_tokens": 99352817.0,
      "step": 2965
    },
    {
      "epoch": 0.1772922636103152,
      "grad_norm": 0.6834815740585327,
      "learning_rate": 4.981700423363584e-05,
      "loss": 0.6473,
      "mean_token_accuracy": 0.8011988520622253,
      "num_tokens": 99520497.0,
      "step": 2970
    },
    {
      "epoch": 0.177590735434575,
      "grad_norm": 0.8949568867683411,
      "learning_rate": 4.9815588156926553e-05,
      "loss": 0.6138,
      "mean_token_accuracy": 0.8123881816864014,
      "num_tokens": 99688177.0,
      "step": 2975
    },
    {
      "epoch": 0.17788920725883475,
      "grad_norm": 0.6223493814468384,
      "learning_rate": 4.981416664484387e-05,
      "loss": 0.5837,
      "mean_token_accuracy": 0.8190564274787903,
      "num_tokens": 99855857.0,
      "step": 2980
    },
    {
      "epoch": 0.17818767908309455,
      "grad_norm": 0.6669052839279175,
      "learning_rate": 4.9812739697734024e-05,
      "loss": 0.6004,
      "mean_token_accuracy": 0.8143504738807679,
      "num_tokens": 100023537.0,
      "step": 2985
    },
    {
      "epoch": 0.17848615090735434,
      "grad_norm": 0.6864182949066162,
      "learning_rate": 4.9811307315944585e-05,
      "loss": 0.6337,
      "mean_token_accuracy": 0.8044136762619019,
      "num_tokens": 100191217.0,
      "step": 2990
    },
    {
      "epoch": 0.17878462273161413,
      "grad_norm": 0.6830667853355408,
      "learning_rate": 4.980986949982443e-05,
      "loss": 0.6015,
      "mean_token_accuracy": 0.8124478101730347,
      "num_tokens": 100358897.0,
      "step": 2995
    },
    {
      "epoch": 0.17908309455587393,
      "grad_norm": 0.6315885186195374,
      "learning_rate": 4.980842624972376e-05,
      "loss": 0.5787,
      "mean_token_accuracy": 0.820577347278595,
      "num_tokens": 100526577.0,
      "step": 3000
    },
    {
      "epoch": 0.17938156638013372,
      "grad_norm": 0.6943024396896362,
      "learning_rate": 4.9806977565994113e-05,
      "loss": 0.5997,
      "mean_token_accuracy": 0.8127636313438416,
      "num_tokens": 100688735.0,
      "step": 3005
    },
    {
      "epoch": 0.17968003820439352,
      "grad_norm": 0.7130090594291687,
      "learning_rate": 4.980552344898834e-05,
      "loss": 0.5806,
      "mean_token_accuracy": 0.8186448812484741,
      "num_tokens": 100856415.0,
      "step": 3010
    },
    {
      "epoch": 0.17997851002865328,
      "grad_norm": 0.7095332741737366,
      "learning_rate": 4.980406389906062e-05,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8244721531867981,
      "num_tokens": 101024095.0,
      "step": 3015
    },
    {
      "epoch": 0.18027698185291308,
      "grad_norm": 0.7395371198654175,
      "learning_rate": 4.980259891656644e-05,
      "loss": 0.6091,
      "mean_token_accuracy": 0.8115829706192017,
      "num_tokens": 101191775.0,
      "step": 3020
    },
    {
      "epoch": 0.18057545367717287,
      "grad_norm": 0.6890239715576172,
      "learning_rate": 4.980112850186265e-05,
      "loss": 0.5699,
      "mean_token_accuracy": 0.8231659293174743,
      "num_tokens": 101359455.0,
      "step": 3025
    },
    {
      "epoch": 0.18087392550143266,
      "grad_norm": 0.614687979221344,
      "learning_rate": 4.979965265530738e-05,
      "loss": 0.599,
      "mean_token_accuracy": 0.8175116300582885,
      "num_tokens": 101527135.0,
      "step": 3030
    },
    {
      "epoch": 0.18117239732569246,
      "grad_norm": 0.6636961698532104,
      "learning_rate": 4.979817137726009e-05,
      "loss": 0.5665,
      "mean_token_accuracy": 0.8247405409812927,
      "num_tokens": 101694815.0,
      "step": 3035
    },
    {
      "epoch": 0.18147086914995225,
      "grad_norm": 0.6953255534172058,
      "learning_rate": 4.97966846680816e-05,
      "loss": 0.5906,
      "mean_token_accuracy": 0.8145651936531066,
      "num_tokens": 101862495.0,
      "step": 3040
    },
    {
      "epoch": 0.18176934097421205,
      "grad_norm": 0.7299449443817139,
      "learning_rate": 4.979519252813399e-05,
      "loss": 0.607,
      "mean_token_accuracy": 0.8143743276596069,
      "num_tokens": 102030175.0,
      "step": 3045
    },
    {
      "epoch": 0.1820678127984718,
      "grad_norm": 0.644241988658905,
      "learning_rate": 4.979369495778074e-05,
      "loss": 0.5778,
      "mean_token_accuracy": 0.8193546414375306,
      "num_tokens": 102197855.0,
      "step": 3050
    },
    {
      "epoch": 0.1823662846227316,
      "grad_norm": 0.7699316740036011,
      "learning_rate": 4.979219195738658e-05,
      "loss": 0.5912,
      "mean_token_accuracy": 0.815036392211914,
      "num_tokens": 102365535.0,
      "step": 3055
    },
    {
      "epoch": 0.1826647564469914,
      "grad_norm": 0.7355851531028748,
      "learning_rate": 4.9790683527317594e-05,
      "loss": 0.604,
      "mean_token_accuracy": 0.8115889430046082,
      "num_tokens": 102533215.0,
      "step": 3060
    },
    {
      "epoch": 0.1829632282712512,
      "grad_norm": 0.7343382239341736,
      "learning_rate": 4.9789169667941196e-05,
      "loss": 0.6401,
      "mean_token_accuracy": 0.8039067268371582,
      "num_tokens": 102700895.0,
      "step": 3065
    },
    {
      "epoch": 0.183261700095511,
      "grad_norm": 0.7042121887207031,
      "learning_rate": 4.978765037962612e-05,
      "loss": 0.603,
      "mean_token_accuracy": 0.8136764883995056,
      "num_tokens": 102868575.0,
      "step": 3070
    },
    {
      "epoch": 0.18356017191977078,
      "grad_norm": 0.6647803783416748,
      "learning_rate": 4.978612566274241e-05,
      "loss": 0.6811,
      "mean_token_accuracy": 0.7922700762748718,
      "num_tokens": 103036255.0,
      "step": 3075
    },
    {
      "epoch": 0.18385864374403058,
      "grad_norm": 0.7692590355873108,
      "learning_rate": 4.9784595517661445e-05,
      "loss": 0.6412,
      "mean_token_accuracy": 0.8045568346977234,
      "num_tokens": 103203935.0,
      "step": 3080
    },
    {
      "epoch": 0.18415711556829034,
      "grad_norm": 0.6967044472694397,
      "learning_rate": 4.978305994475591e-05,
      "loss": 0.6041,
      "mean_token_accuracy": 0.8128474235534668,
      "num_tokens": 103371615.0,
      "step": 3085
    },
    {
      "epoch": 0.18445558739255014,
      "grad_norm": 0.6589059233665466,
      "learning_rate": 4.978151894439983e-05,
      "loss": 0.6225,
      "mean_token_accuracy": 0.8084814548492432,
      "num_tokens": 103539295.0,
      "step": 3090
    },
    {
      "epoch": 0.18475405921680993,
      "grad_norm": 0.6136396527290344,
      "learning_rate": 4.977997251696854e-05,
      "loss": 0.5857,
      "mean_token_accuracy": 0.8194679617881775,
      "num_tokens": 103706975.0,
      "step": 3095
    },
    {
      "epoch": 0.18505253104106972,
      "grad_norm": 0.6578737497329712,
      "learning_rate": 4.97784206628387e-05,
      "loss": 0.6558,
      "mean_token_accuracy": 0.7991649985313416,
      "num_tokens": 103874655.0,
      "step": 3100
    },
    {
      "epoch": 0.18535100286532952,
      "grad_norm": 0.6422058343887329,
      "learning_rate": 4.9776863382388294e-05,
      "loss": 0.6234,
      "mean_token_accuracy": 0.8081474423408508,
      "num_tokens": 104042335.0,
      "step": 3105
    },
    {
      "epoch": 0.1856494746895893,
      "grad_norm": 0.7381056547164917,
      "learning_rate": 4.977530067599664e-05,
      "loss": 0.6141,
      "mean_token_accuracy": 0.8085709333419799,
      "num_tokens": 104210015.0,
      "step": 3110
    },
    {
      "epoch": 0.1859479465138491,
      "grad_norm": 0.7627916932106018,
      "learning_rate": 4.977373254404434e-05,
      "loss": 0.6362,
      "mean_token_accuracy": 0.8067756175994873,
      "num_tokens": 104377695.0,
      "step": 3115
    },
    {
      "epoch": 0.18624641833810887,
      "grad_norm": 0.6562960147857666,
      "learning_rate": 4.977215898691335e-05,
      "loss": 0.6414,
      "mean_token_accuracy": 0.8028689026832581,
      "num_tokens": 104545375.0,
      "step": 3120
    },
    {
      "epoch": 0.18654489016236867,
      "grad_norm": 0.7071141004562378,
      "learning_rate": 4.977058000498696e-05,
      "loss": 0.6396,
      "mean_token_accuracy": 0.8021650910377502,
      "num_tokens": 104713055.0,
      "step": 3125
    },
    {
      "epoch": 0.18684336198662846,
      "grad_norm": 0.6516364216804504,
      "learning_rate": 4.976899559864972e-05,
      "loss": 0.6281,
      "mean_token_accuracy": 0.805702006816864,
      "num_tokens": 104880735.0,
      "step": 3130
    },
    {
      "epoch": 0.18714183381088825,
      "grad_norm": 0.6863721013069153,
      "learning_rate": 4.9767405768287575e-05,
      "loss": 0.641,
      "mean_token_accuracy": 0.8012048006057739,
      "num_tokens": 105048415.0,
      "step": 3135
    },
    {
      "epoch": 0.18744030563514805,
      "grad_norm": 0.6565121412277222,
      "learning_rate": 4.9765810514287734e-05,
      "loss": 0.643,
      "mean_token_accuracy": 0.8026303172111511,
      "num_tokens": 105216095.0,
      "step": 3140
    },
    {
      "epoch": 0.18773877745940784,
      "grad_norm": 0.6837484836578369,
      "learning_rate": 4.976420983703878e-05,
      "loss": 0.65,
      "mean_token_accuracy": 0.8001610398292541,
      "num_tokens": 105383775.0,
      "step": 3145
    },
    {
      "epoch": 0.18803724928366763,
      "grad_norm": 0.6963145732879639,
      "learning_rate": 4.976260373693056e-05,
      "loss": 0.6584,
      "mean_token_accuracy": 0.8001789212226867,
      "num_tokens": 105551455.0,
      "step": 3150
    },
    {
      "epoch": 0.1883357211079274,
      "grad_norm": 0.7384754419326782,
      "learning_rate": 4.976099221435428e-05,
      "loss": 0.6259,
      "mean_token_accuracy": 0.808672308921814,
      "num_tokens": 105719135.0,
      "step": 3155
    },
    {
      "epoch": 0.1886341929321872,
      "grad_norm": 0.7297983765602112,
      "learning_rate": 4.9759375269702455e-05,
      "loss": 0.637,
      "mean_token_accuracy": 0.8058511257171631,
      "num_tokens": 105886815.0,
      "step": 3160
    },
    {
      "epoch": 0.188932664756447,
      "grad_norm": 0.6801038384437561,
      "learning_rate": 4.975775290336893e-05,
      "loss": 0.649,
      "mean_token_accuracy": 0.8018847703933716,
      "num_tokens": 106054495.0,
      "step": 3165
    },
    {
      "epoch": 0.18923113658070678,
      "grad_norm": 0.6962680220603943,
      "learning_rate": 4.975612511574884e-05,
      "loss": 0.6377,
      "mean_token_accuracy": 0.8036621570587158,
      "num_tokens": 106222175.0,
      "step": 3170
    },
    {
      "epoch": 0.18952960840496658,
      "grad_norm": 0.6041184067726135,
      "learning_rate": 4.975449190723869e-05,
      "loss": 0.5969,
      "mean_token_accuracy": 0.8162590861320496,
      "num_tokens": 106389855.0,
      "step": 3175
    },
    {
      "epoch": 0.18982808022922637,
      "grad_norm": 0.7075260281562805,
      "learning_rate": 4.9752853278236256e-05,
      "loss": 0.6197,
      "mean_token_accuracy": 0.8071334838867188,
      "num_tokens": 106557535.0,
      "step": 3180
    },
    {
      "epoch": 0.19012655205348616,
      "grad_norm": 0.7273681163787842,
      "learning_rate": 4.975120922914067e-05,
      "loss": 0.6314,
      "mean_token_accuracy": 0.8052725791931152,
      "num_tokens": 106725215.0,
      "step": 3185
    },
    {
      "epoch": 0.19042502387774593,
      "grad_norm": 0.8485504388809204,
      "learning_rate": 4.974955976035236e-05,
      "loss": 0.6224,
      "mean_token_accuracy": 0.8077001094818115,
      "num_tokens": 106892895.0,
      "step": 3190
    },
    {
      "epoch": 0.19072349570200572,
      "grad_norm": 0.7034767270088196,
      "learning_rate": 4.97479048722731e-05,
      "loss": 0.5787,
      "mean_token_accuracy": 0.8210724115371704,
      "num_tokens": 107060575.0,
      "step": 3195
    },
    {
      "epoch": 0.19102196752626552,
      "grad_norm": 0.7601812481880188,
      "learning_rate": 4.9746244565305956e-05,
      "loss": 0.6551,
      "mean_token_accuracy": 0.8004055857658386,
      "num_tokens": 107228255.0,
      "step": 3200
    },
    {
      "epoch": 0.1913204393505253,
      "grad_norm": 0.5988834500312805,
      "learning_rate": 4.974457883985533e-05,
      "loss": 0.5668,
      "mean_token_accuracy": 0.8238458752632141,
      "num_tokens": 107395935.0,
      "step": 3205
    },
    {
      "epoch": 0.1916189111747851,
      "grad_norm": 0.6376057267189026,
      "learning_rate": 4.9742907696326935e-05,
      "loss": 0.5963,
      "mean_token_accuracy": 0.8158354997634888,
      "num_tokens": 107558903.0,
      "step": 3210
    },
    {
      "epoch": 0.1919173829990449,
      "grad_norm": 0.6359529495239258,
      "learning_rate": 4.974123113512782e-05,
      "loss": 0.574,
      "mean_token_accuracy": 0.8197363615036011,
      "num_tokens": 107726583.0,
      "step": 3215
    },
    {
      "epoch": 0.1922158548233047,
      "grad_norm": 0.7011116743087769,
      "learning_rate": 4.9739549156666334e-05,
      "loss": 0.6012,
      "mean_token_accuracy": 0.8162233233451843,
      "num_tokens": 107894263.0,
      "step": 3220
    },
    {
      "epoch": 0.19251432664756446,
      "grad_norm": 0.7204210162162781,
      "learning_rate": 4.973786176135215e-05,
      "loss": 0.6283,
      "mean_token_accuracy": 0.8073661088943481,
      "num_tokens": 108061943.0,
      "step": 3225
    },
    {
      "epoch": 0.19281279847182425,
      "grad_norm": 0.6594974994659424,
      "learning_rate": 4.9736168949596276e-05,
      "loss": 0.5804,
      "mean_token_accuracy": 0.819086229801178,
      "num_tokens": 108229623.0,
      "step": 3230
    },
    {
      "epoch": 0.19311127029608405,
      "grad_norm": 0.6742432713508606,
      "learning_rate": 4.973447072181102e-05,
      "loss": 0.5868,
      "mean_token_accuracy": 0.8188238143920898,
      "num_tokens": 108397303.0,
      "step": 3235
    },
    {
      "epoch": 0.19340974212034384,
      "grad_norm": 0.6995094418525696,
      "learning_rate": 4.973276707841001e-05,
      "loss": 0.6454,
      "mean_token_accuracy": 0.8005725860595703,
      "num_tokens": 108564983.0,
      "step": 3240
    },
    {
      "epoch": 0.19370821394460364,
      "grad_norm": 0.6854166984558105,
      "learning_rate": 4.9731058019808225e-05,
      "loss": 0.5931,
      "mean_token_accuracy": 0.816115927696228,
      "num_tokens": 108732663.0,
      "step": 3245
    },
    {
      "epoch": 0.19400668576886343,
      "grad_norm": 0.7256622314453125,
      "learning_rate": 4.972934354642191e-05,
      "loss": 0.5934,
      "mean_token_accuracy": 0.8162113785743713,
      "num_tokens": 108900343.0,
      "step": 3250
    },
    {
      "epoch": 0.1943051575931232,
      "grad_norm": 0.6493568420410156,
      "learning_rate": 4.9727623658668674e-05,
      "loss": 0.6268,
      "mean_token_accuracy": 0.8067100048065186,
      "num_tokens": 109068023.0,
      "step": 3255
    },
    {
      "epoch": 0.194603629417383,
      "grad_norm": 0.8499940633773804,
      "learning_rate": 4.972589835696742e-05,
      "loss": 0.6285,
      "mean_token_accuracy": 0.8055290460586548,
      "num_tokens": 109235703.0,
      "step": 3260
    },
    {
      "epoch": 0.19490210124164278,
      "grad_norm": 0.6785118579864502,
      "learning_rate": 4.9724167641738375e-05,
      "loss": 0.6376,
      "mean_token_accuracy": 0.8027734637260437,
      "num_tokens": 109403383.0,
      "step": 3265
    },
    {
      "epoch": 0.19520057306590258,
      "grad_norm": 0.6749463677406311,
      "learning_rate": 4.972243151340309e-05,
      "loss": 0.5759,
      "mean_token_accuracy": 0.8217225313186646,
      "num_tokens": 109571063.0,
      "step": 3270
    },
    {
      "epoch": 0.19549904489016237,
      "grad_norm": 0.6525623202323914,
      "learning_rate": 4.9720689972384435e-05,
      "loss": 0.6043,
      "mean_token_accuracy": 0.8143862724304199,
      "num_tokens": 109738743.0,
      "step": 3275
    },
    {
      "epoch": 0.19579751671442217,
      "grad_norm": 0.6807870864868164,
      "learning_rate": 4.971894301910658e-05,
      "loss": 0.6436,
      "mean_token_accuracy": 0.80174161195755,
      "num_tokens": 109906423.0,
      "step": 3280
    },
    {
      "epoch": 0.19609598853868196,
      "grad_norm": 0.6803872585296631,
      "learning_rate": 4.971719065399505e-05,
      "loss": 0.5847,
      "mean_token_accuracy": 0.8211976647377014,
      "num_tokens": 110074103.0,
      "step": 3285
    },
    {
      "epoch": 0.19639446036294173,
      "grad_norm": 0.6901847124099731,
      "learning_rate": 4.971543287747665e-05,
      "loss": 0.6033,
      "mean_token_accuracy": 0.8121905088424682,
      "num_tokens": 110233485.0,
      "step": 3290
    },
    {
      "epoch": 0.19669293218720152,
      "grad_norm": 0.7427123785018921,
      "learning_rate": 4.971366968997951e-05,
      "loss": 0.6064,
      "mean_token_accuracy": 0.8110998511314392,
      "num_tokens": 110401165.0,
      "step": 3295
    },
    {
      "epoch": 0.1969914040114613,
      "grad_norm": 0.7191442847251892,
      "learning_rate": 4.971190109193312e-05,
      "loss": 0.5735,
      "mean_token_accuracy": 0.8219611287117005,
      "num_tokens": 110568845.0,
      "step": 3300
    },
    {
      "epoch": 0.1972898758357211,
      "grad_norm": 0.66484534740448,
      "learning_rate": 4.971012708376822e-05,
      "loss": 0.6363,
      "mean_token_accuracy": 0.8046343803405762,
      "num_tokens": 110736525.0,
      "step": 3305
    },
    {
      "epoch": 0.1975883476599809,
      "grad_norm": 0.6512221693992615,
      "learning_rate": 4.970834766591692e-05,
      "loss": 0.5841,
      "mean_token_accuracy": 0.8190087080001831,
      "num_tokens": 110904205.0,
      "step": 3310
    },
    {
      "epoch": 0.1978868194842407,
      "grad_norm": 0.5992953181266785,
      "learning_rate": 4.9706562838812634e-05,
      "loss": 0.6738,
      "mean_token_accuracy": 0.7961708307266235,
      "num_tokens": 111071885.0,
      "step": 3315
    },
    {
      "epoch": 0.1981852913085005,
      "grad_norm": 0.6900875568389893,
      "learning_rate": 4.9704772602890075e-05,
      "loss": 0.6207,
      "mean_token_accuracy": 0.8099308133125305,
      "num_tokens": 111239565.0,
      "step": 3320
    },
    {
      "epoch": 0.19848376313276025,
      "grad_norm": 0.6812499165534973,
      "learning_rate": 4.97029769585853e-05,
      "loss": 0.6232,
      "mean_token_accuracy": 0.8077537894248963,
      "num_tokens": 111407245.0,
      "step": 3325
    },
    {
      "epoch": 0.19878223495702005,
      "grad_norm": 0.6209660172462463,
      "learning_rate": 4.9701175906335665e-05,
      "loss": 0.6258,
      "mean_token_accuracy": 0.806203031539917,
      "num_tokens": 111574925.0,
      "step": 3330
    },
    {
      "epoch": 0.19908070678127984,
      "grad_norm": 0.7421411275863647,
      "learning_rate": 4.969936944657985e-05,
      "loss": 0.6517,
      "mean_token_accuracy": 0.8002862930297852,
      "num_tokens": 111742605.0,
      "step": 3335
    },
    {
      "epoch": 0.19937917860553964,
      "grad_norm": 0.6757756471633911,
      "learning_rate": 4.969755757975785e-05,
      "loss": 0.6736,
      "mean_token_accuracy": 0.7945544481277466,
      "num_tokens": 111910285.0,
      "step": 3340
    },
    {
      "epoch": 0.19967765042979943,
      "grad_norm": 0.6517324447631836,
      "learning_rate": 4.969574030631098e-05,
      "loss": 0.6072,
      "mean_token_accuracy": 0.8123881578445434,
      "num_tokens": 112077965.0,
      "step": 3345
    },
    {
      "epoch": 0.19997612225405922,
      "grad_norm": 0.7713308930397034,
      "learning_rate": 4.969391762668188e-05,
      "loss": 0.5744,
      "mean_token_accuracy": 0.8211737990379333,
      "num_tokens": 112245645.0,
      "step": 3350
    },
    {
      "epoch": 0.20027459407831902,
      "grad_norm": 0.6571708917617798,
      "learning_rate": 4.9692089541314485e-05,
      "loss": 0.5806,
      "mean_token_accuracy": 0.8187880158424378,
      "num_tokens": 112413325.0,
      "step": 3355
    },
    {
      "epoch": 0.20057306590257878,
      "grad_norm": 0.5897857546806335,
      "learning_rate": 4.969025605065407e-05,
      "loss": 0.5863,
      "mean_token_accuracy": 0.8179649233818054,
      "num_tokens": 112581005.0,
      "step": 3360
    },
    {
      "epoch": 0.20087153772683858,
      "grad_norm": 0.6495668292045593,
      "learning_rate": 4.9688417155147207e-05,
      "loss": 0.5984,
      "mean_token_accuracy": 0.8132231712341309,
      "num_tokens": 112748685.0,
      "step": 3365
    },
    {
      "epoch": 0.20117000955109837,
      "grad_norm": 0.7834144830703735,
      "learning_rate": 4.9686572855241784e-05,
      "loss": 0.6144,
      "mean_token_accuracy": 0.8103304386138916,
      "num_tokens": 112916365.0,
      "step": 3370
    },
    {
      "epoch": 0.20146848137535817,
      "grad_norm": 0.7511405348777771,
      "learning_rate": 4.968472315138703e-05,
      "loss": 0.5892,
      "mean_token_accuracy": 0.8186700940132141,
      "num_tokens": 113079045.0,
      "step": 3375
    },
    {
      "epoch": 0.20176695319961796,
      "grad_norm": 0.6783428192138672,
      "learning_rate": 4.968286804403347e-05,
      "loss": 0.6151,
      "mean_token_accuracy": 0.8093164801597595,
      "num_tokens": 113246725.0,
      "step": 3380
    },
    {
      "epoch": 0.20206542502387775,
      "grad_norm": 0.6980903148651123,
      "learning_rate": 4.968100753363295e-05,
      "loss": 0.6271,
      "mean_token_accuracy": 0.8070857763290405,
      "num_tokens": 113414405.0,
      "step": 3385
    },
    {
      "epoch": 0.20236389684813755,
      "grad_norm": 0.6501882672309875,
      "learning_rate": 4.9679141620638634e-05,
      "loss": 0.6166,
      "mean_token_accuracy": 0.8123693943023682,
      "num_tokens": 113575308.0,
      "step": 3390
    },
    {
      "epoch": 0.20266236867239731,
      "grad_norm": 0.6438986659049988,
      "learning_rate": 4.9677270305504995e-05,
      "loss": 0.6031,
      "mean_token_accuracy": 0.8148276209831238,
      "num_tokens": 113742988.0,
      "step": 3395
    },
    {
      "epoch": 0.2029608404966571,
      "grad_norm": 0.7259078025817871,
      "learning_rate": 4.967539358868783e-05,
      "loss": 0.603,
      "mean_token_accuracy": 0.8132351279258728,
      "num_tokens": 113910668.0,
      "step": 3400
    },
    {
      "epoch": 0.2032593123209169,
      "grad_norm": 0.659095048904419,
      "learning_rate": 4.9673511470644236e-05,
      "loss": 0.6468,
      "mean_token_accuracy": 0.8014254927635193,
      "num_tokens": 114078348.0,
      "step": 3405
    },
    {
      "epoch": 0.2035577841451767,
      "grad_norm": 0.6263108849525452,
      "learning_rate": 4.9671623951832644e-05,
      "loss": 0.5949,
      "mean_token_accuracy": 0.8181438565254211,
      "num_tokens": 114246028.0,
      "step": 3410
    },
    {
      "epoch": 0.2038562559694365,
      "grad_norm": 0.8719925880432129,
      "learning_rate": 4.9669731032712814e-05,
      "loss": 0.6868,
      "mean_token_accuracy": 0.7903018116950988,
      "num_tokens": 114413708.0,
      "step": 3415
    },
    {
      "epoch": 0.20415472779369628,
      "grad_norm": 0.7731387615203857,
      "learning_rate": 4.9667832713745774e-05,
      "loss": 0.6717,
      "mean_token_accuracy": 0.7944709539413453,
      "num_tokens": 114581388.0,
      "step": 3420
    },
    {
      "epoch": 0.20445319961795608,
      "grad_norm": 0.6292768716812134,
      "learning_rate": 4.9665928995393904e-05,
      "loss": 0.5709,
      "mean_token_accuracy": 0.823332941532135,
      "num_tokens": 114749068.0,
      "step": 3425
    },
    {
      "epoch": 0.20475167144221584,
      "grad_norm": 0.6553537845611572,
      "learning_rate": 4.9664019878120896e-05,
      "loss": 0.6282,
      "mean_token_accuracy": 0.8055290460586548,
      "num_tokens": 114916748.0,
      "step": 3430
    },
    {
      "epoch": 0.20505014326647564,
      "grad_norm": 0.6622532606124878,
      "learning_rate": 4.966210536239175e-05,
      "loss": 0.6241,
      "mean_token_accuracy": 0.8065608859062194,
      "num_tokens": 115084428.0,
      "step": 3435
    },
    {
      "epoch": 0.20534861509073543,
      "grad_norm": 0.6557806730270386,
      "learning_rate": 4.9660185448672764e-05,
      "loss": 0.6005,
      "mean_token_accuracy": 0.8137540221214294,
      "num_tokens": 115252108.0,
      "step": 3440
    },
    {
      "epoch": 0.20564708691499523,
      "grad_norm": 0.6983974575996399,
      "learning_rate": 4.9658260137431586e-05,
      "loss": 0.6516,
      "mean_token_accuracy": 0.8017356514930725,
      "num_tokens": 115419788.0,
      "step": 3445
    },
    {
      "epoch": 0.20594555873925502,
      "grad_norm": 0.7253179550170898,
      "learning_rate": 4.965632942913717e-05,
      "loss": 0.6372,
      "mean_token_accuracy": 0.8035786628723145,
      "num_tokens": 115587468.0,
      "step": 3450
    },
    {
      "epoch": 0.2062440305635148,
      "grad_norm": 0.6325120329856873,
      "learning_rate": 4.965439332425975e-05,
      "loss": 0.622,
      "mean_token_accuracy": 0.8079983353614807,
      "num_tokens": 115755148.0,
      "step": 3455
    },
    {
      "epoch": 0.2065425023877746,
      "grad_norm": 0.6540398001670837,
      "learning_rate": 4.965245182327092e-05,
      "loss": 0.6245,
      "mean_token_accuracy": 0.8069963097572327,
      "num_tokens": 115922828.0,
      "step": 3460
    },
    {
      "epoch": 0.20684097421203437,
      "grad_norm": 0.6008408069610596,
      "learning_rate": 4.9650504926643554e-05,
      "loss": 0.5675,
      "mean_token_accuracy": 0.8241619944572449,
      "num_tokens": 116090508.0,
      "step": 3465
    },
    {
      "epoch": 0.20713944603629417,
      "grad_norm": 1.1994218826293945,
      "learning_rate": 4.9648552634851864e-05,
      "loss": 0.6094,
      "mean_token_accuracy": 0.8137122631072998,
      "num_tokens": 116258188.0,
      "step": 3470
    },
    {
      "epoch": 0.20743791786055396,
      "grad_norm": 0.6906866431236267,
      "learning_rate": 4.964659494837136e-05,
      "loss": 0.6335,
      "mean_token_accuracy": 0.8043421149253845,
      "num_tokens": 116425868.0,
      "step": 3475
    },
    {
      "epoch": 0.20773638968481375,
      "grad_norm": 0.6637681126594543,
      "learning_rate": 4.964463186767888e-05,
      "loss": 0.6357,
      "mean_token_accuracy": 0.804962420463562,
      "num_tokens": 116593548.0,
      "step": 3480
    },
    {
      "epoch": 0.20803486150907355,
      "grad_norm": 0.6916055083274841,
      "learning_rate": 4.964266339325257e-05,
      "loss": 0.6221,
      "mean_token_accuracy": 0.806465458869934,
      "num_tokens": 116761228.0,
      "step": 3485
    },
    {
      "epoch": 0.20833333333333334,
      "grad_norm": 0.6898741722106934,
      "learning_rate": 4.964068952557188e-05,
      "loss": 0.6408,
      "mean_token_accuracy": 0.8034593820571899,
      "num_tokens": 116928908.0,
      "step": 3490
    },
    {
      "epoch": 0.20863180515759314,
      "grad_norm": 0.6792036294937134,
      "learning_rate": 4.963871026511759e-05,
      "loss": 0.6599,
      "mean_token_accuracy": 0.7973160028457642,
      "num_tokens": 117096588.0,
      "step": 3495
    },
    {
      "epoch": 0.2089302769818529,
      "grad_norm": 0.6463459730148315,
      "learning_rate": 4.963672561237177e-05,
      "loss": 0.6135,
      "mean_token_accuracy": 0.8103841066360473,
      "num_tokens": 117264268.0,
      "step": 3500
    },
    {
      "epoch": 0.2092287488061127,
      "grad_norm": 0.6414973735809326,
      "learning_rate": 4.9634735567817843e-05,
      "loss": 0.6016,
      "mean_token_accuracy": 0.8125849962234497,
      "num_tokens": 117431948.0,
      "step": 3505
    },
    {
      "epoch": 0.2095272206303725,
      "grad_norm": 0.6723977327346802,
      "learning_rate": 4.9632740131940506e-05,
      "loss": 0.6232,
      "mean_token_accuracy": 0.8084635496139526,
      "num_tokens": 117599628.0,
      "step": 3510
    },
    {
      "epoch": 0.20982569245463228,
      "grad_norm": 0.7419140338897705,
      "learning_rate": 4.963073930522578e-05,
      "loss": 0.6242,
      "mean_token_accuracy": 0.8072468042373657,
      "num_tokens": 117767308.0,
      "step": 3515
    },
    {
      "epoch": 0.21012416427889208,
      "grad_norm": 0.6374903321266174,
      "learning_rate": 4.962873308816101e-05,
      "loss": 0.5918,
      "mean_token_accuracy": 0.8160920739173889,
      "num_tokens": 117934988.0,
      "step": 3520
    },
    {
      "epoch": 0.21042263610315187,
      "grad_norm": 0.6475487947463989,
      "learning_rate": 4.962672148123486e-05,
      "loss": 0.584,
      "mean_token_accuracy": 0.8184003353118896,
      "num_tokens": 118102668.0,
      "step": 3525
    },
    {
      "epoch": 0.21072110792741167,
      "grad_norm": 0.6644144654273987,
      "learning_rate": 4.962470448493726e-05,
      "loss": 0.5722,
      "mean_token_accuracy": 0.8227364778518677,
      "num_tokens": 118270348.0,
      "step": 3530
    },
    {
      "epoch": 0.21101957975167143,
      "grad_norm": 0.6092865467071533,
      "learning_rate": 4.962268209975952e-05,
      "loss": 0.5983,
      "mean_token_accuracy": 0.8152511119842529,
      "num_tokens": 118438028.0,
      "step": 3535
    },
    {
      "epoch": 0.21131805157593123,
      "grad_norm": 0.6891133785247803,
      "learning_rate": 4.9620654326194216e-05,
      "loss": 0.6472,
      "mean_token_accuracy": 0.8025468230247498,
      "num_tokens": 118605708.0,
      "step": 3540
    },
    {
      "epoch": 0.21161652340019102,
      "grad_norm": 0.7612316012382507,
      "learning_rate": 4.961862116473525e-05,
      "loss": 0.6036,
      "mean_token_accuracy": 0.8143266320228577,
      "num_tokens": 118773388.0,
      "step": 3545
    },
    {
      "epoch": 0.21191499522445081,
      "grad_norm": 0.6373063325881958,
      "learning_rate": 4.961658261587783e-05,
      "loss": 0.6244,
      "mean_token_accuracy": 0.8092806816101075,
      "num_tokens": 118941068.0,
      "step": 3550
    },
    {
      "epoch": 0.2122134670487106,
      "grad_norm": 0.7420687675476074,
      "learning_rate": 4.961453868011849e-05,
      "loss": 0.6076,
      "mean_token_accuracy": 0.8117857575416565,
      "num_tokens": 119108748.0,
      "step": 3555
    },
    {
      "epoch": 0.2125119388729704,
      "grad_norm": 0.6555312871932983,
      "learning_rate": 4.9612489357955076e-05,
      "loss": 0.5866,
      "mean_token_accuracy": 0.8182929873466491,
      "num_tokens": 119276428.0,
      "step": 3560
    },
    {
      "epoch": 0.21281041069723017,
      "grad_norm": 0.6671750545501709,
      "learning_rate": 4.961043464988673e-05,
      "loss": 0.6288,
      "mean_token_accuracy": 0.8062447786331177,
      "num_tokens": 119444108.0,
      "step": 3565
    },
    {
      "epoch": 0.21310888252148996,
      "grad_norm": 0.6754035353660583,
      "learning_rate": 4.96083745564139e-05,
      "loss": 0.5585,
      "mean_token_accuracy": 0.8251043796539307,
      "num_tokens": 119611788.0,
      "step": 3570
    },
    {
      "epoch": 0.21340735434574976,
      "grad_norm": 0.602658212184906,
      "learning_rate": 4.960630907803838e-05,
      "loss": 0.609,
      "mean_token_accuracy": 0.8133842349052429,
      "num_tokens": 119779468.0,
      "step": 3575
    },
    {
      "epoch": 0.21370582617000955,
      "grad_norm": 0.7140071392059326,
      "learning_rate": 4.960423821526325e-05,
      "loss": 0.6103,
      "mean_token_accuracy": 0.811177396774292,
      "num_tokens": 119947148.0,
      "step": 3580
    },
    {
      "epoch": 0.21400429799426934,
      "grad_norm": 0.6357149481773376,
      "learning_rate": 4.96021619685929e-05,
      "loss": 0.6191,
      "mean_token_accuracy": 0.8099486947059631,
      "num_tokens": 120114828.0,
      "step": 3585
    },
    {
      "epoch": 0.21430276981852914,
      "grad_norm": 0.6583658456802368,
      "learning_rate": 4.9600080338533046e-05,
      "loss": 0.5944,
      "mean_token_accuracy": 0.8169867515563964,
      "num_tokens": 120282508.0,
      "step": 3590
    },
    {
      "epoch": 0.21460124164278893,
      "grad_norm": 0.7762123346328735,
      "learning_rate": 4.959799332559071e-05,
      "loss": 0.6848,
      "mean_token_accuracy": 0.7904628276824951,
      "num_tokens": 120450188.0,
      "step": 3595
    },
    {
      "epoch": 0.2148997134670487,
      "grad_norm": 0.6695089340209961,
      "learning_rate": 4.9595900930274206e-05,
      "loss": 0.6312,
      "mean_token_accuracy": 0.8058749794960022,
      "num_tokens": 120617868.0,
      "step": 3600
    },
    {
      "epoch": 0.2151981852913085,
      "grad_norm": 0.645816445350647,
      "learning_rate": 4.95938031530932e-05,
      "loss": 0.6231,
      "mean_token_accuracy": 0.8087200284004211,
      "num_tokens": 120785548.0,
      "step": 3605
    },
    {
      "epoch": 0.21549665711556829,
      "grad_norm": 1.1988598108291626,
      "learning_rate": 4.959169999455862e-05,
      "loss": 0.6063,
      "mean_token_accuracy": 0.8146964073181152,
      "num_tokens": 120953228.0,
      "step": 3610
    },
    {
      "epoch": 0.21579512893982808,
      "grad_norm": 0.7384743094444275,
      "learning_rate": 4.958959145518275e-05,
      "loss": 0.618,
      "mean_token_accuracy": 0.8079864025115967,
      "num_tokens": 121120908.0,
      "step": 3615
    },
    {
      "epoch": 0.21609360076408787,
      "grad_norm": 0.7345996499061584,
      "learning_rate": 4.958747753547915e-05,
      "loss": 0.629,
      "mean_token_accuracy": 0.8074317097663879,
      "num_tokens": 121288588.0,
      "step": 3620
    },
    {
      "epoch": 0.21639207258834767,
      "grad_norm": 0.6977571249008179,
      "learning_rate": 4.9585358235962704e-05,
      "loss": 0.6473,
      "mean_token_accuracy": 0.800614333152771,
      "num_tokens": 121456268.0,
      "step": 3625
    },
    {
      "epoch": 0.21669054441260746,
      "grad_norm": 0.7001806497573853,
      "learning_rate": 4.9583233557149616e-05,
      "loss": 0.6342,
      "mean_token_accuracy": 0.8061672568321228,
      "num_tokens": 121623948.0,
      "step": 3630
    },
    {
      "epoch": 0.21698901623686723,
      "grad_norm": 0.6496277451515198,
      "learning_rate": 4.958110349955738e-05,
      "loss": 0.6127,
      "mean_token_accuracy": 0.8109745860099793,
      "num_tokens": 121791628.0,
      "step": 3635
    },
    {
      "epoch": 0.21728748806112702,
      "grad_norm": 0.7329492568969727,
      "learning_rate": 4.957896806370482e-05,
      "loss": 0.6363,
      "mean_token_accuracy": 0.8030239701271057,
      "num_tokens": 121959308.0,
      "step": 3640
    },
    {
      "epoch": 0.21758595988538681,
      "grad_norm": 0.6131376028060913,
      "learning_rate": 4.9576827250112055e-05,
      "loss": 0.5979,
      "mean_token_accuracy": 0.8146188855171204,
      "num_tokens": 122126988.0,
      "step": 3645
    },
    {
      "epoch": 0.2178844317096466,
      "grad_norm": 0.6058253049850464,
      "learning_rate": 4.9574681059300525e-05,
      "loss": 0.5701,
      "mean_token_accuracy": 0.822330904006958,
      "num_tokens": 122294668.0,
      "step": 3650
    },
    {
      "epoch": 0.2181829035339064,
      "grad_norm": 0.6051797270774841,
      "learning_rate": 4.9572529491792966e-05,
      "loss": 0.5564,
      "mean_token_accuracy": 0.8265835642814636,
      "num_tokens": 122462348.0,
      "step": 3655
    },
    {
      "epoch": 0.2184813753581662,
      "grad_norm": 0.6974071860313416,
      "learning_rate": 4.957037254811344e-05,
      "loss": 0.6101,
      "mean_token_accuracy": 0.8135273575782775,
      "num_tokens": 122630028.0,
      "step": 3660
    },
    {
      "epoch": 0.218779847182426,
      "grad_norm": 0.6367104053497314,
      "learning_rate": 4.956821022878731e-05,
      "loss": 0.6145,
      "mean_token_accuracy": 0.8118096113204956,
      "num_tokens": 122797708.0,
      "step": 3665
    },
    {
      "epoch": 0.21907831900668576,
      "grad_norm": 0.6408621668815613,
      "learning_rate": 4.956604253434124e-05,
      "loss": 0.5942,
      "mean_token_accuracy": 0.8145830750465393,
      "num_tokens": 122965388.0,
      "step": 3670
    },
    {
      "epoch": 0.21937679083094555,
      "grad_norm": 0.6505680680274963,
      "learning_rate": 4.956386946530323e-05,
      "loss": 0.5939,
      "mean_token_accuracy": 0.8148634195327759,
      "num_tokens": 123133068.0,
      "step": 3675
    },
    {
      "epoch": 0.21967526265520534,
      "grad_norm": 0.646426260471344,
      "learning_rate": 4.9561691022202555e-05,
      "loss": 0.6046,
      "mean_token_accuracy": 0.8117618918418884,
      "num_tokens": 123300748.0,
      "step": 3680
    },
    {
      "epoch": 0.21997373447946514,
      "grad_norm": 0.6544671654701233,
      "learning_rate": 4.955950720556981e-05,
      "loss": 0.6677,
      "mean_token_accuracy": 0.7957055926322937,
      "num_tokens": 123468428.0,
      "step": 3685
    },
    {
      "epoch": 0.22027220630372493,
      "grad_norm": 0.683693528175354,
      "learning_rate": 4.955731801593692e-05,
      "loss": 0.6198,
      "mean_token_accuracy": 0.8088333606719971,
      "num_tokens": 123636108.0,
      "step": 3690
    },
    {
      "epoch": 0.22057067812798473,
      "grad_norm": 0.6387953162193298,
      "learning_rate": 4.95551234538371e-05,
      "loss": 0.5699,
      "mean_token_accuracy": 0.8228319406509399,
      "num_tokens": 123803788.0,
      "step": 3695
    },
    {
      "epoch": 0.22086914995224452,
      "grad_norm": 0.6805548071861267,
      "learning_rate": 4.955292351980487e-05,
      "loss": 0.6391,
      "mean_token_accuracy": 0.8038768887519836,
      "num_tokens": 123971468.0,
      "step": 3700
    },
    {
      "epoch": 0.2211676217765043,
      "grad_norm": 0.5999314188957214,
      "learning_rate": 4.955071821437605e-05,
      "loss": 0.5867,
      "mean_token_accuracy": 0.8186746954917907,
      "num_tokens": 124139148.0,
      "step": 3705
    },
    {
      "epoch": 0.22146609360076408,
      "grad_norm": 0.7274695634841919,
      "learning_rate": 4.954850753808782e-05,
      "loss": 0.6324,
      "mean_token_accuracy": 0.8037397027015686,
      "num_tokens": 124306828.0,
      "step": 3710
    },
    {
      "epoch": 0.22176456542502387,
      "grad_norm": 0.6396534442901611,
      "learning_rate": 4.95462914914786e-05,
      "loss": 0.6739,
      "mean_token_accuracy": 0.7941429018974304,
      "num_tokens": 124474508.0,
      "step": 3715
    },
    {
      "epoch": 0.22206303724928367,
      "grad_norm": 0.6167986989021301,
      "learning_rate": 4.9544070075088166e-05,
      "loss": 0.6011,
      "mean_token_accuracy": 0.815537405014038,
      "num_tokens": 124642188.0,
      "step": 3720
    },
    {
      "epoch": 0.22236150907354346,
      "grad_norm": 0.5847646594047546,
      "learning_rate": 4.954184328945757e-05,
      "loss": 0.6379,
      "mean_token_accuracy": 0.802719795703888,
      "num_tokens": 124809868.0,
      "step": 3725
    },
    {
      "epoch": 0.22265998089780326,
      "grad_norm": 1.0940500497817993,
      "learning_rate": 4.9539611135129205e-05,
      "loss": 0.6194,
      "mean_token_accuracy": 0.8100858807563782,
      "num_tokens": 124977548.0,
      "step": 3730
    },
    {
      "epoch": 0.22295845272206305,
      "grad_norm": 0.7148606181144714,
      "learning_rate": 4.953737361264674e-05,
      "loss": 0.6337,
      "mean_token_accuracy": 0.8053322196006775,
      "num_tokens": 125145228.0,
      "step": 3735
    },
    {
      "epoch": 0.22325692454632282,
      "grad_norm": 0.7344508171081543,
      "learning_rate": 4.953513072255516e-05,
      "loss": 0.574,
      "mean_token_accuracy": 0.8218537449836731,
      "num_tokens": 125312908.0,
      "step": 3740
    },
    {
      "epoch": 0.2235553963705826,
      "grad_norm": 0.7283851504325867,
      "learning_rate": 4.9532882465400774e-05,
      "loss": 0.6111,
      "mean_token_accuracy": 0.8125074505805969,
      "num_tokens": 125480588.0,
      "step": 3745
    },
    {
      "epoch": 0.2238538681948424,
      "grad_norm": 0.7857192158699036,
      "learning_rate": 4.9530628841731185e-05,
      "loss": 0.6268,
      "mean_token_accuracy": 0.8052487254142762,
      "num_tokens": 125648268.0,
      "step": 3750
    },
    {
      "epoch": 0.2241523400191022,
      "grad_norm": 0.6330496668815613,
      "learning_rate": 4.952836985209531e-05,
      "loss": 0.6035,
      "mean_token_accuracy": 0.8128772377967834,
      "num_tokens": 125815948.0,
      "step": 3755
    },
    {
      "epoch": 0.224450811843362,
      "grad_norm": 0.6812946200370789,
      "learning_rate": 4.952610549704336e-05,
      "loss": 0.5655,
      "mean_token_accuracy": 0.823708713054657,
      "num_tokens": 125983628.0,
      "step": 3760
    },
    {
      "epoch": 0.22474928366762179,
      "grad_norm": 0.6766478419303894,
      "learning_rate": 4.9523835777126855e-05,
      "loss": 0.6546,
      "mean_token_accuracy": 0.7995705604553223,
      "num_tokens": 126151308.0,
      "step": 3765
    },
    {
      "epoch": 0.22504775549188158,
      "grad_norm": 0.8571595549583435,
      "learning_rate": 4.952156069289863e-05,
      "loss": 0.6042,
      "mean_token_accuracy": 0.8122211694717407,
      "num_tokens": 126318988.0,
      "step": 3770
    },
    {
      "epoch": 0.22534622731614135,
      "grad_norm": 0.605312705039978,
      "learning_rate": 4.9519280244912825e-05,
      "loss": 0.5674,
      "mean_token_accuracy": 0.8233567953109742,
      "num_tokens": 126486668.0,
      "step": 3775
    },
    {
      "epoch": 0.22564469914040114,
      "grad_norm": 0.6619325876235962,
      "learning_rate": 4.951699443372489e-05,
      "loss": 0.616,
      "mean_token_accuracy": 0.8100679874420166,
      "num_tokens": 126654348.0,
      "step": 3780
    },
    {
      "epoch": 0.22594317096466093,
      "grad_norm": 0.6574784517288208,
      "learning_rate": 4.951470325989158e-05,
      "loss": 0.5418,
      "mean_token_accuracy": 0.8305976390838623,
      "num_tokens": 126822028.0,
      "step": 3785
    },
    {
      "epoch": 0.22624164278892073,
      "grad_norm": 0.5957261919975281,
      "learning_rate": 4.9512406723970946e-05,
      "loss": 0.5998,
      "mean_token_accuracy": 0.8132172226905823,
      "num_tokens": 126989708.0,
      "step": 3790
    },
    {
      "epoch": 0.22654011461318052,
      "grad_norm": 0.8026285171508789,
      "learning_rate": 4.951010482652234e-05,
      "loss": 0.599,
      "mean_token_accuracy": 0.8172134160995483,
      "num_tokens": 127157388.0,
      "step": 3795
    },
    {
      "epoch": 0.22683858643744031,
      "grad_norm": 0.6566646099090576,
      "learning_rate": 4.9507797568106454e-05,
      "loss": 0.649,
      "mean_token_accuracy": 0.8028092503547668,
      "num_tokens": 127325068.0,
      "step": 3800
    },
    {
      "epoch": 0.2271370582617001,
      "grad_norm": 0.6030519604682922,
      "learning_rate": 4.950548494928525e-05,
      "loss": 0.5754,
      "mean_token_accuracy": 0.8199928522109985,
      "num_tokens": 127492748.0,
      "step": 3805
    },
    {
      "epoch": 0.22743553008595987,
      "grad_norm": 0.7191128730773926,
      "learning_rate": 4.950316697062201e-05,
      "loss": 0.6478,
      "mean_token_accuracy": 0.8025825977325439,
      "num_tokens": 127660428.0,
      "step": 3810
    },
    {
      "epoch": 0.22773400191021967,
      "grad_norm": 0.7019906044006348,
      "learning_rate": 4.950084363268133e-05,
      "loss": 0.6325,
      "mean_token_accuracy": 0.8053620457649231,
      "num_tokens": 127828108.0,
      "step": 3815
    },
    {
      "epoch": 0.22803247373447946,
      "grad_norm": 0.6270615458488464,
      "learning_rate": 4.949851493602909e-05,
      "loss": 0.6229,
      "mean_token_accuracy": 0.8078253626823425,
      "num_tokens": 127995788.0,
      "step": 3820
    },
    {
      "epoch": 0.22833094555873926,
      "grad_norm": 0.6706055402755737,
      "learning_rate": 4.9496180881232486e-05,
      "loss": 0.5751,
      "mean_token_accuracy": 0.8208576798439026,
      "num_tokens": 128163468.0,
      "step": 3825
    },
    {
      "epoch": 0.22862941738299905,
      "grad_norm": 0.6419352889060974,
      "learning_rate": 4.9493841468860034e-05,
      "loss": 0.6171,
      "mean_token_accuracy": 0.8114517450332641,
      "num_tokens": 128331148.0,
      "step": 3830
    },
    {
      "epoch": 0.22892788920725884,
      "grad_norm": 0.5844231843948364,
      "learning_rate": 4.949149669948152e-05,
      "loss": 0.6126,
      "mean_token_accuracy": 0.8111296653747558,
      "num_tokens": 128498828.0,
      "step": 3835
    },
    {
      "epoch": 0.22922636103151864,
      "grad_norm": 0.657459020614624,
      "learning_rate": 4.948914657366808e-05,
      "loss": 0.6175,
      "mean_token_accuracy": 0.8085530281066895,
      "num_tokens": 128666508.0,
      "step": 3840
    },
    {
      "epoch": 0.2295248328557784,
      "grad_norm": 0.7376065850257874,
      "learning_rate": 4.948679109199212e-05,
      "loss": 0.64,
      "mean_token_accuracy": 0.8036323666572571,
      "num_tokens": 128834188.0,
      "step": 3845
    },
    {
      "epoch": 0.2298233046800382,
      "grad_norm": 0.7029840350151062,
      "learning_rate": 4.948443025502735e-05,
      "loss": 0.5621,
      "mean_token_accuracy": 0.8264105796813965,
      "num_tokens": 129001868.0,
      "step": 3850
    },
    {
      "epoch": 0.230121776504298,
      "grad_norm": 0.6539315581321716,
      "learning_rate": 4.948206406334881e-05,
      "loss": 0.6402,
      "mean_token_accuracy": 0.8021770238876342,
      "num_tokens": 129169548.0,
      "step": 3855
    },
    {
      "epoch": 0.2304202483285578,
      "grad_norm": 0.6318440437316895,
      "learning_rate": 4.9479692517532826e-05,
      "loss": 0.6407,
      "mean_token_accuracy": 0.800864839553833,
      "num_tokens": 129337228.0,
      "step": 3860
    },
    {
      "epoch": 0.23071872015281758,
      "grad_norm": 0.6926705837249756,
      "learning_rate": 4.9477315618157024e-05,
      "loss": 0.6569,
      "mean_token_accuracy": 0.7985446691513062,
      "num_tokens": 129504908.0,
      "step": 3865
    },
    {
      "epoch": 0.23101719197707737,
      "grad_norm": 0.6548948884010315,
      "learning_rate": 4.947493336580036e-05,
      "loss": 0.5607,
      "mean_token_accuracy": 0.8236311554908753,
      "num_tokens": 129672588.0,
      "step": 3870
    },
    {
      "epoch": 0.23131566380133714,
      "grad_norm": 0.7093825340270996,
      "learning_rate": 4.947254576104306e-05,
      "loss": 0.6198,
      "mean_token_accuracy": 0.8086961507797241,
      "num_tokens": 129840268.0,
      "step": 3875
    },
    {
      "epoch": 0.23161413562559693,
      "grad_norm": 0.7560344338417053,
      "learning_rate": 4.947015280446667e-05,
      "loss": 0.6544,
      "mean_token_accuracy": 0.7972981214523316,
      "num_tokens": 130007948.0,
      "step": 3880
    },
    {
      "epoch": 0.23191260744985673,
      "grad_norm": 0.6496684551239014,
      "learning_rate": 4.946775449665404e-05,
      "loss": 0.6367,
      "mean_token_accuracy": 0.8032744884490967,
      "num_tokens": 130175628.0,
      "step": 3885
    },
    {
      "epoch": 0.23221107927411652,
      "grad_norm": 0.6434205770492554,
      "learning_rate": 4.9465350838189326e-05,
      "loss": 0.6565,
      "mean_token_accuracy": 0.7995884537696838,
      "num_tokens": 130343308.0,
      "step": 3890
    },
    {
      "epoch": 0.23250955109837632,
      "grad_norm": 0.6018290519714355,
      "learning_rate": 4.9462941829657987e-05,
      "loss": 0.6235,
      "mean_token_accuracy": 0.8085709333419799,
      "num_tokens": 130510988.0,
      "step": 3895
    },
    {
      "epoch": 0.2328080229226361,
      "grad_norm": 0.573374330997467,
      "learning_rate": 4.9460527471646775e-05,
      "loss": 0.6335,
      "mean_token_accuracy": 0.8070976972579956,
      "num_tokens": 130678668.0,
      "step": 3900
    },
    {
      "epoch": 0.2331064947468959,
      "grad_norm": 0.6306058764457703,
      "learning_rate": 4.945810776474375e-05,
      "loss": 0.5795,
      "mean_token_accuracy": 0.8211857318878174,
      "num_tokens": 130846348.0,
      "step": 3905
    },
    {
      "epoch": 0.23340496657115567,
      "grad_norm": 0.652961790561676,
      "learning_rate": 4.945568270953827e-05,
      "loss": 0.6291,
      "mean_token_accuracy": 0.8067636847496032,
      "num_tokens": 131014028.0,
      "step": 3910
    },
    {
      "epoch": 0.23370343839541546,
      "grad_norm": 0.7349520325660706,
      "learning_rate": 4.9453252306621026e-05,
      "loss": 0.6554,
      "mean_token_accuracy": 0.7977335095405579,
      "num_tokens": 131181708.0,
      "step": 3915
    },
    {
      "epoch": 0.23400191021967526,
      "grad_norm": 0.6300112009048462,
      "learning_rate": 4.9450816556583955e-05,
      "loss": 0.5391,
      "mean_token_accuracy": 0.831003212928772,
      "num_tokens": 131349388.0,
      "step": 3920
    },
    {
      "epoch": 0.23430038204393505,
      "grad_norm": 0.6644827127456665,
      "learning_rate": 4.944837546002036e-05,
      "loss": 0.6238,
      "mean_token_accuracy": 0.8076285362243653,
      "num_tokens": 131517068.0,
      "step": 3925
    },
    {
      "epoch": 0.23459885386819485,
      "grad_norm": 0.5999575257301331,
      "learning_rate": 4.944592901752479e-05,
      "loss": 0.5803,
      "mean_token_accuracy": 0.8205177187919617,
      "num_tokens": 131684748.0,
      "step": 3930
    },
    {
      "epoch": 0.23489732569245464,
      "grad_norm": 0.5984506607055664,
      "learning_rate": 4.944347722969314e-05,
      "loss": 0.5472,
      "mean_token_accuracy": 0.829023027420044,
      "num_tokens": 131852428.0,
      "step": 3935
    },
    {
      "epoch": 0.23519579751671443,
      "grad_norm": 0.6539884805679321,
      "learning_rate": 4.9441020097122584e-05,
      "loss": 0.6394,
      "mean_token_accuracy": 0.8013181447982788,
      "num_tokens": 132020108.0,
      "step": 3940
    },
    {
      "epoch": 0.2354942693409742,
      "grad_norm": 0.6690188050270081,
      "learning_rate": 4.943855762041159e-05,
      "loss": 0.5877,
      "mean_token_accuracy": 0.8185554146766663,
      "num_tokens": 132187788.0,
      "step": 3945
    },
    {
      "epoch": 0.235792741165234,
      "grad_norm": 0.7017703652381897,
      "learning_rate": 4.9436089800159956e-05,
      "loss": 0.6142,
      "mean_token_accuracy": 0.8101097464561462,
      "num_tokens": 132355468.0,
      "step": 3950
    },
    {
      "epoch": 0.2360912129894938,
      "grad_norm": 0.6887523531913757,
      "learning_rate": 4.943361663696875e-05,
      "loss": 0.6289,
      "mean_token_accuracy": 0.8065489649772644,
      "num_tokens": 132523148.0,
      "step": 3955
    },
    {
      "epoch": 0.23638968481375358,
      "grad_norm": 0.6603471636772156,
      "learning_rate": 4.9431138131440376e-05,
      "loss": 0.6517,
      "mean_token_accuracy": 0.8001968145370484,
      "num_tokens": 132690828.0,
      "step": 3960
    },
    {
      "epoch": 0.23668815663801338,
      "grad_norm": 0.602262020111084,
      "learning_rate": 4.94286542841785e-05,
      "loss": 0.6092,
      "mean_token_accuracy": 0.812429916858673,
      "num_tokens": 132858508.0,
      "step": 3965
    },
    {
      "epoch": 0.23698662846227317,
      "grad_norm": 0.5906766057014465,
      "learning_rate": 4.9426165095788105e-05,
      "loss": 0.5855,
      "mean_token_accuracy": 0.8167839646339417,
      "num_tokens": 133026188.0,
      "step": 3970
    },
    {
      "epoch": 0.23728510028653296,
      "grad_norm": 0.7193766832351685,
      "learning_rate": 4.942367056687551e-05,
      "loss": 0.6409,
      "mean_token_accuracy": 0.8023977041244507,
      "num_tokens": 133193868.0,
      "step": 3975
    },
    {
      "epoch": 0.23758357211079273,
      "grad_norm": 0.6253636479377747,
      "learning_rate": 4.942117069804827e-05,
      "loss": 0.6299,
      "mean_token_accuracy": 0.8059704303741455,
      "num_tokens": 133361548.0,
      "step": 3980
    },
    {
      "epoch": 0.23788204393505252,
      "grad_norm": 0.6552958488464355,
      "learning_rate": 4.941866548991529e-05,
      "loss": 0.6087,
      "mean_token_accuracy": 0.8109447598457337,
      "num_tokens": 133529228.0,
      "step": 3985
    },
    {
      "epoch": 0.23818051575931232,
      "grad_norm": 0.6816372871398926,
      "learning_rate": 4.9416154943086765e-05,
      "loss": 0.6263,
      "mean_token_accuracy": 0.8083025217056274,
      "num_tokens": 133696908.0,
      "step": 3990
    },
    {
      "epoch": 0.2384789875835721,
      "grad_norm": 0.613618791103363,
      "learning_rate": 4.941363905817417e-05,
      "loss": 0.6176,
      "mean_token_accuracy": 0.811302638053894,
      "num_tokens": 133864588.0,
      "step": 3995
    },
    {
      "epoch": 0.2387774594078319,
      "grad_norm": 0.6847855448722839,
      "learning_rate": 4.9411117835790304e-05,
      "loss": 0.5766,
      "mean_token_accuracy": 0.8193307876586914,
      "num_tokens": 134032268.0,
      "step": 4000
    },
    {
      "epoch": 0.2390759312320917,
      "grad_norm": 0.682997465133667,
      "learning_rate": 4.9408591276549266e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.8106763601303101,
      "num_tokens": 134199948.0,
      "step": 4005
    },
    {
      "epoch": 0.2393744030563515,
      "grad_norm": 0.7522417306900024,
      "learning_rate": 4.940605938106643e-05,
      "loss": 0.5928,
      "mean_token_accuracy": 0.8140880346298218,
      "num_tokens": 134367628.0,
      "step": 4010
    },
    {
      "epoch": 0.23967287488061126,
      "grad_norm": 0.7058261632919312,
      "learning_rate": 4.9403522149958485e-05,
      "loss": 0.6209,
      "mean_token_accuracy": 0.8093761205673218,
      "num_tokens": 134535308.0,
      "step": 4015
    },
    {
      "epoch": 0.23997134670487105,
      "grad_norm": 0.5665675401687622,
      "learning_rate": 4.940097958384343e-05,
      "loss": 0.5954,
      "mean_token_accuracy": 0.8134498476982117,
      "num_tokens": 134702988.0,
      "step": 4020
    },
    {
      "epoch": 0.24026981852913085,
      "grad_norm": 0.6552174091339111,
      "learning_rate": 4.939843168334056e-05,
      "loss": 0.6219,
      "mean_token_accuracy": 0.8063819646835327,
      "num_tokens": 134870668.0,
      "step": 4025
    },
    {
      "epoch": 0.24056829035339064,
      "grad_norm": 0.7946946620941162,
      "learning_rate": 4.939587844907045e-05,
      "loss": 0.592,
      "mean_token_accuracy": 0.8169211506843567,
      "num_tokens": 135038348.0,
      "step": 4030
    },
    {
      "epoch": 0.24086676217765043,
      "grad_norm": 0.6194802522659302,
      "learning_rate": 4.9393319881655e-05,
      "loss": 0.6307,
      "mean_token_accuracy": 0.8075032711029053,
      "num_tokens": 135206028.0,
      "step": 4035
    },
    {
      "epoch": 0.24116523400191023,
      "grad_norm": 0.6646754741668701,
      "learning_rate": 4.939075598171737e-05,
      "loss": 0.6366,
      "mean_token_accuracy": 0.8035011291503906,
      "num_tokens": 135373708.0,
      "step": 4040
    },
    {
      "epoch": 0.24146370582617002,
      "grad_norm": 0.6231975555419922,
      "learning_rate": 4.9388186749882074e-05,
      "loss": 0.6283,
      "mean_token_accuracy": 0.8051354050636291,
      "num_tokens": 135541388.0,
      "step": 4045
    },
    {
      "epoch": 0.2417621776504298,
      "grad_norm": 0.6142847537994385,
      "learning_rate": 4.9385612186774895e-05,
      "loss": 0.5823,
      "mean_token_accuracy": 0.8176905632019043,
      "num_tokens": 135709068.0,
      "step": 4050
    },
    {
      "epoch": 0.24206064947468958,
      "grad_norm": 0.6034107804298401,
      "learning_rate": 4.938303229302289e-05,
      "loss": 0.5621,
      "mean_token_accuracy": 0.8236550092697144,
      "num_tokens": 135876748.0,
      "step": 4055
    },
    {
      "epoch": 0.24235912129894938,
      "grad_norm": 0.6483427882194519,
      "learning_rate": 4.9380447069254465e-05,
      "loss": 0.5761,
      "mean_token_accuracy": 0.818370521068573,
      "num_tokens": 136044428.0,
      "step": 4060
    },
    {
      "epoch": 0.24265759312320917,
      "grad_norm": 0.6459857821464539,
      "learning_rate": 4.937785651609929e-05,
      "loss": 0.6138,
      "mean_token_accuracy": 0.8086663603782653,
      "num_tokens": 136212108.0,
      "step": 4065
    },
    {
      "epoch": 0.24295606494746896,
      "grad_norm": 0.7154399156570435,
      "learning_rate": 4.937526063418834e-05,
      "loss": 0.5977,
      "mean_token_accuracy": 0.813998568058014,
      "num_tokens": 136379788.0,
      "step": 4070
    },
    {
      "epoch": 0.24325453677172876,
      "grad_norm": 0.7276933789253235,
      "learning_rate": 4.937265942415388e-05,
      "loss": 0.6303,
      "mean_token_accuracy": 0.8070678591728211,
      "num_tokens": 136547468.0,
      "step": 4075
    },
    {
      "epoch": 0.24355300859598855,
      "grad_norm": 0.6441953182220459,
      "learning_rate": 4.937005288662951e-05,
      "loss": 0.6076,
      "mean_token_accuracy": 0.8098473072052002,
      "num_tokens": 136715148.0,
      "step": 4080
    },
    {
      "epoch": 0.24385148042024832,
      "grad_norm": 0.6848450303077698,
      "learning_rate": 4.9367441022250075e-05,
      "loss": 0.6334,
      "mean_token_accuracy": 0.8028569698333741,
      "num_tokens": 136882828.0,
      "step": 4085
    },
    {
      "epoch": 0.2441499522445081,
      "grad_norm": 0.69948810338974,
      "learning_rate": 4.936482383165175e-05,
      "loss": 0.6412,
      "mean_token_accuracy": 0.8035249829292297,
      "num_tokens": 137050508.0,
      "step": 4090
    },
    {
      "epoch": 0.2444484240687679,
      "grad_norm": 0.6532810926437378,
      "learning_rate": 4.936220131547201e-05,
      "loss": 0.6337,
      "mean_token_accuracy": 0.8033341407775879,
      "num_tokens": 137218188.0,
      "step": 4095
    },
    {
      "epoch": 0.2447468958930277,
      "grad_norm": 0.6094919443130493,
      "learning_rate": 4.9359573474349605e-05,
      "loss": 0.5844,
      "mean_token_accuracy": 0.8183228015899658,
      "num_tokens": 137385868.0,
      "step": 4100
    },
    {
      "epoch": 0.2450453677172875,
      "grad_norm": 0.6443773508071899,
      "learning_rate": 4.93569403089246e-05,
      "loss": 0.6021,
      "mean_token_accuracy": 0.8111415982246399,
      "num_tokens": 137553548.0,
      "step": 4105
    },
    {
      "epoch": 0.2453438395415473,
      "grad_norm": 0.6929174065589905,
      "learning_rate": 4.9354301819838364e-05,
      "loss": 0.605,
      "mean_token_accuracy": 0.812895143032074,
      "num_tokens": 137721228.0,
      "step": 4110
    },
    {
      "epoch": 0.24564231136580708,
      "grad_norm": 0.6365860104560852,
      "learning_rate": 4.935165800773352e-05,
      "loss": 0.6466,
      "mean_token_accuracy": 0.7997017621994018,
      "num_tokens": 137888908.0,
      "step": 4115
    },
    {
      "epoch": 0.24594078319006685,
      "grad_norm": 0.636452317237854,
      "learning_rate": 4.934900887325404e-05,
      "loss": 0.6349,
      "mean_token_accuracy": 0.8062028527259827,
      "num_tokens": 138049187.0,
      "step": 4120
    },
    {
      "epoch": 0.24623925501432664,
      "grad_norm": 0.6648581624031067,
      "learning_rate": 4.934635441704515e-05,
      "loss": 0.5835,
      "mean_token_accuracy": 0.8195753335952759,
      "num_tokens": 138216867.0,
      "step": 4125
    },
    {
      "epoch": 0.24653772683858644,
      "grad_norm": 0.6769698858261108,
      "learning_rate": 4.934369463975341e-05,
      "loss": 0.5938,
      "mean_token_accuracy": 0.8174996972084045,
      "num_tokens": 138384547.0,
      "step": 4130
    },
    {
      "epoch": 0.24683619866284623,
      "grad_norm": 0.5818044543266296,
      "learning_rate": 4.9341029542026656e-05,
      "loss": 0.6322,
      "mean_token_accuracy": 0.8060300588607788,
      "num_tokens": 138552227.0,
      "step": 4135
    },
    {
      "epoch": 0.24713467048710602,
      "grad_norm": 1.1713244915008545,
      "learning_rate": 4.9338359124514016e-05,
      "loss": 0.685,
      "mean_token_accuracy": 0.7939281821250915,
      "num_tokens": 138719907.0,
      "step": 4140
    },
    {
      "epoch": 0.24743314231136582,
      "grad_norm": 0.8352034091949463,
      "learning_rate": 4.9335683387865924e-05,
      "loss": 0.6586,
      "mean_token_accuracy": 0.7974352955818176,
      "num_tokens": 138887587.0,
      "step": 4145
    },
    {
      "epoch": 0.2477316141356256,
      "grad_norm": 0.6547311544418335,
      "learning_rate": 4.93330023327341e-05,
      "loss": 0.5677,
      "mean_token_accuracy": 0.8241619825363159,
      "num_tokens": 139055267.0,
      "step": 4150
    },
    {
      "epoch": 0.24803008595988538,
      "grad_norm": 0.6543726921081543,
      "learning_rate": 4.933031595977157e-05,
      "loss": 0.5974,
      "mean_token_accuracy": 0.8156030058860779,
      "num_tokens": 139222947.0,
      "step": 4155
    },
    {
      "epoch": 0.24832855778414517,
      "grad_norm": 0.6652186512947083,
      "learning_rate": 4.932762426963264e-05,
      "loss": 0.6245,
      "mean_token_accuracy": 0.8084337472915649,
      "num_tokens": 139390627.0,
      "step": 4160
    },
    {
      "epoch": 0.24862702960840496,
      "grad_norm": 0.7863958477973938,
      "learning_rate": 4.932492726297293e-05,
      "loss": 0.64,
      "mean_token_accuracy": 0.8058869123458863,
      "num_tokens": 139558307.0,
      "step": 4165
    },
    {
      "epoch": 0.24892550143266476,
      "grad_norm": 0.6341575384140015,
      "learning_rate": 4.9322224940449344e-05,
      "loss": 0.6417,
      "mean_token_accuracy": 0.8025945425033569,
      "num_tokens": 139725987.0,
      "step": 4170
    },
    {
      "epoch": 0.24922397325692455,
      "grad_norm": 0.6730334758758545,
      "learning_rate": 4.931951730272008e-05,
      "loss": 0.6251,
      "mean_token_accuracy": 0.8078849911689758,
      "num_tokens": 139893667.0,
      "step": 4175
    },
    {
      "epoch": 0.24952244508118435,
      "grad_norm": 0.6012901663780212,
      "learning_rate": 4.931680435044464e-05,
      "loss": 0.5785,
      "mean_token_accuracy": 0.8199510931968689,
      "num_tokens": 140061347.0,
      "step": 4180
    },
    {
      "epoch": 0.2498209169054441,
      "grad_norm": 0.6753756403923035,
      "learning_rate": 4.931408608428381e-05,
      "loss": 0.6032,
      "mean_token_accuracy": 0.8137811779975891,
      "num_tokens": 140224430.0,
      "step": 4185
    },
    {
      "epoch": 0.2501193887297039,
      "grad_norm": 0.8917942643165588,
      "learning_rate": 4.931136250489966e-05,
      "loss": 0.581,
      "mean_token_accuracy": 0.8197602391242981,
      "num_tokens": 140392110.0,
      "step": 4190
    },
    {
      "epoch": 0.2504178605539637,
      "grad_norm": 0.656954288482666,
      "learning_rate": 4.930863361295559e-05,
      "loss": 0.629,
      "mean_token_accuracy": 0.8068591117858886,
      "num_tokens": 140559790.0,
      "step": 4195
    },
    {
      "epoch": 0.2507163323782235,
      "grad_norm": 0.7210436463356018,
      "learning_rate": 4.930589940911626e-05,
      "loss": 0.5872,
      "mean_token_accuracy": 0.8177144169807434,
      "num_tokens": 140727470.0,
      "step": 4200
    },
    {
      "epoch": 0.2510148042024833,
      "grad_norm": 0.6364195346832275,
      "learning_rate": 4.930315989404763e-05,
      "loss": 0.5738,
      "mean_token_accuracy": 0.8197661995887756,
      "num_tokens": 140895150.0,
      "step": 4205
    },
    {
      "epoch": 0.2513132760267431,
      "grad_norm": 0.6811306476593018,
      "learning_rate": 4.930041506841698e-05,
      "loss": 0.586,
      "mean_token_accuracy": 0.8177382707595825,
      "num_tokens": 141062830.0,
      "step": 4210
    },
    {
      "epoch": 0.2516117478510029,
      "grad_norm": 0.6453559398651123,
      "learning_rate": 4.9297664932892844e-05,
      "loss": 0.6279,
      "mean_token_accuracy": 0.805588698387146,
      "num_tokens": 141230510.0,
      "step": 4215
    },
    {
      "epoch": 0.25191021967526267,
      "grad_norm": 0.6769728064537048,
      "learning_rate": 4.929490948814508e-05,
      "loss": 0.5905,
      "mean_token_accuracy": 0.8151079773902893,
      "num_tokens": 141398190.0,
      "step": 4220
    },
    {
      "epoch": 0.25220869149952246,
      "grad_norm": 0.6134464740753174,
      "learning_rate": 4.9292148734844816e-05,
      "loss": 0.5882,
      "mean_token_accuracy": 0.8175219058990478,
      "num_tokens": 141559577.0,
      "step": 4225
    },
    {
      "epoch": 0.25250716332378226,
      "grad_norm": 0.6419605612754822,
      "learning_rate": 4.92893826736645e-05,
      "loss": 0.6042,
      "mean_token_accuracy": 0.8128474235534668,
      "num_tokens": 141727257.0,
      "step": 4230
    },
    {
      "epoch": 0.25280563514804205,
      "grad_norm": 0.6950979828834534,
      "learning_rate": 4.9286611305277844e-05,
      "loss": 0.6199,
      "mean_token_accuracy": 0.8077239632606507,
      "num_tokens": 141894937.0,
      "step": 4235
    },
    {
      "epoch": 0.2531041069723018,
      "grad_norm": 0.6773786544799805,
      "learning_rate": 4.9283834630359856e-05,
      "loss": 0.5812,
      "mean_token_accuracy": 0.8211857318878174,
      "num_tokens": 142062617.0,
      "step": 4240
    },
    {
      "epoch": 0.2534025787965616,
      "grad_norm": 0.7095805406570435,
      "learning_rate": 4.928105264958687e-05,
      "loss": 0.667,
      "mean_token_accuracy": 0.7949660062789917,
      "num_tokens": 142230297.0,
      "step": 4245
    },
    {
      "epoch": 0.2537010506208214,
      "grad_norm": 0.7544382214546204,
      "learning_rate": 4.9278265363636485e-05,
      "loss": 0.6297,
      "mean_token_accuracy": 0.8055350065231324,
      "num_tokens": 142397977.0,
      "step": 4250
    },
    {
      "epoch": 0.25399952244508117,
      "grad_norm": 0.6471272706985474,
      "learning_rate": 4.927547277318759e-05,
      "loss": 0.581,
      "mean_token_accuracy": 0.8175414562225342,
      "num_tokens": 142565657.0,
      "step": 4255
    },
    {
      "epoch": 0.25429799426934097,
      "grad_norm": 0.7374337911605835,
      "learning_rate": 4.927267487892036e-05,
      "loss": 0.6279,
      "mean_token_accuracy": 0.8062562942504883,
      "num_tokens": 142727426.0,
      "step": 4260
    },
    {
      "epoch": 0.25459646609360076,
      "grad_norm": 0.7466955780982971,
      "learning_rate": 4.92698716815163e-05,
      "loss": 0.6534,
      "mean_token_accuracy": 0.7972504019737243,
      "num_tokens": 142895106.0,
      "step": 4265
    },
    {
      "epoch": 0.25489493791786055,
      "grad_norm": 0.6746529936790466,
      "learning_rate": 4.926706318165815e-05,
      "loss": 0.636,
      "mean_token_accuracy": 0.8036800622940063,
      "num_tokens": 143062786.0,
      "step": 4270
    },
    {
      "epoch": 0.25519340974212035,
      "grad_norm": 1.364909052848816,
      "learning_rate": 4.926424938003e-05,
      "loss": 0.6147,
      "mean_token_accuracy": 0.8098294258117675,
      "num_tokens": 143230466.0,
      "step": 4275
    },
    {
      "epoch": 0.25549188156638014,
      "grad_norm": 0.6397042274475098,
      "learning_rate": 4.9261430277317196e-05,
      "loss": 0.6297,
      "mean_token_accuracy": 0.8068173646926879,
      "num_tokens": 143398146.0,
      "step": 4280
    },
    {
      "epoch": 0.25579035339063994,
      "grad_norm": 0.5720118880271912,
      "learning_rate": 4.9258605874206364e-05,
      "loss": 0.6016,
      "mean_token_accuracy": 0.8148157000541687,
      "num_tokens": 143565826.0,
      "step": 4285
    },
    {
      "epoch": 0.25608882521489973,
      "grad_norm": 0.6098594665527344,
      "learning_rate": 4.925577617138546e-05,
      "loss": 0.5976,
      "mean_token_accuracy": 0.8132112741470336,
      "num_tokens": 143733506.0,
      "step": 4290
    },
    {
      "epoch": 0.2563872970391595,
      "grad_norm": 0.5886116623878479,
      "learning_rate": 4.925294116954371e-05,
      "loss": 0.5775,
      "mean_token_accuracy": 0.820589280128479,
      "num_tokens": 143901186.0,
      "step": 4295
    },
    {
      "epoch": 0.2566857688634193,
      "grad_norm": 0.6583279371261597,
      "learning_rate": 4.925010086937161e-05,
      "loss": 0.6053,
      "mean_token_accuracy": 0.8142192482948303,
      "num_tokens": 144068866.0,
      "step": 4300
    },
    {
      "epoch": 0.2569842406876791,
      "grad_norm": 0.6619601845741272,
      "learning_rate": 4.9247255271560994e-05,
      "loss": 0.5762,
      "mean_token_accuracy": 0.8225515842437744,
      "num_tokens": 144236546.0,
      "step": 4305
    },
    {
      "epoch": 0.25728271251193885,
      "grad_norm": 0.8144528865814209,
      "learning_rate": 4.924440437680495e-05,
      "loss": 0.6199,
      "mean_token_accuracy": 0.8079506039619446,
      "num_tokens": 144404226.0,
      "step": 4310
    },
    {
      "epoch": 0.25758118433619864,
      "grad_norm": 0.6154191493988037,
      "learning_rate": 4.924154818579786e-05,
      "loss": 0.5922,
      "mean_token_accuracy": 0.814887273311615,
      "num_tokens": 144571906.0,
      "step": 4315
    },
    {
      "epoch": 0.25787965616045844,
      "grad_norm": 0.6959332823753357,
      "learning_rate": 4.92386866992354e-05,
      "loss": 0.6252,
      "mean_token_accuracy": 0.80547536611557,
      "num_tokens": 144739586.0,
      "step": 4320
    },
    {
      "epoch": 0.25817812798471823,
      "grad_norm": 0.642940104007721,
      "learning_rate": 4.923581991781456e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.8142550230026245,
      "num_tokens": 144907266.0,
      "step": 4325
    },
    {
      "epoch": 0.258476599808978,
      "grad_norm": 0.6554369926452637,
      "learning_rate": 4.9232947842233567e-05,
      "loss": 0.5374,
      "mean_token_accuracy": 0.8312060236930847,
      "num_tokens": 145074946.0,
      "step": 4330
    },
    {
      "epoch": 0.2587750716332378,
      "grad_norm": 0.6094810962677002,
      "learning_rate": 4.9230070473191994e-05,
      "loss": 0.573,
      "mean_token_accuracy": 0.820589292049408,
      "num_tokens": 145242626.0,
      "step": 4335
    },
    {
      "epoch": 0.2590735434574976,
      "grad_norm": 0.634344220161438,
      "learning_rate": 4.9227187811390654e-05,
      "loss": 0.5919,
      "mean_token_accuracy": 0.8159250855445862,
      "num_tokens": 145410306.0,
      "step": 4340
    },
    {
      "epoch": 0.2593720152817574,
      "grad_norm": 0.6004631519317627,
      "learning_rate": 4.922429985753169e-05,
      "loss": 0.5849,
      "mean_token_accuracy": 0.8194620132446289,
      "num_tokens": 145577986.0,
      "step": 4345
    },
    {
      "epoch": 0.2596704871060172,
      "grad_norm": 0.6163318157196045,
      "learning_rate": 4.922140661231852e-05,
      "loss": 0.5744,
      "mean_token_accuracy": 0.819724440574646,
      "num_tokens": 145745666.0,
      "step": 4350
    },
    {
      "epoch": 0.259968958930277,
      "grad_norm": 0.6415529251098633,
      "learning_rate": 4.921850807645582e-05,
      "loss": 0.5937,
      "mean_token_accuracy": 0.8149588346481323,
      "num_tokens": 145913346.0,
      "step": 4355
    },
    {
      "epoch": 0.2602674307545368,
      "grad_norm": 0.7055518627166748,
      "learning_rate": 4.9215604250649613e-05,
      "loss": 0.6781,
      "mean_token_accuracy": 0.7928009033203125,
      "num_tokens": 146081026.0,
      "step": 4360
    },
    {
      "epoch": 0.2605659025787966,
      "grad_norm": 0.6350159049034119,
      "learning_rate": 4.9212695135607166e-05,
      "loss": 0.5556,
      "mean_token_accuracy": 0.8243946194648742,
      "num_tokens": 146248706.0,
      "step": 4365
    },
    {
      "epoch": 0.2608643744030564,
      "grad_norm": 0.5864571332931519,
      "learning_rate": 4.920978073203705e-05,
      "loss": 0.5874,
      "mean_token_accuracy": 0.8168495893478394,
      "num_tokens": 146416386.0,
      "step": 4370
    },
    {
      "epoch": 0.2611628462273161,
      "grad_norm": 0.6074321866035461,
      "learning_rate": 4.9206861040649116e-05,
      "loss": 0.6377,
      "mean_token_accuracy": 0.8045389533042908,
      "num_tokens": 146584066.0,
      "step": 4375
    },
    {
      "epoch": 0.2614613180515759,
      "grad_norm": 0.6155579090118408,
      "learning_rate": 4.920393606215451e-05,
      "loss": 0.5663,
      "mean_token_accuracy": 0.8229452490806579,
      "num_tokens": 146751746.0,
      "step": 4380
    },
    {
      "epoch": 0.2617597898758357,
      "grad_norm": 0.7060921788215637,
      "learning_rate": 4.920100579726567e-05,
      "loss": 0.6268,
      "mean_token_accuracy": 0.8044196724891662,
      "num_tokens": 146919426.0,
      "step": 4385
    },
    {
      "epoch": 0.2620582617000955,
      "grad_norm": 0.743566632270813,
      "learning_rate": 4.919807024669632e-05,
      "loss": 0.5814,
      "mean_token_accuracy": 0.8192711353302002,
      "num_tokens": 147087106.0,
      "step": 4390
    },
    {
      "epoch": 0.2623567335243553,
      "grad_norm": 0.6009845733642578,
      "learning_rate": 4.919512941116145e-05,
      "loss": 0.5866,
      "mean_token_accuracy": 0.8185792565345764,
      "num_tokens": 147254786.0,
      "step": 4395
    },
    {
      "epoch": 0.2626552053486151,
      "grad_norm": 0.5867632031440735,
      "learning_rate": 4.919218329137737e-05,
      "loss": 0.6278,
      "mean_token_accuracy": 0.8053262591361999,
      "num_tokens": 147422466.0,
      "step": 4400
    },
    {
      "epoch": 0.2629536771728749,
      "grad_norm": 0.7206937670707703,
      "learning_rate": 4.918923188806166e-05,
      "loss": 0.605,
      "mean_token_accuracy": 0.8124836087226868,
      "num_tokens": 147590146.0,
      "step": 4405
    },
    {
      "epoch": 0.26325214899713467,
      "grad_norm": 0.5884151458740234,
      "learning_rate": 4.918627520193319e-05,
      "loss": 0.6189,
      "mean_token_accuracy": 0.8088274002075195,
      "num_tokens": 147757826.0,
      "step": 4410
    },
    {
      "epoch": 0.26355062082139447,
      "grad_norm": 0.7009053826332092,
      "learning_rate": 4.91833132337121e-05,
      "loss": 0.6253,
      "mean_token_accuracy": 0.8057199120521545,
      "num_tokens": 147925506.0,
      "step": 4415
    },
    {
      "epoch": 0.26384909264565426,
      "grad_norm": 0.5944790840148926,
      "learning_rate": 4.918034598411985e-05,
      "loss": 0.5852,
      "mean_token_accuracy": 0.8170404195785522,
      "num_tokens": 148093186.0,
      "step": 4420
    },
    {
      "epoch": 0.26414756446991405,
      "grad_norm": 0.659602165222168,
      "learning_rate": 4.917737345387916e-05,
      "loss": 0.577,
      "mean_token_accuracy": 0.8198437333106995,
      "num_tokens": 148260866.0,
      "step": 4425
    },
    {
      "epoch": 0.26444603629417385,
      "grad_norm": 0.6222054362297058,
      "learning_rate": 4.9174395643714046e-05,
      "loss": 0.5624,
      "mean_token_accuracy": 0.8251282334327698,
      "num_tokens": 148428546.0,
      "step": 4430
    },
    {
      "epoch": 0.26474450811843364,
      "grad_norm": 0.6671966910362244,
      "learning_rate": 4.917141255434982e-05,
      "loss": 0.6131,
      "mean_token_accuracy": 0.8097936272621155,
      "num_tokens": 148596226.0,
      "step": 4435
    },
    {
      "epoch": 0.26504297994269344,
      "grad_norm": 0.6124293804168701,
      "learning_rate": 4.916842418651305e-05,
      "loss": 0.5991,
      "mean_token_accuracy": 0.813897168636322,
      "num_tokens": 148763906.0,
      "step": 4440
    },
    {
      "epoch": 0.2653414517669532,
      "grad_norm": 0.7368993759155273,
      "learning_rate": 4.916543054093161e-05,
      "loss": 0.5929,
      "mean_token_accuracy": 0.8160443663597107,
      "num_tokens": 148931586.0,
      "step": 4445
    },
    {
      "epoch": 0.26563992359121297,
      "grad_norm": 0.6545404195785522,
      "learning_rate": 4.916243161833468e-05,
      "loss": 0.5994,
      "mean_token_accuracy": 0.813396155834198,
      "num_tokens": 149099266.0,
      "step": 4450
    },
    {
      "epoch": 0.26593839541547276,
      "grad_norm": 0.722064733505249,
      "learning_rate": 4.9159427419452683e-05,
      "loss": 0.6337,
      "mean_token_accuracy": 0.8045628190040588,
      "num_tokens": 149266946.0,
      "step": 4455
    },
    {
      "epoch": 0.26623686723973256,
      "grad_norm": 0.6071226596832275,
      "learning_rate": 4.9156417945017335e-05,
      "loss": 0.6157,
      "mean_token_accuracy": 0.8071223258972168,
      "num_tokens": 149432368.0,
      "step": 4460
    },
    {
      "epoch": 0.26653533906399235,
      "grad_norm": 0.7124296426773071,
      "learning_rate": 4.915340319576168e-05,
      "loss": 0.6045,
      "mean_token_accuracy": 0.8115591168403625,
      "num_tokens": 149600048.0,
      "step": 4465
    },
    {
      "epoch": 0.26683381088825214,
      "grad_norm": 0.6773062348365784,
      "learning_rate": 4.915038317242e-05,
      "loss": 0.6336,
      "mean_token_accuracy": 0.8048848986625672,
      "num_tokens": 149767728.0,
      "step": 4470
    },
    {
      "epoch": 0.26713228271251194,
      "grad_norm": 0.5971149206161499,
      "learning_rate": 4.914735787572787e-05,
      "loss": 0.611,
      "mean_token_accuracy": 0.8102946400642395,
      "num_tokens": 149935408.0,
      "step": 4475
    },
    {
      "epoch": 0.26743075453677173,
      "grad_norm": 0.6414796710014343,
      "learning_rate": 4.914432730642217e-05,
      "loss": 0.606,
      "mean_token_accuracy": 0.814410126209259,
      "num_tokens": 150103088.0,
      "step": 4480
    },
    {
      "epoch": 0.2677292263610315,
      "grad_norm": 0.6498230695724487,
      "learning_rate": 4.914129146524104e-05,
      "loss": 0.6119,
      "mean_token_accuracy": 0.8103244781494141,
      "num_tokens": 150270768.0,
      "step": 4485
    },
    {
      "epoch": 0.2680276981852913,
      "grad_norm": 0.7054505944252014,
      "learning_rate": 4.913825035292392e-05,
      "loss": 0.6122,
      "mean_token_accuracy": 0.8097757458686828,
      "num_tokens": 150438448.0,
      "step": 4490
    },
    {
      "epoch": 0.2683261700095511,
      "grad_norm": 0.6897138357162476,
      "learning_rate": 4.913520397021154e-05,
      "loss": 0.6458,
      "mean_token_accuracy": 0.7985923767089844,
      "num_tokens": 150606128.0,
      "step": 4495
    },
    {
      "epoch": 0.2686246418338109,
      "grad_norm": 0.6120455861091614,
      "learning_rate": 4.913215231784589e-05,
      "loss": 0.6003,
      "mean_token_accuracy": 0.8152928471565246,
      "num_tokens": 150773808.0,
      "step": 4500
    },
    {
      "epoch": 0.2689231136580707,
      "grad_norm": 0.7380448579788208,
      "learning_rate": 4.912909539657026e-05,
      "loss": 0.65,
      "mean_token_accuracy": 0.7993319749832153,
      "num_tokens": 150941488.0,
      "step": 4505
    },
    {
      "epoch": 0.2692215854823305,
      "grad_norm": 0.6018991470336914,
      "learning_rate": 4.9126033207129216e-05,
      "loss": 0.5659,
      "mean_token_accuracy": 0.8222354888916016,
      "num_tokens": 151109168.0,
      "step": 4510
    },
    {
      "epoch": 0.26952005730659023,
      "grad_norm": 0.6238842010498047,
      "learning_rate": 4.912296575026861e-05,
      "loss": 0.572,
      "mean_token_accuracy": 0.8205714106559754,
      "num_tokens": 151276848.0,
      "step": 4515
    },
    {
      "epoch": 0.26981852913085,
      "grad_norm": 0.6419805288314819,
      "learning_rate": 4.911989302673558e-05,
      "loss": 0.6238,
      "mean_token_accuracy": 0.8058690190315246,
      "num_tokens": 151444528.0,
      "step": 4520
    },
    {
      "epoch": 0.2701170009551098,
      "grad_norm": 0.6212139129638672,
      "learning_rate": 4.911681503727856e-05,
      "loss": 0.6277,
      "mean_token_accuracy": 0.8063998579978943,
      "num_tokens": 151612208.0,
      "step": 4525
    },
    {
      "epoch": 0.2704154727793696,
      "grad_norm": 0.6858535408973694,
      "learning_rate": 4.911373178264723e-05,
      "loss": 0.5831,
      "mean_token_accuracy": 0.8192097425460816,
      "num_tokens": 151772383.0,
      "step": 4530
    },
    {
      "epoch": 0.2707139446036294,
      "grad_norm": 0.6183403730392456,
      "learning_rate": 4.911064326359258e-05,
      "loss": 0.5956,
      "mean_token_accuracy": 0.8123822093009949,
      "num_tokens": 151940063.0,
      "step": 4535
    },
    {
      "epoch": 0.2710124164278892,
      "grad_norm": 0.6655312180519104,
      "learning_rate": 4.910754948086688e-05,
      "loss": 0.7022,
      "mean_token_accuracy": 0.7862400054931641,
      "num_tokens": 152107743.0,
      "step": 4540
    },
    {
      "epoch": 0.271310888252149,
      "grad_norm": 0.6107160449028015,
      "learning_rate": 4.910445043522366e-05,
      "loss": 0.5805,
      "mean_token_accuracy": 0.8194679856300354,
      "num_tokens": 152275423.0,
      "step": 4545
    },
    {
      "epoch": 0.2716093600764088,
      "grad_norm": 0.6021928191184998,
      "learning_rate": 4.910134612741777e-05,
      "loss": 0.6332,
      "mean_token_accuracy": 0.8057378053665161,
      "num_tokens": 152443103.0,
      "step": 4550
    },
    {
      "epoch": 0.2719078319006686,
      "grad_norm": 0.6171747446060181,
      "learning_rate": 4.909823655820533e-05,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8300071597099304,
      "num_tokens": 152610783.0,
      "step": 4555
    },
    {
      "epoch": 0.2722063037249284,
      "grad_norm": 0.6629837155342102,
      "learning_rate": 4.9095121728343703e-05,
      "loss": 0.6131,
      "mean_token_accuracy": 0.8109447717666626,
      "num_tokens": 152778463.0,
      "step": 4560
    },
    {
      "epoch": 0.27250477554918817,
      "grad_norm": 0.6519883871078491,
      "learning_rate": 4.909200163859158e-05,
      "loss": 0.6218,
      "mean_token_accuracy": 0.8087200284004211,
      "num_tokens": 152946143.0,
      "step": 4565
    },
    {
      "epoch": 0.27280324737344797,
      "grad_norm": 0.7325212955474854,
      "learning_rate": 4.908887628970892e-05,
      "loss": 0.6511,
      "mean_token_accuracy": 0.798783254623413,
      "num_tokens": 153113823.0,
      "step": 4570
    },
    {
      "epoch": 0.27310171919770776,
      "grad_norm": 0.7060492634773254,
      "learning_rate": 4.908574568245696e-05,
      "loss": 0.6135,
      "mean_token_accuracy": 0.8075808167457581,
      "num_tokens": 153281503.0,
      "step": 4575
    },
    {
      "epoch": 0.27340019102196755,
      "grad_norm": 2.399169921875,
      "learning_rate": 4.90826098175982e-05,
      "loss": 0.6074,
      "mean_token_accuracy": 0.8116068243980408,
      "num_tokens": 153449183.0,
      "step": 4580
    },
    {
      "epoch": 0.2736986628462273,
      "grad_norm": 0.7301245927810669,
      "learning_rate": 4.907946869589646e-05,
      "loss": 0.6652,
      "mean_token_accuracy": 0.7932243824005127,
      "num_tokens": 153616863.0,
      "step": 4585
    },
    {
      "epoch": 0.2739971346704871,
      "grad_norm": 0.8387046456336975,
      "learning_rate": 4.90763223181168e-05,
      "loss": 0.638,
      "mean_token_accuracy": 0.8020458102226258,
      "num_tokens": 153784543.0,
      "step": 4590
    },
    {
      "epoch": 0.2742956064947469,
      "grad_norm": 0.6804915070533752,
      "learning_rate": 4.9073170685025585e-05,
      "loss": 0.6487,
      "mean_token_accuracy": 0.7996182918548584,
      "num_tokens": 153952223.0,
      "step": 4595
    },
    {
      "epoch": 0.2745940783190067,
      "grad_norm": 0.6785187125205994,
      "learning_rate": 4.9070013797390466e-05,
      "loss": 0.6194,
      "mean_token_accuracy": 0.8091673493385315,
      "num_tokens": 154119903.0,
      "step": 4600
    },
    {
      "epoch": 0.27489255014326647,
      "grad_norm": 0.6009164452552795,
      "learning_rate": 4.906685165598034e-05,
      "loss": 0.6108,
      "mean_token_accuracy": 0.8108791589736939,
      "num_tokens": 154287583.0,
      "step": 4605
    },
    {
      "epoch": 0.27519102196752626,
      "grad_norm": 0.6155039072036743,
      "learning_rate": 4.906368426156542e-05,
      "loss": 0.6083,
      "mean_token_accuracy": 0.8118394494056702,
      "num_tokens": 154455263.0,
      "step": 4610
    },
    {
      "epoch": 0.27548949379178606,
      "grad_norm": 0.779868483543396,
      "learning_rate": 4.9060511614917174e-05,
      "loss": 0.6843,
      "mean_token_accuracy": 0.7925682783126831,
      "num_tokens": 154622943.0,
      "step": 4615
    },
    {
      "epoch": 0.27578796561604585,
      "grad_norm": 0.6313185691833496,
      "learning_rate": 4.905733371680836e-05,
      "loss": 0.6082,
      "mean_token_accuracy": 0.8114636778831482,
      "num_tokens": 154790623.0,
      "step": 4620
    },
    {
      "epoch": 0.27608643744030564,
      "grad_norm": 0.7433581352233887,
      "learning_rate": 4.9054150568013015e-05,
      "loss": 0.6507,
      "mean_token_accuracy": 0.7985446810722351,
      "num_tokens": 154958303.0,
      "step": 4625
    },
    {
      "epoch": 0.27638490926456544,
      "grad_norm": 0.6007549166679382,
      "learning_rate": 4.905096216930647e-05,
      "loss": 0.5948,
      "mean_token_accuracy": 0.8150483131408691,
      "num_tokens": 155125983.0,
      "step": 4630
    },
    {
      "epoch": 0.27668338108882523,
      "grad_norm": 0.64377760887146,
      "learning_rate": 4.904776852146529e-05,
      "loss": 0.6639,
      "mean_token_accuracy": 0.7965108036994935,
      "num_tokens": 155293663.0,
      "step": 4635
    },
    {
      "epoch": 0.276981852913085,
      "grad_norm": 0.6571340560913086,
      "learning_rate": 4.904456962526736e-05,
      "loss": 0.5969,
      "mean_token_accuracy": 0.8157521247863769,
      "num_tokens": 155461343.0,
      "step": 4640
    },
    {
      "epoch": 0.2772803247373448,
      "grad_norm": 0.8081356287002563,
      "learning_rate": 4.904136548149184e-05,
      "loss": 0.6486,
      "mean_token_accuracy": 0.7998512029647827,
      "num_tokens": 155621705.0,
      "step": 4645
    },
    {
      "epoch": 0.27757879656160456,
      "grad_norm": 0.5961964726448059,
      "learning_rate": 4.903815609091915e-05,
      "loss": 0.5854,
      "mean_token_accuracy": 0.8178158164024353,
      "num_tokens": 155789385.0,
      "step": 4650
    },
    {
      "epoch": 0.27787726838586435,
      "grad_norm": 0.6135856509208679,
      "learning_rate": 4.903494145433098e-05,
      "loss": 0.5794,
      "mean_token_accuracy": 0.8214601039886474,
      "num_tokens": 155957065.0,
      "step": 4655
    },
    {
      "epoch": 0.27817574021012414,
      "grad_norm": 0.625612199306488,
      "learning_rate": 4.903172157251034e-05,
      "loss": 0.646,
      "mean_token_accuracy": 0.8011093854904174,
      "num_tokens": 156124745.0,
      "step": 4660
    },
    {
      "epoch": 0.27847421203438394,
      "grad_norm": 0.6212636828422546,
      "learning_rate": 4.9028496446241475e-05,
      "loss": 0.5763,
      "mean_token_accuracy": 0.8192711353302002,
      "num_tokens": 156292425.0,
      "step": 4665
    },
    {
      "epoch": 0.27877268385864373,
      "grad_norm": 0.6626037359237671,
      "learning_rate": 4.902526607630994e-05,
      "loss": 0.6119,
      "mean_token_accuracy": 0.8083144426345825,
      "num_tokens": 156460105.0,
      "step": 4670
    },
    {
      "epoch": 0.2790711556829035,
      "grad_norm": 0.6458653211593628,
      "learning_rate": 4.9022030463502535e-05,
      "loss": 0.5655,
      "mean_token_accuracy": 0.8228319168090821,
      "num_tokens": 156627785.0,
      "step": 4675
    },
    {
      "epoch": 0.2793696275071633,
      "grad_norm": 0.7207940220832825,
      "learning_rate": 4.901878960860736e-05,
      "loss": 0.6332,
      "mean_token_accuracy": 0.8043540477752685,
      "num_tokens": 156795465.0,
      "step": 4680
    },
    {
      "epoch": 0.2796680993314231,
      "grad_norm": 0.6036509871482849,
      "learning_rate": 4.9015543512413784e-05,
      "loss": 0.6246,
      "mean_token_accuracy": 0.8071394443511963,
      "num_tokens": 156963145.0,
      "step": 4685
    },
    {
      "epoch": 0.2799665711556829,
      "grad_norm": 0.6261608600616455,
      "learning_rate": 4.9012292175712455e-05,
      "loss": 0.6058,
      "mean_token_accuracy": 0.8114457845687866,
      "num_tokens": 157130825.0,
      "step": 4690
    },
    {
      "epoch": 0.2802650429799427,
      "grad_norm": 0.5815515518188477,
      "learning_rate": 4.900903559929529e-05,
      "loss": 0.5283,
      "mean_token_accuracy": 0.8339675664901733,
      "num_tokens": 157298505.0,
      "step": 4695
    },
    {
      "epoch": 0.2805635148042025,
      "grad_norm": 0.8820967078208923,
      "learning_rate": 4.90057737839555e-05,
      "loss": 0.6077,
      "mean_token_accuracy": 0.8129726767539978,
      "num_tokens": 157466185.0,
      "step": 4700
    },
    {
      "epoch": 0.2808619866284623,
      "grad_norm": 0.6315600275993347,
      "learning_rate": 4.900250673048755e-05,
      "loss": 0.5957,
      "mean_token_accuracy": 0.8143325924873352,
      "num_tokens": 157633865.0,
      "step": 4705
    },
    {
      "epoch": 0.2811604584527221,
      "grad_norm": 0.6613008975982666,
      "learning_rate": 4.89992344396872e-05,
      "loss": 0.6525,
      "mean_token_accuracy": 0.7995705485343934,
      "num_tokens": 157801545.0,
      "step": 4710
    },
    {
      "epoch": 0.2814589302769819,
      "grad_norm": 0.6789300441741943,
      "learning_rate": 4.899595691235147e-05,
      "loss": 0.6126,
      "mean_token_accuracy": 0.8100262403488159,
      "num_tokens": 157969225.0,
      "step": 4715
    },
    {
      "epoch": 0.2817574021012416,
      "grad_norm": 0.5721486806869507,
      "learning_rate": 4.8992674149278666e-05,
      "loss": 0.5895,
      "mean_token_accuracy": 0.8170821905136109,
      "num_tokens": 158136905.0,
      "step": 4720
    },
    {
      "epoch": 0.2820558739255014,
      "grad_norm": 0.658048152923584,
      "learning_rate": 4.898938615126836e-05,
      "loss": 0.5828,
      "mean_token_accuracy": 0.8190743207931519,
      "num_tokens": 158304585.0,
      "step": 4725
    },
    {
      "epoch": 0.2823543457497612,
      "grad_norm": 0.6302525997161865,
      "learning_rate": 4.898609291912141e-05,
      "loss": 0.5807,
      "mean_token_accuracy": 0.8197602272033692,
      "num_tokens": 158472265.0,
      "step": 4730
    },
    {
      "epoch": 0.282652817574021,
      "grad_norm": 0.7022088766098022,
      "learning_rate": 4.898279445363994e-05,
      "loss": 0.6114,
      "mean_token_accuracy": 0.810401999950409,
      "num_tokens": 158639945.0,
      "step": 4735
    },
    {
      "epoch": 0.2829512893982808,
      "grad_norm": 0.686716616153717,
      "learning_rate": 4.897949075562735e-05,
      "loss": 0.6057,
      "mean_token_accuracy": 0.8107419848442078,
      "num_tokens": 158807625.0,
      "step": 4740
    },
    {
      "epoch": 0.2832497612225406,
      "grad_norm": 0.6530365347862244,
      "learning_rate": 4.897618182588832e-05,
      "loss": 0.6132,
      "mean_token_accuracy": 0.809799587726593,
      "num_tokens": 158975305.0,
      "step": 4745
    },
    {
      "epoch": 0.2835482330468004,
      "grad_norm": 0.6255146861076355,
      "learning_rate": 4.8972867665228814e-05,
      "loss": 0.6057,
      "mean_token_accuracy": 0.8126386761665344,
      "num_tokens": 159142985.0,
      "step": 4750
    },
    {
      "epoch": 0.2838467048710602,
      "grad_norm": 0.6031068563461304,
      "learning_rate": 4.896954827445604e-05,
      "loss": 0.5822,
      "mean_token_accuracy": 0.8170821785926818,
      "num_tokens": 159310665.0,
      "step": 4755
    },
    {
      "epoch": 0.28414517669531997,
      "grad_norm": 0.5631386637687683,
      "learning_rate": 4.896622365437849e-05,
      "loss": 0.5838,
      "mean_token_accuracy": 0.8180186152458191,
      "num_tokens": 159478345.0,
      "step": 4760
    },
    {
      "epoch": 0.28444364851957976,
      "grad_norm": 0.5906351804733276,
      "learning_rate": 4.896289380580596e-05,
      "loss": 0.6043,
      "mean_token_accuracy": 0.8128533959388733,
      "num_tokens": 159646025.0,
      "step": 4765
    },
    {
      "epoch": 0.28474212034383956,
      "grad_norm": 0.6804120540618896,
      "learning_rate": 4.8959558729549474e-05,
      "loss": 0.5715,
      "mean_token_accuracy": 0.8215197443962097,
      "num_tokens": 159813705.0,
      "step": 4770
    },
    {
      "epoch": 0.28504059216809935,
      "grad_norm": 0.5899435877799988,
      "learning_rate": 4.8956218426421376e-05,
      "loss": 0.563,
      "mean_token_accuracy": 0.8240963816642761,
      "num_tokens": 159981385.0,
      "step": 4775
    },
    {
      "epoch": 0.28533906399235914,
      "grad_norm": 0.6063807010650635,
      "learning_rate": 4.8952872897235235e-05,
      "loss": 0.6559,
      "mean_token_accuracy": 0.7989741206169129,
      "num_tokens": 160149065.0,
      "step": 4780
    },
    {
      "epoch": 0.28563753581661894,
      "grad_norm": 0.6038150191307068,
      "learning_rate": 4.894952214280592e-05,
      "loss": 0.6318,
      "mean_token_accuracy": 0.804962420463562,
      "num_tokens": 160316745.0,
      "step": 4785
    },
    {
      "epoch": 0.2859360076408787,
      "grad_norm": 0.6020172238349915,
      "learning_rate": 4.894616616394958e-05,
      "loss": 0.611,
      "mean_token_accuracy": 0.8092150688171387,
      "num_tokens": 160484425.0,
      "step": 4790
    },
    {
      "epoch": 0.28623447946513847,
      "grad_norm": 0.6848268508911133,
      "learning_rate": 4.8942804961483625e-05,
      "loss": 0.6041,
      "mean_token_accuracy": 0.8130681157112122,
      "num_tokens": 160652105.0,
      "step": 4795
    },
    {
      "epoch": 0.28653295128939826,
      "grad_norm": 0.6249750256538391,
      "learning_rate": 4.893943853622672e-05,
      "loss": 0.6236,
      "mean_token_accuracy": 0.805588674545288,
      "num_tokens": 160819785.0,
      "step": 4800
    },
    {
      "epoch": 0.28683142311365806,
      "grad_norm": 0.5681930184364319,
      "learning_rate": 4.893606688899884e-05,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8348920345306396,
      "num_tokens": 160987465.0,
      "step": 4805
    },
    {
      "epoch": 0.28712989493791785,
      "grad_norm": 0.5807148218154907,
      "learning_rate": 4.893269002062121e-05,
      "loss": 0.6059,
      "mean_token_accuracy": 0.8145651936531066,
      "num_tokens": 161155145.0,
      "step": 4810
    },
    {
      "epoch": 0.28742836676217765,
      "grad_norm": 0.6384624242782593,
      "learning_rate": 4.892930793191632e-05,
      "loss": 0.5722,
      "mean_token_accuracy": 0.8224919438362122,
      "num_tokens": 161322825.0,
      "step": 4815
    },
    {
      "epoch": 0.28772683858643744,
      "grad_norm": 0.6156942248344421,
      "learning_rate": 4.892592062370795e-05,
      "loss": 0.6309,
      "mean_token_accuracy": 0.8054872870445251,
      "num_tokens": 161490505.0,
      "step": 4820
    },
    {
      "epoch": 0.28802531041069723,
      "grad_norm": 0.5932487845420837,
      "learning_rate": 4.892252809682112e-05,
      "loss": 0.5731,
      "mean_token_accuracy": 0.820827865600586,
      "num_tokens": 161658185.0,
      "step": 4825
    },
    {
      "epoch": 0.288323782234957,
      "grad_norm": 0.5755166411399841,
      "learning_rate": 4.8919130352082166e-05,
      "loss": 0.6025,
      "mean_token_accuracy": 0.8136943817138672,
      "num_tokens": 161825865.0,
      "step": 4830
    },
    {
      "epoch": 0.2886222540592168,
      "grad_norm": 0.60075843334198,
      "learning_rate": 4.8915727390318654e-05,
      "loss": 0.6399,
      "mean_token_accuracy": 0.8035786867141723,
      "num_tokens": 161993545.0,
      "step": 4835
    },
    {
      "epoch": 0.2889207258834766,
      "grad_norm": 0.7384197115898132,
      "learning_rate": 4.8912319212359455e-05,
      "loss": 0.6278,
      "mean_token_accuracy": 0.8037695288658142,
      "num_tokens": 162161225.0,
      "step": 4840
    },
    {
      "epoch": 0.2892191977077364,
      "grad_norm": 0.6177079081535339,
      "learning_rate": 4.890890581903468e-05,
      "loss": 0.6294,
      "mean_token_accuracy": 0.8046045541763306,
      "num_tokens": 162328905.0,
      "step": 4845
    },
    {
      "epoch": 0.2895176695319962,
      "grad_norm": 0.6216784715652466,
      "learning_rate": 4.8905487211175735e-05,
      "loss": 0.6247,
      "mean_token_accuracy": 0.8074018836021424,
      "num_tokens": 162496585.0,
      "step": 4850
    },
    {
      "epoch": 0.289816141356256,
      "grad_norm": 0.6456221342086792,
      "learning_rate": 4.890206338961528e-05,
      "loss": 0.5592,
      "mean_token_accuracy": 0.8251819133758544,
      "num_tokens": 162664265.0,
      "step": 4855
    },
    {
      "epoch": 0.29011461318051573,
      "grad_norm": 0.5484521985054016,
      "learning_rate": 4.8898634355187253e-05,
      "loss": 0.5744,
      "mean_token_accuracy": 0.8210604906082153,
      "num_tokens": 162831945.0,
      "step": 4860
    },
    {
      "epoch": 0.29041308500477553,
      "grad_norm": 0.5259699821472168,
      "learning_rate": 4.889520010872685e-05,
      "loss": 0.56,
      "mean_token_accuracy": 0.8252773523330689,
      "num_tokens": 162999625.0,
      "step": 4865
    },
    {
      "epoch": 0.2907115568290353,
      "grad_norm": 0.6381258964538574,
      "learning_rate": 4.889176065107055e-05,
      "loss": 0.6122,
      "mean_token_accuracy": 0.8114159584045411,
      "num_tokens": 163167305.0,
      "step": 4870
    },
    {
      "epoch": 0.2910100286532951,
      "grad_norm": 0.6049275398254395,
      "learning_rate": 4.88883159830561e-05,
      "loss": 0.5841,
      "mean_token_accuracy": 0.819223427772522,
      "num_tokens": 163334985.0,
      "step": 4875
    },
    {
      "epoch": 0.2913085004775549,
      "grad_norm": 0.7412250638008118,
      "learning_rate": 4.888486610552251e-05,
      "loss": 0.593,
      "mean_token_accuracy": 0.8158833503723144,
      "num_tokens": 163502665.0,
      "step": 4880
    },
    {
      "epoch": 0.2916069723018147,
      "grad_norm": 0.587427020072937,
      "learning_rate": 4.888141101931005e-05,
      "loss": 0.5552,
      "mean_token_accuracy": 0.8264642596244812,
      "num_tokens": 163670345.0,
      "step": 4885
    },
    {
      "epoch": 0.2919054441260745,
      "grad_norm": 0.5680304765701294,
      "learning_rate": 4.887795072526029e-05,
      "loss": 0.5707,
      "mean_token_accuracy": 0.8226350903511047,
      "num_tokens": 163838025.0,
      "step": 4890
    },
    {
      "epoch": 0.2922039159503343,
      "grad_norm": 0.6085522770881653,
      "learning_rate": 4.8874485224216045e-05,
      "loss": 0.6243,
      "mean_token_accuracy": 0.8064058303833008,
      "num_tokens": 164005705.0,
      "step": 4895
    },
    {
      "epoch": 0.2925023877745941,
      "grad_norm": 0.580993115901947,
      "learning_rate": 4.887101451702139e-05,
      "loss": 0.5771,
      "mean_token_accuracy": 0.821000850200653,
      "num_tokens": 164173385.0,
      "step": 4900
    },
    {
      "epoch": 0.2928008595988539,
      "grad_norm": 0.6264471411705017,
      "learning_rate": 4.886753860452168e-05,
      "loss": 0.5935,
      "mean_token_accuracy": 0.8143146872520447,
      "num_tokens": 164341065.0,
      "step": 4905
    },
    {
      "epoch": 0.2930993314231137,
      "grad_norm": 0.6031619310379028,
      "learning_rate": 4.886405748756354e-05,
      "loss": 0.5942,
      "mean_token_accuracy": 0.8155314326286316,
      "num_tokens": 164508745.0,
      "step": 4910
    },
    {
      "epoch": 0.29339780324737347,
      "grad_norm": 0.6414000391960144,
      "learning_rate": 4.886057116699488e-05,
      "loss": 0.6379,
      "mean_token_accuracy": 0.8034951686859131,
      "num_tokens": 164676425.0,
      "step": 4915
    },
    {
      "epoch": 0.29369627507163326,
      "grad_norm": 0.7670672535896301,
      "learning_rate": 4.8857079643664835e-05,
      "loss": 0.6457,
      "mean_token_accuracy": 0.8030001282691955,
      "num_tokens": 164844105.0,
      "step": 4920
    },
    {
      "epoch": 0.29399474689589306,
      "grad_norm": 0.7135143876075745,
      "learning_rate": 4.885358291842383e-05,
      "loss": 0.6023,
      "mean_token_accuracy": 0.8126386761665344,
      "num_tokens": 165011785.0,
      "step": 4925
    },
    {
      "epoch": 0.2942932187201528,
      "grad_norm": 0.5915950536727905,
      "learning_rate": 4.8850080992123566e-05,
      "loss": 0.617,
      "mean_token_accuracy": 0.8075629234313965,
      "num_tokens": 165179465.0,
      "step": 4930
    },
    {
      "epoch": 0.2945916905444126,
      "grad_norm": 0.6435756683349609,
      "learning_rate": 4.8846573865617e-05,
      "loss": 0.5822,
      "mean_token_accuracy": 0.8188059210777283,
      "num_tokens": 165347145.0,
      "step": 4935
    },
    {
      "epoch": 0.2948901623686724,
      "grad_norm": 0.6211757063865662,
      "learning_rate": 4.8843061539758356e-05,
      "loss": 0.5978,
      "mean_token_accuracy": 0.81315758228302,
      "num_tokens": 165514825.0,
      "step": 4940
    },
    {
      "epoch": 0.2951886341929322,
      "grad_norm": 0.5791261196136475,
      "learning_rate": 4.8839544015403124e-05,
      "loss": 0.6208,
      "mean_token_accuracy": 0.8100761532783508,
      "num_tokens": 165682265.0,
      "step": 4945
    },
    {
      "epoch": 0.29548710601719197,
      "grad_norm": 0.5903180837631226,
      "learning_rate": 4.883602129340807e-05,
      "loss": 0.5966,
      "mean_token_accuracy": 0.8133126497268677,
      "num_tokens": 165849945.0,
      "step": 4950
    },
    {
      "epoch": 0.29578557784145176,
      "grad_norm": 0.6267656683921814,
      "learning_rate": 4.883249337463121e-05,
      "loss": 0.5853,
      "mean_token_accuracy": 0.8176667094230652,
      "num_tokens": 166017625.0,
      "step": 4955
    },
    {
      "epoch": 0.29608404966571156,
      "grad_norm": 0.7100227475166321,
      "learning_rate": 4.882896025993183e-05,
      "loss": 0.5917,
      "mean_token_accuracy": 0.8154658198356628,
      "num_tokens": 166185305.0,
      "step": 4960
    },
    {
      "epoch": 0.29638252148997135,
      "grad_norm": 0.6177758574485779,
      "learning_rate": 4.8825421950170504e-05,
      "loss": 0.5757,
      "mean_token_accuracy": 0.8209352374076844,
      "num_tokens": 166352985.0,
      "step": 4965
    },
    {
      "epoch": 0.29668099331423115,
      "grad_norm": 0.6195228099822998,
      "learning_rate": 4.882187844620902e-05,
      "loss": 0.6513,
      "mean_token_accuracy": 0.7995347738265991,
      "num_tokens": 166520665.0,
      "step": 4970
    },
    {
      "epoch": 0.29697946513849094,
      "grad_norm": 0.6087715029716492,
      "learning_rate": 4.88183297489105e-05,
      "loss": 0.6273,
      "mean_token_accuracy": 0.8037814617156982,
      "num_tokens": 166688345.0,
      "step": 4975
    },
    {
      "epoch": 0.29727793696275073,
      "grad_norm": 0.6276805996894836,
      "learning_rate": 4.8814775859139275e-05,
      "loss": 0.6761,
      "mean_token_accuracy": 0.7932720899581909,
      "num_tokens": 166856025.0,
      "step": 4980
    },
    {
      "epoch": 0.2975764087870105,
      "grad_norm": 0.6047452688217163,
      "learning_rate": 4.881121677776097e-05,
      "loss": 0.5799,
      "mean_token_accuracy": 0.8176786303520203,
      "num_tokens": 167023705.0,
      "step": 4985
    },
    {
      "epoch": 0.2978748806112703,
      "grad_norm": 0.6401591897010803,
      "learning_rate": 4.880765250564245e-05,
      "loss": 0.5929,
      "mean_token_accuracy": 0.8156089663505555,
      "num_tokens": 167191385.0,
      "step": 4990
    },
    {
      "epoch": 0.29817335243553006,
      "grad_norm": 0.5788501501083374,
      "learning_rate": 4.8804083043651866e-05,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8146785020828247,
      "num_tokens": 167359065.0,
      "step": 4995
    },
    {
      "epoch": 0.29847182425978985,
      "grad_norm": 0.6091424226760864,
      "learning_rate": 4.880050839265863e-05,
      "loss": 0.6354,
      "mean_token_accuracy": 0.8067947149276733,
      "num_tokens": 167519785.0,
      "step": 5000
    },
    {
      "epoch": 0.29877029608404965,
      "grad_norm": 0.6005545854568481,
      "learning_rate": 4.879692855353342e-05,
      "loss": 0.6073,
      "mean_token_accuracy": 0.811767864227295,
      "num_tokens": 167687465.0,
      "step": 5005
    },
    {
      "epoch": 0.29906876790830944,
      "grad_norm": 0.5670431852340698,
      "learning_rate": 4.879334352714815e-05,
      "loss": 0.5295,
      "mean_token_accuracy": 0.8326553702354431,
      "num_tokens": 167855145.0,
      "step": 5010
    },
    {
      "epoch": 0.29936723973256923,
      "grad_norm": 0.6601032018661499,
      "learning_rate": 4.878975331437605e-05,
      "loss": 0.5877,
      "mean_token_accuracy": 0.8166706323623657,
      "num_tokens": 168022825.0,
      "step": 5015
    },
    {
      "epoch": 0.29966571155682903,
      "grad_norm": 0.6273783445358276,
      "learning_rate": 4.8786157916091557e-05,
      "loss": 0.6016,
      "mean_token_accuracy": 0.8141297698020935,
      "num_tokens": 168190505.0,
      "step": 5020
    },
    {
      "epoch": 0.2999641833810888,
      "grad_norm": 0.6342858672142029,
      "learning_rate": 4.8782557333170407e-05,
      "loss": 0.5616,
      "mean_token_accuracy": 0.8260228872299195,
      "num_tokens": 168358185.0,
      "step": 5025
    },
    {
      "epoch": 0.3002626552053486,
      "grad_norm": 0.6288208961486816,
      "learning_rate": 4.877895156648959e-05,
      "loss": 0.6302,
      "mean_token_accuracy": 0.804348087310791,
      "num_tokens": 168525865.0,
      "step": 5030
    },
    {
      "epoch": 0.3005611270296084,
      "grad_norm": 0.6230400204658508,
      "learning_rate": 4.8775340616927357e-05,
      "loss": 0.5905,
      "mean_token_accuracy": 0.8134021043777466,
      "num_tokens": 168693545.0,
      "step": 5035
    },
    {
      "epoch": 0.3008595988538682,
      "grad_norm": 0.6192215085029602,
      "learning_rate": 4.877172448536322e-05,
      "loss": 0.5606,
      "mean_token_accuracy": 0.8259632587432861,
      "num_tokens": 168861225.0,
      "step": 5040
    },
    {
      "epoch": 0.301158070678128,
      "grad_norm": 0.6162911057472229,
      "learning_rate": 4.8768103172677946e-05,
      "loss": 0.614,
      "mean_token_accuracy": 0.8094655871391296,
      "num_tokens": 169028905.0,
      "step": 5045
    },
    {
      "epoch": 0.3014565425023878,
      "grad_norm": 0.6504818797111511,
      "learning_rate": 4.876447667975358e-05,
      "loss": 0.5949,
      "mean_token_accuracy": 0.8127400755882264,
      "num_tokens": 169196585.0,
      "step": 5050
    },
    {
      "epoch": 0.3017550143266476,
      "grad_norm": 0.5875588059425354,
      "learning_rate": 4.8760845007473425e-05,
      "loss": 0.5729,
      "mean_token_accuracy": 0.8196349740028381,
      "num_tokens": 169364265.0,
      "step": 5055
    },
    {
      "epoch": 0.3020534861509074,
      "grad_norm": 0.5812243819236755,
      "learning_rate": 4.875720815672204e-05,
      "loss": 0.578,
      "mean_token_accuracy": 0.8185196161270142,
      "num_tokens": 169531945.0,
      "step": 5060
    },
    {
      "epoch": 0.3023519579751671,
      "grad_norm": 0.5958006381988525,
      "learning_rate": 4.8753566128385245e-05,
      "loss": 0.6153,
      "mean_token_accuracy": 0.8077537894248963,
      "num_tokens": 169699625.0,
      "step": 5065
    },
    {
      "epoch": 0.3026504297994269,
      "grad_norm": 0.6520281434059143,
      "learning_rate": 4.8749918923350116e-05,
      "loss": 0.5923,
      "mean_token_accuracy": 0.8155552864074707,
      "num_tokens": 169867305.0,
      "step": 5070
    },
    {
      "epoch": 0.3029489016236867,
      "grad_norm": 0.6548631191253662,
      "learning_rate": 4.874626654250502e-05,
      "loss": 0.6496,
      "mean_token_accuracy": 0.800375759601593,
      "num_tokens": 170034985.0,
      "step": 5075
    },
    {
      "epoch": 0.3032473734479465,
      "grad_norm": 0.5772848725318909,
      "learning_rate": 4.874260898673953e-05,
      "loss": 0.6061,
      "mean_token_accuracy": 0.811040198802948,
      "num_tokens": 170202665.0,
      "step": 5080
    },
    {
      "epoch": 0.3035458452722063,
      "grad_norm": 0.6648582220077515,
      "learning_rate": 4.873894625694453e-05,
      "loss": 0.6712,
      "mean_token_accuracy": 0.7928247570991516,
      "num_tokens": 170370345.0,
      "step": 5085
    },
    {
      "epoch": 0.3038443170964661,
      "grad_norm": 0.5819846391677856,
      "learning_rate": 4.873527835401215e-05,
      "loss": 0.569,
      "mean_token_accuracy": 0.8232553958892822,
      "num_tokens": 170538025.0,
      "step": 5090
    },
    {
      "epoch": 0.3041427889207259,
      "grad_norm": 0.613950252532959,
      "learning_rate": 4.8731605278835765e-05,
      "loss": 0.63,
      "mean_token_accuracy": 0.8058272719383239,
      "num_tokens": 170705705.0,
      "step": 5095
    },
    {
      "epoch": 0.3044412607449857,
      "grad_norm": 0.6041838526725769,
      "learning_rate": 4.872792703231003e-05,
      "loss": 0.6195,
      "mean_token_accuracy": 0.8056960463523865,
      "num_tokens": 170873385.0,
      "step": 5100
    },
    {
      "epoch": 0.30473973256924547,
      "grad_norm": 0.5953035354614258,
      "learning_rate": 4.872424361533083e-05,
      "loss": 0.548,
      "mean_token_accuracy": 0.8270905375480652,
      "num_tokens": 171041065.0,
      "step": 5105
    },
    {
      "epoch": 0.30503820439350526,
      "grad_norm": 0.6270595192909241,
      "learning_rate": 4.872055502879536e-05,
      "loss": 0.6101,
      "mean_token_accuracy": 0.8102767467498779,
      "num_tokens": 171208745.0,
      "step": 5110
    },
    {
      "epoch": 0.30533667621776506,
      "grad_norm": 1.1966700553894043,
      "learning_rate": 4.871686127360202e-05,
      "loss": 0.6191,
      "mean_token_accuracy": 0.8083681106567383,
      "num_tokens": 171376425.0,
      "step": 5115
    },
    {
      "epoch": 0.30563514804202485,
      "grad_norm": 0.6643148064613342,
      "learning_rate": 4.8713162350650494e-05,
      "loss": 0.6257,
      "mean_token_accuracy": 0.806465458869934,
      "num_tokens": 171544105.0,
      "step": 5120
    },
    {
      "epoch": 0.30593361986628465,
      "grad_norm": 0.5752717852592468,
      "learning_rate": 4.8709458260841726e-05,
      "loss": 0.5948,
      "mean_token_accuracy": 0.8158594846725464,
      "num_tokens": 171711785.0,
      "step": 5125
    },
    {
      "epoch": 0.30623209169054444,
      "grad_norm": 0.6883658170700073,
      "learning_rate": 4.870574900507792e-05,
      "loss": 0.5731,
      "mean_token_accuracy": 0.8207443714141845,
      "num_tokens": 171879465.0,
      "step": 5130
    },
    {
      "epoch": 0.3065305635148042,
      "grad_norm": 0.5872220993041992,
      "learning_rate": 4.8702034584262526e-05,
      "loss": 0.5396,
      "mean_token_accuracy": 0.8304544806480407,
      "num_tokens": 172047145.0,
      "step": 5135
    },
    {
      "epoch": 0.30682903533906397,
      "grad_norm": 0.7178948521614075,
      "learning_rate": 4.869831499930028e-05,
      "loss": 0.6339,
      "mean_token_accuracy": 0.8045568346977234,
      "num_tokens": 172214825.0,
      "step": 5140
    },
    {
      "epoch": 0.30712750716332377,
      "grad_norm": 0.6904928684234619,
      "learning_rate": 4.869459025109712e-05,
      "loss": 0.6241,
      "mean_token_accuracy": 0.8038709282875061,
      "num_tokens": 172382505.0,
      "step": 5145
    },
    {
      "epoch": 0.30742597898758356,
      "grad_norm": 0.5749028325080872,
      "learning_rate": 4.8690860340560305e-05,
      "loss": 0.5499,
      "mean_token_accuracy": 0.8282416939735413,
      "num_tokens": 172550185.0,
      "step": 5150
    },
    {
      "epoch": 0.30772445081184335,
      "grad_norm": 0.6001327633857727,
      "learning_rate": 4.8687125268598313e-05,
      "loss": 0.555,
      "mean_token_accuracy": 0.8251520872116089,
      "num_tokens": 172717865.0,
      "step": 5155
    },
    {
      "epoch": 0.30802292263610315,
      "grad_norm": 0.5972523093223572,
      "learning_rate": 4.86833850361209e-05,
      "loss": 0.5873,
      "mean_token_accuracy": 0.817756175994873,
      "num_tokens": 172885545.0,
      "step": 5160
    },
    {
      "epoch": 0.30832139446036294,
      "grad_norm": 0.5958548784255981,
      "learning_rate": 4.867963964403906e-05,
      "loss": 0.5543,
      "mean_token_accuracy": 0.825915539264679,
      "num_tokens": 173053225.0,
      "step": 5165
    },
    {
      "epoch": 0.30861986628462273,
      "grad_norm": 0.6663253903388977,
      "learning_rate": 4.8675889093265045e-05,
      "loss": 0.5804,
      "mean_token_accuracy": 0.8201359868049621,
      "num_tokens": 173220905.0,
      "step": 5170
    },
    {
      "epoch": 0.30891833810888253,
      "grad_norm": 0.6232882738113403,
      "learning_rate": 4.867213338471239e-05,
      "loss": 0.5736,
      "mean_token_accuracy": 0.8201598525047302,
      "num_tokens": 173388585.0,
      "step": 5175
    },
    {
      "epoch": 0.3092168099331423,
      "grad_norm": 0.6426844000816345,
      "learning_rate": 4.866837251929585e-05,
      "loss": 0.6398,
      "mean_token_accuracy": 0.8025706648826599,
      "num_tokens": 173556265.0,
      "step": 5180
    },
    {
      "epoch": 0.3095152817574021,
      "grad_norm": 1.2553555965423584,
      "learning_rate": 4.866460649793146e-05,
      "loss": 0.6617,
      "mean_token_accuracy": 0.7992246150970459,
      "num_tokens": 173723945.0,
      "step": 5185
    },
    {
      "epoch": 0.3098137535816619,
      "grad_norm": 0.7143647074699402,
      "learning_rate": 4.866083532153651e-05,
      "loss": 0.5855,
      "mean_token_accuracy": 0.8164678454399109,
      "num_tokens": 173891625.0,
      "step": 5190
    },
    {
      "epoch": 0.3101122254059217,
      "grad_norm": 0.672615647315979,
      "learning_rate": 4.865705899102953e-05,
      "loss": 0.6071,
      "mean_token_accuracy": 0.8126148223876953,
      "num_tokens": 174059305.0,
      "step": 5195
    },
    {
      "epoch": 0.3104106972301815,
      "grad_norm": 0.6389151215553284,
      "learning_rate": 4.865327750733031e-05,
      "loss": 0.6086,
      "mean_token_accuracy": 0.8121197700500489,
      "num_tokens": 174226985.0,
      "step": 5200
    },
    {
      "epoch": 0.31070916905444124,
      "grad_norm": 0.6401156783103943,
      "learning_rate": 4.864949087135992e-05,
      "loss": 0.562,
      "mean_token_accuracy": 0.8238935947418213,
      "num_tokens": 174394665.0,
      "step": 5205
    },
    {
      "epoch": 0.31100764087870103,
      "grad_norm": 0.5634213089942932,
      "learning_rate": 4.8645699084040635e-05,
      "loss": 0.5407,
      "mean_token_accuracy": 0.8304485321044922,
      "num_tokens": 174562345.0,
      "step": 5210
    },
    {
      "epoch": 0.3113061127029608,
      "grad_norm": 0.6145051717758179,
      "learning_rate": 4.8641902146296046e-05,
      "loss": 0.5903,
      "mean_token_accuracy": 0.8170046448707581,
      "num_tokens": 174730025.0,
      "step": 5215
    },
    {
      "epoch": 0.3116045845272206,
      "grad_norm": 0.6142057776451111,
      "learning_rate": 4.8638100059050954e-05,
      "loss": 0.5961,
      "mean_token_accuracy": 0.8149648070335388,
      "num_tokens": 174897705.0,
      "step": 5220
    },
    {
      "epoch": 0.3119030563514804,
      "grad_norm": 0.6189945936203003,
      "learning_rate": 4.863429282323143e-05,
      "loss": 0.5458,
      "mean_token_accuracy": 0.8285876154899597,
      "num_tokens": 175065385.0,
      "step": 5225
    },
    {
      "epoch": 0.3122015281757402,
      "grad_norm": 0.6174318790435791,
      "learning_rate": 4.8630480439764795e-05,
      "loss": 0.5513,
      "mean_token_accuracy": 0.8268877506256104,
      "num_tokens": 175233065.0,
      "step": 5230
    },
    {
      "epoch": 0.3125,
      "grad_norm": 0.7125818133354187,
      "learning_rate": 4.862666290957962e-05,
      "loss": 0.6382,
      "mean_token_accuracy": 0.8030537962913513,
      "num_tokens": 175400745.0,
      "step": 5235
    },
    {
      "epoch": 0.3127984718242598,
      "grad_norm": 0.6511279344558716,
      "learning_rate": 4.862284023360574e-05,
      "loss": 0.6085,
      "mean_token_accuracy": 0.8101157069206237,
      "num_tokens": 175568425.0,
      "step": 5240
    },
    {
      "epoch": 0.3130969436485196,
      "grad_norm": 0.5827562808990479,
      "learning_rate": 4.861901241277424e-05,
      "loss": 0.634,
      "mean_token_accuracy": 0.8035011410713195,
      "num_tokens": 175736105.0,
      "step": 5245
    },
    {
      "epoch": 0.3133954154727794,
      "grad_norm": 0.6435595154762268,
      "learning_rate": 4.8615179448017455e-05,
      "loss": 0.6515,
      "mean_token_accuracy": 0.8005535006523132,
      "num_tokens": 175903562.0,
      "step": 5250
    },
    {
      "epoch": 0.3136938872970392,
      "grad_norm": 0.6507923603057861,
      "learning_rate": 4.8611341340268977e-05,
      "loss": 0.6437,
      "mean_token_accuracy": 0.8026899814605712,
      "num_tokens": 176071242.0,
      "step": 5255
    },
    {
      "epoch": 0.31399235912129897,
      "grad_norm": 0.5808089375495911,
      "learning_rate": 4.860749809046364e-05,
      "loss": 0.6198,
      "mean_token_accuracy": 0.8061851263046265,
      "num_tokens": 176238922.0,
      "step": 5260
    },
    {
      "epoch": 0.31429083094555876,
      "grad_norm": 0.6327818632125854,
      "learning_rate": 4.860364969953754e-05,
      "loss": 0.6264,
      "mean_token_accuracy": 0.8052010059356689,
      "num_tokens": 176406602.0,
      "step": 5265
    },
    {
      "epoch": 0.3145893027698185,
      "grad_norm": 0.6396894454956055,
      "learning_rate": 4.859979616842803e-05,
      "loss": 0.5811,
      "mean_token_accuracy": 0.8183526277542115,
      "num_tokens": 176574282.0,
      "step": 5270
    },
    {
      "epoch": 0.3148877745940783,
      "grad_norm": 1.0681463479995728,
      "learning_rate": 4.859593749807371e-05,
      "loss": 0.6022,
      "mean_token_accuracy": 0.8144280195236206,
      "num_tokens": 176741962.0,
      "step": 5275
    },
    {
      "epoch": 0.3151862464183381,
      "grad_norm": 0.5932404398918152,
      "learning_rate": 4.8592073689414425e-05,
      "loss": 0.6206,
      "mean_token_accuracy": 0.8069306969642639,
      "num_tokens": 176909642.0,
      "step": 5280
    },
    {
      "epoch": 0.3154847182425979,
      "grad_norm": 0.624931812286377,
      "learning_rate": 4.858820474339128e-05,
      "loss": 0.6432,
      "mean_token_accuracy": 0.8028987169265747,
      "num_tokens": 177077322.0,
      "step": 5285
    },
    {
      "epoch": 0.3157831900668577,
      "grad_norm": 0.6247636079788208,
      "learning_rate": 4.8584330660946616e-05,
      "loss": 0.5895,
      "mean_token_accuracy": 0.8153167128562927,
      "num_tokens": 177245002.0,
      "step": 5290
    },
    {
      "epoch": 0.31608166189111747,
      "grad_norm": 0.7553396224975586,
      "learning_rate": 4.8580451443024055e-05,
      "loss": 0.6278,
      "mean_token_accuracy": 0.8084098815917968,
      "num_tokens": 177412682.0,
      "step": 5295
    },
    {
      "epoch": 0.31638013371537727,
      "grad_norm": 0.7019616365432739,
      "learning_rate": 4.8576567090568444e-05,
      "loss": 0.6217,
      "mean_token_accuracy": 0.8063640713691711,
      "num_tokens": 177580362.0,
      "step": 5300
    },
    {
      "epoch": 0.31667860553963706,
      "grad_norm": 0.6351245045661926,
      "learning_rate": 4.857267760452589e-05,
      "loss": 0.6409,
      "mean_token_accuracy": 0.8039186477661133,
      "num_tokens": 177748042.0,
      "step": 5305
    },
    {
      "epoch": 0.31697707736389685,
      "grad_norm": 0.6876610517501831,
      "learning_rate": 4.856878298584375e-05,
      "loss": 0.6451,
      "mean_token_accuracy": 0.8022128224372864,
      "num_tokens": 177915722.0,
      "step": 5310
    },
    {
      "epoch": 0.31727554918815665,
      "grad_norm": 0.6411224603652954,
      "learning_rate": 4.856488323547062e-05,
      "loss": 0.5567,
      "mean_token_accuracy": 0.8240248203277588,
      "num_tokens": 178083402.0,
      "step": 5315
    },
    {
      "epoch": 0.31757402101241644,
      "grad_norm": 0.5736415982246399,
      "learning_rate": 4.856097835435637e-05,
      "loss": 0.5964,
      "mean_token_accuracy": 0.8160384178161622,
      "num_tokens": 178251082.0,
      "step": 5320
    },
    {
      "epoch": 0.31787249283667623,
      "grad_norm": 0.6970974802970886,
      "learning_rate": 4.85570683434521e-05,
      "loss": 0.6301,
      "mean_token_accuracy": 0.8065429925918579,
      "num_tokens": 178418762.0,
      "step": 5325
    },
    {
      "epoch": 0.31817096466093603,
      "grad_norm": 0.5778356790542603,
      "learning_rate": 4.855315320371017e-05,
      "loss": 0.6304,
      "mean_token_accuracy": 0.8064356327056885,
      "num_tokens": 178586442.0,
      "step": 5330
    },
    {
      "epoch": 0.3184694364851958,
      "grad_norm": 0.6154081225395203,
      "learning_rate": 4.854923293608418e-05,
      "loss": 0.5991,
      "mean_token_accuracy": 0.8138733148574829,
      "num_tokens": 178754122.0,
      "step": 5335
    },
    {
      "epoch": 0.31876790830945556,
      "grad_norm": 0.5807360410690308,
      "learning_rate": 4.8545307541528976e-05,
      "loss": 0.5316,
      "mean_token_accuracy": 0.8323934197425842,
      "num_tokens": 178919522.0,
      "step": 5340
    },
    {
      "epoch": 0.31906638013371535,
      "grad_norm": 0.5846946239471436,
      "learning_rate": 4.854137702100068e-05,
      "loss": 0.6195,
      "mean_token_accuracy": 0.8068889498710632,
      "num_tokens": 179087202.0,
      "step": 5345
    },
    {
      "epoch": 0.31936485195797515,
      "grad_norm": 0.679115355014801,
      "learning_rate": 4.8537441375456616e-05,
      "loss": 0.5838,
      "mean_token_accuracy": 0.8179350972175599,
      "num_tokens": 179254882.0,
      "step": 5350
    },
    {
      "epoch": 0.31966332378223494,
      "grad_norm": 0.6877555251121521,
      "learning_rate": 4.853350060585541e-05,
      "loss": 0.5909,
      "mean_token_accuracy": 0.815525472164154,
      "num_tokens": 179422562.0,
      "step": 5355
    },
    {
      "epoch": 0.31996179560649474,
      "grad_norm": 0.5969734787940979,
      "learning_rate": 4.8529554713156894e-05,
      "loss": 0.5847,
      "mean_token_accuracy": 0.8171123147010804,
      "num_tokens": 179587954.0,
      "step": 5360
    },
    {
      "epoch": 0.32026026743075453,
      "grad_norm": 0.6116472482681274,
      "learning_rate": 4.852560369832217e-05,
      "loss": 0.562,
      "mean_token_accuracy": 0.8221042513847351,
      "num_tokens": 179755634.0,
      "step": 5365
    },
    {
      "epoch": 0.3205587392550143,
      "grad_norm": 0.5308366417884827,
      "learning_rate": 4.852164756231357e-05,
      "loss": 0.5794,
      "mean_token_accuracy": 0.8180126547813416,
      "num_tokens": 179923314.0,
      "step": 5370
    },
    {
      "epoch": 0.3208572110792741,
      "grad_norm": 0.652714729309082,
      "learning_rate": 4.85176863060947e-05,
      "loss": 0.5416,
      "mean_token_accuracy": 0.8298818945884705,
      "num_tokens": 180090994.0,
      "step": 5375
    },
    {
      "epoch": 0.3211556829035339,
      "grad_norm": 0.5556767582893372,
      "learning_rate": 4.8513719930630384e-05,
      "loss": 0.5427,
      "mean_token_accuracy": 0.8290886402130127,
      "num_tokens": 180258674.0,
      "step": 5380
    },
    {
      "epoch": 0.3214541547277937,
      "grad_norm": 0.7110474109649658,
      "learning_rate": 4.850974843688671e-05,
      "loss": 0.6025,
      "mean_token_accuracy": 0.8142908215522766,
      "num_tokens": 180426354.0,
      "step": 5385
    },
    {
      "epoch": 0.3217526265520535,
      "grad_norm": 0.5711827874183655,
      "learning_rate": 4.850577182583101e-05,
      "loss": 0.5814,
      "mean_token_accuracy": 0.8186866283416748,
      "num_tokens": 180594034.0,
      "step": 5390
    },
    {
      "epoch": 0.3220510983763133,
      "grad_norm": 0.5733540654182434,
      "learning_rate": 4.850179009843185e-05,
      "loss": 0.5854,
      "mean_token_accuracy": 0.8165752053260803,
      "num_tokens": 180761714.0,
      "step": 5395
    },
    {
      "epoch": 0.3223495702005731,
      "grad_norm": 0.6314049959182739,
      "learning_rate": 4.849780325565908e-05,
      "loss": 0.6026,
      "mean_token_accuracy": 0.8126088500022888,
      "num_tokens": 180929394.0,
      "step": 5400
    },
    {
      "epoch": 0.3226480420248329,
      "grad_norm": 0.6054881811141968,
      "learning_rate": 4.8493811298483734e-05,
      "loss": 0.5932,
      "mean_token_accuracy": 0.8160324454307556,
      "num_tokens": 181097074.0,
      "step": 5405
    },
    {
      "epoch": 0.3229465138490926,
      "grad_norm": 0.7026839256286621,
      "learning_rate": 4.848981422787816e-05,
      "loss": 0.6158,
      "mean_token_accuracy": 0.8087021350860596,
      "num_tokens": 181264754.0,
      "step": 5410
    },
    {
      "epoch": 0.3232449856733524,
      "grad_norm": 0.6497235298156738,
      "learning_rate": 4.848581204481589e-05,
      "loss": 0.5707,
      "mean_token_accuracy": 0.8209889173507691,
      "num_tokens": 181432434.0,
      "step": 5415
    },
    {
      "epoch": 0.3235434574976122,
      "grad_norm": 0.6246758699417114,
      "learning_rate": 4.848180475027175e-05,
      "loss": 0.5663,
      "mean_token_accuracy": 0.8242335677146911,
      "num_tokens": 181600114.0,
      "step": 5420
    },
    {
      "epoch": 0.323841929321872,
      "grad_norm": 0.5725294351577759,
      "learning_rate": 4.847779234522179e-05,
      "loss": 0.5731,
      "mean_token_accuracy": 0.818722414970398,
      "num_tokens": 181767794.0,
      "step": 5425
    },
    {
      "epoch": 0.3241404011461318,
      "grad_norm": 0.6300048828125,
      "learning_rate": 4.8473774830643285e-05,
      "loss": 0.5898,
      "mean_token_accuracy": 0.8144757270812988,
      "num_tokens": 181935474.0,
      "step": 5430
    },
    {
      "epoch": 0.3244388729703916,
      "grad_norm": 0.6033457517623901,
      "learning_rate": 4.84697522075148e-05,
      "loss": 0.549,
      "mean_token_accuracy": 0.8275855898857116,
      "num_tokens": 182103154.0,
      "step": 5435
    },
    {
      "epoch": 0.3247373447946514,
      "grad_norm": 0.6648246645927429,
      "learning_rate": 4.84657244768161e-05,
      "loss": 0.6028,
      "mean_token_accuracy": 0.8138315558433533,
      "num_tokens": 182270834.0,
      "step": 5440
    },
    {
      "epoch": 0.3250358166189112,
      "grad_norm": 0.6584506630897522,
      "learning_rate": 4.846169163952823e-05,
      "loss": 0.6648,
      "mean_token_accuracy": 0.7933377027511597,
      "num_tokens": 182438514.0,
      "step": 5445
    },
    {
      "epoch": 0.32533428844317097,
      "grad_norm": 0.6389214992523193,
      "learning_rate": 4.8457653696633445e-05,
      "loss": 0.548,
      "mean_token_accuracy": 0.8272515892982483,
      "num_tokens": 182606194.0,
      "step": 5450
    },
    {
      "epoch": 0.32563276026743077,
      "grad_norm": 0.5182211995124817,
      "learning_rate": 4.845361064911526e-05,
      "loss": 0.518,
      "mean_token_accuracy": 0.8367708563804627,
      "num_tokens": 182773874.0,
      "step": 5455
    },
    {
      "epoch": 0.32593123209169056,
      "grad_norm": 0.558883547782898,
      "learning_rate": 4.844956249795846e-05,
      "loss": 0.5726,
      "mean_token_accuracy": 0.8197960257530212,
      "num_tokens": 182941554.0,
      "step": 5460
    },
    {
      "epoch": 0.32622970391595035,
      "grad_norm": 0.5931894183158875,
      "learning_rate": 4.844550924414902e-05,
      "loss": 0.5469,
      "mean_token_accuracy": 0.8291184544563294,
      "num_tokens": 183109234.0,
      "step": 5465
    },
    {
      "epoch": 0.32652817574021015,
      "grad_norm": 0.5589408874511719,
      "learning_rate": 4.84414508886742e-05,
      "loss": 0.5301,
      "mean_token_accuracy": 0.8336275815963745,
      "num_tokens": 183276914.0,
      "step": 5470
    },
    {
      "epoch": 0.32682664756446994,
      "grad_norm": 0.5838602781295776,
      "learning_rate": 4.843738743252247e-05,
      "loss": 0.5677,
      "mean_token_accuracy": 0.8215972661972046,
      "num_tokens": 183444594.0,
      "step": 5475
    },
    {
      "epoch": 0.3271251193887297,
      "grad_norm": 0.5662837624549866,
      "learning_rate": 4.843331887668358e-05,
      "loss": 0.5726,
      "mean_token_accuracy": 0.8211618661880493,
      "num_tokens": 183612274.0,
      "step": 5480
    },
    {
      "epoch": 0.3274235912129895,
      "grad_norm": 0.5298907160758972,
      "learning_rate": 4.842924522214849e-05,
      "loss": 0.6208,
      "mean_token_accuracy": 0.8065310716629028,
      "num_tokens": 183779954.0,
      "step": 5485
    },
    {
      "epoch": 0.32772206303724927,
      "grad_norm": 0.6623823642730713,
      "learning_rate": 4.842516646990941e-05,
      "loss": 0.5948,
      "mean_token_accuracy": 0.8137659430503845,
      "num_tokens": 183947634.0,
      "step": 5490
    },
    {
      "epoch": 0.32802053486150906,
      "grad_norm": 0.6267683506011963,
      "learning_rate": 4.842108262095981e-05,
      "loss": 0.5876,
      "mean_token_accuracy": 0.8164260983467102,
      "num_tokens": 184115314.0,
      "step": 5495
    },
    {
      "epoch": 0.32831900668576885,
      "grad_norm": 0.5885863900184631,
      "learning_rate": 4.8416993676294376e-05,
      "loss": 0.577,
      "mean_token_accuracy": 0.8195395350456238,
      "num_tokens": 184282994.0,
      "step": 5500
    },
    {
      "epoch": 0.32861747851002865,
      "grad_norm": 0.6000836491584778,
      "learning_rate": 4.8412899636909046e-05,
      "loss": 0.5741,
      "mean_token_accuracy": 0.8211916923522949,
      "num_tokens": 184450674.0,
      "step": 5505
    },
    {
      "epoch": 0.32891595033428844,
      "grad_norm": 0.7105187773704529,
      "learning_rate": 4.840880050380101e-05,
      "loss": 0.6194,
      "mean_token_accuracy": 0.8071275234222413,
      "num_tokens": 184618354.0,
      "step": 5510
    },
    {
      "epoch": 0.32921442215854824,
      "grad_norm": 0.5919845700263977,
      "learning_rate": 4.840469627796867e-05,
      "loss": 0.5702,
      "mean_token_accuracy": 0.8209590792655945,
      "num_tokens": 184786034.0,
      "step": 5515
    },
    {
      "epoch": 0.32951289398280803,
      "grad_norm": 0.7687093019485474,
      "learning_rate": 4.84005869604117e-05,
      "loss": 0.6197,
      "mean_token_accuracy": 0.8077418565750122,
      "num_tokens": 184953714.0,
      "step": 5520
    },
    {
      "epoch": 0.3298113658070678,
      "grad_norm": 0.5766890645027161,
      "learning_rate": 4.839647255213099e-05,
      "loss": 0.5799,
      "mean_token_accuracy": 0.8185673356056213,
      "num_tokens": 185121394.0,
      "step": 5525
    },
    {
      "epoch": 0.3301098376313276,
      "grad_norm": 0.5770844221115112,
      "learning_rate": 4.839235305412869e-05,
      "loss": 0.5586,
      "mean_token_accuracy": 0.8244423151016236,
      "num_tokens": 185289074.0,
      "step": 5530
    },
    {
      "epoch": 0.3304083094555874,
      "grad_norm": 0.628121018409729,
      "learning_rate": 4.838822846740818e-05,
      "loss": 0.6078,
      "mean_token_accuracy": 0.8108910918235779,
      "num_tokens": 185456754.0,
      "step": 5535
    },
    {
      "epoch": 0.3307067812798472,
      "grad_norm": 0.6534087657928467,
      "learning_rate": 4.838409879297407e-05,
      "loss": 0.5987,
      "mean_token_accuracy": 0.8138912081718445,
      "num_tokens": 185624434.0,
      "step": 5540
    },
    {
      "epoch": 0.33100525310410694,
      "grad_norm": 0.6166808009147644,
      "learning_rate": 4.837996403183221e-05,
      "loss": 0.5865,
      "mean_token_accuracy": 0.8158236861228942,
      "num_tokens": 185792114.0,
      "step": 5545
    },
    {
      "epoch": 0.33130372492836674,
      "grad_norm": 0.6812239289283752,
      "learning_rate": 4.837582418498972e-05,
      "loss": 0.5766,
      "mean_token_accuracy": 0.8192234396934509,
      "num_tokens": 185959794.0,
      "step": 5550
    },
    {
      "epoch": 0.33160219675262653,
      "grad_norm": 0.969882607460022,
      "learning_rate": 4.837167925345493e-05,
      "loss": 0.5693,
      "mean_token_accuracy": 0.8214481711387634,
      "num_tokens": 186127474.0,
      "step": 5555
    },
    {
      "epoch": 0.3319006685768863,
      "grad_norm": 0.6157154440879822,
      "learning_rate": 4.8367529238237415e-05,
      "loss": 0.6312,
      "mean_token_accuracy": 0.8026601552963257,
      "num_tokens": 186295154.0,
      "step": 5560
    },
    {
      "epoch": 0.3321991404011461,
      "grad_norm": 0.5648495554924011,
      "learning_rate": 4.836337414034798e-05,
      "loss": 0.5949,
      "mean_token_accuracy": 0.8149349808692932,
      "num_tokens": 186462834.0,
      "step": 5565
    },
    {
      "epoch": 0.3324976122254059,
      "grad_norm": 0.6061586737632751,
      "learning_rate": 4.835921396079867e-05,
      "loss": 0.6278,
      "mean_token_accuracy": 0.8072468042373657,
      "num_tokens": 186630514.0,
      "step": 5570
    },
    {
      "epoch": 0.3327960840496657,
      "grad_norm": 0.5938173532485962,
      "learning_rate": 4.835504870060279e-05,
      "loss": 0.6261,
      "mean_token_accuracy": 0.806191086769104,
      "num_tokens": 186798194.0,
      "step": 5575
    },
    {
      "epoch": 0.3330945558739255,
      "grad_norm": 0.580461859703064,
      "learning_rate": 4.835087836077486e-05,
      "loss": 0.6032,
      "mean_token_accuracy": 0.8120541572570801,
      "num_tokens": 186965874.0,
      "step": 5580
    },
    {
      "epoch": 0.3333930276981853,
      "grad_norm": 0.5872578024864197,
      "learning_rate": 4.834670294233064e-05,
      "loss": 0.57,
      "mean_token_accuracy": 0.8211857318878174,
      "num_tokens": 187133554.0,
      "step": 5585
    },
    {
      "epoch": 0.3336914995224451,
      "grad_norm": 0.6603530645370483,
      "learning_rate": 4.834252244628714e-05,
      "loss": 0.5981,
      "mean_token_accuracy": 0.8149349808692932,
      "num_tokens": 187301234.0,
      "step": 5590
    },
    {
      "epoch": 0.3339899713467049,
      "grad_norm": 0.587928056716919,
      "learning_rate": 4.8338336873662575e-05,
      "loss": 0.5821,
      "mean_token_accuracy": 0.8193904399871826,
      "num_tokens": 187468914.0,
      "step": 5595
    },
    {
      "epoch": 0.3342884431709647,
      "grad_norm": 0.582267165184021,
      "learning_rate": 4.833414622547643e-05,
      "loss": 0.6273,
      "mean_token_accuracy": 0.8049087524414062,
      "num_tokens": 187636594.0,
      "step": 5600
    },
    {
      "epoch": 0.33458691499522447,
      "grad_norm": 0.6343382596969604,
      "learning_rate": 4.832995050274943e-05,
      "loss": 0.6159,
      "mean_token_accuracy": 0.8087200284004211,
      "num_tokens": 187804274.0,
      "step": 5605
    },
    {
      "epoch": 0.33488538681948427,
      "grad_norm": 0.8218846917152405,
      "learning_rate": 4.832574970650351e-05,
      "loss": 0.6026,
      "mean_token_accuracy": 0.8121257305145264,
      "num_tokens": 187971954.0,
      "step": 5610
    },
    {
      "epoch": 0.335183858643744,
      "grad_norm": 0.5850387811660767,
      "learning_rate": 4.8321543837761845e-05,
      "loss": 0.5869,
      "mean_token_accuracy": 0.8166825652122498,
      "num_tokens": 188139634.0,
      "step": 5615
    },
    {
      "epoch": 0.3354823304680038,
      "grad_norm": 0.6085715293884277,
      "learning_rate": 4.8317332897548854e-05,
      "loss": 0.6095,
      "mean_token_accuracy": 0.8088845252990723,
      "num_tokens": 188304602.0,
      "step": 5620
    },
    {
      "epoch": 0.3357808022922636,
      "grad_norm": 0.6014330983161926,
      "learning_rate": 4.83131168868902e-05,
      "loss": 0.5904,
      "mean_token_accuracy": 0.8155970335006714,
      "num_tokens": 188472282.0,
      "step": 5625
    },
    {
      "epoch": 0.3360792741165234,
      "grad_norm": 0.6113285422325134,
      "learning_rate": 4.830889580681276e-05,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8278301358222961,
      "num_tokens": 188639962.0,
      "step": 5630
    },
    {
      "epoch": 0.3363777459407832,
      "grad_norm": 0.5936959385871887,
      "learning_rate": 4.830466965834465e-05,
      "loss": 0.6193,
      "mean_token_accuracy": 0.8090242147445679,
      "num_tokens": 188807642.0,
      "step": 5635
    },
    {
      "epoch": 0.336676217765043,
      "grad_norm": 0.9623625874519348,
      "learning_rate": 4.830043844251524e-05,
      "loss": 0.5776,
      "mean_token_accuracy": 0.8192413210868835,
      "num_tokens": 188975322.0,
      "step": 5640
    },
    {
      "epoch": 0.33697468958930277,
      "grad_norm": 0.6620087623596191,
      "learning_rate": 4.829620216035511e-05,
      "loss": 0.6114,
      "mean_token_accuracy": 0.8087617874145507,
      "num_tokens": 189143002.0,
      "step": 5645
    },
    {
      "epoch": 0.33727316141356256,
      "grad_norm": 0.6530949473381042,
      "learning_rate": 4.82919608128961e-05,
      "loss": 0.6088,
      "mean_token_accuracy": 0.808946681022644,
      "num_tokens": 189310682.0,
      "step": 5650
    },
    {
      "epoch": 0.33757163323782235,
      "grad_norm": 0.5739794969558716,
      "learning_rate": 4.828771440117125e-05,
      "loss": 0.5597,
      "mean_token_accuracy": 0.8250924468040466,
      "num_tokens": 189478362.0,
      "step": 5655
    },
    {
      "epoch": 0.33787010506208215,
      "grad_norm": 0.7298800349235535,
      "learning_rate": 4.828346292621487e-05,
      "loss": 0.6046,
      "mean_token_accuracy": 0.8116187572479248,
      "num_tokens": 189646042.0,
      "step": 5660
    },
    {
      "epoch": 0.33816857688634194,
      "grad_norm": 0.8617460131645203,
      "learning_rate": 4.827920638906246e-05,
      "loss": 0.6807,
      "mean_token_accuracy": 0.7946856856346131,
      "num_tokens": 189813722.0,
      "step": 5665
    },
    {
      "epoch": 0.33846704871060174,
      "grad_norm": 0.6872591376304626,
      "learning_rate": 4.82749447907508e-05,
      "loss": 0.6192,
      "mean_token_accuracy": 0.8082965612411499,
      "num_tokens": 189981402.0,
      "step": 5670
    },
    {
      "epoch": 0.33876552053486153,
      "grad_norm": 0.7133093476295471,
      "learning_rate": 4.827067813231788e-05,
      "loss": 0.6134,
      "mean_token_accuracy": 0.8093224287033081,
      "num_tokens": 190149082.0,
      "step": 5675
    },
    {
      "epoch": 0.3390639923591213,
      "grad_norm": 0.627886176109314,
      "learning_rate": 4.8266406414802916e-05,
      "loss": 0.5919,
      "mean_token_accuracy": 0.8165155649185181,
      "num_tokens": 190316762.0,
      "step": 5680
    },
    {
      "epoch": 0.33936246418338106,
      "grad_norm": 0.5913046002388,
      "learning_rate": 4.8262129639246366e-05,
      "loss": 0.5536,
      "mean_token_accuracy": 0.8268459916114808,
      "num_tokens": 190484442.0,
      "step": 5685
    },
    {
      "epoch": 0.33966093600764086,
      "grad_norm": 0.5674902200698853,
      "learning_rate": 4.825784780668991e-05,
      "loss": 0.5862,
      "mean_token_accuracy": 0.8183526158332824,
      "num_tokens": 190652122.0,
      "step": 5690
    },
    {
      "epoch": 0.33995940783190065,
      "grad_norm": 0.5599184632301331,
      "learning_rate": 4.825356091817648e-05,
      "loss": 0.5884,
      "mean_token_accuracy": 0.8159370064735413,
      "num_tokens": 190819802.0,
      "step": 5695
    },
    {
      "epoch": 0.34025787965616044,
      "grad_norm": 0.5582353472709656,
      "learning_rate": 4.8249268974750227e-05,
      "loss": 0.5732,
      "mean_token_accuracy": 0.8200942397117614,
      "num_tokens": 190987482.0,
      "step": 5700
    },
    {
      "epoch": 0.34055635148042024,
      "grad_norm": 0.6247255206108093,
      "learning_rate": 4.824497197745653e-05,
      "loss": 0.5789,
      "mean_token_accuracy": 0.8181975364685059,
      "num_tokens": 191155162.0,
      "step": 5705
    },
    {
      "epoch": 0.34085482330468003,
      "grad_norm": 0.695816159248352,
      "learning_rate": 4.8240669927341986e-05,
      "loss": 0.5873,
      "mean_token_accuracy": 0.8188416957855225,
      "num_tokens": 191322842.0,
      "step": 5710
    },
    {
      "epoch": 0.3411532951289398,
      "grad_norm": 0.5786401033401489,
      "learning_rate": 4.823636282545446e-05,
      "loss": 0.5789,
      "mean_token_accuracy": 0.8193904399871826,
      "num_tokens": 191490522.0,
      "step": 5715
    },
    {
      "epoch": 0.3414517669531996,
      "grad_norm": 0.5475450754165649,
      "learning_rate": 4.823205067284301e-05,
      "loss": 0.5965,
      "mean_token_accuracy": 0.8146308064460754,
      "num_tokens": 191658202.0,
      "step": 5720
    },
    {
      "epoch": 0.3417502387774594,
      "grad_norm": 0.9496316313743591,
      "learning_rate": 4.822773347055796e-05,
      "loss": 0.5691,
      "mean_token_accuracy": 0.823947274684906,
      "num_tokens": 191825882.0,
      "step": 5725
    },
    {
      "epoch": 0.3420487106017192,
      "grad_norm": 0.703660786151886,
      "learning_rate": 4.8223411219650835e-05,
      "loss": 0.6503,
      "mean_token_accuracy": 0.8019205451011657,
      "num_tokens": 191993562.0,
      "step": 5730
    },
    {
      "epoch": 0.342347182425979,
      "grad_norm": 0.56575608253479,
      "learning_rate": 4.821908392117439e-05,
      "loss": 0.5818,
      "mean_token_accuracy": 0.8168614864349365,
      "num_tokens": 192161242.0,
      "step": 5735
    },
    {
      "epoch": 0.3426456542502388,
      "grad_norm": 0.5935403108596802,
      "learning_rate": 4.8214751576182626e-05,
      "loss": 0.5999,
      "mean_token_accuracy": 0.8140880346298218,
      "num_tokens": 192328922.0,
      "step": 5740
    },
    {
      "epoch": 0.3429441260744986,
      "grad_norm": 0.5520789623260498,
      "learning_rate": 4.821041418573077e-05,
      "loss": 0.5297,
      "mean_token_accuracy": 0.8318740248680114,
      "num_tokens": 192496602.0,
      "step": 5745
    },
    {
      "epoch": 0.3432425978987584,
      "grad_norm": 0.6562302708625793,
      "learning_rate": 4.820607175087526e-05,
      "loss": 0.5844,
      "mean_token_accuracy": 0.816754150390625,
      "num_tokens": 192664282.0,
      "step": 5750
    },
    {
      "epoch": 0.3435410697230181,
      "grad_norm": 0.6340336203575134,
      "learning_rate": 4.820172427267379e-05,
      "loss": 0.6331,
      "mean_token_accuracy": 0.805099618434906,
      "num_tokens": 192831962.0,
      "step": 5755
    },
    {
      "epoch": 0.3438395415472779,
      "grad_norm": 0.6429221034049988,
      "learning_rate": 4.819737175218527e-05,
      "loss": 0.6264,
      "mean_token_accuracy": 0.8077299356460571,
      "num_tokens": 192999642.0,
      "step": 5760
    },
    {
      "epoch": 0.3441380133715377,
      "grad_norm": 1.1211243867874146,
      "learning_rate": 4.8193014190469815e-05,
      "loss": 0.555,
      "mean_token_accuracy": 0.824060595035553,
      "num_tokens": 193167322.0,
      "step": 5765
    },
    {
      "epoch": 0.3444364851957975,
      "grad_norm": 0.66719651222229,
      "learning_rate": 4.818865158858882e-05,
      "loss": 0.6682,
      "mean_token_accuracy": 0.7946558475494385,
      "num_tokens": 193335002.0,
      "step": 5770
    },
    {
      "epoch": 0.3447349570200573,
      "grad_norm": 0.6126347780227661,
      "learning_rate": 4.818428394760485e-05,
      "loss": 0.5907,
      "mean_token_accuracy": 0.8166944980621338,
      "num_tokens": 193502682.0,
      "step": 5775
    },
    {
      "epoch": 0.3450334288443171,
      "grad_norm": 0.627385139465332,
      "learning_rate": 4.817991126858173e-05,
      "loss": 0.6272,
      "mean_token_accuracy": 0.8069843769073486,
      "num_tokens": 193670362.0,
      "step": 5780
    },
    {
      "epoch": 0.3453319006685769,
      "grad_norm": 0.6703435182571411,
      "learning_rate": 4.817553355258453e-05,
      "loss": 0.6911,
      "mean_token_accuracy": 0.7877192020416259,
      "num_tokens": 193838042.0,
      "step": 5785
    },
    {
      "epoch": 0.3456303724928367,
      "grad_norm": 0.5937191247940063,
      "learning_rate": 4.81711508006795e-05,
      "loss": 0.6407,
      "mean_token_accuracy": 0.8059823393821717,
      "num_tokens": 194005722.0,
      "step": 5790
    },
    {
      "epoch": 0.3459288443170965,
      "grad_norm": 0.6775105595588684,
      "learning_rate": 4.816676301393415e-05,
      "loss": 0.5873,
      "mean_token_accuracy": 0.8178337097167969,
      "num_tokens": 194173402.0,
      "step": 5795
    },
    {
      "epoch": 0.34622731614135627,
      "grad_norm": 0.6267567873001099,
      "learning_rate": 4.81623701934172e-05,
      "loss": 0.5597,
      "mean_token_accuracy": 0.8242633819580079,
      "num_tokens": 194341082.0,
      "step": 5800
    },
    {
      "epoch": 0.34652578796561606,
      "grad_norm": 0.6258201003074646,
      "learning_rate": 4.8157972340198605e-05,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8259095788002014,
      "num_tokens": 194508762.0,
      "step": 5805
    },
    {
      "epoch": 0.34682425978987586,
      "grad_norm": 0.5671982169151306,
      "learning_rate": 4.8153569455349545e-05,
      "loss": 0.5331,
      "mean_token_accuracy": 0.831414771080017,
      "num_tokens": 194676442.0,
      "step": 5810
    },
    {
      "epoch": 0.34712273161413565,
      "grad_norm": 0.6403409242630005,
      "learning_rate": 4.814916153994243e-05,
      "loss": 0.5619,
      "mean_token_accuracy": 0.8223845720291137,
      "num_tokens": 194844122.0,
      "step": 5815
    },
    {
      "epoch": 0.34742120343839544,
      "grad_norm": 0.6637452840805054,
      "learning_rate": 4.814474859505087e-05,
      "loss": 0.624,
      "mean_token_accuracy": 0.8084635496139526,
      "num_tokens": 195011802.0,
      "step": 5820
    },
    {
      "epoch": 0.3477196752626552,
      "grad_norm": 0.6146729588508606,
      "learning_rate": 4.8140330621749735e-05,
      "loss": 0.5998,
      "mean_token_accuracy": 0.8148813128471375,
      "num_tokens": 195179482.0,
      "step": 5825
    },
    {
      "epoch": 0.348018147086915,
      "grad_norm": 0.5506935715675354,
      "learning_rate": 4.8135907621115105e-05,
      "loss": 0.5396,
      "mean_token_accuracy": 0.8299057602882385,
      "num_tokens": 195347162.0,
      "step": 5830
    },
    {
      "epoch": 0.34831661891117477,
      "grad_norm": 0.612335205078125,
      "learning_rate": 4.813147959422427e-05,
      "loss": 0.5724,
      "mean_token_accuracy": 0.821591317653656,
      "num_tokens": 195514842.0,
      "step": 5835
    },
    {
      "epoch": 0.34861509073543456,
      "grad_norm": 0.6685415506362915,
      "learning_rate": 4.812704654215577e-05,
      "loss": 0.5801,
      "mean_token_accuracy": 0.818758201599121,
      "num_tokens": 195682522.0,
      "step": 5840
    },
    {
      "epoch": 0.34891356255969436,
      "grad_norm": 0.6858870387077332,
      "learning_rate": 4.812260846598936e-05,
      "loss": 0.578,
      "mean_token_accuracy": 0.8179887890815735,
      "num_tokens": 195850202.0,
      "step": 5845
    },
    {
      "epoch": 0.34921203438395415,
      "grad_norm": 0.7021246552467346,
      "learning_rate": 4.8118165366806e-05,
      "loss": 0.5741,
      "mean_token_accuracy": 0.8202373862266541,
      "num_tokens": 196017882.0,
      "step": 5850
    },
    {
      "epoch": 0.34951050620821394,
      "grad_norm": 0.6232729554176331,
      "learning_rate": 4.81137172456879e-05,
      "loss": 0.6022,
      "mean_token_accuracy": 0.8130025029182434,
      "num_tokens": 196185562.0,
      "step": 5855
    },
    {
      "epoch": 0.34980897803247374,
      "grad_norm": 1.3280320167541504,
      "learning_rate": 4.8109264103718485e-05,
      "loss": 0.5551,
      "mean_token_accuracy": 0.8264940977096558,
      "num_tokens": 196353242.0,
      "step": 5860
    },
    {
      "epoch": 0.35010744985673353,
      "grad_norm": 0.7067484855651855,
      "learning_rate": 4.810480594198239e-05,
      "loss": 0.6009,
      "mean_token_accuracy": 0.8130800485610962,
      "num_tokens": 196520922.0,
      "step": 5865
    },
    {
      "epoch": 0.3504059216809933,
      "grad_norm": 0.5749301314353943,
      "learning_rate": 4.8100342761565486e-05,
      "loss": 0.5474,
      "mean_token_accuracy": 0.8267028570175171,
      "num_tokens": 196688602.0,
      "step": 5870
    },
    {
      "epoch": 0.3507043935052531,
      "grad_norm": 0.5796326994895935,
      "learning_rate": 4.809587456355486e-05,
      "loss": 0.5897,
      "mean_token_accuracy": 0.8188715219497681,
      "num_tokens": 196856282.0,
      "step": 5875
    },
    {
      "epoch": 0.3510028653295129,
      "grad_norm": 0.6365723013877869,
      "learning_rate": 4.8091401349038826e-05,
      "loss": 0.6236,
      "mean_token_accuracy": 0.8061314463615418,
      "num_tokens": 197023962.0,
      "step": 5880
    },
    {
      "epoch": 0.3513013371537727,
      "grad_norm": 0.545111358165741,
      "learning_rate": 4.808692311910693e-05,
      "loss": 0.5416,
      "mean_token_accuracy": 0.8279911637306213,
      "num_tokens": 197191642.0,
      "step": 5885
    },
    {
      "epoch": 0.35159980897803245,
      "grad_norm": 0.6369797587394714,
      "learning_rate": 4.808243987484991e-05,
      "loss": 0.6055,
      "mean_token_accuracy": 0.8125849962234497,
      "num_tokens": 197359322.0,
      "step": 5890
    },
    {
      "epoch": 0.35189828080229224,
      "grad_norm": 0.5923847556114197,
      "learning_rate": 4.8077951617359756e-05,
      "loss": 0.6146,
      "mean_token_accuracy": 0.8082548022270203,
      "num_tokens": 197527002.0,
      "step": 5895
    },
    {
      "epoch": 0.35219675262655203,
      "grad_norm": 0.652245283126831,
      "learning_rate": 4.8073458347729656e-05,
      "loss": 0.6368,
      "mean_token_accuracy": 0.8015030503273011,
      "num_tokens": 197694682.0,
      "step": 5900
    },
    {
      "epoch": 0.35249522445081183,
      "grad_norm": 0.5620662569999695,
      "learning_rate": 4.806896006705403e-05,
      "loss": 0.5735,
      "mean_token_accuracy": 0.8208934664726257,
      "num_tokens": 197862362.0,
      "step": 5905
    },
    {
      "epoch": 0.3527936962750716,
      "grad_norm": 0.6360174417495728,
      "learning_rate": 4.806445677642852e-05,
      "loss": 0.5581,
      "mean_token_accuracy": 0.8255397796630859,
      "num_tokens": 198030042.0,
      "step": 5910
    },
    {
      "epoch": 0.3530921680993314,
      "grad_norm": 0.6562075018882751,
      "learning_rate": 4.805994847695e-05,
      "loss": 0.569,
      "mean_token_accuracy": 0.8215137839317321,
      "num_tokens": 198197722.0,
      "step": 5915
    },
    {
      "epoch": 0.3533906399235912,
      "grad_norm": 0.6920533776283264,
      "learning_rate": 4.8055435169716525e-05,
      "loss": 0.5723,
      "mean_token_accuracy": 0.8186985611915588,
      "num_tokens": 198365402.0,
      "step": 5920
    },
    {
      "epoch": 0.353689111747851,
      "grad_norm": 0.5866786241531372,
      "learning_rate": 4.80509168558274e-05,
      "loss": 0.56,
      "mean_token_accuracy": 0.8237142205238343,
      "num_tokens": 198527639.0,
      "step": 5925
    },
    {
      "epoch": 0.3539875835721108,
      "grad_norm": 0.5532835721969604,
      "learning_rate": 4.804639353638315e-05,
      "loss": 0.6067,
      "mean_token_accuracy": 0.8116366386413574,
      "num_tokens": 198695319.0,
      "step": 5930
    },
    {
      "epoch": 0.3542860553963706,
      "grad_norm": 0.5809913277626038,
      "learning_rate": 4.8041865212485515e-05,
      "loss": 0.5903,
      "mean_token_accuracy": 0.8160861253738403,
      "num_tokens": 198862999.0,
      "step": 5935
    },
    {
      "epoch": 0.3545845272206304,
      "grad_norm": 0.689055323600769,
      "learning_rate": 4.8037331885237447e-05,
      "loss": 0.6276,
      "mean_token_accuracy": 0.8048729658126831,
      "num_tokens": 199030679.0,
      "step": 5940
    },
    {
      "epoch": 0.3548829990448902,
      "grad_norm": 0.5742608904838562,
      "learning_rate": 4.8032793555743114e-05,
      "loss": 0.5871,
      "mean_token_accuracy": 0.8169748306274414,
      "num_tokens": 199198359.0,
      "step": 5945
    },
    {
      "epoch": 0.35518147086915,
      "grad_norm": 0.6180241107940674,
      "learning_rate": 4.802825022510793e-05,
      "loss": 0.5768,
      "mean_token_accuracy": 0.8204461455345153,
      "num_tokens": 199366039.0,
      "step": 5950
    },
    {
      "epoch": 0.35547994269340977,
      "grad_norm": 1.0495320558547974,
      "learning_rate": 4.802370189443849e-05,
      "loss": 0.5744,
      "mean_token_accuracy": 0.8227603673934937,
      "num_tokens": 199533719.0,
      "step": 5955
    },
    {
      "epoch": 0.3557784145176695,
      "grad_norm": 0.6751816868782043,
      "learning_rate": 4.8019148564842634e-05,
      "loss": 0.625,
      "mean_token_accuracy": 0.8050518989562988,
      "num_tokens": 199701399.0,
      "step": 5960
    },
    {
      "epoch": 0.3560768863419293,
      "grad_norm": 0.5694411396980286,
      "learning_rate": 4.8014590237429405e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.8127042770385742,
      "num_tokens": 199869079.0,
      "step": 5965
    },
    {
      "epoch": 0.3563753581661891,
      "grad_norm": 0.5997521281242371,
      "learning_rate": 4.801002691330906e-05,
      "loss": 0.6002,
      "mean_token_accuracy": 0.8116902589797974,
      "num_tokens": 200036676.0,
      "step": 5970
    },
    {
      "epoch": 0.3566738299904489,
      "grad_norm": 0.6004998683929443,
      "learning_rate": 4.80054585935931e-05,
      "loss": 0.6003,
      "mean_token_accuracy": 0.8130382895469666,
      "num_tokens": 200204356.0,
      "step": 5975
    },
    {
      "epoch": 0.3569723018147087,
      "grad_norm": 0.7464216351509094,
      "learning_rate": 4.8000885279394206e-05,
      "loss": 0.5725,
      "mean_token_accuracy": 0.8201896786689759,
      "num_tokens": 200372036.0,
      "step": 5980
    },
    {
      "epoch": 0.3572707736389685,
      "grad_norm": 0.6673312783241272,
      "learning_rate": 4.79963069718263e-05,
      "loss": 0.5677,
      "mean_token_accuracy": 0.8203149199485779,
      "num_tokens": 200539716.0,
      "step": 5985
    },
    {
      "epoch": 0.35756924546322827,
      "grad_norm": 0.5537075996398926,
      "learning_rate": 4.799172367200451e-05,
      "loss": 0.5545,
      "mean_token_accuracy": 0.824185848236084,
      "num_tokens": 200707396.0,
      "step": 5990
    },
    {
      "epoch": 0.35786771728748806,
      "grad_norm": 0.5983226299285889,
      "learning_rate": 4.798713538104519e-05,
      "loss": 0.5804,
      "mean_token_accuracy": 0.818108081817627,
      "num_tokens": 200875076.0,
      "step": 5995
    },
    {
      "epoch": 0.35816618911174786,
      "grad_norm": 0.5817372798919678,
      "learning_rate": 4.798254210006589e-05,
      "loss": 0.5443,
      "mean_token_accuracy": 0.8290826559066773,
      "num_tokens": 201042756.0,
      "step": 6000
    },
    {
      "epoch": 0.35846466093600765,
      "grad_norm": 0.5799314975738525,
      "learning_rate": 4.7977943830185406e-05,
      "loss": 0.5388,
      "mean_token_accuracy": 0.8296015620231628,
      "num_tokens": 201210436.0,
      "step": 6005
    },
    {
      "epoch": 0.35876313276026744,
      "grad_norm": 0.6342442035675049,
      "learning_rate": 4.7973340572523724e-05,
      "loss": 0.6185,
      "mean_token_accuracy": 0.8091793060302734,
      "num_tokens": 201378116.0,
      "step": 6010
    },
    {
      "epoch": 0.35906160458452724,
      "grad_norm": 0.619718611240387,
      "learning_rate": 4.7968732328202046e-05,
      "loss": 0.5994,
      "mean_token_accuracy": 0.8141953945159912,
      "num_tokens": 201545796.0,
      "step": 6015
    },
    {
      "epoch": 0.35936007640878703,
      "grad_norm": 0.5960201621055603,
      "learning_rate": 4.7964119098342804e-05,
      "loss": 0.5697,
      "mean_token_accuracy": 0.822354769706726,
      "num_tokens": 201713476.0,
      "step": 6020
    },
    {
      "epoch": 0.3596585482330468,
      "grad_norm": 0.5542184114456177,
      "learning_rate": 4.795950088406962e-05,
      "loss": 0.5548,
      "mean_token_accuracy": 0.8247345805168151,
      "num_tokens": 201881156.0,
      "step": 6025
    },
    {
      "epoch": 0.35995702005730656,
      "grad_norm": 0.5916186571121216,
      "learning_rate": 4.795487768650737e-05,
      "loss": 0.5881,
      "mean_token_accuracy": 0.8167601108551026,
      "num_tokens": 202048836.0,
      "step": 6030
    },
    {
      "epoch": 0.36025549188156636,
      "grad_norm": 0.6291908025741577,
      "learning_rate": 4.7950249506782104e-05,
      "loss": 0.6296,
      "mean_token_accuracy": 0.8032906413078308,
      "num_tokens": 202213547.0,
      "step": 6035
    },
    {
      "epoch": 0.36055396370582615,
      "grad_norm": 0.6131892204284668,
      "learning_rate": 4.794561634602109e-05,
      "loss": 0.5751,
      "mean_token_accuracy": 0.8217881441116333,
      "num_tokens": 202381227.0,
      "step": 6040
    },
    {
      "epoch": 0.36085243553008595,
      "grad_norm": 0.5936862826347351,
      "learning_rate": 4.7940978205352845e-05,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8248419404029846,
      "num_tokens": 202548907.0,
      "step": 6045
    },
    {
      "epoch": 0.36115090735434574,
      "grad_norm": 0.5770553946495056,
      "learning_rate": 4.7936335085907064e-05,
      "loss": 0.5773,
      "mean_token_accuracy": 0.8193128824234008,
      "num_tokens": 202716587.0,
      "step": 6050
    },
    {
      "epoch": 0.36144937917860553,
      "grad_norm": 0.6152001023292542,
      "learning_rate": 4.7931686988814644e-05,
      "loss": 0.5468,
      "mean_token_accuracy": 0.8282595753669739,
      "num_tokens": 202884267.0,
      "step": 6055
    },
    {
      "epoch": 0.36174785100286533,
      "grad_norm": 0.5983825922012329,
      "learning_rate": 4.792703391520775e-05,
      "loss": 0.5844,
      "mean_token_accuracy": 0.8193784952163696,
      "num_tokens": 203051947.0,
      "step": 6060
    },
    {
      "epoch": 0.3620463228271251,
      "grad_norm": 0.6283164620399475,
      "learning_rate": 4.79223758662197e-05,
      "loss": 0.5762,
      "mean_token_accuracy": 0.8198198676109314,
      "num_tokens": 203219627.0,
      "step": 6065
    },
    {
      "epoch": 0.3623447946513849,
      "grad_norm": 0.6504350900650024,
      "learning_rate": 4.791771284298505e-05,
      "loss": 0.5752,
      "mean_token_accuracy": 0.8200107336044311,
      "num_tokens": 203387307.0,
      "step": 6070
    },
    {
      "epoch": 0.3626432664756447,
      "grad_norm": 0.7191824316978455,
      "learning_rate": 4.791304484663957e-05,
      "loss": 0.5735,
      "mean_token_accuracy": 0.8221042633056641,
      "num_tokens": 203554987.0,
      "step": 6075
    },
    {
      "epoch": 0.3629417382999045,
      "grad_norm": 0.5728492140769958,
      "learning_rate": 4.7908371878320234e-05,
      "loss": 0.5967,
      "mean_token_accuracy": 0.8138315677642822,
      "num_tokens": 203722667.0,
      "step": 6080
    },
    {
      "epoch": 0.3632402101241643,
      "grad_norm": 0.5692455172538757,
      "learning_rate": 4.790369393916524e-05,
      "loss": 0.5384,
      "mean_token_accuracy": 0.8291065335273743,
      "num_tokens": 203890347.0,
      "step": 6085
    },
    {
      "epoch": 0.3635386819484241,
      "grad_norm": 0.5722931623458862,
      "learning_rate": 4.789901103031397e-05,
      "loss": 0.5352,
      "mean_token_accuracy": 0.8326374769210816,
      "num_tokens": 204058027.0,
      "step": 6090
    },
    {
      "epoch": 0.3638371537726839,
      "grad_norm": 0.6488029956817627,
      "learning_rate": 4.789432315290705e-05,
      "loss": 0.5678,
      "mean_token_accuracy": 0.8214243054389954,
      "num_tokens": 204225707.0,
      "step": 6095
    },
    {
      "epoch": 0.3641356255969436,
      "grad_norm": 0.7144589424133301,
      "learning_rate": 4.788963030808628e-05,
      "loss": 0.5958,
      "mean_token_accuracy": 0.8143087267875672,
      "num_tokens": 204393387.0,
      "step": 6100
    },
    {
      "epoch": 0.3644340974212034,
      "grad_norm": 0.6138399839401245,
      "learning_rate": 4.78849324969947e-05,
      "loss": 0.5498,
      "mean_token_accuracy": 0.8279136300086976,
      "num_tokens": 204561067.0,
      "step": 6105
    },
    {
      "epoch": 0.3647325692454632,
      "grad_norm": 0.5629082322120667,
      "learning_rate": 4.7880229720776556e-05,
      "loss": 0.5881,
      "mean_token_accuracy": 0.8167958855628967,
      "num_tokens": 204728747.0,
      "step": 6110
    },
    {
      "epoch": 0.365031041069723,
      "grad_norm": 0.6037213206291199,
      "learning_rate": 4.787552198057728e-05,
      "loss": 0.5968,
      "mean_token_accuracy": 0.8123762488365174,
      "num_tokens": 204896427.0,
      "step": 6115
    },
    {
      "epoch": 0.3653295128939828,
      "grad_norm": 0.5730690360069275,
      "learning_rate": 4.787080927754354e-05,
      "loss": 0.543,
      "mean_token_accuracy": 0.8294703722000122,
      "num_tokens": 205064107.0,
      "step": 6120
    },
    {
      "epoch": 0.3656279847182426,
      "grad_norm": 0.5808693170547485,
      "learning_rate": 4.78660916128232e-05,
      "loss": 0.6218,
      "mean_token_accuracy": 0.8055827260017395,
      "num_tokens": 205231787.0,
      "step": 6125
    },
    {
      "epoch": 0.3659264565425024,
      "grad_norm": 0.5321030616760254,
      "learning_rate": 4.786136898756533e-05,
      "loss": 0.5775,
      "mean_token_accuracy": 0.8188536286354064,
      "num_tokens": 205399467.0,
      "step": 6130
    },
    {
      "epoch": 0.3662249283667622,
      "grad_norm": 0.5482165813446045,
      "learning_rate": 4.785664140292021e-05,
      "loss": 0.6068,
      "mean_token_accuracy": 0.8114696383476258,
      "num_tokens": 205567147.0,
      "step": 6135
    },
    {
      "epoch": 0.366523400191022,
      "grad_norm": 0.5584302544593811,
      "learning_rate": 4.785190886003934e-05,
      "loss": 0.5869,
      "mean_token_accuracy": 0.8161099672317504,
      "num_tokens": 205734827.0,
      "step": 6140
    },
    {
      "epoch": 0.36682187201528177,
      "grad_norm": 0.5813570618629456,
      "learning_rate": 4.784717136007542e-05,
      "loss": 0.584,
      "mean_token_accuracy": 0.8184301614761352,
      "num_tokens": 205902507.0,
      "step": 6145
    },
    {
      "epoch": 0.36712034383954156,
      "grad_norm": 0.57797771692276,
      "learning_rate": 4.7842428904182346e-05,
      "loss": 0.5445,
      "mean_token_accuracy": 0.8281761169433594,
      "num_tokens": 206065792.0,
      "step": 6150
    },
    {
      "epoch": 0.36741881566380136,
      "grad_norm": 0.6327470541000366,
      "learning_rate": 4.783768149351523e-05,
      "loss": 0.6095,
      "mean_token_accuracy": 0.8096922397613525,
      "num_tokens": 206233472.0,
      "step": 6155
    },
    {
      "epoch": 0.36771728748806115,
      "grad_norm": 0.5569842457771301,
      "learning_rate": 4.78329291292304e-05,
      "loss": 0.5963,
      "mean_token_accuracy": 0.8137421011924744,
      "num_tokens": 206401152.0,
      "step": 6160
    },
    {
      "epoch": 0.3680157593123209,
      "grad_norm": 0.6722360849380493,
      "learning_rate": 4.782817181248537e-05,
      "loss": 0.5768,
      "mean_token_accuracy": 0.8195932149887085,
      "num_tokens": 206568832.0,
      "step": 6165
    },
    {
      "epoch": 0.3683142311365807,
      "grad_norm": 0.6275756359100342,
      "learning_rate": 4.782340954443889e-05,
      "loss": 0.584,
      "mean_token_accuracy": 0.8166110038757324,
      "num_tokens": 206736512.0,
      "step": 6170
    },
    {
      "epoch": 0.3686127029608405,
      "grad_norm": 0.6040167212486267,
      "learning_rate": 4.7818642326250886e-05,
      "loss": 0.5611,
      "mean_token_accuracy": 0.8252415657043457,
      "num_tokens": 206904192.0,
      "step": 6175
    },
    {
      "epoch": 0.36891117478510027,
      "grad_norm": 0.5967251062393188,
      "learning_rate": 4.78138701590825e-05,
      "loss": 0.5449,
      "mean_token_accuracy": 0.8273828029632568,
      "num_tokens": 207071872.0,
      "step": 6180
    },
    {
      "epoch": 0.36920964660936006,
      "grad_norm": 0.5543979406356812,
      "learning_rate": 4.780909304409609e-05,
      "loss": 0.5373,
      "mean_token_accuracy": 0.8296134948730469,
      "num_tokens": 207239552.0,
      "step": 6185
    },
    {
      "epoch": 0.36950811843361986,
      "grad_norm": 0.5790201425552368,
      "learning_rate": 4.7804310982455206e-05,
      "loss": 0.6047,
      "mean_token_accuracy": 0.8101813077926636,
      "num_tokens": 207407232.0,
      "step": 6190
    },
    {
      "epoch": 0.36980659025787965,
      "grad_norm": 0.6514462232589722,
      "learning_rate": 4.779952397532461e-05,
      "loss": 0.6143,
      "mean_token_accuracy": 0.8098652124404907,
      "num_tokens": 207574912.0,
      "step": 6195
    },
    {
      "epoch": 0.37010506208213945,
      "grad_norm": 0.553922712802887,
      "learning_rate": 4.779473202387026e-05,
      "loss": 0.6114,
      "mean_token_accuracy": 0.8086484551429749,
      "num_tokens": 207742592.0,
      "step": 6200
    },
    {
      "epoch": 0.37040353390639924,
      "grad_norm": 0.6184931397438049,
      "learning_rate": 4.7789935129259325e-05,
      "loss": 0.6002,
      "mean_token_accuracy": 0.8134378910064697,
      "num_tokens": 207910272.0,
      "step": 6205
    },
    {
      "epoch": 0.37070200573065903,
      "grad_norm": 0.599904477596283,
      "learning_rate": 4.778513329266018e-05,
      "loss": 0.5428,
      "mean_token_accuracy": 0.827180016040802,
      "num_tokens": 208077952.0,
      "step": 6210
    },
    {
      "epoch": 0.37100047755491883,
      "grad_norm": 0.5533261299133301,
      "learning_rate": 4.778032651524241e-05,
      "loss": 0.547,
      "mean_token_accuracy": 0.8287009477615357,
      "num_tokens": 208245632.0,
      "step": 6215
    },
    {
      "epoch": 0.3712989493791786,
      "grad_norm": 0.63448166847229,
      "learning_rate": 4.777551479817677e-05,
      "loss": 0.6222,
      "mean_token_accuracy": 0.8068472027778626,
      "num_tokens": 208413312.0,
      "step": 6220
    },
    {
      "epoch": 0.3715974212034384,
      "grad_norm": 0.593621015548706,
      "learning_rate": 4.7770698142635274e-05,
      "loss": 0.5822,
      "mean_token_accuracy": 0.8181200146675109,
      "num_tokens": 208580992.0,
      "step": 6225
    },
    {
      "epoch": 0.3718958930276982,
      "grad_norm": 0.603478729724884,
      "learning_rate": 4.776587654979108e-05,
      "loss": 0.5902,
      "mean_token_accuracy": 0.8145592331886291,
      "num_tokens": 208748672.0,
      "step": 6230
    },
    {
      "epoch": 0.37219436485195795,
      "grad_norm": 0.6217207908630371,
      "learning_rate": 4.776105002081859e-05,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8302874684333801,
      "num_tokens": 208916352.0,
      "step": 6235
    },
    {
      "epoch": 0.37249283667621774,
      "grad_norm": 0.5804060101509094,
      "learning_rate": 4.7756218556893396e-05,
      "loss": 0.5486,
      "mean_token_accuracy": 0.826667058467865,
      "num_tokens": 209084032.0,
      "step": 6240
    },
    {
      "epoch": 0.37279130850047754,
      "grad_norm": 0.5571052432060242,
      "learning_rate": 4.7751382159192276e-05,
      "loss": 0.5739,
      "mean_token_accuracy": 0.8195097208023071,
      "num_tokens": 209251712.0,
      "step": 6245
    },
    {
      "epoch": 0.37308978032473733,
      "grad_norm": 1.8832085132598877,
      "learning_rate": 4.7746540828893246e-05,
      "loss": 0.5994,
      "mean_token_accuracy": 0.8152332186698914,
      "num_tokens": 209419392.0,
      "step": 6250
    },
    {
      "epoch": 0.3733882521489971,
      "grad_norm": 0.6478602886199951,
      "learning_rate": 4.774169456717548e-05,
      "loss": 0.5896,
      "mean_token_accuracy": 0.8154598593711853,
      "num_tokens": 209587072.0,
      "step": 6255
    },
    {
      "epoch": 0.3736867239732569,
      "grad_norm": 0.5612985491752625,
      "learning_rate": 4.7736843375219384e-05,
      "loss": 0.5836,
      "mean_token_accuracy": 0.8176607370376587,
      "num_tokens": 209754752.0,
      "step": 6260
    },
    {
      "epoch": 0.3739851957975167,
      "grad_norm": 0.5482649207115173,
      "learning_rate": 4.7731987254206555e-05,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8265477776527405,
      "num_tokens": 209922432.0,
      "step": 6265
    },
    {
      "epoch": 0.3742836676217765,
      "grad_norm": 0.5954232215881348,
      "learning_rate": 4.7727126205319794e-05,
      "loss": 0.5789,
      "mean_token_accuracy": 0.8219231247901917,
      "num_tokens": 210084696.0,
      "step": 6270
    },
    {
      "epoch": 0.3745821394460363,
      "grad_norm": 0.5713602304458618,
      "learning_rate": 4.772226022974309e-05,
      "loss": 0.5537,
      "mean_token_accuracy": 0.825426459312439,
      "num_tokens": 210252376.0,
      "step": 6275
    },
    {
      "epoch": 0.3748806112702961,
      "grad_norm": 0.5526502728462219,
      "learning_rate": 4.771738932866165e-05,
      "loss": 0.5746,
      "mean_token_accuracy": 0.8212215185165406,
      "num_tokens": 210420056.0,
      "step": 6280
    },
    {
      "epoch": 0.3751790830945559,
      "grad_norm": 0.5443446636199951,
      "learning_rate": 4.771251350326187e-05,
      "loss": 0.5628,
      "mean_token_accuracy": 0.8232554078102112,
      "num_tokens": 210587736.0,
      "step": 6285
    },
    {
      "epoch": 0.3754775549188157,
      "grad_norm": 0.5898823738098145,
      "learning_rate": 4.7707632754731345e-05,
      "loss": 0.5655,
      "mean_token_accuracy": 0.8227961301803589,
      "num_tokens": 210755416.0,
      "step": 6290
    },
    {
      "epoch": 0.3757760267430755,
      "grad_norm": 0.5943527221679688,
      "learning_rate": 4.770274708425887e-05,
      "loss": 0.62,
      "mean_token_accuracy": 0.8066026449203492,
      "num_tokens": 210923096.0,
      "step": 6295
    },
    {
      "epoch": 0.37607449856733527,
      "grad_norm": 0.5689024329185486,
      "learning_rate": 4.769785649303445e-05,
      "loss": 0.547,
      "mean_token_accuracy": 0.8284027218818665,
      "num_tokens": 211090776.0,
      "step": 6300
    },
    {
      "epoch": 0.376372970391595,
      "grad_norm": 0.5758464336395264,
      "learning_rate": 4.7692960982249265e-05,
      "loss": 0.5664,
      "mean_token_accuracy": 0.8225813984870911,
      "num_tokens": 211258456.0,
      "step": 6305
    },
    {
      "epoch": 0.3766714422158548,
      "grad_norm": 0.5908043384552002,
      "learning_rate": 4.768806055309571e-05,
      "loss": 0.6244,
      "mean_token_accuracy": 0.804723858833313,
      "num_tokens": 211426136.0,
      "step": 6310
    },
    {
      "epoch": 0.3769699140401146,
      "grad_norm": 0.6135730147361755,
      "learning_rate": 4.7683155206767385e-05,
      "loss": 0.5471,
      "mean_token_accuracy": 0.8274305105209351,
      "num_tokens": 211593816.0,
      "step": 6315
    },
    {
      "epoch": 0.3772683858643744,
      "grad_norm": 0.5765379667282104,
      "learning_rate": 4.7678244944459064e-05,
      "loss": 0.5968,
      "mean_token_accuracy": 0.814010500907898,
      "num_tokens": 211761496.0,
      "step": 6320
    },
    {
      "epoch": 0.3775668576886342,
      "grad_norm": 0.6675630211830139,
      "learning_rate": 4.767332976736675e-05,
      "loss": 0.6126,
      "mean_token_accuracy": 0.8097280144691468,
      "num_tokens": 211929176.0,
      "step": 6325
    },
    {
      "epoch": 0.377865329512894,
      "grad_norm": 0.5652087330818176,
      "learning_rate": 4.766840967668761e-05,
      "loss": 0.5985,
      "mean_token_accuracy": 0.8130979418754578,
      "num_tokens": 212096856.0,
      "step": 6330
    },
    {
      "epoch": 0.37816380133715377,
      "grad_norm": 0.597262978553772,
      "learning_rate": 4.7663484673620026e-05,
      "loss": 0.5925,
      "mean_token_accuracy": 0.8139210343360901,
      "num_tokens": 212264536.0,
      "step": 6335
    },
    {
      "epoch": 0.37846227316141356,
      "grad_norm": 0.5369839668273926,
      "learning_rate": 4.765855475936357e-05,
      "loss": 0.5344,
      "mean_token_accuracy": 0.830138373374939,
      "num_tokens": 212432216.0,
      "step": 6340
    },
    {
      "epoch": 0.37876074498567336,
      "grad_norm": 0.5923442840576172,
      "learning_rate": 4.7653619935119027e-05,
      "loss": 0.5988,
      "mean_token_accuracy": 0.8131396889686584,
      "num_tokens": 212599896.0,
      "step": 6345
    },
    {
      "epoch": 0.37905921680993315,
      "grad_norm": 0.6336514353752136,
      "learning_rate": 4.764868020208835e-05,
      "loss": 0.6201,
      "mean_token_accuracy": 0.808535122871399,
      "num_tokens": 212767576.0,
      "step": 6350
    },
    {
      "epoch": 0.37935768863419295,
      "grad_norm": 0.5950546860694885,
      "learning_rate": 4.76437355614747e-05,
      "loss": 0.6075,
      "mean_token_accuracy": 0.81177978515625,
      "num_tokens": 212935256.0,
      "step": 6355
    },
    {
      "epoch": 0.37965616045845274,
      "grad_norm": 0.5491812229156494,
      "learning_rate": 4.763878601448246e-05,
      "loss": 0.6138,
      "mean_token_accuracy": 0.8096564650535584,
      "num_tokens": 213098151.0,
      "step": 6360
    },
    {
      "epoch": 0.37995463228271253,
      "grad_norm": 0.5339094400405884,
      "learning_rate": 4.763383156231716e-05,
      "loss": 0.5647,
      "mean_token_accuracy": 0.8233806610107421,
      "num_tokens": 213265831.0,
      "step": 6365
    },
    {
      "epoch": 0.38025310410697233,
      "grad_norm": 0.5538163185119629,
      "learning_rate": 4.762887220618555e-05,
      "loss": 0.576,
      "mean_token_accuracy": 0.8189192414283752,
      "num_tokens": 213433511.0,
      "step": 6370
    },
    {
      "epoch": 0.38055157593123207,
      "grad_norm": 0.6217353940010071,
      "learning_rate": 4.762390794729558e-05,
      "loss": 0.5597,
      "mean_token_accuracy": 0.8234402894973755,
      "num_tokens": 213601191.0,
      "step": 6375
    },
    {
      "epoch": 0.38085004775549186,
      "grad_norm": 0.5830212235450745,
      "learning_rate": 4.761893878685639e-05,
      "loss": 0.5635,
      "mean_token_accuracy": 0.8230645298957825,
      "num_tokens": 213768871.0,
      "step": 6380
    },
    {
      "epoch": 0.38114851957975165,
      "grad_norm": 0.7095737457275391,
      "learning_rate": 4.76139647260783e-05,
      "loss": 0.6056,
      "mean_token_accuracy": 0.8110700249671936,
      "num_tokens": 213936551.0,
      "step": 6385
    },
    {
      "epoch": 0.38144699140401145,
      "grad_norm": 0.6289616823196411,
      "learning_rate": 4.7608985766172834e-05,
      "loss": 0.5641,
      "mean_token_accuracy": 0.8217463850975036,
      "num_tokens": 214104231.0,
      "step": 6390
    },
    {
      "epoch": 0.38174546322827124,
      "grad_norm": 0.6377663016319275,
      "learning_rate": 4.760400190835273e-05,
      "loss": 0.6435,
      "mean_token_accuracy": 0.8008230924606323,
      "num_tokens": 214271911.0,
      "step": 6395
    },
    {
      "epoch": 0.38204393505253104,
      "grad_norm": 0.5742936134338379,
      "learning_rate": 4.7599013153831875e-05,
      "loss": 0.5704,
      "mean_token_accuracy": 0.8208159446716309,
      "num_tokens": 214439591.0,
      "step": 6400
    },
    {
      "epoch": 0.38234240687679083,
      "grad_norm": 0.548237144947052,
      "learning_rate": 4.759401950382538e-05,
      "loss": 0.5801,
      "mean_token_accuracy": 0.8183406949043274,
      "num_tokens": 214607271.0,
      "step": 6405
    },
    {
      "epoch": 0.3826408787010506,
      "grad_norm": 0.6111705899238586,
      "learning_rate": 4.758902095954954e-05,
      "loss": 0.5866,
      "mean_token_accuracy": 0.8150065660476684,
      "num_tokens": 214774951.0,
      "step": 6410
    },
    {
      "epoch": 0.3829393505253104,
      "grad_norm": 0.6521761417388916,
      "learning_rate": 4.758401752222185e-05,
      "loss": 0.6238,
      "mean_token_accuracy": 0.80608971118927,
      "num_tokens": 214942631.0,
      "step": 6415
    },
    {
      "epoch": 0.3832378223495702,
      "grad_norm": 0.5645994544029236,
      "learning_rate": 4.757900919306098e-05,
      "loss": 0.5575,
      "mean_token_accuracy": 0.8229870080947876,
      "num_tokens": 215110311.0,
      "step": 6420
    },
    {
      "epoch": 0.38353629417383,
      "grad_norm": 0.6383602023124695,
      "learning_rate": 4.757399597328681e-05,
      "loss": 0.6136,
      "mean_token_accuracy": 0.8083502411842346,
      "num_tokens": 215277991.0,
      "step": 6425
    },
    {
      "epoch": 0.3838347659980898,
      "grad_norm": 0.5976201891899109,
      "learning_rate": 4.7568977864120395e-05,
      "loss": 0.6209,
      "mean_token_accuracy": 0.8083502411842346,
      "num_tokens": 215445671.0,
      "step": 6430
    },
    {
      "epoch": 0.3841332378223496,
      "grad_norm": 0.574621856212616,
      "learning_rate": 4.756395486678399e-05,
      "loss": 0.5663,
      "mean_token_accuracy": 0.8204580664634704,
      "num_tokens": 215613351.0,
      "step": 6435
    },
    {
      "epoch": 0.3844317096466094,
      "grad_norm": 0.6360025405883789,
      "learning_rate": 4.755892698250104e-05,
      "loss": 0.5734,
      "mean_token_accuracy": 0.8212871313095093,
      "num_tokens": 215781031.0,
      "step": 6440
    },
    {
      "epoch": 0.3847301814708691,
      "grad_norm": 0.5911943912506104,
      "learning_rate": 4.755389421249618e-05,
      "loss": 0.5859,
      "mean_token_accuracy": 0.8161934852600098,
      "num_tokens": 215948711.0,
      "step": 6445
    },
    {
      "epoch": 0.3850286532951289,
      "grad_norm": 0.6358640789985657,
      "learning_rate": 4.7548856557995234e-05,
      "loss": 0.5775,
      "mean_token_accuracy": 0.8203805208206176,
      "num_tokens": 216116391.0,
      "step": 6450
    },
    {
      "epoch": 0.3853271251193887,
      "grad_norm": 0.546363890171051,
      "learning_rate": 4.75438140202252e-05,
      "loss": 0.5743,
      "mean_token_accuracy": 0.8201717615127564,
      "num_tokens": 216284071.0,
      "step": 6455
    },
    {
      "epoch": 0.3856255969436485,
      "grad_norm": 0.6215122938156128,
      "learning_rate": 4.7538766600414304e-05,
      "loss": 0.5798,
      "mean_token_accuracy": 0.8191995739936828,
      "num_tokens": 216451751.0,
      "step": 6460
    },
    {
      "epoch": 0.3859240687679083,
      "grad_norm": 0.5524929165840149,
      "learning_rate": 4.753371429979192e-05,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8284802556037902,
      "num_tokens": 216619431.0,
      "step": 6465
    },
    {
      "epoch": 0.3862225405921681,
      "grad_norm": 0.5880683064460754,
      "learning_rate": 4.752865711958866e-05,
      "loss": 0.6087,
      "mean_token_accuracy": 0.8109447598457337,
      "num_tokens": 216787111.0,
      "step": 6470
    },
    {
      "epoch": 0.3865210124164279,
      "grad_norm": 0.5194839239120483,
      "learning_rate": 4.752359506103626e-05,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8333591818809509,
      "num_tokens": 216954791.0,
      "step": 6475
    },
    {
      "epoch": 0.3868194842406877,
      "grad_norm": 0.6575645208358765,
      "learning_rate": 4.7518528125367685e-05,
      "loss": 0.6105,
      "mean_token_accuracy": 0.808320415019989,
      "num_tokens": 217122471.0,
      "step": 6480
    },
    {
      "epoch": 0.3871179560649475,
      "grad_norm": 0.6227360367774963,
      "learning_rate": 4.751345631381709e-05,
      "loss": 0.6125,
      "mean_token_accuracy": 0.8094178676605225,
      "num_tokens": 217290151.0,
      "step": 6485
    },
    {
      "epoch": 0.38741642788920727,
      "grad_norm": 0.5925331711769104,
      "learning_rate": 4.7508379627619806e-05,
      "loss": 0.6085,
      "mean_token_accuracy": 0.8091971755027771,
      "num_tokens": 217457831.0,
      "step": 6490
    },
    {
      "epoch": 0.38771489971346706,
      "grad_norm": 0.5787959098815918,
      "learning_rate": 4.750329806801234e-05,
      "loss": 0.5641,
      "mean_token_accuracy": 0.8212453842163085,
      "num_tokens": 217625511.0,
      "step": 6495
    },
    {
      "epoch": 0.38801337153772686,
      "grad_norm": 0.5762561559677124,
      "learning_rate": 4.749821163623242e-05,
      "loss": 0.5683,
      "mean_token_accuracy": 0.8214004516601563,
      "num_tokens": 217793191.0,
      "step": 6500
    },
    {
      "epoch": 0.38831184336198665,
      "grad_norm": 0.580866813659668,
      "learning_rate": 4.7493120333518934e-05,
      "loss": 0.5739,
      "mean_token_accuracy": 0.8196469068527221,
      "num_tokens": 217960871.0,
      "step": 6505
    },
    {
      "epoch": 0.3886103151862464,
      "grad_norm": 0.6367342472076416,
      "learning_rate": 4.748802416111196e-05,
      "loss": 0.5974,
      "mean_token_accuracy": 0.811803650856018,
      "num_tokens": 218128551.0,
      "step": 6510
    },
    {
      "epoch": 0.3889087870105062,
      "grad_norm": 0.5629777312278748,
      "learning_rate": 4.7482923120252754e-05,
      "loss": 0.5564,
      "mean_token_accuracy": 0.8240427136421203,
      "num_tokens": 218296231.0,
      "step": 6515
    },
    {
      "epoch": 0.389207258834766,
      "grad_norm": 0.6516347527503967,
      "learning_rate": 4.747781721218379e-05,
      "loss": 0.6249,
      "mean_token_accuracy": 0.8065803408622741,
      "num_tokens": 218461234.0,
      "step": 6520
    },
    {
      "epoch": 0.3895057306590258,
      "grad_norm": 0.5853041410446167,
      "learning_rate": 4.747270643814869e-05,
      "loss": 0.6331,
      "mean_token_accuracy": 0.8044912338256835,
      "num_tokens": 218628914.0,
      "step": 6525
    },
    {
      "epoch": 0.38980420248328557,
      "grad_norm": 0.5444105267524719,
      "learning_rate": 4.746759079939229e-05,
      "loss": 0.5686,
      "mean_token_accuracy": 0.8207026243209838,
      "num_tokens": 218796594.0,
      "step": 6530
    },
    {
      "epoch": 0.39010267430754536,
      "grad_norm": 0.627741813659668,
      "learning_rate": 4.746247029716059e-05,
      "loss": 0.5789,
      "mean_token_accuracy": 0.817857563495636,
      "num_tokens": 218964274.0,
      "step": 6535
    },
    {
      "epoch": 0.39040114613180515,
      "grad_norm": 0.641633927822113,
      "learning_rate": 4.745734493270077e-05,
      "loss": 0.604,
      "mean_token_accuracy": 0.812417995929718,
      "num_tokens": 219131954.0,
      "step": 6540
    },
    {
      "epoch": 0.39069961795606495,
      "grad_norm": 0.7165398001670837,
      "learning_rate": 4.745221470726124e-05,
      "loss": 0.5866,
      "mean_token_accuracy": 0.8165394306182862,
      "num_tokens": 219299634.0,
      "step": 6545
    },
    {
      "epoch": 0.39099808978032474,
      "grad_norm": 0.6099076271057129,
      "learning_rate": 4.7447079622091535e-05,
      "loss": 0.5649,
      "mean_token_accuracy": 0.8223189830780029,
      "num_tokens": 219467314.0,
      "step": 6550
    },
    {
      "epoch": 0.39129656160458454,
      "grad_norm": 0.5561268329620361,
      "learning_rate": 4.744193967844241e-05,
      "loss": 0.5832,
      "mean_token_accuracy": 0.8191160678863525,
      "num_tokens": 219634994.0,
      "step": 6555
    },
    {
      "epoch": 0.39159503342884433,
      "grad_norm": 0.6390789151191711,
      "learning_rate": 4.7436794877565784e-05,
      "loss": 0.5905,
      "mean_token_accuracy": 0.8159906983375549,
      "num_tokens": 219802674.0,
      "step": 6560
    },
    {
      "epoch": 0.3918935052531041,
      "grad_norm": 0.604485034942627,
      "learning_rate": 4.7431645220714775e-05,
      "loss": 0.5844,
      "mean_token_accuracy": 0.8169330954551697,
      "num_tokens": 219970354.0,
      "step": 6565
    },
    {
      "epoch": 0.3921919770773639,
      "grad_norm": 0.6159762740135193,
      "learning_rate": 4.742649070914368e-05,
      "loss": 0.5583,
      "mean_token_accuracy": 0.8233389019966125,
      "num_tokens": 220138034.0,
      "step": 6570
    },
    {
      "epoch": 0.3924904489016237,
      "grad_norm": 0.6343648433685303,
      "learning_rate": 4.742133134410797e-05,
      "loss": 0.5981,
      "mean_token_accuracy": 0.8126863956451416,
      "num_tokens": 220305714.0,
      "step": 6575
    },
    {
      "epoch": 0.39278892072588345,
      "grad_norm": 0.5860236287117004,
      "learning_rate": 4.741616712686431e-05,
      "loss": 0.6068,
      "mean_token_accuracy": 0.8143018960952759,
      "num_tokens": 220466307.0,
      "step": 6580
    },
    {
      "epoch": 0.39308739255014324,
      "grad_norm": 0.5698738694190979,
      "learning_rate": 4.741099805867053e-05,
      "loss": 0.531,
      "mean_token_accuracy": 0.8326016902923584,
      "num_tokens": 220633987.0,
      "step": 6585
    },
    {
      "epoch": 0.39338586437440304,
      "grad_norm": 0.5755841135978699,
      "learning_rate": 4.740582414078566e-05,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8273887634277344,
      "num_tokens": 220801667.0,
      "step": 6590
    },
    {
      "epoch": 0.39368433619866283,
      "grad_norm": 0.5453025698661804,
      "learning_rate": 4.7400645374469896e-05,
      "loss": 0.5596,
      "mean_token_accuracy": 0.8226649045944214,
      "num_tokens": 220969347.0,
      "step": 6595
    },
    {
      "epoch": 0.3939828080229226,
      "grad_norm": 0.5659153461456299,
      "learning_rate": 4.739546176098464e-05,
      "loss": 0.579,
      "mean_token_accuracy": 0.8178933501243592,
      "num_tokens": 221137027.0,
      "step": 6600
    },
    {
      "epoch": 0.3942812798471824,
      "grad_norm": 0.5193234086036682,
      "learning_rate": 4.7390273301592436e-05,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8320708513259888,
      "num_tokens": 221304707.0,
      "step": 6605
    },
    {
      "epoch": 0.3945797516714422,
      "grad_norm": 0.6077717542648315,
      "learning_rate": 4.738507999755704e-05,
      "loss": 0.6023,
      "mean_token_accuracy": 0.8145830869674683,
      "num_tokens": 221472387.0,
      "step": 6610
    },
    {
      "epoch": 0.394878223495702,
      "grad_norm": 0.5752969980239868,
      "learning_rate": 4.737988185014337e-05,
      "loss": 0.5823,
      "mean_token_accuracy": 0.8173061609268188,
      "num_tokens": 221637199.0,
      "step": 6615
    },
    {
      "epoch": 0.3951766953199618,
      "grad_norm": 0.541581928730011,
      "learning_rate": 4.737467886061753e-05,
      "loss": 0.5808,
      "mean_token_accuracy": 0.817368483543396,
      "num_tokens": 221804879.0,
      "step": 6620
    },
    {
      "epoch": 0.3954751671442216,
      "grad_norm": 0.7009362578392029,
      "learning_rate": 4.736947103024682e-05,
      "loss": 0.6138,
      "mean_token_accuracy": 0.8095013737678528,
      "num_tokens": 221972559.0,
      "step": 6625
    },
    {
      "epoch": 0.3957736389684814,
      "grad_norm": 0.5366577506065369,
      "learning_rate": 4.7364258360299696e-05,
      "loss": 0.5713,
      "mean_token_accuracy": 0.8189550399780273,
      "num_tokens": 222140239.0,
      "step": 6630
    },
    {
      "epoch": 0.3960721107927412,
      "grad_norm": 0.6163941621780396,
      "learning_rate": 4.7359040852045784e-05,
      "loss": 0.5877,
      "mean_token_accuracy": 0.8163246989250184,
      "num_tokens": 222307919.0,
      "step": 6635
    },
    {
      "epoch": 0.396370582617001,
      "grad_norm": 0.6006356477737427,
      "learning_rate": 4.735381850675592e-05,
      "loss": 0.5771,
      "mean_token_accuracy": 0.8192174673080445,
      "num_tokens": 222475599.0,
      "step": 6640
    },
    {
      "epoch": 0.39666905444126077,
      "grad_norm": 0.5882141590118408,
      "learning_rate": 4.7348591325702096e-05,
      "loss": 0.633,
      "mean_token_accuracy": 0.8048968195915223,
      "num_tokens": 222643279.0,
      "step": 6645
    },
    {
      "epoch": 0.3969675262655205,
      "grad_norm": 0.725095808506012,
      "learning_rate": 4.73433593101575e-05,
      "loss": 0.5831,
      "mean_token_accuracy": 0.8178635358810424,
      "num_tokens": 222810959.0,
      "step": 6650
    },
    {
      "epoch": 0.3972659980897803,
      "grad_norm": 0.5529491305351257,
      "learning_rate": 4.733812246139647e-05,
      "loss": 0.579,
      "mean_token_accuracy": 0.8188297748565674,
      "num_tokens": 222978639.0,
      "step": 6655
    },
    {
      "epoch": 0.3975644699140401,
      "grad_norm": 0.5802292823791504,
      "learning_rate": 4.733288078069455e-05,
      "loss": 0.5735,
      "mean_token_accuracy": 0.8197721600532532,
      "num_tokens": 223146319.0,
      "step": 6660
    },
    {
      "epoch": 0.3978629417382999,
      "grad_norm": 0.7063968777656555,
      "learning_rate": 4.732763426932844e-05,
      "loss": 0.5993,
      "mean_token_accuracy": 0.8129905819892883,
      "num_tokens": 223313999.0,
      "step": 6665
    },
    {
      "epoch": 0.3981614135625597,
      "grad_norm": 0.6847906708717346,
      "learning_rate": 4.7322382928576033e-05,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8226708889007568,
      "num_tokens": 223481679.0,
      "step": 6670
    },
    {
      "epoch": 0.3984598853868195,
      "grad_norm": 0.5250869393348694,
      "learning_rate": 4.731712675971637e-05,
      "loss": 0.5679,
      "mean_token_accuracy": 0.8215734362602234,
      "num_tokens": 223649359.0,
      "step": 6675
    },
    {
      "epoch": 0.3987583572110793,
      "grad_norm": 0.5753973722457886,
      "learning_rate": 4.7311865764029716e-05,
      "loss": 0.5904,
      "mean_token_accuracy": 0.8140641689300537,
      "num_tokens": 223817039.0,
      "step": 6680
    },
    {
      "epoch": 0.39905682903533907,
      "grad_norm": 0.6264418363571167,
      "learning_rate": 4.730659994279747e-05,
      "loss": 0.6066,
      "mean_token_accuracy": 0.8108910918235779,
      "num_tokens": 223984719.0,
      "step": 6685
    },
    {
      "epoch": 0.39935530085959886,
      "grad_norm": 0.6425790786743164,
      "learning_rate": 4.730132929730221e-05,
      "loss": 0.6017,
      "mean_token_accuracy": 0.8102588653564453,
      "num_tokens": 224152399.0,
      "step": 6690
    },
    {
      "epoch": 0.39965377268385865,
      "grad_norm": 0.612373948097229,
      "learning_rate": 4.7296053828827716e-05,
      "loss": 0.6163,
      "mean_token_accuracy": 0.8088214278221131,
      "num_tokens": 224320079.0,
      "step": 6695
    },
    {
      "epoch": 0.39995224450811845,
      "grad_norm": 0.5595587491989136,
      "learning_rate": 4.7290773538658925e-05,
      "loss": 0.6089,
      "mean_token_accuracy": 0.8089130282402038,
      "num_tokens": 224483175.0,
      "step": 6700
    },
    {
      "epoch": 0.40025071633237824,
      "grad_norm": 0.555931806564331,
      "learning_rate": 4.7285488428081934e-05,
      "loss": 0.5823,
      "mean_token_accuracy": 0.8164738297462464,
      "num_tokens": 224650855.0,
      "step": 6705
    },
    {
      "epoch": 0.40054918815663804,
      "grad_norm": 0.6434721946716309,
      "learning_rate": 4.728019849838404e-05,
      "loss": 0.5982,
      "mean_token_accuracy": 0.8130502343177796,
      "num_tokens": 224818535.0,
      "step": 6710
    },
    {
      "epoch": 0.40084765998089783,
      "grad_norm": 0.6346591114997864,
      "learning_rate": 4.7274903750853696e-05,
      "loss": 0.5638,
      "mean_token_accuracy": 0.82160325050354,
      "num_tokens": 224986215.0,
      "step": 6715
    },
    {
      "epoch": 0.40114613180515757,
      "grad_norm": 0.5797615051269531,
      "learning_rate": 4.7269604186780546e-05,
      "loss": 0.57,
      "mean_token_accuracy": 0.8205057978630066,
      "num_tokens": 225153895.0,
      "step": 6720
    },
    {
      "epoch": 0.40144460362941736,
      "grad_norm": 0.5798340439796448,
      "learning_rate": 4.72642998074554e-05,
      "loss": 0.56,
      "mean_token_accuracy": 0.8243588209152222,
      "num_tokens": 225321575.0,
      "step": 6725
    },
    {
      "epoch": 0.40174307545367716,
      "grad_norm": 0.5631234049797058,
      "learning_rate": 4.725899061417021e-05,
      "loss": 0.5925,
      "mean_token_accuracy": 0.8168436169624329,
      "num_tokens": 225489255.0,
      "step": 6730
    },
    {
      "epoch": 0.40204154727793695,
      "grad_norm": 0.5207034945487976,
      "learning_rate": 4.725367660821815e-05,
      "loss": 0.5435,
      "mean_token_accuracy": 0.829875934123993,
      "num_tokens": 225656935.0,
      "step": 6735
    },
    {
      "epoch": 0.40234001910219674,
      "grad_norm": 0.6461813449859619,
      "learning_rate": 4.724835779089355e-05,
      "loss": 0.5983,
      "mean_token_accuracy": 0.8127281427383423,
      "num_tokens": 225824615.0,
      "step": 6740
    },
    {
      "epoch": 0.40263849092645654,
      "grad_norm": 0.5818511843681335,
      "learning_rate": 4.72430341634919e-05,
      "loss": 0.5735,
      "mean_token_accuracy": 0.8205236911773681,
      "num_tokens": 225992295.0,
      "step": 6745
    },
    {
      "epoch": 0.40293696275071633,
      "grad_norm": 0.525959312915802,
      "learning_rate": 4.723770572730986e-05,
      "loss": 0.5518,
      "mean_token_accuracy": 0.8268280982971191,
      "num_tokens": 226159975.0,
      "step": 6750
    },
    {
      "epoch": 0.4032354345749761,
      "grad_norm": 0.5492711067199707,
      "learning_rate": 4.723237248364527e-05,
      "loss": 0.5415,
      "mean_token_accuracy": 0.8284027218818665,
      "num_tokens": 226327655.0,
      "step": 6755
    },
    {
      "epoch": 0.4035339063992359,
      "grad_norm": 0.6119962930679321,
      "learning_rate": 4.722703443379714e-05,
      "loss": 0.5662,
      "mean_token_accuracy": 0.8214779853820801,
      "num_tokens": 226495335.0,
      "step": 6760
    },
    {
      "epoch": 0.4038323782234957,
      "grad_norm": 0.6557056307792664,
      "learning_rate": 4.722169157906567e-05,
      "loss": 0.5571,
      "mean_token_accuracy": 0.8267744302749633,
      "num_tokens": 226663015.0,
      "step": 6765
    },
    {
      "epoch": 0.4041308500477555,
      "grad_norm": 0.5651956796646118,
      "learning_rate": 4.7216343920752185e-05,
      "loss": 0.5733,
      "mean_token_accuracy": 0.8198258399963378,
      "num_tokens": 226830695.0,
      "step": 6770
    },
    {
      "epoch": 0.4044293218720153,
      "grad_norm": 0.5722317695617676,
      "learning_rate": 4.721099146015921e-05,
      "loss": 0.6007,
      "mean_token_accuracy": 0.812429916858673,
      "num_tokens": 226998375.0,
      "step": 6775
    },
    {
      "epoch": 0.4047277936962751,
      "grad_norm": 0.5798092484474182,
      "learning_rate": 4.7205634198590446e-05,
      "loss": 0.559,
      "mean_token_accuracy": 0.8246212601661682,
      "num_tokens": 227166055.0,
      "step": 6780
    },
    {
      "epoch": 0.40502626552053483,
      "grad_norm": 0.58822101354599,
      "learning_rate": 4.720027213735076e-05,
      "loss": 0.5171,
      "mean_token_accuracy": 0.8375820040702819,
      "num_tokens": 227333735.0,
      "step": 6785
    },
    {
      "epoch": 0.40532473734479463,
      "grad_norm": 0.5558359026908875,
      "learning_rate": 4.719490527774616e-05,
      "loss": 0.5695,
      "mean_token_accuracy": 0.8221817970275879,
      "num_tokens": 227501415.0,
      "step": 6790
    },
    {
      "epoch": 0.4056232091690544,
      "grad_norm": 0.5774255990982056,
      "learning_rate": 4.7189533621083856e-05,
      "loss": 0.577,
      "mean_token_accuracy": 0.816628885269165,
      "num_tokens": 227669095.0,
      "step": 6795
    },
    {
      "epoch": 0.4059216809933142,
      "grad_norm": 0.6004100441932678,
      "learning_rate": 4.71841571686722e-05,
      "loss": 0.6043,
      "mean_token_accuracy": 0.8094954013824462,
      "num_tokens": 227836775.0,
      "step": 6800
    },
    {
      "epoch": 0.406220152817574,
      "grad_norm": 0.622145414352417,
      "learning_rate": 4.7178775921820754e-05,
      "loss": 0.6366,
      "mean_token_accuracy": 0.8022784113883972,
      "num_tokens": 228004455.0,
      "step": 6805
    },
    {
      "epoch": 0.4065186246418338,
      "grad_norm": 0.5691177845001221,
      "learning_rate": 4.71733898818402e-05,
      "loss": 0.5774,
      "mean_token_accuracy": 0.819479787349701,
      "num_tokens": 228169230.0,
      "step": 6810
    },
    {
      "epoch": 0.4068170964660936,
      "grad_norm": 0.566308856010437,
      "learning_rate": 4.7167999050042405e-05,
      "loss": 0.5765,
      "mean_token_accuracy": 0.8199809193611145,
      "num_tokens": 228336910.0,
      "step": 6815
    },
    {
      "epoch": 0.4071155682903534,
      "grad_norm": 0.6092915534973145,
      "learning_rate": 4.716260342774041e-05,
      "loss": 0.6006,
      "mean_token_accuracy": 0.8117917299270629,
      "num_tokens": 228504590.0,
      "step": 6820
    },
    {
      "epoch": 0.4074140401146132,
      "grad_norm": 0.5469214916229248,
      "learning_rate": 4.715720301624843e-05,
      "loss": 0.5788,
      "mean_token_accuracy": 0.821227490901947,
      "num_tokens": 228672270.0,
      "step": 6825
    },
    {
      "epoch": 0.407712511938873,
      "grad_norm": 0.5775798559188843,
      "learning_rate": 4.7151797816881823e-05,
      "loss": 0.5487,
      "mean_token_accuracy": 0.8256352186203003,
      "num_tokens": 228839950.0,
      "step": 6830
    },
    {
      "epoch": 0.4080109837631328,
      "grad_norm": 0.5717394948005676,
      "learning_rate": 4.714638783095712e-05,
      "loss": 0.5429,
      "mean_token_accuracy": 0.8299296140670777,
      "num_tokens": 229007630.0,
      "step": 6835
    },
    {
      "epoch": 0.40830945558739257,
      "grad_norm": 0.5832452774047852,
      "learning_rate": 4.714097305979203e-05,
      "loss": 0.5517,
      "mean_token_accuracy": 0.8249135136604309,
      "num_tokens": 229175310.0,
      "step": 6840
    },
    {
      "epoch": 0.40860792741165236,
      "grad_norm": 0.6015300750732422,
      "learning_rate": 4.713555350470542e-05,
      "loss": 0.5962,
      "mean_token_accuracy": 0.8129547834396362,
      "num_tokens": 229342990.0,
      "step": 6845
    },
    {
      "epoch": 0.40890639923591215,
      "grad_norm": 0.5837094783782959,
      "learning_rate": 4.713012916701734e-05,
      "loss": 0.5924,
      "mean_token_accuracy": 0.8144041538238526,
      "num_tokens": 229510670.0,
      "step": 6850
    },
    {
      "epoch": 0.4092048710601719,
      "grad_norm": 0.6850583553314209,
      "learning_rate": 4.712470004804895e-05,
      "loss": 0.6467,
      "mean_token_accuracy": 0.8003697991371155,
      "num_tokens": 229678350.0,
      "step": 6855
    },
    {
      "epoch": 0.4095033428844317,
      "grad_norm": 0.6050103902816772,
      "learning_rate": 4.7119266149122646e-05,
      "loss": 0.6092,
      "mean_token_accuracy": 0.8098234534263611,
      "num_tokens": 229846030.0,
      "step": 6860
    },
    {
      "epoch": 0.4098018147086915,
      "grad_norm": 0.5681546330451965,
      "learning_rate": 4.7113827471561934e-05,
      "loss": 0.5642,
      "mean_token_accuracy": 0.8228677153587342,
      "num_tokens": 230013710.0,
      "step": 6865
    },
    {
      "epoch": 0.4101002865329513,
      "grad_norm": 0.5979629755020142,
      "learning_rate": 4.710838401669152e-05,
      "loss": 0.5581,
      "mean_token_accuracy": 0.8249135136604309,
      "num_tokens": 230181390.0,
      "step": 6870
    },
    {
      "epoch": 0.41039875835721107,
      "grad_norm": 0.5612121224403381,
      "learning_rate": 4.7102935785837244e-05,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8320708632469177,
      "num_tokens": 230349070.0,
      "step": 6875
    },
    {
      "epoch": 0.41069723018147086,
      "grad_norm": 0.8240800499916077,
      "learning_rate": 4.7097482780326126e-05,
      "loss": 0.557,
      "mean_token_accuracy": 0.8262256979942322,
      "num_tokens": 230516750.0,
      "step": 6880
    },
    {
      "epoch": 0.41099570200573066,
      "grad_norm": 0.6205378770828247,
      "learning_rate": 4.7092025001486344e-05,
      "loss": 0.609,
      "mean_token_accuracy": 0.8097220659255981,
      "num_tokens": 230684430.0,
      "step": 6885
    },
    {
      "epoch": 0.41129417382999045,
      "grad_norm": 0.5555119514465332,
      "learning_rate": 4.708656245064726e-05,
      "loss": 0.5588,
      "mean_token_accuracy": 0.8247882723808289,
      "num_tokens": 230852110.0,
      "step": 6890
    },
    {
      "epoch": 0.41159264565425024,
      "grad_norm": 0.5389045476913452,
      "learning_rate": 4.708109512913935e-05,
      "loss": 0.5726,
      "mean_token_accuracy": 0.8197184681892395,
      "num_tokens": 231019790.0,
      "step": 6895
    },
    {
      "epoch": 0.41189111747851004,
      "grad_norm": 0.5762794613838196,
      "learning_rate": 4.707562303829431e-05,
      "loss": 0.5791,
      "mean_token_accuracy": 0.8184605956077575,
      "num_tokens": 231182870.0,
      "step": 6900
    },
    {
      "epoch": 0.41218958930276983,
      "grad_norm": 0.5309786796569824,
      "learning_rate": 4.707014617944495e-05,
      "loss": 0.5582,
      "mean_token_accuracy": 0.8255874872207641,
      "num_tokens": 231350550.0,
      "step": 6905
    },
    {
      "epoch": 0.4124880611270296,
      "grad_norm": 0.5733486413955688,
      "learning_rate": 4.706466455392526e-05,
      "loss": 0.5947,
      "mean_token_accuracy": 0.8153226852416993,
      "num_tokens": 231518230.0,
      "step": 6910
    },
    {
      "epoch": 0.4127865329512894,
      "grad_norm": 0.5413124561309814,
      "learning_rate": 4.7059178163070395e-05,
      "loss": 0.5789,
      "mean_token_accuracy": 0.8188118815422059,
      "num_tokens": 231685910.0,
      "step": 6915
    },
    {
      "epoch": 0.4130850047755492,
      "grad_norm": 0.626871645450592,
      "learning_rate": 4.705368700821667e-05,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8269951105117798,
      "num_tokens": 231853590.0,
      "step": 6920
    },
    {
      "epoch": 0.41338347659980895,
      "grad_norm": 0.5655709505081177,
      "learning_rate": 4.704819109070155e-05,
      "loss": 0.564,
      "mean_token_accuracy": 0.8230585813522339,
      "num_tokens": 232021270.0,
      "step": 6925
    },
    {
      "epoch": 0.41368194842406875,
      "grad_norm": 0.5631151795387268,
      "learning_rate": 4.7042690411863674e-05,
      "loss": 0.5781,
      "mean_token_accuracy": 0.8179649353027344,
      "num_tokens": 232188950.0,
      "step": 6930
    },
    {
      "epoch": 0.41398042024832854,
      "grad_norm": 1.1137640476226807,
      "learning_rate": 4.7037184973042834e-05,
      "loss": 0.6052,
      "mean_token_accuracy": 0.8105809450149536,
      "num_tokens": 232356630.0,
      "step": 6935
    },
    {
      "epoch": 0.41427889207258833,
      "grad_norm": 0.6091469526290894,
      "learning_rate": 4.703167477557997e-05,
      "loss": 0.5846,
      "mean_token_accuracy": 0.8158236980438233,
      "num_tokens": 232524310.0,
      "step": 6940
    },
    {
      "epoch": 0.41457736389684813,
      "grad_norm": 0.7949170470237732,
      "learning_rate": 4.702615982081722e-05,
      "loss": 0.6359,
      "mean_token_accuracy": 0.8052422165870666,
      "num_tokens": 232686315.0,
      "step": 6945
    },
    {
      "epoch": 0.4148758357211079,
      "grad_norm": 0.5734235644340515,
      "learning_rate": 4.702064011009782e-05,
      "loss": 0.6142,
      "mean_token_accuracy": 0.8074973225593567,
      "num_tokens": 232853995.0,
      "step": 6950
    },
    {
      "epoch": 0.4151743075453677,
      "grad_norm": 0.5585874319076538,
      "learning_rate": 4.7015115644766214e-05,
      "loss": 0.574,
      "mean_token_accuracy": 0.820100212097168,
      "num_tokens": 233021675.0,
      "step": 6955
    },
    {
      "epoch": 0.4154727793696275,
      "grad_norm": 0.5752120614051819,
      "learning_rate": 4.700958642616799e-05,
      "loss": 0.5851,
      "mean_token_accuracy": 0.8153346061706543,
      "num_tokens": 233189355.0,
      "step": 6960
    },
    {
      "epoch": 0.4157712511938873,
      "grad_norm": 0.6133684515953064,
      "learning_rate": 4.7004052455649876e-05,
      "loss": 0.607,
      "mean_token_accuracy": 0.8133245825767517,
      "num_tokens": 233357035.0,
      "step": 6965
    },
    {
      "epoch": 0.4160697230181471,
      "grad_norm": 0.5853560566902161,
      "learning_rate": 4.699851373455979e-05,
      "loss": 0.6178,
      "mean_token_accuracy": 0.8065907239913941,
      "num_tokens": 233524715.0,
      "step": 6970
    },
    {
      "epoch": 0.4163681948424069,
      "grad_norm": 0.5792456865310669,
      "learning_rate": 4.6992970264246775e-05,
      "loss": 0.546,
      "mean_token_accuracy": 0.826428484916687,
      "num_tokens": 233692395.0,
      "step": 6975
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.6258230805397034,
      "learning_rate": 4.698742204606107e-05,
      "loss": 0.6231,
      "mean_token_accuracy": 0.8080758690834046,
      "num_tokens": 233860075.0,
      "step": 6980
    },
    {
      "epoch": 0.4169651384909265,
      "grad_norm": 0.5646248459815979,
      "learning_rate": 4.698186908135401e-05,
      "loss": 0.5963,
      "mean_token_accuracy": 0.8156209111213684,
      "num_tokens": 234027755.0,
      "step": 6985
    },
    {
      "epoch": 0.4172636103151863,
      "grad_norm": 0.5563027262687683,
      "learning_rate": 4.697631137147815e-05,
      "loss": 0.5796,
      "mean_token_accuracy": 0.8195276021957397,
      "num_tokens": 234195435.0,
      "step": 6990
    },
    {
      "epoch": 0.417562082139446,
      "grad_norm": 0.5151941180229187,
      "learning_rate": 4.697074891778716e-05,
      "loss": 0.594,
      "mean_token_accuracy": 0.8130442500114441,
      "num_tokens": 234363115.0,
      "step": 6995
    },
    {
      "epoch": 0.4178605539637058,
      "grad_norm": 0.6522302627563477,
      "learning_rate": 4.696518172163589e-05,
      "loss": 0.6223,
      "mean_token_accuracy": 0.8075330972671508,
      "num_tokens": 234530795.0,
      "step": 7000
    },
    {
      "epoch": 0.4181590257879656,
      "grad_norm": 0.5927281975746155,
      "learning_rate": 4.695960978438033e-05,
      "loss": 0.5562,
      "mean_token_accuracy": 0.8239830613136292,
      "num_tokens": 234698475.0,
      "step": 7005
    },
    {
      "epoch": 0.4184574976122254,
      "grad_norm": 0.6028661727905273,
      "learning_rate": 4.69540331073776e-05,
      "loss": 0.5674,
      "mean_token_accuracy": 0.8212095975875855,
      "num_tokens": 234866155.0,
      "step": 7010
    },
    {
      "epoch": 0.4187559694364852,
      "grad_norm": 0.5302370190620422,
      "learning_rate": 4.6948451691986045e-05,
      "loss": 0.5316,
      "mean_token_accuracy": 0.8314267039299011,
      "num_tokens": 235033835.0,
      "step": 7015
    },
    {
      "epoch": 0.419054441260745,
      "grad_norm": 0.5614240169525146,
      "learning_rate": 4.69428655395651e-05,
      "loss": 0.5691,
      "mean_token_accuracy": 0.8193367481231689,
      "num_tokens": 235201515.0,
      "step": 7020
    },
    {
      "epoch": 0.4193529130850048,
      "grad_norm": 0.632599413394928,
      "learning_rate": 4.6937274651475385e-05,
      "loss": 0.6344,
      "mean_token_accuracy": 0.8035667419433594,
      "num_tokens": 235369195.0,
      "step": 7025
    },
    {
      "epoch": 0.41965138490926457,
      "grad_norm": 0.5534483790397644,
      "learning_rate": 4.693167902907865e-05,
      "loss": 0.5897,
      "mean_token_accuracy": 0.8155195116996765,
      "num_tokens": 235536875.0,
      "step": 7030
    },
    {
      "epoch": 0.41994985673352436,
      "grad_norm": 0.5887237787246704,
      "learning_rate": 4.692607867373781e-05,
      "loss": 0.5653,
      "mean_token_accuracy": 0.8224442362785339,
      "num_tokens": 235704555.0,
      "step": 7035
    },
    {
      "epoch": 0.42024832855778416,
      "grad_norm": 0.6462065577507019,
      "learning_rate": 4.692047358681696e-05,
      "loss": 0.625,
      "mean_token_accuracy": 0.8062925100326538,
      "num_tokens": 235872235.0,
      "step": 7040
    },
    {
      "epoch": 0.42054680038204395,
      "grad_norm": 0.5301601886749268,
      "learning_rate": 4.69148637696813e-05,
      "loss": 0.5651,
      "mean_token_accuracy": 0.8209352374076844,
      "num_tokens": 236039915.0,
      "step": 7045
    },
    {
      "epoch": 0.42084527220630374,
      "grad_norm": 0.5547974109649658,
      "learning_rate": 4.6909249223697216e-05,
      "loss": 0.5528,
      "mean_token_accuracy": 0.8276810050010681,
      "num_tokens": 236207595.0,
      "step": 7050
    },
    {
      "epoch": 0.42114374403056354,
      "grad_norm": 0.6368840932846069,
      "learning_rate": 4.6903629950232225e-05,
      "loss": 0.5941,
      "mean_token_accuracy": 0.8133245825767517,
      "num_tokens": 236375275.0,
      "step": 7055
    },
    {
      "epoch": 0.42144221585482333,
      "grad_norm": 0.6618936061859131,
      "learning_rate": 4.689800595065501e-05,
      "loss": 0.5829,
      "mean_token_accuracy": 0.8166944980621338,
      "num_tokens": 236542955.0,
      "step": 7060
    },
    {
      "epoch": 0.42174068767908307,
      "grad_norm": 0.630326509475708,
      "learning_rate": 4.6892377226335405e-05,
      "loss": 0.592,
      "mean_token_accuracy": 0.8146367788314819,
      "num_tokens": 236710635.0,
      "step": 7065
    },
    {
      "epoch": 0.42203915950334286,
      "grad_norm": 0.6314406394958496,
      "learning_rate": 4.6886743778644385e-05,
      "loss": 0.5805,
      "mean_token_accuracy": 0.8181378960609436,
      "num_tokens": 236878315.0,
      "step": 7070
    },
    {
      "epoch": 0.42233763132760266,
      "grad_norm": 0.5336928367614746,
      "learning_rate": 4.688110560895407e-05,
      "loss": 0.5381,
      "mean_token_accuracy": 0.8289812684059144,
      "num_tokens": 237045995.0,
      "step": 7075
    },
    {
      "epoch": 0.42263610315186245,
      "grad_norm": 0.5644545555114746,
      "learning_rate": 4.687546271863776e-05,
      "loss": 0.5634,
      "mean_token_accuracy": 0.8218716502189636,
      "num_tokens": 237213675.0,
      "step": 7080
    },
    {
      "epoch": 0.42293457497612225,
      "grad_norm": 0.5761260986328125,
      "learning_rate": 4.686981510906986e-05,
      "loss": 0.5914,
      "mean_token_accuracy": 0.8146844744682312,
      "num_tokens": 237381355.0,
      "step": 7085
    },
    {
      "epoch": 0.42323304680038204,
      "grad_norm": 0.6816636323928833,
      "learning_rate": 4.686416278162598e-05,
      "loss": 0.6247,
      "mean_token_accuracy": 0.8046164870262146,
      "num_tokens": 237549035.0,
      "step": 7090
    },
    {
      "epoch": 0.42353151862464183,
      "grad_norm": 0.5634966492652893,
      "learning_rate": 4.685850573768283e-05,
      "loss": 0.5327,
      "mean_token_accuracy": 0.831766664981842,
      "num_tokens": 237716715.0,
      "step": 7095
    },
    {
      "epoch": 0.42382999044890163,
      "grad_norm": 0.5552952885627747,
      "learning_rate": 4.685284397861828e-05,
      "loss": 0.5691,
      "mean_token_accuracy": 0.8212573051452636,
      "num_tokens": 237884395.0,
      "step": 7100
    },
    {
      "epoch": 0.4241284622731614,
      "grad_norm": 0.5909750461578369,
      "learning_rate": 4.684717750581138e-05,
      "loss": 0.6142,
      "mean_token_accuracy": 0.8085888147354126,
      "num_tokens": 238052075.0,
      "step": 7105
    },
    {
      "epoch": 0.4244269340974212,
      "grad_norm": 0.5910950303077698,
      "learning_rate": 4.684150632064228e-05,
      "loss": 0.614,
      "mean_token_accuracy": 0.8090182662010192,
      "num_tokens": 238219755.0,
      "step": 7110
    },
    {
      "epoch": 0.424725405921681,
      "grad_norm": 0.5288212299346924,
      "learning_rate": 4.683583042449232e-05,
      "loss": 0.5398,
      "mean_token_accuracy": 0.8288918137550354,
      "num_tokens": 238387435.0,
      "step": 7115
    },
    {
      "epoch": 0.4250238777459408,
      "grad_norm": 0.637802243232727,
      "learning_rate": 4.6830149818743956e-05,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8223010897636414,
      "num_tokens": 238555115.0,
      "step": 7120
    },
    {
      "epoch": 0.4253223495702006,
      "grad_norm": 0.6213273406028748,
      "learning_rate": 4.682446450478082e-05,
      "loss": 0.597,
      "mean_token_accuracy": 0.81300847530365,
      "num_tokens": 238722795.0,
      "step": 7125
    },
    {
      "epoch": 0.42562082139446034,
      "grad_norm": 0.5779014229774475,
      "learning_rate": 4.6818774483987655e-05,
      "loss": 0.6167,
      "mean_token_accuracy": 0.8072706699371338,
      "num_tokens": 238890475.0,
      "step": 7130
    },
    {
      "epoch": 0.42591929321872013,
      "grad_norm": 0.6011877059936523,
      "learning_rate": 4.6813079757750386e-05,
      "loss": 0.5877,
      "mean_token_accuracy": 0.8154956459999084,
      "num_tokens": 239058155.0,
      "step": 7135
    },
    {
      "epoch": 0.4262177650429799,
      "grad_norm": 0.6756574511528015,
      "learning_rate": 4.680738032745606e-05,
      "loss": 0.5846,
      "mean_token_accuracy": 0.8168197631835937,
      "num_tokens": 239225835.0,
      "step": 7140
    },
    {
      "epoch": 0.4265162368672397,
      "grad_norm": 0.523808479309082,
      "learning_rate": 4.6801676194492884e-05,
      "loss": 0.5749,
      "mean_token_accuracy": 0.8182392954826355,
      "num_tokens": 239393515.0,
      "step": 7145
    },
    {
      "epoch": 0.4268147086914995,
      "grad_norm": 0.5292114019393921,
      "learning_rate": 4.67959673602502e-05,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8359119772911072,
      "num_tokens": 239561195.0,
      "step": 7150
    },
    {
      "epoch": 0.4271131805157593,
      "grad_norm": 0.6238457560539246,
      "learning_rate": 4.67902538261185e-05,
      "loss": 0.5963,
      "mean_token_accuracy": 0.81266850233078,
      "num_tokens": 239728875.0,
      "step": 7155
    },
    {
      "epoch": 0.4274116523400191,
      "grad_norm": 0.5610671639442444,
      "learning_rate": 4.678453559348943e-05,
      "loss": 0.6089,
      "mean_token_accuracy": 0.811314582824707,
      "num_tokens": 239896555.0,
      "step": 7160
    },
    {
      "epoch": 0.4277101241642789,
      "grad_norm": 0.569820761680603,
      "learning_rate": 4.677881266375576e-05,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8295717597007751,
      "num_tokens": 240064235.0,
      "step": 7165
    },
    {
      "epoch": 0.4280085959885387,
      "grad_norm": 0.5484038591384888,
      "learning_rate": 4.677308503831143e-05,
      "loss": 0.5154,
      "mean_token_accuracy": 0.8342180609703064,
      "num_tokens": 240231915.0,
      "step": 7170
    },
    {
      "epoch": 0.4283070678127985,
      "grad_norm": 0.5385307669639587,
      "learning_rate": 4.6767352718551484e-05,
      "loss": 0.5743,
      "mean_token_accuracy": 0.8187880158424378,
      "num_tokens": 240399595.0,
      "step": 7175
    },
    {
      "epoch": 0.4286055396370583,
      "grad_norm": 0.5796418786048889,
      "learning_rate": 4.676161570587216e-05,
      "loss": 0.5965,
      "mean_token_accuracy": 0.8141775131225586,
      "num_tokens": 240567275.0,
      "step": 7180
    },
    {
      "epoch": 0.42890401146131807,
      "grad_norm": 0.5219092965126038,
      "learning_rate": 4.675587400167079e-05,
      "loss": 0.5903,
      "mean_token_accuracy": 0.8163843512535095,
      "num_tokens": 240734955.0,
      "step": 7185
    },
    {
      "epoch": 0.42920248328557786,
      "grad_norm": 0.6324279308319092,
      "learning_rate": 4.675012760734589e-05,
      "loss": 0.5435,
      "mean_token_accuracy": 0.8295657873153687,
      "num_tokens": 240902635.0,
      "step": 7190
    },
    {
      "epoch": 0.42950095510983766,
      "grad_norm": 0.5694840550422668,
      "learning_rate": 4.67443765242971e-05,
      "loss": 0.5701,
      "mean_token_accuracy": 0.8202791452407837,
      "num_tokens": 241070315.0,
      "step": 7195
    },
    {
      "epoch": 0.4297994269340974,
      "grad_norm": 0.588840901851654,
      "learning_rate": 4.6738620753925197e-05,
      "loss": 0.5626,
      "mean_token_accuracy": 0.8215137720108032,
      "num_tokens": 241237995.0,
      "step": 7200
    },
    {
      "epoch": 0.4300978987583572,
      "grad_norm": 0.5361875891685486,
      "learning_rate": 4.6732860297632095e-05,
      "loss": 0.5586,
      "mean_token_accuracy": 0.8234164476394653,
      "num_tokens": 241405675.0,
      "step": 7205
    },
    {
      "epoch": 0.430396370582617,
      "grad_norm": 0.63075852394104,
      "learning_rate": 4.6727095156820874e-05,
      "loss": 0.5828,
      "mean_token_accuracy": 0.817517614364624,
      "num_tokens": 241573355.0,
      "step": 7210
    },
    {
      "epoch": 0.4306948424068768,
      "grad_norm": 0.6125807166099548,
      "learning_rate": 4.6721325332895743e-05,
      "loss": 0.565,
      "mean_token_accuracy": 0.8224263429641724,
      "num_tokens": 241741035.0,
      "step": 7215
    },
    {
      "epoch": 0.43099331423113657,
      "grad_norm": 0.5737380981445312,
      "learning_rate": 4.671555082726204e-05,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8244542479515076,
      "num_tokens": 241908715.0,
      "step": 7220
    },
    {
      "epoch": 0.43129178605539636,
      "grad_norm": 0.5836515426635742,
      "learning_rate": 4.6709771641326244e-05,
      "loss": 0.5587,
      "mean_token_accuracy": 0.8233389019966125,
      "num_tokens": 242076395.0,
      "step": 7225
    },
    {
      "epoch": 0.43159025787965616,
      "grad_norm": 0.5387793183326721,
      "learning_rate": 4.6703987776496004e-05,
      "loss": 0.5468,
      "mean_token_accuracy": 0.8256829380989075,
      "num_tokens": 242244075.0,
      "step": 7230
    },
    {
      "epoch": 0.43188872970391595,
      "grad_norm": 0.5948957800865173,
      "learning_rate": 4.669819923418008e-05,
      "loss": 0.6127,
      "mean_token_accuracy": 0.8066623091697693,
      "num_tokens": 242411755.0,
      "step": 7235
    },
    {
      "epoch": 0.43218720152817575,
      "grad_norm": 0.5518760085105896,
      "learning_rate": 4.669240601578835e-05,
      "loss": 0.5872,
      "mean_token_accuracy": 0.8159429788589477,
      "num_tokens": 242579435.0,
      "step": 7240
    },
    {
      "epoch": 0.43248567335243554,
      "grad_norm": 0.5337597131729126,
      "learning_rate": 4.6686608122731906e-05,
      "loss": 0.5195,
      "mean_token_accuracy": 0.8343254208564759,
      "num_tokens": 242747115.0,
      "step": 7245
    },
    {
      "epoch": 0.43278414517669533,
      "grad_norm": 0.551590085029602,
      "learning_rate": 4.6680805556422905e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8353870749473572,
      "num_tokens": 242914795.0,
      "step": 7250
    },
    {
      "epoch": 0.43308261700095513,
      "grad_norm": 0.5611405372619629,
      "learning_rate": 4.6674998318274674e-05,
      "loss": 0.5933,
      "mean_token_accuracy": 0.8142669677734375,
      "num_tokens": 243082475.0,
      "step": 7255
    },
    {
      "epoch": 0.4333810888252149,
      "grad_norm": 0.5417499542236328,
      "learning_rate": 4.666918640970166e-05,
      "loss": 0.5946,
      "mean_token_accuracy": 0.8130681157112122,
      "num_tokens": 243250155.0,
      "step": 7260
    },
    {
      "epoch": 0.4336795606494747,
      "grad_norm": 0.600557267665863,
      "learning_rate": 4.666336983211949e-05,
      "loss": 0.6019,
      "mean_token_accuracy": 0.8110342264175415,
      "num_tokens": 243417835.0,
      "step": 7265
    },
    {
      "epoch": 0.43397803247373445,
      "grad_norm": 0.5825746059417725,
      "learning_rate": 4.6657548586944884e-05,
      "loss": 0.5592,
      "mean_token_accuracy": 0.8250805139541626,
      "num_tokens": 243585515.0,
      "step": 7270
    },
    {
      "epoch": 0.43427650429799425,
      "grad_norm": 0.5456596612930298,
      "learning_rate": 4.66517226755957e-05,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8250387787818909,
      "num_tokens": 243753195.0,
      "step": 7275
    },
    {
      "epoch": 0.43457497612225404,
      "grad_norm": 0.5647388100624084,
      "learning_rate": 4.6645892099490963e-05,
      "loss": 0.5776,
      "mean_token_accuracy": 0.8185196280479431,
      "num_tokens": 243920875.0,
      "step": 7280
    },
    {
      "epoch": 0.43487344794651384,
      "grad_norm": 0.5186007618904114,
      "learning_rate": 4.6640056860050814e-05,
      "loss": 0.5511,
      "mean_token_accuracy": 0.825790286064148,
      "num_tokens": 244088555.0,
      "step": 7285
    },
    {
      "epoch": 0.43517191977077363,
      "grad_norm": 0.5805457234382629,
      "learning_rate": 4.6634216958696534e-05,
      "loss": 0.5575,
      "mean_token_accuracy": 0.8240307807922364,
      "num_tokens": 244256235.0,
      "step": 7290
    },
    {
      "epoch": 0.4354703915950334,
      "grad_norm": 0.5633774995803833,
      "learning_rate": 4.662837239685052e-05,
      "loss": 0.6312,
      "mean_token_accuracy": 0.8053202867507935,
      "num_tokens": 244423915.0,
      "step": 7295
    },
    {
      "epoch": 0.4357688634192932,
      "grad_norm": 0.7506545186042786,
      "learning_rate": 4.662252317593636e-05,
      "loss": 0.577,
      "mean_token_accuracy": 0.8191458821296692,
      "num_tokens": 244591595.0,
      "step": 7300
    },
    {
      "epoch": 0.436067335243553,
      "grad_norm": 0.6479071378707886,
      "learning_rate": 4.6616669297378705e-05,
      "loss": 0.5864,
      "mean_token_accuracy": 0.8154181122779847,
      "num_tokens": 244759275.0,
      "step": 7305
    },
    {
      "epoch": 0.4363658070678128,
      "grad_norm": 0.5674999356269836,
      "learning_rate": 4.6610810762603405e-05,
      "loss": 0.5496,
      "mean_token_accuracy": 0.8271919369697571,
      "num_tokens": 244926955.0,
      "step": 7310
    },
    {
      "epoch": 0.4366642788920726,
      "grad_norm": 0.554707944393158,
      "learning_rate": 4.6604947573037386e-05,
      "loss": 0.5783,
      "mean_token_accuracy": 0.8181140422821045,
      "num_tokens": 245094635.0,
      "step": 7315
    },
    {
      "epoch": 0.4369627507163324,
      "grad_norm": 0.5681338906288147,
      "learning_rate": 4.6599079730108745e-05,
      "loss": 0.6185,
      "mean_token_accuracy": 0.8066563248634339,
      "num_tokens": 245262315.0,
      "step": 7320
    },
    {
      "epoch": 0.4372612225405922,
      "grad_norm": 0.6053183078765869,
      "learning_rate": 4.659320723524672e-05,
      "loss": 0.5287,
      "mean_token_accuracy": 0.8312119722366333,
      "num_tokens": 245429995.0,
      "step": 7325
    },
    {
      "epoch": 0.437559694364852,
      "grad_norm": 0.5857833027839661,
      "learning_rate": 4.658733008988164e-05,
      "loss": 0.5705,
      "mean_token_accuracy": 0.8193784952163696,
      "num_tokens": 245597675.0,
      "step": 7330
    },
    {
      "epoch": 0.4378581661891118,
      "grad_norm": 0.5568007826805115,
      "learning_rate": 4.6581448295445016e-05,
      "loss": 0.5547,
      "mean_token_accuracy": 0.8239949822425843,
      "num_tokens": 245765355.0,
      "step": 7335
    },
    {
      "epoch": 0.4381566380133715,
      "grad_norm": 0.5761382579803467,
      "learning_rate": 4.657556185336945e-05,
      "loss": 0.5857,
      "mean_token_accuracy": 0.814988660812378,
      "num_tokens": 245933035.0,
      "step": 7340
    },
    {
      "epoch": 0.4384551098376313,
      "grad_norm": 0.5609983801841736,
      "learning_rate": 4.6569670765088703e-05,
      "loss": 0.595,
      "mean_token_accuracy": 0.813533341884613,
      "num_tokens": 246100715.0,
      "step": 7345
    },
    {
      "epoch": 0.4387535816618911,
      "grad_norm": 0.5667198300361633,
      "learning_rate": 4.6563775032037654e-05,
      "loss": 0.5683,
      "mean_token_accuracy": 0.820055615901947,
      "num_tokens": 246262211.0,
      "step": 7350
    },
    {
      "epoch": 0.4390520534861509,
      "grad_norm": 0.5638264417648315,
      "learning_rate": 4.6557874655652316e-05,
      "loss": 0.5684,
      "mean_token_accuracy": 0.8206250667572021,
      "num_tokens": 246429891.0,
      "step": 7355
    },
    {
      "epoch": 0.4393505253104107,
      "grad_norm": 0.5304365754127502,
      "learning_rate": 4.655196963736985e-05,
      "loss": 0.5498,
      "mean_token_accuracy": 0.8263509631156921,
      "num_tokens": 246597571.0,
      "step": 7360
    },
    {
      "epoch": 0.4396489971346705,
      "grad_norm": 0.6020424365997314,
      "learning_rate": 4.65460599786285e-05,
      "loss": 0.5994,
      "mean_token_accuracy": 0.813634741306305,
      "num_tokens": 246765251.0,
      "step": 7365
    },
    {
      "epoch": 0.4399474689589303,
      "grad_norm": 0.5858073830604553,
      "learning_rate": 4.654014568086771e-05,
      "loss": 0.6441,
      "mean_token_accuracy": 0.7982822299003601,
      "num_tokens": 246932931.0,
      "step": 7370
    },
    {
      "epoch": 0.44024594078319007,
      "grad_norm": 0.5727011561393738,
      "learning_rate": 4.653422674552799e-05,
      "loss": 0.6163,
      "mean_token_accuracy": 0.8085649490356446,
      "num_tokens": 247100611.0,
      "step": 7375
    },
    {
      "epoch": 0.44054441260744986,
      "grad_norm": 0.5156970024108887,
      "learning_rate": 4.6528303174051015e-05,
      "loss": 0.5541,
      "mean_token_accuracy": 0.8256352186203003,
      "num_tokens": 247268291.0,
      "step": 7380
    },
    {
      "epoch": 0.44084288443170966,
      "grad_norm": 0.5324041843414307,
      "learning_rate": 4.652237496787958e-05,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8214839696884155,
      "num_tokens": 247435971.0,
      "step": 7385
    },
    {
      "epoch": 0.44114135625596945,
      "grad_norm": 0.6281463503837585,
      "learning_rate": 4.6516442128457604e-05,
      "loss": 0.5869,
      "mean_token_accuracy": 0.8148037791252136,
      "num_tokens": 247603651.0,
      "step": 7390
    },
    {
      "epoch": 0.44143982808022925,
      "grad_norm": 0.5490525364875793,
      "learning_rate": 4.6510504657230156e-05,
      "loss": 0.5446,
      "mean_token_accuracy": 0.8279553771018981,
      "num_tokens": 247771331.0,
      "step": 7395
    },
    {
      "epoch": 0.44173829990448904,
      "grad_norm": 0.6098350882530212,
      "learning_rate": 4.65045625556434e-05,
      "loss": 0.5919,
      "mean_token_accuracy": 0.8160861253738403,
      "num_tokens": 247939011.0,
      "step": 7400
    },
    {
      "epoch": 0.4420367717287488,
      "grad_norm": 0.6032289862632751,
      "learning_rate": 4.6498615825144644e-05,
      "loss": 0.5706,
      "mean_token_accuracy": 0.8224740505218506,
      "num_tokens": 248106691.0,
      "step": 7405
    },
    {
      "epoch": 0.4423352435530086,
      "grad_norm": 0.5556846261024475,
      "learning_rate": 4.6492664467182334e-05,
      "loss": 0.5565,
      "mean_token_accuracy": 0.825050699710846,
      "num_tokens": 248274371.0,
      "step": 7410
    },
    {
      "epoch": 0.44263371537726837,
      "grad_norm": 0.6704667806625366,
      "learning_rate": 4.648670848320603e-05,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8290823459625244,
      "num_tokens": 248435093.0,
      "step": 7415
    },
    {
      "epoch": 0.44293218720152816,
      "grad_norm": 0.5523572564125061,
      "learning_rate": 4.648074787466642e-05,
      "loss": 0.5641,
      "mean_token_accuracy": 0.8217821836471557,
      "num_tokens": 248602773.0,
      "step": 7420
    },
    {
      "epoch": 0.44323065902578795,
      "grad_norm": 0.5404214859008789,
      "learning_rate": 4.647478264301532e-05,
      "loss": 0.562,
      "mean_token_accuracy": 0.8221996784210205,
      "num_tokens": 248770453.0,
      "step": 7425
    },
    {
      "epoch": 0.44352913085004775,
      "grad_norm": 0.5859255194664001,
      "learning_rate": 4.646881278970566e-05,
      "loss": 0.612,
      "mean_token_accuracy": 0.811541211605072,
      "num_tokens": 248938133.0,
      "step": 7430
    },
    {
      "epoch": 0.44382760267430754,
      "grad_norm": 0.6272110342979431,
      "learning_rate": 4.646283831619154e-05,
      "loss": 0.6051,
      "mean_token_accuracy": 0.8101753473281861,
      "num_tokens": 249105813.0,
      "step": 7435
    },
    {
      "epoch": 0.44412607449856734,
      "grad_norm": 0.671910285949707,
      "learning_rate": 4.6456859223928135e-05,
      "loss": 0.6462,
      "mean_token_accuracy": 0.7992126941680908,
      "num_tokens": 249273493.0,
      "step": 7440
    },
    {
      "epoch": 0.44442454632282713,
      "grad_norm": 0.5461686849594116,
      "learning_rate": 4.645087551437175e-05,
      "loss": 0.5788,
      "mean_token_accuracy": 0.8167720317840577,
      "num_tokens": 249441173.0,
      "step": 7445
    },
    {
      "epoch": 0.4447230181470869,
      "grad_norm": 0.5986597537994385,
      "learning_rate": 4.644488718897984e-05,
      "loss": 0.5986,
      "mean_token_accuracy": 0.8116545438766479,
      "num_tokens": 249608853.0,
      "step": 7450
    },
    {
      "epoch": 0.4450214899713467,
      "grad_norm": 0.5439218878746033,
      "learning_rate": 4.643889424921098e-05,
      "loss": 0.5844,
      "mean_token_accuracy": 0.8167660713195801,
      "num_tokens": 249776533.0,
      "step": 7455
    },
    {
      "epoch": 0.4453199617956065,
      "grad_norm": 0.6067100763320923,
      "learning_rate": 4.643289669652485e-05,
      "loss": 0.58,
      "mean_token_accuracy": 0.8171835899353027,
      "num_tokens": 249944213.0,
      "step": 7460
    },
    {
      "epoch": 0.4456184336198663,
      "grad_norm": 0.5906630754470825,
      "learning_rate": 4.6426894532382275e-05,
      "loss": 0.5721,
      "mean_token_accuracy": 0.8190743327140808,
      "num_tokens": 250111893.0,
      "step": 7465
    },
    {
      "epoch": 0.4459169054441261,
      "grad_norm": 0.5301469564437866,
      "learning_rate": 4.642088775824517e-05,
      "loss": 0.6097,
      "mean_token_accuracy": 0.8108731865882873,
      "num_tokens": 250279573.0,
      "step": 7470
    },
    {
      "epoch": 0.44621537726838584,
      "grad_norm": 0.5304574966430664,
      "learning_rate": 4.6414876375576634e-05,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8259871125221252,
      "num_tokens": 250447253.0,
      "step": 7475
    },
    {
      "epoch": 0.44651384909264563,
      "grad_norm": 0.6436883211135864,
      "learning_rate": 4.6408860385840813e-05,
      "loss": 0.6137,
      "mean_token_accuracy": 0.8074078321456909,
      "num_tokens": 250614933.0,
      "step": 7480
    },
    {
      "epoch": 0.4468123209169054,
      "grad_norm": 0.6863070726394653,
      "learning_rate": 4.6402839790503035e-05,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8223965048789978,
      "num_tokens": 250782613.0,
      "step": 7485
    },
    {
      "epoch": 0.4471107927411652,
      "grad_norm": 0.5239238739013672,
      "learning_rate": 4.6396814591029715e-05,
      "loss": 0.5508,
      "mean_token_accuracy": 0.8264463901519775,
      "num_tokens": 250950293.0,
      "step": 7490
    },
    {
      "epoch": 0.447409264565425,
      "grad_norm": 0.5771133899688721,
      "learning_rate": 4.639078478888841e-05,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8247763276100158,
      "num_tokens": 251117973.0,
      "step": 7495
    },
    {
      "epoch": 0.4477077363896848,
      "grad_norm": 0.5490459203720093,
      "learning_rate": 4.638475038554778e-05,
      "loss": 0.5546,
      "mean_token_accuracy": 0.8253906726837158,
      "num_tokens": 251285653.0,
      "step": 7500
    },
    {
      "epoch": 0.4480062082139446,
      "grad_norm": 0.6839123964309692,
      "learning_rate": 4.6378711382477626e-05,
      "loss": 0.6101,
      "mean_token_accuracy": 0.8080877900123596,
      "num_tokens": 251453333.0,
      "step": 7505
    },
    {
      "epoch": 0.4483046800382044,
      "grad_norm": 0.573498010635376,
      "learning_rate": 4.6372667781148855e-05,
      "loss": 0.6186,
      "mean_token_accuracy": 0.8067040324211121,
      "num_tokens": 251621013.0,
      "step": 7510
    },
    {
      "epoch": 0.4486031518624642,
      "grad_norm": 0.5457788705825806,
      "learning_rate": 4.636661958303348e-05,
      "loss": 0.5432,
      "mean_token_accuracy": 0.8286114811897278,
      "num_tokens": 251788693.0,
      "step": 7515
    },
    {
      "epoch": 0.448901623686724,
      "grad_norm": 0.5477657914161682,
      "learning_rate": 4.636056678960469e-05,
      "loss": 0.5486,
      "mean_token_accuracy": 0.8247763276100158,
      "num_tokens": 251956373.0,
      "step": 7520
    },
    {
      "epoch": 0.4492000955109838,
      "grad_norm": 0.5851730108261108,
      "learning_rate": 4.635450940233672e-05,
      "loss": 0.5535,
      "mean_token_accuracy": 0.8250268459320068,
      "num_tokens": 252124053.0,
      "step": 7525
    },
    {
      "epoch": 0.44949856733524357,
      "grad_norm": 0.5567428469657898,
      "learning_rate": 4.634844742270497e-05,
      "loss": 0.5484,
      "mean_token_accuracy": 0.8279434561729431,
      "num_tokens": 252291733.0,
      "step": 7530
    },
    {
      "epoch": 0.44979703915950336,
      "grad_norm": 0.5958724021911621,
      "learning_rate": 4.6342380852185954e-05,
      "loss": 0.6196,
      "mean_token_accuracy": 0.8064833641052246,
      "num_tokens": 252459413.0,
      "step": 7535
    },
    {
      "epoch": 0.45009551098376316,
      "grad_norm": 0.6218613982200623,
      "learning_rate": 4.633630969225729e-05,
      "loss": 0.6107,
      "mean_token_accuracy": 0.8095788955688477,
      "num_tokens": 252627093.0,
      "step": 7540
    },
    {
      "epoch": 0.4503939828080229,
      "grad_norm": 0.5165412425994873,
      "learning_rate": 4.633023394439772e-05,
      "loss": 0.5437,
      "mean_token_accuracy": 0.8296671986579895,
      "num_tokens": 252794773.0,
      "step": 7545
    },
    {
      "epoch": 0.4506924546322827,
      "grad_norm": 0.5493040084838867,
      "learning_rate": 4.632415361008711e-05,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8295299887657166,
      "num_tokens": 252962453.0,
      "step": 7550
    },
    {
      "epoch": 0.4509909264565425,
      "grad_norm": 0.5395864844322205,
      "learning_rate": 4.6318068690806426e-05,
      "loss": 0.5483,
      "mean_token_accuracy": 0.8269950985908509,
      "num_tokens": 253130133.0,
      "step": 7555
    },
    {
      "epoch": 0.4512893982808023,
      "grad_norm": 0.5615440011024475,
      "learning_rate": 4.631197918803778e-05,
      "loss": 0.5749,
      "mean_token_accuracy": 0.8188894271850586,
      "num_tokens": 253297813.0,
      "step": 7560
    },
    {
      "epoch": 0.4515878701050621,
      "grad_norm": 0.6137975454330444,
      "learning_rate": 4.630588510326437e-05,
      "loss": 0.5627,
      "mean_token_accuracy": 0.8240904211997986,
      "num_tokens": 253465493.0,
      "step": 7565
    },
    {
      "epoch": 0.45188634192932187,
      "grad_norm": 0.6349193453788757,
      "learning_rate": 4.6299786437970524e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8202373862266541,
      "num_tokens": 253633173.0,
      "step": 7570
    },
    {
      "epoch": 0.45218481375358166,
      "grad_norm": 0.9740370512008667,
      "learning_rate": 4.6293683193641687e-05,
      "loss": 0.653,
      "mean_token_accuracy": 0.7981808423995972,
      "num_tokens": 253800853.0,
      "step": 7575
    },
    {
      "epoch": 0.45248328557784145,
      "grad_norm": 0.7178189158439636,
      "learning_rate": 4.628757537176442e-05,
      "loss": 0.612,
      "mean_token_accuracy": 0.8102111458778382,
      "num_tokens": 253968533.0,
      "step": 7580
    },
    {
      "epoch": 0.45278175740210125,
      "grad_norm": 0.5762429237365723,
      "learning_rate": 4.628146297382638e-05,
      "loss": 0.5698,
      "mean_token_accuracy": 0.820326852798462,
      "num_tokens": 254136213.0,
      "step": 7585
    },
    {
      "epoch": 0.45308022922636104,
      "grad_norm": 0.585010290145874,
      "learning_rate": 4.627534600131639e-05,
      "loss": 0.573,
      "mean_token_accuracy": 0.820199978351593,
      "num_tokens": 254301665.0,
      "step": 7590
    },
    {
      "epoch": 0.45337870105062084,
      "grad_norm": 0.6211780309677124,
      "learning_rate": 4.6269224455724305e-05,
      "loss": 0.552,
      "mean_token_accuracy": 0.8250626206398011,
      "num_tokens": 254469345.0,
      "step": 7595
    },
    {
      "epoch": 0.45367717287488063,
      "grad_norm": 0.5588411688804626,
      "learning_rate": 4.626309833854118e-05,
      "loss": 0.5663,
      "mean_token_accuracy": 0.8222593426704407,
      "num_tokens": 254637025.0,
      "step": 7600
    },
    {
      "epoch": 0.4539756446991404,
      "grad_norm": 0.7048266530036926,
      "learning_rate": 4.625696765125912e-05,
      "loss": 0.6264,
      "mean_token_accuracy": 0.807431697845459,
      "num_tokens": 254804705.0,
      "step": 7605
    },
    {
      "epoch": 0.4542741165234002,
      "grad_norm": 0.52033931016922,
      "learning_rate": 4.625083239537137e-05,
      "loss": 0.5326,
      "mean_token_accuracy": 0.830889880657196,
      "num_tokens": 254972385.0,
      "step": 7610
    },
    {
      "epoch": 0.45457258834765996,
      "grad_norm": 0.6067246794700623,
      "learning_rate": 4.62446925723723e-05,
      "loss": 0.5693,
      "mean_token_accuracy": 0.8232494235038758,
      "num_tokens": 255140065.0,
      "step": 7615
    },
    {
      "epoch": 0.45487106017191975,
      "grad_norm": 0.5551836490631104,
      "learning_rate": 4.6238548183757366e-05,
      "loss": 0.5984,
      "mean_token_accuracy": 0.812531316280365,
      "num_tokens": 255307745.0,
      "step": 7620
    },
    {
      "epoch": 0.45516953199617954,
      "grad_norm": 0.5718890428543091,
      "learning_rate": 4.623239923102314e-05,
      "loss": 0.5818,
      "mean_token_accuracy": 0.8176667094230652,
      "num_tokens": 255475425.0,
      "step": 7625
    },
    {
      "epoch": 0.45546800382043934,
      "grad_norm": 0.6421093940734863,
      "learning_rate": 4.6226245715667323e-05,
      "loss": 0.5761,
      "mean_token_accuracy": 0.8189311861991883,
      "num_tokens": 255643105.0,
      "step": 7630
    },
    {
      "epoch": 0.45576647564469913,
      "grad_norm": 0.6722767949104309,
      "learning_rate": 4.6220087639188713e-05,
      "loss": 0.5495,
      "mean_token_accuracy": 0.826792311668396,
      "num_tokens": 255810785.0,
      "step": 7635
    },
    {
      "epoch": 0.4560649474689589,
      "grad_norm": 0.5936805009841919,
      "learning_rate": 4.621392500308723e-05,
      "loss": 0.5401,
      "mean_token_accuracy": 0.8289812684059144,
      "num_tokens": 255978465.0,
      "step": 7640
    },
    {
      "epoch": 0.4563634192932187,
      "grad_norm": 0.5662449598312378,
      "learning_rate": 4.620775780886389e-05,
      "loss": 0.5726,
      "mean_token_accuracy": 0.8206429600715637,
      "num_tokens": 256146145.0,
      "step": 7645
    },
    {
      "epoch": 0.4566618911174785,
      "grad_norm": 0.6059299111366272,
      "learning_rate": 4.620158605802083e-05,
      "loss": 0.5853,
      "mean_token_accuracy": 0.8163962841033936,
      "num_tokens": 256313825.0,
      "step": 7650
    },
    {
      "epoch": 0.4569603629417383,
      "grad_norm": 0.5768796801567078,
      "learning_rate": 4.61954097520613e-05,
      "loss": 0.5785,
      "mean_token_accuracy": 0.8180603623390198,
      "num_tokens": 256481505.0,
      "step": 7655
    },
    {
      "epoch": 0.4572588347659981,
      "grad_norm": 0.5965715050697327,
      "learning_rate": 4.618922889248965e-05,
      "loss": 0.584,
      "mean_token_accuracy": 0.8151437401771545,
      "num_tokens": 256649185.0,
      "step": 7660
    },
    {
      "epoch": 0.4575573065902579,
      "grad_norm": 0.5589231848716736,
      "learning_rate": 4.6183043480811326e-05,
      "loss": 0.557,
      "mean_token_accuracy": 0.8250502586364746,
      "num_tokens": 256810593.0,
      "step": 7665
    },
    {
      "epoch": 0.4578557784145177,
      "grad_norm": 0.5479607582092285,
      "learning_rate": 4.617685351853292e-05,
      "loss": 0.5593,
      "mean_token_accuracy": 0.8241798758506775,
      "num_tokens": 256978273.0,
      "step": 7670
    },
    {
      "epoch": 0.4581542502387775,
      "grad_norm": 0.5506994724273682,
      "learning_rate": 4.617065900716211e-05,
      "loss": 0.6136,
      "mean_token_accuracy": 0.8089586019515991,
      "num_tokens": 257145953.0,
      "step": 7675
    },
    {
      "epoch": 0.4584527220630373,
      "grad_norm": 0.5962391495704651,
      "learning_rate": 4.616445994820766e-05,
      "loss": 0.5625,
      "mean_token_accuracy": 0.8232673168182373,
      "num_tokens": 257313633.0,
      "step": 7680
    },
    {
      "epoch": 0.458751193887297,
      "grad_norm": 0.5734214782714844,
      "learning_rate": 4.615825634317949e-05,
      "loss": 0.5455,
      "mean_token_accuracy": 0.8281999230384827,
      "num_tokens": 257481313.0,
      "step": 7685
    },
    {
      "epoch": 0.4590496657115568,
      "grad_norm": 0.596531867980957,
      "learning_rate": 4.615204819358861e-05,
      "loss": 0.6096,
      "mean_token_accuracy": 0.8095729351043701,
      "num_tokens": 257648993.0,
      "step": 7690
    },
    {
      "epoch": 0.4593481375358166,
      "grad_norm": 0.5632465481758118,
      "learning_rate": 4.6145835500947107e-05,
      "loss": 0.5528,
      "mean_token_accuracy": 0.8254026055335999,
      "num_tokens": 257816673.0,
      "step": 7695
    },
    {
      "epoch": 0.4596466093600764,
      "grad_norm": 0.6089497208595276,
      "learning_rate": 4.61396182667682e-05,
      "loss": 0.582,
      "mean_token_accuracy": 0.8167243242263794,
      "num_tokens": 257984353.0,
      "step": 7700
    },
    {
      "epoch": 0.4599450811843362,
      "grad_norm": 0.6036685109138489,
      "learning_rate": 4.613339649256622e-05,
      "loss": 0.5787,
      "mean_token_accuracy": 0.8192771077156067,
      "num_tokens": 258152033.0,
      "step": 7705
    },
    {
      "epoch": 0.460243553008596,
      "grad_norm": 0.553791880607605,
      "learning_rate": 4.612717017985659e-05,
      "loss": 0.5908,
      "mean_token_accuracy": 0.814762020111084,
      "num_tokens": 258319713.0,
      "step": 7710
    },
    {
      "epoch": 0.4605420248328558,
      "grad_norm": 0.563846230506897,
      "learning_rate": 4.6120939330155846e-05,
      "loss": 0.566,
      "mean_token_accuracy": 0.8215078234672546,
      "num_tokens": 258487393.0,
      "step": 7715
    },
    {
      "epoch": 0.4608404966571156,
      "grad_norm": 0.5484046936035156,
      "learning_rate": 4.611470394498162e-05,
      "loss": 0.5902,
      "mean_token_accuracy": 0.8158058047294616,
      "num_tokens": 258655073.0,
      "step": 7720
    },
    {
      "epoch": 0.46113896848137537,
      "grad_norm": 0.6056272387504578,
      "learning_rate": 4.610846402585267e-05,
      "loss": 0.5704,
      "mean_token_accuracy": 0.821579384803772,
      "num_tokens": 258822753.0,
      "step": 7725
    },
    {
      "epoch": 0.46143744030563516,
      "grad_norm": 0.5771185755729675,
      "learning_rate": 4.6102219574288825e-05,
      "loss": 0.5905,
      "mean_token_accuracy": 0.8153763532638549,
      "num_tokens": 258990433.0,
      "step": 7730
    },
    {
      "epoch": 0.46173591212989495,
      "grad_norm": 0.6007190942764282,
      "learning_rate": 4.609597059181104e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.8155493140220642,
      "num_tokens": 259158113.0,
      "step": 7735
    },
    {
      "epoch": 0.46203438395415475,
      "grad_norm": 0.5840270519256592,
      "learning_rate": 4.608971707994139e-05,
      "loss": 0.5755,
      "mean_token_accuracy": 0.8209054112434387,
      "num_tokens": 259325793.0,
      "step": 7740
    },
    {
      "epoch": 0.46233285577841454,
      "grad_norm": 0.6028809547424316,
      "learning_rate": 4.608345904020301e-05,
      "loss": 0.5914,
      "mean_token_accuracy": 0.8142132878303527,
      "num_tokens": 259493473.0,
      "step": 7745
    },
    {
      "epoch": 0.4626313276026743,
      "grad_norm": 0.5474963188171387,
      "learning_rate": 4.6077196474120176e-05,
      "loss": 0.5578,
      "mean_token_accuracy": 0.8243886470794678,
      "num_tokens": 259661153.0,
      "step": 7750
    },
    {
      "epoch": 0.4629297994269341,
      "grad_norm": 0.5470784306526184,
      "learning_rate": 4.607092938321824e-05,
      "loss": 0.5923,
      "mean_token_accuracy": 0.8146009802818298,
      "num_tokens": 259828833.0,
      "step": 7755
    },
    {
      "epoch": 0.46322827125119387,
      "grad_norm": 0.6136278510093689,
      "learning_rate": 4.606465776902367e-05,
      "loss": 0.534,
      "mean_token_accuracy": 0.8311582922935485,
      "num_tokens": 259996513.0,
      "step": 7760
    },
    {
      "epoch": 0.46352674307545366,
      "grad_norm": 0.5415297746658325,
      "learning_rate": 4.6058381633064036e-05,
      "loss": 0.5611,
      "mean_token_accuracy": 0.8241023659706116,
      "num_tokens": 260164193.0,
      "step": 7765
    },
    {
      "epoch": 0.46382521489971346,
      "grad_norm": 0.5938883423805237,
      "learning_rate": 4.6052100976868006e-05,
      "loss": 0.6202,
      "mean_token_accuracy": 0.8048968195915223,
      "num_tokens": 260331873.0,
      "step": 7770
    },
    {
      "epoch": 0.46412368672397325,
      "grad_norm": 0.5361422300338745,
      "learning_rate": 4.604581580196536e-05,
      "loss": 0.6023,
      "mean_token_accuracy": 0.8137242197990417,
      "num_tokens": 260499553.0,
      "step": 7775
    },
    {
      "epoch": 0.46442215854823304,
      "grad_norm": 0.5868556499481201,
      "learning_rate": 4.603952610988695e-05,
      "loss": 0.5991,
      "mean_token_accuracy": 0.8122450351715088,
      "num_tokens": 260667233.0,
      "step": 7780
    },
    {
      "epoch": 0.46472063037249284,
      "grad_norm": 0.6046263575553894,
      "learning_rate": 4.603323190216476e-05,
      "loss": 0.5591,
      "mean_token_accuracy": 0.8242932081222534,
      "num_tokens": 260834913.0,
      "step": 7785
    },
    {
      "epoch": 0.46501910219675263,
      "grad_norm": 0.7702558636665344,
      "learning_rate": 4.602693318033185e-05,
      "loss": 0.5488,
      "mean_token_accuracy": 0.8276690721511841,
      "num_tokens": 261002593.0,
      "step": 7790
    },
    {
      "epoch": 0.4653175740210124,
      "grad_norm": 0.5694584846496582,
      "learning_rate": 4.602062994592241e-05,
      "loss": 0.5991,
      "mean_token_accuracy": 0.8131158351898193,
      "num_tokens": 261170273.0,
      "step": 7795
    },
    {
      "epoch": 0.4656160458452722,
      "grad_norm": 0.5280610918998718,
      "learning_rate": 4.601432220047168e-05,
      "loss": 0.5554,
      "mean_token_accuracy": 0.8255040049552917,
      "num_tokens": 261337953.0,
      "step": 7800
    },
    {
      "epoch": 0.465914517669532,
      "grad_norm": 0.5759225487709045,
      "learning_rate": 4.6008009945516054e-05,
      "loss": 0.5879,
      "mean_token_accuracy": 0.816742217540741,
      "num_tokens": 261505633.0,
      "step": 7805
    },
    {
      "epoch": 0.4662129894937918,
      "grad_norm": 0.5653066635131836,
      "learning_rate": 4.600169318259297e-05,
      "loss": 0.5402,
      "mean_token_accuracy": 0.827269458770752,
      "num_tokens": 261673313.0,
      "step": 7810
    },
    {
      "epoch": 0.4665114613180516,
      "grad_norm": 0.5461359620094299,
      "learning_rate": 4.599537191324102e-05,
      "loss": 0.5447,
      "mean_token_accuracy": 0.8271979093551636,
      "num_tokens": 261840993.0,
      "step": 7815
    },
    {
      "epoch": 0.46680993314231134,
      "grad_norm": 0.5427731871604919,
      "learning_rate": 4.598904613899985e-05,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8306722164154052,
      "num_tokens": 262004519.0,
      "step": 7820
    },
    {
      "epoch": 0.46710840496657113,
      "grad_norm": 0.5485845804214478,
      "learning_rate": 4.598271586141024e-05,
      "loss": 0.5831,
      "mean_token_accuracy": 0.8175772428512573,
      "num_tokens": 262172199.0,
      "step": 7825
    },
    {
      "epoch": 0.4674068767908309,
      "grad_norm": 0.5768746733665466,
      "learning_rate": 4.5976381082014016e-05,
      "loss": 0.5683,
      "mean_token_accuracy": 0.8193904399871826,
      "num_tokens": 262339879.0,
      "step": 7830
    },
    {
      "epoch": 0.4677053486150907,
      "grad_norm": 0.5218663215637207,
      "learning_rate": 4.597004180235415e-05,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8376476287841796,
      "num_tokens": 262507559.0,
      "step": 7835
    },
    {
      "epoch": 0.4680038204393505,
      "grad_norm": 0.5930300354957581,
      "learning_rate": 4.596369802397468e-05,
      "loss": 0.5666,
      "mean_token_accuracy": 0.8224144220352173,
      "num_tokens": 262675239.0,
      "step": 7840
    },
    {
      "epoch": 0.4683022922636103,
      "grad_norm": 0.5332184433937073,
      "learning_rate": 4.595734974842076e-05,
      "loss": 0.6273,
      "mean_token_accuracy": 0.8054932475090026,
      "num_tokens": 262842919.0,
      "step": 7845
    },
    {
      "epoch": 0.4686007640878701,
      "grad_norm": 0.5373886227607727,
      "learning_rate": 4.5950996977238625e-05,
      "loss": 0.569,
      "mean_token_accuracy": 0.8213765859603882,
      "num_tokens": 263010599.0,
      "step": 7850
    },
    {
      "epoch": 0.4688992359121299,
      "grad_norm": 0.5584951639175415,
      "learning_rate": 4.594463971197561e-05,
      "loss": 0.5568,
      "mean_token_accuracy": 0.8220267176628113,
      "num_tokens": 263178279.0,
      "step": 7855
    },
    {
      "epoch": 0.4691977077363897,
      "grad_norm": 0.510094165802002,
      "learning_rate": 4.593827795418014e-05,
      "loss": 0.545,
      "mean_token_accuracy": 0.8298282146453857,
      "num_tokens": 263345959.0,
      "step": 7860
    },
    {
      "epoch": 0.4694961795606495,
      "grad_norm": 0.5799757242202759,
      "learning_rate": 4.593191170540175e-05,
      "loss": 0.5829,
      "mean_token_accuracy": 0.8147143006324769,
      "num_tokens": 263513639.0,
      "step": 7865
    },
    {
      "epoch": 0.4697946513849093,
      "grad_norm": 1.1775456666946411,
      "learning_rate": 4.5925540967191044e-05,
      "loss": 0.6029,
      "mean_token_accuracy": 0.8124657034873962,
      "num_tokens": 263681319.0,
      "step": 7870
    },
    {
      "epoch": 0.4700931232091691,
      "grad_norm": 0.9137069582939148,
      "learning_rate": 4.591916574109974e-05,
      "loss": 0.6164,
      "mean_token_accuracy": 0.8097817063331604,
      "num_tokens": 263848999.0,
      "step": 7875
    },
    {
      "epoch": 0.47039159503342887,
      "grad_norm": 0.6295863389968872,
      "learning_rate": 4.591278602868064e-05,
      "loss": 0.5946,
      "mean_token_accuracy": 0.814887273311615,
      "num_tokens": 264016679.0,
      "step": 7880
    },
    {
      "epoch": 0.47069006685768866,
      "grad_norm": 0.5366268754005432,
      "learning_rate": 4.5906401831487644e-05,
      "loss": 0.5492,
      "mean_token_accuracy": 0.8271084308624268,
      "num_tokens": 264184359.0,
      "step": 7885
    },
    {
      "epoch": 0.4709885386819484,
      "grad_norm": 0.5900691151618958,
      "learning_rate": 4.590001315107573e-05,
      "loss": 0.5808,
      "mean_token_accuracy": 0.8175786972045899,
      "num_tokens": 264347485.0,
      "step": 7890
    },
    {
      "epoch": 0.4712870105062082,
      "grad_norm": 0.5910937190055847,
      "learning_rate": 4.5893619989001005e-05,
      "loss": 0.5607,
      "mean_token_accuracy": 0.8234223961830139,
      "num_tokens": 264515165.0,
      "step": 7895
    },
    {
      "epoch": 0.471585482330468,
      "grad_norm": 0.5602463483810425,
      "learning_rate": 4.588722234682061e-05,
      "loss": 0.5952,
      "mean_token_accuracy": 0.8131635427474976,
      "num_tokens": 264682845.0,
      "step": 7900
    },
    {
      "epoch": 0.4718839541547278,
      "grad_norm": 0.5950137376785278,
      "learning_rate": 4.5880820226092824e-05,
      "loss": 0.5476,
      "mean_token_accuracy": 0.8288262009620666,
      "num_tokens": 264850525.0,
      "step": 7905
    },
    {
      "epoch": 0.4721824259789876,
      "grad_norm": 0.4922201633453369,
      "learning_rate": 4.5874413628377014e-05,
      "loss": 0.6,
      "mean_token_accuracy": 0.8148097395896912,
      "num_tokens": 265018205.0,
      "step": 7910
    },
    {
      "epoch": 0.47248089780324737,
      "grad_norm": 0.5623324513435364,
      "learning_rate": 4.5868002555233594e-05,
      "loss": 0.5545,
      "mean_token_accuracy": 0.8247047781944274,
      "num_tokens": 265185885.0,
      "step": 7915
    },
    {
      "epoch": 0.47277936962750716,
      "grad_norm": 0.5618387460708618,
      "learning_rate": 4.586158700822413e-05,
      "loss": 0.5672,
      "mean_token_accuracy": 0.8208696126937867,
      "num_tokens": 265353565.0,
      "step": 7920
    },
    {
      "epoch": 0.47307784145176696,
      "grad_norm": 0.5495238304138184,
      "learning_rate": 4.585516698891123e-05,
      "loss": 0.553,
      "mean_token_accuracy": 0.8243349671363831,
      "num_tokens": 265521245.0,
      "step": 7925
    },
    {
      "epoch": 0.47337631327602675,
      "grad_norm": 0.5636260509490967,
      "learning_rate": 4.584874249885861e-05,
      "loss": 0.5642,
      "mean_token_accuracy": 0.8226350903511047,
      "num_tokens": 265688925.0,
      "step": 7930
    },
    {
      "epoch": 0.47367478510028654,
      "grad_norm": 0.5569517016410828,
      "learning_rate": 4.584231353963108e-05,
      "loss": 0.551,
      "mean_token_accuracy": 0.8252833008766174,
      "num_tokens": 265856605.0,
      "step": 7935
    },
    {
      "epoch": 0.47397325692454634,
      "grad_norm": 0.6061868667602539,
      "learning_rate": 4.5835880112794524e-05,
      "loss": 0.5728,
      "mean_token_accuracy": 0.8192174673080445,
      "num_tokens": 266024285.0,
      "step": 7940
    },
    {
      "epoch": 0.47427172874880613,
      "grad_norm": 0.5769259333610535,
      "learning_rate": 4.582944221991592e-05,
      "loss": 0.6292,
      "mean_token_accuracy": 0.8031611561775207,
      "num_tokens": 266191965.0,
      "step": 7945
    },
    {
      "epoch": 0.4745702005730659,
      "grad_norm": 0.5785863995552063,
      "learning_rate": 4.582299986256335e-05,
      "loss": 0.5473,
      "mean_token_accuracy": 0.8263986706733704,
      "num_tokens": 266359645.0,
      "step": 7950
    },
    {
      "epoch": 0.4748686723973257,
      "grad_norm": 0.5971212387084961,
      "learning_rate": 4.581655304230596e-05,
      "loss": 0.586,
      "mean_token_accuracy": 0.81563880443573,
      "num_tokens": 266527325.0,
      "step": 7955
    },
    {
      "epoch": 0.47516714422158546,
      "grad_norm": 0.608715832233429,
      "learning_rate": 4.581010176071399e-05,
      "loss": 0.5468,
      "mean_token_accuracy": 0.8272098302841187,
      "num_tokens": 266695005.0,
      "step": 7960
    },
    {
      "epoch": 0.47546561604584525,
      "grad_norm": 0.5763922929763794,
      "learning_rate": 4.5803646019358764e-05,
      "loss": 0.5603,
      "mean_token_accuracy": 0.8245735406875611,
      "num_tokens": 266862685.0,
      "step": 7965
    },
    {
      "epoch": 0.47576408787010505,
      "grad_norm": 0.5652427673339844,
      "learning_rate": 4.5797185819812705e-05,
      "loss": 0.5712,
      "mean_token_accuracy": 0.8206310391426086,
      "num_tokens": 267030365.0,
      "step": 7970
    },
    {
      "epoch": 0.47606255969436484,
      "grad_norm": 0.5526332855224609,
      "learning_rate": 4.579072116364932e-05,
      "loss": 0.5373,
      "mean_token_accuracy": 0.8294405341148376,
      "num_tokens": 267198045.0,
      "step": 7975
    },
    {
      "epoch": 0.47636103151862463,
      "grad_norm": 0.5545656085014343,
      "learning_rate": 4.578425205244318e-05,
      "loss": 0.5415,
      "mean_token_accuracy": 0.8296552538871765,
      "num_tokens": 267365725.0,
      "step": 7980
    },
    {
      "epoch": 0.4766595033428844,
      "grad_norm": 0.5866954922676086,
      "learning_rate": 4.577777848776997e-05,
      "loss": 0.559,
      "mean_token_accuracy": 0.8245079278945923,
      "num_tokens": 267533405.0,
      "step": 7985
    },
    {
      "epoch": 0.4769579751671442,
      "grad_norm": 0.631314218044281,
      "learning_rate": 4.577130047120643e-05,
      "loss": 0.5939,
      "mean_token_accuracy": 0.8154777526855469,
      "num_tokens": 267701085.0,
      "step": 7990
    },
    {
      "epoch": 0.477256446991404,
      "grad_norm": 0.6295937299728394,
      "learning_rate": 4.576481800433042e-05,
      "loss": 0.5576,
      "mean_token_accuracy": 0.8249135136604309,
      "num_tokens": 267868765.0,
      "step": 7995
    },
    {
      "epoch": 0.4775549188156638,
      "grad_norm": 0.6489912271499634,
      "learning_rate": 4.575833108872085e-05,
      "loss": 0.6152,
      "mean_token_accuracy": 0.81052725315094,
      "num_tokens": 268036445.0,
      "step": 8000
    },
    {
      "epoch": 0.4778533906399236,
      "grad_norm": 0.533629834651947,
      "learning_rate": 4.575183972595774e-05,
      "loss": 0.6079,
      "mean_token_accuracy": 0.806453537940979,
      "num_tokens": 268204125.0,
      "step": 8005
    },
    {
      "epoch": 0.4781518624641834,
      "grad_norm": 0.5676308274269104,
      "learning_rate": 4.574534391762216e-05,
      "loss": 0.6181,
      "mean_token_accuracy": 0.8083084940910339,
      "num_tokens": 268371805.0,
      "step": 8010
    },
    {
      "epoch": 0.4784503342884432,
      "grad_norm": 0.5404603481292725,
      "learning_rate": 4.573884366529632e-05,
      "loss": 0.5824,
      "mean_token_accuracy": 0.8169569373130798,
      "num_tokens": 268539485.0,
      "step": 8015
    },
    {
      "epoch": 0.478748806112703,
      "grad_norm": 0.5885908007621765,
      "learning_rate": 4.573233897056344e-05,
      "loss": 0.5595,
      "mean_token_accuracy": 0.8206489205360412,
      "num_tokens": 268707165.0,
      "step": 8020
    },
    {
      "epoch": 0.4790472779369627,
      "grad_norm": 0.5349931716918945,
      "learning_rate": 4.572582983500787e-05,
      "loss": 0.6162,
      "mean_token_accuracy": 0.8094596147537232,
      "num_tokens": 268874845.0,
      "step": 8025
    },
    {
      "epoch": 0.4793457497612225,
      "grad_norm": 0.541793704032898,
      "learning_rate": 4.5719316260215045e-05,
      "loss": 0.5807,
      "mean_token_accuracy": 0.8177561640739441,
      "num_tokens": 269042525.0,
      "step": 8030
    },
    {
      "epoch": 0.4796442215854823,
      "grad_norm": 0.5695494413375854,
      "learning_rate": 4.571279824777145e-05,
      "loss": 0.5976,
      "mean_token_accuracy": 0.8117559313774109,
      "num_tokens": 269210205.0,
      "step": 8035
    },
    {
      "epoch": 0.4799426934097421,
      "grad_norm": 0.5621160864830017,
      "learning_rate": 4.570627579926468e-05,
      "loss": 0.6179,
      "mean_token_accuracy": 0.8079923748970032,
      "num_tokens": 269377885.0,
      "step": 8040
    },
    {
      "epoch": 0.4802411652340019,
      "grad_norm": 0.5891603231430054,
      "learning_rate": 4.569974891628337e-05,
      "loss": 0.585,
      "mean_token_accuracy": 0.8176070570945739,
      "num_tokens": 269545565.0,
      "step": 8045
    },
    {
      "epoch": 0.4805396370582617,
      "grad_norm": 0.5767423510551453,
      "learning_rate": 4.56932176004173e-05,
      "loss": 0.591,
      "mean_token_accuracy": 0.8160443782806397,
      "num_tokens": 269713245.0,
      "step": 8050
    },
    {
      "epoch": 0.4808381088825215,
      "grad_norm": 0.5442712903022766,
      "learning_rate": 4.568668185325726e-05,
      "loss": 0.5447,
      "mean_token_accuracy": 0.8274364709854126,
      "num_tokens": 269880925.0,
      "step": 8055
    },
    {
      "epoch": 0.4811365807067813,
      "grad_norm": 0.5646932125091553,
      "learning_rate": 4.568014167639518e-05,
      "loss": 0.5992,
      "mean_token_accuracy": 0.814135754108429,
      "num_tokens": 270048605.0,
      "step": 8060
    },
    {
      "epoch": 0.4814350525310411,
      "grad_norm": 0.5659125447273254,
      "learning_rate": 4.567359707142402e-05,
      "loss": 0.5365,
      "mean_token_accuracy": 0.8292675614356995,
      "num_tokens": 270216285.0,
      "step": 8065
    },
    {
      "epoch": 0.48173352435530087,
      "grad_norm": 0.5524060726165771,
      "learning_rate": 4.5667048039937854e-05,
      "loss": 0.5121,
      "mean_token_accuracy": 0.8350172877311707,
      "num_tokens": 270383965.0,
      "step": 8070
    },
    {
      "epoch": 0.48203199617956066,
      "grad_norm": 0.6202818155288696,
      "learning_rate": 4.566049458353181e-05,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8290289998054504,
      "num_tokens": 270551645.0,
      "step": 8075
    },
    {
      "epoch": 0.48233046800382046,
      "grad_norm": 0.6832270622253418,
      "learning_rate": 4.56539367038021e-05,
      "loss": 0.5961,
      "mean_token_accuracy": 0.8125969290733337,
      "num_tokens": 270719325.0,
      "step": 8080
    },
    {
      "epoch": 0.48262893982808025,
      "grad_norm": 0.6123245358467102,
      "learning_rate": 4.564737440234604e-05,
      "loss": 0.6067,
      "mean_token_accuracy": 0.8086663484573364,
      "num_tokens": 270887005.0,
      "step": 8085
    },
    {
      "epoch": 0.48292741165234004,
      "grad_norm": 0.7667441368103027,
      "learning_rate": 4.564080768076196e-05,
      "loss": 0.5875,
      "mean_token_accuracy": 0.8150602459907532,
      "num_tokens": 271054685.0,
      "step": 8090
    },
    {
      "epoch": 0.4832258834765998,
      "grad_norm": 0.5241187214851379,
      "learning_rate": 4.563423654064934e-05,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8302934527397156,
      "num_tokens": 271222365.0,
      "step": 8095
    },
    {
      "epoch": 0.4835243553008596,
      "grad_norm": 0.6005362272262573,
      "learning_rate": 4.5627660983608696e-05,
      "loss": 0.5878,
      "mean_token_accuracy": 0.8154359936714173,
      "num_tokens": 271390045.0,
      "step": 8100
    },
    {
      "epoch": 0.48382282712511937,
      "grad_norm": 0.6519680023193359,
      "learning_rate": 4.5621081011241615e-05,
      "loss": 0.5725,
      "mean_token_accuracy": 0.8208815455436707,
      "num_tokens": 271557725.0,
      "step": 8105
    },
    {
      "epoch": 0.48412129894937916,
      "grad_norm": 0.6114306449890137,
      "learning_rate": 4.5614496625150774e-05,
      "loss": 0.6172,
      "mean_token_accuracy": 0.809811532497406,
      "num_tokens": 271725405.0,
      "step": 8110
    },
    {
      "epoch": 0.48441977077363896,
      "grad_norm": 0.527331531047821,
      "learning_rate": 4.560790782693993e-05,
      "loss": 0.5839,
      "mean_token_accuracy": 0.8166587233543396,
      "num_tokens": 271893085.0,
      "step": 8115
    },
    {
      "epoch": 0.48471824259789875,
      "grad_norm": 0.5360596179962158,
      "learning_rate": 4.560131461821389e-05,
      "loss": 0.563,
      "mean_token_accuracy": 0.824424433708191,
      "num_tokens": 272060765.0,
      "step": 8120
    },
    {
      "epoch": 0.48501671442215855,
      "grad_norm": 0.6610274910926819,
      "learning_rate": 4.5594717000578577e-05,
      "loss": 0.565,
      "mean_token_accuracy": 0.8247763395309449,
      "num_tokens": 272228445.0,
      "step": 8125
    },
    {
      "epoch": 0.48531518624641834,
      "grad_norm": 0.583880603313446,
      "learning_rate": 4.5588114975640944e-05,
      "loss": 0.5612,
      "mean_token_accuracy": 0.823434317111969,
      "num_tokens": 272396125.0,
      "step": 8130
    },
    {
      "epoch": 0.48561365807067813,
      "grad_norm": 0.5621193647384644,
      "learning_rate": 4.558150854500903e-05,
      "loss": 0.5396,
      "mean_token_accuracy": 0.8299176931381226,
      "num_tokens": 272563805.0,
      "step": 8135
    },
    {
      "epoch": 0.4859121298949379,
      "grad_norm": 0.5357057452201843,
      "learning_rate": 4.5574897710291975e-05,
      "loss": 0.5575,
      "mean_token_accuracy": 0.8243170738220215,
      "num_tokens": 272731485.0,
      "step": 8140
    },
    {
      "epoch": 0.4862106017191977,
      "grad_norm": 0.5147784948348999,
      "learning_rate": 4.556828247309995e-05,
      "loss": 0.5261,
      "mean_token_accuracy": 0.8347488880157471,
      "num_tokens": 272899165.0,
      "step": 8145
    },
    {
      "epoch": 0.4865090735434575,
      "grad_norm": 0.580520510673523,
      "learning_rate": 4.556166283504424e-05,
      "loss": 0.576,
      "mean_token_accuracy": 0.820350706577301,
      "num_tokens": 273066845.0,
      "step": 8150
    },
    {
      "epoch": 0.4868075453677173,
      "grad_norm": 0.5211859345436096,
      "learning_rate": 4.555503879773715e-05,
      "loss": 0.5627,
      "mean_token_accuracy": 0.8230347037315369,
      "num_tokens": 273234525.0,
      "step": 8155
    },
    {
      "epoch": 0.4871060171919771,
      "grad_norm": 0.5352399945259094,
      "learning_rate": 4.554841036279212e-05,
      "loss": 0.5684,
      "mean_token_accuracy": 0.8206429839134216,
      "num_tokens": 273402205.0,
      "step": 8160
    },
    {
      "epoch": 0.48740448901623684,
      "grad_norm": 0.5476245284080505,
      "learning_rate": 4.55417775318236e-05,
      "loss": 0.5893,
      "mean_token_accuracy": 0.8178754568099975,
      "num_tokens": 273569885.0,
      "step": 8165
    },
    {
      "epoch": 0.48770296084049664,
      "grad_norm": 0.5912073850631714,
      "learning_rate": 4.553514030644715e-05,
      "loss": 0.5861,
      "mean_token_accuracy": 0.8170821905136109,
      "num_tokens": 273737565.0,
      "step": 8170
    },
    {
      "epoch": 0.48800143266475643,
      "grad_norm": 0.5961964726448059,
      "learning_rate": 4.552849868827939e-05,
      "loss": 0.5728,
      "mean_token_accuracy": 0.8198854804039002,
      "num_tokens": 273905245.0,
      "step": 8175
    },
    {
      "epoch": 0.4882999044890162,
      "grad_norm": 0.5517949461936951,
      "learning_rate": 4.5521852678938e-05,
      "loss": 0.6329,
      "mean_token_accuracy": 0.8060718059539795,
      "num_tokens": 274072925.0,
      "step": 8180
    },
    {
      "epoch": 0.488598376313276,
      "grad_norm": 0.5162144899368286,
      "learning_rate": 4.551520228004175e-05,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8275498151779175,
      "num_tokens": 274240605.0,
      "step": 8185
    },
    {
      "epoch": 0.4888968481375358,
      "grad_norm": 0.5340211391448975,
      "learning_rate": 4.550854749321046e-05,
      "loss": 0.5828,
      "mean_token_accuracy": 0.8140701413154602,
      "num_tokens": 274408285.0,
      "step": 8190
    },
    {
      "epoch": 0.4891953199617956,
      "grad_norm": 0.5133636593818665,
      "learning_rate": 4.550188832006503e-05,
      "loss": 0.568,
      "mean_token_accuracy": 0.8207324504852295,
      "num_tokens": 274575965.0,
      "step": 8195
    },
    {
      "epoch": 0.4894937917860554,
      "grad_norm": 0.6545477509498596,
      "learning_rate": 4.549522476222742e-05,
      "loss": 0.6044,
      "mean_token_accuracy": 0.8120183706283569,
      "num_tokens": 274743645.0,
      "step": 8200
    },
    {
      "epoch": 0.4897922636103152,
      "grad_norm": 0.4933609366416931,
      "learning_rate": 4.5488556821320676e-05,
      "loss": 0.5966,
      "mean_token_accuracy": 0.814911139011383,
      "num_tokens": 274911325.0,
      "step": 8205
    },
    {
      "epoch": 0.490090735434575,
      "grad_norm": 0.6106254458427429,
      "learning_rate": 4.548188449896888e-05,
      "loss": 0.6006,
      "mean_token_accuracy": 0.8145294070243836,
      "num_tokens": 275079005.0,
      "step": 8210
    },
    {
      "epoch": 0.4903892072588348,
      "grad_norm": 0.540390133857727,
      "learning_rate": 4.5475207796797206e-05,
      "loss": 0.5864,
      "mean_token_accuracy": 0.8152689933776855,
      "num_tokens": 275246685.0,
      "step": 8215
    },
    {
      "epoch": 0.4906876790830946,
      "grad_norm": 0.5468898415565491,
      "learning_rate": 4.546852671643189e-05,
      "loss": 0.5263,
      "mean_token_accuracy": 0.8331027150154113,
      "num_tokens": 275414365.0,
      "step": 8220
    },
    {
      "epoch": 0.49098615090735437,
      "grad_norm": 0.5242897868156433,
      "learning_rate": 4.546184125950023e-05,
      "loss": 0.5768,
      "mean_token_accuracy": 0.8187641620635986,
      "num_tokens": 275582045.0,
      "step": 8225
    },
    {
      "epoch": 0.49128462273161416,
      "grad_norm": 0.5289314389228821,
      "learning_rate": 4.54551514276306e-05,
      "loss": 0.5213,
      "mean_token_accuracy": 0.833633553981781,
      "num_tokens": 275749725.0,
      "step": 8230
    },
    {
      "epoch": 0.4915830945558739,
      "grad_norm": 0.7056505084037781,
      "learning_rate": 4.544845722245242e-05,
      "loss": 0.6329,
      "mean_token_accuracy": 0.8035309553146363,
      "num_tokens": 275917405.0,
      "step": 8235
    },
    {
      "epoch": 0.4918815663801337,
      "grad_norm": 0.560248851776123,
      "learning_rate": 4.544175864559619e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8206310272216797,
      "num_tokens": 276085085.0,
      "step": 8240
    },
    {
      "epoch": 0.4921800382043935,
      "grad_norm": 0.7740751504898071,
      "learning_rate": 4.543505569869348e-05,
      "loss": 0.5649,
      "mean_token_accuracy": 0.8229213953018188,
      "num_tokens": 276252765.0,
      "step": 8245
    },
    {
      "epoch": 0.4924785100286533,
      "grad_norm": 0.5696920156478882,
      "learning_rate": 4.54283483833769e-05,
      "loss": 0.5675,
      "mean_token_accuracy": 0.8218060255050659,
      "num_tokens": 276420445.0,
      "step": 8250
    },
    {
      "epoch": 0.4927769818529131,
      "grad_norm": 0.5968448519706726,
      "learning_rate": 4.542163670128017e-05,
      "loss": 0.605,
      "mean_token_accuracy": 0.8120124220848084,
      "num_tokens": 276588125.0,
      "step": 8255
    },
    {
      "epoch": 0.49307545367717287,
      "grad_norm": 0.54146808385849,
      "learning_rate": 4.5414920654038014e-05,
      "loss": 0.5598,
      "mean_token_accuracy": 0.8222295045852661,
      "num_tokens": 276755805.0,
      "step": 8260
    },
    {
      "epoch": 0.49337392550143266,
      "grad_norm": 0.555019199848175,
      "learning_rate": 4.540820024328627e-05,
      "loss": 0.5739,
      "mean_token_accuracy": 0.8204163312911987,
      "num_tokens": 276923485.0,
      "step": 8265
    },
    {
      "epoch": 0.49367239732569246,
      "grad_norm": 0.5188547968864441,
      "learning_rate": 4.54014754706618e-05,
      "loss": 0.6033,
      "mean_token_accuracy": 0.8122867822647095,
      "num_tokens": 277091165.0,
      "step": 8270
    },
    {
      "epoch": 0.49397086914995225,
      "grad_norm": 0.525219738483429,
      "learning_rate": 4.539474633780256e-05,
      "loss": 0.5942,
      "mean_token_accuracy": 0.8140939950942994,
      "num_tokens": 277258845.0,
      "step": 8275
    },
    {
      "epoch": 0.49426934097421205,
      "grad_norm": 0.5421309471130371,
      "learning_rate": 4.538801284634755e-05,
      "loss": 0.543,
      "mean_token_accuracy": 0.8292735338211059,
      "num_tokens": 277426525.0,
      "step": 8280
    },
    {
      "epoch": 0.49456781279847184,
      "grad_norm": 0.4715145528316498,
      "learning_rate": 4.538127499793683e-05,
      "loss": 0.5421,
      "mean_token_accuracy": 0.8290826678276062,
      "num_tokens": 277594205.0,
      "step": 8285
    },
    {
      "epoch": 0.49486628462273163,
      "grad_norm": 0.5990399122238159,
      "learning_rate": 4.5374532794211544e-05,
      "loss": 0.6005,
      "mean_token_accuracy": 0.8137480616569519,
      "num_tokens": 277761885.0,
      "step": 8290
    },
    {
      "epoch": 0.4951647564469914,
      "grad_norm": 0.6102491021156311,
      "learning_rate": 4.5367786236813856e-05,
      "loss": 0.5773,
      "mean_token_accuracy": 0.8196528553962708,
      "num_tokens": 277929565.0,
      "step": 8295
    },
    {
      "epoch": 0.4954632282712512,
      "grad_norm": 0.5606467127799988,
      "learning_rate": 4.536103532738704e-05,
      "loss": 0.5709,
      "mean_token_accuracy": 0.8191995739936828,
      "num_tokens": 278097245.0,
      "step": 8300
    },
    {
      "epoch": 0.49576170009551096,
      "grad_norm": 0.5498657822608948,
      "learning_rate": 4.535428006757539e-05,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8396516799926758,
      "num_tokens": 278264925.0,
      "step": 8305
    },
    {
      "epoch": 0.49606017191977075,
      "grad_norm": 0.5730018615722656,
      "learning_rate": 4.5347520459024285e-05,
      "loss": 0.6072,
      "mean_token_accuracy": 0.8089765071868896,
      "num_tokens": 278432605.0,
      "step": 8310
    },
    {
      "epoch": 0.49635864374403055,
      "grad_norm": 0.5413969159126282,
      "learning_rate": 4.534075650338014e-05,
      "loss": 0.5681,
      "mean_token_accuracy": 0.8203984379768372,
      "num_tokens": 278600285.0,
      "step": 8315
    },
    {
      "epoch": 0.49665711556829034,
      "grad_norm": 0.5423916578292847,
      "learning_rate": 4.533398820229044e-05,
      "loss": 0.5756,
      "mean_token_accuracy": 0.8190504550933838,
      "num_tokens": 278767965.0,
      "step": 8320
    },
    {
      "epoch": 0.49695558739255014,
      "grad_norm": 0.5244458317756653,
      "learning_rate": 4.532721555740374e-05,
      "loss": 0.5436,
      "mean_token_accuracy": 0.8271263241767883,
      "num_tokens": 278935645.0,
      "step": 8325
    },
    {
      "epoch": 0.49725405921680993,
      "grad_norm": 0.6236850023269653,
      "learning_rate": 4.5320438570369636e-05,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8316831588745117,
      "num_tokens": 279103325.0,
      "step": 8330
    },
    {
      "epoch": 0.4975525310410697,
      "grad_norm": 0.543181300163269,
      "learning_rate": 4.53136572428388e-05,
      "loss": 0.5671,
      "mean_token_accuracy": 0.8195753216743469,
      "num_tokens": 279271005.0,
      "step": 8335
    },
    {
      "epoch": 0.4978510028653295,
      "grad_norm": 0.5347437262535095,
      "learning_rate": 4.530687157646293e-05,
      "loss": 0.5683,
      "mean_token_accuracy": 0.8222831964492798,
      "num_tokens": 279438685.0,
      "step": 8340
    },
    {
      "epoch": 0.4981494746895893,
      "grad_norm": 0.5316510796546936,
      "learning_rate": 4.530008157289481e-05,
      "loss": 0.6315,
      "mean_token_accuracy": 0.8046284198760987,
      "num_tokens": 279606365.0,
      "step": 8345
    },
    {
      "epoch": 0.4984479465138491,
      "grad_norm": 0.6096969842910767,
      "learning_rate": 4.5293287233788276e-05,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8230406880378723,
      "num_tokens": 279774045.0,
      "step": 8350
    },
    {
      "epoch": 0.4987464183381089,
      "grad_norm": 0.6531689167022705,
      "learning_rate": 4.528648856079821e-05,
      "loss": 0.5523,
      "mean_token_accuracy": 0.8257604598999023,
      "num_tokens": 279941725.0,
      "step": 8355
    },
    {
      "epoch": 0.4990448901623687,
      "grad_norm": 0.5208596587181091,
      "learning_rate": 4.527968555558055e-05,
      "loss": 0.5436,
      "mean_token_accuracy": 0.8273291230201721,
      "num_tokens": 280109405.0,
      "step": 8360
    },
    {
      "epoch": 0.4993433619866285,
      "grad_norm": 0.5364457368850708,
      "learning_rate": 4.52728782197923e-05,
      "loss": 0.5945,
      "mean_token_accuracy": 0.8145174741744995,
      "num_tokens": 280277085.0,
      "step": 8365
    },
    {
      "epoch": 0.4996418338108882,
      "grad_norm": 0.574241578578949,
      "learning_rate": 4.5266066555091506e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8201956391334534,
      "num_tokens": 280444765.0,
      "step": 8370
    },
    {
      "epoch": 0.499940305635148,
      "grad_norm": 0.5839861631393433,
      "learning_rate": 4.5259250563137284e-05,
      "loss": 0.5826,
      "mean_token_accuracy": 0.8177144289016723,
      "num_tokens": 280612445.0,
      "step": 8375
    },
    {
      "epoch": 0.5002387774594078,
      "grad_norm": 0.6105102896690369,
      "learning_rate": 4.5252430245589786e-05,
      "loss": 0.5819,
      "mean_token_accuracy": 0.8183048963546753,
      "num_tokens": 280780125.0,
      "step": 8380
    },
    {
      "epoch": 0.5005372492836676,
      "grad_norm": 0.6635046005249023,
      "learning_rate": 4.524560560411023e-05,
      "loss": 0.5976,
      "mean_token_accuracy": 0.8133543968200684,
      "num_tokens": 280947805.0,
      "step": 8385
    },
    {
      "epoch": 0.5008357211079274,
      "grad_norm": 0.6643062233924866,
      "learning_rate": 4.523877664036088e-05,
      "loss": 0.5587,
      "mean_token_accuracy": 0.8244661808013916,
      "num_tokens": 281115485.0,
      "step": 8390
    },
    {
      "epoch": 0.5011341929321872,
      "grad_norm": 0.5526153445243835,
      "learning_rate": 4.523194335600507e-05,
      "loss": 0.5678,
      "mean_token_accuracy": 0.8207622408866883,
      "num_tokens": 281283165.0,
      "step": 8395
    },
    {
      "epoch": 0.501432664756447,
      "grad_norm": 0.5357623100280762,
      "learning_rate": 4.522510575270715e-05,
      "loss": 0.5762,
      "mean_token_accuracy": 0.8183287620544434,
      "num_tokens": 281450845.0,
      "step": 8400
    },
    {
      "epoch": 0.5017311365807068,
      "grad_norm": 0.4978879988193512,
      "learning_rate": 4.521826383213256e-05,
      "loss": 0.5427,
      "mean_token_accuracy": 0.8289156675338745,
      "num_tokens": 281618525.0,
      "step": 8405
    },
    {
      "epoch": 0.5020296084049666,
      "grad_norm": 0.5199429392814636,
      "learning_rate": 4.521141759594776e-05,
      "loss": 0.5524,
      "mean_token_accuracy": 0.8254741668701172,
      "num_tokens": 281786205.0,
      "step": 8410
    },
    {
      "epoch": 0.5023280802292264,
      "grad_norm": 0.559115469455719,
      "learning_rate": 4.520456704582031e-05,
      "loss": 0.5762,
      "mean_token_accuracy": 0.8197184920310974,
      "num_tokens": 281953885.0,
      "step": 8415
    },
    {
      "epoch": 0.5026265520534862,
      "grad_norm": 0.59633868932724,
      "learning_rate": 4.519771218341876e-05,
      "loss": 0.6122,
      "mean_token_accuracy": 0.8071632981300354,
      "num_tokens": 282121565.0,
      "step": 8420
    },
    {
      "epoch": 0.502925023877746,
      "grad_norm": 0.6832320690155029,
      "learning_rate": 4.5190853010412745e-05,
      "loss": 0.5991,
      "mean_token_accuracy": 0.8122509837150573,
      "num_tokens": 282289245.0,
      "step": 8425
    },
    {
      "epoch": 0.5032234957020058,
      "grad_norm": 0.5692151784896851,
      "learning_rate": 4.518398952847294e-05,
      "loss": 0.5938,
      "mean_token_accuracy": 0.8133544206619263,
      "num_tokens": 282456925.0,
      "step": 8430
    },
    {
      "epoch": 0.5035219675262655,
      "grad_norm": 0.5501351952552795,
      "learning_rate": 4.5177121739271073e-05,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8299713730812073,
      "num_tokens": 282624605.0,
      "step": 8435
    },
    {
      "epoch": 0.5038204393505253,
      "grad_norm": 0.557152509689331,
      "learning_rate": 4.5170249644479937e-05,
      "loss": 0.5753,
      "mean_token_accuracy": 0.8187999606132508,
      "num_tokens": 282792285.0,
      "step": 8440
    },
    {
      "epoch": 0.5041189111747851,
      "grad_norm": 0.6025996208190918,
      "learning_rate": 4.516337324577333e-05,
      "loss": 0.6139,
      "mean_token_accuracy": 0.80866037607193,
      "num_tokens": 282959965.0,
      "step": 8445
    },
    {
      "epoch": 0.5044173829990449,
      "grad_norm": 0.5659865140914917,
      "learning_rate": 4.5156492544826146e-05,
      "loss": 0.5736,
      "mean_token_accuracy": 0.8184062838554382,
      "num_tokens": 283127645.0,
      "step": 8450
    },
    {
      "epoch": 0.5047158548233047,
      "grad_norm": 0.5491483211517334,
      "learning_rate": 4.5149607543314306e-05,
      "loss": 0.541,
      "mean_token_accuracy": 0.8302159190177918,
      "num_tokens": 283295325.0,
      "step": 8455
    },
    {
      "epoch": 0.5050143266475645,
      "grad_norm": 0.6142468452453613,
      "learning_rate": 4.514271824291477e-05,
      "loss": 0.579,
      "mean_token_accuracy": 0.8175414562225342,
      "num_tokens": 283463005.0,
      "step": 8460
    },
    {
      "epoch": 0.5053127984718243,
      "grad_norm": 0.6395441889762878,
      "learning_rate": 4.513582464530556e-05,
      "loss": 0.5975,
      "mean_token_accuracy": 0.814022445678711,
      "num_tokens": 283630685.0,
      "step": 8465
    },
    {
      "epoch": 0.5056112702960841,
      "grad_norm": 0.5647761821746826,
      "learning_rate": 4.512892675216574e-05,
      "loss": 0.5598,
      "mean_token_accuracy": 0.8255397796630859,
      "num_tokens": 283798365.0,
      "step": 8470
    },
    {
      "epoch": 0.5059097421203438,
      "grad_norm": 0.5564920902252197,
      "learning_rate": 4.5122024565175414e-05,
      "loss": 0.5952,
      "mean_token_accuracy": 0.8134021162986755,
      "num_tokens": 283966045.0,
      "step": 8475
    },
    {
      "epoch": 0.5062082139446036,
      "grad_norm": 0.5405991673469543,
      "learning_rate": 4.511511808601574e-05,
      "loss": 0.5433,
      "mean_token_accuracy": 0.8270010828971863,
      "num_tokens": 284133725.0,
      "step": 8480
    },
    {
      "epoch": 0.5065066857688634,
      "grad_norm": 0.546653151512146,
      "learning_rate": 4.5108207316368915e-05,
      "loss": 0.5887,
      "mean_token_accuracy": 0.8135870218276977,
      "num_tokens": 284301405.0,
      "step": 8485
    },
    {
      "epoch": 0.5068051575931232,
      "grad_norm": 0.6025099754333496,
      "learning_rate": 4.5101292257918184e-05,
      "loss": 0.5453,
      "mean_token_accuracy": 0.8282893896102905,
      "num_tokens": 284469085.0,
      "step": 8490
    },
    {
      "epoch": 0.507103629417383,
      "grad_norm": 0.5866109132766724,
      "learning_rate": 4.509437291234786e-05,
      "loss": 0.6112,
      "mean_token_accuracy": 0.8091912150382996,
      "num_tokens": 284636765.0,
      "step": 8495
    },
    {
      "epoch": 0.5074021012416428,
      "grad_norm": 0.5114835500717163,
      "learning_rate": 4.508744928134324e-05,
      "loss": 0.5474,
      "mean_token_accuracy": 0.8278122425079346,
      "num_tokens": 284804445.0,
      "step": 8500
    },
    {
      "epoch": 0.5077005730659025,
      "grad_norm": 0.54168701171875,
      "learning_rate": 4.5080521366590724e-05,
      "loss": 0.5361,
      "mean_token_accuracy": 0.833245849609375,
      "num_tokens": 284972125.0,
      "step": 8505
    },
    {
      "epoch": 0.5079990448901623,
      "grad_norm": 0.5267937183380127,
      "learning_rate": 4.507358916977774e-05,
      "loss": 0.5503,
      "mean_token_accuracy": 0.8275855898857116,
      "num_tokens": 285139805.0,
      "step": 8510
    },
    {
      "epoch": 0.5082975167144221,
      "grad_norm": 0.5776879787445068,
      "learning_rate": 4.5066652692592734e-05,
      "loss": 0.5735,
      "mean_token_accuracy": 0.8207443714141845,
      "num_tokens": 285307485.0,
      "step": 8515
    },
    {
      "epoch": 0.5085959885386819,
      "grad_norm": 0.6104937791824341,
      "learning_rate": 4.505971193672522e-05,
      "loss": 0.6164,
      "mean_token_accuracy": 0.8079148411750794,
      "num_tokens": 285475165.0,
      "step": 8520
    },
    {
      "epoch": 0.5088944603629417,
      "grad_norm": 0.6085206866264343,
      "learning_rate": 4.505276690386575e-05,
      "loss": 0.559,
      "mean_token_accuracy": 0.824561607837677,
      "num_tokens": 285642845.0,
      "step": 8525
    },
    {
      "epoch": 0.5091929321872015,
      "grad_norm": 0.5954418182373047,
      "learning_rate": 4.504581759570591e-05,
      "loss": 0.5224,
      "mean_token_accuracy": 0.834122633934021,
      "num_tokens": 285810525.0,
      "step": 8530
    },
    {
      "epoch": 0.5094914040114613,
      "grad_norm": 0.5252765417098999,
      "learning_rate": 4.503886401393834e-05,
      "loss": 0.5973,
      "mean_token_accuracy": 0.8123881697654725,
      "num_tokens": 285978205.0,
      "step": 8535
    },
    {
      "epoch": 0.5097898758357211,
      "grad_norm": 0.6340962052345276,
      "learning_rate": 4.503190616025672e-05,
      "loss": 0.5953,
      "mean_token_accuracy": 0.8133245825767517,
      "num_tokens": 286145885.0,
      "step": 8540
    },
    {
      "epoch": 0.5100883476599809,
      "grad_norm": 0.5511323809623718,
      "learning_rate": 4.5024944036355735e-05,
      "loss": 0.5703,
      "mean_token_accuracy": 0.8201662182807923,
      "num_tokens": 286305497.0,
      "step": 8545
    },
    {
      "epoch": 0.5103868194842407,
      "grad_norm": 0.6255331635475159,
      "learning_rate": 4.501797764393116e-05,
      "loss": 0.601,
      "mean_token_accuracy": 0.8114756107330322,
      "num_tokens": 286473177.0,
      "step": 8550
    },
    {
      "epoch": 0.5106852913085005,
      "grad_norm": 0.5895081758499146,
      "learning_rate": 4.50110069846798e-05,
      "loss": 0.5892,
      "mean_token_accuracy": 0.8159310460090637,
      "num_tokens": 286640857.0,
      "step": 8555
    },
    {
      "epoch": 0.5109837631327603,
      "grad_norm": 0.555617094039917,
      "learning_rate": 4.5004032060299455e-05,
      "loss": 0.5521,
      "mean_token_accuracy": 0.8255636334419251,
      "num_tokens": 286808537.0,
      "step": 8560
    },
    {
      "epoch": 0.5112822349570201,
      "grad_norm": 0.5435903072357178,
      "learning_rate": 4.4997052872489023e-05,
      "loss": 0.5364,
      "mean_token_accuracy": 0.8305618405342102,
      "num_tokens": 286976217.0,
      "step": 8565
    },
    {
      "epoch": 0.5115807067812799,
      "grad_norm": 0.6119329333305359,
      "learning_rate": 4.49900694229484e-05,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8312954783439637,
      "num_tokens": 287143897.0,
      "step": 8570
    },
    {
      "epoch": 0.5118791786055397,
      "grad_norm": 0.5741212368011475,
      "learning_rate": 4.498308171337854e-05,
      "loss": 0.5755,
      "mean_token_accuracy": 0.8186866164207458,
      "num_tokens": 287311577.0,
      "step": 8575
    },
    {
      "epoch": 0.5121776504297995,
      "grad_norm": 0.5080263614654541,
      "learning_rate": 4.4976089745481437e-05,
      "loss": 0.5224,
      "mean_token_accuracy": 0.8341643571853637,
      "num_tokens": 287479257.0,
      "step": 8580
    },
    {
      "epoch": 0.5124761222540593,
      "grad_norm": 0.6037417054176331,
      "learning_rate": 4.4969093520960096e-05,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8187522411346435,
      "num_tokens": 287646937.0,
      "step": 8585
    },
    {
      "epoch": 0.512774594078319,
      "grad_norm": 0.5993456840515137,
      "learning_rate": 4.496209304151859e-05,
      "loss": 0.5495,
      "mean_token_accuracy": 0.8264583110809326,
      "num_tokens": 287814617.0,
      "step": 8590
    },
    {
      "epoch": 0.5130730659025788,
      "grad_norm": 0.6401481628417969,
      "learning_rate": 4.495508830886201e-05,
      "loss": 0.5836,
      "mean_token_accuracy": 0.8159489393234253,
      "num_tokens": 287982297.0,
      "step": 8595
    },
    {
      "epoch": 0.5133715377268386,
      "grad_norm": 0.5762320756912231,
      "learning_rate": 4.494807932469649e-05,
      "loss": 0.59,
      "mean_token_accuracy": 0.813421881198883,
      "num_tokens": 288141450.0,
      "step": 8600
    },
    {
      "epoch": 0.5136700095510984,
      "grad_norm": 0.5389105677604675,
      "learning_rate": 4.494106609072919e-05,
      "loss": 0.5402,
      "mean_token_accuracy": 0.829046881198883,
      "num_tokens": 288309130.0,
      "step": 8605
    },
    {
      "epoch": 0.5139684813753582,
      "grad_norm": 0.5098772644996643,
      "learning_rate": 4.493404860866832e-05,
      "loss": 0.5181,
      "mean_token_accuracy": 0.8354884862899781,
      "num_tokens": 288476810.0,
      "step": 8610
    },
    {
      "epoch": 0.5142669531996179,
      "grad_norm": 0.5440981984138489,
      "learning_rate": 4.4927026880223117e-05,
      "loss": 0.5643,
      "mean_token_accuracy": 0.8219611048698425,
      "num_tokens": 288644490.0,
      "step": 8615
    },
    {
      "epoch": 0.5145654250238777,
      "grad_norm": 0.5773564577102661,
      "learning_rate": 4.492000090710385e-05,
      "loss": 0.5593,
      "mean_token_accuracy": 0.8247703671455383,
      "num_tokens": 288812170.0,
      "step": 8620
    },
    {
      "epoch": 0.5148638968481375,
      "grad_norm": 0.6357622742652893,
      "learning_rate": 4.4912970691021824e-05,
      "loss": 0.5455,
      "mean_token_accuracy": 0.8289633870124817,
      "num_tokens": 288979850.0,
      "step": 8625
    },
    {
      "epoch": 0.5151623686723973,
      "grad_norm": 0.6056704521179199,
      "learning_rate": 4.490593623368938e-05,
      "loss": 0.5649,
      "mean_token_accuracy": 0.8217642784118653,
      "num_tokens": 289147530.0,
      "step": 8630
    },
    {
      "epoch": 0.5154608404966571,
      "grad_norm": 0.5196501612663269,
      "learning_rate": 4.489889753681989e-05,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8234880089759826,
      "num_tokens": 289315210.0,
      "step": 8635
    },
    {
      "epoch": 0.5157593123209169,
      "grad_norm": 0.5886099338531494,
      "learning_rate": 4.489185460212777e-05,
      "loss": 0.5608,
      "mean_token_accuracy": 0.8236609697341919,
      "num_tokens": 289482890.0,
      "step": 8640
    },
    {
      "epoch": 0.5160577841451767,
      "grad_norm": 0.5845518708229065,
      "learning_rate": 4.488480743132843e-05,
      "loss": 0.5837,
      "mean_token_accuracy": 0.8166706323623657,
      "num_tokens": 289650570.0,
      "step": 8645
    },
    {
      "epoch": 0.5163562559694365,
      "grad_norm": 0.5307056903839111,
      "learning_rate": 4.487775602613836e-05,
      "loss": 0.5353,
      "mean_token_accuracy": 0.8278703212738037,
      "num_tokens": 289815987.0,
      "step": 8650
    },
    {
      "epoch": 0.5166547277936963,
      "grad_norm": 0.5496380925178528,
      "learning_rate": 4.487070038827505e-05,
      "loss": 0.5607,
      "mean_token_accuracy": 0.8231361150741577,
      "num_tokens": 289983667.0,
      "step": 8655
    },
    {
      "epoch": 0.516953199617956,
      "grad_norm": 0.5331660509109497,
      "learning_rate": 4.4863640519457034e-05,
      "loss": 0.552,
      "mean_token_accuracy": 0.8254204988479614,
      "num_tokens": 290151347.0,
      "step": 8660
    },
    {
      "epoch": 0.5172516714422158,
      "grad_norm": 0.5752771496772766,
      "learning_rate": 4.485657642140387e-05,
      "loss": 0.5264,
      "mean_token_accuracy": 0.8333114743232727,
      "num_tokens": 290319027.0,
      "step": 8665
    },
    {
      "epoch": 0.5175501432664756,
      "grad_norm": 0.47481390833854675,
      "learning_rate": 4.484950809583616e-05,
      "loss": 0.509,
      "mean_token_accuracy": 0.8388835072517395,
      "num_tokens": 290482019.0,
      "step": 8670
    },
    {
      "epoch": 0.5178486150907354,
      "grad_norm": 0.5849865674972534,
      "learning_rate": 4.484243554447552e-05,
      "loss": 0.6146,
      "mean_token_accuracy": 0.8089228272438049,
      "num_tokens": 290649699.0,
      "step": 8675
    },
    {
      "epoch": 0.5181470869149952,
      "grad_norm": 0.6237294673919678,
      "learning_rate": 4.483535876904459e-05,
      "loss": 0.5992,
      "mean_token_accuracy": 0.8107419848442078,
      "num_tokens": 290817379.0,
      "step": 8680
    },
    {
      "epoch": 0.518445558739255,
      "grad_norm": 0.554248034954071,
      "learning_rate": 4.482827777126706e-05,
      "loss": 0.5654,
      "mean_token_accuracy": 0.8227603554725647,
      "num_tokens": 290985059.0,
      "step": 8685
    },
    {
      "epoch": 0.5187440305635148,
      "grad_norm": 0.6273401975631714,
      "learning_rate": 4.4821192552867636e-05,
      "loss": 0.5713,
      "mean_token_accuracy": 0.8194798946380615,
      "num_tokens": 291152739.0,
      "step": 8690
    },
    {
      "epoch": 0.5190425023877746,
      "grad_norm": 0.563341498374939,
      "learning_rate": 4.481410311557206e-05,
      "loss": 0.5913,
      "mean_token_accuracy": 0.814034354686737,
      "num_tokens": 291320419.0,
      "step": 8695
    },
    {
      "epoch": 0.5193409742120344,
      "grad_norm": 0.5764652490615845,
      "learning_rate": 4.480700946110708e-05,
      "loss": 0.5654,
      "mean_token_accuracy": 0.8213109850883484,
      "num_tokens": 291488099.0,
      "step": 8700
    },
    {
      "epoch": 0.5196394460362942,
      "grad_norm": 0.5594584345817566,
      "learning_rate": 4.47999115912005e-05,
      "loss": 0.5275,
      "mean_token_accuracy": 0.831641411781311,
      "num_tokens": 291655779.0,
      "step": 8705
    },
    {
      "epoch": 0.519937917860554,
      "grad_norm": 0.5473382472991943,
      "learning_rate": 4.4792809507581126e-05,
      "loss": 0.5885,
      "mean_token_accuracy": 0.8138613820075988,
      "num_tokens": 291823459.0,
      "step": 8710
    },
    {
      "epoch": 0.5202363896848138,
      "grad_norm": 0.5849176049232483,
      "learning_rate": 4.478570321197881e-05,
      "loss": 0.5742,
      "mean_token_accuracy": 0.8186761736869812,
      "num_tokens": 291986614.0,
      "step": 8715
    },
    {
      "epoch": 0.5205348615090736,
      "grad_norm": 0.4745195209980011,
      "learning_rate": 4.4778592706124426e-05,
      "loss": 0.5296,
      "mean_token_accuracy": 0.8316831707954406,
      "num_tokens": 292154294.0,
      "step": 8720
    },
    {
      "epoch": 0.5208333333333334,
      "grad_norm": 0.5845153331756592,
      "learning_rate": 4.477147799174986e-05,
      "loss": 0.6311,
      "mean_token_accuracy": 0.8032446503639221,
      "num_tokens": 292321974.0,
      "step": 8725
    },
    {
      "epoch": 0.5211318051575932,
      "grad_norm": 0.5913403034210205,
      "learning_rate": 4.476435907058802e-05,
      "loss": 0.6044,
      "mean_token_accuracy": 0.8127877831459045,
      "num_tokens": 292489654.0,
      "step": 8730
    },
    {
      "epoch": 0.521430276981853,
      "grad_norm": 0.5641769170761108,
      "learning_rate": 4.475723594437289e-05,
      "loss": 0.5772,
      "mean_token_accuracy": 0.8175951361656189,
      "num_tokens": 292657334.0,
      "step": 8735
    },
    {
      "epoch": 0.5217287488061128,
      "grad_norm": 0.5563650727272034,
      "learning_rate": 4.475010861483939e-05,
      "loss": 0.6175,
      "mean_token_accuracy": 0.8094894409179687,
      "num_tokens": 292825014.0,
      "step": 8740
    },
    {
      "epoch": 0.5220272206303725,
      "grad_norm": 0.5406704545021057,
      "learning_rate": 4.474297708372355e-05,
      "loss": 0.579,
      "mean_token_accuracy": 0.8181498289108277,
      "num_tokens": 292992694.0,
      "step": 8745
    },
    {
      "epoch": 0.5223256924546322,
      "grad_norm": 0.6333975195884705,
      "learning_rate": 4.4735841352762367e-05,
      "loss": 0.6105,
      "mean_token_accuracy": 0.8097817182540894,
      "num_tokens": 293160374.0,
      "step": 8750
    },
    {
      "epoch": 0.522624164278892,
      "grad_norm": 0.6653600931167603,
      "learning_rate": 4.472870142369388e-05,
      "loss": 0.5534,
      "mean_token_accuracy": 0.8252177119255066,
      "num_tokens": 293328054.0,
      "step": 8755
    },
    {
      "epoch": 0.5229226361031518,
      "grad_norm": 0.5845597386360168,
      "learning_rate": 4.472155729825717e-05,
      "loss": 0.5636,
      "mean_token_accuracy": 0.8220207691192627,
      "num_tokens": 293495734.0,
      "step": 8760
    },
    {
      "epoch": 0.5232211079274116,
      "grad_norm": 0.459767609834671,
      "learning_rate": 4.4714408978192294e-05,
      "loss": 0.5244,
      "mean_token_accuracy": 0.8326732754707337,
      "num_tokens": 293663414.0,
      "step": 8765
    },
    {
      "epoch": 0.5235195797516714,
      "grad_norm": 0.521870493888855,
      "learning_rate": 4.470725646524038e-05,
      "loss": 0.5627,
      "mean_token_accuracy": 0.8234760761260986,
      "num_tokens": 293831094.0,
      "step": 8770
    },
    {
      "epoch": 0.5238180515759312,
      "grad_norm": 0.5080561637878418,
      "learning_rate": 4.4700099761143535e-05,
      "loss": 0.5466,
      "mean_token_accuracy": 0.8269712448120117,
      "num_tokens": 293998774.0,
      "step": 8775
    },
    {
      "epoch": 0.524116523400191,
      "grad_norm": 0.5342180132865906,
      "learning_rate": 4.469293886764493e-05,
      "loss": 0.5558,
      "mean_token_accuracy": 0.8248896598815918,
      "num_tokens": 294166454.0,
      "step": 8780
    },
    {
      "epoch": 0.5244149952244508,
      "grad_norm": 0.599281907081604,
      "learning_rate": 4.4685773786488704e-05,
      "loss": 0.5696,
      "mean_token_accuracy": 0.8206191062927246,
      "num_tokens": 294334134.0,
      "step": 8785
    },
    {
      "epoch": 0.5247134670487106,
      "grad_norm": 0.513864278793335,
      "learning_rate": 4.467860451942006e-05,
      "loss": 0.5804,
      "mean_token_accuracy": 0.81725515127182,
      "num_tokens": 294501814.0,
      "step": 8790
    },
    {
      "epoch": 0.5250119388729704,
      "grad_norm": 0.4977003335952759,
      "learning_rate": 4.467143106818521e-05,
      "loss": 0.5485,
      "mean_token_accuracy": 0.8274842023849487,
      "num_tokens": 294669494.0,
      "step": 8795
    },
    {
      "epoch": 0.5253104106972302,
      "grad_norm": 0.6562310457229614,
      "learning_rate": 4.4664253434531375e-05,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8244960069656372,
      "num_tokens": 294837174.0,
      "step": 8800
    },
    {
      "epoch": 0.52560888252149,
      "grad_norm": 0.53802090883255,
      "learning_rate": 4.46570716202068e-05,
      "loss": 0.5485,
      "mean_token_accuracy": 0.8254324197769165,
      "num_tokens": 295004854.0,
      "step": 8805
    },
    {
      "epoch": 0.5259073543457498,
      "grad_norm": 0.5729022026062012,
      "learning_rate": 4.464988562696075e-05,
      "loss": 0.5364,
      "mean_token_accuracy": 0.8288023352622986,
      "num_tokens": 295172534.0,
      "step": 8810
    },
    {
      "epoch": 0.5262058261700095,
      "grad_norm": 0.5360018014907837,
      "learning_rate": 4.46426954565435e-05,
      "loss": 0.5878,
      "mean_token_accuracy": 0.8148276329040527,
      "num_tokens": 295340214.0,
      "step": 8815
    },
    {
      "epoch": 0.5265042979942693,
      "grad_norm": 0.5254443287849426,
      "learning_rate": 4.463550111070635e-05,
      "loss": 0.5877,
      "mean_token_accuracy": 0.8135631561279297,
      "num_tokens": 295507894.0,
      "step": 8820
    },
    {
      "epoch": 0.5268027698185291,
      "grad_norm": 0.5107590556144714,
      "learning_rate": 4.462830259120163e-05,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8352976202964782,
      "num_tokens": 295675574.0,
      "step": 8825
    },
    {
      "epoch": 0.5271012416427889,
      "grad_norm": 0.5577269792556763,
      "learning_rate": 4.462109989978265e-05,
      "loss": 0.5534,
      "mean_token_accuracy": 0.8250566601753235,
      "num_tokens": 295843254.0,
      "step": 8830
    },
    {
      "epoch": 0.5273997134670487,
      "grad_norm": 0.603557288646698,
      "learning_rate": 4.461389303820378e-05,
      "loss": 0.5516,
      "mean_token_accuracy": 0.8260050177574157,
      "num_tokens": 296010934.0,
      "step": 8835
    },
    {
      "epoch": 0.5276981852913085,
      "grad_norm": 0.5219425559043884,
      "learning_rate": 4.460668200822037e-05,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8235178351402282,
      "num_tokens": 296178614.0,
      "step": 8840
    },
    {
      "epoch": 0.5279966571155683,
      "grad_norm": 0.5341835021972656,
      "learning_rate": 4.4599466811588805e-05,
      "loss": 0.5579,
      "mean_token_accuracy": 0.8254085659980774,
      "num_tokens": 296346294.0,
      "step": 8845
    },
    {
      "epoch": 0.5282951289398281,
      "grad_norm": 0.5543561577796936,
      "learning_rate": 4.459224745006648e-05,
      "loss": 0.5876,
      "mean_token_accuracy": 0.8151079535484314,
      "num_tokens": 296513974.0,
      "step": 8850
    },
    {
      "epoch": 0.5285936007640879,
      "grad_norm": 0.5444608926773071,
      "learning_rate": 4.45850239254118e-05,
      "loss": 0.5706,
      "mean_token_accuracy": 0.819837772846222,
      "num_tokens": 296681654.0,
      "step": 8855
    },
    {
      "epoch": 0.5288920725883477,
      "grad_norm": 0.5355438590049744,
      "learning_rate": 4.45777962393842e-05,
      "loss": 0.5358,
      "mean_token_accuracy": 0.830788505077362,
      "num_tokens": 296849334.0,
      "step": 8860
    },
    {
      "epoch": 0.5291905444126075,
      "grad_norm": 0.5682732462882996,
      "learning_rate": 4.457056439374408e-05,
      "loss": 0.5848,
      "mean_token_accuracy": 0.8136884212493897,
      "num_tokens": 297017014.0,
      "step": 8865
    },
    {
      "epoch": 0.5294890162368673,
      "grad_norm": 0.5435397624969482,
      "learning_rate": 4.456332839025293e-05,
      "loss": 0.5794,
      "mean_token_accuracy": 0.817117965221405,
      "num_tokens": 297184694.0,
      "step": 8870
    },
    {
      "epoch": 0.5297874880611271,
      "grad_norm": 0.5456133484840393,
      "learning_rate": 4.4556088230673207e-05,
      "loss": 0.5567,
      "mean_token_accuracy": 0.82181795835495,
      "num_tokens": 297352374.0,
      "step": 8875
    },
    {
      "epoch": 0.5300859598853869,
      "grad_norm": 0.5780700445175171,
      "learning_rate": 4.4548843916768364e-05,
      "loss": 0.6033,
      "mean_token_accuracy": 0.8101037859916687,
      "num_tokens": 297520054.0,
      "step": 8880
    },
    {
      "epoch": 0.5303844317096467,
      "grad_norm": 0.5610108971595764,
      "learning_rate": 4.45415954503029e-05,
      "loss": 0.5674,
      "mean_token_accuracy": 0.8212632656097412,
      "num_tokens": 297687734.0,
      "step": 8885
    },
    {
      "epoch": 0.5306829035339063,
      "grad_norm": 0.5550732612609863,
      "learning_rate": 4.453434283304232e-05,
      "loss": 0.6066,
      "mean_token_accuracy": 0.810050094127655,
      "num_tokens": 297855414.0,
      "step": 8890
    },
    {
      "epoch": 0.5309813753581661,
      "grad_norm": 0.5507266521453857,
      "learning_rate": 4.452708606675311e-05,
      "loss": 0.5778,
      "mean_token_accuracy": 0.8170404314994812,
      "num_tokens": 298023094.0,
      "step": 8895
    },
    {
      "epoch": 0.5312798471824259,
      "grad_norm": 0.5527105331420898,
      "learning_rate": 4.4519825153202824e-05,
      "loss": 0.6087,
      "mean_token_accuracy": 0.8090003609657288,
      "num_tokens": 298190774.0,
      "step": 8900
    },
    {
      "epoch": 0.5315783190066857,
      "grad_norm": 0.5168611407279968,
      "learning_rate": 4.451256009415996e-05,
      "loss": 0.5489,
      "mean_token_accuracy": 0.8256888866424561,
      "num_tokens": 298358454.0,
      "step": 8905
    },
    {
      "epoch": 0.5318767908309455,
      "grad_norm": 0.5142448544502258,
      "learning_rate": 4.450529089139408e-05,
      "loss": 0.572,
      "mean_token_accuracy": 0.8191995859146118,
      "num_tokens": 298526134.0,
      "step": 8910
    },
    {
      "epoch": 0.5321752626552053,
      "grad_norm": 0.5494459271430969,
      "learning_rate": 4.4498017546675714e-05,
      "loss": 0.5446,
      "mean_token_accuracy": 0.8277227640151977,
      "num_tokens": 298693814.0,
      "step": 8915
    },
    {
      "epoch": 0.5324737344794651,
      "grad_norm": 0.6558916568756104,
      "learning_rate": 4.449074006177642e-05,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8241262197494507,
      "num_tokens": 298861494.0,
      "step": 8920
    },
    {
      "epoch": 0.5327722063037249,
      "grad_norm": 0.642120897769928,
      "learning_rate": 4.448345843846878e-05,
      "loss": 0.6123,
      "mean_token_accuracy": 0.8097459197044372,
      "num_tokens": 299029174.0,
      "step": 8925
    },
    {
      "epoch": 0.5330706781279847,
      "grad_norm": 0.5801108479499817,
      "learning_rate": 4.447617267852635e-05,
      "loss": 0.5795,
      "mean_token_accuracy": 0.8184778809547424,
      "num_tokens": 299196854.0,
      "step": 8930
    },
    {
      "epoch": 0.5333691499522445,
      "grad_norm": 0.5552683472633362,
      "learning_rate": 4.446888278372373e-05,
      "loss": 0.5623,
      "mean_token_accuracy": 0.8219969034194946,
      "num_tokens": 299364534.0,
      "step": 8935
    },
    {
      "epoch": 0.5336676217765043,
      "grad_norm": 0.5098156929016113,
      "learning_rate": 4.446158875583649e-05,
      "loss": 0.5423,
      "mean_token_accuracy": 0.8268102049827576,
      "num_tokens": 299532214.0,
      "step": 8940
    },
    {
      "epoch": 0.5339660936007641,
      "grad_norm": 0.6488260626792908,
      "learning_rate": 4.445429059664124e-05,
      "loss": 0.6259,
      "mean_token_accuracy": 0.8039365291595459,
      "num_tokens": 299699894.0,
      "step": 8945
    },
    {
      "epoch": 0.5342645654250239,
      "grad_norm": 0.531151294708252,
      "learning_rate": 4.444698830791557e-05,
      "loss": 0.5675,
      "mean_token_accuracy": 0.8217225313186646,
      "num_tokens": 299867574.0,
      "step": 8950
    },
    {
      "epoch": 0.5345630372492837,
      "grad_norm": 0.5506296157836914,
      "learning_rate": 4.4439681891438096e-05,
      "loss": 0.5476,
      "mean_token_accuracy": 0.8268996834754944,
      "num_tokens": 300035254.0,
      "step": 8955
    },
    {
      "epoch": 0.5348615090735435,
      "grad_norm": 0.5589233040809631,
      "learning_rate": 4.443237134898843e-05,
      "loss": 0.5473,
      "mean_token_accuracy": 0.8274662971496582,
      "num_tokens": 300202934.0,
      "step": 8960
    },
    {
      "epoch": 0.5351599808978033,
      "grad_norm": 0.5993244647979736,
      "learning_rate": 4.442505668234718e-05,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8296969890594482,
      "num_tokens": 300370614.0,
      "step": 8965
    },
    {
      "epoch": 0.535458452722063,
      "grad_norm": 0.5460119843482971,
      "learning_rate": 4.441773789329598e-05,
      "loss": 0.6338,
      "mean_token_accuracy": 0.8036979675292969,
      "num_tokens": 300538294.0,
      "step": 8970
    },
    {
      "epoch": 0.5357569245463228,
      "grad_norm": 0.6032574772834778,
      "learning_rate": 4.4410414983617446e-05,
      "loss": 0.5501,
      "mean_token_accuracy": 0.8245258331298828,
      "num_tokens": 300705974.0,
      "step": 8975
    },
    {
      "epoch": 0.5360553963705826,
      "grad_norm": 0.6453452110290527,
      "learning_rate": 4.44030879550952e-05,
      "loss": 0.5489,
      "mean_token_accuracy": 0.82506263256073,
      "num_tokens": 300873654.0,
      "step": 8980
    },
    {
      "epoch": 0.5363538681948424,
      "grad_norm": 0.5219424366950989,
      "learning_rate": 4.4395756809513906e-05,
      "loss": 0.5633,
      "mean_token_accuracy": 0.8222533583641052,
      "num_tokens": 301041334.0,
      "step": 8985
    },
    {
      "epoch": 0.5366523400191022,
      "grad_norm": 0.5612058043479919,
      "learning_rate": 4.438842154865917e-05,
      "loss": 0.6,
      "mean_token_accuracy": 0.8105272650718689,
      "num_tokens": 301209014.0,
      "step": 8990
    },
    {
      "epoch": 0.536950811843362,
      "grad_norm": 0.6003037095069885,
      "learning_rate": 4.438108217431765e-05,
      "loss": 0.621,
      "mean_token_accuracy": 0.8066145658493042,
      "num_tokens": 301376694.0,
      "step": 8995
    },
    {
      "epoch": 0.5372492836676218,
      "grad_norm": 0.5480248332023621,
      "learning_rate": 4.437373868827698e-05,
      "loss": 0.5523,
      "mean_token_accuracy": 0.8252296328544617,
      "num_tokens": 301544374.0,
      "step": 9000
    },
    {
      "epoch": 0.5375477554918816,
      "grad_norm": 0.5819894671440125,
      "learning_rate": 4.436639109232579e-05,
      "loss": 0.5816,
      "mean_token_accuracy": 0.8197363615036011,
      "num_tokens": 301712054.0,
      "step": 9005
    },
    {
      "epoch": 0.5378462273161414,
      "grad_norm": 0.5392967462539673,
      "learning_rate": 4.4359039388253726e-05,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8287963747978211,
      "num_tokens": 301879734.0,
      "step": 9010
    },
    {
      "epoch": 0.5381446991404012,
      "grad_norm": 0.5332432985305786,
      "learning_rate": 4.435168357785145e-05,
      "loss": 0.5237,
      "mean_token_accuracy": 0.8327225804328918,
      "num_tokens": 302046313.0,
      "step": 9015
    },
    {
      "epoch": 0.538443170964661,
      "grad_norm": 0.5505347847938538,
      "learning_rate": 4.434432366291058e-05,
      "loss": 0.5227,
      "mean_token_accuracy": 0.833860182762146,
      "num_tokens": 302213993.0,
      "step": 9020
    },
    {
      "epoch": 0.5387416427889207,
      "grad_norm": 0.5260748267173767,
      "learning_rate": 4.433695964522378e-05,
      "loss": 0.5976,
      "mean_token_accuracy": 0.8123762369155884,
      "num_tokens": 302381673.0,
      "step": 9025
    },
    {
      "epoch": 0.5390401146131805,
      "grad_norm": 0.5036420226097107,
      "learning_rate": 4.432959152658469e-05,
      "loss": 0.5824,
      "mean_token_accuracy": 0.8154956459999084,
      "num_tokens": 302549353.0,
      "step": 9030
    },
    {
      "epoch": 0.5393385864374403,
      "grad_norm": 0.5186771750450134,
      "learning_rate": 4.432221930878793e-05,
      "loss": 0.5687,
      "mean_token_accuracy": 0.820237398147583,
      "num_tokens": 302717033.0,
      "step": 9035
    },
    {
      "epoch": 0.5396370582617,
      "grad_norm": 0.5144567489624023,
      "learning_rate": 4.431484299362916e-05,
      "loss": 0.5814,
      "mean_token_accuracy": 0.8172253370285034,
      "num_tokens": 302884713.0,
      "step": 9040
    },
    {
      "epoch": 0.5399355300859598,
      "grad_norm": 0.5339194536209106,
      "learning_rate": 4.430746258290502e-05,
      "loss": 0.5156,
      "mean_token_accuracy": 0.835965633392334,
      "num_tokens": 303052393.0,
      "step": 9045
    },
    {
      "epoch": 0.5402340019102196,
      "grad_norm": 0.560391366481781,
      "learning_rate": 4.4300078078413126e-05,
      "loss": 0.5771,
      "mean_token_accuracy": 0.8182512283325195,
      "num_tokens": 303220073.0,
      "step": 9050
    },
    {
      "epoch": 0.5405324737344794,
      "grad_norm": 0.6061593890190125,
      "learning_rate": 4.429268948195213e-05,
      "loss": 0.5785,
      "mean_token_accuracy": 0.8181557774543762,
      "num_tokens": 303387753.0,
      "step": 9055
    },
    {
      "epoch": 0.5408309455587392,
      "grad_norm": 0.6637635231018066,
      "learning_rate": 4.428529679532165e-05,
      "loss": 0.5747,
      "mean_token_accuracy": 0.8197304129600524,
      "num_tokens": 303555433.0,
      "step": 9060
    },
    {
      "epoch": 0.541129417382999,
      "grad_norm": 0.5166680812835693,
      "learning_rate": 4.4277900020322325e-05,
      "loss": 0.5716,
      "mean_token_accuracy": 0.8197423338890075,
      "num_tokens": 303723113.0,
      "step": 9065
    },
    {
      "epoch": 0.5414278892072588,
      "grad_norm": 0.5843801498413086,
      "learning_rate": 4.427049915875575e-05,
      "loss": 0.5736,
      "mean_token_accuracy": 0.8185255885124206,
      "num_tokens": 303890793.0,
      "step": 9070
    },
    {
      "epoch": 0.5417263610315186,
      "grad_norm": 0.5218236446380615,
      "learning_rate": 4.4263094212424564e-05,
      "loss": 0.5723,
      "mean_token_accuracy": 0.8196767210960388,
      "num_tokens": 304058473.0,
      "step": 9075
    },
    {
      "epoch": 0.5420248328557784,
      "grad_norm": 0.525384783744812,
      "learning_rate": 4.425568518313237e-05,
      "loss": 0.5496,
      "mean_token_accuracy": 0.8261362314224243,
      "num_tokens": 304226153.0,
      "step": 9080
    },
    {
      "epoch": 0.5423233046800382,
      "grad_norm": 0.563209593296051,
      "learning_rate": 4.424827207268377e-05,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8216509580612182,
      "num_tokens": 304393833.0,
      "step": 9085
    },
    {
      "epoch": 0.542621776504298,
      "grad_norm": 0.576561450958252,
      "learning_rate": 4.424085488288438e-05,
      "loss": 0.5772,
      "mean_token_accuracy": 0.8167839646339417,
      "num_tokens": 304561513.0,
      "step": 9090
    },
    {
      "epoch": 0.5429202483285578,
      "grad_norm": 0.49649518728256226,
      "learning_rate": 4.423343361554077e-05,
      "loss": 0.5389,
      "mean_token_accuracy": 0.829410707950592,
      "num_tokens": 304729193.0,
      "step": 9095
    },
    {
      "epoch": 0.5432187201528176,
      "grad_norm": 0.6102119088172913,
      "learning_rate": 4.422600827246055e-05,
      "loss": 0.5425,
      "mean_token_accuracy": 0.8270070314407348,
      "num_tokens": 304896873.0,
      "step": 9100
    },
    {
      "epoch": 0.5435171919770774,
      "grad_norm": 0.5200836658477783,
      "learning_rate": 4.421857885545229e-05,
      "loss": 0.5408,
      "mean_token_accuracy": 0.8289395093917846,
      "num_tokens": 305064553.0,
      "step": 9105
    },
    {
      "epoch": 0.5438156638013372,
      "grad_norm": 0.5582632422447205,
      "learning_rate": 4.4211145366325545e-05,
      "loss": 0.5569,
      "mean_token_accuracy": 0.823833966255188,
      "num_tokens": 305232233.0,
      "step": 9110
    },
    {
      "epoch": 0.544114135625597,
      "grad_norm": 0.5659685730934143,
      "learning_rate": 4.42037078068909e-05,
      "loss": 0.5809,
      "mean_token_accuracy": 0.8184301614761352,
      "num_tokens": 305399913.0,
      "step": 9115
    },
    {
      "epoch": 0.5444126074498568,
      "grad_norm": 0.5390270948410034,
      "learning_rate": 4.419626617895991e-05,
      "loss": 0.5738,
      "mean_token_accuracy": 0.8193188667297363,
      "num_tokens": 305567593.0,
      "step": 9120
    },
    {
      "epoch": 0.5447110792741165,
      "grad_norm": 0.5471552014350891,
      "learning_rate": 4.4188820484345116e-05,
      "loss": 0.5273,
      "mean_token_accuracy": 0.8319873452186585,
      "num_tokens": 305735273.0,
      "step": 9125
    },
    {
      "epoch": 0.5450095510983763,
      "grad_norm": 0.5766529440879822,
      "learning_rate": 4.418137072486005e-05,
      "loss": 0.5263,
      "mean_token_accuracy": 0.83276869058609,
      "num_tokens": 305902953.0,
      "step": 9130
    },
    {
      "epoch": 0.5453080229226361,
      "grad_norm": 0.5826560258865356,
      "learning_rate": 4.417391690231924e-05,
      "loss": 0.5802,
      "mean_token_accuracy": 0.8160622596740723,
      "num_tokens": 306070633.0,
      "step": 9135
    },
    {
      "epoch": 0.5456064947468959,
      "grad_norm": 0.5199210047721863,
      "learning_rate": 4.4166459018538196e-05,
      "loss": 0.5091,
      "mean_token_accuracy": 0.8379756569862366,
      "num_tokens": 306238313.0,
      "step": 9140
    },
    {
      "epoch": 0.5459049665711557,
      "grad_norm": 0.5484384894371033,
      "learning_rate": 4.415899707533344e-05,
      "loss": 0.5642,
      "mean_token_accuracy": 0.8205177307128906,
      "num_tokens": 306405993.0,
      "step": 9145
    },
    {
      "epoch": 0.5462034383954155,
      "grad_norm": 0.551522970199585,
      "learning_rate": 4.415153107452245e-05,
      "loss": 0.5351,
      "mean_token_accuracy": 0.8285518169403077,
      "num_tokens": 306573673.0,
      "step": 9150
    },
    {
      "epoch": 0.5465019102196753,
      "grad_norm": 0.5362953543663025,
      "learning_rate": 4.414406101792372e-05,
      "loss": 0.5528,
      "mean_token_accuracy": 0.8251699924468994,
      "num_tokens": 306741353.0,
      "step": 9155
    },
    {
      "epoch": 0.5468003820439351,
      "grad_norm": 0.5488634705543518,
      "learning_rate": 4.413658690735671e-05,
      "loss": 0.544,
      "mean_token_accuracy": 0.8298699736595154,
      "num_tokens": 306909033.0,
      "step": 9160
    },
    {
      "epoch": 0.5470988538681948,
      "grad_norm": 0.5442475080490112,
      "learning_rate": 4.412910874464189e-05,
      "loss": 0.5801,
      "mean_token_accuracy": 0.8183048963546753,
      "num_tokens": 307076713.0,
      "step": 9165
    },
    {
      "epoch": 0.5473973256924546,
      "grad_norm": 0.6032072305679321,
      "learning_rate": 4.412162653160068e-05,
      "loss": 0.5838,
      "mean_token_accuracy": 0.8158356428146363,
      "num_tokens": 307244393.0,
      "step": 9170
    },
    {
      "epoch": 0.5476957975167144,
      "grad_norm": 0.5607478022575378,
      "learning_rate": 4.411414027005555e-05,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8199689745903015,
      "num_tokens": 307412073.0,
      "step": 9175
    },
    {
      "epoch": 0.5479942693409742,
      "grad_norm": 0.5132443904876709,
      "learning_rate": 4.410664996182989e-05,
      "loss": 0.554,
      "mean_token_accuracy": 0.8261720180511475,
      "num_tokens": 307579753.0,
      "step": 9180
    },
    {
      "epoch": 0.548292741165234,
      "grad_norm": 0.5602428913116455,
      "learning_rate": 4.40991556087481e-05,
      "loss": 0.5751,
      "mean_token_accuracy": 0.8194560289382935,
      "num_tokens": 307747433.0,
      "step": 9185
    },
    {
      "epoch": 0.5485912129894938,
      "grad_norm": 0.5555048584938049,
      "learning_rate": 4.4091657212635586e-05,
      "loss": 0.5654,
      "mean_token_accuracy": 0.8209352254867553,
      "num_tokens": 307915113.0,
      "step": 9190
    },
    {
      "epoch": 0.5488896848137536,
      "grad_norm": 0.611984133720398,
      "learning_rate": 4.40841547753187e-05,
      "loss": 0.5451,
      "mean_token_accuracy": 0.827669084072113,
      "num_tokens": 308082793.0,
      "step": 9195
    },
    {
      "epoch": 0.5491881566380133,
      "grad_norm": 0.4953276515007019,
      "learning_rate": 4.4076648298624815e-05,
      "loss": 0.569,
      "mean_token_accuracy": 0.8204163312911987,
      "num_tokens": 308250473.0,
      "step": 9200
    },
    {
      "epoch": 0.5494866284622731,
      "grad_norm": 0.5716685056686401,
      "learning_rate": 4.4069137784382266e-05,
      "loss": 0.5322,
      "mean_token_accuracy": 0.8322796106338501,
      "num_tokens": 308418153.0,
      "step": 9205
    },
    {
      "epoch": 0.5497851002865329,
      "grad_norm": 0.5615285038948059,
      "learning_rate": 4.406162323442038e-05,
      "loss": 0.5546,
      "mean_token_accuracy": 0.8238518476486206,
      "num_tokens": 308585833.0,
      "step": 9210
    },
    {
      "epoch": 0.5500835721107927,
      "grad_norm": 0.5403125882148743,
      "learning_rate": 4.405410465056946e-05,
      "loss": 0.5466,
      "mean_token_accuracy": 0.8265835642814636,
      "num_tokens": 308753513.0,
      "step": 9215
    },
    {
      "epoch": 0.5503820439350525,
      "grad_norm": 0.5557153820991516,
      "learning_rate": 4.40465820346608e-05,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8229631543159485,
      "num_tokens": 308921193.0,
      "step": 9220
    },
    {
      "epoch": 0.5506805157593123,
      "grad_norm": 0.5201708078384399,
      "learning_rate": 4.4039055388526674e-05,
      "loss": 0.5721,
      "mean_token_accuracy": 0.8206608533859253,
      "num_tokens": 309088873.0,
      "step": 9225
    },
    {
      "epoch": 0.5509789875835721,
      "grad_norm": 0.5418476462364197,
      "learning_rate": 4.403152471400033e-05,
      "loss": 0.5611,
      "mean_token_accuracy": 0.8228677034378051,
      "num_tokens": 309256553.0,
      "step": 9230
    },
    {
      "epoch": 0.5512774594078319,
      "grad_norm": 0.5212776064872742,
      "learning_rate": 4.4023990012916e-05,
      "loss": 0.5779,
      "mean_token_accuracy": 0.8170285105705262,
      "num_tokens": 309424233.0,
      "step": 9235
    },
    {
      "epoch": 0.5515759312320917,
      "grad_norm": 0.5821331739425659,
      "learning_rate": 4.401645128710892e-05,
      "loss": 0.5783,
      "mean_token_accuracy": 0.8175652980804443,
      "num_tokens": 309591913.0,
      "step": 9240
    },
    {
      "epoch": 0.5518744030563515,
      "grad_norm": 0.5345621109008789,
      "learning_rate": 4.400890853841525e-05,
      "loss": 0.5586,
      "mean_token_accuracy": 0.8244960069656372,
      "num_tokens": 309759593.0,
      "step": 9245
    },
    {
      "epoch": 0.5521728748806113,
      "grad_norm": 0.557401716709137,
      "learning_rate": 4.40013617686722e-05,
      "loss": 0.5821,
      "mean_token_accuracy": 0.8174639105796814,
      "num_tokens": 309927273.0,
      "step": 9250
    },
    {
      "epoch": 0.5524713467048711,
      "grad_norm": 0.5907803177833557,
      "learning_rate": 4.399381097971791e-05,
      "loss": 0.6042,
      "mean_token_accuracy": 0.8101932406425476,
      "num_tokens": 310094953.0,
      "step": 9255
    },
    {
      "epoch": 0.5527698185291309,
      "grad_norm": 0.5467956066131592,
      "learning_rate": 4.39862561733915e-05,
      "loss": 0.5613,
      "mean_token_accuracy": 0.8219969034194946,
      "num_tokens": 310262633.0,
      "step": 9260
    },
    {
      "epoch": 0.5530682903533907,
      "grad_norm": 0.49250057339668274,
      "learning_rate": 4.397869735153312e-05,
      "loss": 0.5403,
      "mean_token_accuracy": 0.8290886282920837,
      "num_tokens": 310430313.0,
      "step": 9265
    },
    {
      "epoch": 0.5533667621776505,
      "grad_norm": 0.47722843289375305,
      "learning_rate": 4.3971134515983824e-05,
      "loss": 0.5538,
      "mean_token_accuracy": 0.8251759529113769,
      "num_tokens": 310597993.0,
      "step": 9270
    },
    {
      "epoch": 0.5536652340019103,
      "grad_norm": 0.5117371082305908,
      "learning_rate": 4.39635676685857e-05,
      "loss": 0.5841,
      "mean_token_accuracy": 0.8160324454307556,
      "num_tokens": 310765673.0,
      "step": 9275
    },
    {
      "epoch": 0.55396370582617,
      "grad_norm": 0.5404113531112671,
      "learning_rate": 4.395599681118178e-05,
      "loss": 0.5415,
      "mean_token_accuracy": 0.8286114692687988,
      "num_tokens": 310933353.0,
      "step": 9280
    },
    {
      "epoch": 0.5542621776504298,
      "grad_norm": 0.4832151532173157,
      "learning_rate": 4.3948421945616095e-05,
      "loss": 0.5158,
      "mean_token_accuracy": 0.835840392112732,
      "num_tokens": 311101033.0,
      "step": 9285
    },
    {
      "epoch": 0.5545606494746896,
      "grad_norm": 0.5472707748413086,
      "learning_rate": 4.394084307373365e-05,
      "loss": 0.5756,
      "mean_token_accuracy": 0.8209352135658264,
      "num_tokens": 311268713.0,
      "step": 9290
    },
    {
      "epoch": 0.5548591212989494,
      "grad_norm": 1.157281517982483,
      "learning_rate": 4.393326019738041e-05,
      "loss": 0.569,
      "mean_token_accuracy": 0.8219909429550171,
      "num_tokens": 311436393.0,
      "step": 9295
    },
    {
      "epoch": 0.5551575931232091,
      "grad_norm": 0.5492601990699768,
      "learning_rate": 4.392567331840332e-05,
      "loss": 0.5624,
      "mean_token_accuracy": 0.823696768283844,
      "num_tokens": 311604073.0,
      "step": 9300
    },
    {
      "epoch": 0.5554560649474689,
      "grad_norm": 0.5792446136474609,
      "learning_rate": 4.391808243865031e-05,
      "loss": 0.5631,
      "mean_token_accuracy": 0.8220625042915344,
      "num_tokens": 311771753.0,
      "step": 9305
    },
    {
      "epoch": 0.5557545367717287,
      "grad_norm": 0.493204265832901,
      "learning_rate": 4.391048755997028e-05,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8309495449066162,
      "num_tokens": 311939433.0,
      "step": 9310
    },
    {
      "epoch": 0.5560530085959885,
      "grad_norm": 0.5801738500595093,
      "learning_rate": 4.390288868421311e-05,
      "loss": 0.542,
      "mean_token_accuracy": 0.8293331742286683,
      "num_tokens": 312107113.0,
      "step": 9315
    },
    {
      "epoch": 0.5563514804202483,
      "grad_norm": 0.6259301900863647,
      "learning_rate": 4.389528581322963e-05,
      "loss": 0.6023,
      "mean_token_accuracy": 0.8118394255638123,
      "num_tokens": 312274793.0,
      "step": 9320
    },
    {
      "epoch": 0.5566499522445081,
      "grad_norm": 0.5844714045524597,
      "learning_rate": 4.388767894887167e-05,
      "loss": 0.6053,
      "mean_token_accuracy": 0.8097101211547851,
      "num_tokens": 312442473.0,
      "step": 9325
    },
    {
      "epoch": 0.5569484240687679,
      "grad_norm": 0.5438195466995239,
      "learning_rate": 4.388006809299202e-05,
      "loss": 0.5275,
      "mean_token_accuracy": 0.8327627301216125,
      "num_tokens": 312610153.0,
      "step": 9330
    },
    {
      "epoch": 0.5572468958930277,
      "grad_norm": 0.5652937889099121,
      "learning_rate": 4.387245324744444e-05,
      "loss": 0.5673,
      "mean_token_accuracy": 0.8214183568954467,
      "num_tokens": 312777833.0,
      "step": 9335
    },
    {
      "epoch": 0.5575453677172875,
      "grad_norm": 0.5891305804252625,
      "learning_rate": 4.386483441408368e-05,
      "loss": 0.5322,
      "mean_token_accuracy": 0.8325539708137513,
      "num_tokens": 312945513.0,
      "step": 9340
    },
    {
      "epoch": 0.5578438395415473,
      "grad_norm": 0.5884708762168884,
      "learning_rate": 4.3857211594765435e-05,
      "loss": 0.6028,
      "mean_token_accuracy": 0.8108314394950866,
      "num_tokens": 313113193.0,
      "step": 9345
    },
    {
      "epoch": 0.558142311365807,
      "grad_norm": 0.5042849183082581,
      "learning_rate": 4.384958479134638e-05,
      "loss": 0.5473,
      "mean_token_accuracy": 0.8275975227355957,
      "num_tokens": 313280873.0,
      "step": 9350
    },
    {
      "epoch": 0.5584407831900668,
      "grad_norm": 0.4993573725223541,
      "learning_rate": 4.384195400568418e-05,
      "loss": 0.5372,
      "mean_token_accuracy": 0.8292258024215698,
      "num_tokens": 313448553.0,
      "step": 9355
    },
    {
      "epoch": 0.5587392550143266,
      "grad_norm": 0.5659687519073486,
      "learning_rate": 4.383431923963743e-05,
      "loss": 0.648,
      "mean_token_accuracy": 0.7970893502235412,
      "num_tokens": 313616233.0,
      "step": 9360
    },
    {
      "epoch": 0.5590377268385864,
      "grad_norm": 0.5630433559417725,
      "learning_rate": 4.382668049506575e-05,
      "loss": 0.5855,
      "mean_token_accuracy": 0.8163246989250184,
      "num_tokens": 313783913.0,
      "step": 9365
    },
    {
      "epoch": 0.5593361986628462,
      "grad_norm": 0.6412773728370667,
      "learning_rate": 4.3819037773829666e-05,
      "loss": 0.5516,
      "mean_token_accuracy": 0.8258380055427551,
      "num_tokens": 313951593.0,
      "step": 9370
    },
    {
      "epoch": 0.559634670487106,
      "grad_norm": 0.5726149678230286,
      "learning_rate": 4.381139107779073e-05,
      "loss": 0.5945,
      "mean_token_accuracy": 0.8127221822738647,
      "num_tokens": 314119273.0,
      "step": 9375
    },
    {
      "epoch": 0.5599331423113658,
      "grad_norm": 0.5358670353889465,
      "learning_rate": 4.380374040881141e-05,
      "loss": 0.5302,
      "mean_token_accuracy": 0.8315400123596192,
      "num_tokens": 314286953.0,
      "step": 9380
    },
    {
      "epoch": 0.5602316141356256,
      "grad_norm": 0.4920586943626404,
      "learning_rate": 4.379608576875519e-05,
      "loss": 0.5699,
      "mean_token_accuracy": 0.8202254414558411,
      "num_tokens": 314454633.0,
      "step": 9385
    },
    {
      "epoch": 0.5605300859598854,
      "grad_norm": 0.5719877481460571,
      "learning_rate": 4.378842715948648e-05,
      "loss": 0.538,
      "mean_token_accuracy": 0.8295538663864136,
      "num_tokens": 314622313.0,
      "step": 9390
    },
    {
      "epoch": 0.5608285577841452,
      "grad_norm": 0.5797923803329468,
      "learning_rate": 4.37807645828707e-05,
      "loss": 0.5424,
      "mean_token_accuracy": 0.8275376915931701,
      "num_tokens": 314788279.0,
      "step": 9395
    },
    {
      "epoch": 0.561127029608405,
      "grad_norm": 0.581619918346405,
      "learning_rate": 4.3773098040774185e-05,
      "loss": 0.512,
      "mean_token_accuracy": 0.8385482430458069,
      "num_tokens": 314955959.0,
      "step": 9400
    },
    {
      "epoch": 0.5614255014326648,
      "grad_norm": 0.5046367645263672,
      "learning_rate": 4.3765427535064275e-05,
      "loss": 0.5298,
      "mean_token_accuracy": 0.8310807704925537,
      "num_tokens": 315123639.0,
      "step": 9405
    },
    {
      "epoch": 0.5617239732569246,
      "grad_norm": 0.6082244515419006,
      "learning_rate": 4.375775306760927e-05,
      "loss": 0.5583,
      "mean_token_accuracy": 0.8237564086914062,
      "num_tokens": 315291319.0,
      "step": 9410
    },
    {
      "epoch": 0.5620224450811844,
      "grad_norm": 0.5142558813095093,
      "learning_rate": 4.3750074640278414e-05,
      "loss": 0.5056,
      "mean_token_accuracy": 0.8406477332115173,
      "num_tokens": 315458999.0,
      "step": 9415
    },
    {
      "epoch": 0.5623209169054442,
      "grad_norm": 0.5200819373130798,
      "learning_rate": 4.374239225494195e-05,
      "loss": 0.5209,
      "mean_token_accuracy": 0.8346773266792298,
      "num_tokens": 315626679.0,
      "step": 9420
    },
    {
      "epoch": 0.562619388729704,
      "grad_norm": 0.466137170791626,
      "learning_rate": 4.373470591347104e-05,
      "loss": 0.5232,
      "mean_token_accuracy": 0.8334725141525269,
      "num_tokens": 315794359.0,
      "step": 9425
    },
    {
      "epoch": 0.5629178605539638,
      "grad_norm": 0.6146972179412842,
      "learning_rate": 4.3727015617737846e-05,
      "loss": 0.5821,
      "mean_token_accuracy": 0.8179410696029663,
      "num_tokens": 315962039.0,
      "step": 9430
    },
    {
      "epoch": 0.5632163323782235,
      "grad_norm": 0.5308025479316711,
      "learning_rate": 4.371932136961549e-05,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8369080185890198,
      "num_tokens": 316129719.0,
      "step": 9435
    },
    {
      "epoch": 0.5635148042024832,
      "grad_norm": 0.7136004567146301,
      "learning_rate": 4.371162317097804e-05,
      "loss": 0.5591,
      "mean_token_accuracy": 0.8240427017211914,
      "num_tokens": 316297399.0,
      "step": 9440
    },
    {
      "epoch": 0.563813276026743,
      "grad_norm": 0.6727559566497803,
      "learning_rate": 4.3703921023700535e-05,
      "loss": 0.5811,
      "mean_token_accuracy": 0.8192592024803161,
      "num_tokens": 316465079.0,
      "step": 9445
    },
    {
      "epoch": 0.5641117478510028,
      "grad_norm": 0.4987635016441345,
      "learning_rate": 4.369621492965896e-05,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8316951036453247,
      "num_tokens": 316632759.0,
      "step": 9450
    },
    {
      "epoch": 0.5644102196752626,
      "grad_norm": 0.5753521919250488,
      "learning_rate": 4.368850489073031e-05,
      "loss": 0.5824,
      "mean_token_accuracy": 0.8179172158241272,
      "num_tokens": 316800439.0,
      "step": 9455
    },
    {
      "epoch": 0.5647086914995224,
      "grad_norm": 0.5425297021865845,
      "learning_rate": 4.368079090879248e-05,
      "loss": 0.5921,
      "mean_token_accuracy": 0.8140403270721436,
      "num_tokens": 316968119.0,
      "step": 9460
    },
    {
      "epoch": 0.5650071633237822,
      "grad_norm": 0.5226919651031494,
      "learning_rate": 4.3673072985724364e-05,
      "loss": 0.5893,
      "mean_token_accuracy": 0.8144518494606018,
      "num_tokens": 317135799.0,
      "step": 9465
    },
    {
      "epoch": 0.565305635148042,
      "grad_norm": 0.5343566536903381,
      "learning_rate": 4.366535112340579e-05,
      "loss": 0.5539,
      "mean_token_accuracy": 0.8235655546188354,
      "num_tokens": 317303479.0,
      "step": 9470
    },
    {
      "epoch": 0.5656041069723018,
      "grad_norm": 0.6416198015213013,
      "learning_rate": 4.365762532371758e-05,
      "loss": 0.5733,
      "mean_token_accuracy": 0.8221039295196533,
      "num_tokens": 317462968.0,
      "step": 9475
    },
    {
      "epoch": 0.5659025787965616,
      "grad_norm": 0.5042870044708252,
      "learning_rate": 4.364989558854149e-05,
      "loss": 0.5394,
      "mean_token_accuracy": 0.8275856018066406,
      "num_tokens": 317630648.0,
      "step": 9480
    },
    {
      "epoch": 0.5662010506208214,
      "grad_norm": 0.5087573528289795,
      "learning_rate": 4.364216191976024e-05,
      "loss": 0.5347,
      "mean_token_accuracy": 0.8297566652297974,
      "num_tokens": 317798328.0,
      "step": 9485
    },
    {
      "epoch": 0.5664995224450812,
      "grad_norm": 0.5444625616073608,
      "learning_rate": 4.3634424319257495e-05,
      "loss": 0.5488,
      "mean_token_accuracy": 0.8262435913085937,
      "num_tokens": 317966008.0,
      "step": 9490
    },
    {
      "epoch": 0.566797994269341,
      "grad_norm": 0.5101728439331055,
      "learning_rate": 4.36266827889179e-05,
      "loss": 0.5384,
      "mean_token_accuracy": 0.8292794942855835,
      "num_tokens": 318133688.0,
      "step": 9495
    },
    {
      "epoch": 0.5670964660936008,
      "grad_norm": 0.5687821507453918,
      "learning_rate": 4.361893733062705e-05,
      "loss": 0.5564,
      "mean_token_accuracy": 0.8238279700279236,
      "num_tokens": 318301368.0,
      "step": 9500
    },
    {
      "epoch": 0.5673949379178606,
      "grad_norm": 0.5257329940795898,
      "learning_rate": 4.3611187946271495e-05,
      "loss": 0.5781,
      "mean_token_accuracy": 0.8187760829925537,
      "num_tokens": 318469048.0,
      "step": 9505
    },
    {
      "epoch": 0.5676934097421203,
      "grad_norm": 0.6087618470191956,
      "learning_rate": 4.360343463773873e-05,
      "loss": 0.5528,
      "mean_token_accuracy": 0.8239353418350219,
      "num_tokens": 318636728.0,
      "step": 9510
    },
    {
      "epoch": 0.5679918815663801,
      "grad_norm": 0.5090599060058594,
      "learning_rate": 4.359567740691723e-05,
      "loss": 0.5327,
      "mean_token_accuracy": 0.8291423320770264,
      "num_tokens": 318804408.0,
      "step": 9515
    },
    {
      "epoch": 0.5682903533906399,
      "grad_norm": 0.6896741986274719,
      "learning_rate": 4.35879162556964e-05,
      "loss": 0.574,
      "mean_token_accuracy": 0.8179172039031982,
      "num_tokens": 318972088.0,
      "step": 9520
    },
    {
      "epoch": 0.5685888252148997,
      "grad_norm": 0.5273687839508057,
      "learning_rate": 4.3580151185966625e-05,
      "loss": 0.5561,
      "mean_token_accuracy": 0.8233985424041748,
      "num_tokens": 319139768.0,
      "step": 9525
    },
    {
      "epoch": 0.5688872970391595,
      "grad_norm": 0.5132014751434326,
      "learning_rate": 4.357238219961922e-05,
      "loss": 0.54,
      "mean_token_accuracy": 0.8274961352348328,
      "num_tokens": 319307448.0,
      "step": 9530
    },
    {
      "epoch": 0.5691857688634193,
      "grad_norm": 0.5733283758163452,
      "learning_rate": 4.3564609298546463e-05,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8231062889099121,
      "num_tokens": 319475128.0,
      "step": 9535
    },
    {
      "epoch": 0.5694842406876791,
      "grad_norm": 0.5522651672363281,
      "learning_rate": 4.3556832484641594e-05,
      "loss": 0.5233,
      "mean_token_accuracy": 0.8317189574241638,
      "num_tokens": 319642808.0,
      "step": 9540
    },
    {
      "epoch": 0.5697827125119389,
      "grad_norm": 0.5674959421157837,
      "learning_rate": 4.3549051759798805e-05,
      "loss": 0.5505,
      "mean_token_accuracy": 0.8244125008583069,
      "num_tokens": 319810488.0,
      "step": 9545
    },
    {
      "epoch": 0.5700811843361987,
      "grad_norm": 0.5781137347221375,
      "learning_rate": 4.354126712591322e-05,
      "loss": 0.551,
      "mean_token_accuracy": 0.8258797526359558,
      "num_tokens": 319978168.0,
      "step": 9550
    },
    {
      "epoch": 0.5703796561604585,
      "grad_norm": 0.6012808680534363,
      "learning_rate": 4.3533478584880945e-05,
      "loss": 0.5749,
      "mean_token_accuracy": 0.8197065591812134,
      "num_tokens": 320145848.0,
      "step": 9555
    },
    {
      "epoch": 0.5706781279847183,
      "grad_norm": 0.5958951711654663,
      "learning_rate": 4.352568613859901e-05,
      "loss": 0.558,
      "mean_token_accuracy": 0.8245258212089539,
      "num_tokens": 320313528.0,
      "step": 9560
    },
    {
      "epoch": 0.5709765998089781,
      "grad_norm": 0.564767062664032,
      "learning_rate": 4.3517889788965426e-05,
      "loss": 0.555,
      "mean_token_accuracy": 0.8242634057998657,
      "num_tokens": 320481208.0,
      "step": 9565
    },
    {
      "epoch": 0.5712750716332379,
      "grad_norm": 0.5786231160163879,
      "learning_rate": 4.3510089537879113e-05,
      "loss": 0.5642,
      "mean_token_accuracy": 0.8220267176628113,
      "num_tokens": 320648888.0,
      "step": 9570
    },
    {
      "epoch": 0.5715735434574976,
      "grad_norm": 0.6010572910308838,
      "learning_rate": 4.350228538723999e-05,
      "loss": 0.6235,
      "mean_token_accuracy": 0.8046940326690674,
      "num_tokens": 320816568.0,
      "step": 9575
    },
    {
      "epoch": 0.5718720152817574,
      "grad_norm": 0.6927845478057861,
      "learning_rate": 4.349447733894889e-05,
      "loss": 0.5852,
      "mean_token_accuracy": 0.8170523524284363,
      "num_tokens": 320984248.0,
      "step": 9580
    },
    {
      "epoch": 0.5721704871060171,
      "grad_norm": 0.517074465751648,
      "learning_rate": 4.348666539490761e-05,
      "loss": 0.5453,
      "mean_token_accuracy": 0.8278778553009033,
      "num_tokens": 321151928.0,
      "step": 9585
    },
    {
      "epoch": 0.5724689589302769,
      "grad_norm": 0.5346642732620239,
      "learning_rate": 4.34788495570189e-05,
      "loss": 0.5664,
      "mean_token_accuracy": 0.8198437333106995,
      "num_tokens": 321319608.0,
      "step": 9590
    },
    {
      "epoch": 0.5727674307545367,
      "grad_norm": 0.5233079791069031,
      "learning_rate": 4.347102982718644e-05,
      "loss": 0.5575,
      "mean_token_accuracy": 0.8219968914985657,
      "num_tokens": 321487288.0,
      "step": 9595
    },
    {
      "epoch": 0.5730659025787965,
      "grad_norm": 0.5271203517913818,
      "learning_rate": 4.346320620731487e-05,
      "loss": 0.5725,
      "mean_token_accuracy": 0.8205177068710328,
      "num_tokens": 321654968.0,
      "step": 9600
    },
    {
      "epoch": 0.5733643744030563,
      "grad_norm": 0.5608517527580261,
      "learning_rate": 4.345537869930978e-05,
      "loss": 0.5646,
      "mean_token_accuracy": 0.8202791333198547,
      "num_tokens": 321822648.0,
      "step": 9605
    },
    {
      "epoch": 0.5736628462273161,
      "grad_norm": 0.49211350083351135,
      "learning_rate": 4.3447547305077716e-05,
      "loss": 0.5173,
      "mean_token_accuracy": 0.8362519383430481,
      "num_tokens": 321990328.0,
      "step": 9610
    },
    {
      "epoch": 0.5739613180515759,
      "grad_norm": 0.5640000700950623,
      "learning_rate": 4.3439712026526134e-05,
      "loss": 0.5237,
      "mean_token_accuracy": 0.8342419147491456,
      "num_tokens": 322158008.0,
      "step": 9615
    },
    {
      "epoch": 0.5742597898758357,
      "grad_norm": 0.5405912399291992,
      "learning_rate": 4.3431872865563484e-05,
      "loss": 0.5623,
      "mean_token_accuracy": 0.821728503704071,
      "num_tokens": 322325688.0,
      "step": 9620
    },
    {
      "epoch": 0.5745582617000955,
      "grad_norm": 0.5111109018325806,
      "learning_rate": 4.3424029824099126e-05,
      "loss": 0.5811,
      "mean_token_accuracy": 0.8195753216743469,
      "num_tokens": 322493368.0,
      "step": 9625
    },
    {
      "epoch": 0.5748567335243553,
      "grad_norm": 0.48165133595466614,
      "learning_rate": 4.341618290404338e-05,
      "loss": 0.5331,
      "mean_token_accuracy": 0.82977454662323,
      "num_tokens": 322661048.0,
      "step": 9630
    },
    {
      "epoch": 0.5751552053486151,
      "grad_norm": 0.6553030610084534,
      "learning_rate": 4.340833210730751e-05,
      "loss": 0.5654,
      "mean_token_accuracy": 0.8210724115371704,
      "num_tokens": 322828728.0,
      "step": 9635
    },
    {
      "epoch": 0.5754536771728749,
      "grad_norm": 0.5732831358909607,
      "learning_rate": 4.340047743580372e-05,
      "loss": 0.6124,
      "mean_token_accuracy": 0.8088810801506042,
      "num_tokens": 322996408.0,
      "step": 9640
    },
    {
      "epoch": 0.5757521489971347,
      "grad_norm": 0.5754977464675903,
      "learning_rate": 4.339261889144516e-05,
      "loss": 0.5964,
      "mean_token_accuracy": 0.8130919814109803,
      "num_tokens": 323164088.0,
      "step": 9645
    },
    {
      "epoch": 0.5760506208213945,
      "grad_norm": 0.5848853588104248,
      "learning_rate": 4.338475647614593e-05,
      "loss": 0.5491,
      "mean_token_accuracy": 0.8263748049736023,
      "num_tokens": 323331768.0,
      "step": 9650
    },
    {
      "epoch": 0.5763490926456543,
      "grad_norm": 0.6020246744155884,
      "learning_rate": 4.3376890191821054e-05,
      "loss": 0.5781,
      "mean_token_accuracy": 0.8168436169624329,
      "num_tokens": 323499448.0,
      "step": 9655
    },
    {
      "epoch": 0.576647564469914,
      "grad_norm": 0.6079437732696533,
      "learning_rate": 4.336902004038653e-05,
      "loss": 0.545,
      "mean_token_accuracy": 0.8260706305503845,
      "num_tokens": 323667128.0,
      "step": 9660
    },
    {
      "epoch": 0.5769460362941738,
      "grad_norm": 0.5589531064033508,
      "learning_rate": 4.336114602375926e-05,
      "loss": 0.5385,
      "mean_token_accuracy": 0.8265060067176819,
      "num_tokens": 323834808.0,
      "step": 9665
    },
    {
      "epoch": 0.5772445081184336,
      "grad_norm": 0.5987696647644043,
      "learning_rate": 4.3353268143857104e-05,
      "loss": 0.5751,
      "mean_token_accuracy": 0.8176845788955689,
      "num_tokens": 324002488.0,
      "step": 9670
    },
    {
      "epoch": 0.5775429799426934,
      "grad_norm": 0.5677342414855957,
      "learning_rate": 4.334538640259888e-05,
      "loss": 0.5416,
      "mean_token_accuracy": 0.8283788681030273,
      "num_tokens": 324170168.0,
      "step": 9675
    },
    {
      "epoch": 0.5778414517669532,
      "grad_norm": 0.526494562625885,
      "learning_rate": 4.3337500801904336e-05,
      "loss": 0.6107,
      "mean_token_accuracy": 0.8081235766410828,
      "num_tokens": 324337848.0,
      "step": 9680
    },
    {
      "epoch": 0.578139923591213,
      "grad_norm": 0.5406463742256165,
      "learning_rate": 4.332961134369413e-05,
      "loss": 0.569,
      "mean_token_accuracy": 0.8202552795410156,
      "num_tokens": 324505528.0,
      "step": 9685
    },
    {
      "epoch": 0.5784383954154728,
      "grad_norm": 0.6007771492004395,
      "learning_rate": 4.332171802988991e-05,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8211857318878174,
      "num_tokens": 324673208.0,
      "step": 9690
    },
    {
      "epoch": 0.5787368672397326,
      "grad_norm": 0.6053233742713928,
      "learning_rate": 4.3313820862414225e-05,
      "loss": 0.5868,
      "mean_token_accuracy": 0.8157401919364929,
      "num_tokens": 324840888.0,
      "step": 9695
    },
    {
      "epoch": 0.5790353390639924,
      "grad_norm": 0.6275655627250671,
      "learning_rate": 4.3305919843190576e-05,
      "loss": 0.582,
      "mean_token_accuracy": 0.8180722951889038,
      "num_tokens": 325008568.0,
      "step": 9700
    },
    {
      "epoch": 0.5793338108882522,
      "grad_norm": 0.573482096195221,
      "learning_rate": 4.32980149741434e-05,
      "loss": 0.5506,
      "mean_token_accuracy": 0.8243409276008606,
      "num_tokens": 325176248.0,
      "step": 9705
    },
    {
      "epoch": 0.579632282712512,
      "grad_norm": 0.5884882211685181,
      "learning_rate": 4.329010625719809e-05,
      "loss": 0.5934,
      "mean_token_accuracy": 0.8156030178070068,
      "num_tokens": 325343928.0,
      "step": 9710
    },
    {
      "epoch": 0.5799307545367717,
      "grad_norm": 0.5336214303970337,
      "learning_rate": 4.328219369428094e-05,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8237444877624511,
      "num_tokens": 325511608.0,
      "step": 9715
    },
    {
      "epoch": 0.5802292263610315,
      "grad_norm": 0.5747628808021545,
      "learning_rate": 4.327427728731922e-05,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8334784746170044,
      "num_tokens": 325679288.0,
      "step": 9720
    },
    {
      "epoch": 0.5805276981852913,
      "grad_norm": 0.5531827807426453,
      "learning_rate": 4.3266357038241104e-05,
      "loss": 0.5738,
      "mean_token_accuracy": 0.8195393085479736,
      "num_tokens": 325846270.0,
      "step": 9725
    },
    {
      "epoch": 0.5808261700095511,
      "grad_norm": 0.5761283040046692,
      "learning_rate": 4.3258432948975714e-05,
      "loss": 0.5596,
      "mean_token_accuracy": 0.8210008263587951,
      "num_tokens": 326013950.0,
      "step": 9730
    },
    {
      "epoch": 0.5811246418338109,
      "grad_norm": 0.551312267780304,
      "learning_rate": 4.325050502145311e-05,
      "loss": 0.577,
      "mean_token_accuracy": 0.8192174553871154,
      "num_tokens": 326181630.0,
      "step": 9735
    },
    {
      "epoch": 0.5814231136580706,
      "grad_norm": 0.6245538592338562,
      "learning_rate": 4.32425732576043e-05,
      "loss": 0.5884,
      "mean_token_accuracy": 0.8152570724487305,
      "num_tokens": 326349310.0,
      "step": 9740
    },
    {
      "epoch": 0.5817215854823304,
      "grad_norm": 0.5730269551277161,
      "learning_rate": 4.3234637659361186e-05,
      "loss": 0.5705,
      "mean_token_accuracy": 0.8192592144012452,
      "num_tokens": 326516990.0,
      "step": 9745
    },
    {
      "epoch": 0.5820200573065902,
      "grad_norm": 0.5351265072822571,
      "learning_rate": 4.322669822865665e-05,
      "loss": 0.5542,
      "mean_token_accuracy": 0.8247405409812927,
      "num_tokens": 326684670.0,
      "step": 9750
    },
    {
      "epoch": 0.58231852913085,
      "grad_norm": 0.48421576619148254,
      "learning_rate": 4.3218754967424465e-05,
      "loss": 0.5154,
      "mean_token_accuracy": 0.8354944586753845,
      "num_tokens": 326852350.0,
      "step": 9755
    },
    {
      "epoch": 0.5826170009551098,
      "grad_norm": 0.5339226126670837,
      "learning_rate": 4.321080787759938e-05,
      "loss": 0.5855,
      "mean_token_accuracy": 0.8142848610877991,
      "num_tokens": 327020030.0,
      "step": 9760
    },
    {
      "epoch": 0.5829154727793696,
      "grad_norm": 0.46684908866882324,
      "learning_rate": 4.3202856961117056e-05,
      "loss": 0.5203,
      "mean_token_accuracy": 0.8335202217102051,
      "num_tokens": 327187710.0,
      "step": 9765
    },
    {
      "epoch": 0.5832139446036294,
      "grad_norm": 0.534252405166626,
      "learning_rate": 4.3194902219914074e-05,
      "loss": 0.55,
      "mean_token_accuracy": 0.82533700466156,
      "num_tokens": 327355390.0,
      "step": 9770
    },
    {
      "epoch": 0.5835124164278892,
      "grad_norm": 0.5405399203300476,
      "learning_rate": 4.318694365592795e-05,
      "loss": 0.5434,
      "mean_token_accuracy": 0.826273787021637,
      "num_tokens": 327518373.0,
      "step": 9775
    },
    {
      "epoch": 0.583810888252149,
      "grad_norm": 0.5407677292823792,
      "learning_rate": 4.317898127109717e-05,
      "loss": 0.5783,
      "mean_token_accuracy": 0.8166765928268432,
      "num_tokens": 327686053.0,
      "step": 9780
    },
    {
      "epoch": 0.5841093600764088,
      "grad_norm": 0.5031781196594238,
      "learning_rate": 4.3171015067361086e-05,
      "loss": 0.5592,
      "mean_token_accuracy": 0.8218119859695434,
      "num_tokens": 327853733.0,
      "step": 9785
    },
    {
      "epoch": 0.5844078319006686,
      "grad_norm": 0.5705164670944214,
      "learning_rate": 4.316304504666004e-05,
      "loss": 0.575,
      "mean_token_accuracy": 0.8173446297645569,
      "num_tokens": 328021413.0,
      "step": 9790
    },
    {
      "epoch": 0.5847063037249284,
      "grad_norm": 0.5071836113929749,
      "learning_rate": 4.315507121093525e-05,
      "loss": 0.5842,
      "mean_token_accuracy": 0.8162352323532105,
      "num_tokens": 328189093.0,
      "step": 9795
    },
    {
      "epoch": 0.5850047755491882,
      "grad_norm": 0.49702900648117065,
      "learning_rate": 4.314709356212893e-05,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8300668120384216,
      "num_tokens": 328356773.0,
      "step": 9800
    },
    {
      "epoch": 0.585303247373448,
      "grad_norm": 0.5357098579406738,
      "learning_rate": 4.313911210218414e-05,
      "loss": 0.567,
      "mean_token_accuracy": 0.8208934783935546,
      "num_tokens": 328524453.0,
      "step": 9805
    },
    {
      "epoch": 0.5856017191977078,
      "grad_norm": 0.6276393532752991,
      "learning_rate": 4.313112683304494e-05,
      "loss": 0.5671,
      "mean_token_accuracy": 0.8203610420227051,
      "num_tokens": 328689970.0,
      "step": 9810
    },
    {
      "epoch": 0.5859001910219676,
      "grad_norm": 0.5248390436172485,
      "learning_rate": 4.312313775665628e-05,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8238339424133301,
      "num_tokens": 328857650.0,
      "step": 9815
    },
    {
      "epoch": 0.5861986628462273,
      "grad_norm": 0.5058759450912476,
      "learning_rate": 4.311514487496405e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8338363528251648,
      "num_tokens": 329025330.0,
      "step": 9820
    },
    {
      "epoch": 0.5864971346704871,
      "grad_norm": 0.5292270183563232,
      "learning_rate": 4.310714818991507e-05,
      "loss": 0.5457,
      "mean_token_accuracy": 0.8273947358131408,
      "num_tokens": 329193010.0,
      "step": 9825
    },
    {
      "epoch": 0.5867956064947469,
      "grad_norm": 0.5985789895057678,
      "learning_rate": 4.3099147703457065e-05,
      "loss": 0.5591,
      "mean_token_accuracy": 0.8240128874778747,
      "num_tokens": 329360690.0,
      "step": 9830
    },
    {
      "epoch": 0.5870940783190067,
      "grad_norm": 0.4757097363471985,
      "learning_rate": 4.3091143417538714e-05,
      "loss": 0.5555,
      "mean_token_accuracy": 0.8237862229347229,
      "num_tokens": 329528370.0,
      "step": 9835
    },
    {
      "epoch": 0.5873925501432665,
      "grad_norm": 0.550864577293396,
      "learning_rate": 4.3083135334109595e-05,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8260825514793396,
      "num_tokens": 329696050.0,
      "step": 9840
    },
    {
      "epoch": 0.5876910219675263,
      "grad_norm": 0.5893641710281372,
      "learning_rate": 4.307512345512025e-05,
      "loss": 0.5191,
      "mean_token_accuracy": 0.8350016593933105,
      "num_tokens": 329855467.0,
      "step": 9845
    },
    {
      "epoch": 0.5879894937917861,
      "grad_norm": 0.5553637146949768,
      "learning_rate": 4.306710778252209e-05,
      "loss": 0.5652,
      "mean_token_accuracy": 0.8219551563262939,
      "num_tokens": 330023147.0,
      "step": 9850
    },
    {
      "epoch": 0.5882879656160458,
      "grad_norm": 0.5286957025527954,
      "learning_rate": 4.30590883182675e-05,
      "loss": 0.5326,
      "mean_token_accuracy": 0.8308302640914917,
      "num_tokens": 330190827.0,
      "step": 9855
    },
    {
      "epoch": 0.5885864374403056,
      "grad_norm": 0.5817766189575195,
      "learning_rate": 4.305106506430976e-05,
      "loss": 0.6034,
      "mean_token_accuracy": 0.8110819578170776,
      "num_tokens": 330358507.0,
      "step": 9860
    },
    {
      "epoch": 0.5888849092645654,
      "grad_norm": 0.518365204334259,
      "learning_rate": 4.3043038022603086e-05,
      "loss": 0.5446,
      "mean_token_accuracy": 0.8270607233047486,
      "num_tokens": 330526187.0,
      "step": 9865
    },
    {
      "epoch": 0.5891833810888252,
      "grad_norm": 0.5628482103347778,
      "learning_rate": 4.303500719510261e-05,
      "loss": 0.5875,
      "mean_token_accuracy": 0.8142371416091919,
      "num_tokens": 330693867.0,
      "step": 9870
    },
    {
      "epoch": 0.589481852913085,
      "grad_norm": 0.5530689358711243,
      "learning_rate": 4.3026972583764384e-05,
      "loss": 0.5415,
      "mean_token_accuracy": 0.827657163143158,
      "num_tokens": 330861547.0,
      "step": 9875
    },
    {
      "epoch": 0.5897803247373448,
      "grad_norm": 0.5350573062896729,
      "learning_rate": 4.3018934190545397e-05,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8293570399284362,
      "num_tokens": 331029227.0,
      "step": 9880
    },
    {
      "epoch": 0.5900787965616046,
      "grad_norm": 0.4851717948913574,
      "learning_rate": 4.3010892017403535e-05,
      "loss": 0.5421,
      "mean_token_accuracy": 0.8271263360977172,
      "num_tokens": 331196907.0,
      "step": 9885
    },
    {
      "epoch": 0.5903772683858644,
      "grad_norm": 0.564995288848877,
      "learning_rate": 4.300284606629763e-05,
      "loss": 0.6306,
      "mean_token_accuracy": 0.8031850099563599,
      "num_tokens": 331364587.0,
      "step": 9890
    },
    {
      "epoch": 0.5906757402101241,
      "grad_norm": 0.563106119632721,
      "learning_rate": 4.299479633918741e-05,
      "loss": 0.6004,
      "mean_token_accuracy": 0.8085470557212829,
      "num_tokens": 331532267.0,
      "step": 9895
    },
    {
      "epoch": 0.5909742120343839,
      "grad_norm": 0.5216640830039978,
      "learning_rate": 4.298674283803354e-05,
      "loss": 0.5502,
      "mean_token_accuracy": 0.8246331930160522,
      "num_tokens": 331699947.0,
      "step": 9900
    },
    {
      "epoch": 0.5912726838586437,
      "grad_norm": 0.5671737194061279,
      "learning_rate": 4.2978685564797606e-05,
      "loss": 0.5878,
      "mean_token_accuracy": 0.815626859664917,
      "num_tokens": 331867627.0,
      "step": 9905
    },
    {
      "epoch": 0.5915711556829035,
      "grad_norm": 0.5278947949409485,
      "learning_rate": 4.297062452144209e-05,
      "loss": 0.5746,
      "mean_token_accuracy": 0.8176488161087037,
      "num_tokens": 332035307.0,
      "step": 9910
    },
    {
      "epoch": 0.5918696275071633,
      "grad_norm": 0.5447534918785095,
      "learning_rate": 4.296255970993042e-05,
      "loss": 0.5442,
      "mean_token_accuracy": 0.8262256979942322,
      "num_tokens": 332202987.0,
      "step": 9915
    },
    {
      "epoch": 0.5921680993314231,
      "grad_norm": 0.5424688458442688,
      "learning_rate": 4.295449113222693e-05,
      "loss": 0.5458,
      "mean_token_accuracy": 0.8272754192352295,
      "num_tokens": 332370667.0,
      "step": 9920
    },
    {
      "epoch": 0.5924665711556829,
      "grad_norm": 0.507956326007843,
      "learning_rate": 4.2946418790296866e-05,
      "loss": 0.511,
      "mean_token_accuracy": 0.8354944467544556,
      "num_tokens": 332538347.0,
      "step": 9925
    },
    {
      "epoch": 0.5927650429799427,
      "grad_norm": 0.5435249209403992,
      "learning_rate": 4.2938342686106394e-05,
      "loss": 0.5337,
      "mean_token_accuracy": 0.8304545044898987,
      "num_tokens": 332706027.0,
      "step": 9930
    },
    {
      "epoch": 0.5930635148042025,
      "grad_norm": 0.5713160634040833,
      "learning_rate": 4.293026282162261e-05,
      "loss": 0.5512,
      "mean_token_accuracy": 0.826142179965973,
      "num_tokens": 332873707.0,
      "step": 9935
    },
    {
      "epoch": 0.5933619866284623,
      "grad_norm": 0.578020453453064,
      "learning_rate": 4.29221791988135e-05,
      "loss": 0.5546,
      "mean_token_accuracy": 0.8232732892036438,
      "num_tokens": 333041387.0,
      "step": 9940
    },
    {
      "epoch": 0.5936604584527221,
      "grad_norm": 0.5681842565536499,
      "learning_rate": 4.291409181964799e-05,
      "loss": 0.5804,
      "mean_token_accuracy": 0.8164141774177551,
      "num_tokens": 333209067.0,
      "step": 9945
    },
    {
      "epoch": 0.5939589302769819,
      "grad_norm": 0.5614805817604065,
      "learning_rate": 4.290600068609589e-05,
      "loss": 0.5503,
      "mean_token_accuracy": 0.8254622459411621,
      "num_tokens": 333376747.0,
      "step": 9950
    },
    {
      "epoch": 0.5942574021012417,
      "grad_norm": 0.6109113693237305,
      "learning_rate": 4.289790580012797e-05,
      "loss": 0.6079,
      "mean_token_accuracy": 0.8090242147445679,
      "num_tokens": 333544427.0,
      "step": 9955
    },
    {
      "epoch": 0.5945558739255015,
      "grad_norm": 0.530504047870636,
      "learning_rate": 4.2889807163715886e-05,
      "loss": 0.555,
      "mean_token_accuracy": 0.8229869961738586,
      "num_tokens": 333712107.0,
      "step": 9960
    },
    {
      "epoch": 0.5948543457497613,
      "grad_norm": 0.5134127140045166,
      "learning_rate": 4.2881704778832196e-05,
      "loss": 0.6026,
      "mean_token_accuracy": 0.8102051734924316,
      "num_tokens": 333879787.0,
      "step": 9965
    },
    {
      "epoch": 0.595152817574021,
      "grad_norm": 0.6299505233764648,
      "learning_rate": 4.2873598647450406e-05,
      "loss": 0.6735,
      "mean_token_accuracy": 0.7928307175636291,
      "num_tokens": 334047467.0,
      "step": 9970
    },
    {
      "epoch": 0.5954512893982808,
      "grad_norm": 0.5211063027381897,
      "learning_rate": 4.286548877154489e-05,
      "loss": 0.5522,
      "mean_token_accuracy": 0.8243946075439453,
      "num_tokens": 334215147.0,
      "step": 9975
    },
    {
      "epoch": 0.5957497612225406,
      "grad_norm": 0.5255116820335388,
      "learning_rate": 4.285737515309097e-05,
      "loss": 0.5315,
      "mean_token_accuracy": 0.8329178094863892,
      "num_tokens": 334382827.0,
      "step": 9980
    },
    {
      "epoch": 0.5960482330468004,
      "grad_norm": 0.5419749617576599,
      "learning_rate": 4.284925779406487e-05,
      "loss": 0.5967,
      "mean_token_accuracy": 0.8149648070335388,
      "num_tokens": 334550507.0,
      "step": 9985
    },
    {
      "epoch": 0.5963467048710601,
      "grad_norm": 0.602005660533905,
      "learning_rate": 4.284113669644372e-05,
      "loss": 0.6162,
      "mean_token_accuracy": 0.8094417214393616,
      "num_tokens": 334718187.0,
      "step": 9990
    },
    {
      "epoch": 0.5966451766953199,
      "grad_norm": 0.5772136449813843,
      "learning_rate": 4.2833011862205555e-05,
      "loss": 0.5873,
      "mean_token_accuracy": 0.8133782625198365,
      "num_tokens": 334885867.0,
      "step": 9995
    },
    {
      "epoch": 0.5969436485195797,
      "grad_norm": 0.6226353049278259,
      "learning_rate": 4.282488329332934e-05,
      "loss": 0.5509,
      "mean_token_accuracy": 0.8242872595787049,
      "num_tokens": 335053547.0,
      "step": 10000
    },
    {
      "epoch": 0.5972421203438395,
      "grad_norm": 0.5816977620124817,
      "learning_rate": 4.281675099179494e-05,
      "loss": 0.5349,
      "mean_token_accuracy": 0.8323511958122254,
      "num_tokens": 335221227.0,
      "step": 10005
    },
    {
      "epoch": 0.5975405921680993,
      "grad_norm": 0.5523040890693665,
      "learning_rate": 4.2808614959583115e-05,
      "loss": 0.599,
      "mean_token_accuracy": 0.8114696383476258,
      "num_tokens": 335388907.0,
      "step": 10010
    },
    {
      "epoch": 0.5978390639923591,
      "grad_norm": 0.5386897921562195,
      "learning_rate": 4.280047519867555e-05,
      "loss": 0.5731,
      "mean_token_accuracy": 0.8187999486923218,
      "num_tokens": 335556587.0,
      "step": 10015
    },
    {
      "epoch": 0.5981375358166189,
      "grad_norm": 0.5356436967849731,
      "learning_rate": 4.2792331711054843e-05,
      "loss": 0.54,
      "mean_token_accuracy": 0.8271024703979493,
      "num_tokens": 335724267.0,
      "step": 10020
    },
    {
      "epoch": 0.5984360076408787,
      "grad_norm": 0.5296046733856201,
      "learning_rate": 4.2784184498704484e-05,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8252773523330689,
      "num_tokens": 335891947.0,
      "step": 10025
    },
    {
      "epoch": 0.5987344794651385,
      "grad_norm": 0.5193471312522888,
      "learning_rate": 4.277603356360887e-05,
      "loss": 0.5421,
      "mean_token_accuracy": 0.8285756945610047,
      "num_tokens": 336059627.0,
      "step": 10030
    },
    {
      "epoch": 0.5990329512893983,
      "grad_norm": 0.5423997640609741,
      "learning_rate": 4.2767878907753325e-05,
      "loss": 0.5492,
      "mean_token_accuracy": 0.8246033787727356,
      "num_tokens": 336227307.0,
      "step": 10035
    },
    {
      "epoch": 0.5993314231136581,
      "grad_norm": 0.536270022392273,
      "learning_rate": 4.275972053312405e-05,
      "loss": 0.6006,
      "mean_token_accuracy": 0.8123165965080261,
      "num_tokens": 336394987.0,
      "step": 10040
    },
    {
      "epoch": 0.5996298949379179,
      "grad_norm": 0.5258265137672424,
      "learning_rate": 4.2751558441708184e-05,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8293331742286683,
      "num_tokens": 336562667.0,
      "step": 10045
    },
    {
      "epoch": 0.5999283667621776,
      "grad_norm": 0.5426004528999329,
      "learning_rate": 4.274339263549375e-05,
      "loss": 0.5603,
      "mean_token_accuracy": 0.8236192345619202,
      "num_tokens": 336730347.0,
      "step": 10050
    },
    {
      "epoch": 0.6002268385864374,
      "grad_norm": 0.509745180606842,
      "learning_rate": 4.2735223116469666e-05,
      "loss": 0.5711,
      "mean_token_accuracy": 0.8194977879524231,
      "num_tokens": 336898027.0,
      "step": 10055
    },
    {
      "epoch": 0.6005253104106972,
      "grad_norm": 0.5620666146278381,
      "learning_rate": 4.2727049886625795e-05,
      "loss": 0.5621,
      "mean_token_accuracy": 0.8229034900665283,
      "num_tokens": 337065707.0,
      "step": 10060
    },
    {
      "epoch": 0.600823782234957,
      "grad_norm": 0.5186175107955933,
      "learning_rate": 4.271887294795286e-05,
      "loss": 0.5713,
      "mean_token_accuracy": 0.8214362502098084,
      "num_tokens": 337233387.0,
      "step": 10065
    },
    {
      "epoch": 0.6011222540592168,
      "grad_norm": 0.5515191555023193,
      "learning_rate": 4.2710692302442515e-05,
      "loss": 0.5673,
      "mean_token_accuracy": 0.8205057978630066,
      "num_tokens": 337401067.0,
      "step": 10070
    },
    {
      "epoch": 0.6014207258834766,
      "grad_norm": 0.5776181817054749,
      "learning_rate": 4.270250795208729e-05,
      "loss": 0.5578,
      "mean_token_accuracy": 0.825462257862091,
      "num_tokens": 337568747.0,
      "step": 10075
    },
    {
      "epoch": 0.6017191977077364,
      "grad_norm": 0.5542211532592773,
      "learning_rate": 4.269431989888066e-05,
      "loss": 0.535,
      "mean_token_accuracy": 0.8300608396530151,
      "num_tokens": 337736427.0,
      "step": 10080
    },
    {
      "epoch": 0.6020176695319962,
      "grad_norm": 0.5703225135803223,
      "learning_rate": 4.268612814481696e-05,
      "loss": 0.5956,
      "mean_token_accuracy": 0.8141119003295898,
      "num_tokens": 337904107.0,
      "step": 10085
    },
    {
      "epoch": 0.602316141356256,
      "grad_norm": 0.5758171081542969,
      "learning_rate": 4.267793269189145e-05,
      "loss": 0.5867,
      "mean_token_accuracy": 0.8173765897750854,
      "num_tokens": 338067957.0,
      "step": 10090
    },
    {
      "epoch": 0.6026146131805158,
      "grad_norm": 0.5809513330459595,
      "learning_rate": 4.266973354210028e-05,
      "loss": 0.5476,
      "mean_token_accuracy": 0.8254025936126709,
      "num_tokens": 338235637.0,
      "step": 10095
    },
    {
      "epoch": 0.6029130850047756,
      "grad_norm": 0.6540355682373047,
      "learning_rate": 4.26615306974405e-05,
      "loss": 0.6,
      "mean_token_accuracy": 0.812793743610382,
      "num_tokens": 338403317.0,
      "step": 10100
    },
    {
      "epoch": 0.6032115568290354,
      "grad_norm": 0.5405868291854858,
      "learning_rate": 4.265332415991007e-05,
      "loss": 0.5713,
      "mean_token_accuracy": 0.8203745603561401,
      "num_tokens": 338570997.0,
      "step": 10105
    },
    {
      "epoch": 0.6035100286532952,
      "grad_norm": 0.5001184940338135,
      "learning_rate": 4.264511393150784e-05,
      "loss": 0.5794,
      "mean_token_accuracy": 0.8165871262550354,
      "num_tokens": 338738677.0,
      "step": 10110
    },
    {
      "epoch": 0.603808500477555,
      "grad_norm": 0.5327947735786438,
      "learning_rate": 4.263690001423357e-05,
      "loss": 0.5694,
      "mean_token_accuracy": 0.8189490675926209,
      "num_tokens": 338906357.0,
      "step": 10115
    },
    {
      "epoch": 0.6041069723018148,
      "grad_norm": 0.5442296862602234,
      "learning_rate": 4.26286824100879e-05,
      "loss": 0.6141,
      "mean_token_accuracy": 0.8060837388038635,
      "num_tokens": 339074037.0,
      "step": 10120
    },
    {
      "epoch": 0.6044054441260746,
      "grad_norm": 0.49894338846206665,
      "learning_rate": 4.262046112107239e-05,
      "loss": 0.5575,
      "mean_token_accuracy": 0.8232255816459656,
      "num_tokens": 339241717.0,
      "step": 10125
    },
    {
      "epoch": 0.6047039159503342,
      "grad_norm": 0.5164963603019714,
      "learning_rate": 4.261223614918949e-05,
      "loss": 0.538,
      "mean_token_accuracy": 0.8288500428199768,
      "num_tokens": 339409397.0,
      "step": 10130
    },
    {
      "epoch": 0.605002387774594,
      "grad_norm": 0.8744543194770813,
      "learning_rate": 4.260400749644252e-05,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8351723790168762,
      "num_tokens": 339577077.0,
      "step": 10135
    },
    {
      "epoch": 0.6053008595988538,
      "grad_norm": 0.5206270813941956,
      "learning_rate": 4.2595775164835743e-05,
      "loss": 0.5927,
      "mean_token_accuracy": 0.8134796619415283,
      "num_tokens": 339744757.0,
      "step": 10140
    },
    {
      "epoch": 0.6055993314231136,
      "grad_norm": 0.5476736426353455,
      "learning_rate": 4.2587539156374295e-05,
      "loss": 0.5633,
      "mean_token_accuracy": 0.8216390252113343,
      "num_tokens": 339912437.0,
      "step": 10145
    },
    {
      "epoch": 0.6058978032473734,
      "grad_norm": 0.5553728938102722,
      "learning_rate": 4.257929947306419e-05,
      "loss": 0.5881,
      "mean_token_accuracy": 0.8156865119934082,
      "num_tokens": 340080117.0,
      "step": 10150
    },
    {
      "epoch": 0.6061962750716332,
      "grad_norm": 0.5113295316696167,
      "learning_rate": 4.2571056116912376e-05,
      "loss": 0.5476,
      "mean_token_accuracy": 0.8250268578529358,
      "num_tokens": 340247797.0,
      "step": 10155
    },
    {
      "epoch": 0.606494746895893,
      "grad_norm": 0.5264196395874023,
      "learning_rate": 4.2562809089926665e-05,
      "loss": 0.567,
      "mean_token_accuracy": 0.819861626625061,
      "num_tokens": 340415477.0,
      "step": 10160
    },
    {
      "epoch": 0.6067932187201528,
      "grad_norm": 0.5544784069061279,
      "learning_rate": 4.255455839411577e-05,
      "loss": 0.6149,
      "mean_token_accuracy": 0.8072709560394287,
      "num_tokens": 340578191.0,
      "step": 10165
    },
    {
      "epoch": 0.6070916905444126,
      "grad_norm": 0.6234328150749207,
      "learning_rate": 4.254630403148929e-05,
      "loss": 0.5562,
      "mean_token_accuracy": 0.8248061537742615,
      "num_tokens": 340745871.0,
      "step": 10170
    },
    {
      "epoch": 0.6073901623686724,
      "grad_norm": 0.5575888752937317,
      "learning_rate": 4.2538046004057766e-05,
      "loss": 0.5947,
      "mean_token_accuracy": 0.8110819578170776,
      "num_tokens": 340913551.0,
      "step": 10175
    },
    {
      "epoch": 0.6076886341929322,
      "grad_norm": 0.5839712619781494,
      "learning_rate": 4.252978431383255e-05,
      "loss": 0.529,
      "mean_token_accuracy": 0.8309077858924866,
      "num_tokens": 341081231.0,
      "step": 10180
    },
    {
      "epoch": 0.607987106017192,
      "grad_norm": 0.5024023056030273,
      "learning_rate": 4.2521518962825955e-05,
      "loss": 0.5259,
      "mean_token_accuracy": 0.8329063534736634,
      "num_tokens": 341247330.0,
      "step": 10185
    },
    {
      "epoch": 0.6082855778414518,
      "grad_norm": 0.5004905462265015,
      "learning_rate": 4.251324995305114e-05,
      "loss": 0.558,
      "mean_token_accuracy": 0.8245437264442443,
      "num_tokens": 341415010.0,
      "step": 10190
    },
    {
      "epoch": 0.6085840496657116,
      "grad_norm": 0.5296902060508728,
      "learning_rate": 4.250497728652219e-05,
      "loss": 0.5686,
      "mean_token_accuracy": 0.8214004635810852,
      "num_tokens": 341582690.0,
      "step": 10195
    },
    {
      "epoch": 0.6088825214899714,
      "grad_norm": 0.6006523370742798,
      "learning_rate": 4.249670096525406e-05,
      "loss": 0.5812,
      "mean_token_accuracy": 0.8172730445861817,
      "num_tokens": 341750370.0,
      "step": 10200
    },
    {
      "epoch": 0.6091809933142311,
      "grad_norm": 0.623622477054596,
      "learning_rate": 4.2488420991262607e-05,
      "loss": 0.6192,
      "mean_token_accuracy": 0.8064117908477784,
      "num_tokens": 341918050.0,
      "step": 10205
    },
    {
      "epoch": 0.6094794651384909,
      "grad_norm": 0.5206360220909119,
      "learning_rate": 4.248013736656457e-05,
      "loss": 0.5798,
      "mean_token_accuracy": 0.8177382826805115,
      "num_tokens": 342085730.0,
      "step": 10210
    },
    {
      "epoch": 0.6097779369627507,
      "grad_norm": 0.5445969700813293,
      "learning_rate": 4.247185009317757e-05,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8214422106742859,
      "num_tokens": 342253410.0,
      "step": 10215
    },
    {
      "epoch": 0.6100764087870105,
      "grad_norm": 0.5614628195762634,
      "learning_rate": 4.246355917312013e-05,
      "loss": 0.6046,
      "mean_token_accuracy": 0.8115829586982727,
      "num_tokens": 342421090.0,
      "step": 10220
    },
    {
      "epoch": 0.6103748806112703,
      "grad_norm": 0.49875083565711975,
      "learning_rate": 4.245526460841165e-05,
      "loss": 0.5316,
      "mean_token_accuracy": 0.8295061349868774,
      "num_tokens": 342588770.0,
      "step": 10225
    },
    {
      "epoch": 0.6106733524355301,
      "grad_norm": 0.6407276391983032,
      "learning_rate": 4.2446966401072454e-05,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8199511051177979,
      "num_tokens": 342756450.0,
      "step": 10230
    },
    {
      "epoch": 0.6109718242597899,
      "grad_norm": 0.5701903700828552,
      "learning_rate": 4.2438664553123694e-05,
      "loss": 0.5697,
      "mean_token_accuracy": 0.8202970385551452,
      "num_tokens": 342924130.0,
      "step": 10235
    },
    {
      "epoch": 0.6112702960840497,
      "grad_norm": 0.5019197463989258,
      "learning_rate": 4.2430359066587436e-05,
      "loss": 0.6115,
      "mean_token_accuracy": 0.8076762557029724,
      "num_tokens": 343091810.0,
      "step": 10240
    },
    {
      "epoch": 0.6115687679083095,
      "grad_norm": 0.4857781231403351,
      "learning_rate": 4.242204994348665e-05,
      "loss": 0.506,
      "mean_token_accuracy": 0.8365740180015564,
      "num_tokens": 343259490.0,
      "step": 10245
    },
    {
      "epoch": 0.6118672397325693,
      "grad_norm": 0.5388410091400146,
      "learning_rate": 4.241373718584517e-05,
      "loss": 0.5497,
      "mean_token_accuracy": 0.8249015808105469,
      "num_tokens": 343427170.0,
      "step": 10250
    },
    {
      "epoch": 0.6121657115568291,
      "grad_norm": 0.5654188394546509,
      "learning_rate": 4.240542079568772e-05,
      "loss": 0.5509,
      "mean_token_accuracy": 0.8270547389984131,
      "num_tokens": 343594850.0,
      "step": 10255
    },
    {
      "epoch": 0.6124641833810889,
      "grad_norm": 0.5905010104179382,
      "learning_rate": 4.2397100775039925e-05,
      "loss": 0.6175,
      "mean_token_accuracy": 0.8060479521751404,
      "num_tokens": 343762530.0,
      "step": 10260
    },
    {
      "epoch": 0.6127626552053486,
      "grad_norm": 0.5358613729476929,
      "learning_rate": 4.238877712592826e-05,
      "loss": 0.5519,
      "mean_token_accuracy": 0.8270965099334717,
      "num_tokens": 343930210.0,
      "step": 10265
    },
    {
      "epoch": 0.6130611270296084,
      "grad_norm": 0.5065945386886597,
      "learning_rate": 4.238044985038011e-05,
      "loss": 0.518,
      "mean_token_accuracy": 0.83399738073349,
      "num_tokens": 344097890.0,
      "step": 10270
    },
    {
      "epoch": 0.6133595988538681,
      "grad_norm": 0.526066780090332,
      "learning_rate": 4.237211895042373e-05,
      "loss": 0.5203,
      "mean_token_accuracy": 0.833472490310669,
      "num_tokens": 344265570.0,
      "step": 10275
    },
    {
      "epoch": 0.6136580706781279,
      "grad_norm": 0.5284095406532288,
      "learning_rate": 4.2363784428088275e-05,
      "loss": 0.5978,
      "mean_token_accuracy": 0.8105749726295471,
      "num_tokens": 344433250.0,
      "step": 10280
    },
    {
      "epoch": 0.6139565425023877,
      "grad_norm": 0.5001017451286316,
      "learning_rate": 4.235544628540378e-05,
      "loss": 0.5417,
      "mean_token_accuracy": 0.8277048707008362,
      "num_tokens": 344600930.0,
      "step": 10285
    },
    {
      "epoch": 0.6142550143266475,
      "grad_norm": 0.6117889285087585,
      "learning_rate": 4.234710452440113e-05,
      "loss": 0.5615,
      "mean_token_accuracy": 0.8219014644622803,
      "num_tokens": 344768610.0,
      "step": 10290
    },
    {
      "epoch": 0.6145534861509073,
      "grad_norm": 0.5988826155662537,
      "learning_rate": 4.233875914711213e-05,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8210246920585632,
      "num_tokens": 344936290.0,
      "step": 10295
    },
    {
      "epoch": 0.6148519579751671,
      "grad_norm": 0.5304162502288818,
      "learning_rate": 4.233041015556945e-05,
      "loss": 0.5361,
      "mean_token_accuracy": 0.8294286012649537,
      "num_tokens": 345103970.0,
      "step": 10300
    },
    {
      "epoch": 0.6151504297994269,
      "grad_norm": 0.5469595789909363,
      "learning_rate": 4.2322057551806635e-05,
      "loss": 0.5844,
      "mean_token_accuracy": 0.8154956579208374,
      "num_tokens": 345271650.0,
      "step": 10305
    },
    {
      "epoch": 0.6154489016236867,
      "grad_norm": 0.5119556784629822,
      "learning_rate": 4.231370133785812e-05,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8262555241584778,
      "num_tokens": 345439330.0,
      "step": 10310
    },
    {
      "epoch": 0.6157473734479465,
      "grad_norm": 0.5373770594596863,
      "learning_rate": 4.2305341515759224e-05,
      "loss": 0.5816,
      "mean_token_accuracy": 0.8163604855537414,
      "num_tokens": 345607010.0,
      "step": 10315
    },
    {
      "epoch": 0.6160458452722063,
      "grad_norm": 0.5381370782852173,
      "learning_rate": 4.229697808754612e-05,
      "loss": 0.522,
      "mean_token_accuracy": 0.8334307551383973,
      "num_tokens": 345774690.0,
      "step": 10320
    },
    {
      "epoch": 0.6163443170964661,
      "grad_norm": 0.5179805159568787,
      "learning_rate": 4.228861105525587e-05,
      "loss": 0.5626,
      "mean_token_accuracy": 0.8212215304374695,
      "num_tokens": 345942370.0,
      "step": 10325
    },
    {
      "epoch": 0.6166427889207259,
      "grad_norm": 0.5689610838890076,
      "learning_rate": 4.228024042092645e-05,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8277108550071717,
      "num_tokens": 346110050.0,
      "step": 10330
    },
    {
      "epoch": 0.6169412607449857,
      "grad_norm": 0.5430788993835449,
      "learning_rate": 4.227186618659665e-05,
      "loss": 0.5552,
      "mean_token_accuracy": 0.8233269691467285,
      "num_tokens": 346277730.0,
      "step": 10335
    },
    {
      "epoch": 0.6172397325692455,
      "grad_norm": 0.56238853931427,
      "learning_rate": 4.226348835430619e-05,
      "loss": 0.5572,
      "mean_token_accuracy": 0.8235536217689514,
      "num_tokens": 346445410.0,
      "step": 10340
    },
    {
      "epoch": 0.6175382043935053,
      "grad_norm": 0.5491074323654175,
      "learning_rate": 4.225510692609564e-05,
      "loss": 0.5899,
      "mean_token_accuracy": 0.814523434638977,
      "num_tokens": 346613090.0,
      "step": 10345
    },
    {
      "epoch": 0.6178366762177651,
      "grad_norm": 0.5272854566574097,
      "learning_rate": 4.224672190400645e-05,
      "loss": 0.5534,
      "mean_token_accuracy": 0.8263449907302857,
      "num_tokens": 346780770.0,
      "step": 10350
    },
    {
      "epoch": 0.6181351480420249,
      "grad_norm": 0.4945986270904541,
      "learning_rate": 4.2238333290080936e-05,
      "loss": 0.569,
      "mean_token_accuracy": 0.82071453332901,
      "num_tokens": 346948450.0,
      "step": 10355
    },
    {
      "epoch": 0.6184336198662846,
      "grad_norm": 0.5312017202377319,
      "learning_rate": 4.222994108636232e-05,
      "loss": 0.5825,
      "mean_token_accuracy": 0.8145174860954285,
      "num_tokens": 347116130.0,
      "step": 10360
    },
    {
      "epoch": 0.6187320916905444,
      "grad_norm": 0.5052494406700134,
      "learning_rate": 4.222154529489467e-05,
      "loss": 0.554,
      "mean_token_accuracy": 0.8242872357368469,
      "num_tokens": 347283810.0,
      "step": 10365
    },
    {
      "epoch": 0.6190305635148042,
      "grad_norm": 0.5967806577682495,
      "learning_rate": 4.221314591772293e-05,
      "loss": 0.5797,
      "mean_token_accuracy": 0.818370521068573,
      "num_tokens": 347451490.0,
      "step": 10370
    },
    {
      "epoch": 0.619329035339064,
      "grad_norm": 0.5441157817840576,
      "learning_rate": 4.2204742956892925e-05,
      "loss": 0.5403,
      "mean_token_accuracy": 0.8269712567329407,
      "num_tokens": 347619170.0,
      "step": 10375
    },
    {
      "epoch": 0.6196275071633238,
      "grad_norm": 0.61036616563797,
      "learning_rate": 4.219633641445136e-05,
      "loss": 0.5407,
      "mean_token_accuracy": 0.8299654006958008,
      "num_tokens": 347786850.0,
      "step": 10380
    },
    {
      "epoch": 0.6199259789875836,
      "grad_norm": 0.5069438815116882,
      "learning_rate": 4.2187926292445803e-05,
      "loss": 0.5658,
      "mean_token_accuracy": 0.8200286269187927,
      "num_tokens": 347954530.0,
      "step": 10385
    },
    {
      "epoch": 0.6202244508118434,
      "grad_norm": 0.5739896297454834,
      "learning_rate": 4.217951259292468e-05,
      "loss": 0.5697,
      "mean_token_accuracy": 0.8203745722770691,
      "num_tokens": 348122210.0,
      "step": 10390
    },
    {
      "epoch": 0.6205229226361032,
      "grad_norm": 0.5449885725975037,
      "learning_rate": 4.217109531793731e-05,
      "loss": 0.5705,
      "mean_token_accuracy": 0.818972933292389,
      "num_tokens": 348289890.0,
      "step": 10395
    },
    {
      "epoch": 0.620821394460363,
      "grad_norm": 0.46956634521484375,
      "learning_rate": 4.2162674469533894e-05,
      "loss": 0.559,
      "mean_token_accuracy": 0.8203805446624756,
      "num_tokens": 348457570.0,
      "step": 10400
    },
    {
      "epoch": 0.6211198662846227,
      "grad_norm": 0.550395131111145,
      "learning_rate": 4.2154250049765456e-05,
      "loss": 0.553,
      "mean_token_accuracy": 0.8255338191986084,
      "num_tokens": 348625250.0,
      "step": 10405
    },
    {
      "epoch": 0.6214183381088825,
      "grad_norm": 0.5533021688461304,
      "learning_rate": 4.214582206068394e-05,
      "loss": 0.5911,
      "mean_token_accuracy": 0.8165095925331116,
      "num_tokens": 348792930.0,
      "step": 10410
    },
    {
      "epoch": 0.6217168099331423,
      "grad_norm": 0.5339484810829163,
      "learning_rate": 4.213739050434212e-05,
      "loss": 0.5581,
      "mean_token_accuracy": 0.8240128755569458,
      "num_tokens": 348960610.0,
      "step": 10415
    },
    {
      "epoch": 0.6220152817574021,
      "grad_norm": 0.5239465832710266,
      "learning_rate": 4.212895538279367e-05,
      "loss": 0.5512,
      "mean_token_accuracy": 0.8253310322761536,
      "num_tokens": 349128290.0,
      "step": 10420
    },
    {
      "epoch": 0.6223137535816619,
      "grad_norm": 0.5749861598014832,
      "learning_rate": 4.212051669809312e-05,
      "loss": 0.5316,
      "mean_token_accuracy": 0.8337230205535888,
      "num_tokens": 349295970.0,
      "step": 10425
    },
    {
      "epoch": 0.6226122254059216,
      "grad_norm": 0.5575063824653625,
      "learning_rate": 4.211207445229586e-05,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8276750564575195,
      "num_tokens": 349463650.0,
      "step": 10430
    },
    {
      "epoch": 0.6229106972301814,
      "grad_norm": 0.5194706320762634,
      "learning_rate": 4.210362864745817e-05,
      "loss": 0.543,
      "mean_token_accuracy": 0.8282297611236572,
      "num_tokens": 349631330.0,
      "step": 10435
    },
    {
      "epoch": 0.6232091690544412,
      "grad_norm": 0.5498590469360352,
      "learning_rate": 4.209517928563717e-05,
      "loss": 0.5916,
      "mean_token_accuracy": 0.8153167247772217,
      "num_tokens": 349799010.0,
      "step": 10440
    },
    {
      "epoch": 0.623507640878701,
      "grad_norm": 0.5149429440498352,
      "learning_rate": 4.208672636889085e-05,
      "loss": 0.5607,
      "mean_token_accuracy": 0.8220565438270568,
      "num_tokens": 349966690.0,
      "step": 10445
    },
    {
      "epoch": 0.6238061127029608,
      "grad_norm": 0.644813060760498,
      "learning_rate": 4.207826989927808e-05,
      "loss": 0.5805,
      "mean_token_accuracy": 0.8166110038757324,
      "num_tokens": 350134370.0,
      "step": 10450
    },
    {
      "epoch": 0.6241045845272206,
      "grad_norm": 0.5750823616981506,
      "learning_rate": 4.2069809878858604e-05,
      "loss": 0.5757,
      "mean_token_accuracy": 0.8192413210868835,
      "num_tokens": 350302050.0,
      "step": 10455
    },
    {
      "epoch": 0.6244030563514804,
      "grad_norm": 0.48424893617630005,
      "learning_rate": 4.2061346309692995e-05,
      "loss": 0.554,
      "mean_token_accuracy": 0.8266253113746643,
      "num_tokens": 350469730.0,
      "step": 10460
    },
    {
      "epoch": 0.6247015281757402,
      "grad_norm": 0.5818796753883362,
      "learning_rate": 4.205287919384272e-05,
      "loss": 0.5738,
      "mean_token_accuracy": 0.8195037603378296,
      "num_tokens": 350637410.0,
      "step": 10465
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.571145236492157,
      "learning_rate": 4.20444085333701e-05,
      "loss": 0.5706,
      "mean_token_accuracy": 0.8211141586303711,
      "num_tokens": 350805090.0,
      "step": 10470
    },
    {
      "epoch": 0.6252984718242598,
      "grad_norm": 0.5542051196098328,
      "learning_rate": 4.203593433033833e-05,
      "loss": 0.5609,
      "mean_token_accuracy": 0.8236907958984375,
      "num_tokens": 350972770.0,
      "step": 10475
    },
    {
      "epoch": 0.6255969436485196,
      "grad_norm": 0.4784558415412903,
      "learning_rate": 4.202745658681145e-05,
      "loss": 0.5789,
      "mean_token_accuracy": 0.8159131407737732,
      "num_tokens": 351140450.0,
      "step": 10480
    },
    {
      "epoch": 0.6258954154727794,
      "grad_norm": 0.5490285754203796,
      "learning_rate": 4.201897530485437e-05,
      "loss": 0.5333,
      "mean_token_accuracy": 0.8304783463478088,
      "num_tokens": 351308130.0,
      "step": 10485
    },
    {
      "epoch": 0.6261938872970392,
      "grad_norm": 0.5519647598266602,
      "learning_rate": 4.201049048653287e-05,
      "loss": 0.5832,
      "mean_token_accuracy": 0.8183048963546753,
      "num_tokens": 351475810.0,
      "step": 10490
    },
    {
      "epoch": 0.626492359121299,
      "grad_norm": 0.5074359178543091,
      "learning_rate": 4.200200213391358e-05,
      "loss": 0.5563,
      "mean_token_accuracy": 0.823583447933197,
      "num_tokens": 351643490.0,
      "step": 10495
    },
    {
      "epoch": 0.6267908309455588,
      "grad_norm": 0.4795893430709839,
      "learning_rate": 4.1993510249063985e-05,
      "loss": 0.5408,
      "mean_token_accuracy": 0.8303053736686706,
      "num_tokens": 351811170.0,
      "step": 10500
    },
    {
      "epoch": 0.6270893027698186,
      "grad_norm": 0.5287749767303467,
      "learning_rate": 4.1985014834052456e-05,
      "loss": 0.588,
      "mean_token_accuracy": 0.8165752053260803,
      "num_tokens": 351978850.0,
      "step": 10505
    },
    {
      "epoch": 0.6273877745940784,
      "grad_norm": 0.5833807587623596,
      "learning_rate": 4.197651589094819e-05,
      "loss": 0.5891,
      "mean_token_accuracy": 0.8154717803001403,
      "num_tokens": 352146530.0,
      "step": 10510
    },
    {
      "epoch": 0.6276862464183381,
      "grad_norm": 0.50035560131073,
      "learning_rate": 4.196801342182129e-05,
      "loss": 0.5339,
      "mean_token_accuracy": 0.8296791076660156,
      "num_tokens": 352314210.0,
      "step": 10515
    },
    {
      "epoch": 0.6279847182425979,
      "grad_norm": 0.5603047013282776,
      "learning_rate": 4.195950742874267e-05,
      "loss": 0.5628,
      "mean_token_accuracy": 0.8212453722953796,
      "num_tokens": 352481890.0,
      "step": 10520
    },
    {
      "epoch": 0.6282831900668577,
      "grad_norm": 0.5571617484092712,
      "learning_rate": 4.195099791378411e-05,
      "loss": 0.5333,
      "mean_token_accuracy": 0.8330788493156434,
      "num_tokens": 352649570.0,
      "step": 10525
    },
    {
      "epoch": 0.6285816618911175,
      "grad_norm": 0.5076308846473694,
      "learning_rate": 4.194248487901828e-05,
      "loss": 0.5507,
      "mean_token_accuracy": 0.8260773777961731,
      "num_tokens": 352810928.0,
      "step": 10530
    },
    {
      "epoch": 0.6288801337153773,
      "grad_norm": 0.6115785241127014,
      "learning_rate": 4.193396832651868e-05,
      "loss": 0.5769,
      "mean_token_accuracy": 0.8177800297737121,
      "num_tokens": 352978608.0,
      "step": 10535
    },
    {
      "epoch": 0.629178605539637,
      "grad_norm": 0.5612708926200867,
      "learning_rate": 4.192544825835968e-05,
      "loss": 0.5804,
      "mean_token_accuracy": 0.8171000719070435,
      "num_tokens": 353146288.0,
      "step": 10540
    },
    {
      "epoch": 0.6294770773638968,
      "grad_norm": 0.5526641607284546,
      "learning_rate": 4.191692467661649e-05,
      "loss": 0.5779,
      "mean_token_accuracy": 0.8166587114334106,
      "num_tokens": 353313968.0,
      "step": 10545
    },
    {
      "epoch": 0.6297755491881566,
      "grad_norm": 0.5095444321632385,
      "learning_rate": 4.190839758336518e-05,
      "loss": 0.5053,
      "mean_token_accuracy": 0.8365501642227173,
      "num_tokens": 353481648.0,
      "step": 10550
    },
    {
      "epoch": 0.6300740210124164,
      "grad_norm": 0.529994547367096,
      "learning_rate": 4.1899866980682693e-05,
      "loss": 0.5899,
      "mean_token_accuracy": 0.8148037672042847,
      "num_tokens": 353649328.0,
      "step": 10555
    },
    {
      "epoch": 0.6303724928366762,
      "grad_norm": 0.5498631000518799,
      "learning_rate": 4.189133287064682e-05,
      "loss": 0.5621,
      "mean_token_accuracy": 0.8221400499343872,
      "num_tokens": 353817008.0,
      "step": 10560
    },
    {
      "epoch": 0.630670964660936,
      "grad_norm": 0.5215694308280945,
      "learning_rate": 4.188279525533619e-05,
      "loss": 0.5442,
      "mean_token_accuracy": 0.8271501779556274,
      "num_tokens": 353984688.0,
      "step": 10565
    },
    {
      "epoch": 0.6309694364851958,
      "grad_norm": 0.5187599658966064,
      "learning_rate": 4.18742541368303e-05,
      "loss": 0.521,
      "mean_token_accuracy": 0.8338959813117981,
      "num_tokens": 354152368.0,
      "step": 10570
    },
    {
      "epoch": 0.6312679083094556,
      "grad_norm": 0.5384235978126526,
      "learning_rate": 4.186570951720949e-05,
      "loss": 0.5305,
      "mean_token_accuracy": 0.8320947170257569,
      "num_tokens": 354320048.0,
      "step": 10575
    },
    {
      "epoch": 0.6315663801337154,
      "grad_norm": 0.4864861071109772,
      "learning_rate": 4.185716139855497e-05,
      "loss": 0.5608,
      "mean_token_accuracy": 0.8226231694221496,
      "num_tokens": 354487728.0,
      "step": 10580
    },
    {
      "epoch": 0.6318648519579751,
      "grad_norm": 0.49161964654922485,
      "learning_rate": 4.184860978294879e-05,
      "loss": 0.52,
      "mean_token_accuracy": 0.8338065147399902,
      "num_tokens": 354655408.0,
      "step": 10585
    },
    {
      "epoch": 0.6321633237822349,
      "grad_norm": 0.5725266933441162,
      "learning_rate": 4.184005467247386e-05,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8218418240547181,
      "num_tokens": 354823088.0,
      "step": 10590
    },
    {
      "epoch": 0.6324617956064947,
      "grad_norm": 0.5377210378646851,
      "learning_rate": 4.183149606921392e-05,
      "loss": 0.6134,
      "mean_token_accuracy": 0.8093761205673218,
      "num_tokens": 354990768.0,
      "step": 10595
    },
    {
      "epoch": 0.6327602674307545,
      "grad_norm": 0.534552276134491,
      "learning_rate": 4.182293397525358e-05,
      "loss": 0.4982,
      "mean_token_accuracy": 0.8397769212722779,
      "num_tokens": 355158448.0,
      "step": 10600
    },
    {
      "epoch": 0.6330587392550143,
      "grad_norm": 0.4438951909542084,
      "learning_rate": 4.1814368392678304e-05,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8311761736869812,
      "num_tokens": 355326128.0,
      "step": 10605
    },
    {
      "epoch": 0.6333572110792741,
      "grad_norm": 0.5355693697929382,
      "learning_rate": 4.180579932357439e-05,
      "loss": 0.5512,
      "mean_token_accuracy": 0.8268519639968872,
      "num_tokens": 355493808.0,
      "step": 10610
    },
    {
      "epoch": 0.6336556829035339,
      "grad_norm": 0.5087995529174805,
      "learning_rate": 4.1797226770029e-05,
      "loss": 0.5515,
      "mean_token_accuracy": 0.8248658061027527,
      "num_tokens": 355661488.0,
      "step": 10615
    },
    {
      "epoch": 0.6339541547277937,
      "grad_norm": 0.6685255169868469,
      "learning_rate": 4.178865073413012e-05,
      "loss": 0.6056,
      "mean_token_accuracy": 0.8093522667884827,
      "num_tokens": 355829168.0,
      "step": 10620
    },
    {
      "epoch": 0.6342526265520535,
      "grad_norm": 0.5254787802696228,
      "learning_rate": 4.178007121796663e-05,
      "loss": 0.5419,
      "mean_token_accuracy": 0.8273947238922119,
      "num_tokens": 355996848.0,
      "step": 10625
    },
    {
      "epoch": 0.6345510983763133,
      "grad_norm": 0.5092074275016785,
      "learning_rate": 4.1771488223628216e-05,
      "loss": 0.563,
      "mean_token_accuracy": 0.8221400499343872,
      "num_tokens": 356164528.0,
      "step": 10630
    },
    {
      "epoch": 0.6348495702005731,
      "grad_norm": 0.6070569753646851,
      "learning_rate": 4.1762901753205416e-05,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8147679328918457,
      "num_tokens": 356327278.0,
      "step": 10635
    },
    {
      "epoch": 0.6351480420248329,
      "grad_norm": 0.5248810052871704,
      "learning_rate": 4.175431180878964e-05,
      "loss": 0.602,
      "mean_token_accuracy": 0.8114099979400635,
      "num_tokens": 356494958.0,
      "step": 10640
    },
    {
      "epoch": 0.6354465138490927,
      "grad_norm": 0.5092383623123169,
      "learning_rate": 4.174571839247311e-05,
      "loss": 0.5323,
      "mean_token_accuracy": 0.8320529580116272,
      "num_tokens": 356662638.0,
      "step": 10645
    },
    {
      "epoch": 0.6357449856733525,
      "grad_norm": 0.5364642143249512,
      "learning_rate": 4.173712150634892e-05,
      "loss": 0.547,
      "mean_token_accuracy": 0.8257246851921082,
      "num_tokens": 356830318.0,
      "step": 10650
    },
    {
      "epoch": 0.6360434574976123,
      "grad_norm": 0.5233772993087769,
      "learning_rate": 4.1728521152511e-05,
      "loss": 0.5457,
      "mean_token_accuracy": 0.826690936088562,
      "num_tokens": 356997998.0,
      "step": 10655
    },
    {
      "epoch": 0.6363419293218721,
      "grad_norm": 0.5759458541870117,
      "learning_rate": 4.171991733305413e-05,
      "loss": 0.5373,
      "mean_token_accuracy": 0.8284742951393127,
      "num_tokens": 357165678.0,
      "step": 10660
    },
    {
      "epoch": 0.6366404011461319,
      "grad_norm": 0.5486500859260559,
      "learning_rate": 4.171131005007392e-05,
      "loss": 0.5473,
      "mean_token_accuracy": 0.8266968846321106,
      "num_tokens": 357333358.0,
      "step": 10665
    },
    {
      "epoch": 0.6369388729703916,
      "grad_norm": 0.5107816457748413,
      "learning_rate": 4.1702699305666835e-05,
      "loss": 0.5545,
      "mean_token_accuracy": 0.8247822880744934,
      "num_tokens": 357501038.0,
      "step": 10670
    },
    {
      "epoch": 0.6372373447946514,
      "grad_norm": 0.5290648341178894,
      "learning_rate": 4.1694085101930187e-05,
      "loss": 0.5761,
      "mean_token_accuracy": 0.8195693612098693,
      "num_tokens": 357668718.0,
      "step": 10675
    },
    {
      "epoch": 0.6375358166189111,
      "grad_norm": 0.5462967753410339,
      "learning_rate": 4.168546744096211e-05,
      "loss": 0.5499,
      "mean_token_accuracy": 0.8227126240730286,
      "num_tokens": 357836398.0,
      "step": 10680
    },
    {
      "epoch": 0.6378342884431709,
      "grad_norm": 0.47816041111946106,
      "learning_rate": 4.167684632486162e-05,
      "loss": 0.5247,
      "mean_token_accuracy": 0.8340212345123291,
      "num_tokens": 358004078.0,
      "step": 10685
    },
    {
      "epoch": 0.6381327602674307,
      "grad_norm": 0.6931360960006714,
      "learning_rate": 4.166822175572852e-05,
      "loss": 0.5992,
      "mean_token_accuracy": 0.8122330904006958,
      "num_tokens": 358171758.0,
      "step": 10690
    },
    {
      "epoch": 0.6384312320916905,
      "grad_norm": 0.5212947130203247,
      "learning_rate": 4.16595937356635e-05,
      "loss": 0.503,
      "mean_token_accuracy": 0.8401049733161926,
      "num_tokens": 358339438.0,
      "step": 10695
    },
    {
      "epoch": 0.6387297039159503,
      "grad_norm": 0.49074992537498474,
      "learning_rate": 4.1650962266768074e-05,
      "loss": 0.5247,
      "mean_token_accuracy": 0.8335381150245667,
      "num_tokens": 358507118.0,
      "step": 10700
    },
    {
      "epoch": 0.6390281757402101,
      "grad_norm": 0.5606208443641663,
      "learning_rate": 4.164232735114458e-05,
      "loss": 0.5611,
      "mean_token_accuracy": 0.8220565319061279,
      "num_tokens": 358674798.0,
      "step": 10705
    },
    {
      "epoch": 0.6393266475644699,
      "grad_norm": 0.5256921648979187,
      "learning_rate": 4.1633688990896236e-05,
      "loss": 0.5841,
      "mean_token_accuracy": 0.8178038954734802,
      "num_tokens": 358842478.0,
      "step": 10710
    },
    {
      "epoch": 0.6396251193887297,
      "grad_norm": 0.4700057804584503,
      "learning_rate": 4.162504718812705e-05,
      "loss": 0.5507,
      "mean_token_accuracy": 0.8245437145233154,
      "num_tokens": 359010158.0,
      "step": 10715
    },
    {
      "epoch": 0.6399235912129895,
      "grad_norm": 0.5408313870429993,
      "learning_rate": 4.16164019449419e-05,
      "loss": 0.5965,
      "mean_token_accuracy": 0.8115710258483887,
      "num_tokens": 359177838.0,
      "step": 10720
    },
    {
      "epoch": 0.6402220630372493,
      "grad_norm": 0.5639321208000183,
      "learning_rate": 4.1607753263446515e-05,
      "loss": 0.5579,
      "mean_token_accuracy": 0.823833954334259,
      "num_tokens": 359345518.0,
      "step": 10725
    },
    {
      "epoch": 0.6405205348615091,
      "grad_norm": 0.5468165278434753,
      "learning_rate": 4.159910114574741e-05,
      "loss": 0.5111,
      "mean_token_accuracy": 0.8363891243934631,
      "num_tokens": 359513198.0,
      "step": 10730
    },
    {
      "epoch": 0.6408190066857689,
      "grad_norm": 0.5450208783149719,
      "learning_rate": 4.1590445593951975e-05,
      "loss": 0.5808,
      "mean_token_accuracy": 0.8147978067398072,
      "num_tokens": 359680878.0,
      "step": 10735
    },
    {
      "epoch": 0.6411174785100286,
      "grad_norm": 0.46378010511398315,
      "learning_rate": 4.158178661016845e-05,
      "loss": 0.5623,
      "mean_token_accuracy": 0.8223726630210877,
      "num_tokens": 359848558.0,
      "step": 10740
    },
    {
      "epoch": 0.6414159503342884,
      "grad_norm": 0.48744845390319824,
      "learning_rate": 4.1573124196505866e-05,
      "loss": 0.4943,
      "mean_token_accuracy": 0.8421329021453857,
      "num_tokens": 360016238.0,
      "step": 10745
    },
    {
      "epoch": 0.6417144221585482,
      "grad_norm": 0.5765150189399719,
      "learning_rate": 4.156445835507413e-05,
      "loss": 0.5382,
      "mean_token_accuracy": 0.8284921884536743,
      "num_tokens": 360183918.0,
      "step": 10750
    },
    {
      "epoch": 0.642012893982808,
      "grad_norm": 0.5338862538337708,
      "learning_rate": 4.155578908798396e-05,
      "loss": 0.5858,
      "mean_token_accuracy": 0.8138196349143982,
      "num_tokens": 360351598.0,
      "step": 10755
    },
    {
      "epoch": 0.6423113658070678,
      "grad_norm": 0.4562791585922241,
      "learning_rate": 4.154711639734692e-05,
      "loss": 0.5297,
      "mean_token_accuracy": 0.8298878788948059,
      "num_tokens": 360519278.0,
      "step": 10760
    },
    {
      "epoch": 0.6426098376313276,
      "grad_norm": 0.5389201641082764,
      "learning_rate": 4.1538440285275396e-05,
      "loss": 0.548,
      "mean_token_accuracy": 0.8253071784973145,
      "num_tokens": 360686958.0,
      "step": 10765
    },
    {
      "epoch": 0.6429083094555874,
      "grad_norm": 0.5104871392250061,
      "learning_rate": 4.152976075388263e-05,
      "loss": 0.5234,
      "mean_token_accuracy": 0.8333770751953125,
      "num_tokens": 360854638.0,
      "step": 10770
    },
    {
      "epoch": 0.6432067812798472,
      "grad_norm": 0.5955930948257446,
      "learning_rate": 4.1521077805282654e-05,
      "loss": 0.6227,
      "mean_token_accuracy": 0.8061076045036316,
      "num_tokens": 361022318.0,
      "step": 10775
    },
    {
      "epoch": 0.643505253104107,
      "grad_norm": 0.4999445378780365,
      "learning_rate": 4.1512391441590396e-05,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8257962584495544,
      "num_tokens": 361189998.0,
      "step": 10780
    },
    {
      "epoch": 0.6438037249283668,
      "grad_norm": 0.4865644872188568,
      "learning_rate": 4.150370166492156e-05,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8348980069160461,
      "num_tokens": 361357678.0,
      "step": 10785
    },
    {
      "epoch": 0.6441021967526266,
      "grad_norm": 0.5511912703514099,
      "learning_rate": 4.14950084773927e-05,
      "loss": 0.5518,
      "mean_token_accuracy": 0.8262495517730712,
      "num_tokens": 361525358.0,
      "step": 10790
    },
    {
      "epoch": 0.6444006685768864,
      "grad_norm": 0.5373526215553284,
      "learning_rate": 4.148631188112121e-05,
      "loss": 0.5492,
      "mean_token_accuracy": 0.8248956203460693,
      "num_tokens": 361693038.0,
      "step": 10795
    },
    {
      "epoch": 0.6446991404011462,
      "grad_norm": 0.6302651166915894,
      "learning_rate": 4.147761187822531e-05,
      "loss": 0.5821,
      "mean_token_accuracy": 0.8165573358535767,
      "num_tokens": 361860718.0,
      "step": 10800
    },
    {
      "epoch": 0.644997612225406,
      "grad_norm": 0.4753219187259674,
      "learning_rate": 4.146890847082403e-05,
      "loss": 0.5445,
      "mean_token_accuracy": 0.827669095993042,
      "num_tokens": 362028398.0,
      "step": 10805
    },
    {
      "epoch": 0.6452960840496658,
      "grad_norm": 0.591619074344635,
      "learning_rate": 4.1460201661037254e-05,
      "loss": 0.5795,
      "mean_token_accuracy": 0.8173506021499634,
      "num_tokens": 362196078.0,
      "step": 10810
    },
    {
      "epoch": 0.6455945558739254,
      "grad_norm": 0.5333344340324402,
      "learning_rate": 4.1451491450985694e-05,
      "loss": 0.55,
      "mean_token_accuracy": 0.827555775642395,
      "num_tokens": 362363758.0,
      "step": 10815
    },
    {
      "epoch": 0.6458930276981852,
      "grad_norm": 0.5531866550445557,
      "learning_rate": 4.1442777842790884e-05,
      "loss": 0.5108,
      "mean_token_accuracy": 0.836365258693695,
      "num_tokens": 362531438.0,
      "step": 10820
    },
    {
      "epoch": 0.646191499522445,
      "grad_norm": 0.5012946128845215,
      "learning_rate": 4.143406083857516e-05,
      "loss": 0.5928,
      "mean_token_accuracy": 0.8163425922393799,
      "num_tokens": 362699118.0,
      "step": 10825
    },
    {
      "epoch": 0.6464899713467048,
      "grad_norm": 0.5352234840393066,
      "learning_rate": 4.142534044046173e-05,
      "loss": 0.5593,
      "mean_token_accuracy": 0.8233329296112061,
      "num_tokens": 362866798.0,
      "step": 10830
    },
    {
      "epoch": 0.6467884431709646,
      "grad_norm": 0.563372790813446,
      "learning_rate": 4.14166166505746e-05,
      "loss": 0.5755,
      "mean_token_accuracy": 0.8181259751319885,
      "num_tokens": 363034478.0,
      "step": 10835
    },
    {
      "epoch": 0.6470869149952244,
      "grad_norm": 0.5415008664131165,
      "learning_rate": 4.1407889471038624e-05,
      "loss": 0.5324,
      "mean_token_accuracy": 0.8302338004112244,
      "num_tokens": 363202158.0,
      "step": 10840
    },
    {
      "epoch": 0.6473853868194842,
      "grad_norm": 0.48782598972320557,
      "learning_rate": 4.139915890397946e-05,
      "loss": 0.5191,
      "mean_token_accuracy": 0.8339914083480835,
      "num_tokens": 363369838.0,
      "step": 10845
    },
    {
      "epoch": 0.647683858643744,
      "grad_norm": 0.6319615840911865,
      "learning_rate": 4.1390424951523584e-05,
      "loss": 0.5812,
      "mean_token_accuracy": 0.8150304079055786,
      "num_tokens": 363537518.0,
      "step": 10850
    },
    {
      "epoch": 0.6479823304680038,
      "grad_norm": 0.5104392766952515,
      "learning_rate": 4.138168761579833e-05,
      "loss": 0.5683,
      "mean_token_accuracy": 0.8196170926094055,
      "num_tokens": 363705198.0,
      "step": 10855
    },
    {
      "epoch": 0.6482808022922636,
      "grad_norm": 0.5604403018951416,
      "learning_rate": 4.137294689893182e-05,
      "loss": 0.5432,
      "mean_token_accuracy": 0.827191948890686,
      "num_tokens": 363872878.0,
      "step": 10860
    },
    {
      "epoch": 0.6485792741165234,
      "grad_norm": 0.5327582955360413,
      "learning_rate": 4.136420280305304e-05,
      "loss": 0.5867,
      "mean_token_accuracy": 0.8157044053077698,
      "num_tokens": 364040558.0,
      "step": 10865
    },
    {
      "epoch": 0.6488777459407832,
      "grad_norm": 0.47843194007873535,
      "learning_rate": 4.135545533029176e-05,
      "loss": 0.5149,
      "mean_token_accuracy": 0.8352916717529297,
      "num_tokens": 364208238.0,
      "step": 10870
    },
    {
      "epoch": 0.649176217765043,
      "grad_norm": 0.5262420177459717,
      "learning_rate": 4.134670448277859e-05,
      "loss": 0.5578,
      "mean_token_accuracy": 0.8243289947509765,
      "num_tokens": 364375918.0,
      "step": 10875
    },
    {
      "epoch": 0.6494746895893028,
      "grad_norm": 0.4862832725048065,
      "learning_rate": 4.133795026264497e-05,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8269652843475341,
      "num_tokens": 364543598.0,
      "step": 10880
    },
    {
      "epoch": 0.6497731614135626,
      "grad_norm": 0.5758223533630371,
      "learning_rate": 4.132919267202313e-05,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8212990522384643,
      "num_tokens": 364711278.0,
      "step": 10885
    },
    {
      "epoch": 0.6500716332378224,
      "grad_norm": 0.5943029522895813,
      "learning_rate": 4.132043171304616e-05,
      "loss": 0.5612,
      "mean_token_accuracy": 0.8238474488258362,
      "num_tokens": 364877455.0,
      "step": 10890
    },
    {
      "epoch": 0.6503701050620821,
      "grad_norm": 0.5892188549041748,
      "learning_rate": 4.131166738784796e-05,
      "loss": 0.6368,
      "mean_token_accuracy": 0.8013360381126404,
      "num_tokens": 365045135.0,
      "step": 10895
    },
    {
      "epoch": 0.6506685768863419,
      "grad_norm": 0.5199084281921387,
      "learning_rate": 4.130289969856322e-05,
      "loss": 0.5624,
      "mean_token_accuracy": 0.8210306525230407,
      "num_tokens": 365212815.0,
      "step": 10900
    },
    {
      "epoch": 0.6509670487106017,
      "grad_norm": 0.534847617149353,
      "learning_rate": 4.12941286473275e-05,
      "loss": 0.5597,
      "mean_token_accuracy": 0.8198974013328553,
      "num_tokens": 365380495.0,
      "step": 10905
    },
    {
      "epoch": 0.6512655205348615,
      "grad_norm": 0.5123786926269531,
      "learning_rate": 4.1285354236277134e-05,
      "loss": 0.6039,
      "mean_token_accuracy": 0.8097638010978698,
      "num_tokens": 365548175.0,
      "step": 10910
    },
    {
      "epoch": 0.6515639923591213,
      "grad_norm": 0.5008794665336609,
      "learning_rate": 4.127657646754929e-05,
      "loss": 0.4995,
      "mean_token_accuracy": 0.8405940532684326,
      "num_tokens": 365715855.0,
      "step": 10915
    },
    {
      "epoch": 0.6518624641833811,
      "grad_norm": 0.524743914604187,
      "learning_rate": 4.126779534328196e-05,
      "loss": 0.5536,
      "mean_token_accuracy": 0.8236729025840759,
      "num_tokens": 365883535.0,
      "step": 10920
    },
    {
      "epoch": 0.6521609360076409,
      "grad_norm": 0.5517908930778503,
      "learning_rate": 4.125901086561396e-05,
      "loss": 0.5624,
      "mean_token_accuracy": 0.8214481711387634,
      "num_tokens": 366051215.0,
      "step": 10925
    },
    {
      "epoch": 0.6524594078319007,
      "grad_norm": 0.5227869749069214,
      "learning_rate": 4.125022303668489e-05,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8237325549125671,
      "num_tokens": 366218895.0,
      "step": 10930
    },
    {
      "epoch": 0.6527578796561605,
      "grad_norm": 0.5112661123275757,
      "learning_rate": 4.1241431858635204e-05,
      "loss": 0.5888,
      "mean_token_accuracy": 0.8149707794189454,
      "num_tokens": 366386575.0,
      "step": 10935
    },
    {
      "epoch": 0.6530563514804203,
      "grad_norm": 0.5125031471252441,
      "learning_rate": 4.123263733360615e-05,
      "loss": 0.5368,
      "mean_token_accuracy": 0.8308600783348083,
      "num_tokens": 366554255.0,
      "step": 10940
    },
    {
      "epoch": 0.6533548233046801,
      "grad_norm": 0.5317060947418213,
      "learning_rate": 4.122383946373981e-05,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8230526208877563,
      "num_tokens": 366721935.0,
      "step": 10945
    },
    {
      "epoch": 0.6536532951289399,
      "grad_norm": 0.5169796347618103,
      "learning_rate": 4.121503825117904e-05,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8330848097801209,
      "num_tokens": 366889615.0,
      "step": 10950
    },
    {
      "epoch": 0.6539517669531996,
      "grad_norm": 0.6890859603881836,
      "learning_rate": 4.120623369806756e-05,
      "loss": 0.5837,
      "mean_token_accuracy": 0.816730284690857,
      "num_tokens": 367057295.0,
      "step": 10955
    },
    {
      "epoch": 0.6542502387774594,
      "grad_norm": 0.5225417017936707,
      "learning_rate": 4.1197425806549876e-05,
      "loss": 0.5546,
      "mean_token_accuracy": 0.8240904211997986,
      "num_tokens": 367224975.0,
      "step": 10960
    },
    {
      "epoch": 0.6545487106017192,
      "grad_norm": 0.5400025248527527,
      "learning_rate": 4.11886145787713e-05,
      "loss": 0.5688,
      "mean_token_accuracy": 0.8195097208023071,
      "num_tokens": 367392655.0,
      "step": 10965
    },
    {
      "epoch": 0.654847182425979,
      "grad_norm": 0.5433284640312195,
      "learning_rate": 4.117980001687799e-05,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8300966262817383,
      "num_tokens": 367560335.0,
      "step": 10970
    },
    {
      "epoch": 0.6551456542502387,
      "grad_norm": 0.5311228632926941,
      "learning_rate": 4.117098212301689e-05,
      "loss": 0.5347,
      "mean_token_accuracy": 0.828533935546875,
      "num_tokens": 367728015.0,
      "step": 10975
    },
    {
      "epoch": 0.6554441260744985,
      "grad_norm": 0.5552988052368164,
      "learning_rate": 4.116216089933575e-05,
      "loss": 0.591,
      "mean_token_accuracy": 0.8147262215614319,
      "num_tokens": 367895695.0,
      "step": 10980
    },
    {
      "epoch": 0.6557425978987583,
      "grad_norm": 0.494136244058609,
      "learning_rate": 4.1153336347983154e-05,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8228140354156495,
      "num_tokens": 368063375.0,
      "step": 10985
    },
    {
      "epoch": 0.6560410697230181,
      "grad_norm": 0.5637604594230652,
      "learning_rate": 4.114450847110848e-05,
      "loss": 0.5518,
      "mean_token_accuracy": 0.8252773404121398,
      "num_tokens": 368231055.0,
      "step": 10990
    },
    {
      "epoch": 0.6563395415472779,
      "grad_norm": 0.5620095729827881,
      "learning_rate": 4.11356772708619e-05,
      "loss": 0.5966,
      "mean_token_accuracy": 0.8131516218185425,
      "num_tokens": 368398735.0,
      "step": 10995
    },
    {
      "epoch": 0.6566380133715377,
      "grad_norm": 0.5555197596549988,
      "learning_rate": 4.112684274939445e-05,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8292198538780212,
      "num_tokens": 368566415.0,
      "step": 11000
    },
    {
      "epoch": 0.6569364851957975,
      "grad_norm": 0.5233317613601685,
      "learning_rate": 4.111800490885793e-05,
      "loss": 0.5794,
      "mean_token_accuracy": 0.8216628909111023,
      "num_tokens": 368734095.0,
      "step": 11005
    },
    {
      "epoch": 0.6572349570200573,
      "grad_norm": 0.5886181592941284,
      "learning_rate": 4.110916375140494e-05,
      "loss": 0.553,
      "mean_token_accuracy": 0.8228378772735596,
      "num_tokens": 368901775.0,
      "step": 11010
    },
    {
      "epoch": 0.6575334288443171,
      "grad_norm": 0.6114769577980042,
      "learning_rate": 4.110031927918894e-05,
      "loss": 0.5488,
      "mean_token_accuracy": 0.8270488023757935,
      "num_tokens": 369069455.0,
      "step": 11015
    },
    {
      "epoch": 0.6578319006685769,
      "grad_norm": 0.5818997025489807,
      "learning_rate": 4.109147149436414e-05,
      "loss": 0.5491,
      "mean_token_accuracy": 0.8271859645843506,
      "num_tokens": 369237135.0,
      "step": 11020
    },
    {
      "epoch": 0.6581303724928367,
      "grad_norm": 0.4692915380001068,
      "learning_rate": 4.108262039908558e-05,
      "loss": 0.5042,
      "mean_token_accuracy": 0.8374388813972473,
      "num_tokens": 369404815.0,
      "step": 11025
    },
    {
      "epoch": 0.6584288443170965,
      "grad_norm": 0.61583012342453,
      "learning_rate": 4.107376599550912e-05,
      "loss": 0.5627,
      "mean_token_accuracy": 0.8220305800437927,
      "num_tokens": 369566416.0,
      "step": 11030
    },
    {
      "epoch": 0.6587273161413563,
      "grad_norm": 0.5520955920219421,
      "learning_rate": 4.106490828579141e-05,
      "loss": 0.55,
      "mean_token_accuracy": 0.824323034286499,
      "num_tokens": 369734096.0,
      "step": 11035
    },
    {
      "epoch": 0.6590257879656161,
      "grad_norm": 0.5377001166343689,
      "learning_rate": 4.10560472720899e-05,
      "loss": 0.5508,
      "mean_token_accuracy": 0.8276750564575195,
      "num_tokens": 369901776.0,
      "step": 11040
    },
    {
      "epoch": 0.6593242597898759,
      "grad_norm": 0.5589100122451782,
      "learning_rate": 4.104718295656287e-05,
      "loss": 0.5195,
      "mean_token_accuracy": 0.8350590467453003,
      "num_tokens": 370069456.0,
      "step": 11045
    },
    {
      "epoch": 0.6596227316141356,
      "grad_norm": 0.5749718546867371,
      "learning_rate": 4.103831534136938e-05,
      "loss": 0.5403,
      "mean_token_accuracy": 0.8269652843475341,
      "num_tokens": 370237136.0,
      "step": 11050
    },
    {
      "epoch": 0.6599212034383954,
      "grad_norm": 0.5292873382568359,
      "learning_rate": 4.102944442866929e-05,
      "loss": 0.5078,
      "mean_token_accuracy": 0.8370392560958863,
      "num_tokens": 370404816.0,
      "step": 11055
    },
    {
      "epoch": 0.6602196752626552,
      "grad_norm": 0.5220714211463928,
      "learning_rate": 4.102057022062329e-05,
      "loss": 0.5398,
      "mean_token_accuracy": 0.8292496681213379,
      "num_tokens": 370572496.0,
      "step": 11060
    },
    {
      "epoch": 0.660518147086915,
      "grad_norm": 0.47866925597190857,
      "learning_rate": 4.101169271939286e-05,
      "loss": 0.5166,
      "mean_token_accuracy": 0.8339794754981995,
      "num_tokens": 370740176.0,
      "step": 11065
    },
    {
      "epoch": 0.6608166189111748,
      "grad_norm": 0.4852518141269684,
      "learning_rate": 4.100281192714026e-05,
      "loss": 0.5341,
      "mean_token_accuracy": 0.8323750495910645,
      "num_tokens": 370907856.0,
      "step": 11070
    },
    {
      "epoch": 0.6611150907354346,
      "grad_norm": 0.6355628371238708,
      "learning_rate": 4.0993927846028584e-05,
      "loss": 0.5367,
      "mean_token_accuracy": 0.8273529648780823,
      "num_tokens": 371075536.0,
      "step": 11075
    },
    {
      "epoch": 0.6614135625596944,
      "grad_norm": 0.47717875242233276,
      "learning_rate": 4.0985040478221715e-05,
      "loss": 0.5267,
      "mean_token_accuracy": 0.8321245431900024,
      "num_tokens": 371243216.0,
      "step": 11080
    },
    {
      "epoch": 0.6617120343839542,
      "grad_norm": 0.5030310750007629,
      "learning_rate": 4.097614982588433e-05,
      "loss": 0.582,
      "mean_token_accuracy": 0.8168614983558655,
      "num_tokens": 371410896.0,
      "step": 11085
    },
    {
      "epoch": 0.6620105062082139,
      "grad_norm": 0.5493571162223816,
      "learning_rate": 4.096725589118191e-05,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8234760880470275,
      "num_tokens": 371578576.0,
      "step": 11090
    },
    {
      "epoch": 0.6623089780324737,
      "grad_norm": 0.6388803124427795,
      "learning_rate": 4.0958358676280743e-05,
      "loss": 0.5603,
      "mean_token_accuracy": 0.822569477558136,
      "num_tokens": 371746256.0,
      "step": 11095
    },
    {
      "epoch": 0.6626074498567335,
      "grad_norm": 0.5451918840408325,
      "learning_rate": 4.094945818334791e-05,
      "loss": 0.5391,
      "mean_token_accuracy": 0.827806270122528,
      "num_tokens": 371913936.0,
      "step": 11100
    },
    {
      "epoch": 0.6629059216809933,
      "grad_norm": 0.5010148882865906,
      "learning_rate": 4.0940554414551294e-05,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8250984191894531,
      "num_tokens": 372081616.0,
      "step": 11105
    },
    {
      "epoch": 0.6632043935052531,
      "grad_norm": 0.5547599792480469,
      "learning_rate": 4.0931647372059556e-05,
      "loss": 0.5618,
      "mean_token_accuracy": 0.8234104633331298,
      "num_tokens": 372249296.0,
      "step": 11110
    },
    {
      "epoch": 0.6635028653295129,
      "grad_norm": 0.5289527773857117,
      "learning_rate": 4.092273705804218e-05,
      "loss": 0.5502,
      "mean_token_accuracy": 0.8256829261779786,
      "num_tokens": 372416976.0,
      "step": 11115
    },
    {
      "epoch": 0.6638013371537727,
      "grad_norm": 0.5747686624526978,
      "learning_rate": 4.091382347466945e-05,
      "loss": 0.5508,
      "mean_token_accuracy": 0.8247942209243775,
      "num_tokens": 372584656.0,
      "step": 11120
    },
    {
      "epoch": 0.6640998089780324,
      "grad_norm": 0.5888289213180542,
      "learning_rate": 4.090490662411241e-05,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8328581571578979,
      "num_tokens": 372752336.0,
      "step": 11125
    },
    {
      "epoch": 0.6643982808022922,
      "grad_norm": 0.5585777759552002,
      "learning_rate": 4.089598650854294e-05,
      "loss": 0.5967,
      "mean_token_accuracy": 0.8128116369247437,
      "num_tokens": 372920016.0,
      "step": 11130
    },
    {
      "epoch": 0.664696752626552,
      "grad_norm": 0.5006683468818665,
      "learning_rate": 4.088706313013369e-05,
      "loss": 0.5254,
      "mean_token_accuracy": 0.8320708513259888,
      "num_tokens": 373087696.0,
      "step": 11135
    },
    {
      "epoch": 0.6649952244508118,
      "grad_norm": 0.5076429843902588,
      "learning_rate": 4.0878136491058125e-05,
      "loss": 0.4872,
      "mean_token_accuracy": 0.8438029408454895,
      "num_tokens": 373255376.0,
      "step": 11140
    },
    {
      "epoch": 0.6652936962750716,
      "grad_norm": 0.5112064480781555,
      "learning_rate": 4.0869206593490475e-05,
      "loss": 0.5505,
      "mean_token_accuracy": 0.8246809005737304,
      "num_tokens": 373423056.0,
      "step": 11145
    },
    {
      "epoch": 0.6655921680993314,
      "grad_norm": 0.5198042392730713,
      "learning_rate": 4.086027343960581e-05,
      "loss": 0.5783,
      "mean_token_accuracy": 0.8192055344581604,
      "num_tokens": 373590736.0,
      "step": 11150
    },
    {
      "epoch": 0.6658906399235912,
      "grad_norm": 0.48146018385887146,
      "learning_rate": 4.085133703157993e-05,
      "loss": 0.5239,
      "mean_token_accuracy": 0.8302696108818054,
      "num_tokens": 373758416.0,
      "step": 11155
    },
    {
      "epoch": 0.666189111747851,
      "grad_norm": 0.49789804220199585,
      "learning_rate": 4.0842397371589484e-05,
      "loss": 0.5291,
      "mean_token_accuracy": 0.8303948521614075,
      "num_tokens": 373926096.0,
      "step": 11160
    },
    {
      "epoch": 0.6664875835721108,
      "grad_norm": 0.5661185383796692,
      "learning_rate": 4.083345446181188e-05,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8365084171295166,
      "num_tokens": 374093776.0,
      "step": 11165
    },
    {
      "epoch": 0.6667860553963706,
      "grad_norm": 0.5354464054107666,
      "learning_rate": 4.082450830442534e-05,
      "loss": 0.5018,
      "mean_token_accuracy": 0.8397828936576843,
      "num_tokens": 374261456.0,
      "step": 11170
    },
    {
      "epoch": 0.6670845272206304,
      "grad_norm": 0.556145191192627,
      "learning_rate": 4.081555890160886e-05,
      "loss": 0.5456,
      "mean_token_accuracy": 0.8248717546463012,
      "num_tokens": 374429136.0,
      "step": 11175
    },
    {
      "epoch": 0.6673829990448902,
      "grad_norm": 0.5707510113716125,
      "learning_rate": 4.080660625554223e-05,
      "loss": 0.5323,
      "mean_token_accuracy": 0.8308541059494019,
      "num_tokens": 374596816.0,
      "step": 11180
    },
    {
      "epoch": 0.66768147086915,
      "grad_norm": 0.549965500831604,
      "learning_rate": 4.0797650368406044e-05,
      "loss": 0.5618,
      "mean_token_accuracy": 0.820213520526886,
      "num_tokens": 374764496.0,
      "step": 11185
    },
    {
      "epoch": 0.6679799426934098,
      "grad_norm": 0.5679465532302856,
      "learning_rate": 4.078869124238165e-05,
      "loss": 0.5411,
      "mean_token_accuracy": 0.828647255897522,
      "num_tokens": 374932176.0,
      "step": 11190
    },
    {
      "epoch": 0.6682784145176696,
      "grad_norm": 0.511679470539093,
      "learning_rate": 4.077972887965123e-05,
      "loss": 0.5272,
      "mean_token_accuracy": 0.8318680644035339,
      "num_tokens": 375099856.0,
      "step": 11195
    },
    {
      "epoch": 0.6685768863419294,
      "grad_norm": 0.4986500144004822,
      "learning_rate": 4.077076328239773e-05,
      "loss": 0.5589,
      "mean_token_accuracy": 0.8212096095085144,
      "num_tokens": 375267536.0,
      "step": 11200
    },
    {
      "epoch": 0.6688753581661891,
      "grad_norm": 0.6039528846740723,
      "learning_rate": 4.076179445280488e-05,
      "loss": 0.5612,
      "mean_token_accuracy": 0.8201720952987671,
      "num_tokens": 375433898.0,
      "step": 11205
    },
    {
      "epoch": 0.6691738299904489,
      "grad_norm": 0.535342276096344,
      "learning_rate": 4.07528223930572e-05,
      "loss": 0.5632,
      "mean_token_accuracy": 0.8205833315849305,
      "num_tokens": 375601578.0,
      "step": 11210
    },
    {
      "epoch": 0.6694723018147087,
      "grad_norm": 0.540589451789856,
      "learning_rate": 4.074384710534001e-05,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8301443338394165,
      "num_tokens": 375769258.0,
      "step": 11215
    },
    {
      "epoch": 0.6697707736389685,
      "grad_norm": 0.510288417339325,
      "learning_rate": 4.0734868591839404e-05,
      "loss": 0.5401,
      "mean_token_accuracy": 0.8273291230201721,
      "num_tokens": 375936938.0,
      "step": 11220
    },
    {
      "epoch": 0.6700692454632283,
      "grad_norm": 0.5015286207199097,
      "learning_rate": 4.0725886854742266e-05,
      "loss": 0.5084,
      "mean_token_accuracy": 0.8369080305099488,
      "num_tokens": 376104618.0,
      "step": 11225
    },
    {
      "epoch": 0.670367717287488,
      "grad_norm": 0.5854654908180237,
      "learning_rate": 4.0716901896236264e-05,
      "loss": 0.542,
      "mean_token_accuracy": 0.827794349193573,
      "num_tokens": 376272298.0,
      "step": 11230
    },
    {
      "epoch": 0.6706661891117478,
      "grad_norm": 0.6088428497314453,
      "learning_rate": 4.070791371850985e-05,
      "loss": 0.5794,
      "mean_token_accuracy": 0.8176786422729492,
      "num_tokens": 376439978.0,
      "step": 11235
    },
    {
      "epoch": 0.6709646609360076,
      "grad_norm": 0.5472153425216675,
      "learning_rate": 4.069892232375226e-05,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8329714894294739,
      "num_tokens": 376607658.0,
      "step": 11240
    },
    {
      "epoch": 0.6712631327602674,
      "grad_norm": 0.5092966556549072,
      "learning_rate": 4.0689927714153514e-05,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8273112297058105,
      "num_tokens": 376775338.0,
      "step": 11245
    },
    {
      "epoch": 0.6715616045845272,
      "grad_norm": 0.5294577479362488,
      "learning_rate": 4.068092989190441e-05,
      "loss": 0.5336,
      "mean_token_accuracy": 0.831003212928772,
      "num_tokens": 376943018.0,
      "step": 11250
    },
    {
      "epoch": 0.671860076408787,
      "grad_norm": 0.5168683528900146,
      "learning_rate": 4.067192885919654e-05,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8299057602882385,
      "num_tokens": 377110698.0,
      "step": 11255
    },
    {
      "epoch": 0.6721585482330468,
      "grad_norm": 0.5447584986686707,
      "learning_rate": 4.0662924618222276e-05,
      "loss": 0.5544,
      "mean_token_accuracy": 0.8227782368659973,
      "num_tokens": 377278378.0,
      "step": 11260
    },
    {
      "epoch": 0.6724570200573066,
      "grad_norm": 0.5891069173812866,
      "learning_rate": 4.0653917171174754e-05,
      "loss": 0.6083,
      "mean_token_accuracy": 0.8079506158828735,
      "num_tokens": 377446058.0,
      "step": 11265
    },
    {
      "epoch": 0.6727554918815664,
      "grad_norm": 0.5154567956924438,
      "learning_rate": 4.064490652024791e-05,
      "loss": 0.555,
      "mean_token_accuracy": 0.8246033549308777,
      "num_tokens": 377613738.0,
      "step": 11270
    },
    {
      "epoch": 0.6730539637058262,
      "grad_norm": 0.5120232701301575,
      "learning_rate": 4.063589266763646e-05,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8392043352127075,
      "num_tokens": 377781418.0,
      "step": 11275
    },
    {
      "epoch": 0.673352435530086,
      "grad_norm": 0.5150737762451172,
      "learning_rate": 4.062687561553589e-05,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8282416820526123,
      "num_tokens": 377949098.0,
      "step": 11280
    },
    {
      "epoch": 0.6736509073543457,
      "grad_norm": 0.6296200156211853,
      "learning_rate": 4.061785536614248e-05,
      "loss": 0.5853,
      "mean_token_accuracy": 0.8143564343452454,
      "num_tokens": 378116778.0,
      "step": 11285
    },
    {
      "epoch": 0.6739493791786055,
      "grad_norm": 0.5804410576820374,
      "learning_rate": 4.060883192165326e-05,
      "loss": 0.5638,
      "mean_token_accuracy": 0.8228199720382691,
      "num_tokens": 378284458.0,
      "step": 11290
    },
    {
      "epoch": 0.6742478510028653,
      "grad_norm": 0.4958733320236206,
      "learning_rate": 4.059980528426606e-05,
      "loss": 0.5386,
      "mean_token_accuracy": 0.8267982840538025,
      "num_tokens": 378452138.0,
      "step": 11295
    },
    {
      "epoch": 0.6745463228271251,
      "grad_norm": 0.5601612329483032,
      "learning_rate": 4.059077545617948e-05,
      "loss": 0.5519,
      "mean_token_accuracy": 0.825187873840332,
      "num_tokens": 378619818.0,
      "step": 11300
    },
    {
      "epoch": 0.6748447946513849,
      "grad_norm": 0.5521968603134155,
      "learning_rate": 4.058174243959292e-05,
      "loss": 0.552,
      "mean_token_accuracy": 0.8240665555000305,
      "num_tokens": 378787498.0,
      "step": 11305
    },
    {
      "epoch": 0.6751432664756447,
      "grad_norm": 0.5210601687431335,
      "learning_rate": 4.057270623670653e-05,
      "loss": 0.5915,
      "mean_token_accuracy": 0.8142132878303527,
      "num_tokens": 378955178.0,
      "step": 11310
    },
    {
      "epoch": 0.6754417382999045,
      "grad_norm": 0.4831472635269165,
      "learning_rate": 4.0563666849721246e-05,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8285160541534424,
      "num_tokens": 379122858.0,
      "step": 11315
    },
    {
      "epoch": 0.6757402101241643,
      "grad_norm": 0.5644519925117493,
      "learning_rate": 4.0554624280838766e-05,
      "loss": 0.5395,
      "mean_token_accuracy": 0.8277525901794434,
      "num_tokens": 379290538.0,
      "step": 11320
    },
    {
      "epoch": 0.6760386819484241,
      "grad_norm": 0.5638044476509094,
      "learning_rate": 4.054557853226158e-05,
      "loss": 0.5692,
      "mean_token_accuracy": 0.8201180934906006,
      "num_tokens": 379458218.0,
      "step": 11325
    },
    {
      "epoch": 0.6763371537726839,
      "grad_norm": 0.4986534118652344,
      "learning_rate": 4.0536529606192945e-05,
      "loss": 0.503,
      "mean_token_accuracy": 0.8384468674659729,
      "num_tokens": 379625898.0,
      "step": 11330
    },
    {
      "epoch": 0.6766356255969437,
      "grad_norm": 0.5392152070999146,
      "learning_rate": 4.05274775048369e-05,
      "loss": 0.5476,
      "mean_token_accuracy": 0.8245914459228516,
      "num_tokens": 379793578.0,
      "step": 11335
    },
    {
      "epoch": 0.6769340974212035,
      "grad_norm": 0.5276528596878052,
      "learning_rate": 4.0518422230398256e-05,
      "loss": 0.5765,
      "mean_token_accuracy": 0.8172611355781555,
      "num_tokens": 379961258.0,
      "step": 11340
    },
    {
      "epoch": 0.6772325692454633,
      "grad_norm": 0.6615886092185974,
      "learning_rate": 4.0509363785082585e-05,
      "loss": 0.5776,
      "mean_token_accuracy": 0.8180245637893677,
      "num_tokens": 380128938.0,
      "step": 11345
    },
    {
      "epoch": 0.6775310410697231,
      "grad_norm": 0.6085119247436523,
      "learning_rate": 4.050030217109623e-05,
      "loss": 0.5224,
      "mean_token_accuracy": 0.8313193321228027,
      "num_tokens": 380296618.0,
      "step": 11350
    },
    {
      "epoch": 0.6778295128939829,
      "grad_norm": 0.5373436808586121,
      "learning_rate": 4.049123739064633e-05,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8414588928222656,
      "num_tokens": 380464298.0,
      "step": 11355
    },
    {
      "epoch": 0.6781279847182426,
      "grad_norm": 0.5334434509277344,
      "learning_rate": 4.048216944594077e-05,
      "loss": 0.5479,
      "mean_token_accuracy": 0.8263449788093566,
      "num_tokens": 380631978.0,
      "step": 11360
    },
    {
      "epoch": 0.6784264565425024,
      "grad_norm": 0.530987560749054,
      "learning_rate": 4.047309833918822e-05,
      "loss": 0.5911,
      "mean_token_accuracy": 0.8143027663230896,
      "num_tokens": 380799658.0,
      "step": 11365
    },
    {
      "epoch": 0.6787249283667621,
      "grad_norm": 0.5713772177696228,
      "learning_rate": 4.046402407259812e-05,
      "loss": 0.5414,
      "mean_token_accuracy": 0.8291661739349365,
      "num_tokens": 380967338.0,
      "step": 11370
    },
    {
      "epoch": 0.6790234001910219,
      "grad_norm": 0.5532450079917908,
      "learning_rate": 4.045494664838066e-05,
      "loss": 0.5429,
      "mean_token_accuracy": 0.8269891500473022,
      "num_tokens": 381135018.0,
      "step": 11375
    },
    {
      "epoch": 0.6793218720152817,
      "grad_norm": 0.5362998843193054,
      "learning_rate": 4.044586606874682e-05,
      "loss": 0.5108,
      "mean_token_accuracy": 0.8373195767402649,
      "num_tokens": 381302698.0,
      "step": 11380
    },
    {
      "epoch": 0.6796203438395415,
      "grad_norm": 0.4848956763744354,
      "learning_rate": 4.043678233590835e-05,
      "loss": 0.5415,
      "mean_token_accuracy": 0.8279255628585815,
      "num_tokens": 381470378.0,
      "step": 11385
    },
    {
      "epoch": 0.6799188156638013,
      "grad_norm": 0.49829065799713135,
      "learning_rate": 4.042769545207775e-05,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8315101981163024,
      "num_tokens": 381638058.0,
      "step": 11390
    },
    {
      "epoch": 0.6802172874880611,
      "grad_norm": 0.4904751479625702,
      "learning_rate": 4.041860541946831e-05,
      "loss": 0.5361,
      "mean_token_accuracy": 0.8296194791793823,
      "num_tokens": 381805738.0,
      "step": 11395
    },
    {
      "epoch": 0.6805157593123209,
      "grad_norm": 0.5338941812515259,
      "learning_rate": 4.040951224029407e-05,
      "loss": 0.5671,
      "mean_token_accuracy": 0.8181259751319885,
      "num_tokens": 381973418.0,
      "step": 11400
    },
    {
      "epoch": 0.6808142311365807,
      "grad_norm": 0.5032311081886292,
      "learning_rate": 4.040041591676985e-05,
      "loss": 0.5246,
      "mean_token_accuracy": 0.831754732131958,
      "num_tokens": 382141098.0,
      "step": 11405
    },
    {
      "epoch": 0.6811127029608405,
      "grad_norm": 0.6189849376678467,
      "learning_rate": 4.0391316451111206e-05,
      "loss": 0.5464,
      "mean_token_accuracy": 0.8272933244705201,
      "num_tokens": 382308778.0,
      "step": 11410
    },
    {
      "epoch": 0.6814111747851003,
      "grad_norm": 0.5293188095092773,
      "learning_rate": 4.03822138455345e-05,
      "loss": 0.5415,
      "mean_token_accuracy": 0.8288500547409058,
      "num_tokens": 382476458.0,
      "step": 11415
    },
    {
      "epoch": 0.6817096466093601,
      "grad_norm": 0.46318134665489197,
      "learning_rate": 4.037310810225684e-05,
      "loss": 0.5277,
      "mean_token_accuracy": 0.8333233952522278,
      "num_tokens": 382644138.0,
      "step": 11420
    },
    {
      "epoch": 0.6820081184336199,
      "grad_norm": 0.46766602993011475,
      "learning_rate": 4.0363999223496086e-05,
      "loss": 0.5148,
      "mean_token_accuracy": 0.8358284711837769,
      "num_tokens": 382811818.0,
      "step": 11425
    },
    {
      "epoch": 0.6823065902578797,
      "grad_norm": 0.5994117856025696,
      "learning_rate": 4.035488721147087e-05,
      "loss": 0.6112,
      "mean_token_accuracy": 0.8090047240257263,
      "num_tokens": 382978660.0,
      "step": 11430
    },
    {
      "epoch": 0.6826050620821394,
      "grad_norm": 0.670329749584198,
      "learning_rate": 4.0345772068400613e-05,
      "loss": 0.5322,
      "mean_token_accuracy": 0.8300548791885376,
      "num_tokens": 383146340.0,
      "step": 11435
    },
    {
      "epoch": 0.6829035339063992,
      "grad_norm": 0.45337140560150146,
      "learning_rate": 4.0336653796505454e-05,
      "loss": 0.5051,
      "mean_token_accuracy": 0.839311707019806,
      "num_tokens": 383314020.0,
      "step": 11440
    },
    {
      "epoch": 0.683202005730659,
      "grad_norm": 0.6317613124847412,
      "learning_rate": 4.032753239800633e-05,
      "loss": 0.5602,
      "mean_token_accuracy": 0.8210187315940857,
      "num_tokens": 383481700.0,
      "step": 11445
    },
    {
      "epoch": 0.6835004775549188,
      "grad_norm": 3.444387912750244,
      "learning_rate": 4.031840787512492e-05,
      "loss": 0.5405,
      "mean_token_accuracy": 0.8279613614082336,
      "num_tokens": 383649380.0,
      "step": 11450
    },
    {
      "epoch": 0.6837989493791786,
      "grad_norm": 0.49650055170059204,
      "learning_rate": 4.0309280230083675e-05,
      "loss": 0.5693,
      "mean_token_accuracy": 0.8220326781272889,
      "num_tokens": 383817060.0,
      "step": 11455
    },
    {
      "epoch": 0.6840974212034384,
      "grad_norm": 0.49125024676322937,
      "learning_rate": 4.030014946510578e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8342896342277527,
      "num_tokens": 383984740.0,
      "step": 11460
    },
    {
      "epoch": 0.6843958930276982,
      "grad_norm": 0.5134462118148804,
      "learning_rate": 4.029101558241523e-05,
      "loss": 0.4904,
      "mean_token_accuracy": 0.842305850982666,
      "num_tokens": 384152420.0,
      "step": 11465
    },
    {
      "epoch": 0.684694364851958,
      "grad_norm": 0.5404857397079468,
      "learning_rate": 4.028187858423675e-05,
      "loss": 0.5382,
      "mean_token_accuracy": 0.830752718448639,
      "num_tokens": 384320100.0,
      "step": 11470
    },
    {
      "epoch": 0.6849928366762178,
      "grad_norm": 0.5220535397529602,
      "learning_rate": 4.027273847279581e-05,
      "loss": 0.551,
      "mean_token_accuracy": 0.8270487904548645,
      "num_tokens": 384487780.0,
      "step": 11475
    },
    {
      "epoch": 0.6852913085004776,
      "grad_norm": 0.5012786984443665,
      "learning_rate": 4.026359525031865e-05,
      "loss": 0.5389,
      "mean_token_accuracy": 0.8275796294212341,
      "num_tokens": 384655460.0,
      "step": 11480
    },
    {
      "epoch": 0.6855897803247374,
      "grad_norm": 0.5386480093002319,
      "learning_rate": 4.025444891903228e-05,
      "loss": 0.593,
      "mean_token_accuracy": 0.8125670909881592,
      "num_tokens": 384823140.0,
      "step": 11485
    },
    {
      "epoch": 0.6858882521489972,
      "grad_norm": 0.5205801725387573,
      "learning_rate": 4.024529948116446e-05,
      "loss": 0.5183,
      "mean_token_accuracy": 0.8351306200027466,
      "num_tokens": 384990820.0,
      "step": 11490
    },
    {
      "epoch": 0.686186723973257,
      "grad_norm": 0.5129269361495972,
      "learning_rate": 4.02361469389437e-05,
      "loss": 0.5559,
      "mean_token_accuracy": 0.8248896479606629,
      "num_tokens": 385158500.0,
      "step": 11495
    },
    {
      "epoch": 0.6864851957975168,
      "grad_norm": 0.5198088884353638,
      "learning_rate": 4.022699129459927e-05,
      "loss": 0.5297,
      "mean_token_accuracy": 0.8319456100463867,
      "num_tokens": 385326180.0,
      "step": 11500
    },
    {
      "epoch": 0.6867836676217765,
      "grad_norm": 0.5159403681755066,
      "learning_rate": 4.021783255036119e-05,
      "loss": 0.5388,
      "mean_token_accuracy": 0.8272396564483643,
      "num_tokens": 385493860.0,
      "step": 11505
    },
    {
      "epoch": 0.6870821394460362,
      "grad_norm": 0.5392984747886658,
      "learning_rate": 4.020867070846025e-05,
      "loss": 0.5173,
      "mean_token_accuracy": 0.8346176862716674,
      "num_tokens": 385661540.0,
      "step": 11510
    },
    {
      "epoch": 0.687380611270296,
      "grad_norm": 0.5313904285430908,
      "learning_rate": 4.0199505771127994e-05,
      "loss": 0.5205,
      "mean_token_accuracy": 0.8346057534217834,
      "num_tokens": 385829220.0,
      "step": 11515
    },
    {
      "epoch": 0.6876790830945558,
      "grad_norm": 0.5097572207450867,
      "learning_rate": 4.019033774059669e-05,
      "loss": 0.5435,
      "mean_token_accuracy": 0.8272277235984802,
      "num_tokens": 385996900.0,
      "step": 11520
    },
    {
      "epoch": 0.6879775549188156,
      "grad_norm": 0.5073562860488892,
      "learning_rate": 4.01811666190994e-05,
      "loss": 0.4907,
      "mean_token_accuracy": 0.843045449256897,
      "num_tokens": 386164580.0,
      "step": 11525
    },
    {
      "epoch": 0.6882760267430754,
      "grad_norm": 0.4953153431415558,
      "learning_rate": 4.017199240886991e-05,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8365501642227173,
      "num_tokens": 386332260.0,
      "step": 11530
    },
    {
      "epoch": 0.6885744985673352,
      "grad_norm": 0.5245486497879028,
      "learning_rate": 4.0162815112142767e-05,
      "loss": 0.511,
      "mean_token_accuracy": 0.8372300863265991,
      "num_tokens": 386499940.0,
      "step": 11535
    },
    {
      "epoch": 0.688872970391595,
      "grad_norm": 0.6272040009498596,
      "learning_rate": 4.015363473115328e-05,
      "loss": 0.5482,
      "mean_token_accuracy": 0.826303231716156,
      "num_tokens": 386667620.0,
      "step": 11540
    },
    {
      "epoch": 0.6891714422158548,
      "grad_norm": 0.5627403855323792,
      "learning_rate": 4.014445126813749e-05,
      "loss": 0.5539,
      "mean_token_accuracy": 0.8240188479423523,
      "num_tokens": 386835300.0,
      "step": 11545
    },
    {
      "epoch": 0.6894699140401146,
      "grad_norm": 0.5106717348098755,
      "learning_rate": 4.0135264725332205e-05,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8307348132133484,
      "num_tokens": 387002980.0,
      "step": 11550
    },
    {
      "epoch": 0.6897683858643744,
      "grad_norm": 0.4849369525909424,
      "learning_rate": 4.0126075104974975e-05,
      "loss": 0.5139,
      "mean_token_accuracy": 0.8371466040611267,
      "num_tokens": 387170660.0,
      "step": 11555
    },
    {
      "epoch": 0.6900668576886342,
      "grad_norm": 0.5245375633239746,
      "learning_rate": 4.011688240930411e-05,
      "loss": 0.5589,
      "mean_token_accuracy": 0.8229213953018188,
      "num_tokens": 387338340.0,
      "step": 11560
    },
    {
      "epoch": 0.690365329512894,
      "grad_norm": 0.5323561429977417,
      "learning_rate": 4.010768664055865e-05,
      "loss": 0.5137,
      "mean_token_accuracy": 0.8351425528526306,
      "num_tokens": 387506020.0,
      "step": 11565
    },
    {
      "epoch": 0.6906638013371538,
      "grad_norm": 0.5747860074043274,
      "learning_rate": 4.009848780097839e-05,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8349517107009887,
      "num_tokens": 387673700.0,
      "step": 11570
    },
    {
      "epoch": 0.6909622731614136,
      "grad_norm": 0.5796414613723755,
      "learning_rate": 4.008928589280389e-05,
      "loss": 0.5634,
      "mean_token_accuracy": 0.8215495467185974,
      "num_tokens": 387841380.0,
      "step": 11575
    },
    {
      "epoch": 0.6912607449856734,
      "grad_norm": 0.5231610536575317,
      "learning_rate": 4.008008091827644e-05,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8329392671585083,
      "num_tokens": 388001766.0,
      "step": 11580
    },
    {
      "epoch": 0.6915592168099332,
      "grad_norm": 0.49011021852493286,
      "learning_rate": 4.0070872879638084e-05,
      "loss": 0.5026,
      "mean_token_accuracy": 0.8392520546913147,
      "num_tokens": 388169446.0,
      "step": 11585
    },
    {
      "epoch": 0.691857688634193,
      "grad_norm": 0.5687540173530579,
      "learning_rate": 4.0061661779131604e-05,
      "loss": 0.5524,
      "mean_token_accuracy": 0.8256352066993713,
      "num_tokens": 388337126.0,
      "step": 11590
    },
    {
      "epoch": 0.6921561604584527,
      "grad_norm": 0.5706913471221924,
      "learning_rate": 4.0052447619000536e-05,
      "loss": 0.5793,
      "mean_token_accuracy": 0.8191113352775574,
      "num_tokens": 388504676.0,
      "step": 11595
    },
    {
      "epoch": 0.6924546322827125,
      "grad_norm": 0.4936061203479767,
      "learning_rate": 4.004323040148914e-05,
      "loss": 0.5478,
      "mean_token_accuracy": 0.8254204988479614,
      "num_tokens": 388672356.0,
      "step": 11600
    },
    {
      "epoch": 0.6927531041069723,
      "grad_norm": 0.5241875052452087,
      "learning_rate": 4.0034010128842484e-05,
      "loss": 0.5732,
      "mean_token_accuracy": 0.8186508536338806,
      "num_tokens": 388840036.0,
      "step": 11605
    },
    {
      "epoch": 0.6930515759312321,
      "grad_norm": 0.5459261536598206,
      "learning_rate": 4.00247868033063e-05,
      "loss": 0.551,
      "mean_token_accuracy": 0.8260706067085266,
      "num_tokens": 389007716.0,
      "step": 11610
    },
    {
      "epoch": 0.6933500477554919,
      "grad_norm": 0.5209773182868958,
      "learning_rate": 4.001556042712711e-05,
      "loss": 0.5897,
      "mean_token_accuracy": 0.8160682201385498,
      "num_tokens": 389175396.0,
      "step": 11615
    },
    {
      "epoch": 0.6936485195797517,
      "grad_norm": 0.4841661751270294,
      "learning_rate": 4.000633100255217e-05,
      "loss": 0.4998,
      "mean_token_accuracy": 0.8390492677688599,
      "num_tokens": 389343076.0,
      "step": 11620
    },
    {
      "epoch": 0.6939469914040115,
      "grad_norm": 0.5271602272987366,
      "learning_rate": 3.9997098531829475e-05,
      "loss": 0.5545,
      "mean_token_accuracy": 0.8236132621765136,
      "num_tokens": 389510756.0,
      "step": 11625
    },
    {
      "epoch": 0.6942454632282713,
      "grad_norm": 0.45626330375671387,
      "learning_rate": 3.9987863017207755e-05,
      "loss": 0.5747,
      "mean_token_accuracy": 0.8168734312057495,
      "num_tokens": 389678436.0,
      "step": 11630
    },
    {
      "epoch": 0.6945439350525311,
      "grad_norm": 0.5206072926521301,
      "learning_rate": 3.99786244609365e-05,
      "loss": 0.5613,
      "mean_token_accuracy": 0.8222772240638733,
      "num_tokens": 389846116.0,
      "step": 11635
    },
    {
      "epoch": 0.6948424068767909,
      "grad_norm": 0.5700442790985107,
      "learning_rate": 3.9969382865265936e-05,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8291244149208069,
      "num_tokens": 390013796.0,
      "step": 11640
    },
    {
      "epoch": 0.6951408787010506,
      "grad_norm": 0.5254628658294678,
      "learning_rate": 3.9960138232447004e-05,
      "loss": 0.5192,
      "mean_token_accuracy": 0.8340987682342529,
      "num_tokens": 390181476.0,
      "step": 11645
    },
    {
      "epoch": 0.6954393505253104,
      "grad_norm": 0.5238580107688904,
      "learning_rate": 3.995089056473143e-05,
      "loss": 0.5276,
      "mean_token_accuracy": 0.831742811203003,
      "num_tokens": 390349156.0,
      "step": 11650
    },
    {
      "epoch": 0.6957378223495702,
      "grad_norm": 0.5199779272079468,
      "learning_rate": 3.994163986437163e-05,
      "loss": 0.5507,
      "mean_token_accuracy": 0.8270249485969543,
      "num_tokens": 390516836.0,
      "step": 11655
    },
    {
      "epoch": 0.69603629417383,
      "grad_norm": 0.7571831941604614,
      "learning_rate": 3.9932386133620795e-05,
      "loss": 0.5396,
      "mean_token_accuracy": 0.8308421850204468,
      "num_tokens": 390684516.0,
      "step": 11660
    },
    {
      "epoch": 0.6963347659980897,
      "grad_norm": 0.5050497651100159,
      "learning_rate": 3.9923129374732854e-05,
      "loss": 0.5254,
      "mean_token_accuracy": 0.8310747861862182,
      "num_tokens": 390852196.0,
      "step": 11665
    },
    {
      "epoch": 0.6966332378223495,
      "grad_norm": 0.4855859577655792,
      "learning_rate": 3.9913869589962445e-05,
      "loss": 0.5368,
      "mean_token_accuracy": 0.8278778433799744,
      "num_tokens": 391019876.0,
      "step": 11670
    },
    {
      "epoch": 0.6969317096466093,
      "grad_norm": 0.5321979522705078,
      "learning_rate": 3.9904606781564965e-05,
      "loss": 0.5205,
      "mean_token_accuracy": 0.832142424583435,
      "num_tokens": 391187556.0,
      "step": 11675
    },
    {
      "epoch": 0.6972301814708691,
      "grad_norm": 0.536324143409729,
      "learning_rate": 3.989534095179654e-05,
      "loss": 0.5885,
      "mean_token_accuracy": 0.8154181122779847,
      "num_tokens": 391355236.0,
      "step": 11680
    },
    {
      "epoch": 0.6975286532951289,
      "grad_norm": 0.5377629399299622,
      "learning_rate": 3.9886072102914016e-05,
      "loss": 0.5614,
      "mean_token_accuracy": 0.8204342246055603,
      "num_tokens": 391522916.0,
      "step": 11685
    },
    {
      "epoch": 0.6978271251193887,
      "grad_norm": 0.4855128228664398,
      "learning_rate": 3.987680023717503e-05,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8315936923027039,
      "num_tokens": 391690596.0,
      "step": 11690
    },
    {
      "epoch": 0.6981255969436485,
      "grad_norm": 0.5176255702972412,
      "learning_rate": 3.98675253568379e-05,
      "loss": 0.5914,
      "mean_token_accuracy": 0.8139866471290589,
      "num_tokens": 391858276.0,
      "step": 11695
    },
    {
      "epoch": 0.6984240687679083,
      "grad_norm": 0.5208262801170349,
      "learning_rate": 3.985824746416168e-05,
      "loss": 0.5236,
      "mean_token_accuracy": 0.8321722507476806,
      "num_tokens": 392025956.0,
      "step": 11700
    },
    {
      "epoch": 0.6987225405921681,
      "grad_norm": 0.6102920770645142,
      "learning_rate": 3.9848966561406185e-05,
      "loss": 0.5612,
      "mean_token_accuracy": 0.8202373743057251,
      "num_tokens": 392193636.0,
      "step": 11705
    },
    {
      "epoch": 0.6990210124164279,
      "grad_norm": 0.5241305232048035,
      "learning_rate": 3.9839682650831956e-05,
      "loss": 0.559,
      "mean_token_accuracy": 0.8223368763923645,
      "num_tokens": 392361316.0,
      "step": 11710
    },
    {
      "epoch": 0.6993194842406877,
      "grad_norm": 0.44423869252204895,
      "learning_rate": 3.983039573470024e-05,
      "loss": 0.5544,
      "mean_token_accuracy": 0.8246928215026855,
      "num_tokens": 392528996.0,
      "step": 11715
    },
    {
      "epoch": 0.6996179560649475,
      "grad_norm": 0.5356616377830505,
      "learning_rate": 3.982110581527306e-05,
      "loss": 0.5148,
      "mean_token_accuracy": 0.8359656572341919,
      "num_tokens": 392696676.0,
      "step": 11720
    },
    {
      "epoch": 0.6999164278892073,
      "grad_norm": 0.5257697701454163,
      "learning_rate": 3.981181289481313e-05,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8279494166374206,
      "num_tokens": 392864356.0,
      "step": 11725
    },
    {
      "epoch": 0.7002148997134671,
      "grad_norm": 0.5748518705368042,
      "learning_rate": 3.980251697558392e-05,
      "loss": 0.5771,
      "mean_token_accuracy": 0.8179112434387207,
      "num_tokens": 393032036.0,
      "step": 11730
    },
    {
      "epoch": 0.7005133715377269,
      "grad_norm": 0.5754997134208679,
      "learning_rate": 3.9793218059849627e-05,
      "loss": 0.5829,
      "mean_token_accuracy": 0.8174519777297974,
      "num_tokens": 393199716.0,
      "step": 11735
    },
    {
      "epoch": 0.7008118433619867,
      "grad_norm": 0.6012305617332458,
      "learning_rate": 3.978391614987517e-05,
      "loss": 0.5569,
      "mean_token_accuracy": 0.8231122493743896,
      "num_tokens": 393367396.0,
      "step": 11740
    },
    {
      "epoch": 0.7011103151862464,
      "grad_norm": 0.5065628290176392,
      "learning_rate": 3.97746112479262e-05,
      "loss": 0.5348,
      "mean_token_accuracy": 0.8282297492027283,
      "num_tokens": 393535076.0,
      "step": 11745
    },
    {
      "epoch": 0.7014087870105062,
      "grad_norm": 0.5431389808654785,
      "learning_rate": 3.976530335626909e-05,
      "loss": 0.5798,
      "mean_token_accuracy": 0.8172134160995483,
      "num_tokens": 393702756.0,
      "step": 11750
    },
    {
      "epoch": 0.701707258834766,
      "grad_norm": 0.5676679611206055,
      "learning_rate": 3.975599247717096e-05,
      "loss": 0.5718,
      "mean_token_accuracy": 0.8192114949226379,
      "num_tokens": 393870436.0,
      "step": 11755
    },
    {
      "epoch": 0.7020057306590258,
      "grad_norm": 0.5505884885787964,
      "learning_rate": 3.974667861289964e-05,
      "loss": 0.5709,
      "mean_token_accuracy": 0.8195395469665527,
      "num_tokens": 394038116.0,
      "step": 11760
    },
    {
      "epoch": 0.7023042024832856,
      "grad_norm": 0.5596762895584106,
      "learning_rate": 3.9737361765723703e-05,
      "loss": 0.5106,
      "mean_token_accuracy": 0.8344387531280517,
      "num_tokens": 394205796.0,
      "step": 11765
    },
    {
      "epoch": 0.7026026743075454,
      "grad_norm": 0.48200860619544983,
      "learning_rate": 3.972804193791242e-05,
      "loss": 0.5475,
      "mean_token_accuracy": 0.8268400430679321,
      "num_tokens": 394373476.0,
      "step": 11770
    },
    {
      "epoch": 0.7029011461318052,
      "grad_norm": 0.5312851667404175,
      "learning_rate": 3.9718719131735824e-05,
      "loss": 0.5471,
      "mean_token_accuracy": 0.8260825395584106,
      "num_tokens": 394541156.0,
      "step": 11775
    },
    {
      "epoch": 0.7031996179560649,
      "grad_norm": 0.49796268343925476,
      "learning_rate": 3.9709393349464656e-05,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8380353093147278,
      "num_tokens": 394708836.0,
      "step": 11780
    },
    {
      "epoch": 0.7034980897803247,
      "grad_norm": 0.5282920598983765,
      "learning_rate": 3.970006459337038e-05,
      "loss": 0.5596,
      "mean_token_accuracy": 0.8240128993988037,
      "num_tokens": 394876516.0,
      "step": 11785
    },
    {
      "epoch": 0.7037965616045845,
      "grad_norm": 0.5482062697410583,
      "learning_rate": 3.9690732865725186e-05,
      "loss": 0.5674,
      "mean_token_accuracy": 0.8192890405654907,
      "num_tokens": 395044196.0,
      "step": 11790
    },
    {
      "epoch": 0.7040950334288443,
      "grad_norm": 0.5222627520561218,
      "learning_rate": 3.9681398168801974e-05,
      "loss": 0.5149,
      "mean_token_accuracy": 0.833997368812561,
      "num_tokens": 395211876.0,
      "step": 11795
    },
    {
      "epoch": 0.7043935052531041,
      "grad_norm": 0.5341181755065918,
      "learning_rate": 3.967206050487441e-05,
      "loss": 0.5683,
      "mean_token_accuracy": 0.8223965167999268,
      "num_tokens": 395379556.0,
      "step": 11800
    },
    {
      "epoch": 0.7046919770773639,
      "grad_norm": 0.5348356366157532,
      "learning_rate": 3.966271987621683e-05,
      "loss": 0.4906,
      "mean_token_accuracy": 0.8417690515518188,
      "num_tokens": 395547236.0,
      "step": 11805
    },
    {
      "epoch": 0.7049904489016237,
      "grad_norm": 0.5016529560089111,
      "learning_rate": 3.965337628510433e-05,
      "loss": 0.4846,
      "mean_token_accuracy": 0.8448347806930542,
      "num_tokens": 395714916.0,
      "step": 11810
    },
    {
      "epoch": 0.7052889207258835,
      "grad_norm": 0.5226614475250244,
      "learning_rate": 3.964402973381271e-05,
      "loss": 0.5207,
      "mean_token_accuracy": 0.8325957298278809,
      "num_tokens": 395882596.0,
      "step": 11815
    },
    {
      "epoch": 0.7055873925501432,
      "grad_norm": 0.5080669522285461,
      "learning_rate": 3.9634680224618495e-05,
      "loss": 0.6092,
      "mean_token_accuracy": 0.8084575891494751,
      "num_tokens": 396050276.0,
      "step": 11820
    },
    {
      "epoch": 0.705885864374403,
      "grad_norm": 0.5736380219459534,
      "learning_rate": 3.962532775979893e-05,
      "loss": 0.569,
      "mean_token_accuracy": 0.8204342126846313,
      "num_tokens": 396217956.0,
      "step": 11825
    },
    {
      "epoch": 0.7061843361986628,
      "grad_norm": 0.48929014801979065,
      "learning_rate": 3.961597234163199e-05,
      "loss": 0.5596,
      "mean_token_accuracy": 0.822957181930542,
      "num_tokens": 396385636.0,
      "step": 11830
    },
    {
      "epoch": 0.7064828080229226,
      "grad_norm": 0.47653937339782715,
      "learning_rate": 3.9606613972396334e-05,
      "loss": 0.5443,
      "mean_token_accuracy": 0.8255397796630859,
      "num_tokens": 396553316.0,
      "step": 11835
    },
    {
      "epoch": 0.7067812798471824,
      "grad_norm": 0.5823026299476624,
      "learning_rate": 3.959725265437139e-05,
      "loss": 0.5712,
      "mean_token_accuracy": 0.8201896786689759,
      "num_tokens": 396720996.0,
      "step": 11840
    },
    {
      "epoch": 0.7070797516714422,
      "grad_norm": 0.5395498275756836,
      "learning_rate": 3.958788838983727e-05,
      "loss": 0.5117,
      "mean_token_accuracy": 0.834862220287323,
      "num_tokens": 396888676.0,
      "step": 11845
    },
    {
      "epoch": 0.707378223495702,
      "grad_norm": 0.599995493888855,
      "learning_rate": 3.9578521181074825e-05,
      "loss": 0.5234,
      "mean_token_accuracy": 0.8309496760368347,
      "num_tokens": 397041315.0,
      "step": 11850
    },
    {
      "epoch": 0.7076766953199618,
      "grad_norm": 0.6038875579833984,
      "learning_rate": 3.9569151030365596e-05,
      "loss": 0.5794,
      "mean_token_accuracy": 0.8185613512992859,
      "num_tokens": 397208995.0,
      "step": 11855
    },
    {
      "epoch": 0.7079751671442216,
      "grad_norm": 0.5417625904083252,
      "learning_rate": 3.955977793999186e-05,
      "loss": 0.5957,
      "mean_token_accuracy": 0.8149170875549316,
      "num_tokens": 397376675.0,
      "step": 11860
    },
    {
      "epoch": 0.7082736389684814,
      "grad_norm": 0.5411232113838196,
      "learning_rate": 3.955040191223661e-05,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8206071734428406,
      "num_tokens": 397544355.0,
      "step": 11865
    },
    {
      "epoch": 0.7085721107927412,
      "grad_norm": 0.5124090909957886,
      "learning_rate": 3.9541022949383563e-05,
      "loss": 0.5332,
      "mean_token_accuracy": 0.8303113579750061,
      "num_tokens": 397712035.0,
      "step": 11870
    },
    {
      "epoch": 0.708870582617001,
      "grad_norm": 0.5989068150520325,
      "learning_rate": 3.9531641053717107e-05,
      "loss": 0.5753,
      "mean_token_accuracy": 0.8204282402992249,
      "num_tokens": 397879715.0,
      "step": 11875
    },
    {
      "epoch": 0.7091690544412608,
      "grad_norm": 0.5417442321777344,
      "learning_rate": 3.95222562275224e-05,
      "loss": 0.543,
      "mean_token_accuracy": 0.8272575497627258,
      "num_tokens": 398047395.0,
      "step": 11880
    },
    {
      "epoch": 0.7094675262655206,
      "grad_norm": 0.5037710666656494,
      "learning_rate": 3.951286847308528e-05,
      "loss": 0.5674,
      "mean_token_accuracy": 0.8210068106651306,
      "num_tokens": 398215075.0,
      "step": 11885
    },
    {
      "epoch": 0.7097659980897804,
      "grad_norm": 0.5060092210769653,
      "learning_rate": 3.950347779269232e-05,
      "loss": 0.5724,
      "mean_token_accuracy": 0.8195335865020752,
      "num_tokens": 398382755.0,
      "step": 11890
    },
    {
      "epoch": 0.7100644699140402,
      "grad_norm": 0.4879721701145172,
      "learning_rate": 3.9494084188630776e-05,
      "loss": 0.5205,
      "mean_token_accuracy": 0.834009301662445,
      "num_tokens": 398550435.0,
      "step": 11895
    },
    {
      "epoch": 0.7103629417383,
      "grad_norm": 0.49785342812538147,
      "learning_rate": 3.948468766318864e-05,
      "loss": 0.5832,
      "mean_token_accuracy": 0.815513551235199,
      "num_tokens": 398718115.0,
      "step": 11900
    },
    {
      "epoch": 0.7106614135625597,
      "grad_norm": 0.5198028087615967,
      "learning_rate": 3.9475288218654615e-05,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8280090689659119,
      "num_tokens": 398885795.0,
      "step": 11905
    },
    {
      "epoch": 0.7109598853868195,
      "grad_norm": 0.5034233927726746,
      "learning_rate": 3.946588585731812e-05,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8311761975288391,
      "num_tokens": 399053475.0,
      "step": 11910
    },
    {
      "epoch": 0.7112583572110793,
      "grad_norm": 0.523741602897644,
      "learning_rate": 3.945648058146924e-05,
      "loss": 0.5224,
      "mean_token_accuracy": 0.8322975039482117,
      "num_tokens": 399221155.0,
      "step": 11915
    },
    {
      "epoch": 0.711556829035339,
      "grad_norm": 0.49971213936805725,
      "learning_rate": 3.9447072393398835e-05,
      "loss": 0.5481,
      "mean_token_accuracy": 0.8265239119529724,
      "num_tokens": 399388835.0,
      "step": 11920
    },
    {
      "epoch": 0.7118553008595988,
      "grad_norm": 0.5266532897949219,
      "learning_rate": 3.943766129539843e-05,
      "loss": 0.5608,
      "mean_token_accuracy": 0.8214004516601563,
      "num_tokens": 399556515.0,
      "step": 11925
    },
    {
      "epoch": 0.7121537726838586,
      "grad_norm": 0.5646588802337646,
      "learning_rate": 3.9428247289760266e-05,
      "loss": 0.5514,
      "mean_token_accuracy": 0.8242395162582398,
      "num_tokens": 399724195.0,
      "step": 11930
    },
    {
      "epoch": 0.7124522445081184,
      "grad_norm": 0.5337622761726379,
      "learning_rate": 3.94188303787773e-05,
      "loss": 0.5632,
      "mean_token_accuracy": 0.8218060255050659,
      "num_tokens": 399891875.0,
      "step": 11935
    },
    {
      "epoch": 0.7127507163323782,
      "grad_norm": 0.4910246729850769,
      "learning_rate": 3.9409410564743215e-05,
      "loss": 0.5686,
      "mean_token_accuracy": 0.8197960138320923,
      "num_tokens": 400059555.0,
      "step": 11940
    },
    {
      "epoch": 0.713049188156638,
      "grad_norm": 0.5019068121910095,
      "learning_rate": 3.939998784995236e-05,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8304067730903626,
      "num_tokens": 400227235.0,
      "step": 11945
    },
    {
      "epoch": 0.7133476599808978,
      "grad_norm": 0.5448180437088013,
      "learning_rate": 3.9390562236699804e-05,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8271501898765564,
      "num_tokens": 400394915.0,
      "step": 11950
    },
    {
      "epoch": 0.7136461318051576,
      "grad_norm": 0.5013893842697144,
      "learning_rate": 3.938113372728135e-05,
      "loss": 0.5295,
      "mean_token_accuracy": 0.8310032248497009,
      "num_tokens": 400562595.0,
      "step": 11955
    },
    {
      "epoch": 0.7139446036294174,
      "grad_norm": 0.5113101601600647,
      "learning_rate": 3.937170232399348e-05,
      "loss": 0.4928,
      "mean_token_accuracy": 0.8408922672271728,
      "num_tokens": 400730275.0,
      "step": 11960
    },
    {
      "epoch": 0.7142430754536772,
      "grad_norm": 0.4909002482891083,
      "learning_rate": 3.936226802913337e-05,
      "loss": 0.5759,
      "mean_token_accuracy": 0.8190206289291382,
      "num_tokens": 400897955.0,
      "step": 11965
    },
    {
      "epoch": 0.714541547277937,
      "grad_norm": 0.5437666773796082,
      "learning_rate": 3.935283084499894e-05,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8215435981750489,
      "num_tokens": 401065635.0,
      "step": 11970
    },
    {
      "epoch": 0.7148400191021967,
      "grad_norm": 0.5017220973968506,
      "learning_rate": 3.934339077388878e-05,
      "loss": 0.549,
      "mean_token_accuracy": 0.8266372323036194,
      "num_tokens": 401233315.0,
      "step": 11975
    },
    {
      "epoch": 0.7151384909264565,
      "grad_norm": 0.49572592973709106,
      "learning_rate": 3.9333947818102186e-05,
      "loss": 0.5548,
      "mean_token_accuracy": 0.8233866214752197,
      "num_tokens": 401400995.0,
      "step": 11980
    },
    {
      "epoch": 0.7154369627507163,
      "grad_norm": 0.48280057311058044,
      "learning_rate": 3.932450197993917e-05,
      "loss": 0.5287,
      "mean_token_accuracy": 0.8322199702262878,
      "num_tokens": 401568675.0,
      "step": 11985
    },
    {
      "epoch": 0.7157354345749761,
      "grad_norm": 0.54119873046875,
      "learning_rate": 3.931505326170044e-05,
      "loss": 0.5695,
      "mean_token_accuracy": 0.8198675870895386,
      "num_tokens": 401736355.0,
      "step": 11990
    },
    {
      "epoch": 0.7160339063992359,
      "grad_norm": 0.5073707103729248,
      "learning_rate": 3.9305601665687416e-05,
      "loss": 0.5307,
      "mean_token_accuracy": 0.8299296259880066,
      "num_tokens": 401904035.0,
      "step": 11995
    },
    {
      "epoch": 0.7163323782234957,
      "grad_norm": 0.4858524203300476,
      "learning_rate": 3.92961471942022e-05,
      "loss": 0.5213,
      "mean_token_accuracy": 0.8352320075035096,
      "num_tokens": 402071715.0,
      "step": 12000
    },
    {
      "epoch": 0.7166308500477555,
      "grad_norm": 0.502540647983551,
      "learning_rate": 3.92866898495476e-05,
      "loss": 0.5117,
      "mean_token_accuracy": 0.8379339218139649,
      "num_tokens": 402239395.0,
      "step": 12005
    },
    {
      "epoch": 0.7169293218720153,
      "grad_norm": 0.5087219476699829,
      "learning_rate": 3.927722963402712e-05,
      "loss": 0.5334,
      "mean_token_accuracy": 0.8295299887657166,
      "num_tokens": 402407075.0,
      "step": 12010
    },
    {
      "epoch": 0.7172277936962751,
      "grad_norm": 0.5110599398612976,
      "learning_rate": 3.926776654994499e-05,
      "loss": 0.5683,
      "mean_token_accuracy": 0.8199093461036682,
      "num_tokens": 402574755.0,
      "step": 12015
    },
    {
      "epoch": 0.7175262655205349,
      "grad_norm": 0.47929057478904724,
      "learning_rate": 3.925830059960611e-05,
      "loss": 0.5232,
      "mean_token_accuracy": 0.8329834103584289,
      "num_tokens": 402742435.0,
      "step": 12020
    },
    {
      "epoch": 0.7178247373447947,
      "grad_norm": 0.624451220035553,
      "learning_rate": 3.924883178531608e-05,
      "loss": 0.4865,
      "mean_token_accuracy": 0.8451986193656922,
      "num_tokens": 402910115.0,
      "step": 12025
    },
    {
      "epoch": 0.7181232091690545,
      "grad_norm": 0.6565896272659302,
      "learning_rate": 3.923936010938122e-05,
      "loss": 0.5543,
      "mean_token_accuracy": 0.8263302445411682,
      "num_tokens": 403070069.0,
      "step": 12030
    },
    {
      "epoch": 0.7184216809933143,
      "grad_norm": 0.5351299047470093,
      "learning_rate": 3.922988557410852e-05,
      "loss": 0.558,
      "mean_token_accuracy": 0.8221161842346192,
      "num_tokens": 403237749.0,
      "step": 12035
    },
    {
      "epoch": 0.7187201528175741,
      "grad_norm": 0.5642988681793213,
      "learning_rate": 3.922040818180567e-05,
      "loss": 0.5823,
      "mean_token_accuracy": 0.8155075788497925,
      "num_tokens": 403405429.0,
      "step": 12040
    },
    {
      "epoch": 0.7190186246418339,
      "grad_norm": 0.46795371174812317,
      "learning_rate": 3.921092793478107e-05,
      "loss": 0.5247,
      "mean_token_accuracy": 0.8317189455032349,
      "num_tokens": 403573109.0,
      "step": 12045
    },
    {
      "epoch": 0.7193170964660937,
      "grad_norm": 0.5519506335258484,
      "learning_rate": 3.920144483534382e-05,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8202493190765381,
      "num_tokens": 403740789.0,
      "step": 12050
    },
    {
      "epoch": 0.7196155682903533,
      "grad_norm": 0.6325945854187012,
      "learning_rate": 3.9191958885803704e-05,
      "loss": 0.6071,
      "mean_token_accuracy": 0.8072289228439331,
      "num_tokens": 403908469.0,
      "step": 12055
    },
    {
      "epoch": 0.7199140401146131,
      "grad_norm": 0.5446462035179138,
      "learning_rate": 3.9182470088471186e-05,
      "loss": 0.5619,
      "mean_token_accuracy": 0.8210604786872864,
      "num_tokens": 404076149.0,
      "step": 12060
    },
    {
      "epoch": 0.7202125119388729,
      "grad_norm": 0.4894942343235016,
      "learning_rate": 3.9172978445657446e-05,
      "loss": 0.5261,
      "mean_token_accuracy": 0.8332219958305359,
      "num_tokens": 404243829.0,
      "step": 12065
    },
    {
      "epoch": 0.7205109837631327,
      "grad_norm": 0.5383883714675903,
      "learning_rate": 3.9163483959674335e-05,
      "loss": 0.534,
      "mean_token_accuracy": 0.8312239050865173,
      "num_tokens": 404411509.0,
      "step": 12070
    },
    {
      "epoch": 0.7208094555873925,
      "grad_norm": 0.5403720736503601,
      "learning_rate": 3.9153986632834436e-05,
      "loss": 0.5839,
      "mean_token_accuracy": 0.8152391672134399,
      "num_tokens": 404579189.0,
      "step": 12075
    },
    {
      "epoch": 0.7211079274116523,
      "grad_norm": 0.5110628008842468,
      "learning_rate": 3.9144486467450974e-05,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8216330647468567,
      "num_tokens": 404746869.0,
      "step": 12080
    },
    {
      "epoch": 0.7214063992359121,
      "grad_norm": 0.5106228590011597,
      "learning_rate": 3.913498346583791e-05,
      "loss": 0.517,
      "mean_token_accuracy": 0.8357747673988343,
      "num_tokens": 404914549.0,
      "step": 12085
    },
    {
      "epoch": 0.7217048710601719,
      "grad_norm": 0.4880300462245941,
      "learning_rate": 3.9125477630309865e-05,
      "loss": 0.5361,
      "mean_token_accuracy": 0.8293093085289002,
      "num_tokens": 405082229.0,
      "step": 12090
    },
    {
      "epoch": 0.7220033428844317,
      "grad_norm": 0.530001163482666,
      "learning_rate": 3.911596896318216e-05,
      "loss": 0.5496,
      "mean_token_accuracy": 0.82604079246521,
      "num_tokens": 405249909.0,
      "step": 12095
    },
    {
      "epoch": 0.7223018147086915,
      "grad_norm": 0.5284551978111267,
      "learning_rate": 3.9106457466770805e-05,
      "loss": 0.524,
      "mean_token_accuracy": 0.8315698385238648,
      "num_tokens": 405417589.0,
      "step": 12100
    },
    {
      "epoch": 0.7226002865329513,
      "grad_norm": 0.4907339811325073,
      "learning_rate": 3.9096943143392506e-05,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8217523574829102,
      "num_tokens": 405585269.0,
      "step": 12105
    },
    {
      "epoch": 0.7228987583572111,
      "grad_norm": 0.571054220199585,
      "learning_rate": 3.908742599536465e-05,
      "loss": 0.5781,
      "mean_token_accuracy": 0.8165990591049195,
      "num_tokens": 405752949.0,
      "step": 12110
    },
    {
      "epoch": 0.7231972301814709,
      "grad_norm": 0.516926646232605,
      "learning_rate": 3.9077906025005304e-05,
      "loss": 0.5494,
      "mean_token_accuracy": 0.8263688325881958,
      "num_tokens": 405920629.0,
      "step": 12115
    },
    {
      "epoch": 0.7234957020057307,
      "grad_norm": 0.46530216932296753,
      "learning_rate": 3.9068383234633256e-05,
      "loss": 0.5173,
      "mean_token_accuracy": 0.8344626069068909,
      "num_tokens": 406088309.0,
      "step": 12120
    },
    {
      "epoch": 0.7237941738299905,
      "grad_norm": 0.5585097074508667,
      "learning_rate": 3.905885762656793e-05,
      "loss": 0.5671,
      "mean_token_accuracy": 0.821340811252594,
      "num_tokens": 406255989.0,
      "step": 12125
    },
    {
      "epoch": 0.7240926456542502,
      "grad_norm": 0.534710168838501,
      "learning_rate": 3.904932920312948e-05,
      "loss": 0.5233,
      "mean_token_accuracy": 0.8315579175949097,
      "num_tokens": 406423669.0,
      "step": 12130
    },
    {
      "epoch": 0.72439111747851,
      "grad_norm": 0.4949762225151062,
      "learning_rate": 3.903979796663874e-05,
      "loss": 0.5604,
      "mean_token_accuracy": 0.822229516506195,
      "num_tokens": 406591349.0,
      "step": 12135
    },
    {
      "epoch": 0.7246895893027698,
      "grad_norm": 0.47784188389778137,
      "learning_rate": 3.9030263919417195e-05,
      "loss": 0.5005,
      "mean_token_accuracy": 0.8394965887069702,
      "num_tokens": 406759029.0,
      "step": 12140
    },
    {
      "epoch": 0.7249880611270296,
      "grad_norm": 0.5196817517280579,
      "learning_rate": 3.902072706378705e-05,
      "loss": 0.5057,
      "mean_token_accuracy": 0.8381546020507813,
      "num_tokens": 406926709.0,
      "step": 12145
    },
    {
      "epoch": 0.7252865329512894,
      "grad_norm": 0.5363680124282837,
      "learning_rate": 3.9011187402071186e-05,
      "loss": 0.5608,
      "mean_token_accuracy": 0.8227842092514038,
      "num_tokens": 407094389.0,
      "step": 12150
    },
    {
      "epoch": 0.7255850047755492,
      "grad_norm": 0.5104947686195374,
      "learning_rate": 3.900164493659314e-05,
      "loss": 0.5229,
      "mean_token_accuracy": 0.8333233952522278,
      "num_tokens": 407262069.0,
      "step": 12155
    },
    {
      "epoch": 0.725883476599809,
      "grad_norm": 0.6038317084312439,
      "learning_rate": 3.899209966967719e-05,
      "loss": 0.5708,
      "mean_token_accuracy": 0.8190385341644287,
      "num_tokens": 407429749.0,
      "step": 12160
    },
    {
      "epoch": 0.7261819484240688,
      "grad_norm": 0.5249974131584167,
      "learning_rate": 3.898255160364824e-05,
      "loss": 0.5286,
      "mean_token_accuracy": 0.8315817713737488,
      "num_tokens": 407597429.0,
      "step": 12165
    },
    {
      "epoch": 0.7264804202483286,
      "grad_norm": 0.5322732329368591,
      "learning_rate": 3.89730007408319e-05,
      "loss": 0.5373,
      "mean_token_accuracy": 0.828307282924652,
      "num_tokens": 407765109.0,
      "step": 12170
    },
    {
      "epoch": 0.7267788920725884,
      "grad_norm": 0.6955226063728333,
      "learning_rate": 3.896344708355447e-05,
      "loss": 0.5662,
      "mean_token_accuracy": 0.8209650397300721,
      "num_tokens": 407932789.0,
      "step": 12175
    },
    {
      "epoch": 0.7270773638968482,
      "grad_norm": 0.5516892671585083,
      "learning_rate": 3.8953890634142904e-05,
      "loss": 0.5452,
      "mean_token_accuracy": 0.8284146428108216,
      "num_tokens": 408100469.0,
      "step": 12180
    },
    {
      "epoch": 0.727375835721108,
      "grad_norm": 0.5327548980712891,
      "learning_rate": 3.894433139492487e-05,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8306632518768311,
      "num_tokens": 408268149.0,
      "step": 12185
    },
    {
      "epoch": 0.7276743075453678,
      "grad_norm": 0.5411967039108276,
      "learning_rate": 3.8934769368228675e-05,
      "loss": 0.567,
      "mean_token_accuracy": 0.8178456425666809,
      "num_tokens": 408435829.0,
      "step": 12190
    },
    {
      "epoch": 0.7279727793696275,
      "grad_norm": 0.5814877152442932,
      "learning_rate": 3.8925204556383345e-05,
      "loss": 0.5931,
      "mean_token_accuracy": 0.81239413022995,
      "num_tokens": 408603509.0,
      "step": 12195
    },
    {
      "epoch": 0.7282712511938872,
      "grad_norm": 0.45790383219718933,
      "learning_rate": 3.891563696171856e-05,
      "loss": 0.4856,
      "mean_token_accuracy": 0.8433973550796509,
      "num_tokens": 408771189.0,
      "step": 12200
    },
    {
      "epoch": 0.728569723018147,
      "grad_norm": 0.495868057012558,
      "learning_rate": 3.8906066586564675e-05,
      "loss": 0.5405,
      "mean_token_accuracy": 0.8264404177665711,
      "num_tokens": 408938869.0,
      "step": 12205
    },
    {
      "epoch": 0.7288681948424068,
      "grad_norm": 0.6081781387329102,
      "learning_rate": 3.8896493433252755e-05,
      "loss": 0.617,
      "mean_token_accuracy": 0.806328272819519,
      "num_tokens": 409106549.0,
      "step": 12210
    },
    {
      "epoch": 0.7291666666666666,
      "grad_norm": 0.580693244934082,
      "learning_rate": 3.888691750411449e-05,
      "loss": 0.6039,
      "mean_token_accuracy": 0.8085828423500061,
      "num_tokens": 409274229.0,
      "step": 12215
    },
    {
      "epoch": 0.7294651384909264,
      "grad_norm": 0.5905386805534363,
      "learning_rate": 3.8877338801482306e-05,
      "loss": 0.5733,
      "mean_token_accuracy": 0.8189073085784913,
      "num_tokens": 409441909.0,
      "step": 12220
    },
    {
      "epoch": 0.7297636103151862,
      "grad_norm": 0.5744999647140503,
      "learning_rate": 3.886775732768924e-05,
      "loss": 0.5723,
      "mean_token_accuracy": 0.8189311742782592,
      "num_tokens": 409609589.0,
      "step": 12225
    },
    {
      "epoch": 0.730062082139446,
      "grad_norm": 0.5736605525016785,
      "learning_rate": 3.8858173085069063e-05,
      "loss": 0.5728,
      "mean_token_accuracy": 0.8173565626144409,
      "num_tokens": 409777269.0,
      "step": 12230
    },
    {
      "epoch": 0.7303605539637058,
      "grad_norm": 0.5335548520088196,
      "learning_rate": 3.884858607595617e-05,
      "loss": 0.5707,
      "mean_token_accuracy": 0.8188895583152771,
      "num_tokens": 409937824.0,
      "step": 12235
    },
    {
      "epoch": 0.7306590257879656,
      "grad_norm": 0.4800441265106201,
      "learning_rate": 3.883899630268568e-05,
      "loss": 0.56,
      "mean_token_accuracy": 0.8227305293083191,
      "num_tokens": 410105504.0,
      "step": 12240
    },
    {
      "epoch": 0.7309574976122254,
      "grad_norm": 0.5362090468406677,
      "learning_rate": 3.8829403767593345e-05,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8355302333831787,
      "num_tokens": 410273184.0,
      "step": 12245
    },
    {
      "epoch": 0.7312559694364852,
      "grad_norm": 0.4971427321434021,
      "learning_rate": 3.88198084730156e-05,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8270487904548645,
      "num_tokens": 410440864.0,
      "step": 12250
    },
    {
      "epoch": 0.731554441260745,
      "grad_norm": 0.5260369181632996,
      "learning_rate": 3.881021042128957e-05,
      "loss": 0.5621,
      "mean_token_accuracy": 0.8213348388671875,
      "num_tokens": 410608544.0,
      "step": 12255
    },
    {
      "epoch": 0.7318529130850048,
      "grad_norm": 0.5524857044219971,
      "learning_rate": 3.880060961475302e-05,
      "loss": 0.5543,
      "mean_token_accuracy": 0.824824047088623,
      "num_tokens": 410776224.0,
      "step": 12260
    },
    {
      "epoch": 0.7321513849092646,
      "grad_norm": 0.5446348786354065,
      "learning_rate": 3.8791006055744415e-05,
      "loss": 0.5694,
      "mean_token_accuracy": 0.8190802812576294,
      "num_tokens": 410943904.0,
      "step": 12265
    },
    {
      "epoch": 0.7324498567335244,
      "grad_norm": 0.5169851779937744,
      "learning_rate": 3.878139974660288e-05,
      "loss": 0.5295,
      "mean_token_accuracy": 0.831516170501709,
      "num_tokens": 411111584.0,
      "step": 12270
    },
    {
      "epoch": 0.7327483285577842,
      "grad_norm": 0.49699828028678894,
      "learning_rate": 3.8771790689668205e-05,
      "loss": 0.5374,
      "mean_token_accuracy": 0.8284862160682678,
      "num_tokens": 411279264.0,
      "step": 12275
    },
    {
      "epoch": 0.733046800382044,
      "grad_norm": 0.781423032283783,
      "learning_rate": 3.8762178887280856e-05,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8279136300086976,
      "num_tokens": 411446944.0,
      "step": 12280
    },
    {
      "epoch": 0.7333452722063037,
      "grad_norm": 0.6001463532447815,
      "learning_rate": 3.875256434178196e-05,
      "loss": 0.5704,
      "mean_token_accuracy": 0.8198735594749451,
      "num_tokens": 411614624.0,
      "step": 12285
    },
    {
      "epoch": 0.7336437440305635,
      "grad_norm": 0.5646892786026001,
      "learning_rate": 3.874294705551332e-05,
      "loss": 0.5564,
      "mean_token_accuracy": 0.8235774874687195,
      "num_tokens": 411782304.0,
      "step": 12290
    },
    {
      "epoch": 0.7339422158548233,
      "grad_norm": 0.5249640345573425,
      "learning_rate": 3.873332703081741e-05,
      "loss": 0.5287,
      "mean_token_accuracy": 0.830901837348938,
      "num_tokens": 411949984.0,
      "step": 12295
    },
    {
      "epoch": 0.7342406876790831,
      "grad_norm": 0.49104538559913635,
      "learning_rate": 3.872370427003735e-05,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8361565113067627,
      "num_tokens": 412117664.0,
      "step": 12300
    },
    {
      "epoch": 0.7345391595033429,
      "grad_norm": 0.583514392375946,
      "learning_rate": 3.871407877551695e-05,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8202493190765381,
      "num_tokens": 412285344.0,
      "step": 12305
    },
    {
      "epoch": 0.7348376313276027,
      "grad_norm": 0.4964461922645569,
      "learning_rate": 3.870445054960068e-05,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8318438649177551,
      "num_tokens": 412451956.0,
      "step": 12310
    },
    {
      "epoch": 0.7351361031518625,
      "grad_norm": 0.5448853373527527,
      "learning_rate": 3.869481959463366e-05,
      "loss": 0.5749,
      "mean_token_accuracy": 0.8187760949134827,
      "num_tokens": 412619636.0,
      "step": 12315
    },
    {
      "epoch": 0.7354345749761223,
      "grad_norm": 0.5224013924598694,
      "learning_rate": 3.86851859129617e-05,
      "loss": 0.5762,
      "mean_token_accuracy": 0.8196290135383606,
      "num_tokens": 412787316.0,
      "step": 12320
    },
    {
      "epoch": 0.7357330468003821,
      "grad_norm": 0.5130847692489624,
      "learning_rate": 3.867554950693126e-05,
      "loss": 0.5349,
      "mean_token_accuracy": 0.8297805190086365,
      "num_tokens": 412954996.0,
      "step": 12325
    },
    {
      "epoch": 0.7360315186246418,
      "grad_norm": 0.5278207063674927,
      "learning_rate": 3.866591037888946e-05,
      "loss": 0.5714,
      "mean_token_accuracy": 0.8177084445953369,
      "num_tokens": 413122676.0,
      "step": 12330
    },
    {
      "epoch": 0.7363299904489016,
      "grad_norm": 0.4937397241592407,
      "learning_rate": 3.865626853118409e-05,
      "loss": 0.5633,
      "mean_token_accuracy": 0.8221996903419495,
      "num_tokens": 413290356.0,
      "step": 12335
    },
    {
      "epoch": 0.7366284622731614,
      "grad_norm": 0.46921923756599426,
      "learning_rate": 3.8646623966163584e-05,
      "loss": 0.559,
      "mean_token_accuracy": 0.8241679668426514,
      "num_tokens": 413458036.0,
      "step": 12340
    },
    {
      "epoch": 0.7369269340974212,
      "grad_norm": 0.47525590658187866,
      "learning_rate": 3.863697668617708e-05,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8403316259384155,
      "num_tokens": 413625716.0,
      "step": 12345
    },
    {
      "epoch": 0.737225405921681,
      "grad_norm": 0.4790109395980835,
      "learning_rate": 3.862732669357434e-05,
      "loss": 0.5394,
      "mean_token_accuracy": 0.8282595753669739,
      "num_tokens": 413793396.0,
      "step": 12350
    },
    {
      "epoch": 0.7375238777459407,
      "grad_norm": 0.563318133354187,
      "learning_rate": 3.8617673990705796e-05,
      "loss": 0.5453,
      "mean_token_accuracy": 0.82631516456604,
      "num_tokens": 413961076.0,
      "step": 12355
    },
    {
      "epoch": 0.7378223495702005,
      "grad_norm": 0.5196044445037842,
      "learning_rate": 3.860801857992253e-05,
      "loss": 0.5297,
      "mean_token_accuracy": 0.8309077858924866,
      "num_tokens": 414128756.0,
      "step": 12360
    },
    {
      "epoch": 0.7381208213944603,
      "grad_norm": 0.5479102730751038,
      "learning_rate": 3.8598360463576324e-05,
      "loss": 0.5908,
      "mean_token_accuracy": 0.8142967939376831,
      "num_tokens": 414296436.0,
      "step": 12365
    },
    {
      "epoch": 0.7384192932187201,
      "grad_norm": 0.5005536675453186,
      "learning_rate": 3.858869964401957e-05,
      "loss": 0.5202,
      "mean_token_accuracy": 0.8349457263946534,
      "num_tokens": 414464116.0,
      "step": 12370
    },
    {
      "epoch": 0.7387177650429799,
      "grad_norm": 0.5197511315345764,
      "learning_rate": 3.857903612360535e-05,
      "loss": 0.5323,
      "mean_token_accuracy": 0.8296910524368286,
      "num_tokens": 414631796.0,
      "step": 12375
    },
    {
      "epoch": 0.7390162368672397,
      "grad_norm": 0.5165748596191406,
      "learning_rate": 3.8569369904687375e-05,
      "loss": 0.4909,
      "mean_token_accuracy": 0.8423535704612732,
      "num_tokens": 414799476.0,
      "step": 12380
    },
    {
      "epoch": 0.7393147086914995,
      "grad_norm": 0.5564424395561218,
      "learning_rate": 3.855970098962005e-05,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8253668069839477,
      "num_tokens": 414967156.0,
      "step": 12385
    },
    {
      "epoch": 0.7396131805157593,
      "grad_norm": 0.5473902225494385,
      "learning_rate": 3.855002938075841e-05,
      "loss": 0.5285,
      "mean_token_accuracy": 0.8327090382575989,
      "num_tokens": 415134836.0,
      "step": 12390
    },
    {
      "epoch": 0.7399116523400191,
      "grad_norm": 0.5365471839904785,
      "learning_rate": 3.854035508045815e-05,
      "loss": 0.532,
      "mean_token_accuracy": 0.8312775731086731,
      "num_tokens": 415302516.0,
      "step": 12395
    },
    {
      "epoch": 0.7402101241642789,
      "grad_norm": 0.572266697883606,
      "learning_rate": 3.8530678091075625e-05,
      "loss": 0.5337,
      "mean_token_accuracy": 0.8299952268600463,
      "num_tokens": 415470196.0,
      "step": 12400
    },
    {
      "epoch": 0.7405085959885387,
      "grad_norm": 0.4953557252883911,
      "learning_rate": 3.852099841496786e-05,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8318978905677795,
      "num_tokens": 415637876.0,
      "step": 12405
    },
    {
      "epoch": 0.7408070678127985,
      "grad_norm": 0.5083139538764954,
      "learning_rate": 3.85113160544925e-05,
      "loss": 0.502,
      "mean_token_accuracy": 0.8381367087364197,
      "num_tokens": 415805556.0,
      "step": 12410
    },
    {
      "epoch": 0.7411055396370583,
      "grad_norm": 0.5080996155738831,
      "learning_rate": 3.8501631012007873e-05,
      "loss": 0.518,
      "mean_token_accuracy": 0.8339257955551147,
      "num_tokens": 415973236.0,
      "step": 12415
    },
    {
      "epoch": 0.7414040114613181,
      "grad_norm": 0.5270262360572815,
      "learning_rate": 3.849194328987293e-05,
      "loss": 0.563,
      "mean_token_accuracy": 0.8203029990196228,
      "num_tokens": 416140916.0,
      "step": 12420
    },
    {
      "epoch": 0.7417024832855779,
      "grad_norm": 0.4829140305519104,
      "learning_rate": 3.8482252890447334e-05,
      "loss": 0.5073,
      "mean_token_accuracy": 0.8386042952537537,
      "num_tokens": 416306881.0,
      "step": 12425
    },
    {
      "epoch": 0.7420009551098377,
      "grad_norm": 0.5917607545852661,
      "learning_rate": 3.8472559816091324e-05,
      "loss": 0.6197,
      "mean_token_accuracy": 0.8060181260108947,
      "num_tokens": 416474561.0,
      "step": 12430
    },
    {
      "epoch": 0.7422994269340975,
      "grad_norm": 0.5040755271911621,
      "learning_rate": 3.8462864069165846e-05,
      "loss": 0.5291,
      "mean_token_accuracy": 0.8313909053802491,
      "num_tokens": 416642241.0,
      "step": 12435
    },
    {
      "epoch": 0.7425978987583572,
      "grad_norm": 0.4962681233882904,
      "learning_rate": 3.845316565203247e-05,
      "loss": 0.5377,
      "mean_token_accuracy": 0.8275140047073364,
      "num_tokens": 416809921.0,
      "step": 12440
    },
    {
      "epoch": 0.742896370582617,
      "grad_norm": 0.4777105152606964,
      "learning_rate": 3.844346456705343e-05,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8348204612731933,
      "num_tokens": 416977601.0,
      "step": 12445
    },
    {
      "epoch": 0.7431948424068768,
      "grad_norm": 0.5366834998130798,
      "learning_rate": 3.843376081659159e-05,
      "loss": 0.5567,
      "mean_token_accuracy": 0.8225754499435425,
      "num_tokens": 417145281.0,
      "step": 12450
    },
    {
      "epoch": 0.7434933142311366,
      "grad_norm": 0.4639711380004883,
      "learning_rate": 3.84240544030105e-05,
      "loss": 0.5169,
      "mean_token_accuracy": 0.8342299818992615,
      "num_tokens": 417312961.0,
      "step": 12455
    },
    {
      "epoch": 0.7437917860553964,
      "grad_norm": 0.4997599720954895,
      "learning_rate": 3.841434532867432e-05,
      "loss": 0.5319,
      "mean_token_accuracy": 0.8300071477890014,
      "num_tokens": 417480641.0,
      "step": 12460
    },
    {
      "epoch": 0.7440902578796562,
      "grad_norm": 0.5132542848587036,
      "learning_rate": 3.8404633595947884e-05,
      "loss": 0.5432,
      "mean_token_accuracy": 0.8284742951393127,
      "num_tokens": 417648321.0,
      "step": 12465
    },
    {
      "epoch": 0.7443887297039159,
      "grad_norm": 0.558612048625946,
      "learning_rate": 3.839491920719666e-05,
      "loss": 0.5224,
      "mean_token_accuracy": 0.8330669164657593,
      "num_tokens": 417816001.0,
      "step": 12470
    },
    {
      "epoch": 0.7446872015281757,
      "grad_norm": 0.5155857801437378,
      "learning_rate": 3.838520216478677e-05,
      "loss": 0.5297,
      "mean_token_accuracy": 0.8296671867370605,
      "num_tokens": 417983681.0,
      "step": 12475
    },
    {
      "epoch": 0.7449856733524355,
      "grad_norm": 0.591007649898529,
      "learning_rate": 3.8375482471084966e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8344447016716003,
      "num_tokens": 418151361.0,
      "step": 12480
    },
    {
      "epoch": 0.7452841451766953,
      "grad_norm": 0.5150274634361267,
      "learning_rate": 3.836576012845867e-05,
      "loss": 0.5408,
      "mean_token_accuracy": 0.828146243095398,
      "num_tokens": 418319041.0,
      "step": 12485
    },
    {
      "epoch": 0.7455826170009551,
      "grad_norm": 0.4874996542930603,
      "learning_rate": 3.835603513927593e-05,
      "loss": 0.5344,
      "mean_token_accuracy": 0.8289812803268433,
      "num_tokens": 418486721.0,
      "step": 12490
    },
    {
      "epoch": 0.7458810888252149,
      "grad_norm": 0.5001745820045471,
      "learning_rate": 3.834630750590546e-05,
      "loss": 0.5695,
      "mean_token_accuracy": 0.8206668257713318,
      "num_tokens": 418654401.0,
      "step": 12495
    },
    {
      "epoch": 0.7461795606494747,
      "grad_norm": 0.5270755887031555,
      "learning_rate": 3.8336577230716595e-05,
      "loss": 0.5714,
      "mean_token_accuracy": 0.818108069896698,
      "num_tokens": 418822081.0,
      "step": 12500
    },
    {
      "epoch": 0.7464780324737345,
      "grad_norm": 0.49236175417900085,
      "learning_rate": 3.832684431607933e-05,
      "loss": 0.5372,
      "mean_token_accuracy": 0.8276273250579834,
      "num_tokens": 418989761.0,
      "step": 12505
    },
    {
      "epoch": 0.7467765042979942,
      "grad_norm": 0.525929868221283,
      "learning_rate": 3.831710876436428e-05,
      "loss": 0.5391,
      "mean_token_accuracy": 0.827168083190918,
      "num_tokens": 419157441.0,
      "step": 12510
    },
    {
      "epoch": 0.747074976122254,
      "grad_norm": 0.5231567621231079,
      "learning_rate": 3.830737057794274e-05,
      "loss": 0.5477,
      "mean_token_accuracy": 0.82493736743927,
      "num_tokens": 419325121.0,
      "step": 12515
    },
    {
      "epoch": 0.7473734479465138,
      "grad_norm": 0.49298495054244995,
      "learning_rate": 3.829762975918661e-05,
      "loss": 0.565,
      "mean_token_accuracy": 0.8215435981750489,
      "num_tokens": 419492801.0,
      "step": 12520
    },
    {
      "epoch": 0.7476719197707736,
      "grad_norm": 0.46484676003456116,
      "learning_rate": 3.828788631046845e-05,
      "loss": 0.5007,
      "mean_token_accuracy": 0.837743055820465,
      "num_tokens": 419660481.0,
      "step": 12525
    },
    {
      "epoch": 0.7479703915950334,
      "grad_norm": 0.47831910848617554,
      "learning_rate": 3.8278140234161463e-05,
      "loss": 0.5771,
      "mean_token_accuracy": 0.8174460291862488,
      "num_tokens": 419828161.0,
      "step": 12530
    },
    {
      "epoch": 0.7482688634192932,
      "grad_norm": 0.5279443264007568,
      "learning_rate": 3.8268391532639474e-05,
      "loss": 0.4913,
      "mean_token_accuracy": 0.8427233695983887,
      "num_tokens": 419995841.0,
      "step": 12535
    },
    {
      "epoch": 0.748567335243553,
      "grad_norm": 0.4774021804332733,
      "learning_rate": 3.825864020827697e-05,
      "loss": 0.521,
      "mean_token_accuracy": 0.8333472490310669,
      "num_tokens": 420163521.0,
      "step": 12540
    },
    {
      "epoch": 0.7488658070678128,
      "grad_norm": 0.513725757598877,
      "learning_rate": 3.824888626344905e-05,
      "loss": 0.5341,
      "mean_token_accuracy": 0.8304843187332154,
      "num_tokens": 420331201.0,
      "step": 12545
    },
    {
      "epoch": 0.7491642788920726,
      "grad_norm": 0.4979497790336609,
      "learning_rate": 3.8239129700531495e-05,
      "loss": 0.5466,
      "mean_token_accuracy": 0.825038754940033,
      "num_tokens": 420498881.0,
      "step": 12550
    },
    {
      "epoch": 0.7494627507163324,
      "grad_norm": 0.4993870258331299,
      "learning_rate": 3.822937052190067e-05,
      "loss": 0.535,
      "mean_token_accuracy": 0.827645230293274,
      "num_tokens": 420666561.0,
      "step": 12555
    },
    {
      "epoch": 0.7497612225405922,
      "grad_norm": 0.4990299344062805,
      "learning_rate": 3.821960872993362e-05,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8360014438629151,
      "num_tokens": 420834241.0,
      "step": 12560
    },
    {
      "epoch": 0.750059694364852,
      "grad_norm": 0.5518876314163208,
      "learning_rate": 3.8209844327008e-05,
      "loss": 0.4977,
      "mean_token_accuracy": 0.8402361989021301,
      "num_tokens": 421001921.0,
      "step": 12565
    },
    {
      "epoch": 0.7503581661891118,
      "grad_norm": 0.5664923191070557,
      "learning_rate": 3.820007731550211e-05,
      "loss": 0.5324,
      "mean_token_accuracy": 0.8299952387809754,
      "num_tokens": 421169601.0,
      "step": 12570
    },
    {
      "epoch": 0.7506566380133716,
      "grad_norm": 0.5897797346115112,
      "learning_rate": 3.819030769779489e-05,
      "loss": 0.5176,
      "mean_token_accuracy": 0.834342873096466,
      "num_tokens": 421331895.0,
      "step": 12575
    },
    {
      "epoch": 0.7509551098376314,
      "grad_norm": 0.529281497001648,
      "learning_rate": 3.818053547626591e-05,
      "loss": 0.5762,
      "mean_token_accuracy": 0.8180126428604126,
      "num_tokens": 421499575.0,
      "step": 12580
    },
    {
      "epoch": 0.7512535816618912,
      "grad_norm": 0.47769302129745483,
      "learning_rate": 3.817076065329537e-05,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8178396821022034,
      "num_tokens": 421667255.0,
      "step": 12585
    },
    {
      "epoch": 0.751552053486151,
      "grad_norm": 0.48184019327163696,
      "learning_rate": 3.816098323126413e-05,
      "loss": 0.4995,
      "mean_token_accuracy": 0.8382977485656739,
      "num_tokens": 421834935.0,
      "step": 12590
    },
    {
      "epoch": 0.7518505253104107,
      "grad_norm": 0.49549925327301025,
      "learning_rate": 3.8151203212553624e-05,
      "loss": 0.5548,
      "mean_token_accuracy": 0.8245258212089539,
      "num_tokens": 422002615.0,
      "step": 12595
    },
    {
      "epoch": 0.7521489971346705,
      "grad_norm": 0.4643173813819885,
      "learning_rate": 3.814142059954599e-05,
      "loss": 0.5512,
      "mean_token_accuracy": 0.8241739392280578,
      "num_tokens": 422170295.0,
      "step": 12600
    },
    {
      "epoch": 0.7524474689589303,
      "grad_norm": 0.536861777305603,
      "learning_rate": 3.813163539462395e-05,
      "loss": 0.5535,
      "mean_token_accuracy": 0.8233031153678894,
      "num_tokens": 422337975.0,
      "step": 12605
    },
    {
      "epoch": 0.75274594078319,
      "grad_norm": 0.6138288974761963,
      "learning_rate": 3.8121847600170867e-05,
      "loss": 0.532,
      "mean_token_accuracy": 0.830526065826416,
      "num_tokens": 422505655.0,
      "step": 12610
    },
    {
      "epoch": 0.7530444126074498,
      "grad_norm": 0.48329293727874756,
      "learning_rate": 3.811205721857075e-05,
      "loss": 0.5018,
      "mean_token_accuracy": 0.8385071277618408,
      "num_tokens": 422673218.0,
      "step": 12615
    },
    {
      "epoch": 0.7533428844317096,
      "grad_norm": 0.48226669430732727,
      "learning_rate": 3.810226425220821e-05,
      "loss": 0.5319,
      "mean_token_accuracy": 0.8300966143608093,
      "num_tokens": 422840898.0,
      "step": 12620
    },
    {
      "epoch": 0.7536413562559694,
      "grad_norm": 0.5072008371353149,
      "learning_rate": 3.809246870346853e-05,
      "loss": 0.5599,
      "mean_token_accuracy": 0.8212334632873535,
      "num_tokens": 423008578.0,
      "step": 12625
    },
    {
      "epoch": 0.7539398280802292,
      "grad_norm": 0.6549006700515747,
      "learning_rate": 3.808267057473757e-05,
      "loss": 0.587,
      "mean_token_accuracy": 0.8156208992004395,
      "num_tokens": 423176258.0,
      "step": 12630
    },
    {
      "epoch": 0.754238299904489,
      "grad_norm": 0.5449286103248596,
      "learning_rate": 3.807286986840187e-05,
      "loss": 0.5121,
      "mean_token_accuracy": 0.8350440382957458,
      "num_tokens": 423338896.0,
      "step": 12635
    },
    {
      "epoch": 0.7545367717287488,
      "grad_norm": 0.4926968514919281,
      "learning_rate": 3.806306658684855e-05,
      "loss": 0.5247,
      "mean_token_accuracy": 0.8317487716674805,
      "num_tokens": 423506576.0,
      "step": 12640
    },
    {
      "epoch": 0.7548352435530086,
      "grad_norm": 0.5020460486412048,
      "learning_rate": 3.80532607324654e-05,
      "loss": 0.5433,
      "mean_token_accuracy": 0.8282893776893616,
      "num_tokens": 423674256.0,
      "step": 12645
    },
    {
      "epoch": 0.7551337153772684,
      "grad_norm": 0.5663554668426514,
      "learning_rate": 3.80434523076408e-05,
      "loss": 0.5243,
      "mean_token_accuracy": 0.8324048757553101,
      "num_tokens": 423841936.0,
      "step": 12650
    },
    {
      "epoch": 0.7554321872015282,
      "grad_norm": 0.4894520938396454,
      "learning_rate": 3.8033641314763786e-05,
      "loss": 0.5299,
      "mean_token_accuracy": 0.8319277167320251,
      "num_tokens": 424009616.0,
      "step": 12655
    },
    {
      "epoch": 0.755730659025788,
      "grad_norm": 0.5790097713470459,
      "learning_rate": 3.8023827756223996e-05,
      "loss": 0.61,
      "mean_token_accuracy": 0.8063879370689392,
      "num_tokens": 424177296.0,
      "step": 12660
    },
    {
      "epoch": 0.7560291308500477,
      "grad_norm": 0.4726533591747284,
      "learning_rate": 3.801401163441171e-05,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8300608396530151,
      "num_tokens": 424344976.0,
      "step": 12665
    },
    {
      "epoch": 0.7563276026743075,
      "grad_norm": 0.48812583088874817,
      "learning_rate": 3.8004192951717826e-05,
      "loss": 0.4988,
      "mean_token_accuracy": 0.8403316140174866,
      "num_tokens": 424512656.0,
      "step": 12670
    },
    {
      "epoch": 0.7566260744985673,
      "grad_norm": 0.5608876943588257,
      "learning_rate": 3.799437171053386e-05,
      "loss": 0.5608,
      "mean_token_accuracy": 0.8230108499526978,
      "num_tokens": 424680336.0,
      "step": 12675
    },
    {
      "epoch": 0.7569245463228271,
      "grad_norm": 0.4958289563655853,
      "learning_rate": 3.798454791325196e-05,
      "loss": 0.5745,
      "mean_token_accuracy": 0.8201896667480468,
      "num_tokens": 424848016.0,
      "step": 12680
    },
    {
      "epoch": 0.7572230181470869,
      "grad_norm": 0.5346928834915161,
      "learning_rate": 3.79747215622649e-05,
      "loss": 0.4835,
      "mean_token_accuracy": 0.8438208341598511,
      "num_tokens": 425015696.0,
      "step": 12685
    },
    {
      "epoch": 0.7575214899713467,
      "grad_norm": 0.4995134770870209,
      "learning_rate": 3.796489265996606e-05,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8379756689071656,
      "num_tokens": 425183376.0,
      "step": 12690
    },
    {
      "epoch": 0.7578199617956065,
      "grad_norm": 0.515295147895813,
      "learning_rate": 3.795506120874945e-05,
      "loss": 0.525,
      "mean_token_accuracy": 0.8330490231513977,
      "num_tokens": 425351056.0,
      "step": 12695
    },
    {
      "epoch": 0.7581184336198663,
      "grad_norm": 0.5539008378982544,
      "learning_rate": 3.79452272110097e-05,
      "loss": 0.5785,
      "mean_token_accuracy": 0.8171060442924499,
      "num_tokens": 425518736.0,
      "step": 12700
    },
    {
      "epoch": 0.7584169054441261,
      "grad_norm": 0.683463990688324,
      "learning_rate": 3.793539066914207e-05,
      "loss": 0.5559,
      "mean_token_accuracy": 0.8225396633148193,
      "num_tokens": 425686416.0,
      "step": 12705
    },
    {
      "epoch": 0.7587153772683859,
      "grad_norm": 0.5496516227722168,
      "learning_rate": 3.792555158554243e-05,
      "loss": 0.5912,
      "mean_token_accuracy": 0.8126148104667663,
      "num_tokens": 425854096.0,
      "step": 12710
    },
    {
      "epoch": 0.7590138490926457,
      "grad_norm": 0.6611767411231995,
      "learning_rate": 3.791570996260727e-05,
      "loss": 0.576,
      "mean_token_accuracy": 0.8191399216651917,
      "num_tokens": 426021776.0,
      "step": 12715
    },
    {
      "epoch": 0.7593123209169055,
      "grad_norm": 0.524413526058197,
      "learning_rate": 3.790586580273369e-05,
      "loss": 0.5289,
      "mean_token_accuracy": 0.8310390114784241,
      "num_tokens": 426189456.0,
      "step": 12720
    },
    {
      "epoch": 0.7596107927411653,
      "grad_norm": 0.5060654282569885,
      "learning_rate": 3.789601910831944e-05,
      "loss": 0.5265,
      "mean_token_accuracy": 0.8316712379455566,
      "num_tokens": 426357136.0,
      "step": 12725
    },
    {
      "epoch": 0.7599092645654251,
      "grad_norm": 0.5930113792419434,
      "learning_rate": 3.788616988176282e-05,
      "loss": 0.5499,
      "mean_token_accuracy": 0.824209702014923,
      "num_tokens": 426524816.0,
      "step": 12730
    },
    {
      "epoch": 0.7602077363896849,
      "grad_norm": 0.508378803730011,
      "learning_rate": 3.7876318125462836e-05,
      "loss": 0.5099,
      "mean_token_accuracy": 0.8372360706329346,
      "num_tokens": 426692496.0,
      "step": 12735
    },
    {
      "epoch": 0.7605062082139447,
      "grad_norm": 0.5116025805473328,
      "learning_rate": 3.786646384181905e-05,
      "loss": 0.5252,
      "mean_token_accuracy": 0.831641411781311,
      "num_tokens": 426860176.0,
      "step": 12740
    },
    {
      "epoch": 0.7608046800382043,
      "grad_norm": 0.5730810761451721,
      "learning_rate": 3.785660703323164e-05,
      "loss": 0.5468,
      "mean_token_accuracy": 0.8255099654197693,
      "num_tokens": 427027856.0,
      "step": 12745
    },
    {
      "epoch": 0.7611031518624641,
      "grad_norm": 0.49122607707977295,
      "learning_rate": 3.7846747702101436e-05,
      "loss": 0.5295,
      "mean_token_accuracy": 0.8305678367614746,
      "num_tokens": 427195536.0,
      "step": 12750
    },
    {
      "epoch": 0.7614016236867239,
      "grad_norm": 0.5074653029441833,
      "learning_rate": 3.783688585082985e-05,
      "loss": 0.49,
      "mean_token_accuracy": 0.8436657547950744,
      "num_tokens": 427363216.0,
      "step": 12755
    },
    {
      "epoch": 0.7617000955109837,
      "grad_norm": 0.5781688690185547,
      "learning_rate": 3.7827021481818917e-05,
      "loss": 0.5834,
      "mean_token_accuracy": 0.8172313094139099,
      "num_tokens": 427530896.0,
      "step": 12760
    },
    {
      "epoch": 0.7619985673352435,
      "grad_norm": 0.579497754573822,
      "learning_rate": 3.781715459747128e-05,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8302270293235778,
      "num_tokens": 427693025.0,
      "step": 12765
    },
    {
      "epoch": 0.7622970391595033,
      "grad_norm": 0.5157675743103027,
      "learning_rate": 3.780728520019021e-05,
      "loss": 0.5256,
      "mean_token_accuracy": 0.8338005661964416,
      "num_tokens": 427860705.0,
      "step": 12770
    },
    {
      "epoch": 0.7625955109837631,
      "grad_norm": 0.5437571406364441,
      "learning_rate": 3.7797413292379585e-05,
      "loss": 0.571,
      "mean_token_accuracy": 0.8192413210868835,
      "num_tokens": 428028385.0,
      "step": 12775
    },
    {
      "epoch": 0.7628939828080229,
      "grad_norm": 0.5211789608001709,
      "learning_rate": 3.7787538876443884e-05,
      "loss": 0.5405,
      "mean_token_accuracy": 0.8283072948455811,
      "num_tokens": 428196065.0,
      "step": 12780
    },
    {
      "epoch": 0.7631924546322827,
      "grad_norm": 0.49580714106559753,
      "learning_rate": 3.77776619547882e-05,
      "loss": 0.4809,
      "mean_token_accuracy": 0.8436836481094361,
      "num_tokens": 428363745.0,
      "step": 12785
    },
    {
      "epoch": 0.7634909264565425,
      "grad_norm": 0.5346022844314575,
      "learning_rate": 3.776778252981824e-05,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8198318123817444,
      "num_tokens": 428531425.0,
      "step": 12790
    },
    {
      "epoch": 0.7637893982808023,
      "grad_norm": 0.5726235508918762,
      "learning_rate": 3.7757900603940326e-05,
      "loss": 0.5419,
      "mean_token_accuracy": 0.8279017090797425,
      "num_tokens": 428699105.0,
      "step": 12795
    },
    {
      "epoch": 0.7640878701050621,
      "grad_norm": 0.520103394985199,
      "learning_rate": 3.7748016179561385e-05,
      "loss": 0.543,
      "mean_token_accuracy": 0.8271382570266723,
      "num_tokens": 428866785.0,
      "step": 12800
    },
    {
      "epoch": 0.7643863419293219,
      "grad_norm": 0.49786993861198425,
      "learning_rate": 3.773812925908894e-05,
      "loss": 0.5305,
      "mean_token_accuracy": 0.8287546277046204,
      "num_tokens": 429034465.0,
      "step": 12805
    },
    {
      "epoch": 0.7646848137535817,
      "grad_norm": 0.5412278771400452,
      "learning_rate": 3.772823984493114e-05,
      "loss": 0.5723,
      "mean_token_accuracy": 0.817732310295105,
      "num_tokens": 429202145.0,
      "step": 12810
    },
    {
      "epoch": 0.7649832855778415,
      "grad_norm": 0.546747088432312,
      "learning_rate": 3.7718347939496734e-05,
      "loss": 0.5753,
      "mean_token_accuracy": 0.8174162030220031,
      "num_tokens": 429369825.0,
      "step": 12815
    },
    {
      "epoch": 0.7652817574021012,
      "grad_norm": 0.5095201134681702,
      "learning_rate": 3.7708453545195085e-05,
      "loss": 0.5426,
      "mean_token_accuracy": 0.8274484157562256,
      "num_tokens": 429537505.0,
      "step": 12820
    },
    {
      "epoch": 0.765580229226361,
      "grad_norm": 0.5252211093902588,
      "learning_rate": 3.769855666443614e-05,
      "loss": 0.5833,
      "mean_token_accuracy": 0.815638792514801,
      "num_tokens": 429705185.0,
      "step": 12825
    },
    {
      "epoch": 0.7658787010506208,
      "grad_norm": 0.4762919843196869,
      "learning_rate": 3.7688657299630485e-05,
      "loss": 0.5477,
      "mean_token_accuracy": 0.8255576729774475,
      "num_tokens": 429872865.0,
      "step": 12830
    },
    {
      "epoch": 0.7661771728748806,
      "grad_norm": 0.49446505308151245,
      "learning_rate": 3.767875545318928e-05,
      "loss": 0.5466,
      "mean_token_accuracy": 0.8253489255905151,
      "num_tokens": 430040545.0,
      "step": 12835
    },
    {
      "epoch": 0.7664756446991404,
      "grad_norm": 0.5296342372894287,
      "learning_rate": 3.766885112752431e-05,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8284623742103576,
      "num_tokens": 430208225.0,
      "step": 12840
    },
    {
      "epoch": 0.7667741165234002,
      "grad_norm": 0.47552910447120667,
      "learning_rate": 3.765894432504794e-05,
      "loss": 0.53,
      "mean_token_accuracy": 0.8313968777656555,
      "num_tokens": 430375905.0,
      "step": 12845
    },
    {
      "epoch": 0.76707258834766,
      "grad_norm": 0.5044168829917908,
      "learning_rate": 3.764903504817318e-05,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8256053924560547,
      "num_tokens": 430543585.0,
      "step": 12850
    },
    {
      "epoch": 0.7673710601719198,
      "grad_norm": 0.5925907492637634,
      "learning_rate": 3.7639123299313586e-05,
      "loss": 0.5393,
      "mean_token_accuracy": 0.8274126172065734,
      "num_tokens": 430711265.0,
      "step": 12855
    },
    {
      "epoch": 0.7676695319961796,
      "grad_norm": 0.49887076020240784,
      "learning_rate": 3.7629209080883386e-05,
      "loss": 0.5537,
      "mean_token_accuracy": 0.823332941532135,
      "num_tokens": 430878945.0,
      "step": 12860
    },
    {
      "epoch": 0.7679680038204394,
      "grad_norm": 0.5620576739311218,
      "learning_rate": 3.761929239529734e-05,
      "loss": 0.5605,
      "mean_token_accuracy": 0.8211678385734558,
      "num_tokens": 431046625.0,
      "step": 12865
    },
    {
      "epoch": 0.7682664756446992,
      "grad_norm": 0.5315982103347778,
      "learning_rate": 3.760937324497086e-05,
      "loss": 0.5377,
      "mean_token_accuracy": 0.8285160303115845,
      "num_tokens": 431214305.0,
      "step": 12870
    },
    {
      "epoch": 0.768564947468959,
      "grad_norm": 0.5332516431808472,
      "learning_rate": 3.759945163231992e-05,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8358039379119873,
      "num_tokens": 431374141.0,
      "step": 12875
    },
    {
      "epoch": 0.7688634192932188,
      "grad_norm": 0.4973939061164856,
      "learning_rate": 3.7589527559761124e-05,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8283728957176208,
      "num_tokens": 431541821.0,
      "step": 12880
    },
    {
      "epoch": 0.7691618911174785,
      "grad_norm": 0.5521952509880066,
      "learning_rate": 3.757960102971165e-05,
      "loss": 0.5557,
      "mean_token_accuracy": 0.8238220334053039,
      "num_tokens": 431709501.0,
      "step": 12885
    },
    {
      "epoch": 0.7694603629417383,
      "grad_norm": 0.505549967288971,
      "learning_rate": 3.756967204458931e-05,
      "loss": 0.5522,
      "mean_token_accuracy": 0.8223487973213196,
      "num_tokens": 431877181.0,
      "step": 12890
    },
    {
      "epoch": 0.769758834765998,
      "grad_norm": 0.5413235425949097,
      "learning_rate": 3.7559740606812474e-05,
      "loss": 0.5187,
      "mean_token_accuracy": 0.8329118490219116,
      "num_tokens": 432044861.0,
      "step": 12895
    },
    {
      "epoch": 0.7700573065902578,
      "grad_norm": 0.5566238760948181,
      "learning_rate": 3.754980671880013e-05,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8259751796722412,
      "num_tokens": 432212541.0,
      "step": 12900
    },
    {
      "epoch": 0.7703557784145176,
      "grad_norm": 0.5344740152359009,
      "learning_rate": 3.753987038297187e-05,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8176667094230652,
      "num_tokens": 432380221.0,
      "step": 12905
    },
    {
      "epoch": 0.7706542502387774,
      "grad_norm": 0.5251920819282532,
      "learning_rate": 3.752993160174785e-05,
      "loss": 0.56,
      "mean_token_accuracy": 0.8227603554725647,
      "num_tokens": 432547901.0,
      "step": 12910
    },
    {
      "epoch": 0.7709527220630372,
      "grad_norm": 0.6028419137001038,
      "learning_rate": 3.751999037754886e-05,
      "loss": 0.5314,
      "mean_token_accuracy": 0.830537986755371,
      "num_tokens": 432715581.0,
      "step": 12915
    },
    {
      "epoch": 0.771251193887297,
      "grad_norm": 0.4760540723800659,
      "learning_rate": 3.751004671279627e-05,
      "loss": 0.4909,
      "mean_token_accuracy": 0.8439043283462524,
      "num_tokens": 432883261.0,
      "step": 12920
    },
    {
      "epoch": 0.7715496657115568,
      "grad_norm": 0.49214062094688416,
      "learning_rate": 3.7500100609912034e-05,
      "loss": 0.5249,
      "mean_token_accuracy": 0.8315579056739807,
      "num_tokens": 433050941.0,
      "step": 12925
    },
    {
      "epoch": 0.7718481375358166,
      "grad_norm": 0.5240420699119568,
      "learning_rate": 3.749015207131872e-05,
      "loss": 0.5502,
      "mean_token_accuracy": 0.8248479008674622,
      "num_tokens": 433218621.0,
      "step": 12930
    },
    {
      "epoch": 0.7721466093600764,
      "grad_norm": 0.5015308260917664,
      "learning_rate": 3.7480201099439476e-05,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8387510418891907,
      "num_tokens": 433386301.0,
      "step": 12935
    },
    {
      "epoch": 0.7724450811843362,
      "grad_norm": 0.5115908980369568,
      "learning_rate": 3.747024769669803e-05,
      "loss": 0.5665,
      "mean_token_accuracy": 0.8219849824905395,
      "num_tokens": 433553981.0,
      "step": 12940
    },
    {
      "epoch": 0.772743553008596,
      "grad_norm": 0.5259997248649597,
      "learning_rate": 3.7460291865518736e-05,
      "loss": 0.5739,
      "mean_token_accuracy": 0.8163485646247863,
      "num_tokens": 433721661.0,
      "step": 12945
    },
    {
      "epoch": 0.7730420248328558,
      "grad_norm": 0.4921368658542633,
      "learning_rate": 3.7450333608326506e-05,
      "loss": 0.5285,
      "mean_token_accuracy": 0.8316235303878784,
      "num_tokens": 433889341.0,
      "step": 12950
    },
    {
      "epoch": 0.7733404966571156,
      "grad_norm": 0.649025022983551,
      "learning_rate": 3.744037292754687e-05,
      "loss": 0.5689,
      "mean_token_accuracy": 0.8203984379768372,
      "num_tokens": 434057021.0,
      "step": 12955
    },
    {
      "epoch": 0.7736389684813754,
      "grad_norm": 0.505024254322052,
      "learning_rate": 3.743040982560592e-05,
      "loss": 0.5115,
      "mean_token_accuracy": 0.8339735269546509,
      "num_tokens": 434224701.0,
      "step": 12960
    },
    {
      "epoch": 0.7739374403056352,
      "grad_norm": 0.5172065496444702,
      "learning_rate": 3.742044430493037e-05,
      "loss": 0.5764,
      "mean_token_accuracy": 0.8179178953170776,
      "num_tokens": 434386695.0,
      "step": 12965
    },
    {
      "epoch": 0.774235912129895,
      "grad_norm": 0.5499746203422546,
      "learning_rate": 3.741047636794749e-05,
      "loss": 0.5324,
      "mean_token_accuracy": 0.831015145778656,
      "num_tokens": 434554375.0,
      "step": 12970
    },
    {
      "epoch": 0.7745343839541547,
      "grad_norm": 0.5083252191543579,
      "learning_rate": 3.740050601708518e-05,
      "loss": 0.5371,
      "mean_token_accuracy": 0.8301860928535462,
      "num_tokens": 434722055.0,
      "step": 12975
    },
    {
      "epoch": 0.7748328557784145,
      "grad_norm": 0.5025472640991211,
      "learning_rate": 3.739053325477187e-05,
      "loss": 0.5733,
      "mean_token_accuracy": 0.8194799065589905,
      "num_tokens": 434889735.0,
      "step": 12980
    },
    {
      "epoch": 0.7751313276026743,
      "grad_norm": 0.5539953708648682,
      "learning_rate": 3.738055808343664e-05,
      "loss": 0.5595,
      "mean_token_accuracy": 0.8211618781089782,
      "num_tokens": 435057415.0,
      "step": 12985
    },
    {
      "epoch": 0.7754297994269341,
      "grad_norm": 0.5148086547851562,
      "learning_rate": 3.73705805055091e-05,
      "loss": 0.5403,
      "mean_token_accuracy": 0.8274722695350647,
      "num_tokens": 435225095.0,
      "step": 12990
    },
    {
      "epoch": 0.7757282712511939,
      "grad_norm": 0.45913150906562805,
      "learning_rate": 3.7360600523419506e-05,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8391089081764221,
      "num_tokens": 435392775.0,
      "step": 12995
    },
    {
      "epoch": 0.7760267430754537,
      "grad_norm": 0.5459899306297302,
      "learning_rate": 3.735061813959864e-05,
      "loss": 0.5528,
      "mean_token_accuracy": 0.824460220336914,
      "num_tokens": 435560455.0,
      "step": 13000
    },
    {
      "epoch": 0.7763252148997135,
      "grad_norm": 0.4576101005077362,
      "learning_rate": 3.734063335647791e-05,
      "loss": 0.5177,
      "mean_token_accuracy": 0.8360193133354187,
      "num_tokens": 435728135.0,
      "step": 13005
    },
    {
      "epoch": 0.7766236867239733,
      "grad_norm": 0.49657967686653137,
      "learning_rate": 3.733064617648927e-05,
      "loss": 0.5507,
      "mean_token_accuracy": 0.825676965713501,
      "num_tokens": 435895815.0,
      "step": 13010
    },
    {
      "epoch": 0.7769221585482331,
      "grad_norm": 0.5331529974937439,
      "learning_rate": 3.732065660206532e-05,
      "loss": 0.5531,
      "mean_token_accuracy": 0.8237802624702454,
      "num_tokens": 436063495.0,
      "step": 13015
    },
    {
      "epoch": 0.7772206303724928,
      "grad_norm": 0.4832727313041687,
      "learning_rate": 3.731066463563919e-05,
      "loss": 0.4851,
      "mean_token_accuracy": 0.8433019161224365,
      "num_tokens": 436231175.0,
      "step": 13020
    },
    {
      "epoch": 0.7775191021967526,
      "grad_norm": 0.5183764100074768,
      "learning_rate": 3.730067027964459e-05,
      "loss": 0.5289,
      "mean_token_accuracy": 0.8313252925872803,
      "num_tokens": 436398855.0,
      "step": 13025
    },
    {
      "epoch": 0.7778175740210124,
      "grad_norm": 0.5579322576522827,
      "learning_rate": 3.729067353651585e-05,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8266432166099549,
      "num_tokens": 436566535.0,
      "step": 13030
    },
    {
      "epoch": 0.7781160458452722,
      "grad_norm": 0.5010125041007996,
      "learning_rate": 3.7280674408687847e-05,
      "loss": 0.6078,
      "mean_token_accuracy": 0.8086782813072204,
      "num_tokens": 436734215.0,
      "step": 13035
    },
    {
      "epoch": 0.778414517669532,
      "grad_norm": 0.542412519454956,
      "learning_rate": 3.727067289859607e-05,
      "loss": 0.5562,
      "mean_token_accuracy": 0.8254025936126709,
      "num_tokens": 436901895.0,
      "step": 13040
    },
    {
      "epoch": 0.7787129894937918,
      "grad_norm": 0.538048505783081,
      "learning_rate": 3.726066900867656e-05,
      "loss": 0.499,
      "mean_token_accuracy": 0.838470721244812,
      "num_tokens": 437069575.0,
      "step": 13045
    },
    {
      "epoch": 0.7790114613180515,
      "grad_norm": 0.5493173599243164,
      "learning_rate": 3.7250662741365955e-05,
      "loss": 0.5265,
      "mean_token_accuracy": 0.8312716126441956,
      "num_tokens": 437237255.0,
      "step": 13050
    },
    {
      "epoch": 0.7793099331423113,
      "grad_norm": 0.4959273636341095,
      "learning_rate": 3.7240654099101464e-05,
      "loss": 0.4938,
      "mean_token_accuracy": 0.8409101843833924,
      "num_tokens": 437404935.0,
      "step": 13055
    },
    {
      "epoch": 0.7796084049665711,
      "grad_norm": 0.5687335133552551,
      "learning_rate": 3.723064308432086e-05,
      "loss": 0.5307,
      "mean_token_accuracy": 0.8290289878845215,
      "num_tokens": 437572615.0,
      "step": 13060
    },
    {
      "epoch": 0.7799068767908309,
      "grad_norm": 0.5093406438827515,
      "learning_rate": 3.722062969946254e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8340033411979675,
      "num_tokens": 437740295.0,
      "step": 13065
    },
    {
      "epoch": 0.7802053486150907,
      "grad_norm": 0.5641523599624634,
      "learning_rate": 3.7210613946965435e-05,
      "loss": 0.5554,
      "mean_token_accuracy": 0.8209879159927368,
      "num_tokens": 437901833.0,
      "step": 13070
    },
    {
      "epoch": 0.7805038204393505,
      "grad_norm": 0.4834235608577728,
      "learning_rate": 3.7200595829269064e-05,
      "loss": 0.5262,
      "mean_token_accuracy": 0.8319456100463867,
      "num_tokens": 438069513.0,
      "step": 13075
    },
    {
      "epoch": 0.7808022922636103,
      "grad_norm": 0.47315308451652527,
      "learning_rate": 3.719057534881353e-05,
      "loss": 0.568,
      "mean_token_accuracy": 0.8208159327507019,
      "num_tokens": 438237193.0,
      "step": 13080
    },
    {
      "epoch": 0.7811007640878701,
      "grad_norm": 0.565412163734436,
      "learning_rate": 3.718055250803949e-05,
      "loss": 0.5579,
      "mean_token_accuracy": 0.8213945031166077,
      "num_tokens": 438404873.0,
      "step": 13085
    },
    {
      "epoch": 0.7813992359121299,
      "grad_norm": 0.4810662567615509,
      "learning_rate": 3.717052730938821e-05,
      "loss": 0.5718,
      "mean_token_accuracy": 0.8193546414375306,
      "num_tokens": 438572553.0,
      "step": 13090
    },
    {
      "epoch": 0.7816977077363897,
      "grad_norm": 0.5309857726097107,
      "learning_rate": 3.71604997553015e-05,
      "loss": 0.5263,
      "mean_token_accuracy": 0.8315877318382263,
      "num_tokens": 438740233.0,
      "step": 13095
    },
    {
      "epoch": 0.7819961795606495,
      "grad_norm": 0.5122263431549072,
      "learning_rate": 3.715046984822177e-05,
      "loss": 0.5463,
      "mean_token_accuracy": 0.8265000581741333,
      "num_tokens": 438907913.0,
      "step": 13100
    },
    {
      "epoch": 0.7822946513849093,
      "grad_norm": 0.4956192374229431,
      "learning_rate": 3.714043759059198e-05,
      "loss": 0.5092,
      "mean_token_accuracy": 0.837593936920166,
      "num_tokens": 439075593.0,
      "step": 13105
    },
    {
      "epoch": 0.7825931232091691,
      "grad_norm": 0.5619259476661682,
      "learning_rate": 3.7130402984855675e-05,
      "loss": 0.5796,
      "mean_token_accuracy": 0.8169509768486023,
      "num_tokens": 439243273.0,
      "step": 13110
    },
    {
      "epoch": 0.7828915950334289,
      "grad_norm": 0.505506157875061,
      "learning_rate": 3.712036603345697e-05,
      "loss": 0.5013,
      "mean_token_accuracy": 0.8392654895782471,
      "num_tokens": 439404835.0,
      "step": 13115
    },
    {
      "epoch": 0.7831900668576887,
      "grad_norm": 0.4841116964817047,
      "learning_rate": 3.711032673884054e-05,
      "loss": 0.5672,
      "mean_token_accuracy": 0.8238651037216187,
      "num_tokens": 439565475.0,
      "step": 13120
    },
    {
      "epoch": 0.7834885386819485,
      "grad_norm": 0.48040181398391724,
      "learning_rate": 3.7100285103451644e-05,
      "loss": 0.528,
      "mean_token_accuracy": 0.8303113341331482,
      "num_tokens": 439733155.0,
      "step": 13125
    },
    {
      "epoch": 0.7837870105062082,
      "grad_norm": 0.530348002910614,
      "learning_rate": 3.709024112973611e-05,
      "loss": 0.5638,
      "mean_token_accuracy": 0.8213169574737549,
      "num_tokens": 439900835.0,
      "step": 13130
    },
    {
      "epoch": 0.784085482330468,
      "grad_norm": 0.5174166560173035,
      "learning_rate": 3.708019482014034e-05,
      "loss": 0.5284,
      "mean_token_accuracy": 0.8310509324073792,
      "num_tokens": 440068515.0,
      "step": 13135
    },
    {
      "epoch": 0.7843839541547278,
      "grad_norm": 0.5310731530189514,
      "learning_rate": 3.707014617711129e-05,
      "loss": 0.5443,
      "mean_token_accuracy": 0.8263211250305176,
      "num_tokens": 440236195.0,
      "step": 13140
    },
    {
      "epoch": 0.7846824259789876,
      "grad_norm": 0.5151842832565308,
      "learning_rate": 3.706009520309649e-05,
      "loss": 0.5516,
      "mean_token_accuracy": 0.8239114880561829,
      "num_tokens": 440403875.0,
      "step": 13145
    },
    {
      "epoch": 0.7849808978032474,
      "grad_norm": 0.6412427425384521,
      "learning_rate": 3.705004190054405e-05,
      "loss": 0.5855,
      "mean_token_accuracy": 0.8147441029548645,
      "num_tokens": 440571555.0,
      "step": 13150
    },
    {
      "epoch": 0.7852793696275072,
      "grad_norm": 0.4924268424510956,
      "learning_rate": 3.7039986271902624e-05,
      "loss": 0.5108,
      "mean_token_accuracy": 0.8361803650856018,
      "num_tokens": 440739235.0,
      "step": 13155
    },
    {
      "epoch": 0.7855778414517669,
      "grad_norm": 0.518210768699646,
      "learning_rate": 3.702992831962145e-05,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8306095719337463,
      "num_tokens": 440906915.0,
      "step": 13160
    },
    {
      "epoch": 0.7858763132760267,
      "grad_norm": 0.5177653431892395,
      "learning_rate": 3.701986804615034e-05,
      "loss": 0.5405,
      "mean_token_accuracy": 0.8298103213310242,
      "num_tokens": 441074595.0,
      "step": 13165
    },
    {
      "epoch": 0.7861747851002865,
      "grad_norm": 0.5237435102462769,
      "learning_rate": 3.700980545393965e-05,
      "loss": 0.5436,
      "mean_token_accuracy": 0.82692950963974,
      "num_tokens": 441242275.0,
      "step": 13170
    },
    {
      "epoch": 0.7864732569245463,
      "grad_norm": 0.4853658974170685,
      "learning_rate": 3.69997405454403e-05,
      "loss": 0.5055,
      "mean_token_accuracy": 0.8395502805709839,
      "num_tokens": 441409955.0,
      "step": 13175
    },
    {
      "epoch": 0.7867717287488061,
      "grad_norm": 0.5364872813224792,
      "learning_rate": 3.698967332310381e-05,
      "loss": 0.5577,
      "mean_token_accuracy": 0.8237385392189026,
      "num_tokens": 441577635.0,
      "step": 13180
    },
    {
      "epoch": 0.7870702005730659,
      "grad_norm": 0.5034254193305969,
      "learning_rate": 3.697960378938222e-05,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8354348063468933,
      "num_tokens": 441745315.0,
      "step": 13185
    },
    {
      "epoch": 0.7873686723973257,
      "grad_norm": 0.5822336673736572,
      "learning_rate": 3.696953194672815e-05,
      "loss": 0.574,
      "mean_token_accuracy": 0.8178873777389526,
      "num_tokens": 441912995.0,
      "step": 13190
    },
    {
      "epoch": 0.7876671442215855,
      "grad_norm": 0.5720198154449463,
      "learning_rate": 3.695945779759479e-05,
      "loss": 0.5459,
      "mean_token_accuracy": 0.8244960069656372,
      "num_tokens": 442080675.0,
      "step": 13195
    },
    {
      "epoch": 0.7879656160458453,
      "grad_norm": 0.49734410643577576,
      "learning_rate": 3.694938134443588e-05,
      "loss": 0.5094,
      "mean_token_accuracy": 0.8362841367721557,
      "num_tokens": 442248281.0,
      "step": 13200
    },
    {
      "epoch": 0.788264087870105,
      "grad_norm": 0.5321495532989502,
      "learning_rate": 3.693930258970573e-05,
      "loss": 0.5566,
      "mean_token_accuracy": 0.822557556629181,
      "num_tokens": 442415961.0,
      "step": 13205
    },
    {
      "epoch": 0.7885625596943648,
      "grad_norm": 0.48074623942375183,
      "learning_rate": 3.69292215358592e-05,
      "loss": 0.5189,
      "mean_token_accuracy": 0.8337528347969055,
      "num_tokens": 442583641.0,
      "step": 13210
    },
    {
      "epoch": 0.7888610315186246,
      "grad_norm": 0.4866797626018524,
      "learning_rate": 3.691913818535172e-05,
      "loss": 0.5478,
      "mean_token_accuracy": 0.8245735406875611,
      "num_tokens": 442751321.0,
      "step": 13215
    },
    {
      "epoch": 0.7891595033428844,
      "grad_norm": 0.541733980178833,
      "learning_rate": 3.690905254063928e-05,
      "loss": 0.516,
      "mean_token_accuracy": 0.834259819984436,
      "num_tokens": 442919001.0,
      "step": 13220
    },
    {
      "epoch": 0.7894579751671442,
      "grad_norm": 0.49007222056388855,
      "learning_rate": 3.6898964604178425e-05,
      "loss": 0.5126,
      "mean_token_accuracy": 0.8356495261192322,
      "num_tokens": 443086681.0,
      "step": 13225
    },
    {
      "epoch": 0.789756446991404,
      "grad_norm": 0.5745511651039124,
      "learning_rate": 3.688887437842626e-05,
      "loss": 0.5352,
      "mean_token_accuracy": 0.828885841369629,
      "num_tokens": 443254361.0,
      "step": 13230
    },
    {
      "epoch": 0.7900549188156638,
      "grad_norm": 0.5187568664550781,
      "learning_rate": 3.687878186584043e-05,
      "loss": 0.5708,
      "mean_token_accuracy": 0.818859589099884,
      "num_tokens": 443422041.0,
      "step": 13235
    },
    {
      "epoch": 0.7903533906399236,
      "grad_norm": 0.5131155848503113,
      "learning_rate": 3.6868687068879174e-05,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8292616009712219,
      "num_tokens": 443589721.0,
      "step": 13240
    },
    {
      "epoch": 0.7906518624641834,
      "grad_norm": 0.5099996328353882,
      "learning_rate": 3.6858589990001256e-05,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8425265431404114,
      "num_tokens": 443757401.0,
      "step": 13245
    },
    {
      "epoch": 0.7909503342884432,
      "grad_norm": 0.6119116544723511,
      "learning_rate": 3.6848490631666e-05,
      "loss": 0.517,
      "mean_token_accuracy": 0.8364189386367797,
      "num_tokens": 443925081.0,
      "step": 13250
    },
    {
      "epoch": 0.791248806112703,
      "grad_norm": 0.5421985983848572,
      "learning_rate": 3.68383889963333e-05,
      "loss": 0.5515,
      "mean_token_accuracy": 0.8247405409812927,
      "num_tokens": 444092761.0,
      "step": 13255
    },
    {
      "epoch": 0.7915472779369628,
      "grad_norm": 0.522763729095459,
      "learning_rate": 3.682828508646359e-05,
      "loss": 0.5229,
      "mean_token_accuracy": 0.8328343152999877,
      "num_tokens": 444260441.0,
      "step": 13260
    },
    {
      "epoch": 0.7918457497612226,
      "grad_norm": 0.5403424501419067,
      "learning_rate": 3.6818178904517874e-05,
      "loss": 0.569,
      "mean_token_accuracy": 0.8193785071372985,
      "num_tokens": 444428121.0,
      "step": 13265
    },
    {
      "epoch": 0.7921442215854824,
      "grad_norm": 0.5987299084663391,
      "learning_rate": 3.680807045295769e-05,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8293868660926819,
      "num_tokens": 444595801.0,
      "step": 13270
    },
    {
      "epoch": 0.7924426934097422,
      "grad_norm": 0.4669605791568756,
      "learning_rate": 3.6797959734245134e-05,
      "loss": 0.5191,
      "mean_token_accuracy": 0.8352439403533936,
      "num_tokens": 444763481.0,
      "step": 13275
    },
    {
      "epoch": 0.792741165234002,
      "grad_norm": 0.48344823718070984,
      "learning_rate": 3.678784675084287e-05,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8264821529388428,
      "num_tokens": 444931161.0,
      "step": 13280
    },
    {
      "epoch": 0.7930396370582617,
      "grad_norm": 0.5072746276855469,
      "learning_rate": 3.6777731505214086e-05,
      "loss": 0.5407,
      "mean_token_accuracy": 0.8270130038261414,
      "num_tokens": 445098841.0,
      "step": 13285
    },
    {
      "epoch": 0.7933381088825215,
      "grad_norm": 0.4818067252635956,
      "learning_rate": 3.676761399982255e-05,
      "loss": 0.5083,
      "mean_token_accuracy": 0.8379756450653076,
      "num_tokens": 445266521.0,
      "step": 13290
    },
    {
      "epoch": 0.7936365807067812,
      "grad_norm": 0.48458853363990784,
      "learning_rate": 3.675749423713257e-05,
      "loss": 0.5707,
      "mean_token_accuracy": 0.8203864932060242,
      "num_tokens": 445434201.0,
      "step": 13295
    },
    {
      "epoch": 0.793935052531041,
      "grad_norm": 0.5793253183364868,
      "learning_rate": 3.6747372219608996e-05,
      "loss": 0.6197,
      "mean_token_accuracy": 0.8065430164337158,
      "num_tokens": 445601881.0,
      "step": 13300
    },
    {
      "epoch": 0.7942335243553008,
      "grad_norm": 0.47477298974990845,
      "learning_rate": 3.6737247949717227e-05,
      "loss": 0.5257,
      "mean_token_accuracy": 0.8313610911369324,
      "num_tokens": 445769561.0,
      "step": 13305
    },
    {
      "epoch": 0.7945319961795606,
      "grad_norm": 0.5024024248123169,
      "learning_rate": 3.6727121429923214e-05,
      "loss": 0.5567,
      "mean_token_accuracy": 0.8228617548942566,
      "num_tokens": 445937241.0,
      "step": 13310
    },
    {
      "epoch": 0.7948304680038204,
      "grad_norm": 0.6198261380195618,
      "learning_rate": 3.671699266269347e-05,
      "loss": 0.5222,
      "mean_token_accuracy": 0.8339615821838379,
      "num_tokens": 446104921.0,
      "step": 13315
    },
    {
      "epoch": 0.7951289398280802,
      "grad_norm": 0.4861522912979126,
      "learning_rate": 3.6706861650495045e-05,
      "loss": 0.5139,
      "mean_token_accuracy": 0.8368066310882568,
      "num_tokens": 446272601.0,
      "step": 13320
    },
    {
      "epoch": 0.79542741165234,
      "grad_norm": 0.6107427477836609,
      "learning_rate": 3.669672839579552e-05,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8310211181640625,
      "num_tokens": 446440281.0,
      "step": 13325
    },
    {
      "epoch": 0.7957258834765998,
      "grad_norm": 0.5753244757652283,
      "learning_rate": 3.668659290106303e-05,
      "loss": 0.5898,
      "mean_token_accuracy": 0.8134498476982117,
      "num_tokens": 446607961.0,
      "step": 13330
    },
    {
      "epoch": 0.7960243553008596,
      "grad_norm": 0.5145120024681091,
      "learning_rate": 3.667645516876629e-05,
      "loss": 0.5518,
      "mean_token_accuracy": 0.8248956084251404,
      "num_tokens": 446775641.0,
      "step": 13335
    },
    {
      "epoch": 0.7963228271251194,
      "grad_norm": 0.499484121799469,
      "learning_rate": 3.66663152013745e-05,
      "loss": 0.5194,
      "mean_token_accuracy": 0.8345461130142212,
      "num_tokens": 446943321.0,
      "step": 13340
    },
    {
      "epoch": 0.7966212989493792,
      "grad_norm": 0.4748535752296448,
      "learning_rate": 3.665617300135747e-05,
      "loss": 0.54,
      "mean_token_accuracy": 0.8282595753669739,
      "num_tokens": 447111001.0,
      "step": 13345
    },
    {
      "epoch": 0.796919770773639,
      "grad_norm": 0.5834560990333557,
      "learning_rate": 3.6646028571185485e-05,
      "loss": 0.5643,
      "mean_token_accuracy": 0.8213467597961426,
      "num_tokens": 447278681.0,
      "step": 13350
    },
    {
      "epoch": 0.7972182425978988,
      "grad_norm": 0.5450048446655273,
      "learning_rate": 3.663588191332944e-05,
      "loss": 0.5773,
      "mean_token_accuracy": 0.8185434818267823,
      "num_tokens": 447446361.0,
      "step": 13355
    },
    {
      "epoch": 0.7975167144221585,
      "grad_norm": 0.5127323865890503,
      "learning_rate": 3.662573303026072e-05,
      "loss": 0.556,
      "mean_token_accuracy": 0.8222473978996276,
      "num_tokens": 447614041.0,
      "step": 13360
    },
    {
      "epoch": 0.7978151862464183,
      "grad_norm": 0.536323606967926,
      "learning_rate": 3.6615581924451274e-05,
      "loss": 0.5421,
      "mean_token_accuracy": 0.8263986706733704,
      "num_tokens": 447781721.0,
      "step": 13365
    },
    {
      "epoch": 0.7981136580706781,
      "grad_norm": 0.5523877143859863,
      "learning_rate": 3.66054285983736e-05,
      "loss": 0.5522,
      "mean_token_accuracy": 0.8237385392189026,
      "num_tokens": 447949401.0,
      "step": 13370
    },
    {
      "epoch": 0.7984121298949379,
      "grad_norm": 0.6258491277694702,
      "learning_rate": 3.659527305450072e-05,
      "loss": 0.57,
      "mean_token_accuracy": 0.8175951361656189,
      "num_tokens": 448117081.0,
      "step": 13375
    },
    {
      "epoch": 0.7987106017191977,
      "grad_norm": 0.5400011539459229,
      "learning_rate": 3.658511529530622e-05,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8227245688438416,
      "num_tokens": 448284761.0,
      "step": 13380
    },
    {
      "epoch": 0.7990090735434575,
      "grad_norm": 0.5358766913414001,
      "learning_rate": 3.657495532326419e-05,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8327269434928894,
      "num_tokens": 448452441.0,
      "step": 13385
    },
    {
      "epoch": 0.7993075453677173,
      "grad_norm": 0.5900395512580872,
      "learning_rate": 3.656479314084928e-05,
      "loss": 0.5593,
      "mean_token_accuracy": 0.8233150482177735,
      "num_tokens": 448620121.0,
      "step": 13390
    },
    {
      "epoch": 0.7996060171919771,
      "grad_norm": 0.5061820149421692,
      "learning_rate": 3.655462875053669e-05,
      "loss": 0.5324,
      "mean_token_accuracy": 0.8286293625831604,
      "num_tokens": 448787801.0,
      "step": 13395
    },
    {
      "epoch": 0.7999044890162369,
      "grad_norm": 0.567453145980835,
      "learning_rate": 3.654446215480214e-05,
      "loss": 0.6078,
      "mean_token_accuracy": 0.8083919763565064,
      "num_tokens": 448955481.0,
      "step": 13400
    },
    {
      "epoch": 0.8002029608404967,
      "grad_norm": 0.5236828923225403,
      "learning_rate": 3.6534293356121885e-05,
      "loss": 0.5501,
      "mean_token_accuracy": 0.8239234089851379,
      "num_tokens": 449123161.0,
      "step": 13405
    },
    {
      "epoch": 0.8005014326647565,
      "grad_norm": 0.5467981100082397,
      "learning_rate": 3.652412235697274e-05,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8145711541175842,
      "num_tokens": 449290841.0,
      "step": 13410
    },
    {
      "epoch": 0.8007999044890163,
      "grad_norm": 0.536647617816925,
      "learning_rate": 3.6513949159832024e-05,
      "loss": 0.5302,
      "mean_token_accuracy": 0.8306453704833985,
      "num_tokens": 449458521.0,
      "step": 13415
    },
    {
      "epoch": 0.8010983763132761,
      "grad_norm": 0.49470803141593933,
      "learning_rate": 3.650377376717761e-05,
      "loss": 0.5751,
      "mean_token_accuracy": 0.8175891757011413,
      "num_tokens": 449626201.0,
      "step": 13420
    },
    {
      "epoch": 0.8013968481375359,
      "grad_norm": 0.5080121159553528,
      "learning_rate": 3.6493596181487896e-05,
      "loss": 0.4775,
      "mean_token_accuracy": 0.8460634589195252,
      "num_tokens": 449793881.0,
      "step": 13425
    },
    {
      "epoch": 0.8016953199617957,
      "grad_norm": 0.5268555879592896,
      "learning_rate": 3.648341640524184e-05,
      "loss": 0.5126,
      "mean_token_accuracy": 0.8359537243843078,
      "num_tokens": 449961561.0,
      "step": 13430
    },
    {
      "epoch": 0.8019937917860553,
      "grad_norm": 0.5904242396354675,
      "learning_rate": 3.64732344409189e-05,
      "loss": 0.5912,
      "mean_token_accuracy": 0.8128235816955567,
      "num_tokens": 450129241.0,
      "step": 13435
    },
    {
      "epoch": 0.8022922636103151,
      "grad_norm": 0.5389574766159058,
      "learning_rate": 3.646305029099909e-05,
      "loss": 0.54,
      "mean_token_accuracy": 0.8272456169128418,
      "num_tokens": 450296921.0,
      "step": 13440
    },
    {
      "epoch": 0.8025907354345749,
      "grad_norm": 0.5236079096794128,
      "learning_rate": 3.645286395796293e-05,
      "loss": 0.5491,
      "mean_token_accuracy": 0.8264761924743652,
      "num_tokens": 450464601.0,
      "step": 13445
    },
    {
      "epoch": 0.8028892072588347,
      "grad_norm": 0.47309550642967224,
      "learning_rate": 3.6442675444291516e-05,
      "loss": 0.4904,
      "mean_token_accuracy": 0.8411547183990479,
      "num_tokens": 450632281.0,
      "step": 13450
    },
    {
      "epoch": 0.8031876790830945,
      "grad_norm": 0.5088681578636169,
      "learning_rate": 3.6432484752466425e-05,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8232971549034118,
      "num_tokens": 450799961.0,
      "step": 13455
    },
    {
      "epoch": 0.8034861509073543,
      "grad_norm": 0.6903104186058044,
      "learning_rate": 3.64222918849698e-05,
      "loss": 0.5596,
      "mean_token_accuracy": 0.8209889173507691,
      "num_tokens": 450967641.0,
      "step": 13460
    },
    {
      "epoch": 0.8037846227316141,
      "grad_norm": 0.5452691316604614,
      "learning_rate": 3.6412096844284304e-05,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8408445596694947,
      "num_tokens": 451135321.0,
      "step": 13465
    },
    {
      "epoch": 0.8040830945558739,
      "grad_norm": 0.5216230154037476,
      "learning_rate": 3.640189963289312e-05,
      "loss": 0.5723,
      "mean_token_accuracy": 0.8172849655151367,
      "num_tokens": 451303001.0,
      "step": 13470
    },
    {
      "epoch": 0.8043815663801337,
      "grad_norm": 0.5238450169563293,
      "learning_rate": 3.639170025327997e-05,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8273768305778504,
      "num_tokens": 451470681.0,
      "step": 13475
    },
    {
      "epoch": 0.8046800382043935,
      "grad_norm": 0.5196294188499451,
      "learning_rate": 3.63814987079291e-05,
      "loss": 0.5122,
      "mean_token_accuracy": 0.8351186871528625,
      "num_tokens": 451638361.0,
      "step": 13480
    },
    {
      "epoch": 0.8049785100286533,
      "grad_norm": 0.5821755528450012,
      "learning_rate": 3.637129499932529e-05,
      "loss": 0.5896,
      "mean_token_accuracy": 0.8135810613632202,
      "num_tokens": 451806041.0,
      "step": 13485
    },
    {
      "epoch": 0.8052769818529131,
      "grad_norm": 0.5047705173492432,
      "learning_rate": 3.636108912995384e-05,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8399200797080993,
      "num_tokens": 451973721.0,
      "step": 13490
    },
    {
      "epoch": 0.8055754536771729,
      "grad_norm": 0.5244894623756409,
      "learning_rate": 3.6350881102300566e-05,
      "loss": 0.523,
      "mean_token_accuracy": 0.833120596408844,
      "num_tokens": 452141401.0,
      "step": 13495
    },
    {
      "epoch": 0.8058739255014327,
      "grad_norm": 0.4949883222579956,
      "learning_rate": 3.6340670918851835e-05,
      "loss": 0.5102,
      "mean_token_accuracy": 0.8356793522834778,
      "num_tokens": 452309081.0,
      "step": 13500
    },
    {
      "epoch": 0.8061723973256925,
      "grad_norm": 0.5216145515441895,
      "learning_rate": 3.6330458582094525e-05,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8228273987770081,
      "num_tokens": 452469604.0,
      "step": 13505
    },
    {
      "epoch": 0.8064708691499523,
      "grad_norm": 0.4776259958744049,
      "learning_rate": 3.6320244094516035e-05,
      "loss": 0.4624,
      "mean_token_accuracy": 0.849761426448822,
      "num_tokens": 452637284.0,
      "step": 13510
    },
    {
      "epoch": 0.806769340974212,
      "grad_norm": 0.5067180395126343,
      "learning_rate": 3.631002745860429e-05,
      "loss": 0.555,
      "mean_token_accuracy": 0.8254324197769165,
      "num_tokens": 452804964.0,
      "step": 13515
    },
    {
      "epoch": 0.8070678127984718,
      "grad_norm": 0.48812413215637207,
      "learning_rate": 3.629980867684775e-05,
      "loss": 0.5341,
      "mean_token_accuracy": 0.8304604530334473,
      "num_tokens": 452972644.0,
      "step": 13520
    },
    {
      "epoch": 0.8073662846227316,
      "grad_norm": 0.4747433662414551,
      "learning_rate": 3.628958775173537e-05,
      "loss": 0.5353,
      "mean_token_accuracy": 0.8279315114021302,
      "num_tokens": 453140324.0,
      "step": 13525
    },
    {
      "epoch": 0.8076647564469914,
      "grad_norm": 0.559675931930542,
      "learning_rate": 3.627936468575665e-05,
      "loss": 0.5505,
      "mean_token_accuracy": 0.8259036064147949,
      "num_tokens": 453308004.0,
      "step": 13530
    },
    {
      "epoch": 0.8079632282712512,
      "grad_norm": 0.4890124499797821,
      "learning_rate": 3.626913948140163e-05,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8287188410758972,
      "num_tokens": 453475684.0,
      "step": 13535
    },
    {
      "epoch": 0.808261700095511,
      "grad_norm": 0.45533356070518494,
      "learning_rate": 3.6258912141160825e-05,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8260825514793396,
      "num_tokens": 453643364.0,
      "step": 13540
    },
    {
      "epoch": 0.8085601719197708,
      "grad_norm": 0.5417728424072266,
      "learning_rate": 3.62486826675253e-05,
      "loss": 0.6006,
      "mean_token_accuracy": 0.8089944005012513,
      "num_tokens": 453811044.0,
      "step": 13545
    },
    {
      "epoch": 0.8088586437440306,
      "grad_norm": 0.454820454120636,
      "learning_rate": 3.6238451062986625e-05,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8369318962097168,
      "num_tokens": 453978724.0,
      "step": 13550
    },
    {
      "epoch": 0.8091571155682904,
      "grad_norm": 0.5340718626976013,
      "learning_rate": 3.6228217330036894e-05,
      "loss": 0.5442,
      "mean_token_accuracy": 0.8275259494781494,
      "num_tokens": 454146404.0,
      "step": 13555
    },
    {
      "epoch": 0.8094555873925502,
      "grad_norm": 0.5246963500976562,
      "learning_rate": 3.621798147116873e-05,
      "loss": 0.533,
      "mean_token_accuracy": 0.8294703722000122,
      "num_tokens": 454314084.0,
      "step": 13560
    },
    {
      "epoch": 0.80975405921681,
      "grad_norm": 0.5007262825965881,
      "learning_rate": 3.620774348887527e-05,
      "loss": 0.5325,
      "mean_token_accuracy": 0.8301622271537781,
      "num_tokens": 454481764.0,
      "step": 13565
    },
    {
      "epoch": 0.8100525310410697,
      "grad_norm": 0.461511492729187,
      "learning_rate": 3.619750338565015e-05,
      "loss": 0.5062,
      "mean_token_accuracy": 0.8373911499977111,
      "num_tokens": 454649444.0,
      "step": 13570
    },
    {
      "epoch": 0.8103510028653295,
      "grad_norm": 0.5453050136566162,
      "learning_rate": 3.618726116398755e-05,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8277346968650818,
      "num_tokens": 454817124.0,
      "step": 13575
    },
    {
      "epoch": 0.8106494746895893,
      "grad_norm": 0.48418128490448,
      "learning_rate": 3.617701682638214e-05,
      "loss": 0.5479,
      "mean_token_accuracy": 0.8239711165428162,
      "num_tokens": 454984804.0,
      "step": 13580
    },
    {
      "epoch": 0.810947946513849,
      "grad_norm": 1.3153331279754639,
      "learning_rate": 3.616677037532912e-05,
      "loss": 0.5717,
      "mean_token_accuracy": 0.8217762231826782,
      "num_tokens": 455152484.0,
      "step": 13585
    },
    {
      "epoch": 0.8112464183381088,
      "grad_norm": 1.1322860717773438,
      "learning_rate": 3.615652181332421e-05,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8178814291954041,
      "num_tokens": 455320164.0,
      "step": 13590
    },
    {
      "epoch": 0.8115448901623686,
      "grad_norm": 0.4932667911052704,
      "learning_rate": 3.614627114286364e-05,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8264702439308167,
      "num_tokens": 455487844.0,
      "step": 13595
    },
    {
      "epoch": 0.8118433619866284,
      "grad_norm": 0.561107873916626,
      "learning_rate": 3.613601836644412e-05,
      "loss": 0.5329,
      "mean_token_accuracy": 0.8300489068031311,
      "num_tokens": 455655524.0,
      "step": 13600
    },
    {
      "epoch": 0.8121418338108882,
      "grad_norm": 0.534894585609436,
      "learning_rate": 3.612576348656294e-05,
      "loss": 0.5242,
      "mean_token_accuracy": 0.8313491702079773,
      "num_tokens": 455823204.0,
      "step": 13605
    },
    {
      "epoch": 0.812440305635148,
      "grad_norm": 0.5064579248428345,
      "learning_rate": 3.611550650571784e-05,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8381247878074646,
      "num_tokens": 455990884.0,
      "step": 13610
    },
    {
      "epoch": 0.8127387774594078,
      "grad_norm": 0.6063612103462219,
      "learning_rate": 3.6105247426407116e-05,
      "loss": 0.5776,
      "mean_token_accuracy": 0.8201061725616455,
      "num_tokens": 456158564.0,
      "step": 13615
    },
    {
      "epoch": 0.8130372492836676,
      "grad_norm": 0.548568069934845,
      "learning_rate": 3.609498625112954e-05,
      "loss": 0.5621,
      "mean_token_accuracy": 0.8212215185165406,
      "num_tokens": 456326244.0,
      "step": 13620
    },
    {
      "epoch": 0.8133357211079274,
      "grad_norm": 0.4991692900657654,
      "learning_rate": 3.6084722982384425e-05,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8415066242218018,
      "num_tokens": 456493924.0,
      "step": 13625
    },
    {
      "epoch": 0.8136341929321872,
      "grad_norm": 0.5445214509963989,
      "learning_rate": 3.607445762267156e-05,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8310449838638305,
      "num_tokens": 456661604.0,
      "step": 13630
    },
    {
      "epoch": 0.813932664756447,
      "grad_norm": 0.4892699718475342,
      "learning_rate": 3.6064190174491286e-05,
      "loss": 0.5339,
      "mean_token_accuracy": 0.8288082957267762,
      "num_tokens": 456829284.0,
      "step": 13635
    },
    {
      "epoch": 0.8142311365807068,
      "grad_norm": 0.4974958598613739,
      "learning_rate": 3.6053920640344404e-05,
      "loss": 0.5422,
      "mean_token_accuracy": 0.8274364829063415,
      "num_tokens": 456996964.0,
      "step": 13640
    },
    {
      "epoch": 0.8145296084049666,
      "grad_norm": 0.5441718697547913,
      "learning_rate": 3.604364902273227e-05,
      "loss": 0.5087,
      "mean_token_accuracy": 0.8363354325294494,
      "num_tokens": 457164644.0,
      "step": 13645
    },
    {
      "epoch": 0.8148280802292264,
      "grad_norm": 0.5584273338317871,
      "learning_rate": 3.603337532415672e-05,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8359954595565796,
      "num_tokens": 457332324.0,
      "step": 13650
    },
    {
      "epoch": 0.8151265520534862,
      "grad_norm": 0.474702924489975,
      "learning_rate": 3.602309954712011e-05,
      "loss": 0.5502,
      "mean_token_accuracy": 0.8242574334144592,
      "num_tokens": 457500004.0,
      "step": 13655
    },
    {
      "epoch": 0.815425023877746,
      "grad_norm": 0.49339690804481506,
      "learning_rate": 3.601282169412528e-05,
      "loss": 0.5724,
      "mean_token_accuracy": 0.8200524806976318,
      "num_tokens": 457667684.0,
      "step": 13660
    },
    {
      "epoch": 0.8157234957020058,
      "grad_norm": 0.48369187116622925,
      "learning_rate": 3.6002541767675594e-05,
      "loss": 0.5581,
      "mean_token_accuracy": 0.8212871313095093,
      "num_tokens": 457835364.0,
      "step": 13665
    },
    {
      "epoch": 0.8160219675262655,
      "grad_norm": 0.5393469333648682,
      "learning_rate": 3.599225977027492e-05,
      "loss": 0.5437,
      "mean_token_accuracy": 0.8268638968467712,
      "num_tokens": 458003044.0,
      "step": 13670
    },
    {
      "epoch": 0.8163204393505253,
      "grad_norm": 0.5142853856086731,
      "learning_rate": 3.598197570442764e-05,
      "loss": 0.4921,
      "mean_token_accuracy": 0.8428009152412415,
      "num_tokens": 458170724.0,
      "step": 13675
    },
    {
      "epoch": 0.8166189111747851,
      "grad_norm": 0.5485891699790955,
      "learning_rate": 3.597168957263861e-05,
      "loss": 0.5405,
      "mean_token_accuracy": 0.825438380241394,
      "num_tokens": 458338404.0,
      "step": 13680
    },
    {
      "epoch": 0.8169173829990449,
      "grad_norm": 0.5144602656364441,
      "learning_rate": 3.5961401377413216e-05,
      "loss": 0.5245,
      "mean_token_accuracy": 0.8322855830192566,
      "num_tokens": 458506084.0,
      "step": 13685
    },
    {
      "epoch": 0.8172158548233047,
      "grad_norm": 0.5123664736747742,
      "learning_rate": 3.595111112125734e-05,
      "loss": 0.5694,
      "mean_token_accuracy": 0.8184301495552063,
      "num_tokens": 458673764.0,
      "step": 13690
    },
    {
      "epoch": 0.8175143266475645,
      "grad_norm": 0.5427488088607788,
      "learning_rate": 3.594081880667736e-05,
      "loss": 0.5424,
      "mean_token_accuracy": 0.8269533514976501,
      "num_tokens": 458841444.0,
      "step": 13695
    },
    {
      "epoch": 0.8178127984718243,
      "grad_norm": 0.5048689246177673,
      "learning_rate": 3.5930524436180155e-05,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8360670447349549,
      "num_tokens": 459009124.0,
      "step": 13700
    },
    {
      "epoch": 0.8181112702960841,
      "grad_norm": 0.4972262978553772,
      "learning_rate": 3.592022801227312e-05,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8376953363418579,
      "num_tokens": 459176804.0,
      "step": 13705
    },
    {
      "epoch": 0.8184097421203438,
      "grad_norm": 0.5604311227798462,
      "learning_rate": 3.5909929537464136e-05,
      "loss": 0.5472,
      "mean_token_accuracy": 0.8253489255905151,
      "num_tokens": 459344484.0,
      "step": 13710
    },
    {
      "epoch": 0.8187082139446036,
      "grad_norm": 0.500683069229126,
      "learning_rate": 3.589962901426157e-05,
      "loss": 0.5233,
      "mean_token_accuracy": 0.8328223824501038,
      "num_tokens": 459512164.0,
      "step": 13715
    },
    {
      "epoch": 0.8190066857688634,
      "grad_norm": 0.48557719588279724,
      "learning_rate": 3.5889326445174333e-05,
      "loss": 0.5349,
      "mean_token_accuracy": 0.8281701087951661,
      "num_tokens": 459679844.0,
      "step": 13720
    },
    {
      "epoch": 0.8193051575931232,
      "grad_norm": 0.5008373260498047,
      "learning_rate": 3.587902183271178e-05,
      "loss": 0.5681,
      "mean_token_accuracy": 0.81687833070755,
      "num_tokens": 459839940.0,
      "step": 13725
    },
    {
      "epoch": 0.819603629417383,
      "grad_norm": 0.50841224193573,
      "learning_rate": 3.58687151793838e-05,
      "loss": 0.5291,
      "mean_token_accuracy": 0.8292795062065125,
      "num_tokens": 460007620.0,
      "step": 13730
    },
    {
      "epoch": 0.8199021012416428,
      "grad_norm": 0.5785766839981079,
      "learning_rate": 3.585840648770078e-05,
      "loss": 0.5528,
      "mean_token_accuracy": 0.8259513258934021,
      "num_tokens": 460175300.0,
      "step": 13735
    },
    {
      "epoch": 0.8202005730659025,
      "grad_norm": 0.5972203612327576,
      "learning_rate": 3.584809576017358e-05,
      "loss": 0.5865,
      "mean_token_accuracy": 0.8151497125625611,
      "num_tokens": 460342980.0,
      "step": 13740
    },
    {
      "epoch": 0.8204990448901623,
      "grad_norm": 0.5197674036026001,
      "learning_rate": 3.5837782999313556e-05,
      "loss": 0.5408,
      "mean_token_accuracy": 0.8268281102180481,
      "num_tokens": 460510660.0,
      "step": 13745
    },
    {
      "epoch": 0.8207975167144221,
      "grad_norm": 0.48860955238342285,
      "learning_rate": 3.5827468207632605e-05,
      "loss": 0.519,
      "mean_token_accuracy": 0.8344685673713684,
      "num_tokens": 460678340.0,
      "step": 13750
    },
    {
      "epoch": 0.8210959885386819,
      "grad_norm": 0.5184617638587952,
      "learning_rate": 3.581715138764304e-05,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8293570280075073,
      "num_tokens": 460846020.0,
      "step": 13755
    },
    {
      "epoch": 0.8213944603629417,
      "grad_norm": 0.4803366959095001,
      "learning_rate": 3.580683254185776e-05,
      "loss": 0.5861,
      "mean_token_accuracy": 0.8131277561187744,
      "num_tokens": 461013700.0,
      "step": 13760
    },
    {
      "epoch": 0.8216929321872015,
      "grad_norm": 0.5201559066772461,
      "learning_rate": 3.5796511672790076e-05,
      "loss": 0.5589,
      "mean_token_accuracy": 0.8221519708633422,
      "num_tokens": 461181380.0,
      "step": 13765
    },
    {
      "epoch": 0.8219914040114613,
      "grad_norm": 0.5416216850280762,
      "learning_rate": 3.578618878295384e-05,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8290528535842896,
      "num_tokens": 461349060.0,
      "step": 13770
    },
    {
      "epoch": 0.8222898758357211,
      "grad_norm": 0.5434958338737488,
      "learning_rate": 3.577586387486338e-05,
      "loss": 0.5194,
      "mean_token_accuracy": 0.8329416632652282,
      "num_tokens": 461516740.0,
      "step": 13775
    },
    {
      "epoch": 0.8225883476599809,
      "grad_norm": 0.536403477191925,
      "learning_rate": 3.576553695103352e-05,
      "loss": 0.5626,
      "mean_token_accuracy": 0.8210246920585632,
      "num_tokens": 461684420.0,
      "step": 13780
    },
    {
      "epoch": 0.8228868194842407,
      "grad_norm": 0.5481341481208801,
      "learning_rate": 3.575520801397957e-05,
      "loss": 0.5795,
      "mean_token_accuracy": 0.8169628977775574,
      "num_tokens": 461852100.0,
      "step": 13785
    },
    {
      "epoch": 0.8231852913085005,
      "grad_norm": 0.5420999526977539,
      "learning_rate": 3.5744877066217344e-05,
      "loss": 0.5628,
      "mean_token_accuracy": 0.8219789981842041,
      "num_tokens": 462019780.0,
      "step": 13790
    },
    {
      "epoch": 0.8234837631327603,
      "grad_norm": 0.48876407742500305,
      "learning_rate": 3.573454411026311e-05,
      "loss": 0.5514,
      "mean_token_accuracy": 0.8228021025657654,
      "num_tokens": 462187460.0,
      "step": 13795
    },
    {
      "epoch": 0.8237822349570201,
      "grad_norm": 0.5534393191337585,
      "learning_rate": 3.572420914863368e-05,
      "loss": 0.579,
      "mean_token_accuracy": 0.8169867634773255,
      "num_tokens": 462355140.0,
      "step": 13800
    },
    {
      "epoch": 0.8240807067812799,
      "grad_norm": 0.5204468369483948,
      "learning_rate": 3.57138721838463e-05,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8350769400596618,
      "num_tokens": 462522820.0,
      "step": 13805
    },
    {
      "epoch": 0.8243791786055397,
      "grad_norm": 0.5033581852912903,
      "learning_rate": 3.570353321841874e-05,
      "loss": 0.5415,
      "mean_token_accuracy": 0.8270845770835876,
      "num_tokens": 462690500.0,
      "step": 13810
    },
    {
      "epoch": 0.8246776504297995,
      "grad_norm": 0.508622944355011,
      "learning_rate": 3.569319225486924e-05,
      "loss": 0.5371,
      "mean_token_accuracy": 0.8308660387992859,
      "num_tokens": 462858180.0,
      "step": 13815
    },
    {
      "epoch": 0.8249761222540593,
      "grad_norm": 0.5545150637626648,
      "learning_rate": 3.568284929571655e-05,
      "loss": 0.5524,
      "mean_token_accuracy": 0.8241142988204956,
      "num_tokens": 463025860.0,
      "step": 13820
    },
    {
      "epoch": 0.825274594078319,
      "grad_norm": 0.46474355459213257,
      "learning_rate": 3.567250434347987e-05,
      "loss": 0.5105,
      "mean_token_accuracy": 0.8378623366355896,
      "num_tokens": 463193540.0,
      "step": 13825
    },
    {
      "epoch": 0.8255730659025788,
      "grad_norm": 0.5277979373931885,
      "learning_rate": 3.566215740067891e-05,
      "loss": 0.5655,
      "mean_token_accuracy": 0.8230478525161743,
      "num_tokens": 463359917.0,
      "step": 13830
    },
    {
      "epoch": 0.8258715377268386,
      "grad_norm": 0.5198206901550293,
      "learning_rate": 3.5651808469833867e-05,
      "loss": 0.5281,
      "mean_token_accuracy": 0.8314923048019409,
      "num_tokens": 463527597.0,
      "step": 13835
    },
    {
      "epoch": 0.8261700095510984,
      "grad_norm": 0.5170151591300964,
      "learning_rate": 3.564145755346539e-05,
      "loss": 0.5066,
      "mean_token_accuracy": 0.8356316328048706,
      "num_tokens": 463695277.0,
      "step": 13840
    },
    {
      "epoch": 0.8264684813753582,
      "grad_norm": 0.6290221810340881,
      "learning_rate": 3.563110465409466e-05,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8329178094863892,
      "num_tokens": 463862957.0,
      "step": 13845
    },
    {
      "epoch": 0.8267669531996179,
      "grad_norm": 0.5610340237617493,
      "learning_rate": 3.56207497742433e-05,
      "loss": 0.5262,
      "mean_token_accuracy": 0.8316175699234009,
      "num_tokens": 464030637.0,
      "step": 13850
    },
    {
      "epoch": 0.8270654250238777,
      "grad_norm": 0.5238777995109558,
      "learning_rate": 3.561039291643345e-05,
      "loss": 0.5316,
      "mean_token_accuracy": 0.8308063864707946,
      "num_tokens": 464198317.0,
      "step": 13855
    },
    {
      "epoch": 0.8273638968481375,
      "grad_norm": 0.6883141994476318,
      "learning_rate": 3.560003408318771e-05,
      "loss": 0.5343,
      "mean_token_accuracy": 0.829386854171753,
      "num_tokens": 464365997.0,
      "step": 13860
    },
    {
      "epoch": 0.8276623686723973,
      "grad_norm": 0.5681681632995605,
      "learning_rate": 3.5589673277029154e-05,
      "loss": 0.5254,
      "mean_token_accuracy": 0.831754744052887,
      "num_tokens": 464533677.0,
      "step": 13865
    },
    {
      "epoch": 0.8279608404966571,
      "grad_norm": 0.5041458606719971,
      "learning_rate": 3.557931050048135e-05,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8284206032752991,
      "num_tokens": 464701357.0,
      "step": 13870
    },
    {
      "epoch": 0.8282593123209169,
      "grad_norm": 0.5190247893333435,
      "learning_rate": 3.556894575606835e-05,
      "loss": 0.559,
      "mean_token_accuracy": 0.8227007031440735,
      "num_tokens": 464869037.0,
      "step": 13875
    },
    {
      "epoch": 0.8285577841451767,
      "grad_norm": 0.5099976062774658,
      "learning_rate": 3.555857904631467e-05,
      "loss": 0.5243,
      "mean_token_accuracy": 0.8317964911460877,
      "num_tokens": 465036717.0,
      "step": 13880
    },
    {
      "epoch": 0.8288562559694365,
      "grad_norm": 0.6251339912414551,
      "learning_rate": 3.554821037374533e-05,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8280925631523133,
      "num_tokens": 465204397.0,
      "step": 13885
    },
    {
      "epoch": 0.8291547277936963,
      "grad_norm": 0.5084047317504883,
      "learning_rate": 3.5537839740885784e-05,
      "loss": 0.5222,
      "mean_token_accuracy": 0.8342240214347839,
      "num_tokens": 465372077.0,
      "step": 13890
    },
    {
      "epoch": 0.829453199617956,
      "grad_norm": 0.4804095923900604,
      "learning_rate": 3.552746715026202e-05,
      "loss": 0.5087,
      "mean_token_accuracy": 0.8358463644981384,
      "num_tokens": 465539757.0,
      "step": 13895
    },
    {
      "epoch": 0.8297516714422158,
      "grad_norm": 0.47673165798187256,
      "learning_rate": 3.5517092604400465e-05,
      "loss": 0.5533,
      "mean_token_accuracy": 0.8231122374534607,
      "num_tokens": 465707437.0,
      "step": 13900
    },
    {
      "epoch": 0.8300501432664756,
      "grad_norm": 0.5101574659347534,
      "learning_rate": 3.550671610582802e-05,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8393236398696899,
      "num_tokens": 465875117.0,
      "step": 13905
    },
    {
      "epoch": 0.8303486150907354,
      "grad_norm": 0.5689011216163635,
      "learning_rate": 3.549633765707208e-05,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8284027099609375,
      "num_tokens": 466042797.0,
      "step": 13910
    },
    {
      "epoch": 0.8306470869149952,
      "grad_norm": 0.47667232155799866,
      "learning_rate": 3.548595726066051e-05,
      "loss": 0.4812,
      "mean_token_accuracy": 0.8440057277679444,
      "num_tokens": 466210477.0,
      "step": 13915
    },
    {
      "epoch": 0.830945558739255,
      "grad_norm": 0.6022024154663086,
      "learning_rate": 3.547557491912164e-05,
      "loss": 0.5397,
      "mean_token_accuracy": 0.827657151222229,
      "num_tokens": 466378157.0,
      "step": 13920
    },
    {
      "epoch": 0.8312440305635148,
      "grad_norm": 0.5344946980476379,
      "learning_rate": 3.546519063498429e-05,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8267982840538025,
      "num_tokens": 466545837.0,
      "step": 13925
    },
    {
      "epoch": 0.8315425023877746,
      "grad_norm": 0.5445512533187866,
      "learning_rate": 3.5454804410777744e-05,
      "loss": 0.5299,
      "mean_token_accuracy": 0.8288262009620666,
      "num_tokens": 466713517.0,
      "step": 13930
    },
    {
      "epoch": 0.8318409742120344,
      "grad_norm": 0.5051825046539307,
      "learning_rate": 3.544441624903176e-05,
      "loss": 0.5101,
      "mean_token_accuracy": 0.8361326456069946,
      "num_tokens": 466881197.0,
      "step": 13935
    },
    {
      "epoch": 0.8321394460362942,
      "grad_norm": 0.49107885360717773,
      "learning_rate": 3.543402615227655e-05,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8305678129196167,
      "num_tokens": 467048877.0,
      "step": 13940
    },
    {
      "epoch": 0.832437917860554,
      "grad_norm": 0.526508092880249,
      "learning_rate": 3.542363412304283e-05,
      "loss": 0.5097,
      "mean_token_accuracy": 0.8367768168449402,
      "num_tokens": 467216557.0,
      "step": 13945
    },
    {
      "epoch": 0.8327363896848138,
      "grad_norm": 0.535841703414917,
      "learning_rate": 3.541324016386176e-05,
      "loss": 0.534,
      "mean_token_accuracy": 0.8307049989700317,
      "num_tokens": 467384237.0,
      "step": 13950
    },
    {
      "epoch": 0.8330348615090736,
      "grad_norm": 0.48743993043899536,
      "learning_rate": 3.540284427726499e-05,
      "loss": 0.5097,
      "mean_token_accuracy": 0.8360491633415222,
      "num_tokens": 467551917.0,
      "step": 13955
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.4653318524360657,
      "learning_rate": 3.5392446465784627e-05,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8283669233322144,
      "num_tokens": 467719597.0,
      "step": 13960
    },
    {
      "epoch": 0.8336318051575932,
      "grad_norm": 0.5721853375434875,
      "learning_rate": 3.5382046731953256e-05,
      "loss": 0.5323,
      "mean_token_accuracy": 0.8286234140396118,
      "num_tokens": 467887277.0,
      "step": 13965
    },
    {
      "epoch": 0.833930276981853,
      "grad_norm": 0.514631986618042,
      "learning_rate": 3.537164507830391e-05,
      "loss": 0.527,
      "mean_token_accuracy": 0.8319933176040649,
      "num_tokens": 468054957.0,
      "step": 13970
    },
    {
      "epoch": 0.8342287488061128,
      "grad_norm": 0.5097449421882629,
      "learning_rate": 3.536124150737011e-05,
      "loss": 0.5118,
      "mean_token_accuracy": 0.8368066430091858,
      "num_tokens": 468222637.0,
      "step": 13975
    },
    {
      "epoch": 0.8345272206303725,
      "grad_norm": 0.6050868034362793,
      "learning_rate": 3.535083602168584e-05,
      "loss": 0.5203,
      "mean_token_accuracy": 0.834599781036377,
      "num_tokens": 468390317.0,
      "step": 13980
    },
    {
      "epoch": 0.8348256924546322,
      "grad_norm": 0.49385109543800354,
      "learning_rate": 3.534042862378556e-05,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8334009289741516,
      "num_tokens": 468557997.0,
      "step": 13985
    },
    {
      "epoch": 0.835124164278892,
      "grad_norm": 0.4916013777256012,
      "learning_rate": 3.533001931620415e-05,
      "loss": 0.512,
      "mean_token_accuracy": 0.8355600476264954,
      "num_tokens": 468725677.0,
      "step": 13990
    },
    {
      "epoch": 0.8354226361031518,
      "grad_norm": 0.5108382105827332,
      "learning_rate": 3.531960810147702e-05,
      "loss": 0.5244,
      "mean_token_accuracy": 0.8322855710983277,
      "num_tokens": 468893357.0,
      "step": 13995
    },
    {
      "epoch": 0.8357211079274116,
      "grad_norm": 0.5445250272750854,
      "learning_rate": 3.530919498214e-05,
      "loss": 0.5851,
      "mean_token_accuracy": 0.8149469137191773,
      "num_tokens": 469061037.0,
      "step": 14000
    },
    {
      "epoch": 0.8360195797516714,
      "grad_norm": 0.522246241569519,
      "learning_rate": 3.529877996072941e-05,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8337230205535888,
      "num_tokens": 469228717.0,
      "step": 14005
    },
    {
      "epoch": 0.8363180515759312,
      "grad_norm": 0.45550069212913513,
      "learning_rate": 3.5288363039782005e-05,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8324943304061889,
      "num_tokens": 469396397.0,
      "step": 14010
    },
    {
      "epoch": 0.836616523400191,
      "grad_norm": 0.49933379888534546,
      "learning_rate": 3.5277944221835025e-05,
      "loss": 0.5459,
      "mean_token_accuracy": 0.8270607233047486,
      "num_tokens": 469564077.0,
      "step": 14015
    },
    {
      "epoch": 0.8369149952244508,
      "grad_norm": 0.553541362285614,
      "learning_rate": 3.5267523509426164e-05,
      "loss": 0.5202,
      "mean_token_accuracy": 0.8323273301124573,
      "num_tokens": 469731757.0,
      "step": 14020
    },
    {
      "epoch": 0.8372134670487106,
      "grad_norm": 0.43102484941482544,
      "learning_rate": 3.525710090509359e-05,
      "loss": 0.5183,
      "mean_token_accuracy": 0.8342777013778686,
      "num_tokens": 469899437.0,
      "step": 14025
    },
    {
      "epoch": 0.8375119388729704,
      "grad_norm": 0.49941667914390564,
      "learning_rate": 3.52466764113759e-05,
      "loss": 0.5318,
      "mean_token_accuracy": 0.8313193440437316,
      "num_tokens": 470067117.0,
      "step": 14030
    },
    {
      "epoch": 0.8378104106972302,
      "grad_norm": 0.5098081231117249,
      "learning_rate": 3.523625003081218e-05,
      "loss": 0.5078,
      "mean_token_accuracy": 0.8364726305007935,
      "num_tokens": 470234797.0,
      "step": 14035
    },
    {
      "epoch": 0.83810888252149,
      "grad_norm": 0.5632203221321106,
      "learning_rate": 3.5225821765941985e-05,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8249612212181091,
      "num_tokens": 470402477.0,
      "step": 14040
    },
    {
      "epoch": 0.8384073543457498,
      "grad_norm": 0.5137168765068054,
      "learning_rate": 3.5215391619305284e-05,
      "loss": 0.5195,
      "mean_token_accuracy": 0.8332339286804199,
      "num_tokens": 470570157.0,
      "step": 14045
    },
    {
      "epoch": 0.8387058261700095,
      "grad_norm": 0.49517109990119934,
      "learning_rate": 3.520495959344255e-05,
      "loss": 0.5564,
      "mean_token_accuracy": 0.8223786234855652,
      "num_tokens": 470737837.0,
      "step": 14050
    },
    {
      "epoch": 0.8390042979942693,
      "grad_norm": 0.4742959141731262,
      "learning_rate": 3.519452569089468e-05,
      "loss": 0.523,
      "mean_token_accuracy": 0.8328789830207824,
      "num_tokens": 470901313.0,
      "step": 14055
    },
    {
      "epoch": 0.8393027698185291,
      "grad_norm": 0.5067028999328613,
      "learning_rate": 3.5184089914203064e-05,
      "loss": 0.5334,
      "mean_token_accuracy": 0.8302695870399475,
      "num_tokens": 471068993.0,
      "step": 14060
    },
    {
      "epoch": 0.8396012416427889,
      "grad_norm": 0.54022616147995,
      "learning_rate": 3.5173652265909514e-05,
      "loss": 0.5589,
      "mean_token_accuracy": 0.8213945031166077,
      "num_tokens": 471236673.0,
      "step": 14065
    },
    {
      "epoch": 0.8398997134670487,
      "grad_norm": 0.5471249222755432,
      "learning_rate": 3.5163212748556316e-05,
      "loss": 0.5365,
      "mean_token_accuracy": 0.8292138695716857,
      "num_tokens": 471404353.0,
      "step": 14070
    },
    {
      "epoch": 0.8401981852913085,
      "grad_norm": 0.4616066813468933,
      "learning_rate": 3.515277136468621e-05,
      "loss": 0.5391,
      "mean_token_accuracy": 0.8259334325790405,
      "num_tokens": 471572033.0,
      "step": 14075
    },
    {
      "epoch": 0.8404966571155683,
      "grad_norm": 0.5127003788948059,
      "learning_rate": 3.514232811684238e-05,
      "loss": 0.5756,
      "mean_token_accuracy": 0.8191250801086426,
      "num_tokens": 471737001.0,
      "step": 14080
    },
    {
      "epoch": 0.8407951289398281,
      "grad_norm": 0.8175386786460876,
      "learning_rate": 3.513188300756847e-05,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8297101616859436,
      "num_tokens": 471903226.0,
      "step": 14085
    },
    {
      "epoch": 0.8410936007640879,
      "grad_norm": 0.49775615334510803,
      "learning_rate": 3.51214360394086e-05,
      "loss": 0.5201,
      "mean_token_accuracy": 0.8327925682067872,
      "num_tokens": 472070906.0,
      "step": 14090
    },
    {
      "epoch": 0.8413920725883477,
      "grad_norm": 0.5120581984519958,
      "learning_rate": 3.511098721490729e-05,
      "loss": 0.551,
      "mean_token_accuracy": 0.8235417008399963,
      "num_tokens": 472238586.0,
      "step": 14095
    },
    {
      "epoch": 0.8416905444126075,
      "grad_norm": 0.4909849762916565,
      "learning_rate": 3.5100536536609574e-05,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8215436100959778,
      "num_tokens": 472406266.0,
      "step": 14100
    },
    {
      "epoch": 0.8419890162368673,
      "grad_norm": 0.5036343336105347,
      "learning_rate": 3.5090084007060895e-05,
      "loss": 0.5587,
      "mean_token_accuracy": 0.8215533375740052,
      "num_tokens": 472565981.0,
      "step": 14105
    },
    {
      "epoch": 0.8422874880611271,
      "grad_norm": 0.5000784993171692,
      "learning_rate": 3.507962962880715e-05,
      "loss": 0.5474,
      "mean_token_accuracy": 0.8252356052398682,
      "num_tokens": 472733661.0,
      "step": 14110
    },
    {
      "epoch": 0.8425859598853869,
      "grad_norm": 0.566957414150238,
      "learning_rate": 3.5069173404394704e-05,
      "loss": 0.549,
      "mean_token_accuracy": 0.8240427017211914,
      "num_tokens": 472901341.0,
      "step": 14115
    },
    {
      "epoch": 0.8428844317096467,
      "grad_norm": 0.5268917083740234,
      "learning_rate": 3.5058715336370374e-05,
      "loss": 0.54,
      "mean_token_accuracy": 0.8272157907485962,
      "num_tokens": 473069021.0,
      "step": 14120
    },
    {
      "epoch": 0.8431829035339063,
      "grad_norm": 0.5712652206420898,
      "learning_rate": 3.5048255427281404e-05,
      "loss": 0.5677,
      "mean_token_accuracy": 0.8207622647285462,
      "num_tokens": 473236701.0,
      "step": 14125
    },
    {
      "epoch": 0.8434813753581661,
      "grad_norm": 0.48082488775253296,
      "learning_rate": 3.50377936796755e-05,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8411368250846862,
      "num_tokens": 473404381.0,
      "step": 14130
    },
    {
      "epoch": 0.8437798471824259,
      "grad_norm": 0.5146220326423645,
      "learning_rate": 3.5027330096100805e-05,
      "loss": 0.5246,
      "mean_token_accuracy": 0.8308839201927185,
      "num_tokens": 473572061.0,
      "step": 14135
    },
    {
      "epoch": 0.8440783190066857,
      "grad_norm": 0.5012195110321045,
      "learning_rate": 3.501686467910593e-05,
      "loss": 0.546,
      "mean_token_accuracy": 0.8261660456657409,
      "num_tokens": 473739741.0,
      "step": 14140
    },
    {
      "epoch": 0.8443767908309455,
      "grad_norm": 0.5551221966743469,
      "learning_rate": 3.5006397431239926e-05,
      "loss": 0.5524,
      "mean_token_accuracy": 0.8249970078468323,
      "num_tokens": 473907421.0,
      "step": 14145
    },
    {
      "epoch": 0.8446752626552053,
      "grad_norm": 0.5235398411750793,
      "learning_rate": 3.499592835505226e-05,
      "loss": 0.5548,
      "mean_token_accuracy": 0.8232971668243408,
      "num_tokens": 474075101.0,
      "step": 14150
    },
    {
      "epoch": 0.8449737344794651,
      "grad_norm": 0.6222826242446899,
      "learning_rate": 3.4985457453092894e-05,
      "loss": 0.5616,
      "mean_token_accuracy": 0.8232017040252686,
      "num_tokens": 474242781.0,
      "step": 14155
    },
    {
      "epoch": 0.8452722063037249,
      "grad_norm": 0.46936047077178955,
      "learning_rate": 3.497498472791221e-05,
      "loss": 0.5227,
      "mean_token_accuracy": 0.8320648908615113,
      "num_tokens": 474410461.0,
      "step": 14160
    },
    {
      "epoch": 0.8455706781279847,
      "grad_norm": 0.5509169101715088,
      "learning_rate": 3.4964510182061005e-05,
      "loss": 0.6008,
      "mean_token_accuracy": 0.8102409601211548,
      "num_tokens": 474578141.0,
      "step": 14165
    },
    {
      "epoch": 0.8458691499522445,
      "grad_norm": 0.46338951587677,
      "learning_rate": 3.495403381809057e-05,
      "loss": 0.4766,
      "mean_token_accuracy": 0.8434390902519227,
      "num_tokens": 474745821.0,
      "step": 14170
    },
    {
      "epoch": 0.8461676217765043,
      "grad_norm": 0.5123574733734131,
      "learning_rate": 3.4943555638552625e-05,
      "loss": 0.5569,
      "mean_token_accuracy": 0.8220147967338562,
      "num_tokens": 474913501.0,
      "step": 14175
    },
    {
      "epoch": 0.8464660936007641,
      "grad_norm": 0.47622036933898926,
      "learning_rate": 3.49330756459993e-05,
      "loss": 0.5351,
      "mean_token_accuracy": 0.8301741600036621,
      "num_tokens": 475081181.0,
      "step": 14180
    },
    {
      "epoch": 0.8467645654250239,
      "grad_norm": 0.5069224238395691,
      "learning_rate": 3.4922593842983206e-05,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8378183722496033,
      "num_tokens": 475243223.0,
      "step": 14185
    },
    {
      "epoch": 0.8470630372492837,
      "grad_norm": 0.5753136277198792,
      "learning_rate": 3.491211023205737e-05,
      "loss": 0.5591,
      "mean_token_accuracy": 0.8227722644805908,
      "num_tokens": 475410903.0,
      "step": 14190
    },
    {
      "epoch": 0.8473615090735435,
      "grad_norm": 0.540684163570404,
      "learning_rate": 3.490162481577529e-05,
      "loss": 0.5948,
      "mean_token_accuracy": 0.8118394374847412,
      "num_tokens": 475578583.0,
      "step": 14195
    },
    {
      "epoch": 0.8476599808978033,
      "grad_norm": 0.497579962015152,
      "learning_rate": 3.4891137596690845e-05,
      "loss": 0.5088,
      "mean_token_accuracy": 0.8377669095993042,
      "num_tokens": 475746263.0,
      "step": 14200
    },
    {
      "epoch": 0.847958452722063,
      "grad_norm": 0.46217402815818787,
      "learning_rate": 3.488064857735843e-05,
      "loss": 0.492,
      "mean_token_accuracy": 0.8410712122917176,
      "num_tokens": 475913943.0,
      "step": 14205
    },
    {
      "epoch": 0.8482569245463228,
      "grad_norm": 0.44693678617477417,
      "learning_rate": 3.487015776033282e-05,
      "loss": 0.5356,
      "mean_token_accuracy": 0.828921627998352,
      "num_tokens": 476081623.0,
      "step": 14210
    },
    {
      "epoch": 0.8485553963705826,
      "grad_norm": 0.5191335082054138,
      "learning_rate": 3.485966514816925e-05,
      "loss": 0.5178,
      "mean_token_accuracy": 0.8327985167503357,
      "num_tokens": 476249303.0,
      "step": 14215
    },
    {
      "epoch": 0.8488538681948424,
      "grad_norm": 0.5962714552879333,
      "learning_rate": 3.4849170743423385e-05,
      "loss": 0.5494,
      "mean_token_accuracy": 0.8267624974250793,
      "num_tokens": 476416983.0,
      "step": 14220
    },
    {
      "epoch": 0.8491523400191022,
      "grad_norm": 0.5007015466690063,
      "learning_rate": 3.4838674548651346e-05,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8269772052764892,
      "num_tokens": 476584663.0,
      "step": 14225
    },
    {
      "epoch": 0.849450811843362,
      "grad_norm": 0.4992782771587372,
      "learning_rate": 3.482817656640965e-05,
      "loss": 0.5596,
      "mean_token_accuracy": 0.8225515842437744,
      "num_tokens": 476752343.0,
      "step": 14230
    },
    {
      "epoch": 0.8497492836676218,
      "grad_norm": 0.5055899620056152,
      "learning_rate": 3.4817676799255307e-05,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8330251693725585,
      "num_tokens": 476920023.0,
      "step": 14235
    },
    {
      "epoch": 0.8500477554918816,
      "grad_norm": 0.46591854095458984,
      "learning_rate": 3.480717524974569e-05,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8340271949768067,
      "num_tokens": 477087703.0,
      "step": 14240
    },
    {
      "epoch": 0.8503462273161414,
      "grad_norm": 0.49624285101890564,
      "learning_rate": 3.479667192043868e-05,
      "loss": 0.5781,
      "mean_token_accuracy": 0.8159549117088318,
      "num_tokens": 477255383.0,
      "step": 14245
    },
    {
      "epoch": 0.8506446991404012,
      "grad_norm": 0.4981527328491211,
      "learning_rate": 3.4786166813892535e-05,
      "loss": 0.5494,
      "mean_token_accuracy": 0.8254622340202331,
      "num_tokens": 477423063.0,
      "step": 14250
    },
    {
      "epoch": 0.850943170964661,
      "grad_norm": 0.4836324155330658,
      "learning_rate": 3.477565993266598e-05,
      "loss": 0.5056,
      "mean_token_accuracy": 0.8368722319602966,
      "num_tokens": 477590743.0,
      "step": 14255
    },
    {
      "epoch": 0.8512416427889207,
      "grad_norm": 0.49234965443611145,
      "learning_rate": 3.476515127931816e-05,
      "loss": 0.5231,
      "mean_token_accuracy": 0.8327865839004517,
      "num_tokens": 477758423.0,
      "step": 14260
    },
    {
      "epoch": 0.8515401146131805,
      "grad_norm": 0.5004165172576904,
      "learning_rate": 3.475464085640864e-05,
      "loss": 0.5104,
      "mean_token_accuracy": 0.8364845514297485,
      "num_tokens": 477926103.0,
      "step": 14265
    },
    {
      "epoch": 0.8518385864374403,
      "grad_norm": 0.512992799282074,
      "learning_rate": 3.474412866649743e-05,
      "loss": 0.5262,
      "mean_token_accuracy": 0.8305201053619384,
      "num_tokens": 478093783.0,
      "step": 14270
    },
    {
      "epoch": 0.8521370582617,
      "grad_norm": 0.5965930819511414,
      "learning_rate": 3.473361471214498e-05,
      "loss": 0.5759,
      "mean_token_accuracy": 0.8187403082847595,
      "num_tokens": 478261463.0,
      "step": 14275
    },
    {
      "epoch": 0.8524355300859598,
      "grad_norm": 0.5190654993057251,
      "learning_rate": 3.4723098995912135e-05,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8395443201065064,
      "num_tokens": 478429143.0,
      "step": 14280
    },
    {
      "epoch": 0.8527340019102196,
      "grad_norm": 0.5768404006958008,
      "learning_rate": 3.471258152036022e-05,
      "loss": 0.5949,
      "mean_token_accuracy": 0.8119467973709107,
      "num_tokens": 478596823.0,
      "step": 14285
    },
    {
      "epoch": 0.8530324737344794,
      "grad_norm": 0.5373517870903015,
      "learning_rate": 3.4702062288050945e-05,
      "loss": 0.4949,
      "mean_token_accuracy": 0.840838611125946,
      "num_tokens": 478764503.0,
      "step": 14290
    },
    {
      "epoch": 0.8533309455587392,
      "grad_norm": 0.5412562489509583,
      "learning_rate": 3.469154130154647e-05,
      "loss": 0.5872,
      "mean_token_accuracy": 0.8135452747344971,
      "num_tokens": 478932183.0,
      "step": 14295
    },
    {
      "epoch": 0.853629417382999,
      "grad_norm": 0.5922993421554565,
      "learning_rate": 3.4681018563409364e-05,
      "loss": 0.5583,
      "mean_token_accuracy": 0.8255338191986084,
      "num_tokens": 479099863.0,
      "step": 14300
    },
    {
      "epoch": 0.8539278892072588,
      "grad_norm": 0.490708589553833,
      "learning_rate": 3.467049407620264e-05,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8310390114784241,
      "num_tokens": 479267543.0,
      "step": 14305
    },
    {
      "epoch": 0.8542263610315186,
      "grad_norm": 0.5206578373908997,
      "learning_rate": 3.4659967842489733e-05,
      "loss": 0.5026,
      "mean_token_accuracy": 0.8386675357818604,
      "num_tokens": 479435223.0,
      "step": 14310
    },
    {
      "epoch": 0.8545248328557784,
      "grad_norm": 0.5300891399383545,
      "learning_rate": 3.46494398648345e-05,
      "loss": 0.5393,
      "mean_token_accuracy": 0.8279852151870728,
      "num_tokens": 479602903.0,
      "step": 14315
    },
    {
      "epoch": 0.8548233046800382,
      "grad_norm": 0.459747850894928,
      "learning_rate": 3.463891014580122e-05,
      "loss": 0.5407,
      "mean_token_accuracy": 0.8270249485969543,
      "num_tokens": 479770583.0,
      "step": 14320
    },
    {
      "epoch": 0.855121776504298,
      "grad_norm": 0.612096905708313,
      "learning_rate": 3.462837868795461e-05,
      "loss": 0.5268,
      "mean_token_accuracy": 0.8332160353660584,
      "num_tokens": 479938263.0,
      "step": 14325
    },
    {
      "epoch": 0.8554202483285578,
      "grad_norm": 0.5251293778419495,
      "learning_rate": 3.461784549385979e-05,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8328701019287109,
      "num_tokens": 480105943.0,
      "step": 14330
    },
    {
      "epoch": 0.8557187201528176,
      "grad_norm": 0.502863347530365,
      "learning_rate": 3.460731056608233e-05,
      "loss": 0.4748,
      "mean_token_accuracy": 0.8477573752403259,
      "num_tokens": 480273623.0,
      "step": 14335
    },
    {
      "epoch": 0.8560171919770774,
      "grad_norm": 0.5029167532920837,
      "learning_rate": 3.459677390718819e-05,
      "loss": 0.5348,
      "mean_token_accuracy": 0.8268519759178161,
      "num_tokens": 480441303.0,
      "step": 14340
    },
    {
      "epoch": 0.8563156638013372,
      "grad_norm": 0.5670776963233948,
      "learning_rate": 3.4586235519743774e-05,
      "loss": 0.5339,
      "mean_token_accuracy": 0.829899787902832,
      "num_tokens": 480608983.0,
      "step": 14345
    },
    {
      "epoch": 0.856614135625597,
      "grad_norm": 0.4996817111968994,
      "learning_rate": 3.45756954063159e-05,
      "loss": 0.5557,
      "mean_token_accuracy": 0.8222533702850342,
      "num_tokens": 480776663.0,
      "step": 14350
    },
    {
      "epoch": 0.8569126074498568,
      "grad_norm": 0.5385510325431824,
      "learning_rate": 3.456515356947181e-05,
      "loss": 0.5091,
      "mean_token_accuracy": 0.8370869398117066,
      "num_tokens": 480944343.0,
      "step": 14355
    },
    {
      "epoch": 0.8572110792741165,
      "grad_norm": 0.5175701975822449,
      "learning_rate": 3.455461001177915e-05,
      "loss": 0.5766,
      "mean_token_accuracy": 0.8181796431541443,
      "num_tokens": 481112023.0,
      "step": 14360
    },
    {
      "epoch": 0.8575095510983763,
      "grad_norm": 0.5024404525756836,
      "learning_rate": 3.454406473580602e-05,
      "loss": 0.5292,
      "mean_token_accuracy": 0.8305618405342102,
      "num_tokens": 481279703.0,
      "step": 14365
    },
    {
      "epoch": 0.8578080229226361,
      "grad_norm": 0.5207492709159851,
      "learning_rate": 3.4533517744120894e-05,
      "loss": 0.5142,
      "mean_token_accuracy": 0.8351246476173401,
      "num_tokens": 481447383.0,
      "step": 14370
    },
    {
      "epoch": 0.8581064947468959,
      "grad_norm": 0.5085864663124084,
      "learning_rate": 3.452296903929271e-05,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8236311674118042,
      "num_tokens": 481615063.0,
      "step": 14375
    },
    {
      "epoch": 0.8584049665711557,
      "grad_norm": 0.4787059426307678,
      "learning_rate": 3.451241862389078e-05,
      "loss": 0.4988,
      "mean_token_accuracy": 0.8387331366539001,
      "num_tokens": 481782743.0,
      "step": 14380
    },
    {
      "epoch": 0.8587034383954155,
      "grad_norm": 0.5232252478599548,
      "learning_rate": 3.4501866500484856e-05,
      "loss": 0.5561,
      "mean_token_accuracy": 0.8238160490989686,
      "num_tokens": 481950423.0,
      "step": 14385
    },
    {
      "epoch": 0.8590019102196753,
      "grad_norm": 0.4464224874973297,
      "learning_rate": 3.449131267164512e-05,
      "loss": 0.4947,
      "mean_token_accuracy": 0.8401168942451477,
      "num_tokens": 482118103.0,
      "step": 14390
    },
    {
      "epoch": 0.8593003820439351,
      "grad_norm": 0.4599359631538391,
      "learning_rate": 3.448075713994213e-05,
      "loss": 0.5387,
      "mean_token_accuracy": 0.829524040222168,
      "num_tokens": 482285783.0,
      "step": 14395
    },
    {
      "epoch": 0.8595988538681948,
      "grad_norm": 0.4855235815048218,
      "learning_rate": 3.4470199907946886e-05,
      "loss": 0.5435,
      "mean_token_accuracy": 0.8270070433616639,
      "num_tokens": 482453463.0,
      "step": 14400
    },
    {
      "epoch": 0.8598973256924546,
      "grad_norm": 0.5263638496398926,
      "learning_rate": 3.44596409782308e-05,
      "loss": 0.5116,
      "mean_token_accuracy": 0.835351312160492,
      "num_tokens": 482621143.0,
      "step": 14405
    },
    {
      "epoch": 0.8601957975167144,
      "grad_norm": 0.5329366326332092,
      "learning_rate": 3.44490803533657e-05,
      "loss": 0.5174,
      "mean_token_accuracy": 0.8341345667839051,
      "num_tokens": 482788823.0,
      "step": 14410
    },
    {
      "epoch": 0.8604942693409742,
      "grad_norm": 0.5646621584892273,
      "learning_rate": 3.44385180359238e-05,
      "loss": 0.5704,
      "mean_token_accuracy": 0.8182392835617065,
      "num_tokens": 482956503.0,
      "step": 14415
    },
    {
      "epoch": 0.860792741165234,
      "grad_norm": 0.5521789193153381,
      "learning_rate": 3.442795402847778e-05,
      "loss": 0.5225,
      "mean_token_accuracy": 0.8330311298370361,
      "num_tokens": 483124183.0,
      "step": 14420
    },
    {
      "epoch": 0.8610912129894938,
      "grad_norm": 0.5149720907211304,
      "learning_rate": 3.441738833360067e-05,
      "loss": 0.5933,
      "mean_token_accuracy": 0.8144399523735046,
      "num_tokens": 483291863.0,
      "step": 14425
    },
    {
      "epoch": 0.8613896848137536,
      "grad_norm": 0.5148646831512451,
      "learning_rate": 3.440682095386595e-05,
      "loss": 0.5041,
      "mean_token_accuracy": 0.8392878413200379,
      "num_tokens": 483459543.0,
      "step": 14430
    },
    {
      "epoch": 0.8616881566380133,
      "grad_norm": 0.6794476509094238,
      "learning_rate": 3.43962518918475e-05,
      "loss": 0.5928,
      "mean_token_accuracy": 0.8129965424537658,
      "num_tokens": 483627223.0,
      "step": 14435
    },
    {
      "epoch": 0.8619866284622731,
      "grad_norm": 0.4807991683483124,
      "learning_rate": 3.438568115011962e-05,
      "loss": 0.5368,
      "mean_token_accuracy": 0.8297089219093323,
      "num_tokens": 483794903.0,
      "step": 14440
    },
    {
      "epoch": 0.8622851002865329,
      "grad_norm": 0.5064502954483032,
      "learning_rate": 3.437510873125699e-05,
      "loss": 0.5387,
      "mean_token_accuracy": 0.8275080561637879,
      "num_tokens": 483962583.0,
      "step": 14445
    },
    {
      "epoch": 0.8625835721107927,
      "grad_norm": 0.5123314261436462,
      "learning_rate": 3.436453463783474e-05,
      "loss": 0.5533,
      "mean_token_accuracy": 0.8242335796356202,
      "num_tokens": 484130263.0,
      "step": 14450
    },
    {
      "epoch": 0.8628820439350525,
      "grad_norm": 0.49425628781318665,
      "learning_rate": 3.435395887242837e-05,
      "loss": 0.5424,
      "mean_token_accuracy": 0.827317190170288,
      "num_tokens": 484297943.0,
      "step": 14455
    },
    {
      "epoch": 0.8631805157593123,
      "grad_norm": 0.5996102690696716,
      "learning_rate": 3.4343381437613814e-05,
      "loss": 0.515,
      "mean_token_accuracy": 0.8351186990737915,
      "num_tokens": 484465623.0,
      "step": 14460
    },
    {
      "epoch": 0.8634789875835721,
      "grad_norm": 0.501671314239502,
      "learning_rate": 3.433280233596738e-05,
      "loss": 0.5475,
      "mean_token_accuracy": 0.8252475261688232,
      "num_tokens": 484633303.0,
      "step": 14465
    },
    {
      "epoch": 0.8637774594078319,
      "grad_norm": 0.5192108154296875,
      "learning_rate": 3.4322221570065824e-05,
      "loss": 0.5415,
      "mean_token_accuracy": 0.8267028570175171,
      "num_tokens": 484800983.0,
      "step": 14470
    },
    {
      "epoch": 0.8640759312320917,
      "grad_norm": 0.4694269895553589,
      "learning_rate": 3.4311639142486295e-05,
      "loss": 0.4752,
      "mean_token_accuracy": 0.8457533240318298,
      "num_tokens": 484968663.0,
      "step": 14475
    },
    {
      "epoch": 0.8643744030563515,
      "grad_norm": 0.49048593640327454,
      "learning_rate": 3.430105505580631e-05,
      "loss": 0.5337,
      "mean_token_accuracy": 0.8305499315261841,
      "num_tokens": 485136343.0,
      "step": 14480
    },
    {
      "epoch": 0.8646728748806113,
      "grad_norm": 0.4649653434753418,
      "learning_rate": 3.4290469312603826e-05,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8389499068260193,
      "num_tokens": 485302737.0,
      "step": 14485
    },
    {
      "epoch": 0.8649713467048711,
      "grad_norm": 0.5668534636497498,
      "learning_rate": 3.4279881915457216e-05,
      "loss": 0.533,
      "mean_token_accuracy": 0.8292675733566284,
      "num_tokens": 485470417.0,
      "step": 14490
    },
    {
      "epoch": 0.8652698185291309,
      "grad_norm": 0.55607670545578,
      "learning_rate": 3.426929286694522e-05,
      "loss": 0.5253,
      "mean_token_accuracy": 0.8315280914306641,
      "num_tokens": 485638097.0,
      "step": 14495
    },
    {
      "epoch": 0.8655682903533907,
      "grad_norm": 0.5578140616416931,
      "learning_rate": 3.4258702169646994e-05,
      "loss": 0.4983,
      "mean_token_accuracy": 0.8403435707092285,
      "num_tokens": 485805777.0,
      "step": 14500
    },
    {
      "epoch": 0.8658667621776505,
      "grad_norm": 0.5597586631774902,
      "learning_rate": 3.424810982614212e-05,
      "loss": 0.5686,
      "mean_token_accuracy": 0.821501863002777,
      "num_tokens": 485973457.0,
      "step": 14505
    },
    {
      "epoch": 0.8661652340019103,
      "grad_norm": 0.5656439661979675,
      "learning_rate": 3.423751583901052e-05,
      "loss": 0.5552,
      "mean_token_accuracy": 0.8234760880470275,
      "num_tokens": 486141137.0,
      "step": 14510
    },
    {
      "epoch": 0.86646370582617,
      "grad_norm": 0.5375494360923767,
      "learning_rate": 3.422692021083259e-05,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8346415400505066,
      "num_tokens": 486308817.0,
      "step": 14515
    },
    {
      "epoch": 0.8667621776504298,
      "grad_norm": 0.44071945548057556,
      "learning_rate": 3.4216322944189076e-05,
      "loss": 0.5243,
      "mean_token_accuracy": 0.8339437007904053,
      "num_tokens": 486476497.0,
      "step": 14520
    },
    {
      "epoch": 0.8670606494746896,
      "grad_norm": 0.4848516285419464,
      "learning_rate": 3.4205724041661135e-05,
      "loss": 0.5301,
      "mean_token_accuracy": 0.8339735269546509,
      "num_tokens": 486644177.0,
      "step": 14525
    },
    {
      "epoch": 0.8673591212989494,
      "grad_norm": 0.4755540192127228,
      "learning_rate": 3.4195123505830326e-05,
      "loss": 0.5056,
      "mean_token_accuracy": 0.8366217374801636,
      "num_tokens": 486811857.0,
      "step": 14530
    },
    {
      "epoch": 0.8676575931232091,
      "grad_norm": 0.5334329009056091,
      "learning_rate": 3.418452133927862e-05,
      "loss": 0.5868,
      "mean_token_accuracy": 0.8151377677917481,
      "num_tokens": 486979537.0,
      "step": 14535
    },
    {
      "epoch": 0.8679560649474689,
      "grad_norm": 0.5428524017333984,
      "learning_rate": 3.417391754458836e-05,
      "loss": 0.5317,
      "mean_token_accuracy": 0.8293510675430298,
      "num_tokens": 487147217.0,
      "step": 14540
    },
    {
      "epoch": 0.8682545367717287,
      "grad_norm": 0.45626723766326904,
      "learning_rate": 3.41633121243423e-05,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8340331554412842,
      "num_tokens": 487314897.0,
      "step": 14545
    },
    {
      "epoch": 0.8685530085959885,
      "grad_norm": 0.6469405293464661,
      "learning_rate": 3.415270508112357e-05,
      "loss": 0.5368,
      "mean_token_accuracy": 0.8293093204498291,
      "num_tokens": 487482577.0,
      "step": 14550
    },
    {
      "epoch": 0.8688514804202483,
      "grad_norm": 0.5415879487991333,
      "learning_rate": 3.414209641751572e-05,
      "loss": 0.5633,
      "mean_token_accuracy": 0.8217941045761108,
      "num_tokens": 487650257.0,
      "step": 14555
    },
    {
      "epoch": 0.8691499522445081,
      "grad_norm": 0.5342953205108643,
      "learning_rate": 3.41314861361027e-05,
      "loss": 0.5571,
      "mean_token_accuracy": 0.8227722764015197,
      "num_tokens": 487817937.0,
      "step": 14560
    },
    {
      "epoch": 0.8694484240687679,
      "grad_norm": 0.47511962056159973,
      "learning_rate": 3.412087423946883e-05,
      "loss": 0.5073,
      "mean_token_accuracy": 0.8363175392150879,
      "num_tokens": 487985617.0,
      "step": 14565
    },
    {
      "epoch": 0.8697468958930277,
      "grad_norm": 0.49829038977622986,
      "learning_rate": 3.411026073019882e-05,
      "loss": 0.4882,
      "mean_token_accuracy": 0.8436180472373962,
      "num_tokens": 488153297.0,
      "step": 14570
    },
    {
      "epoch": 0.8700453677172875,
      "grad_norm": 0.6791826486587524,
      "learning_rate": 3.409964561087781e-05,
      "loss": 0.544,
      "mean_token_accuracy": 0.8278003096580505,
      "num_tokens": 488320977.0,
      "step": 14575
    },
    {
      "epoch": 0.8703438395415473,
      "grad_norm": 0.46776020526885986,
      "learning_rate": 3.408902888409127e-05,
      "loss": 0.5373,
      "mean_token_accuracy": 0.827281391620636,
      "num_tokens": 488488657.0,
      "step": 14580
    },
    {
      "epoch": 0.870642311365807,
      "grad_norm": 0.5241900086402893,
      "learning_rate": 3.4078410552425154e-05,
      "loss": 0.5243,
      "mean_token_accuracy": 0.8316950917243957,
      "num_tokens": 488656337.0,
      "step": 14585
    },
    {
      "epoch": 0.8709407831900668,
      "grad_norm": 0.47714149951934814,
      "learning_rate": 3.406779061846571e-05,
      "loss": 0.5009,
      "mean_token_accuracy": 0.8374746561050415,
      "num_tokens": 488824017.0,
      "step": 14590
    },
    {
      "epoch": 0.8712392550143266,
      "grad_norm": 0.46514490246772766,
      "learning_rate": 3.4057169084799645e-05,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8257604718208313,
      "num_tokens": 488991697.0,
      "step": 14595
    },
    {
      "epoch": 0.8715377268385864,
      "grad_norm": 0.5989508628845215,
      "learning_rate": 3.404654595401401e-05,
      "loss": 0.5436,
      "mean_token_accuracy": 0.826792311668396,
      "num_tokens": 489159377.0,
      "step": 14600
    },
    {
      "epoch": 0.8718361986628462,
      "grad_norm": 0.47459229826927185,
      "learning_rate": 3.4035921228696275e-05,
      "loss": 0.5097,
      "mean_token_accuracy": 0.8347548723220826,
      "num_tokens": 489327057.0,
      "step": 14605
    },
    {
      "epoch": 0.872134670487106,
      "grad_norm": 0.5450000762939453,
      "learning_rate": 3.402529491143428e-05,
      "loss": 0.5252,
      "mean_token_accuracy": 0.8310688376426697,
      "num_tokens": 489494737.0,
      "step": 14610
    },
    {
      "epoch": 0.8724331423113658,
      "grad_norm": 0.4706822335720062,
      "learning_rate": 3.401466700481627e-05,
      "loss": 0.4926,
      "mean_token_accuracy": 0.8403316259384155,
      "num_tokens": 489662417.0,
      "step": 14615
    },
    {
      "epoch": 0.8727316141356256,
      "grad_norm": 0.4576685130596161,
      "learning_rate": 3.4004037511430866e-05,
      "loss": 0.4926,
      "mean_token_accuracy": 0.8423893690109253,
      "num_tokens": 489830097.0,
      "step": 14620
    },
    {
      "epoch": 0.8730300859598854,
      "grad_norm": 0.4945715665817261,
      "learning_rate": 3.3993406433867076e-05,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8403972387313843,
      "num_tokens": 489997777.0,
      "step": 14625
    },
    {
      "epoch": 0.8733285577841452,
      "grad_norm": 0.5267667770385742,
      "learning_rate": 3.398277377471429e-05,
      "loss": 0.5098,
      "mean_token_accuracy": 0.8352022051811219,
      "num_tokens": 490165457.0,
      "step": 14630
    },
    {
      "epoch": 0.873627029608405,
      "grad_norm": 0.4888465404510498,
      "learning_rate": 3.397213953656229e-05,
      "loss": 0.5227,
      "mean_token_accuracy": 0.8301025986671448,
      "num_tokens": 490333137.0,
      "step": 14635
    },
    {
      "epoch": 0.8739255014326648,
      "grad_norm": 0.4893541932106018,
      "learning_rate": 3.396150372200125e-05,
      "loss": 0.5171,
      "mean_token_accuracy": 0.8331981420516967,
      "num_tokens": 490500817.0,
      "step": 14640
    },
    {
      "epoch": 0.8742239732569246,
      "grad_norm": 0.494081974029541,
      "learning_rate": 3.3950866333621715e-05,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8382440567016601,
      "num_tokens": 490668497.0,
      "step": 14645
    },
    {
      "epoch": 0.8745224450811844,
      "grad_norm": 0.465697318315506,
      "learning_rate": 3.394022737401462e-05,
      "loss": 0.5236,
      "mean_token_accuracy": 0.8325539827346802,
      "num_tokens": 490836177.0,
      "step": 14650
    },
    {
      "epoch": 0.8748209169054442,
      "grad_norm": 0.5026145577430725,
      "learning_rate": 3.392958684577126e-05,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8313551306724548,
      "num_tokens": 491003857.0,
      "step": 14655
    },
    {
      "epoch": 0.875119388729704,
      "grad_norm": 0.4903579652309418,
      "learning_rate": 3.3918944751483353e-05,
      "loss": 0.5932,
      "mean_token_accuracy": 0.8120780110359191,
      "num_tokens": 491171537.0,
      "step": 14660
    },
    {
      "epoch": 0.8754178605539638,
      "grad_norm": 0.4996260106563568,
      "learning_rate": 3.390830109374297e-05,
      "loss": 0.5449,
      "mean_token_accuracy": 0.8251043796539307,
      "num_tokens": 491339217.0,
      "step": 14665
    },
    {
      "epoch": 0.8757163323782235,
      "grad_norm": 0.47404783964157104,
      "learning_rate": 3.389765587514258e-05,
      "loss": 0.515,
      "mean_token_accuracy": 0.8338005423545838,
      "num_tokens": 491506897.0,
      "step": 14670
    },
    {
      "epoch": 0.8760148042024832,
      "grad_norm": 0.5179705023765564,
      "learning_rate": 3.3887009098275006e-05,
      "loss": 0.5422,
      "mean_token_accuracy": 0.8266849517822266,
      "num_tokens": 491674577.0,
      "step": 14675
    },
    {
      "epoch": 0.876313276026743,
      "grad_norm": 0.5372985005378723,
      "learning_rate": 3.387636076573348e-05,
      "loss": 0.5455,
      "mean_token_accuracy": 0.8258857250213623,
      "num_tokens": 491842257.0,
      "step": 14680
    },
    {
      "epoch": 0.8766117478510028,
      "grad_norm": 0.5171255469322205,
      "learning_rate": 3.3865710880111594e-05,
      "loss": 0.5533,
      "mean_token_accuracy": 0.8231241703033447,
      "num_tokens": 492009937.0,
      "step": 14685
    },
    {
      "epoch": 0.8769102196752626,
      "grad_norm": 0.4769144654273987,
      "learning_rate": 3.385505944400332e-05,
      "loss": 0.5268,
      "mean_token_accuracy": 0.830776572227478,
      "num_tokens": 492177617.0,
      "step": 14690
    },
    {
      "epoch": 0.8772086914995224,
      "grad_norm": 0.47414425015449524,
      "learning_rate": 3.384440646000303e-05,
      "loss": 0.51,
      "mean_token_accuracy": 0.8342717409133911,
      "num_tokens": 492345297.0,
      "step": 14695
    },
    {
      "epoch": 0.8775071633237822,
      "grad_norm": 0.4481916129589081,
      "learning_rate": 3.383375193070543e-05,
      "loss": 0.5256,
      "mean_token_accuracy": 0.8328343033790588,
      "num_tokens": 492512977.0,
      "step": 14700
    },
    {
      "epoch": 0.877805635148042,
      "grad_norm": 0.49395641684532166,
      "learning_rate": 3.382309585870564e-05,
      "loss": 0.5421,
      "mean_token_accuracy": 0.8277048826217651,
      "num_tokens": 492680657.0,
      "step": 14705
    },
    {
      "epoch": 0.8781041069723018,
      "grad_norm": 0.45159032940864563,
      "learning_rate": 3.3812438246599154e-05,
      "loss": 0.529,
      "mean_token_accuracy": 0.8305558919906616,
      "num_tokens": 492848337.0,
      "step": 14710
    },
    {
      "epoch": 0.8784025787965616,
      "grad_norm": 0.48325857520103455,
      "learning_rate": 3.380177909698181e-05,
      "loss": 0.4791,
      "mean_token_accuracy": 0.8445902466773987,
      "num_tokens": 493016017.0,
      "step": 14715
    },
    {
      "epoch": 0.8787010506208214,
      "grad_norm": 0.44052574038505554,
      "learning_rate": 3.379111841244985e-05,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8406119704246521,
      "num_tokens": 493183697.0,
      "step": 14720
    },
    {
      "epoch": 0.8789995224450812,
      "grad_norm": 0.5341507792472839,
      "learning_rate": 3.378045619559988e-05,
      "loss": 0.5639,
      "mean_token_accuracy": 0.8211737990379333,
      "num_tokens": 493351377.0,
      "step": 14725
    },
    {
      "epoch": 0.879297994269341,
      "grad_norm": 0.5167533159255981,
      "learning_rate": 3.3769792449028884e-05,
      "loss": 0.4826,
      "mean_token_accuracy": 0.8435106754302979,
      "num_tokens": 493519057.0,
      "step": 14730
    },
    {
      "epoch": 0.8795964660936008,
      "grad_norm": 0.5037776827812195,
      "learning_rate": 3.37591271753342e-05,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8321126103401184,
      "num_tokens": 493686737.0,
      "step": 14735
    },
    {
      "epoch": 0.8798949379178606,
      "grad_norm": 0.4913686513900757,
      "learning_rate": 3.3748460377113584e-05,
      "loss": 0.4983,
      "mean_token_accuracy": 0.8401765346527099,
      "num_tokens": 493854417.0,
      "step": 14740
    },
    {
      "epoch": 0.8801934097421203,
      "grad_norm": 0.5046586394309998,
      "learning_rate": 3.3737792056965086e-05,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8373255372047425,
      "num_tokens": 494022097.0,
      "step": 14745
    },
    {
      "epoch": 0.8804918815663801,
      "grad_norm": 0.5026734471321106,
      "learning_rate": 3.372712221748721e-05,
      "loss": 0.5453,
      "mean_token_accuracy": 0.8256411671638488,
      "num_tokens": 494189777.0,
      "step": 14750
    },
    {
      "epoch": 0.8807903533906399,
      "grad_norm": 0.5048609972000122,
      "learning_rate": 3.371645086127877e-05,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8392878532409668,
      "num_tokens": 494357457.0,
      "step": 14755
    },
    {
      "epoch": 0.8810888252148997,
      "grad_norm": 0.5479912757873535,
      "learning_rate": 3.370577799093899e-05,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8284614205360412,
      "num_tokens": 494518018.0,
      "step": 14760
    },
    {
      "epoch": 0.8813872970391595,
      "grad_norm": 0.44558560848236084,
      "learning_rate": 3.3695103609067427e-05,
      "loss": 0.5714,
      "mean_token_accuracy": 0.8166587233543396,
      "num_tokens": 494685698.0,
      "step": 14765
    },
    {
      "epoch": 0.8816857688634193,
      "grad_norm": 0.4608181118965149,
      "learning_rate": 3.368442771826403e-05,
      "loss": 0.459,
      "mean_token_accuracy": 0.8502147197723389,
      "num_tokens": 494853378.0,
      "step": 14770
    },
    {
      "epoch": 0.8819842406876791,
      "grad_norm": 0.5255980491638184,
      "learning_rate": 3.367375032112911e-05,
      "loss": 0.5358,
      "mean_token_accuracy": 0.8271340608596802,
      "num_tokens": 495021009.0,
      "step": 14775
    },
    {
      "epoch": 0.8822827125119389,
      "grad_norm": 0.5068820714950562,
      "learning_rate": 3.366307142026335e-05,
      "loss": 0.5608,
      "mean_token_accuracy": 0.8233090877532959,
      "num_tokens": 495188689.0,
      "step": 14780
    },
    {
      "epoch": 0.8825811843361987,
      "grad_norm": 0.5439766049385071,
      "learning_rate": 3.365239101826779e-05,
      "loss": 0.5311,
      "mean_token_accuracy": 0.8304664254188537,
      "num_tokens": 495356369.0,
      "step": 14785
    },
    {
      "epoch": 0.8828796561604585,
      "grad_norm": 0.5392639636993408,
      "learning_rate": 3.364170911774384e-05,
      "loss": 0.51,
      "mean_token_accuracy": 0.8345580458641052,
      "num_tokens": 495524049.0,
      "step": 14790
    },
    {
      "epoch": 0.8831781279847183,
      "grad_norm": 0.4509327709674835,
      "learning_rate": 3.3631025721293256e-05,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8354348063468933,
      "num_tokens": 495691729.0,
      "step": 14795
    },
    {
      "epoch": 0.8834765998089781,
      "grad_norm": 0.5160630941390991,
      "learning_rate": 3.362034083151821e-05,
      "loss": 0.554,
      "mean_token_accuracy": 0.8232375025749207,
      "num_tokens": 495859409.0,
      "step": 14800
    },
    {
      "epoch": 0.8837750716332379,
      "grad_norm": 0.47062546014785767,
      "learning_rate": 3.360965445102118e-05,
      "loss": 0.4815,
      "mean_token_accuracy": 0.843033516407013,
      "num_tokens": 496027089.0,
      "step": 14805
    },
    {
      "epoch": 0.8840735434574976,
      "grad_norm": 0.5308709144592285,
      "learning_rate": 3.359896658240503e-05,
      "loss": 0.5249,
      "mean_token_accuracy": 0.8304843068122864,
      "num_tokens": 496194769.0,
      "step": 14810
    },
    {
      "epoch": 0.8843720152817574,
      "grad_norm": 0.5363506078720093,
      "learning_rate": 3.3588277228273e-05,
      "loss": 0.5065,
      "mean_token_accuracy": 0.8359596729278564,
      "num_tokens": 496362449.0,
      "step": 14815
    },
    {
      "epoch": 0.8846704871060171,
      "grad_norm": 0.5338091850280762,
      "learning_rate": 3.357758639122867e-05,
      "loss": 0.5453,
      "mean_token_accuracy": 0.8241619825363159,
      "num_tokens": 496530129.0,
      "step": 14820
    },
    {
      "epoch": 0.8849689589302769,
      "grad_norm": 0.5140828490257263,
      "learning_rate": 3.356689407387601e-05,
      "loss": 0.5149,
      "mean_token_accuracy": 0.8358046054840088,
      "num_tokens": 496697809.0,
      "step": 14825
    },
    {
      "epoch": 0.8852674307545367,
      "grad_norm": 0.5231084823608398,
      "learning_rate": 3.355620027881929e-05,
      "loss": 0.5737,
      "mean_token_accuracy": 0.8169211506843567,
      "num_tokens": 496865489.0,
      "step": 14830
    },
    {
      "epoch": 0.8855659025787965,
      "grad_norm": 0.4676034450531006,
      "learning_rate": 3.354550500866322e-05,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8310091853141784,
      "num_tokens": 497033169.0,
      "step": 14835
    },
    {
      "epoch": 0.8858643744030563,
      "grad_norm": 0.490128755569458,
      "learning_rate": 3.353480826601283e-05,
      "loss": 0.5775,
      "mean_token_accuracy": 0.8170165896415711,
      "num_tokens": 497200849.0,
      "step": 14840
    },
    {
      "epoch": 0.8861628462273161,
      "grad_norm": 0.501987874507904,
      "learning_rate": 3.352411005347348e-05,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8287725210189819,
      "num_tokens": 497368529.0,
      "step": 14845
    },
    {
      "epoch": 0.8864613180515759,
      "grad_norm": 0.4809821546077728,
      "learning_rate": 3.3513410373650943e-05,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8153644442558289,
      "num_tokens": 497536209.0,
      "step": 14850
    },
    {
      "epoch": 0.8867597898758357,
      "grad_norm": 0.5100000500679016,
      "learning_rate": 3.350270922915131e-05,
      "loss": 0.4947,
      "mean_token_accuracy": 0.8402660012245178,
      "num_tokens": 497703889.0,
      "step": 14855
    },
    {
      "epoch": 0.8870582617000955,
      "grad_norm": 0.4943197965621948,
      "learning_rate": 3.3492006622581055e-05,
      "loss": 0.5804,
      "mean_token_accuracy": 0.8168376445770263,
      "num_tokens": 497871569.0,
      "step": 14860
    },
    {
      "epoch": 0.8873567335243553,
      "grad_norm": 0.4699857532978058,
      "learning_rate": 3.348130255654698e-05,
      "loss": 0.4959,
      "mean_token_accuracy": 0.8403614521026611,
      "num_tokens": 498039249.0,
      "step": 14865
    },
    {
      "epoch": 0.8876552053486151,
      "grad_norm": 0.5275648236274719,
      "learning_rate": 3.347059703365627e-05,
      "loss": 0.5614,
      "mean_token_accuracy": 0.8211976647377014,
      "num_tokens": 498206929.0,
      "step": 14870
    },
    {
      "epoch": 0.8879536771728749,
      "grad_norm": 0.44864362478256226,
      "learning_rate": 3.345989005651645e-05,
      "loss": 0.4876,
      "mean_token_accuracy": 0.8426637172698974,
      "num_tokens": 498374609.0,
      "step": 14875
    },
    {
      "epoch": 0.8882521489971347,
      "grad_norm": 0.5722591280937195,
      "learning_rate": 3.34491816277354e-05,
      "loss": 0.5269,
      "mean_token_accuracy": 0.830400812625885,
      "num_tokens": 498542289.0,
      "step": 14880
    },
    {
      "epoch": 0.8885506208213945,
      "grad_norm": 0.470558762550354,
      "learning_rate": 3.343847174992135e-05,
      "loss": 0.5291,
      "mean_token_accuracy": 0.8319873571395874,
      "num_tokens": 498709969.0,
      "step": 14885
    },
    {
      "epoch": 0.8888490926456543,
      "grad_norm": 0.47396644949913025,
      "learning_rate": 3.342776042568289e-05,
      "loss": 0.4856,
      "mean_token_accuracy": 0.8426637291908264,
      "num_tokens": 498877649.0,
      "step": 14890
    },
    {
      "epoch": 0.889147564469914,
      "grad_norm": 0.47508078813552856,
      "learning_rate": 3.341704765762897e-05,
      "loss": 0.5014,
      "mean_token_accuracy": 0.838208281993866,
      "num_tokens": 499045329.0,
      "step": 14895
    },
    {
      "epoch": 0.8894460362941738,
      "grad_norm": 0.49763280153274536,
      "learning_rate": 3.340633344836887e-05,
      "loss": 0.5244,
      "mean_token_accuracy": 0.832118570804596,
      "num_tokens": 499213009.0,
      "step": 14900
    },
    {
      "epoch": 0.8897445081184336,
      "grad_norm": 0.5509679913520813,
      "learning_rate": 3.3395617800512254e-05,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8350829005241394,
      "num_tokens": 499380689.0,
      "step": 14905
    },
    {
      "epoch": 0.8900429799426934,
      "grad_norm": 0.5063443183898926,
      "learning_rate": 3.338490071666908e-05,
      "loss": 0.5612,
      "mean_token_accuracy": 0.822456157207489,
      "num_tokens": 499548369.0,
      "step": 14910
    },
    {
      "epoch": 0.8903414517669532,
      "grad_norm": 0.5468242168426514,
      "learning_rate": 3.337418219944974e-05,
      "loss": 0.5233,
      "mean_token_accuracy": 0.8337766885757446,
      "num_tokens": 499716049.0,
      "step": 14915
    },
    {
      "epoch": 0.890639923591213,
      "grad_norm": 0.6402581930160522,
      "learning_rate": 3.336346225146488e-05,
      "loss": 0.5352,
      "mean_token_accuracy": 0.8274354934692383,
      "num_tokens": 499875788.0,
      "step": 14920
    },
    {
      "epoch": 0.8909383954154728,
      "grad_norm": 0.5565267205238342,
      "learning_rate": 3.3352740875325575e-05,
      "loss": 0.5557,
      "mean_token_accuracy": 0.8220744252204895,
      "num_tokens": 500043468.0,
      "step": 14925
    },
    {
      "epoch": 0.8912368672397326,
      "grad_norm": 0.55153489112854,
      "learning_rate": 3.334201807364319e-05,
      "loss": 0.4965,
      "mean_token_accuracy": 0.8384945750236511,
      "num_tokens": 500211148.0,
      "step": 14930
    },
    {
      "epoch": 0.8915353390639924,
      "grad_norm": 0.44260117411613464,
      "learning_rate": 3.333129384902948e-05,
      "loss": 0.4923,
      "mean_token_accuracy": 0.8412859201431274,
      "num_tokens": 500378828.0,
      "step": 14935
    },
    {
      "epoch": 0.8918338108882522,
      "grad_norm": 0.5871852040290833,
      "learning_rate": 3.332056820409651e-05,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8241441011428833,
      "num_tokens": 500546508.0,
      "step": 14940
    },
    {
      "epoch": 0.892132282712512,
      "grad_norm": 0.48628076910972595,
      "learning_rate": 3.330984114145673e-05,
      "loss": 0.5391,
      "mean_token_accuracy": 0.8266790032386779,
      "num_tokens": 500714188.0,
      "step": 14945
    },
    {
      "epoch": 0.8924307545367717,
      "grad_norm": 0.5118081569671631,
      "learning_rate": 3.32991126637229e-05,
      "loss": 0.5234,
      "mean_token_accuracy": 0.8316950917243957,
      "num_tokens": 500881868.0,
      "step": 14950
    },
    {
      "epoch": 0.8927292263610315,
      "grad_norm": 0.4748677611351013,
      "learning_rate": 3.3288382773508145e-05,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8354825258255005,
      "num_tokens": 501049548.0,
      "step": 14955
    },
    {
      "epoch": 0.8930276981852913,
      "grad_norm": 0.4990309476852417,
      "learning_rate": 3.3277651473425905e-05,
      "loss": 0.5123,
      "mean_token_accuracy": 0.8350590467453003,
      "num_tokens": 501217228.0,
      "step": 14960
    },
    {
      "epoch": 0.8933261700095511,
      "grad_norm": 0.47078853845596313,
      "learning_rate": 3.326691876609002e-05,
      "loss": 0.4982,
      "mean_token_accuracy": 0.8375462174415589,
      "num_tokens": 501384908.0,
      "step": 14965
    },
    {
      "epoch": 0.8936246418338109,
      "grad_norm": 0.4765874445438385,
      "learning_rate": 3.325618465411463e-05,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8348443269729614,
      "num_tokens": 501552588.0,
      "step": 14970
    },
    {
      "epoch": 0.8939231136580706,
      "grad_norm": 0.6424722075462341,
      "learning_rate": 3.324544914011421e-05,
      "loss": 0.5201,
      "mean_token_accuracy": 0.8334903955459595,
      "num_tokens": 501720268.0,
      "step": 14975
    },
    {
      "epoch": 0.8942215854823304,
      "grad_norm": 0.5706070065498352,
      "learning_rate": 3.323471222670361e-05,
      "loss": 0.526,
      "mean_token_accuracy": 0.8302039742469788,
      "num_tokens": 501887948.0,
      "step": 14980
    },
    {
      "epoch": 0.8945200573065902,
      "grad_norm": 0.5030493140220642,
      "learning_rate": 3.322397391649799e-05,
      "loss": 0.5254,
      "mean_token_accuracy": 0.8320827841758728,
      "num_tokens": 502055628.0,
      "step": 14985
    },
    {
      "epoch": 0.89481852913085,
      "grad_norm": 0.44015440344810486,
      "learning_rate": 3.3213234212112875e-05,
      "loss": 0.495,
      "mean_token_accuracy": 0.8401825070381165,
      "num_tokens": 502223308.0,
      "step": 14990
    },
    {
      "epoch": 0.8951170009551098,
      "grad_norm": 0.5025378465652466,
      "learning_rate": 3.3202493116164104e-05,
      "loss": 0.5521,
      "mean_token_accuracy": 0.8257246851921082,
      "num_tokens": 502390988.0,
      "step": 14995
    },
    {
      "epoch": 0.8954154727793696,
      "grad_norm": 0.4992840588092804,
      "learning_rate": 3.319175063126789e-05,
      "loss": 0.5267,
      "mean_token_accuracy": 0.8307884931564331,
      "num_tokens": 502558668.0,
      "step": 15000
    },
    {
      "epoch": 0.8957139446036294,
      "grad_norm": 0.5053568482398987,
      "learning_rate": 3.318100676004075e-05,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8335261702537536,
      "num_tokens": 502726348.0,
      "step": 15005
    },
    {
      "epoch": 0.8960124164278892,
      "grad_norm": 0.5102866291999817,
      "learning_rate": 3.317026150509955e-05,
      "loss": 0.5464,
      "mean_token_accuracy": 0.8258201122283936,
      "num_tokens": 502894028.0,
      "step": 15010
    },
    {
      "epoch": 0.896310888252149,
      "grad_norm": 0.569139838218689,
      "learning_rate": 3.315951486906151e-05,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8341941952705383,
      "num_tokens": 503061708.0,
      "step": 15015
    },
    {
      "epoch": 0.8966093600764088,
      "grad_norm": 0.5200099349021912,
      "learning_rate": 3.314876685454415e-05,
      "loss": 0.527,
      "mean_token_accuracy": 0.8316772103309631,
      "num_tokens": 503229388.0,
      "step": 15020
    },
    {
      "epoch": 0.8969078319006686,
      "grad_norm": 0.5570427179336548,
      "learning_rate": 3.313801746416536e-05,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8284444689750672,
      "num_tokens": 503397068.0,
      "step": 15025
    },
    {
      "epoch": 0.8972063037249284,
      "grad_norm": 0.4885665774345398,
      "learning_rate": 3.312726670054335e-05,
      "loss": 0.4869,
      "mean_token_accuracy": 0.8417809844017029,
      "num_tokens": 503564748.0,
      "step": 15030
    },
    {
      "epoch": 0.8975047755491882,
      "grad_norm": 0.6092052459716797,
      "learning_rate": 3.311651456629667e-05,
      "loss": 0.5227,
      "mean_token_accuracy": 0.8317249178886413,
      "num_tokens": 503732428.0,
      "step": 15035
    },
    {
      "epoch": 0.897803247373448,
      "grad_norm": 0.5439836382865906,
      "learning_rate": 3.31057610640442e-05,
      "loss": 0.5753,
      "mean_token_accuracy": 0.8178217768669128,
      "num_tokens": 503900108.0,
      "step": 15040
    },
    {
      "epoch": 0.8981017191977078,
      "grad_norm": 0.4790586531162262,
      "learning_rate": 3.309500619640515e-05,
      "loss": 0.5009,
      "mean_token_accuracy": 0.838846480846405,
      "num_tokens": 504067788.0,
      "step": 15045
    },
    {
      "epoch": 0.8984001910219676,
      "grad_norm": 0.49048352241516113,
      "learning_rate": 3.308424996599906e-05,
      "loss": 0.5148,
      "mean_token_accuracy": 0.8354288339614868,
      "num_tokens": 504235468.0,
      "step": 15050
    },
    {
      "epoch": 0.8986986628462273,
      "grad_norm": 0.44867396354675293,
      "learning_rate": 3.3073492375445835e-05,
      "loss": 0.5035,
      "mean_token_accuracy": 0.8378921508789062,
      "num_tokens": 504403148.0,
      "step": 15055
    },
    {
      "epoch": 0.8989971346704871,
      "grad_norm": 0.4918707609176636,
      "learning_rate": 3.3062733427365664e-05,
      "loss": 0.5196,
      "mean_token_accuracy": 0.8321305036544799,
      "num_tokens": 504570828.0,
      "step": 15060
    },
    {
      "epoch": 0.8992956064947469,
      "grad_norm": 0.5182741284370422,
      "learning_rate": 3.305197312437908e-05,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8293630003929138,
      "num_tokens": 504738508.0,
      "step": 15065
    },
    {
      "epoch": 0.8995940783190067,
      "grad_norm": 0.4851759076118469,
      "learning_rate": 3.3041211469106966e-05,
      "loss": 0.522,
      "mean_token_accuracy": 0.8325301289558411,
      "num_tokens": 504906188.0,
      "step": 15070
    },
    {
      "epoch": 0.8998925501432665,
      "grad_norm": 0.4884530007839203,
      "learning_rate": 3.3030448464170524e-05,
      "loss": 0.5307,
      "mean_token_accuracy": 0.8286472678184509,
      "num_tokens": 505073868.0,
      "step": 15075
    },
    {
      "epoch": 0.9001910219675263,
      "grad_norm": 0.5374017357826233,
      "learning_rate": 3.301968411219128e-05,
      "loss": 0.5587,
      "mean_token_accuracy": 0.8218418240547181,
      "num_tokens": 505241548.0,
      "step": 15080
    },
    {
      "epoch": 0.9004894937917861,
      "grad_norm": 0.4759073555469513,
      "learning_rate": 3.3008918415791083e-05,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8395144939422607,
      "num_tokens": 505409228.0,
      "step": 15085
    },
    {
      "epoch": 0.9007879656160458,
      "grad_norm": 0.4784984290599823,
      "learning_rate": 3.299815137759213e-05,
      "loss": 0.4691,
      "mean_token_accuracy": 0.848395574092865,
      "num_tokens": 505576908.0,
      "step": 15090
    },
    {
      "epoch": 0.9010864374403056,
      "grad_norm": 0.48732420802116394,
      "learning_rate": 3.298738300021692e-05,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8292556285858155,
      "num_tokens": 505744588.0,
      "step": 15095
    },
    {
      "epoch": 0.9013849092645654,
      "grad_norm": 0.4827894866466522,
      "learning_rate": 3.29766132862883e-05,
      "loss": 0.5146,
      "mean_token_accuracy": 0.8375394463539123,
      "num_tokens": 505909536.0,
      "step": 15100
    },
    {
      "epoch": 0.9016833810888252,
      "grad_norm": 0.5672697424888611,
      "learning_rate": 3.296584223842942e-05,
      "loss": 0.5543,
      "mean_token_accuracy": 0.8234224081039428,
      "num_tokens": 506077216.0,
      "step": 15105
    },
    {
      "epoch": 0.901981852913085,
      "grad_norm": 0.5556638240814209,
      "learning_rate": 3.295506985926377e-05,
      "loss": 0.5398,
      "mean_token_accuracy": 0.829535973072052,
      "num_tokens": 506244896.0,
      "step": 15110
    },
    {
      "epoch": 0.9022803247373448,
      "grad_norm": 0.5258305072784424,
      "learning_rate": 3.294429615141518e-05,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8422641038894654,
      "num_tokens": 506412576.0,
      "step": 15115
    },
    {
      "epoch": 0.9025787965616046,
      "grad_norm": 0.5012132525444031,
      "learning_rate": 3.2933521117507766e-05,
      "loss": 0.5078,
      "mean_token_accuracy": 0.836454713344574,
      "num_tokens": 506580256.0,
      "step": 15120
    },
    {
      "epoch": 0.9028772683858644,
      "grad_norm": 0.4982709586620331,
      "learning_rate": 3.2922744760165994e-05,
      "loss": 0.5279,
      "mean_token_accuracy": 0.8287367343902587,
      "num_tokens": 506747936.0,
      "step": 15125
    },
    {
      "epoch": 0.9031757402101241,
      "grad_norm": 0.4942888915538788,
      "learning_rate": 3.291196708201464e-05,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8272038698196411,
      "num_tokens": 506915616.0,
      "step": 15130
    },
    {
      "epoch": 0.9034742120343839,
      "grad_norm": 0.5942230820655823,
      "learning_rate": 3.290118808567881e-05,
      "loss": 0.5732,
      "mean_token_accuracy": 0.8190623998641968,
      "num_tokens": 507083296.0,
      "step": 15135
    },
    {
      "epoch": 0.9037726838586437,
      "grad_norm": 0.48611149191856384,
      "learning_rate": 3.2890407773783935e-05,
      "loss": 0.5161,
      "mean_token_accuracy": 0.8341763138771057,
      "num_tokens": 507250976.0,
      "step": 15140
    },
    {
      "epoch": 0.9040711556829035,
      "grad_norm": 0.47812941670417786,
      "learning_rate": 3.2879626148955745e-05,
      "loss": 0.5482,
      "mean_token_accuracy": 0.8242932200431824,
      "num_tokens": 507418656.0,
      "step": 15145
    },
    {
      "epoch": 0.9043696275071633,
      "grad_norm": 0.4972628355026245,
      "learning_rate": 3.2868843213820305e-05,
      "loss": 0.4946,
      "mean_token_accuracy": 0.8406954526901245,
      "num_tokens": 507586336.0,
      "step": 15150
    },
    {
      "epoch": 0.9046680993314231,
      "grad_norm": 0.4947787821292877,
      "learning_rate": 3.285805897100401e-05,
      "loss": 0.5503,
      "mean_token_accuracy": 0.8235118746757507,
      "num_tokens": 507754016.0,
      "step": 15155
    },
    {
      "epoch": 0.9049665711556829,
      "grad_norm": 0.5631944537162781,
      "learning_rate": 3.2847273423133546e-05,
      "loss": 0.553,
      "mean_token_accuracy": 0.8239114880561829,
      "num_tokens": 507921696.0,
      "step": 15160
    },
    {
      "epoch": 0.9052650429799427,
      "grad_norm": 0.47290360927581787,
      "learning_rate": 3.283648657283595e-05,
      "loss": 0.5292,
      "mean_token_accuracy": 0.829023027420044,
      "num_tokens": 508089376.0,
      "step": 15165
    },
    {
      "epoch": 0.9055635148042025,
      "grad_norm": 0.6200077533721924,
      "learning_rate": 3.2825698422738546e-05,
      "loss": 0.5008,
      "mean_token_accuracy": 0.838834547996521,
      "num_tokens": 508257056.0,
      "step": 15170
    },
    {
      "epoch": 0.9058619866284623,
      "grad_norm": 0.483733206987381,
      "learning_rate": 3.281490897546899e-05,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8374030828475952,
      "num_tokens": 508424736.0,
      "step": 15175
    },
    {
      "epoch": 0.9061604584527221,
      "grad_norm": 0.46810218691825867,
      "learning_rate": 3.280411823365525e-05,
      "loss": 0.5169,
      "mean_token_accuracy": 0.8342836618423461,
      "num_tokens": 508592416.0,
      "step": 15180
    },
    {
      "epoch": 0.9064589302769819,
      "grad_norm": 0.565943717956543,
      "learning_rate": 3.2793326199925615e-05,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8266611099243164,
      "num_tokens": 508760096.0,
      "step": 15185
    },
    {
      "epoch": 0.9067574021012417,
      "grad_norm": 0.5199659466743469,
      "learning_rate": 3.2782532876908674e-05,
      "loss": 0.5387,
      "mean_token_accuracy": 0.8281760811805725,
      "num_tokens": 508927776.0,
      "step": 15190
    },
    {
      "epoch": 0.9070558739255015,
      "grad_norm": 0.5202977061271667,
      "learning_rate": 3.2771738267233346e-05,
      "loss": 0.4923,
      "mean_token_accuracy": 0.8416557312011719,
      "num_tokens": 509095456.0,
      "step": 15195
    },
    {
      "epoch": 0.9073543457497613,
      "grad_norm": 0.5077999234199524,
      "learning_rate": 3.2760942373528867e-05,
      "loss": 0.537,
      "mean_token_accuracy": 0.8274662971496582,
      "num_tokens": 509263136.0,
      "step": 15200
    },
    {
      "epoch": 0.907652817574021,
      "grad_norm": 0.5351226329803467,
      "learning_rate": 3.275014519842476e-05,
      "loss": 0.5569,
      "mean_token_accuracy": 0.8228319168090821,
      "num_tokens": 509430816.0,
      "step": 15205
    },
    {
      "epoch": 0.9079512893982808,
      "grad_norm": 0.50678950548172,
      "learning_rate": 3.273934674455088e-05,
      "loss": 0.5075,
      "mean_token_accuracy": 0.8355600595474243,
      "num_tokens": 509598496.0,
      "step": 15210
    },
    {
      "epoch": 0.9082497612225406,
      "grad_norm": 0.556103527545929,
      "learning_rate": 3.2728547014537395e-05,
      "loss": 0.5811,
      "mean_token_accuracy": 0.8161576867103577,
      "num_tokens": 509766176.0,
      "step": 15215
    },
    {
      "epoch": 0.9085482330468004,
      "grad_norm": 0.49839338660240173,
      "learning_rate": 3.271774601101478e-05,
      "loss": 0.5233,
      "mean_token_accuracy": 0.8324048638343811,
      "num_tokens": 509933856.0,
      "step": 15220
    },
    {
      "epoch": 0.9088467048710601,
      "grad_norm": 0.5003037452697754,
      "learning_rate": 3.270694373661382e-05,
      "loss": 0.5239,
      "mean_token_accuracy": 0.8318203449249267,
      "num_tokens": 510101536.0,
      "step": 15225
    },
    {
      "epoch": 0.9091451766953199,
      "grad_norm": 0.5407162308692932,
      "learning_rate": 3.26961401939656e-05,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8290170431137085,
      "num_tokens": 510269216.0,
      "step": 15230
    },
    {
      "epoch": 0.9094436485195797,
      "grad_norm": 0.5040706992149353,
      "learning_rate": 3.268533538570153e-05,
      "loss": 0.5319,
      "mean_token_accuracy": 0.830388879776001,
      "num_tokens": 510436896.0,
      "step": 15235
    },
    {
      "epoch": 0.9097421203438395,
      "grad_norm": 0.5021921992301941,
      "learning_rate": 3.2674529314453304e-05,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8309733867645264,
      "num_tokens": 510604576.0,
      "step": 15240
    },
    {
      "epoch": 0.9100405921680993,
      "grad_norm": 0.4788692593574524,
      "learning_rate": 3.266372198285296e-05,
      "loss": 0.5191,
      "mean_token_accuracy": 0.8329297423362731,
      "num_tokens": 510772256.0,
      "step": 15245
    },
    {
      "epoch": 0.9103390639923591,
      "grad_norm": 0.47935912013053894,
      "learning_rate": 3.265291339353282e-05,
      "loss": 0.5138,
      "mean_token_accuracy": 0.8348860740661621,
      "num_tokens": 510939936.0,
      "step": 15250
    },
    {
      "epoch": 0.9106375358166189,
      "grad_norm": 0.5384992361068726,
      "learning_rate": 3.264210354912551e-05,
      "loss": 0.5721,
      "mean_token_accuracy": 0.8200584530830384,
      "num_tokens": 511107616.0,
      "step": 15255
    },
    {
      "epoch": 0.9109360076408787,
      "grad_norm": 0.5259842872619629,
      "learning_rate": 3.263129245226396e-05,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8408326387405396,
      "num_tokens": 511275296.0,
      "step": 15260
    },
    {
      "epoch": 0.9112344794651385,
      "grad_norm": 0.5167257189750671,
      "learning_rate": 3.262048010558143e-05,
      "loss": 0.5191,
      "mean_token_accuracy": 0.8357509255409241,
      "num_tokens": 511442976.0,
      "step": 15265
    },
    {
      "epoch": 0.9115329512893983,
      "grad_norm": 0.4853335916996002,
      "learning_rate": 3.260966651171145e-05,
      "loss": 0.5279,
      "mean_token_accuracy": 0.8301801323890686,
      "num_tokens": 511610656.0,
      "step": 15270
    },
    {
      "epoch": 0.9118314231136581,
      "grad_norm": 0.5171031951904297,
      "learning_rate": 3.259885167328788e-05,
      "loss": 0.5494,
      "mean_token_accuracy": 0.8258797645568847,
      "num_tokens": 511778336.0,
      "step": 15275
    },
    {
      "epoch": 0.9121298949379179,
      "grad_norm": 0.5895015001296997,
      "learning_rate": 3.258803559294487e-05,
      "loss": 0.5226,
      "mean_token_accuracy": 0.832004714012146,
      "num_tokens": 511943286.0,
      "step": 15280
    },
    {
      "epoch": 0.9124283667621776,
      "grad_norm": 0.49766111373901367,
      "learning_rate": 3.257721827331688e-05,
      "loss": 0.504,
      "mean_token_accuracy": 0.8365501642227173,
      "num_tokens": 512110966.0,
      "step": 15285
    },
    {
      "epoch": 0.9127268385864374,
      "grad_norm": 0.4960426390171051,
      "learning_rate": 3.256639971703866e-05,
      "loss": 0.5209,
      "mean_token_accuracy": 0.8328641295433045,
      "num_tokens": 512278646.0,
      "step": 15290
    },
    {
      "epoch": 0.9130253104106972,
      "grad_norm": 0.49136650562286377,
      "learning_rate": 3.255557992674527e-05,
      "loss": 0.4948,
      "mean_token_accuracy": 0.8410950779914856,
      "num_tokens": 512446326.0,
      "step": 15295
    },
    {
      "epoch": 0.913323782234957,
      "grad_norm": 0.4935888648033142,
      "learning_rate": 3.254475890507209e-05,
      "loss": 0.5305,
      "mean_token_accuracy": 0.8296314001083374,
      "num_tokens": 512614006.0,
      "step": 15300
    },
    {
      "epoch": 0.9136222540592168,
      "grad_norm": 1.0330394506454468,
      "learning_rate": 3.253393665465475e-05,
      "loss": 0.532,
      "mean_token_accuracy": 0.8303053855895997,
      "num_tokens": 512781686.0,
      "step": 15305
    },
    {
      "epoch": 0.9139207258834766,
      "grad_norm": 0.4904855787754059,
      "learning_rate": 3.252311317812923e-05,
      "loss": 0.5296,
      "mean_token_accuracy": 0.8292735338211059,
      "num_tokens": 512949366.0,
      "step": 15310
    },
    {
      "epoch": 0.9142191977077364,
      "grad_norm": 0.45952892303466797,
      "learning_rate": 3.251228847813177e-05,
      "loss": 0.4818,
      "mean_token_accuracy": 0.8445723414421081,
      "num_tokens": 513117046.0,
      "step": 15315
    },
    {
      "epoch": 0.9145176695319962,
      "grad_norm": 0.4795607328414917,
      "learning_rate": 3.250146255729895e-05,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8371108174324036,
      "num_tokens": 513284726.0,
      "step": 15320
    },
    {
      "epoch": 0.914816141356256,
      "grad_norm": 0.5202341675758362,
      "learning_rate": 3.2490635418267586e-05,
      "loss": 0.5407,
      "mean_token_accuracy": 0.8274484038352966,
      "num_tokens": 513452406.0,
      "step": 15325
    },
    {
      "epoch": 0.9151146131805158,
      "grad_norm": 0.5209221243858337,
      "learning_rate": 3.247980706367487e-05,
      "loss": 0.5186,
      "mean_token_accuracy": 0.8329595565795899,
      "num_tokens": 513620086.0,
      "step": 15330
    },
    {
      "epoch": 0.9154130850047756,
      "grad_norm": 0.485456258058548,
      "learning_rate": 3.246897749615821e-05,
      "loss": 0.5578,
      "mean_token_accuracy": 0.8226350903511047,
      "num_tokens": 513787766.0,
      "step": 15335
    },
    {
      "epoch": 0.9157115568290354,
      "grad_norm": 0.5021275281906128,
      "learning_rate": 3.2458146718355374e-05,
      "loss": 0.5183,
      "mean_token_accuracy": 0.8338244080543518,
      "num_tokens": 513955446.0,
      "step": 15340
    },
    {
      "epoch": 0.9160100286532952,
      "grad_norm": 0.5215170979499817,
      "learning_rate": 3.244731473290438e-05,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8298341870307923,
      "num_tokens": 514123126.0,
      "step": 15345
    },
    {
      "epoch": 0.916308500477555,
      "grad_norm": 0.5435329675674438,
      "learning_rate": 3.2436481542443574e-05,
      "loss": 0.5249,
      "mean_token_accuracy": 0.8318740367889405,
      "num_tokens": 514290806.0,
      "step": 15350
    },
    {
      "epoch": 0.9166069723018148,
      "grad_norm": 0.5269924998283386,
      "learning_rate": 3.2425647149611564e-05,
      "loss": 0.5777,
      "mean_token_accuracy": 0.81725515127182,
      "num_tokens": 514458486.0,
      "step": 15355
    },
    {
      "epoch": 0.9169054441260746,
      "grad_norm": 0.5256832242012024,
      "learning_rate": 3.2414811557047284e-05,
      "loss": 0.5071,
      "mean_token_accuracy": 0.83711678981781,
      "num_tokens": 514626166.0,
      "step": 15360
    },
    {
      "epoch": 0.9172039159503342,
      "grad_norm": 0.48518332839012146,
      "learning_rate": 3.240397476738993e-05,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8411547183990479,
      "num_tokens": 514793846.0,
      "step": 15365
    },
    {
      "epoch": 0.917502387774594,
      "grad_norm": 0.5374515056610107,
      "learning_rate": 3.2393136783279e-05,
      "loss": 0.5411,
      "mean_token_accuracy": 0.825426459312439,
      "num_tokens": 514961526.0,
      "step": 15370
    },
    {
      "epoch": 0.9178008595988538,
      "grad_norm": 0.4761549234390259,
      "learning_rate": 3.238229760735429e-05,
      "loss": 0.4947,
      "mean_token_accuracy": 0.8396695733070374,
      "num_tokens": 515129206.0,
      "step": 15375
    },
    {
      "epoch": 0.9180993314231136,
      "grad_norm": 0.465145081281662,
      "learning_rate": 3.2371457242255884e-05,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8463378190994263,
      "num_tokens": 515296886.0,
      "step": 15380
    },
    {
      "epoch": 0.9183978032473734,
      "grad_norm": 0.5342923998832703,
      "learning_rate": 3.2360615690624144e-05,
      "loss": 0.5286,
      "mean_token_accuracy": 0.8328283309936524,
      "num_tokens": 515464566.0,
      "step": 15385
    },
    {
      "epoch": 0.9186962750716332,
      "grad_norm": 0.526098370552063,
      "learning_rate": 3.234977295509974e-05,
      "loss": 0.4981,
      "mean_token_accuracy": 0.8372599124908447,
      "num_tokens": 515632246.0,
      "step": 15390
    },
    {
      "epoch": 0.918994746895893,
      "grad_norm": 0.46639159321784973,
      "learning_rate": 3.2338929038323615e-05,
      "loss": 0.5273,
      "mean_token_accuracy": 0.8287307620048523,
      "num_tokens": 515799926.0,
      "step": 15395
    },
    {
      "epoch": 0.9192932187201528,
      "grad_norm": 0.5124784111976624,
      "learning_rate": 3.232808394293701e-05,
      "loss": 0.5627,
      "mean_token_accuracy": 0.8195395469665527,
      "num_tokens": 515967606.0,
      "step": 15400
    },
    {
      "epoch": 0.9195916905444126,
      "grad_norm": 0.5939876437187195,
      "learning_rate": 3.231723767158144e-05,
      "loss": 0.5363,
      "mean_token_accuracy": 0.8277108550071717,
      "num_tokens": 516135286.0,
      "step": 15405
    },
    {
      "epoch": 0.9198901623686724,
      "grad_norm": 0.4634951055049896,
      "learning_rate": 3.230639022689871e-05,
      "loss": 0.5402,
      "mean_token_accuracy": 0.8277406573295594,
      "num_tokens": 516302966.0,
      "step": 15410
    },
    {
      "epoch": 0.9201886341929322,
      "grad_norm": 0.5093323588371277,
      "learning_rate": 3.229554161153094e-05,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8295717358589172,
      "num_tokens": 516470646.0,
      "step": 15415
    },
    {
      "epoch": 0.920487106017192,
      "grad_norm": 0.5386112332344055,
      "learning_rate": 3.228469182812049e-05,
      "loss": 0.5675,
      "mean_token_accuracy": 0.8193844676017761,
      "num_tokens": 516638326.0,
      "step": 15420
    },
    {
      "epoch": 0.9207855778414518,
      "grad_norm": 0.48846355080604553,
      "learning_rate": 3.227384087931003e-05,
      "loss": 0.4985,
      "mean_token_accuracy": 0.8384408950805664,
      "num_tokens": 516806006.0,
      "step": 15425
    },
    {
      "epoch": 0.9210840496657116,
      "grad_norm": 0.569942057132721,
      "learning_rate": 3.226298876774251e-05,
      "loss": 0.5326,
      "mean_token_accuracy": 0.8298818945884705,
      "num_tokens": 516973686.0,
      "step": 15430
    },
    {
      "epoch": 0.9213825214899714,
      "grad_norm": 0.5315511226654053,
      "learning_rate": 3.225213549606116e-05,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8307467341423035,
      "num_tokens": 517141366.0,
      "step": 15435
    },
    {
      "epoch": 0.9216809933142311,
      "grad_norm": 0.5086844563484192,
      "learning_rate": 3.224128106690949e-05,
      "loss": 0.5862,
      "mean_token_accuracy": 0.8138733267784118,
      "num_tokens": 517309046.0,
      "step": 15440
    },
    {
      "epoch": 0.9219794651384909,
      "grad_norm": 0.49994051456451416,
      "learning_rate": 3.223042548293131e-05,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8245019674301147,
      "num_tokens": 517476726.0,
      "step": 15445
    },
    {
      "epoch": 0.9222779369627507,
      "grad_norm": 0.5457820296287537,
      "learning_rate": 3.2219568746770685e-05,
      "loss": 0.5359,
      "mean_token_accuracy": 0.826816177368164,
      "num_tokens": 517644406.0,
      "step": 15450
    },
    {
      "epoch": 0.9225764087870105,
      "grad_norm": 0.46678459644317627,
      "learning_rate": 3.2208710861071986e-05,
      "loss": 0.5475,
      "mean_token_accuracy": 0.8279732942581177,
      "num_tokens": 517812086.0,
      "step": 15455
    },
    {
      "epoch": 0.9228748806112703,
      "grad_norm": 0.5285353660583496,
      "learning_rate": 3.219785182847983e-05,
      "loss": 0.5447,
      "mean_token_accuracy": 0.8274722576141358,
      "num_tokens": 517979766.0,
      "step": 15460
    },
    {
      "epoch": 0.9231733524355301,
      "grad_norm": 0.5083525776863098,
      "learning_rate": 3.218699165163916e-05,
      "loss": 0.5481,
      "mean_token_accuracy": 0.8246152877807618,
      "num_tokens": 518147446.0,
      "step": 15465
    },
    {
      "epoch": 0.9234718242597899,
      "grad_norm": 0.5395824313163757,
      "learning_rate": 3.2176130333195155e-05,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8230585575103759,
      "num_tokens": 518315126.0,
      "step": 15470
    },
    {
      "epoch": 0.9237702960840497,
      "grad_norm": 0.5434145331382751,
      "learning_rate": 3.2165267875793305e-05,
      "loss": 0.5257,
      "mean_token_accuracy": 0.8322080492973327,
      "num_tokens": 518482806.0,
      "step": 15475
    },
    {
      "epoch": 0.9240687679083095,
      "grad_norm": 0.5313842296600342,
      "learning_rate": 3.215440428207934e-05,
      "loss": 0.5773,
      "mean_token_accuracy": 0.8145340085029602,
      "num_tokens": 518650096.0,
      "step": 15480
    },
    {
      "epoch": 0.9243672397325693,
      "grad_norm": 0.5660425424575806,
      "learning_rate": 3.2143539554699314e-05,
      "loss": 0.5672,
      "mean_token_accuracy": 0.8219372391700744,
      "num_tokens": 518817776.0,
      "step": 15485
    },
    {
      "epoch": 0.9246657115568291,
      "grad_norm": 0.49238401651382446,
      "learning_rate": 3.2132673696299515e-05,
      "loss": 0.5408,
      "mean_token_accuracy": 0.8271740436553955,
      "num_tokens": 518985456.0,
      "step": 15490
    },
    {
      "epoch": 0.9249641833810889,
      "grad_norm": 0.5256762504577637,
      "learning_rate": 3.212180670952653e-05,
      "loss": 0.559,
      "mean_token_accuracy": 0.8211797595024108,
      "num_tokens": 519153136.0,
      "step": 15495
    },
    {
      "epoch": 0.9252626552053486,
      "grad_norm": 0.5064694881439209,
      "learning_rate": 3.2110938597027204e-05,
      "loss": 0.5478,
      "mean_token_accuracy": 0.8261481642723083,
      "num_tokens": 519320816.0,
      "step": 15500
    },
    {
      "epoch": 0.9255611270296084,
      "grad_norm": 0.5971306562423706,
      "learning_rate": 3.210006936144869e-05,
      "loss": 0.4997,
      "mean_token_accuracy": 0.83870929479599,
      "num_tokens": 519488496.0,
      "step": 15505
    },
    {
      "epoch": 0.9258595988538681,
      "grad_norm": 0.5045725107192993,
      "learning_rate": 3.208919900543836e-05,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8391924142837525,
      "num_tokens": 519656176.0,
      "step": 15510
    },
    {
      "epoch": 0.9261580706781279,
      "grad_norm": 0.4632502496242523,
      "learning_rate": 3.2078327531643903e-05,
      "loss": 0.5391,
      "mean_token_accuracy": 0.8304144024848938,
      "num_tokens": 519822501.0,
      "step": 15515
    },
    {
      "epoch": 0.9264565425023877,
      "grad_norm": 0.4779451787471771,
      "learning_rate": 3.206745494271328e-05,
      "loss": 0.5252,
      "mean_token_accuracy": 0.8307348132133484,
      "num_tokens": 519990181.0,
      "step": 15520
    },
    {
      "epoch": 0.9267550143266475,
      "grad_norm": 0.5667571425437927,
      "learning_rate": 3.20565812412947e-05,
      "loss": 0.5341,
      "mean_token_accuracy": 0.8276976108551025,
      "num_tokens": 520146945.0,
      "step": 15525
    },
    {
      "epoch": 0.9270534861509073,
      "grad_norm": 0.5411862730979919,
      "learning_rate": 3.204570643003665e-05,
      "loss": 0.5372,
      "mean_token_accuracy": 0.827317190170288,
      "num_tokens": 520314625.0,
      "step": 15530
    },
    {
      "epoch": 0.9273519579751671,
      "grad_norm": 0.5148149132728577,
      "learning_rate": 3.20348305115879e-05,
      "loss": 0.4926,
      "mean_token_accuracy": 0.8397113084793091,
      "num_tokens": 520482305.0,
      "step": 15535
    },
    {
      "epoch": 0.9276504297994269,
      "grad_norm": 0.5398635268211365,
      "learning_rate": 3.2023953488597466e-05,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8288023233413696,
      "num_tokens": 520649985.0,
      "step": 15540
    },
    {
      "epoch": 0.9279489016236867,
      "grad_norm": 0.4518938362598419,
      "learning_rate": 3.201307536371466e-05,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8311642527580261,
      "num_tokens": 520817665.0,
      "step": 15545
    },
    {
      "epoch": 0.9282473734479465,
      "grad_norm": 0.5684276819229126,
      "learning_rate": 3.200219613958905e-05,
      "loss": 0.5576,
      "mean_token_accuracy": 0.8230645418167114,
      "num_tokens": 520985345.0,
      "step": 15550
    },
    {
      "epoch": 0.9285458452722063,
      "grad_norm": 0.5751977562904358,
      "learning_rate": 3.199131581887047e-05,
      "loss": 0.5656,
      "mean_token_accuracy": 0.8203149199485779,
      "num_tokens": 521153025.0,
      "step": 15555
    },
    {
      "epoch": 0.9288443170964661,
      "grad_norm": 0.569940447807312,
      "learning_rate": 3.1980434404209024e-05,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8360431909561157,
      "num_tokens": 521320705.0,
      "step": 15560
    },
    {
      "epoch": 0.9291427889207259,
      "grad_norm": 0.550160825252533,
      "learning_rate": 3.196955189825508e-05,
      "loss": 0.5296,
      "mean_token_accuracy": 0.830013120174408,
      "num_tokens": 521488385.0,
      "step": 15565
    },
    {
      "epoch": 0.9294412607449857,
      "grad_norm": 0.48204129934310913,
      "learning_rate": 3.195866830365927e-05,
      "loss": 0.5179,
      "mean_token_accuracy": 0.8323644518852233,
      "num_tokens": 521652077.0,
      "step": 15570
    },
    {
      "epoch": 0.9297397325692455,
      "grad_norm": 0.5202948451042175,
      "learning_rate": 3.194778362307249e-05,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8378188848495484,
      "num_tokens": 521808106.0,
      "step": 15575
    },
    {
      "epoch": 0.9300382043935053,
      "grad_norm": 0.4583127498626709,
      "learning_rate": 3.193689785914592e-05,
      "loss": 0.5315,
      "mean_token_accuracy": 0.8298461318016053,
      "num_tokens": 521975786.0,
      "step": 15580
    },
    {
      "epoch": 0.9303366762177651,
      "grad_norm": 0.5228418707847595,
      "learning_rate": 3.192601101453098e-05,
      "loss": 0.5433,
      "mean_token_accuracy": 0.8251401662826539,
      "num_tokens": 522143466.0,
      "step": 15585
    },
    {
      "epoch": 0.9306351480420249,
      "grad_norm": 0.5833495855331421,
      "learning_rate": 3.191512309187936e-05,
      "loss": 0.5277,
      "mean_token_accuracy": 0.830764639377594,
      "num_tokens": 522311146.0,
      "step": 15590
    },
    {
      "epoch": 0.9309336198662846,
      "grad_norm": 0.5323729515075684,
      "learning_rate": 3.190423409384302e-05,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8311463713645935,
      "num_tokens": 522478826.0,
      "step": 15595
    },
    {
      "epoch": 0.9312320916905444,
      "grad_norm": 0.4731408953666687,
      "learning_rate": 3.1893344023074176e-05,
      "loss": 0.538,
      "mean_token_accuracy": 0.8266505479812623,
      "num_tokens": 522642120.0,
      "step": 15600
    },
    {
      "epoch": 0.9315305635148042,
      "grad_norm": 0.5499926209449768,
      "learning_rate": 3.1882452882225296e-05,
      "loss": 0.5458,
      "mean_token_accuracy": 0.8248956203460693,
      "num_tokens": 522809800.0,
      "step": 15605
    },
    {
      "epoch": 0.931829035339064,
      "grad_norm": 0.480619877576828,
      "learning_rate": 3.187156067394913e-05,
      "loss": 0.5583,
      "mean_token_accuracy": 0.8228319168090821,
      "num_tokens": 522977480.0,
      "step": 15610
    },
    {
      "epoch": 0.9321275071633238,
      "grad_norm": 0.5242154002189636,
      "learning_rate": 3.1860667400898676e-05,
      "loss": 0.5567,
      "mean_token_accuracy": 0.8220445990562439,
      "num_tokens": 523145160.0,
      "step": 15615
    },
    {
      "epoch": 0.9324259789875836,
      "grad_norm": 0.47804081439971924,
      "learning_rate": 3.184977306572718e-05,
      "loss": 0.5617,
      "mean_token_accuracy": 0.8214481592178344,
      "num_tokens": 523312840.0,
      "step": 15620
    },
    {
      "epoch": 0.9327244508118434,
      "grad_norm": 0.4681434631347656,
      "learning_rate": 3.1838877671088166e-05,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8425921440124512,
      "num_tokens": 523480520.0,
      "step": 15625
    },
    {
      "epoch": 0.9330229226361032,
      "grad_norm": 0.495015412569046,
      "learning_rate": 3.1827981219635415e-05,
      "loss": 0.5559,
      "mean_token_accuracy": 0.8237444877624511,
      "num_tokens": 523648200.0,
      "step": 15630
    },
    {
      "epoch": 0.933321394460363,
      "grad_norm": 0.5059530735015869,
      "learning_rate": 3.181708371402295e-05,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8229034900665283,
      "num_tokens": 523815880.0,
      "step": 15635
    },
    {
      "epoch": 0.9336198662846227,
      "grad_norm": 0.5483002662658691,
      "learning_rate": 3.180618515690507e-05,
      "loss": 0.5117,
      "mean_token_accuracy": 0.8340451002120972,
      "num_tokens": 523983560.0,
      "step": 15640
    },
    {
      "epoch": 0.9339183381088825,
      "grad_norm": 0.6235918402671814,
      "learning_rate": 3.17952855509363e-05,
      "loss": 0.5436,
      "mean_token_accuracy": 0.8260825395584106,
      "num_tokens": 524151240.0,
      "step": 15645
    },
    {
      "epoch": 0.9342168099331423,
      "grad_norm": 0.4676841199398041,
      "learning_rate": 3.178438489877145e-05,
      "loss": 0.5104,
      "mean_token_accuracy": 0.8357926726341247,
      "num_tokens": 524318920.0,
      "step": 15650
    },
    {
      "epoch": 0.9345152817574021,
      "grad_norm": 0.5368183851242065,
      "learning_rate": 3.177348320306558e-05,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8243767142295837,
      "num_tokens": 524486600.0,
      "step": 15655
    },
    {
      "epoch": 0.9348137535816619,
      "grad_norm": 0.4797813594341278,
      "learning_rate": 3.1762580466474005e-05,
      "loss": 0.4834,
      "mean_token_accuracy": 0.843295955657959,
      "num_tokens": 524654280.0,
      "step": 15660
    },
    {
      "epoch": 0.9351122254059216,
      "grad_norm": 0.4770227372646332,
      "learning_rate": 3.175167669165228e-05,
      "loss": 0.501,
      "mean_token_accuracy": 0.8369855761528016,
      "num_tokens": 524821960.0,
      "step": 15665
    },
    {
      "epoch": 0.9354106972301814,
      "grad_norm": 0.49592265486717224,
      "learning_rate": 3.1740771881256216e-05,
      "loss": 0.5539,
      "mean_token_accuracy": 0.8225515842437744,
      "num_tokens": 524989640.0,
      "step": 15670
    },
    {
      "epoch": 0.9357091690544412,
      "grad_norm": 0.527769923210144,
      "learning_rate": 3.172986603794188e-05,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8192532420158386,
      "num_tokens": 525157320.0,
      "step": 15675
    },
    {
      "epoch": 0.936007640878701,
      "grad_norm": 0.6402641534805298,
      "learning_rate": 3.17189591643656e-05,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8278182029724122,
      "num_tokens": 525325000.0,
      "step": 15680
    },
    {
      "epoch": 0.9363061127029608,
      "grad_norm": 0.48500001430511475,
      "learning_rate": 3.170805126318393e-05,
      "loss": 0.515,
      "mean_token_accuracy": 0.8350292205810547,
      "num_tokens": 525492680.0,
      "step": 15685
    },
    {
      "epoch": 0.9366045845272206,
      "grad_norm": 0.5422039031982422,
      "learning_rate": 3.169714233705372e-05,
      "loss": 0.5078,
      "mean_token_accuracy": 0.8368125915527344,
      "num_tokens": 525660360.0,
      "step": 15690
    },
    {
      "epoch": 0.9369030563514804,
      "grad_norm": 0.48064228892326355,
      "learning_rate": 3.1686232388632014e-05,
      "loss": 0.5382,
      "mean_token_accuracy": 0.8278838276863099,
      "num_tokens": 525828040.0,
      "step": 15695
    },
    {
      "epoch": 0.9372015281757402,
      "grad_norm": 0.46511518955230713,
      "learning_rate": 3.167532142057615e-05,
      "loss": 0.541,
      "mean_token_accuracy": 0.8270487785339355,
      "num_tokens": 525995720.0,
      "step": 15700
    },
    {
      "epoch": 0.9375,
      "grad_norm": 0.502470850944519,
      "learning_rate": 3.166440943554368e-05,
      "loss": 0.4972,
      "mean_token_accuracy": 0.8385422945022583,
      "num_tokens": 526163400.0,
      "step": 15705
    },
    {
      "epoch": 0.9377984718242598,
      "grad_norm": 0.5148174166679382,
      "learning_rate": 3.1653496436192423e-05,
      "loss": 0.4927,
      "mean_token_accuracy": 0.840689480304718,
      "num_tokens": 526331080.0,
      "step": 15710
    },
    {
      "epoch": 0.9380969436485196,
      "grad_norm": 0.5600574016571045,
      "learning_rate": 3.1642582425180444e-05,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8386198163032532,
      "num_tokens": 526498760.0,
      "step": 15715
    },
    {
      "epoch": 0.9383954154727794,
      "grad_norm": 0.520237147808075,
      "learning_rate": 3.163166740516607e-05,
      "loss": 0.5432,
      "mean_token_accuracy": 0.8268698573112487,
      "num_tokens": 526666440.0,
      "step": 15720
    },
    {
      "epoch": 0.9386938872970392,
      "grad_norm": 0.45518752932548523,
      "learning_rate": 3.162075137880782e-05,
      "loss": 0.5101,
      "mean_token_accuracy": 0.8360849261283875,
      "num_tokens": 526834120.0,
      "step": 15725
    },
    {
      "epoch": 0.938992359121299,
      "grad_norm": 0.5561493039131165,
      "learning_rate": 3.160983434876452e-05,
      "loss": 0.5609,
      "mean_token_accuracy": 0.8224203705787658,
      "num_tokens": 527001800.0,
      "step": 15730
    },
    {
      "epoch": 0.9392908309455588,
      "grad_norm": 0.4951010048389435,
      "learning_rate": 3.1598916317695194e-05,
      "loss": 0.5134,
      "mean_token_accuracy": 0.8320410370826721,
      "num_tokens": 527169480.0,
      "step": 15735
    },
    {
      "epoch": 0.9395893027698186,
      "grad_norm": 0.5409601330757141,
      "learning_rate": 3.1587997288259144e-05,
      "loss": 0.5259,
      "mean_token_accuracy": 0.8323511958122254,
      "num_tokens": 527337160.0,
      "step": 15740
    },
    {
      "epoch": 0.9398877745940784,
      "grad_norm": 0.5131236910820007,
      "learning_rate": 3.157707726311591e-05,
      "loss": 0.4859,
      "mean_token_accuracy": 0.8410175323486329,
      "num_tokens": 527504840.0,
      "step": 15745
    },
    {
      "epoch": 0.9401862464183381,
      "grad_norm": 0.5314719080924988,
      "learning_rate": 3.156615624492524e-05,
      "loss": 0.5352,
      "mean_token_accuracy": 0.8283251762390137,
      "num_tokens": 527672520.0,
      "step": 15750
    },
    {
      "epoch": 0.9404847182425979,
      "grad_norm": 0.5749266743659973,
      "learning_rate": 3.1555234236347175e-05,
      "loss": 0.567,
      "mean_token_accuracy": 0.8178217887878418,
      "num_tokens": 527840200.0,
      "step": 15755
    },
    {
      "epoch": 0.9407831900668577,
      "grad_norm": 0.5439934134483337,
      "learning_rate": 3.154431124004196e-05,
      "loss": 0.5143,
      "mean_token_accuracy": 0.83399738073349,
      "num_tokens": 528007880.0,
      "step": 15760
    },
    {
      "epoch": 0.9410816618911175,
      "grad_norm": 0.5169892907142639,
      "learning_rate": 3.1533387258670076e-05,
      "loss": 0.5927,
      "mean_token_accuracy": 0.8124776363372803,
      "num_tokens": 528175560.0,
      "step": 15765
    },
    {
      "epoch": 0.9413801337153773,
      "grad_norm": 0.5228572487831116,
      "learning_rate": 3.152246229489229e-05,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8227364897727967,
      "num_tokens": 528343240.0,
      "step": 15770
    },
    {
      "epoch": 0.941678605539637,
      "grad_norm": 0.4965108036994934,
      "learning_rate": 3.151153635136958e-05,
      "loss": 0.4895,
      "mean_token_accuracy": 0.8412859320640564,
      "num_tokens": 528510920.0,
      "step": 15775
    },
    {
      "epoch": 0.9419770773638968,
      "grad_norm": 0.517785370349884,
      "learning_rate": 3.150060943076313e-05,
      "loss": 0.5819,
      "mean_token_accuracy": 0.8136526346206665,
      "num_tokens": 528678600.0,
      "step": 15780
    },
    {
      "epoch": 0.9422755491881566,
      "grad_norm": 0.47368016839027405,
      "learning_rate": 3.148968153573441e-05,
      "loss": 0.5215,
      "mean_token_accuracy": 0.8339496493339539,
      "num_tokens": 528846280.0,
      "step": 15785
    },
    {
      "epoch": 0.9425740210124164,
      "grad_norm": 0.514857828617096,
      "learning_rate": 3.147875266894511e-05,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8382378578186035,
      "num_tokens": 529012034.0,
      "step": 15790
    },
    {
      "epoch": 0.9428724928366762,
      "grad_norm": 0.4875916838645935,
      "learning_rate": 3.146782283305717e-05,
      "loss": 0.499,
      "mean_token_accuracy": 0.8389896273612976,
      "num_tokens": 529179714.0,
      "step": 15795
    },
    {
      "epoch": 0.943170964660936,
      "grad_norm": 0.49777963757514954,
      "learning_rate": 3.145689203073273e-05,
      "loss": 0.5641,
      "mean_token_accuracy": 0.8193785071372985,
      "num_tokens": 529347394.0,
      "step": 15800
    },
    {
      "epoch": 0.9434694364851958,
      "grad_norm": 0.5428093075752258,
      "learning_rate": 3.144596026463421e-05,
      "loss": 0.476,
      "mean_token_accuracy": 0.8452940583229065,
      "num_tokens": 529515074.0,
      "step": 15805
    },
    {
      "epoch": 0.9437679083094556,
      "grad_norm": 0.49488624930381775,
      "learning_rate": 3.1435027537424225e-05,
      "loss": 0.523,
      "mean_token_accuracy": 0.8318382382392884,
      "num_tokens": 529682754.0,
      "step": 15810
    },
    {
      "epoch": 0.9440663801337154,
      "grad_norm": 0.5103921294212341,
      "learning_rate": 3.1424093851765656e-05,
      "loss": 0.5116,
      "mean_token_accuracy": 0.8359835386276245,
      "num_tokens": 529850434.0,
      "step": 15815
    },
    {
      "epoch": 0.9443648519579751,
      "grad_norm": 0.5930900573730469,
      "learning_rate": 3.141315921032159e-05,
      "loss": 0.5744,
      "mean_token_accuracy": 0.8167563557624817,
      "num_tokens": 530014952.0,
      "step": 15820
    },
    {
      "epoch": 0.9446633237822349,
      "grad_norm": 0.5441305041313171,
      "learning_rate": 3.140222361575539e-05,
      "loss": 0.5182,
      "mean_token_accuracy": 0.8344387412071228,
      "num_tokens": 530182632.0,
      "step": 15825
    },
    {
      "epoch": 0.9449617956064947,
      "grad_norm": 0.5473672747612,
      "learning_rate": 3.139128707073058e-05,
      "loss": 0.5651,
      "mean_token_accuracy": 0.8199809193611145,
      "num_tokens": 530350312.0,
      "step": 15830
    },
    {
      "epoch": 0.9452602674307545,
      "grad_norm": 0.5072567462921143,
      "learning_rate": 3.138034957791099e-05,
      "loss": 0.5122,
      "mean_token_accuracy": 0.8345759153366089,
      "num_tokens": 530517992.0,
      "step": 15835
    },
    {
      "epoch": 0.9455587392550143,
      "grad_norm": 0.5286093950271606,
      "learning_rate": 3.1369411139960636e-05,
      "loss": 0.5592,
      "mean_token_accuracy": 0.8219193577766418,
      "num_tokens": 530685672.0,
      "step": 15840
    },
    {
      "epoch": 0.9458572110792741,
      "grad_norm": 0.529873788356781,
      "learning_rate": 3.135847175954378e-05,
      "loss": 0.5487,
      "mean_token_accuracy": 0.8244184613227844,
      "num_tokens": 530853352.0,
      "step": 15845
    },
    {
      "epoch": 0.9461556829035339,
      "grad_norm": 0.5681453943252563,
      "learning_rate": 3.1347531439324896e-05,
      "loss": 0.5504,
      "mean_token_accuracy": 0.8237504601478577,
      "num_tokens": 531021032.0,
      "step": 15850
    },
    {
      "epoch": 0.9464541547277937,
      "grad_norm": 0.8035094738006592,
      "learning_rate": 3.1336590181968725e-05,
      "loss": 0.5877,
      "mean_token_accuracy": 0.8152451515197754,
      "num_tokens": 531188712.0,
      "step": 15855
    },
    {
      "epoch": 0.9467526265520535,
      "grad_norm": 0.5231764912605286,
      "learning_rate": 3.132564799014019e-05,
      "loss": 0.5593,
      "mean_token_accuracy": 0.8210604786872864,
      "num_tokens": 531356392.0,
      "step": 15860
    },
    {
      "epoch": 0.9470510983763133,
      "grad_norm": 0.5790037512779236,
      "learning_rate": 3.1314704866504484e-05,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8264821648597718,
      "num_tokens": 531524072.0,
      "step": 15865
    },
    {
      "epoch": 0.9473495702005731,
      "grad_norm": 0.4739135205745697,
      "learning_rate": 3.1303760813726996e-05,
      "loss": 0.5517,
      "mean_token_accuracy": 0.8243409276008606,
      "num_tokens": 531691752.0,
      "step": 15870
    },
    {
      "epoch": 0.9476480420248329,
      "grad_norm": 0.46261680126190186,
      "learning_rate": 3.129281583447335e-05,
      "loss": 0.5298,
      "mean_token_accuracy": 0.8301860928535462,
      "num_tokens": 531859432.0,
      "step": 15875
    },
    {
      "epoch": 0.9479465138490927,
      "grad_norm": 0.5272606611251831,
      "learning_rate": 3.1281869931409394e-05,
      "loss": 0.5429,
      "mean_token_accuracy": 0.8279673218727112,
      "num_tokens": 532027112.0,
      "step": 15880
    },
    {
      "epoch": 0.9482449856733525,
      "grad_norm": 0.4770931899547577,
      "learning_rate": 3.127092310720121e-05,
      "loss": 0.5055,
      "mean_token_accuracy": 0.8383931636810302,
      "num_tokens": 532194792.0,
      "step": 15885
    },
    {
      "epoch": 0.9485434574976123,
      "grad_norm": 0.48625510931015015,
      "learning_rate": 3.125997536451511e-05,
      "loss": 0.5557,
      "mean_token_accuracy": 0.8261541247367858,
      "num_tokens": 532362472.0,
      "step": 15890
    },
    {
      "epoch": 0.9488419293218721,
      "grad_norm": 0.4478358030319214,
      "learning_rate": 3.1249026706017604e-05,
      "loss": 0.4875,
      "mean_token_accuracy": 0.8413396120071411,
      "num_tokens": 532530152.0,
      "step": 15895
    },
    {
      "epoch": 0.9491404011461319,
      "grad_norm": 0.5235236883163452,
      "learning_rate": 3.123807713437544e-05,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8288023352622986,
      "num_tokens": 532697832.0,
      "step": 15900
    },
    {
      "epoch": 0.9494388729703916,
      "grad_norm": 0.4963972568511963,
      "learning_rate": 3.12271266522556e-05,
      "loss": 0.4765,
      "mean_token_accuracy": 0.844912326335907,
      "num_tokens": 532865512.0,
      "step": 15905
    },
    {
      "epoch": 0.9497373447946514,
      "grad_norm": 0.4907382130622864,
      "learning_rate": 3.121617526232527e-05,
      "loss": 0.5233,
      "mean_token_accuracy": 0.832381010055542,
      "num_tokens": 533033192.0,
      "step": 15910
    },
    {
      "epoch": 0.9500358166189111,
      "grad_norm": 0.5912783145904541,
      "learning_rate": 3.1205222967251857e-05,
      "loss": 0.5655,
      "mean_token_accuracy": 0.8202075719833374,
      "num_tokens": 533200872.0,
      "step": 15915
    },
    {
      "epoch": 0.9503342884431709,
      "grad_norm": 0.5628465414047241,
      "learning_rate": 3.1194269769703e-05,
      "loss": 0.5689,
      "mean_token_accuracy": 0.8188178420066834,
      "num_tokens": 533368552.0,
      "step": 15920
    },
    {
      "epoch": 0.9506327602674307,
      "grad_norm": 0.4964522123336792,
      "learning_rate": 3.1183315672346555e-05,
      "loss": 0.5222,
      "mean_token_accuracy": 0.8315101981163024,
      "num_tokens": 533536232.0,
      "step": 15925
    },
    {
      "epoch": 0.9509312320916905,
      "grad_norm": 0.530226469039917,
      "learning_rate": 3.117236067785059e-05,
      "loss": 0.5255,
      "mean_token_accuracy": 0.832369077205658,
      "num_tokens": 533703912.0,
      "step": 15930
    },
    {
      "epoch": 0.9512297039159503,
      "grad_norm": 0.45650309324264526,
      "learning_rate": 3.116140478888339e-05,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8332160353660584,
      "num_tokens": 533871592.0,
      "step": 15935
    },
    {
      "epoch": 0.9515281757402101,
      "grad_norm": 0.539772093296051,
      "learning_rate": 3.1150448008113495e-05,
      "loss": 0.4838,
      "mean_token_accuracy": 0.8441488742828369,
      "num_tokens": 534039272.0,
      "step": 15940
    },
    {
      "epoch": 0.9518266475644699,
      "grad_norm": 0.5195083618164062,
      "learning_rate": 3.1139490338209585e-05,
      "loss": 0.5189,
      "mean_token_accuracy": 0.8337707281112671,
      "num_tokens": 534206952.0,
      "step": 15945
    },
    {
      "epoch": 0.9521251193887297,
      "grad_norm": 0.4558088779449463,
      "learning_rate": 3.112853178184064e-05,
      "loss": 0.5096,
      "mean_token_accuracy": 0.8356018185615539,
      "num_tokens": 534374632.0,
      "step": 15950
    },
    {
      "epoch": 0.9524235912129895,
      "grad_norm": 0.554297149181366,
      "learning_rate": 3.1117572341675796e-05,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8226768493652343,
      "num_tokens": 534542312.0,
      "step": 15955
    },
    {
      "epoch": 0.9527220630372493,
      "grad_norm": 0.5143594145774841,
      "learning_rate": 3.110661202038443e-05,
      "loss": 0.5554,
      "mean_token_accuracy": 0.8240367412567139,
      "num_tokens": 534709992.0,
      "step": 15960
    },
    {
      "epoch": 0.9530205348615091,
      "grad_norm": 0.47093838453292847,
      "learning_rate": 3.1095650820636146e-05,
      "loss": 0.5316,
      "mean_token_accuracy": 0.8290767073631287,
      "num_tokens": 534877672.0,
      "step": 15965
    },
    {
      "epoch": 0.9533190066857689,
      "grad_norm": 0.4957123398780823,
      "learning_rate": 3.1084688745100736e-05,
      "loss": 0.5579,
      "mean_token_accuracy": 0.8235417008399963,
      "num_tokens": 535045352.0,
      "step": 15970
    },
    {
      "epoch": 0.9536174785100286,
      "grad_norm": 0.49072322249412537,
      "learning_rate": 3.1073725796448216e-05,
      "loss": 0.4749,
      "mean_token_accuracy": 0.844286072254181,
      "num_tokens": 535213032.0,
      "step": 15975
    },
    {
      "epoch": 0.9539159503342884,
      "grad_norm": 0.5057286620140076,
      "learning_rate": 3.1062761977348814e-05,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8388583898544312,
      "num_tokens": 535380712.0,
      "step": 15980
    },
    {
      "epoch": 0.9542144221585482,
      "grad_norm": 0.5004511475563049,
      "learning_rate": 3.1051797290472966e-05,
      "loss": 0.5229,
      "mean_token_accuracy": 0.8323452115058899,
      "num_tokens": 535548392.0,
      "step": 15985
    },
    {
      "epoch": 0.954512893982808,
      "grad_norm": 0.5344756841659546,
      "learning_rate": 3.104083173849134e-05,
      "loss": 0.5385,
      "mean_token_accuracy": 0.8270368576049805,
      "num_tokens": 535716072.0,
      "step": 15990
    },
    {
      "epoch": 0.9548113658070678,
      "grad_norm": 0.4614121913909912,
      "learning_rate": 3.102986532407478e-05,
      "loss": 0.493,
      "mean_token_accuracy": 0.8404210805892944,
      "num_tokens": 535883752.0,
      "step": 15995
    },
    {
      "epoch": 0.9551098376313276,
      "grad_norm": 0.4517372250556946,
      "learning_rate": 3.101889804989437e-05,
      "loss": 0.5317,
      "mean_token_accuracy": 0.8297268271446228,
      "num_tokens": 536051432.0,
      "step": 16000
    },
    {
      "epoch": 0.9554083094555874,
      "grad_norm": 0.655108630657196,
      "learning_rate": 3.1007929918621385e-05,
      "loss": 0.525,
      "mean_token_accuracy": 0.832518196105957,
      "num_tokens": 536219112.0,
      "step": 16005
    },
    {
      "epoch": 0.9557067812798472,
      "grad_norm": 0.5116455554962158,
      "learning_rate": 3.099696093292732e-05,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8291124939918518,
      "num_tokens": 536386792.0,
      "step": 16010
    },
    {
      "epoch": 0.956005253104107,
      "grad_norm": 0.5170709490776062,
      "learning_rate": 3.0985991095483876e-05,
      "loss": 0.5723,
      "mean_token_accuracy": 0.8196230530738831,
      "num_tokens": 536554472.0,
      "step": 16015
    },
    {
      "epoch": 0.9563037249283668,
      "grad_norm": 0.4800170660018921,
      "learning_rate": 3.097502040896296e-05,
      "loss": 0.5137,
      "mean_token_accuracy": 0.834235954284668,
      "num_tokens": 536722152.0,
      "step": 16020
    },
    {
      "epoch": 0.9566021967526266,
      "grad_norm": 0.49619466066360474,
      "learning_rate": 3.096404887603667e-05,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8323869705200195,
      "num_tokens": 536889832.0,
      "step": 16025
    },
    {
      "epoch": 0.9569006685768864,
      "grad_norm": 0.5148215889930725,
      "learning_rate": 3.095307649937735e-05,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8387093067169189,
      "num_tokens": 537057512.0,
      "step": 16030
    },
    {
      "epoch": 0.9571991404011462,
      "grad_norm": 0.5083907246589661,
      "learning_rate": 3.09421032816575e-05,
      "loss": 0.5173,
      "mean_token_accuracy": 0.8319217443466187,
      "num_tokens": 537225192.0,
      "step": 16035
    },
    {
      "epoch": 0.957497612225406,
      "grad_norm": 0.6620160341262817,
      "learning_rate": 3.093112922554986e-05,
      "loss": 0.562,
      "mean_token_accuracy": 0.8210187196731568,
      "num_tokens": 537392872.0,
      "step": 16040
    },
    {
      "epoch": 0.9577960840496658,
      "grad_norm": 0.480284184217453,
      "learning_rate": 3.0920154333727386e-05,
      "loss": 0.5597,
      "mean_token_accuracy": 0.8216748356819152,
      "num_tokens": 537560552.0,
      "step": 16045
    },
    {
      "epoch": 0.9580945558739254,
      "grad_norm": 0.4943378269672394,
      "learning_rate": 3.0909178608863185e-05,
      "loss": 0.5255,
      "mean_token_accuracy": 0.831617558002472,
      "num_tokens": 537728232.0,
      "step": 16050
    },
    {
      "epoch": 0.9583930276981852,
      "grad_norm": 0.4783206582069397,
      "learning_rate": 3.08982020536306e-05,
      "loss": 0.5272,
      "mean_token_accuracy": 0.831402850151062,
      "num_tokens": 537895912.0,
      "step": 16055
    },
    {
      "epoch": 0.958691499522445,
      "grad_norm": 0.5013982653617859,
      "learning_rate": 3.088722467070319e-05,
      "loss": 0.4841,
      "mean_token_accuracy": 0.8433138489723205,
      "num_tokens": 538063592.0,
      "step": 16060
    },
    {
      "epoch": 0.9589899713467048,
      "grad_norm": 0.5668599009513855,
      "learning_rate": 3.0876246462754685e-05,
      "loss": 0.5421,
      "mean_token_accuracy": 0.826177966594696,
      "num_tokens": 538231272.0,
      "step": 16065
    },
    {
      "epoch": 0.9592884431709646,
      "grad_norm": 0.4846465289592743,
      "learning_rate": 3.086526743245904e-05,
      "loss": 0.5476,
      "mean_token_accuracy": 0.8265358448028565,
      "num_tokens": 538398952.0,
      "step": 16070
    },
    {
      "epoch": 0.9595869149952244,
      "grad_norm": 0.5607597231864929,
      "learning_rate": 3.08542875824904e-05,
      "loss": 0.5236,
      "mean_token_accuracy": 0.8326980590820312,
      "num_tokens": 538561362.0,
      "step": 16075
    },
    {
      "epoch": 0.9598853868194842,
      "grad_norm": 0.504673182964325,
      "learning_rate": 3.0843306915523096e-05,
      "loss": 0.5095,
      "mean_token_accuracy": 0.8356793522834778,
      "num_tokens": 538729042.0,
      "step": 16080
    },
    {
      "epoch": 0.960183858643744,
      "grad_norm": 0.5429574847221375,
      "learning_rate": 3.083232543423169e-05,
      "loss": 0.5394,
      "mean_token_accuracy": 0.8269891262054443,
      "num_tokens": 538896722.0,
      "step": 16085
    },
    {
      "epoch": 0.9604823304680038,
      "grad_norm": 0.5217997431755066,
      "learning_rate": 3.082134314129091e-05,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8325420618057251,
      "num_tokens": 539064402.0,
      "step": 16090
    },
    {
      "epoch": 0.9607808022922636,
      "grad_norm": 0.5468534231185913,
      "learning_rate": 3.0810360039375705e-05,
      "loss": 0.5597,
      "mean_token_accuracy": 0.8216807842254639,
      "num_tokens": 539232082.0,
      "step": 16095
    },
    {
      "epoch": 0.9610792741165234,
      "grad_norm": 0.48495370149612427,
      "learning_rate": 3.079937613116121e-05,
      "loss": 0.5468,
      "mean_token_accuracy": 0.8240844488143921,
      "num_tokens": 539399762.0,
      "step": 16100
    },
    {
      "epoch": 0.9613777459407832,
      "grad_norm": 0.4977683424949646,
      "learning_rate": 3.0788391419322766e-05,
      "loss": 0.517,
      "mean_token_accuracy": 0.8354228734970093,
      "num_tokens": 539567442.0,
      "step": 16105
    },
    {
      "epoch": 0.961676217765043,
      "grad_norm": 0.5285274386405945,
      "learning_rate": 3.077740590653588e-05,
      "loss": 0.543,
      "mean_token_accuracy": 0.8270607113838195,
      "num_tokens": 539735122.0,
      "step": 16110
    },
    {
      "epoch": 0.9619746895893028,
      "grad_norm": 0.5071113705635071,
      "learning_rate": 3.076641959547631e-05,
      "loss": 0.5349,
      "mean_token_accuracy": 0.828659188747406,
      "num_tokens": 539902802.0,
      "step": 16115
    },
    {
      "epoch": 0.9622731614135626,
      "grad_norm": 0.4986821115016937,
      "learning_rate": 3.075543248881995e-05,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8289574265480042,
      "num_tokens": 540070482.0,
      "step": 16120
    },
    {
      "epoch": 0.9625716332378224,
      "grad_norm": 0.5541632771492004,
      "learning_rate": 3.074444458924291e-05,
      "loss": 0.5094,
      "mean_token_accuracy": 0.83626229763031,
      "num_tokens": 540236251.0,
      "step": 16125
    },
    {
      "epoch": 0.9628701050620821,
      "grad_norm": 0.5603629350662231,
      "learning_rate": 3.073345589942152e-05,
      "loss": 0.5437,
      "mean_token_accuracy": 0.8257306456565857,
      "num_tokens": 540403931.0,
      "step": 16130
    },
    {
      "epoch": 0.9631685768863419,
      "grad_norm": 0.46965834498405457,
      "learning_rate": 3.0722466422032245e-05,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8326461434364318,
      "num_tokens": 540571254.0,
      "step": 16135
    },
    {
      "epoch": 0.9634670487106017,
      "grad_norm": 0.49990078806877136,
      "learning_rate": 3.07114761597518e-05,
      "loss": 0.5531,
      "mean_token_accuracy": 0.8224203824996948,
      "num_tokens": 540738934.0,
      "step": 16140
    },
    {
      "epoch": 0.9637655205348615,
      "grad_norm": 0.4671829044818878,
      "learning_rate": 3.070048511525707e-05,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8352439522743225,
      "num_tokens": 540906614.0,
      "step": 16145
    },
    {
      "epoch": 0.9640639923591213,
      "grad_norm": 0.5460081100463867,
      "learning_rate": 3.06894932912251e-05,
      "loss": 0.542,
      "mean_token_accuracy": 0.8288560152053833,
      "num_tokens": 541074294.0,
      "step": 16150
    },
    {
      "epoch": 0.9643624641833811,
      "grad_norm": 0.4894111752510071,
      "learning_rate": 3.067850069033318e-05,
      "loss": 0.5317,
      "mean_token_accuracy": 0.8283430695533752,
      "num_tokens": 541241974.0,
      "step": 16155
    },
    {
      "epoch": 0.9646609360076409,
      "grad_norm": 0.47574856877326965,
      "learning_rate": 3.0667507315258734e-05,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8375402450561523,
      "num_tokens": 541409654.0,
      "step": 16160
    },
    {
      "epoch": 0.9649594078319007,
      "grad_norm": 0.49619680643081665,
      "learning_rate": 3.065651316867944e-05,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8350590586662292,
      "num_tokens": 541577334.0,
      "step": 16165
    },
    {
      "epoch": 0.9652578796561605,
      "grad_norm": 0.8324124217033386,
      "learning_rate": 3.0645518253273076e-05,
      "loss": 0.5059,
      "mean_token_accuracy": 0.836967658996582,
      "num_tokens": 541745014.0,
      "step": 16170
    },
    {
      "epoch": 0.9655563514804203,
      "grad_norm": 0.4928455948829651,
      "learning_rate": 3.06345225717177e-05,
      "loss": 0.5547,
      "mean_token_accuracy": 0.8223602414131165,
      "num_tokens": 541906331.0,
      "step": 16175
    },
    {
      "epoch": 0.9658548233046801,
      "grad_norm": 0.5277348160743713,
      "learning_rate": 3.0623526126691484e-05,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8410413861274719,
      "num_tokens": 542074011.0,
      "step": 16180
    },
    {
      "epoch": 0.9661532951289399,
      "grad_norm": 0.4890357553958893,
      "learning_rate": 3.061252892087282e-05,
      "loss": 0.5479,
      "mean_token_accuracy": 0.8254324197769165,
      "num_tokens": 542241691.0,
      "step": 16185
    },
    {
      "epoch": 0.9664517669531996,
      "grad_norm": 0.4640437364578247,
      "learning_rate": 3.06015309569403e-05,
      "loss": 0.5064,
      "mean_token_accuracy": 0.8366992831230163,
      "num_tokens": 542409371.0,
      "step": 16190
    },
    {
      "epoch": 0.9667502387774594,
      "grad_norm": 0.4964997172355652,
      "learning_rate": 3.059053223757267e-05,
      "loss": 0.5496,
      "mean_token_accuracy": 0.8235059022903443,
      "num_tokens": 542577051.0,
      "step": 16195
    },
    {
      "epoch": 0.9670487106017192,
      "grad_norm": 0.5225151777267456,
      "learning_rate": 3.0579532765448857e-05,
      "loss": 0.5344,
      "mean_token_accuracy": 0.8288202285766602,
      "num_tokens": 542744731.0,
      "step": 16200
    },
    {
      "epoch": 0.967347182425979,
      "grad_norm": 0.49010005593299866,
      "learning_rate": 3.0568532543248e-05,
      "loss": 0.5384,
      "mean_token_accuracy": 0.827156150341034,
      "num_tokens": 542912411.0,
      "step": 16205
    },
    {
      "epoch": 0.9676456542502387,
      "grad_norm": 0.5213024616241455,
      "learning_rate": 3.055753157364941e-05,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8323511958122254,
      "num_tokens": 543080091.0,
      "step": 16210
    },
    {
      "epoch": 0.9679441260744985,
      "grad_norm": 0.5314414501190186,
      "learning_rate": 3.0546529859332566e-05,
      "loss": 0.5607,
      "mean_token_accuracy": 0.8219611167907714,
      "num_tokens": 543247771.0,
      "step": 16215
    },
    {
      "epoch": 0.9682425978987583,
      "grad_norm": 0.5320379734039307,
      "learning_rate": 3.0535527402977144e-05,
      "loss": 0.5087,
      "mean_token_accuracy": 0.8350888729095459,
      "num_tokens": 543415451.0,
      "step": 16220
    },
    {
      "epoch": 0.9685410697230181,
      "grad_norm": 0.5203273296356201,
      "learning_rate": 3.0524524207262994e-05,
      "loss": 0.5306,
      "mean_token_accuracy": 0.8277168035507202,
      "num_tokens": 543583131.0,
      "step": 16225
    },
    {
      "epoch": 0.9688395415472779,
      "grad_norm": 0.5501710772514343,
      "learning_rate": 3.0513520274870143e-05,
      "loss": 0.5195,
      "mean_token_accuracy": 0.8340152621269226,
      "num_tokens": 543750811.0,
      "step": 16230
    },
    {
      "epoch": 0.9691380133715377,
      "grad_norm": 0.5646803379058838,
      "learning_rate": 3.0502515608478805e-05,
      "loss": 0.5396,
      "mean_token_accuracy": 0.8267147779464722,
      "num_tokens": 543918491.0,
      "step": 16235
    },
    {
      "epoch": 0.9694364851957975,
      "grad_norm": 0.6729055643081665,
      "learning_rate": 3.0491510210769375e-05,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8442264080047608,
      "num_tokens": 544086171.0,
      "step": 16240
    },
    {
      "epoch": 0.9697349570200573,
      "grad_norm": 0.5303055644035339,
      "learning_rate": 3.0480504084422413e-05,
      "loss": 0.483,
      "mean_token_accuracy": 0.8435882091522217,
      "num_tokens": 544253851.0,
      "step": 16245
    },
    {
      "epoch": 0.9700334288443171,
      "grad_norm": 0.4985578656196594,
      "learning_rate": 3.046949723211867e-05,
      "loss": 0.558,
      "mean_token_accuracy": 0.8207741856575013,
      "num_tokens": 544421531.0,
      "step": 16250
    },
    {
      "epoch": 0.9703319006685769,
      "grad_norm": 0.5357292294502258,
      "learning_rate": 3.0458489656539062e-05,
      "loss": 0.5275,
      "mean_token_accuracy": 0.8285041093826294,
      "num_tokens": 544589211.0,
      "step": 16255
    },
    {
      "epoch": 0.9706303724928367,
      "grad_norm": 0.8215190768241882,
      "learning_rate": 3.0447481360364693e-05,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8250268459320068,
      "num_tokens": 544756891.0,
      "step": 16260
    },
    {
      "epoch": 0.9709288443170965,
      "grad_norm": 0.5300891995429993,
      "learning_rate": 3.0436472346276833e-05,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8356316447257995,
      "num_tokens": 544924571.0,
      "step": 16265
    },
    {
      "epoch": 0.9712273161413563,
      "grad_norm": 0.4493369162082672,
      "learning_rate": 3.042546261695694e-05,
      "loss": 0.4879,
      "mean_token_accuracy": 0.8421567440032959,
      "num_tokens": 545092251.0,
      "step": 16270
    },
    {
      "epoch": 0.9715257879656161,
      "grad_norm": 0.5489478707313538,
      "learning_rate": 3.0414452175086623e-05,
      "loss": 0.556,
      "mean_token_accuracy": 0.8219193696975708,
      "num_tokens": 545259931.0,
      "step": 16275
    },
    {
      "epoch": 0.9718242597898759,
      "grad_norm": 0.5192468762397766,
      "learning_rate": 3.0403441023347677e-05,
      "loss": 0.5085,
      "mean_token_accuracy": 0.835589873790741,
      "num_tokens": 545427611.0,
      "step": 16280
    },
    {
      "epoch": 0.9721227316141356,
      "grad_norm": 0.48073679208755493,
      "learning_rate": 3.039242916442209e-05,
      "loss": 0.4838,
      "mean_token_accuracy": 0.8418764114379883,
      "num_tokens": 545595291.0,
      "step": 16285
    },
    {
      "epoch": 0.9724212034383954,
      "grad_norm": 0.512175440788269,
      "learning_rate": 3.0381416600991984e-05,
      "loss": 0.5298,
      "mean_token_accuracy": 0.8305737733840942,
      "num_tokens": 545762971.0,
      "step": 16290
    },
    {
      "epoch": 0.9727196752626552,
      "grad_norm": 0.5279102325439453,
      "learning_rate": 3.0370403335739677e-05,
      "loss": 0.4732,
      "mean_token_accuracy": 0.8479243636131286,
      "num_tokens": 545930651.0,
      "step": 16295
    },
    {
      "epoch": 0.973018147086915,
      "grad_norm": 0.4833485186100006,
      "learning_rate": 3.035938937134767e-05,
      "loss": 0.5479,
      "mean_token_accuracy": 0.824311101436615,
      "num_tokens": 546098331.0,
      "step": 16300
    },
    {
      "epoch": 0.9733166189111748,
      "grad_norm": 0.4734708070755005,
      "learning_rate": 3.0348374710498596e-05,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8419002652168274,
      "num_tokens": 546266011.0,
      "step": 16305
    },
    {
      "epoch": 0.9736150907354346,
      "grad_norm": 0.5586357116699219,
      "learning_rate": 3.0337359355875282e-05,
      "loss": 0.5214,
      "mean_token_accuracy": 0.8335301876068115,
      "num_tokens": 546427743.0,
      "step": 16310
    },
    {
      "epoch": 0.9739135625596944,
      "grad_norm": 0.5801336765289307,
      "learning_rate": 3.0326343310160726e-05,
      "loss": 0.5105,
      "mean_token_accuracy": 0.8357032060623169,
      "num_tokens": 546595423.0,
      "step": 16315
    },
    {
      "epoch": 0.9742120343839542,
      "grad_norm": 0.5442116260528564,
      "learning_rate": 3.0315326576038094e-05,
      "loss": 0.5267,
      "mean_token_accuracy": 0.8301503181457519,
      "num_tokens": 546763103.0,
      "step": 16320
    },
    {
      "epoch": 0.9745105062082139,
      "grad_norm": 0.5060010552406311,
      "learning_rate": 3.0304309156190702e-05,
      "loss": 0.4704,
      "mean_token_accuracy": 0.8482882022857666,
      "num_tokens": 546930783.0,
      "step": 16325
    },
    {
      "epoch": 0.9748089780324737,
      "grad_norm": 0.4762267768383026,
      "learning_rate": 3.0293291053302064e-05,
      "loss": 0.4872,
      "mean_token_accuracy": 0.8422223567962647,
      "num_tokens": 547098463.0,
      "step": 16330
    },
    {
      "epoch": 0.9751074498567335,
      "grad_norm": 0.5427188277244568,
      "learning_rate": 3.028227227005582e-05,
      "loss": 0.5458,
      "mean_token_accuracy": 0.8237921953201294,
      "num_tokens": 547266143.0,
      "step": 16335
    },
    {
      "epoch": 0.9754059216809933,
      "grad_norm": 0.5529686808586121,
      "learning_rate": 3.027125280913582e-05,
      "loss": 0.5598,
      "mean_token_accuracy": 0.8193665862083435,
      "num_tokens": 547433823.0,
      "step": 16340
    },
    {
      "epoch": 0.9757043935052531,
      "grad_norm": 0.5190616846084595,
      "learning_rate": 3.026023267322604e-05,
      "loss": 0.5099,
      "mean_token_accuracy": 0.8367052435874939,
      "num_tokens": 547601503.0,
      "step": 16345
    },
    {
      "epoch": 0.9760028653295129,
      "grad_norm": 0.5212611556053162,
      "learning_rate": 3.0249211865010647e-05,
      "loss": 0.5526,
      "mean_token_accuracy": 0.8234377026557922,
      "num_tokens": 547764372.0,
      "step": 16350
    },
    {
      "epoch": 0.9763013371537727,
      "grad_norm": 0.5687365531921387,
      "learning_rate": 3.023819038717396e-05,
      "loss": 0.5486,
      "mean_token_accuracy": 0.8242097139358521,
      "num_tokens": 547932052.0,
      "step": 16355
    },
    {
      "epoch": 0.9765998089780324,
      "grad_norm": 0.5005989670753479,
      "learning_rate": 3.022716824240046e-05,
      "loss": 0.5555,
      "mean_token_accuracy": 0.8228519797325134,
      "num_tokens": 548096057.0,
      "step": 16360
    },
    {
      "epoch": 0.9768982808022922,
      "grad_norm": 0.5083593130111694,
      "learning_rate": 3.0216145433374806e-05,
      "loss": 0.4938,
      "mean_token_accuracy": 0.8415125846862793,
      "num_tokens": 548263737.0,
      "step": 16365
    },
    {
      "epoch": 0.977196752626552,
      "grad_norm": 0.45629918575286865,
      "learning_rate": 3.020512196278178e-05,
      "loss": 0.5143,
      "mean_token_accuracy": 0.8360491394996643,
      "num_tokens": 548431417.0,
      "step": 16370
    },
    {
      "epoch": 0.9774952244508118,
      "grad_norm": 0.5211593508720398,
      "learning_rate": 3.0194097833306383e-05,
      "loss": 0.5277,
      "mean_token_accuracy": 0.830239760875702,
      "num_tokens": 548599097.0,
      "step": 16375
    },
    {
      "epoch": 0.9777936962750716,
      "grad_norm": 0.4815974831581116,
      "learning_rate": 3.018307304763373e-05,
      "loss": 0.5232,
      "mean_token_accuracy": 0.8328104376792907,
      "num_tokens": 548766777.0,
      "step": 16380
    },
    {
      "epoch": 0.9780921680993314,
      "grad_norm": 0.49401426315307617,
      "learning_rate": 3.017204760844911e-05,
      "loss": 0.5277,
      "mean_token_accuracy": 0.8308302521705627,
      "num_tokens": 548934457.0,
      "step": 16385
    },
    {
      "epoch": 0.9783906399235912,
      "grad_norm": 0.5188146829605103,
      "learning_rate": 3.0161021518437976e-05,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8375521898269653,
      "num_tokens": 549102137.0,
      "step": 16390
    },
    {
      "epoch": 0.978689111747851,
      "grad_norm": 0.5481185913085938,
      "learning_rate": 3.0149994780285932e-05,
      "loss": 0.5106,
      "mean_token_accuracy": 0.8341226220130921,
      "num_tokens": 549269817.0,
      "step": 16395
    },
    {
      "epoch": 0.9789875835721108,
      "grad_norm": 0.47941240668296814,
      "learning_rate": 3.013896739667874e-05,
      "loss": 0.5481,
      "mean_token_accuracy": 0.8241202473640442,
      "num_tokens": 549437497.0,
      "step": 16400
    },
    {
      "epoch": 0.9792860553963706,
      "grad_norm": 0.497881144285202,
      "learning_rate": 3.0127939370302344e-05,
      "loss": 0.5293,
      "mean_token_accuracy": 0.8290946006774902,
      "num_tokens": 549605177.0,
      "step": 16405
    },
    {
      "epoch": 0.9795845272206304,
      "grad_norm": 0.4580265283584595,
      "learning_rate": 3.0116910703842794e-05,
      "loss": 0.4825,
      "mean_token_accuracy": 0.8440057158470153,
      "num_tokens": 549772857.0,
      "step": 16410
    },
    {
      "epoch": 0.9798829990448902,
      "grad_norm": 0.5197835564613342,
      "learning_rate": 3.0105881399986353e-05,
      "loss": 0.5181,
      "mean_token_accuracy": 0.8343790888786315,
      "num_tokens": 549940537.0,
      "step": 16415
    },
    {
      "epoch": 0.98018147086915,
      "grad_norm": 0.47877657413482666,
      "learning_rate": 3.0094851461419387e-05,
      "loss": 0.4963,
      "mean_token_accuracy": 0.8376416683197021,
      "num_tokens": 550108217.0,
      "step": 16420
    },
    {
      "epoch": 0.9804799426934098,
      "grad_norm": 0.48808667063713074,
      "learning_rate": 3.008382089082845e-05,
      "loss": 0.4884,
      "mean_token_accuracy": 0.8425265431404114,
      "num_tokens": 550275897.0,
      "step": 16425
    },
    {
      "epoch": 0.9807784145176696,
      "grad_norm": 0.46539559960365295,
      "learning_rate": 3.007278969090025e-05,
      "loss": 0.5223,
      "mean_token_accuracy": 0.8330743074417114,
      "num_tokens": 550437385.0,
      "step": 16430
    },
    {
      "epoch": 0.9810768863419294,
      "grad_norm": 0.48131659626960754,
      "learning_rate": 3.0061757864321638e-05,
      "loss": 0.4955,
      "mean_token_accuracy": 0.8383454799652099,
      "num_tokens": 550605065.0,
      "step": 16435
    },
    {
      "epoch": 0.9813753581661891,
      "grad_norm": 0.576423704624176,
      "learning_rate": 3.0050725413779608e-05,
      "loss": 0.5799,
      "mean_token_accuracy": 0.8150900602340698,
      "num_tokens": 550772745.0,
      "step": 16440
    },
    {
      "epoch": 0.9816738299904489,
      "grad_norm": 0.5067886114120483,
      "learning_rate": 3.0039692341961317e-05,
      "loss": 0.5417,
      "mean_token_accuracy": 0.8267624855041504,
      "num_tokens": 550940425.0,
      "step": 16445
    },
    {
      "epoch": 0.9819723018147087,
      "grad_norm": 0.47427457571029663,
      "learning_rate": 3.0028658651554082e-05,
      "loss": 0.5234,
      "mean_token_accuracy": 0.8323452234268188,
      "num_tokens": 551108105.0,
      "step": 16450
    },
    {
      "epoch": 0.9822707736389685,
      "grad_norm": 0.4905551075935364,
      "learning_rate": 3.001762434524536e-05,
      "loss": 0.5356,
      "mean_token_accuracy": 0.8273887634277344,
      "num_tokens": 551275785.0,
      "step": 16455
    },
    {
      "epoch": 0.9825692454632283,
      "grad_norm": 0.45076194405555725,
      "learning_rate": 3.000658942572276e-05,
      "loss": 0.5331,
      "mean_token_accuracy": 0.8284146428108216,
      "num_tokens": 551443465.0,
      "step": 16460
    },
    {
      "epoch": 0.982867717287488,
      "grad_norm": 0.4858933091163635,
      "learning_rate": 2.999555389567404e-05,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8305379986763001,
      "num_tokens": 551611145.0,
      "step": 16465
    },
    {
      "epoch": 0.9831661891117478,
      "grad_norm": 0.5855818390846252,
      "learning_rate": 2.99845177577871e-05,
      "loss": 0.5697,
      "mean_token_accuracy": 0.8184301495552063,
      "num_tokens": 551778825.0,
      "step": 16470
    },
    {
      "epoch": 0.9834646609360076,
      "grad_norm": 0.517839252948761,
      "learning_rate": 2.997348101475001e-05,
      "loss": 0.554,
      "mean_token_accuracy": 0.8234939813613892,
      "num_tokens": 551946505.0,
      "step": 16475
    },
    {
      "epoch": 0.9837631327602674,
      "grad_norm": 0.5414894223213196,
      "learning_rate": 2.996244366925095e-05,
      "loss": 0.5432,
      "mean_token_accuracy": 0.8261660575866699,
      "num_tokens": 552114185.0,
      "step": 16480
    },
    {
      "epoch": 0.9840616045845272,
      "grad_norm": 0.4802056550979614,
      "learning_rate": 2.9951405723978297e-05,
      "loss": 0.5078,
      "mean_token_accuracy": 0.834886085987091,
      "num_tokens": 552281865.0,
      "step": 16485
    },
    {
      "epoch": 0.984360076408787,
      "grad_norm": 0.5247105360031128,
      "learning_rate": 2.9940367181620522e-05,
      "loss": 0.4974,
      "mean_token_accuracy": 0.8404568672180176,
      "num_tokens": 552449545.0,
      "step": 16490
    },
    {
      "epoch": 0.9846585482330468,
      "grad_norm": 0.4956377446651459,
      "learning_rate": 2.9929328044866283e-05,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8321364760398865,
      "num_tokens": 552617225.0,
      "step": 16495
    },
    {
      "epoch": 0.9849570200573066,
      "grad_norm": 0.4788982570171356,
      "learning_rate": 2.9918288316404358e-05,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8384528279304504,
      "num_tokens": 552784905.0,
      "step": 16500
    },
    {
      "epoch": 0.9852554918815664,
      "grad_norm": 0.5005519390106201,
      "learning_rate": 2.9907247998923677e-05,
      "loss": 0.5257,
      "mean_token_accuracy": 0.8299773335456848,
      "num_tokens": 552952585.0,
      "step": 16505
    },
    {
      "epoch": 0.9855539637058262,
      "grad_norm": 0.5599433183670044,
      "learning_rate": 2.989620709511331e-05,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8315817713737488,
      "num_tokens": 553120265.0,
      "step": 16510
    },
    {
      "epoch": 0.985852435530086,
      "grad_norm": 0.4988783299922943,
      "learning_rate": 2.9885165607662474e-05,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8326248407363892,
      "num_tokens": 553281086.0,
      "step": 16515
    },
    {
      "epoch": 0.9861509073543457,
      "grad_norm": 0.46395114064216614,
      "learning_rate": 2.9874123539260533e-05,
      "loss": 0.5206,
      "mean_token_accuracy": 0.8307228922843933,
      "num_tokens": 553448766.0,
      "step": 16520
    },
    {
      "epoch": 0.9864493791786055,
      "grad_norm": 0.5822600722312927,
      "learning_rate": 2.9863080892596978e-05,
      "loss": 0.5299,
      "mean_token_accuracy": 0.8284921765327453,
      "num_tokens": 553616446.0,
      "step": 16525
    },
    {
      "epoch": 0.9867478510028653,
      "grad_norm": 0.5490720272064209,
      "learning_rate": 2.985203767036146e-05,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8197602272033692,
      "num_tokens": 553784126.0,
      "step": 16530
    },
    {
      "epoch": 0.9870463228271251,
      "grad_norm": 0.5024212598800659,
      "learning_rate": 2.9840993875243754e-05,
      "loss": 0.5264,
      "mean_token_accuracy": 0.8299236416816711,
      "num_tokens": 553951806.0,
      "step": 16535
    },
    {
      "epoch": 0.9873447946513849,
      "grad_norm": 0.5087766647338867,
      "learning_rate": 2.9829949509933775e-05,
      "loss": 0.5418,
      "mean_token_accuracy": 0.8258678317070007,
      "num_tokens": 554119486.0,
      "step": 16540
    },
    {
      "epoch": 0.9876432664756447,
      "grad_norm": 0.46259379386901855,
      "learning_rate": 2.9818904577121587e-05,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8383991479873657,
      "num_tokens": 554287166.0,
      "step": 16545
    },
    {
      "epoch": 0.9879417382999045,
      "grad_norm": 0.4664643704891205,
      "learning_rate": 2.9807859079497397e-05,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8427591562271118,
      "num_tokens": 554454846.0,
      "step": 16550
    },
    {
      "epoch": 0.9882402101241643,
      "grad_norm": 0.5438719391822815,
      "learning_rate": 2.979681301975152e-05,
      "loss": 0.5444,
      "mean_token_accuracy": 0.8262555122375488,
      "num_tokens": 554622526.0,
      "step": 16555
    },
    {
      "epoch": 0.9885386819484241,
      "grad_norm": 0.5597907900810242,
      "learning_rate": 2.9785766400574443e-05,
      "loss": 0.5178,
      "mean_token_accuracy": 0.8334903836250305,
      "num_tokens": 554790206.0,
      "step": 16560
    },
    {
      "epoch": 0.9888371537726839,
      "grad_norm": 0.42497727274894714,
      "learning_rate": 2.9774719224656756e-05,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8307348132133484,
      "num_tokens": 554957886.0,
      "step": 16565
    },
    {
      "epoch": 0.9891356255969437,
      "grad_norm": 0.5093929767608643,
      "learning_rate": 2.9763671494689232e-05,
      "loss": 0.5246,
      "mean_token_accuracy": 0.832417094707489,
      "num_tokens": 555119436.0,
      "step": 16570
    },
    {
      "epoch": 0.9894340974212035,
      "grad_norm": 0.49292880296707153,
      "learning_rate": 2.9752623213362724e-05,
      "loss": 0.5195,
      "mean_token_accuracy": 0.8331981420516967,
      "num_tokens": 555287116.0,
      "step": 16575
    },
    {
      "epoch": 0.9897325692454633,
      "grad_norm": 0.49336308240890503,
      "learning_rate": 2.974157438336826e-05,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8432422757148743,
      "num_tokens": 555454796.0,
      "step": 16580
    },
    {
      "epoch": 0.9900310410697231,
      "grad_norm": 0.526658833026886,
      "learning_rate": 2.9730525007396974e-05,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8256948709487915,
      "num_tokens": 555622476.0,
      "step": 16585
    },
    {
      "epoch": 0.9903295128939829,
      "grad_norm": 0.5207714438438416,
      "learning_rate": 2.971947508814016e-05,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8277645111083984,
      "num_tokens": 555790156.0,
      "step": 16590
    },
    {
      "epoch": 0.9906279847182426,
      "grad_norm": 0.5171641707420349,
      "learning_rate": 2.9708424628289207e-05,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8342598080635071,
      "num_tokens": 555957836.0,
      "step": 16595
    },
    {
      "epoch": 0.9909264565425024,
      "grad_norm": 0.487448126077652,
      "learning_rate": 2.969737363053568e-05,
      "loss": 0.5029,
      "mean_token_accuracy": 0.8397351741790772,
      "num_tokens": 556125516.0,
      "step": 16600
    },
    {
      "epoch": 0.9912249283667621,
      "grad_norm": 0.49776336550712585,
      "learning_rate": 2.968632209757125e-05,
      "loss": 0.4881,
      "mean_token_accuracy": 0.8399439334869385,
      "num_tokens": 556293196.0,
      "step": 16605
    },
    {
      "epoch": 0.9915234001910219,
      "grad_norm": 0.500651478767395,
      "learning_rate": 2.9675270032087714e-05,
      "loss": 0.5356,
      "mean_token_accuracy": 0.8271322965621948,
      "num_tokens": 556460876.0,
      "step": 16610
    },
    {
      "epoch": 0.9918218720152817,
      "grad_norm": 0.4809315502643585,
      "learning_rate": 2.966421743677702e-05,
      "loss": 0.5184,
      "mean_token_accuracy": 0.8325181841850281,
      "num_tokens": 556628556.0,
      "step": 16615
    },
    {
      "epoch": 0.9921203438395415,
      "grad_norm": 0.5235933661460876,
      "learning_rate": 2.965316431433121e-05,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8260706067085266,
      "num_tokens": 556796236.0,
      "step": 16620
    },
    {
      "epoch": 0.9924188156638013,
      "grad_norm": 0.5473774671554565,
      "learning_rate": 2.964211066744249e-05,
      "loss": 0.5009,
      "mean_token_accuracy": 0.8375462174415589,
      "num_tokens": 556963916.0,
      "step": 16625
    },
    {
      "epoch": 0.9927172874880611,
      "grad_norm": 0.47868818044662476,
      "learning_rate": 2.9631056498803182e-05,
      "loss": 0.4994,
      "mean_token_accuracy": 0.8415423989295959,
      "num_tokens": 557131596.0,
      "step": 16630
    },
    {
      "epoch": 0.9930157593123209,
      "grad_norm": 0.5369992852210999,
      "learning_rate": 2.962000181110573e-05,
      "loss": 0.5513,
      "mean_token_accuracy": 0.8234522342681885,
      "num_tokens": 557299276.0,
      "step": 16635
    },
    {
      "epoch": 0.9933142311365807,
      "grad_norm": 0.433292418718338,
      "learning_rate": 2.960894660704271e-05,
      "loss": 0.5196,
      "mean_token_accuracy": 0.8330251693725585,
      "num_tokens": 557466956.0,
      "step": 16640
    },
    {
      "epoch": 0.9936127029608405,
      "grad_norm": 0.48996397852897644,
      "learning_rate": 2.9597890889306813e-05,
      "loss": 0.5077,
      "mean_token_accuracy": 0.8355123400688171,
      "num_tokens": 557634636.0,
      "step": 16645
    },
    {
      "epoch": 0.9939111747851003,
      "grad_norm": 0.523293137550354,
      "learning_rate": 2.958683466059086e-05,
      "loss": 0.5123,
      "mean_token_accuracy": 0.8356972455978393,
      "num_tokens": 557802316.0,
      "step": 16650
    },
    {
      "epoch": 0.9942096466093601,
      "grad_norm": 0.48072439432144165,
      "learning_rate": 2.9575777923587815e-05,
      "loss": 0.4863,
      "mean_token_accuracy": 0.8418108105659485,
      "num_tokens": 557969996.0,
      "step": 16655
    },
    {
      "epoch": 0.9945081184336199,
      "grad_norm": 0.4458793103694916,
      "learning_rate": 2.956472068099074e-05,
      "loss": 0.4911,
      "mean_token_accuracy": 0.8398186802864075,
      "num_tokens": 558137676.0,
      "step": 16660
    },
    {
      "epoch": 0.9948065902578797,
      "grad_norm": 0.4677237868309021,
      "learning_rate": 2.955366293549283e-05,
      "loss": 0.4879,
      "mean_token_accuracy": 0.8403912663459778,
      "num_tokens": 558305356.0,
      "step": 16665
    },
    {
      "epoch": 0.9951050620821394,
      "grad_norm": 0.46648067235946655,
      "learning_rate": 2.954260468978741e-05,
      "loss": 0.5127,
      "mean_token_accuracy": 0.8355540990829468,
      "num_tokens": 558473036.0,
      "step": 16670
    },
    {
      "epoch": 0.9954035339063992,
      "grad_norm": 0.47389277815818787,
      "learning_rate": 2.9531545946567917e-05,
      "loss": 0.534,
      "mean_token_accuracy": 0.8278718948364258,
      "num_tokens": 558640716.0,
      "step": 16675
    },
    {
      "epoch": 0.995702005730659,
      "grad_norm": 0.5456295013427734,
      "learning_rate": 2.9520486708527907e-05,
      "loss": 0.5273,
      "mean_token_accuracy": 0.8305558800697327,
      "num_tokens": 558808396.0,
      "step": 16680
    },
    {
      "epoch": 0.9960004775549188,
      "grad_norm": 0.5450401306152344,
      "learning_rate": 2.9509426978361073e-05,
      "loss": 0.5236,
      "mean_token_accuracy": 0.8299296259880066,
      "num_tokens": 558976076.0,
      "step": 16685
    },
    {
      "epoch": 0.9962989493791786,
      "grad_norm": 0.5304602384567261,
      "learning_rate": 2.9498366758761197e-05,
      "loss": 0.5253,
      "mean_token_accuracy": 0.8312477588653564,
      "num_tokens": 559143756.0,
      "step": 16690
    },
    {
      "epoch": 0.9965974212034384,
      "grad_norm": 0.4793808162212372,
      "learning_rate": 2.9487306052422213e-05,
      "loss": 0.4857,
      "mean_token_accuracy": 0.8424490213394165,
      "num_tokens": 559311436.0,
      "step": 16695
    },
    {
      "epoch": 0.9968958930276982,
      "grad_norm": 0.49623870849609375,
      "learning_rate": 2.9476244862038168e-05,
      "loss": 0.4742,
      "mean_token_accuracy": 0.845276153087616,
      "num_tokens": 559479116.0,
      "step": 16700
    },
    {
      "epoch": 0.997194364851958,
      "grad_norm": 0.5237420201301575,
      "learning_rate": 2.94651831903032e-05,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8301801204681396,
      "num_tokens": 559646796.0,
      "step": 16705
    },
    {
      "epoch": 0.9974928366762178,
      "grad_norm": 0.510319173336029,
      "learning_rate": 2.9454121039911597e-05,
      "loss": 0.5267,
      "mean_token_accuracy": 0.8292496800422668,
      "num_tokens": 559814476.0,
      "step": 16710
    },
    {
      "epoch": 0.9977913085004776,
      "grad_norm": 0.5285435914993286,
      "learning_rate": 2.9443058413557746e-05,
      "loss": 0.5276,
      "mean_token_accuracy": 0.8320768117904663,
      "num_tokens": 559982156.0,
      "step": 16715
    },
    {
      "epoch": 0.9980897803247374,
      "grad_norm": 0.49364015460014343,
      "learning_rate": 2.943199531393615e-05,
      "loss": 0.4902,
      "mean_token_accuracy": 0.842454981803894,
      "num_tokens": 560149836.0,
      "step": 16720
    },
    {
      "epoch": 0.9983882521489972,
      "grad_norm": 0.4487300217151642,
      "learning_rate": 2.9420931743741443e-05,
      "loss": 0.5389,
      "mean_token_accuracy": 0.826690924167633,
      "num_tokens": 560317516.0,
      "step": 16725
    },
    {
      "epoch": 0.998686723973257,
      "grad_norm": 0.5413393378257751,
      "learning_rate": 2.9409867705668343e-05,
      "loss": 0.5819,
      "mean_token_accuracy": 0.8143027544021606,
      "num_tokens": 560485196.0,
      "step": 16730
    },
    {
      "epoch": 0.9989851957975168,
      "grad_norm": 0.4478927254676819,
      "learning_rate": 2.9398803202411718e-05,
      "loss": 0.5078,
      "mean_token_accuracy": 0.8353453397750854,
      "num_tokens": 560652876.0,
      "step": 16735
    },
    {
      "epoch": 0.9992836676217765,
      "grad_norm": 0.5136594176292419,
      "learning_rate": 2.9387738236666522e-05,
      "loss": 0.5292,
      "mean_token_accuracy": 0.8307467579841614,
      "num_tokens": 560820556.0,
      "step": 16740
    },
    {
      "epoch": 0.9995821394460362,
      "grad_norm": 0.604115903377533,
      "learning_rate": 2.9376672811127837e-05,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8377668976783752,
      "num_tokens": 560988236.0,
      "step": 16745
    },
    {
      "epoch": 0.999880611270296,
      "grad_norm": 0.498207688331604,
      "learning_rate": 2.9365606928490847e-05,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8410831332206726,
      "num_tokens": 561155916.0,
      "step": 16750
    },
    {
      "epoch": 1.0001790830945558,
      "grad_norm": 0.4530643820762634,
      "learning_rate": 2.9354540591450863e-05,
      "loss": 0.4816,
      "mean_token_accuracy": 0.8472503900527955,
      "num_tokens": 561306828.0,
      "step": 16755
    },
    {
      "epoch": 1.0004775549188156,
      "grad_norm": 0.5342052578926086,
      "learning_rate": 2.934347380270327e-05,
      "loss": 0.4478,
      "mean_token_accuracy": 0.8516342759132385,
      "num_tokens": 561474508.0,
      "step": 16760
    },
    {
      "epoch": 1.0007760267430754,
      "grad_norm": 0.5704651474952698,
      "learning_rate": 2.933240656494362e-05,
      "loss": 0.4743,
      "mean_token_accuracy": 0.8443039655685425,
      "num_tokens": 561642188.0,
      "step": 16765
    },
    {
      "epoch": 1.0010744985673352,
      "grad_norm": 0.507685661315918,
      "learning_rate": 2.9321338880867523e-05,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8531253695487976,
      "num_tokens": 561809868.0,
      "step": 16770
    },
    {
      "epoch": 1.001372970391595,
      "grad_norm": 0.5275362133979797,
      "learning_rate": 2.9310270753170727e-05,
      "loss": 0.45,
      "mean_token_accuracy": 0.8498031616210937,
      "num_tokens": 561977548.0,
      "step": 16775
    },
    {
      "epoch": 1.0016714422158548,
      "grad_norm": 0.6999732851982117,
      "learning_rate": 2.929920218454907e-05,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8510557174682617,
      "num_tokens": 562145228.0,
      "step": 16780
    },
    {
      "epoch": 1.0019699140401146,
      "grad_norm": 0.5578259229660034,
      "learning_rate": 2.9288133177698512e-05,
      "loss": 0.4745,
      "mean_token_accuracy": 0.8445007801055908,
      "num_tokens": 562312908.0,
      "step": 16785
    },
    {
      "epoch": 1.0022683858643744,
      "grad_norm": 0.4876433312892914,
      "learning_rate": 2.927706373531512e-05,
      "loss": 0.4543,
      "mean_token_accuracy": 0.8488130807876587,
      "num_tokens": 562480588.0,
      "step": 16790
    },
    {
      "epoch": 1.0025668576886342,
      "grad_norm": 0.6000683307647705,
      "learning_rate": 2.9265993860095054e-05,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8427889823913575,
      "num_tokens": 562648268.0,
      "step": 16795
    },
    {
      "epoch": 1.002865329512894,
      "grad_norm": 0.485141783952713,
      "learning_rate": 2.9254923554734576e-05,
      "loss": 0.4666,
      "mean_token_accuracy": 0.847439420223236,
      "num_tokens": 562809626.0,
      "step": 16800
    },
    {
      "epoch": 1.0031638013371538,
      "grad_norm": 0.6160366535186768,
      "learning_rate": 2.9243852821930084e-05,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8355898857116699,
      "num_tokens": 562977306.0,
      "step": 16805
    },
    {
      "epoch": 1.0034622731614136,
      "grad_norm": 0.5146437287330627,
      "learning_rate": 2.9232781664378045e-05,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8429142236709595,
      "num_tokens": 563144986.0,
      "step": 16810
    },
    {
      "epoch": 1.0037607449856734,
      "grad_norm": 0.4939850866794586,
      "learning_rate": 2.922171008477505e-05,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8421149969100952,
      "num_tokens": 563312666.0,
      "step": 16815
    },
    {
      "epoch": 1.0040592168099332,
      "grad_norm": 0.582244873046875,
      "learning_rate": 2.9210638085817782e-05,
      "loss": 0.4888,
      "mean_token_accuracy": 0.8381903767585754,
      "num_tokens": 563480346.0,
      "step": 16820
    },
    {
      "epoch": 1.004357688634193,
      "grad_norm": 0.47668758034706116,
      "learning_rate": 2.9199565670203037e-05,
      "loss": 0.4268,
      "mean_token_accuracy": 0.8566980719566345,
      "num_tokens": 563648026.0,
      "step": 16825
    },
    {
      "epoch": 1.0046561604584527,
      "grad_norm": 0.45325109362602234,
      "learning_rate": 2.9188492840627703e-05,
      "loss": 0.4134,
      "mean_token_accuracy": 0.8622331023216248,
      "num_tokens": 563815706.0,
      "step": 16830
    },
    {
      "epoch": 1.0049546322827125,
      "grad_norm": 0.4884812831878662,
      "learning_rate": 2.9177419599788775e-05,
      "loss": 0.4464,
      "mean_token_accuracy": 0.8531015157699585,
      "num_tokens": 563983386.0,
      "step": 16835
    },
    {
      "epoch": 1.0052531041069723,
      "grad_norm": 0.5482917428016663,
      "learning_rate": 2.9166345950383332e-05,
      "loss": 0.4632,
      "mean_token_accuracy": 0.8472265243530274,
      "num_tokens": 564151066.0,
      "step": 16840
    },
    {
      "epoch": 1.0055515759312321,
      "grad_norm": 0.5486830472946167,
      "learning_rate": 2.915527189510858e-05,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8506859183311463,
      "num_tokens": 564318746.0,
      "step": 16845
    },
    {
      "epoch": 1.005850047755492,
      "grad_norm": 0.4940612316131592,
      "learning_rate": 2.9144197436661815e-05,
      "loss": 0.4871,
      "mean_token_accuracy": 0.8405463457107544,
      "num_tokens": 564486426.0,
      "step": 16850
    },
    {
      "epoch": 1.0061485195797517,
      "grad_norm": 0.4869822859764099,
      "learning_rate": 2.91331225777404e-05,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8547298073768616,
      "num_tokens": 564654106.0,
      "step": 16855
    },
    {
      "epoch": 1.0064469914040115,
      "grad_norm": 0.4788769781589508,
      "learning_rate": 2.912204732104185e-05,
      "loss": 0.4439,
      "mean_token_accuracy": 0.8526959300041199,
      "num_tokens": 564821786.0,
      "step": 16860
    },
    {
      "epoch": 1.0067454632282713,
      "grad_norm": 0.4857526123523712,
      "learning_rate": 2.9110971669263727e-05,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8441488742828369,
      "num_tokens": 564989466.0,
      "step": 16865
    },
    {
      "epoch": 1.007043935052531,
      "grad_norm": 0.5107783675193787,
      "learning_rate": 2.9099895625103724e-05,
      "loss": 0.4423,
      "mean_token_accuracy": 0.8519086122512818,
      "num_tokens": 565157146.0,
      "step": 16870
    },
    {
      "epoch": 1.0073424068767909,
      "grad_norm": 0.542470395565033,
      "learning_rate": 2.9088819191259603e-05,
      "loss": 0.4927,
      "mean_token_accuracy": 0.8381546139717102,
      "num_tokens": 565324826.0,
      "step": 16875
    },
    {
      "epoch": 1.0076408787010507,
      "grad_norm": 0.4769206941127777,
      "learning_rate": 2.9077742370429252e-05,
      "loss": 0.4668,
      "mean_token_accuracy": 0.846540629863739,
      "num_tokens": 565492506.0,
      "step": 16880
    },
    {
      "epoch": 1.0079393505253105,
      "grad_norm": 0.5395178198814392,
      "learning_rate": 2.9066665165310614e-05,
      "loss": 0.4516,
      "mean_token_accuracy": 0.8497673988342285,
      "num_tokens": 565660186.0,
      "step": 16885
    },
    {
      "epoch": 1.0082378223495703,
      "grad_norm": 0.5594173669815063,
      "learning_rate": 2.9055587578601767e-05,
      "loss": 0.4627,
      "mean_token_accuracy": 0.8471191644668579,
      "num_tokens": 565827866.0,
      "step": 16890
    },
    {
      "epoch": 1.00853629417383,
      "grad_norm": 0.5374088883399963,
      "learning_rate": 2.904450961300084e-05,
      "loss": 0.4725,
      "mean_token_accuracy": 0.8446319818496704,
      "num_tokens": 565995546.0,
      "step": 16895
    },
    {
      "epoch": 1.0088347659980899,
      "grad_norm": 0.6275510787963867,
      "learning_rate": 2.9033431271206106e-05,
      "loss": 0.4645,
      "mean_token_accuracy": 0.8466241240501404,
      "num_tokens": 566163226.0,
      "step": 16900
    },
    {
      "epoch": 1.0091332378223496,
      "grad_norm": 0.5092877745628357,
      "learning_rate": 2.9022352555915865e-05,
      "loss": 0.4399,
      "mean_token_accuracy": 0.852731728553772,
      "num_tokens": 566330906.0,
      "step": 16905
    },
    {
      "epoch": 1.0094317096466094,
      "grad_norm": 0.5199767351150513,
      "learning_rate": 2.9011273469828577e-05,
      "loss": 0.436,
      "mean_token_accuracy": 0.8531432628631592,
      "num_tokens": 566498586.0,
      "step": 16910
    },
    {
      "epoch": 1.0097301814708692,
      "grad_norm": 0.4958356022834778,
      "learning_rate": 2.9000194015642728e-05,
      "loss": 0.4413,
      "mean_token_accuracy": 0.8539126873016357,
      "num_tokens": 566666266.0,
      "step": 16915
    },
    {
      "epoch": 1.010028653295129,
      "grad_norm": 0.49873092770576477,
      "learning_rate": 2.8989114196056938e-05,
      "loss": 0.4339,
      "mean_token_accuracy": 0.8544793128967285,
      "num_tokens": 566833946.0,
      "step": 16920
    },
    {
      "epoch": 1.0103271251193888,
      "grad_norm": 0.5273438692092896,
      "learning_rate": 2.89780340137699e-05,
      "loss": 0.4748,
      "mean_token_accuracy": 0.8430037021636962,
      "num_tokens": 567001626.0,
      "step": 16925
    },
    {
      "epoch": 1.0106255969436486,
      "grad_norm": 0.4734728932380676,
      "learning_rate": 2.8966953471480396e-05,
      "loss": 0.4599,
      "mean_token_accuracy": 0.8484253764152527,
      "num_tokens": 567169306.0,
      "step": 16930
    },
    {
      "epoch": 1.0109240687679084,
      "grad_norm": 0.4726862609386444,
      "learning_rate": 2.89558725718873e-05,
      "loss": 0.4581,
      "mean_token_accuracy": 0.850614321231842,
      "num_tokens": 567336986.0,
      "step": 16935
    },
    {
      "epoch": 1.0112225405921682,
      "grad_norm": 0.4735446572303772,
      "learning_rate": 2.894479131768957e-05,
      "loss": 0.4558,
      "mean_token_accuracy": 0.8485148429870606,
      "num_tokens": 567504666.0,
      "step": 16940
    },
    {
      "epoch": 1.011521012416428,
      "grad_norm": 0.5605773329734802,
      "learning_rate": 2.8933709711586245e-05,
      "loss": 0.4495,
      "mean_token_accuracy": 0.8506501197814942,
      "num_tokens": 567672346.0,
      "step": 16945
    },
    {
      "epoch": 1.0118194842406876,
      "grad_norm": 0.49589309096336365,
      "learning_rate": 2.892262775627645e-05,
      "loss": 0.4338,
      "mean_token_accuracy": 0.8552129387855529,
      "num_tokens": 567840026.0,
      "step": 16950
    },
    {
      "epoch": 1.0121179560649474,
      "grad_norm": 0.4771466553211212,
      "learning_rate": 2.8911545454459406e-05,
      "loss": 0.4663,
      "mean_token_accuracy": 0.8449660062789917,
      "num_tokens": 568007706.0,
      "step": 16955
    },
    {
      "epoch": 1.0124164278892072,
      "grad_norm": 0.47365638613700867,
      "learning_rate": 2.8900462808834416e-05,
      "loss": 0.4221,
      "mean_token_accuracy": 0.8595013737678527,
      "num_tokens": 568175386.0,
      "step": 16960
    },
    {
      "epoch": 1.012714899713467,
      "grad_norm": 0.4710509181022644,
      "learning_rate": 2.888937982210086e-05,
      "loss": 0.4494,
      "mean_token_accuracy": 0.8508409857749939,
      "num_tokens": 568343066.0,
      "step": 16965
    },
    {
      "epoch": 1.0130133715377267,
      "grad_norm": 0.47972074151039124,
      "learning_rate": 2.8878296496958208e-05,
      "loss": 0.4292,
      "mean_token_accuracy": 0.8572110176086426,
      "num_tokens": 568510746.0,
      "step": 16970
    },
    {
      "epoch": 1.0133118433619865,
      "grad_norm": 0.5294878482818604,
      "learning_rate": 2.8867212836105995e-05,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8439281940460205,
      "num_tokens": 568678426.0,
      "step": 16975
    },
    {
      "epoch": 1.0136103151862463,
      "grad_norm": 0.49053069949150085,
      "learning_rate": 2.8856128842243863e-05,
      "loss": 0.4593,
      "mean_token_accuracy": 0.8476857900619507,
      "num_tokens": 568846106.0,
      "step": 16980
    },
    {
      "epoch": 1.0139087870105061,
      "grad_norm": 0.9311002492904663,
      "learning_rate": 2.8845044518071523e-05,
      "loss": 0.4713,
      "mean_token_accuracy": 0.845526647567749,
      "num_tokens": 569013786.0,
      "step": 16985
    },
    {
      "epoch": 1.014207258834766,
      "grad_norm": 0.4663950502872467,
      "learning_rate": 2.883395986628876e-05,
      "loss": 0.4119,
      "mean_token_accuracy": 0.8621734499931335,
      "num_tokens": 569181466.0,
      "step": 16990
    },
    {
      "epoch": 1.0145057306590257,
      "grad_norm": 0.49617063999176025,
      "learning_rate": 2.8822874889595452e-05,
      "loss": 0.464,
      "mean_token_accuracy": 0.8443337678909302,
      "num_tokens": 569349146.0,
      "step": 16995
    },
    {
      "epoch": 1.0148042024832855,
      "grad_norm": 0.5455309748649597,
      "learning_rate": 2.8811789590691545e-05,
      "loss": 0.4584,
      "mean_token_accuracy": 0.8497912526130676,
      "num_tokens": 569516826.0,
      "step": 17000
    },
    {
      "epoch": 1.0151026743075453,
      "grad_norm": 0.5212569832801819,
      "learning_rate": 2.8800703972277077e-05,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8501013994216919,
      "num_tokens": 569684506.0,
      "step": 17005
    },
    {
      "epoch": 1.015401146131805,
      "grad_norm": 0.5574789047241211,
      "learning_rate": 2.878961803705214e-05,
      "loss": 0.4558,
      "mean_token_accuracy": 0.8503101468086243,
      "num_tokens": 569852186.0,
      "step": 17010
    },
    {
      "epoch": 1.015699617956065,
      "grad_norm": 0.48953911662101746,
      "learning_rate": 2.8778531787716935e-05,
      "loss": 0.4464,
      "mean_token_accuracy": 0.8512883186340332,
      "num_tokens": 570019866.0,
      "step": 17015
    },
    {
      "epoch": 1.0159980897803247,
      "grad_norm": 0.502530038356781,
      "learning_rate": 2.8767445226971706e-05,
      "loss": 0.4501,
      "mean_token_accuracy": 0.8508469581604003,
      "num_tokens": 570187546.0,
      "step": 17020
    },
    {
      "epoch": 1.0162965616045845,
      "grad_norm": 0.4684881567955017,
      "learning_rate": 2.8756358357516805e-05,
      "loss": 0.4372,
      "mean_token_accuracy": 0.854825246334076,
      "num_tokens": 570355226.0,
      "step": 17025
    },
    {
      "epoch": 1.0165950334288443,
      "grad_norm": 0.49216997623443604,
      "learning_rate": 2.874527118205263e-05,
      "loss": 0.4222,
      "mean_token_accuracy": 0.8588094830513,
      "num_tokens": 570522906.0,
      "step": 17030
    },
    {
      "epoch": 1.016893505253104,
      "grad_norm": 0.5277981162071228,
      "learning_rate": 2.8734183703279672e-05,
      "loss": 0.4504,
      "mean_token_accuracy": 0.8510020256042481,
      "num_tokens": 570690586.0,
      "step": 17035
    },
    {
      "epoch": 1.0171919770773639,
      "grad_norm": 0.5185527801513672,
      "learning_rate": 2.8723095923898496e-05,
      "loss": 0.4586,
      "mean_token_accuracy": 0.8484373092651367,
      "num_tokens": 570858266.0,
      "step": 17040
    },
    {
      "epoch": 1.0174904489016237,
      "grad_norm": 0.4920758903026581,
      "learning_rate": 2.871200784660973e-05,
      "loss": 0.4603,
      "mean_token_accuracy": 0.8466778039932251,
      "num_tokens": 571025946.0,
      "step": 17045
    },
    {
      "epoch": 1.0177889207258835,
      "grad_norm": 0.8109400272369385,
      "learning_rate": 2.870091947411408e-05,
      "loss": 0.4408,
      "mean_token_accuracy": 0.8544315934181214,
      "num_tokens": 571193626.0,
      "step": 17050
    },
    {
      "epoch": 1.0180873925501432,
      "grad_norm": 0.45363783836364746,
      "learning_rate": 2.8689830809112332e-05,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8521770238876343,
      "num_tokens": 571361306.0,
      "step": 17055
    },
    {
      "epoch": 1.018385864374403,
      "grad_norm": 0.5127880573272705,
      "learning_rate": 2.8678741854305324e-05,
      "loss": 0.4584,
      "mean_token_accuracy": 0.8477036714553833,
      "num_tokens": 571528986.0,
      "step": 17060
    },
    {
      "epoch": 1.0186843361986628,
      "grad_norm": 0.5394222140312195,
      "learning_rate": 2.8667652612393984e-05,
      "loss": 0.489,
      "mean_token_accuracy": 0.8415179133415223,
      "num_tokens": 571696632.0,
      "step": 17065
    },
    {
      "epoch": 1.0189828080229226,
      "grad_norm": 0.5126343965530396,
      "learning_rate": 2.8656563086079298e-05,
      "loss": 0.425,
      "mean_token_accuracy": 0.8594894409179688,
      "num_tokens": 571864312.0,
      "step": 17070
    },
    {
      "epoch": 1.0192812798471824,
      "grad_norm": 0.4388852119445801,
      "learning_rate": 2.8645473278062324e-05,
      "loss": 0.4298,
      "mean_token_accuracy": 0.8574376583099366,
      "num_tokens": 572031992.0,
      "step": 17075
    },
    {
      "epoch": 1.0195797516714422,
      "grad_norm": 0.4752084016799927,
      "learning_rate": 2.863438319104419e-05,
      "loss": 0.4893,
      "mean_token_accuracy": 0.8369438290596009,
      "num_tokens": 572199672.0,
      "step": 17080
    },
    {
      "epoch": 1.019878223495702,
      "grad_norm": 0.4854961335659027,
      "learning_rate": 2.8623292827726094e-05,
      "loss": 0.4365,
      "mean_token_accuracy": 0.8559226989746094,
      "num_tokens": 572367352.0,
      "step": 17085
    },
    {
      "epoch": 1.0201766953199618,
      "grad_norm": 0.5119810700416565,
      "learning_rate": 2.8612202190809285e-05,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8451747536659241,
      "num_tokens": 572535032.0,
      "step": 17090
    },
    {
      "epoch": 1.0204751671442216,
      "grad_norm": 0.6095016598701477,
      "learning_rate": 2.8601111282995107e-05,
      "loss": 0.4793,
      "mean_token_accuracy": 0.8423833966255188,
      "num_tokens": 572702712.0,
      "step": 17095
    },
    {
      "epoch": 1.0207736389684814,
      "grad_norm": 0.49058955907821655,
      "learning_rate": 2.859002010698496e-05,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8513121843338013,
      "num_tokens": 572870392.0,
      "step": 17100
    },
    {
      "epoch": 1.0210721107927412,
      "grad_norm": 0.5054989457130432,
      "learning_rate": 2.8578928665480297e-05,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8425086498260498,
      "num_tokens": 573038072.0,
      "step": 17105
    },
    {
      "epoch": 1.021370582617001,
      "grad_norm": 0.5695568323135376,
      "learning_rate": 2.856783696118263e-05,
      "loss": 0.4453,
      "mean_token_accuracy": 0.8508708238601684,
      "num_tokens": 573205752.0,
      "step": 17110
    },
    {
      "epoch": 1.0216690544412608,
      "grad_norm": 0.505357563495636,
      "learning_rate": 2.855674499679357e-05,
      "loss": 0.4176,
      "mean_token_accuracy": 0.8603320598602295,
      "num_tokens": 573366307.0,
      "step": 17115
    },
    {
      "epoch": 1.0219675262655206,
      "grad_norm": 0.5061622858047485,
      "learning_rate": 2.854565277501476e-05,
      "loss": 0.4623,
      "mean_token_accuracy": 0.8473219633102417,
      "num_tokens": 573533987.0,
      "step": 17120
    },
    {
      "epoch": 1.0222659980897804,
      "grad_norm": 0.5195006132125854,
      "learning_rate": 2.853456029854792e-05,
      "loss": 0.4728,
      "mean_token_accuracy": 0.8437790751457215,
      "num_tokens": 573701667.0,
      "step": 17125
    },
    {
      "epoch": 1.0225644699140402,
      "grad_norm": 0.4995376169681549,
      "learning_rate": 2.8523467570094816e-05,
      "loss": 0.4497,
      "mean_token_accuracy": 0.8511571049690246,
      "num_tokens": 573869347.0,
      "step": 17130
    },
    {
      "epoch": 1.0228629417383,
      "grad_norm": 0.5369946360588074,
      "learning_rate": 2.8512374592357307e-05,
      "loss": 0.463,
      "mean_token_accuracy": 0.8471847891807556,
      "num_tokens": 574037027.0,
      "step": 17135
    },
    {
      "epoch": 1.0231614135625597,
      "grad_norm": 0.4634990394115448,
      "learning_rate": 2.8501281368037265e-05,
      "loss": 0.4673,
      "mean_token_accuracy": 0.8460455775260926,
      "num_tokens": 574204707.0,
      "step": 17140
    },
    {
      "epoch": 1.0234598853868195,
      "grad_norm": 0.5192978978157043,
      "learning_rate": 2.8490187899836668e-05,
      "loss": 0.4741,
      "mean_token_accuracy": 0.8441548466682434,
      "num_tokens": 574372387.0,
      "step": 17145
    },
    {
      "epoch": 1.0237583572110793,
      "grad_norm": 0.5141440629959106,
      "learning_rate": 2.847909419045754e-05,
      "loss": 0.4441,
      "mean_token_accuracy": 0.8513002395629883,
      "num_tokens": 574540067.0,
      "step": 17150
    },
    {
      "epoch": 1.0240568290353391,
      "grad_norm": 0.4873316287994385,
      "learning_rate": 2.8468000242601938e-05,
      "loss": 0.4313,
      "mean_token_accuracy": 0.8560896992683411,
      "num_tokens": 574707747.0,
      "step": 17155
    },
    {
      "epoch": 1.024355300859599,
      "grad_norm": 0.4926494061946869,
      "learning_rate": 2.8456906058972014e-05,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8442025423049927,
      "num_tokens": 574875427.0,
      "step": 17160
    },
    {
      "epoch": 1.0246537726838587,
      "grad_norm": 0.5193369388580322,
      "learning_rate": 2.844581164226996e-05,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8519265174865722,
      "num_tokens": 575043107.0,
      "step": 17165
    },
    {
      "epoch": 1.0249522445081185,
      "grad_norm": 0.5441789031028748,
      "learning_rate": 2.8434716995198018e-05,
      "loss": 0.4491,
      "mean_token_accuracy": 0.8523917436599732,
      "num_tokens": 575210787.0,
      "step": 17170
    },
    {
      "epoch": 1.0252507163323783,
      "grad_norm": 0.4996625483036041,
      "learning_rate": 2.8423622120458504e-05,
      "loss": 0.4385,
      "mean_token_accuracy": 0.8542705535888672,
      "num_tokens": 575378467.0,
      "step": 17175
    },
    {
      "epoch": 1.025549188156638,
      "grad_norm": 0.4947277903556824,
      "learning_rate": 2.8412527020753778e-05,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8346295952796936,
      "num_tokens": 575546147.0,
      "step": 17180
    },
    {
      "epoch": 1.0258476599808979,
      "grad_norm": 0.475445032119751,
      "learning_rate": 2.840143169878624e-05,
      "loss": 0.422,
      "mean_token_accuracy": 0.8593820810317994,
      "num_tokens": 575713827.0,
      "step": 17185
    },
    {
      "epoch": 1.0261461318051577,
      "grad_norm": 0.5255208015441895,
      "learning_rate": 2.839033615725838e-05,
      "loss": 0.4577,
      "mean_token_accuracy": 0.8482762813568115,
      "num_tokens": 575881507.0,
      "step": 17190
    },
    {
      "epoch": 1.0264446036294175,
      "grad_norm": 0.4992060661315918,
      "learning_rate": 2.837924039887272e-05,
      "loss": 0.4456,
      "mean_token_accuracy": 0.8530239820480346,
      "num_tokens": 576049187.0,
      "step": 17195
    },
    {
      "epoch": 1.0267430754536773,
      "grad_norm": 0.5629310607910156,
      "learning_rate": 2.8368144426331822e-05,
      "loss": 0.4599,
      "mean_token_accuracy": 0.8473577499389648,
      "num_tokens": 576216867.0,
      "step": 17200
    },
    {
      "epoch": 1.027041547277937,
      "grad_norm": 0.49983513355255127,
      "learning_rate": 2.835704824233833e-05,
      "loss": 0.4679,
      "mean_token_accuracy": 0.844787061214447,
      "num_tokens": 576384547.0,
      "step": 17205
    },
    {
      "epoch": 1.0273400191021969,
      "grad_norm": 0.547196090221405,
      "learning_rate": 2.834595184959492e-05,
      "loss": 0.4606,
      "mean_token_accuracy": 0.8481032967567443,
      "num_tokens": 576552227.0,
      "step": 17210
    },
    {
      "epoch": 1.0276384909264566,
      "grad_norm": 1.0569008588790894,
      "learning_rate": 2.833485525080432e-05,
      "loss": 0.4602,
      "mean_token_accuracy": 0.8474352836608887,
      "num_tokens": 576719907.0,
      "step": 17215
    },
    {
      "epoch": 1.0279369627507164,
      "grad_norm": 0.46995216608047485,
      "learning_rate": 2.832375844866932e-05,
      "loss": 0.4678,
      "mean_token_accuracy": 0.8440832614898681,
      "num_tokens": 576887587.0,
      "step": 17220
    },
    {
      "epoch": 1.028235434574976,
      "grad_norm": 0.519990086555481,
      "learning_rate": 2.8312661445892737e-05,
      "loss": 0.4546,
      "mean_token_accuracy": 0.8497554540634156,
      "num_tokens": 577055267.0,
      "step": 17225
    },
    {
      "epoch": 1.0285339063992358,
      "grad_norm": 0.566287100315094,
      "learning_rate": 2.830156424517746e-05,
      "loss": 0.4719,
      "mean_token_accuracy": 0.8484850287437439,
      "num_tokens": 577222947.0,
      "step": 17230
    },
    {
      "epoch": 1.0288323782234956,
      "grad_norm": 0.5435327291488647,
      "learning_rate": 2.829046684922642e-05,
      "loss": 0.4313,
      "mean_token_accuracy": 0.8563760042190551,
      "num_tokens": 577390627.0,
      "step": 17235
    },
    {
      "epoch": 1.0291308500477554,
      "grad_norm": 0.5225843191146851,
      "learning_rate": 2.827936926074259e-05,
      "loss": 0.4533,
      "mean_token_accuracy": 0.8496421456336976,
      "num_tokens": 577558307.0,
      "step": 17240
    },
    {
      "epoch": 1.0294293218720152,
      "grad_norm": 0.5453398823738098,
      "learning_rate": 2.8268271482428986e-05,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8497852802276611,
      "num_tokens": 577725987.0,
      "step": 17245
    },
    {
      "epoch": 1.029727793696275,
      "grad_norm": 0.5361250042915344,
      "learning_rate": 2.8257173516988688e-05,
      "loss": 0.4621,
      "mean_token_accuracy": 0.8470177769660949,
      "num_tokens": 577893667.0,
      "step": 17250
    },
    {
      "epoch": 1.0300262655205348,
      "grad_norm": 0.52862149477005,
      "learning_rate": 2.8246075367124797e-05,
      "loss": 0.4372,
      "mean_token_accuracy": 0.854598593711853,
      "num_tokens": 578061347.0,
      "step": 17255
    },
    {
      "epoch": 1.0303247373447946,
      "grad_norm": 0.5109806656837463,
      "learning_rate": 2.8234977035540483e-05,
      "loss": 0.445,
      "mean_token_accuracy": 0.8528033018112182,
      "num_tokens": 578229027.0,
      "step": 17260
    },
    {
      "epoch": 1.0306232091690544,
      "grad_norm": 0.5901560187339783,
      "learning_rate": 2.822387852493894e-05,
      "loss": 0.518,
      "mean_token_accuracy": 0.829023003578186,
      "num_tokens": 578396707.0,
      "step": 17265
    },
    {
      "epoch": 1.0309216809933142,
      "grad_norm": 0.5068350434303284,
      "learning_rate": 2.8212779838023435e-05,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8478527903556824,
      "num_tokens": 578564387.0,
      "step": 17270
    },
    {
      "epoch": 1.031220152817574,
      "grad_norm": 0.505020260810852,
      "learning_rate": 2.820168097749723e-05,
      "loss": 0.4771,
      "mean_token_accuracy": 0.8409638643264771,
      "num_tokens": 578732067.0,
      "step": 17275
    },
    {
      "epoch": 1.0315186246418337,
      "grad_norm": 0.5027931332588196,
      "learning_rate": 2.8190581946063678e-05,
      "loss": 0.4371,
      "mean_token_accuracy": 0.8546224594116211,
      "num_tokens": 578899747.0,
      "step": 17280
    },
    {
      "epoch": 1.0318170964660935,
      "grad_norm": 0.5489554405212402,
      "learning_rate": 2.817948274642614e-05,
      "loss": 0.4635,
      "mean_token_accuracy": 0.8468626976013184,
      "num_tokens": 579067427.0,
      "step": 17285
    },
    {
      "epoch": 1.0321155682903533,
      "grad_norm": 0.5266851186752319,
      "learning_rate": 2.8168383381288042e-05,
      "loss": 0.4685,
      "mean_token_accuracy": 0.8452701807022095,
      "num_tokens": 579235107.0,
      "step": 17290
    },
    {
      "epoch": 1.0324140401146131,
      "grad_norm": 0.6650678515434265,
      "learning_rate": 2.815728385335284e-05,
      "loss": 0.4371,
      "mean_token_accuracy": 0.8537755012512207,
      "num_tokens": 579402787.0,
      "step": 17295
    },
    {
      "epoch": 1.032712511938873,
      "grad_norm": 0.5936505198478699,
      "learning_rate": 2.8146184165324017e-05,
      "loss": 0.4865,
      "mean_token_accuracy": 0.8394429206848144,
      "num_tokens": 579570467.0,
      "step": 17300
    },
    {
      "epoch": 1.0330109837631327,
      "grad_norm": 0.4832249581813812,
      "learning_rate": 2.8135084319905114e-05,
      "loss": 0.4342,
      "mean_token_accuracy": 0.8561075925827026,
      "num_tokens": 579738147.0,
      "step": 17305
    },
    {
      "epoch": 1.0333094555873925,
      "grad_norm": 0.4816303849220276,
      "learning_rate": 2.81239843197997e-05,
      "loss": 0.4378,
      "mean_token_accuracy": 0.8538649678230286,
      "num_tokens": 579905827.0,
      "step": 17310
    },
    {
      "epoch": 1.0336079274116523,
      "grad_norm": 0.4698982834815979,
      "learning_rate": 2.8112884167711395e-05,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8508588790893554,
      "num_tokens": 580073507.0,
      "step": 17315
    },
    {
      "epoch": 1.033906399235912,
      "grad_norm": 0.5446665287017822,
      "learning_rate": 2.8101783866343834e-05,
      "loss": 0.4853,
      "mean_token_accuracy": 0.8409340500831604,
      "num_tokens": 580241187.0,
      "step": 17320
    },
    {
      "epoch": 1.034204871060172,
      "grad_norm": 0.49912068247795105,
      "learning_rate": 2.8090683418400705e-05,
      "loss": 0.45,
      "mean_token_accuracy": 0.850888705253601,
      "num_tokens": 580408867.0,
      "step": 17325
    },
    {
      "epoch": 1.0345033428844317,
      "grad_norm": 0.5369765758514404,
      "learning_rate": 2.8079582826585726e-05,
      "loss": 0.4346,
      "mean_token_accuracy": 0.8561493039131165,
      "num_tokens": 580571963.0,
      "step": 17330
    },
    {
      "epoch": 1.0348018147086915,
      "grad_norm": 0.527891993522644,
      "learning_rate": 2.8068482093602648e-05,
      "loss": 0.4669,
      "mean_token_accuracy": 0.8471967101097106,
      "num_tokens": 580739643.0,
      "step": 17335
    },
    {
      "epoch": 1.0351002865329513,
      "grad_norm": 0.5228444933891296,
      "learning_rate": 2.8057381222155266e-05,
      "loss": 0.4325,
      "mean_token_accuracy": 0.8545926213264465,
      "num_tokens": 580907323.0,
      "step": 17340
    },
    {
      "epoch": 1.035398758357211,
      "grad_norm": 0.516074001789093,
      "learning_rate": 2.8046280214947402e-05,
      "loss": 0.4469,
      "mean_token_accuracy": 0.8509722113609314,
      "num_tokens": 581075003.0,
      "step": 17345
    },
    {
      "epoch": 1.0356972301814709,
      "grad_norm": 0.5603234171867371,
      "learning_rate": 2.8035179074682903e-05,
      "loss": 0.4401,
      "mean_token_accuracy": 0.8537277817726135,
      "num_tokens": 581242683.0,
      "step": 17350
    },
    {
      "epoch": 1.0359957020057307,
      "grad_norm": 0.5408618450164795,
      "learning_rate": 2.8024077804065675e-05,
      "loss": 0.4893,
      "mean_token_accuracy": 0.8401467323303222,
      "num_tokens": 581410363.0,
      "step": 17355
    },
    {
      "epoch": 1.0362941738299905,
      "grad_norm": 0.5174014568328857,
      "learning_rate": 2.8012976405799613e-05,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8507694125175476,
      "num_tokens": 581578043.0,
      "step": 17360
    },
    {
      "epoch": 1.0365926456542502,
      "grad_norm": 0.5578038692474365,
      "learning_rate": 2.800187488258868e-05,
      "loss": 0.4741,
      "mean_token_accuracy": 0.8429082751274108,
      "num_tokens": 581745723.0,
      "step": 17365
    },
    {
      "epoch": 1.03689111747851,
      "grad_norm": 0.4302588999271393,
      "learning_rate": 2.799077323713685e-05,
      "loss": 0.4451,
      "mean_token_accuracy": 0.8539126873016357,
      "num_tokens": 581913403.0,
      "step": 17370
    },
    {
      "epoch": 1.0371895893027698,
      "grad_norm": 0.5418205261230469,
      "learning_rate": 2.7979671472148155e-05,
      "loss": 0.4419,
      "mean_token_accuracy": 0.8527615427970886,
      "num_tokens": 582081083.0,
      "step": 17375
    },
    {
      "epoch": 1.0374880611270296,
      "grad_norm": 0.4621972441673279,
      "learning_rate": 2.796856959032661e-05,
      "loss": 0.4606,
      "mean_token_accuracy": 0.8478527903556824,
      "num_tokens": 582248763.0,
      "step": 17380
    },
    {
      "epoch": 1.0377865329512894,
      "grad_norm": 0.5263122916221619,
      "learning_rate": 2.7957467594376298e-05,
      "loss": 0.4812,
      "mean_token_accuracy": 0.8401347875595093,
      "num_tokens": 582416443.0,
      "step": 17385
    },
    {
      "epoch": 1.0380850047755492,
      "grad_norm": 0.7780062556266785,
      "learning_rate": 2.7946365487001296e-05,
      "loss": 0.5261,
      "mean_token_accuracy": 0.8289514422416687,
      "num_tokens": 582584123.0,
      "step": 17390
    },
    {
      "epoch": 1.038383476599809,
      "grad_norm": 0.5385981202125549,
      "learning_rate": 2.7935263270905753e-05,
      "loss": 0.4359,
      "mean_token_accuracy": 0.8551890850067139,
      "num_tokens": 582751803.0,
      "step": 17395
    },
    {
      "epoch": 1.0386819484240688,
      "grad_norm": 0.5106160044670105,
      "learning_rate": 2.7924160948793798e-05,
      "loss": 0.4171,
      "mean_token_accuracy": 0.8621495962142944,
      "num_tokens": 582919483.0,
      "step": 17400
    },
    {
      "epoch": 1.0389804202483286,
      "grad_norm": 0.5020478963851929,
      "learning_rate": 2.7913058523369624e-05,
      "loss": 0.4232,
      "mean_token_accuracy": 0.8596325755119324,
      "num_tokens": 583087163.0,
      "step": 17405
    },
    {
      "epoch": 1.0392788920725884,
      "grad_norm": 0.44876614212989807,
      "learning_rate": 2.790195599733742e-05,
      "loss": 0.4332,
      "mean_token_accuracy": 0.855815327167511,
      "num_tokens": 583254843.0,
      "step": 17410
    },
    {
      "epoch": 1.0395773638968482,
      "grad_norm": 0.5497639775276184,
      "learning_rate": 2.7890853373401416e-05,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8499463319778442,
      "num_tokens": 583422523.0,
      "step": 17415
    },
    {
      "epoch": 1.039875835721108,
      "grad_norm": 0.5056437253952026,
      "learning_rate": 2.7879750654265847e-05,
      "loss": 0.452,
      "mean_token_accuracy": 0.8501790165901184,
      "num_tokens": 583589980.0,
      "step": 17420
    },
    {
      "epoch": 1.0401743075453678,
      "grad_norm": 0.5529347062110901,
      "learning_rate": 2.7868647842635008e-05,
      "loss": 0.4595,
      "mean_token_accuracy": 0.8473994970321655,
      "num_tokens": 583757660.0,
      "step": 17425
    },
    {
      "epoch": 1.0404727793696276,
      "grad_norm": 0.5109863877296448,
      "learning_rate": 2.785754494121317e-05,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8398544669151307,
      "num_tokens": 583925340.0,
      "step": 17430
    },
    {
      "epoch": 1.0407712511938874,
      "grad_norm": 0.6398445963859558,
      "learning_rate": 2.784644195270468e-05,
      "loss": 0.4313,
      "mean_token_accuracy": 0.8573422312736512,
      "num_tokens": 584093020.0,
      "step": 17435
    },
    {
      "epoch": 1.0410697230181472,
      "grad_norm": 0.5292453765869141,
      "learning_rate": 2.7835338879813842e-05,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8505725979804992,
      "num_tokens": 584260700.0,
      "step": 17440
    },
    {
      "epoch": 1.041368194842407,
      "grad_norm": 0.521647572517395,
      "learning_rate": 2.7824235725245042e-05,
      "loss": 0.4691,
      "mean_token_accuracy": 0.8443993926048279,
      "num_tokens": 584428380.0,
      "step": 17445
    },
    {
      "epoch": 1.0416666666666667,
      "grad_norm": 0.4575381875038147,
      "learning_rate": 2.7813132491702637e-05,
      "loss": 0.4374,
      "mean_token_accuracy": 0.8541274070739746,
      "num_tokens": 584596060.0,
      "step": 17450
    },
    {
      "epoch": 1.0419651384909265,
      "grad_norm": 0.5322298407554626,
      "learning_rate": 2.7802029181891044e-05,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8572289109230041,
      "num_tokens": 584763740.0,
      "step": 17455
    },
    {
      "epoch": 1.0422636103151863,
      "grad_norm": 0.5360782742500305,
      "learning_rate": 2.7790925798514656e-05,
      "loss": 0.4734,
      "mean_token_accuracy": 0.842556357383728,
      "num_tokens": 584931420.0,
      "step": 17460
    },
    {
      "epoch": 1.0425620821394461,
      "grad_norm": 0.4834948480129242,
      "learning_rate": 2.777982234427793e-05,
      "loss": 0.4472,
      "mean_token_accuracy": 0.8519145965576171,
      "num_tokens": 585099100.0,
      "step": 17465
    },
    {
      "epoch": 1.042860553963706,
      "grad_norm": 0.5424057841300964,
      "learning_rate": 2.7768718821885305e-05,
      "loss": 0.423,
      "mean_token_accuracy": 0.8586365342140198,
      "num_tokens": 585266780.0,
      "step": 17470
    },
    {
      "epoch": 1.0431590257879657,
      "grad_norm": 0.5537320971488953,
      "learning_rate": 2.7757615234041245e-05,
      "loss": 0.4408,
      "mean_token_accuracy": 0.8541095018386841,
      "num_tokens": 585434460.0,
      "step": 17475
    },
    {
      "epoch": 1.0434574976122255,
      "grad_norm": 0.49758070707321167,
      "learning_rate": 2.7746511583450246e-05,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8515865445137024,
      "num_tokens": 585602140.0,
      "step": 17480
    },
    {
      "epoch": 1.0437559694364853,
      "grad_norm": 0.5076404809951782,
      "learning_rate": 2.7735407872816797e-05,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8494453072547913,
      "num_tokens": 585769820.0,
      "step": 17485
    },
    {
      "epoch": 1.044054441260745,
      "grad_norm": 0.4706971347332001,
      "learning_rate": 2.772430410484543e-05,
      "loss": 0.4553,
      "mean_token_accuracy": 0.8479541897773742,
      "num_tokens": 585937500.0,
      "step": 17490
    },
    {
      "epoch": 1.0443529130850049,
      "grad_norm": 0.517845869064331,
      "learning_rate": 2.7713200282240635e-05,
      "loss": 0.4884,
      "mean_token_accuracy": 0.8387987613677979,
      "num_tokens": 586105180.0,
      "step": 17495
    },
    {
      "epoch": 1.0446513849092645,
      "grad_norm": 0.5571268200874329,
      "learning_rate": 2.7702096407706996e-05,
      "loss": 0.472,
      "mean_token_accuracy": 0.8445544481277466,
      "num_tokens": 586272860.0,
      "step": 17500
    },
    {
      "epoch": 1.0449498567335243,
      "grad_norm": 0.5131720900535583,
      "learning_rate": 2.769099248394903e-05,
      "loss": 0.4354,
      "mean_token_accuracy": 0.8560419797897338,
      "num_tokens": 586440540.0,
      "step": 17505
    },
    {
      "epoch": 1.045248328557784,
      "grad_norm": 0.623945415019989,
      "learning_rate": 2.7679888513671326e-05,
      "loss": 0.4667,
      "mean_token_accuracy": 0.8485029220581055,
      "num_tokens": 586608220.0,
      "step": 17510
    },
    {
      "epoch": 1.0455468003820438,
      "grad_norm": 0.49267029762268066,
      "learning_rate": 2.7668784499578455e-05,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8474233627319336,
      "num_tokens": 586775900.0,
      "step": 17515
    },
    {
      "epoch": 1.0458452722063036,
      "grad_norm": 0.587166428565979,
      "learning_rate": 2.7657680444375005e-05,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8388584017753601,
      "num_tokens": 586943580.0,
      "step": 17520
    },
    {
      "epoch": 1.0461437440305634,
      "grad_norm": 0.5883978009223938,
      "learning_rate": 2.7646576350765574e-05,
      "loss": 0.4841,
      "mean_token_accuracy": 0.8425503969192505,
      "num_tokens": 587111260.0,
      "step": 17525
    },
    {
      "epoch": 1.0464422158548232,
      "grad_norm": 0.5067137479782104,
      "learning_rate": 2.763547222145476e-05,
      "loss": 0.4581,
      "mean_token_accuracy": 0.8483001470565796,
      "num_tokens": 587278940.0,
      "step": 17530
    },
    {
      "epoch": 1.046740687679083,
      "grad_norm": 0.509401798248291,
      "learning_rate": 2.7624368059147188e-05,
      "loss": 0.4315,
      "mean_token_accuracy": 0.8547178864479065,
      "num_tokens": 587446620.0,
      "step": 17535
    },
    {
      "epoch": 1.0470391595033428,
      "grad_norm": 0.5064797401428223,
      "learning_rate": 2.7613263866547483e-05,
      "loss": 0.4243,
      "mean_token_accuracy": 0.8578074693679809,
      "num_tokens": 587614300.0,
      "step": 17540
    },
    {
      "epoch": 1.0473376313276026,
      "grad_norm": 0.5510956645011902,
      "learning_rate": 2.760215964636027e-05,
      "loss": 0.4865,
      "mean_token_accuracy": 0.8381546020507813,
      "num_tokens": 587781980.0,
      "step": 17545
    },
    {
      "epoch": 1.0476361031518624,
      "grad_norm": 0.45521649718284607,
      "learning_rate": 2.7591055401290195e-05,
      "loss": 0.4367,
      "mean_token_accuracy": 0.8536383271217346,
      "num_tokens": 587949660.0,
      "step": 17550
    },
    {
      "epoch": 1.0479345749761222,
      "grad_norm": 0.4933571517467499,
      "learning_rate": 2.7579951134041893e-05,
      "loss": 0.4357,
      "mean_token_accuracy": 0.856328284740448,
      "num_tokens": 588117340.0,
      "step": 17555
    },
    {
      "epoch": 1.048233046800382,
      "grad_norm": 0.5497868657112122,
      "learning_rate": 2.7568846847320012e-05,
      "loss": 0.4411,
      "mean_token_accuracy": 0.8524573564529419,
      "num_tokens": 588285020.0,
      "step": 17560
    },
    {
      "epoch": 1.0485315186246418,
      "grad_norm": 0.5534296035766602,
      "learning_rate": 2.7557742543829213e-05,
      "loss": 0.4775,
      "mean_token_accuracy": 0.8417750120162963,
      "num_tokens": 588452700.0,
      "step": 17565
    },
    {
      "epoch": 1.0488299904489016,
      "grad_norm": 0.5633777976036072,
      "learning_rate": 2.754663822627415e-05,
      "loss": 0.463,
      "mean_token_accuracy": 0.8476798295974731,
      "num_tokens": 588620380.0,
      "step": 17570
    },
    {
      "epoch": 1.0491284622731614,
      "grad_norm": 0.515777051448822,
      "learning_rate": 2.7535533897359496e-05,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8480734825134277,
      "num_tokens": 588788060.0,
      "step": 17575
    },
    {
      "epoch": 1.0494269340974212,
      "grad_norm": 0.5349860191345215,
      "learning_rate": 2.75244295597899e-05,
      "loss": 0.502,
      "mean_token_accuracy": 0.832756757736206,
      "num_tokens": 588955740.0,
      "step": 17580
    },
    {
      "epoch": 1.049725405921681,
      "grad_norm": 0.5449660420417786,
      "learning_rate": 2.7513325216270024e-05,
      "loss": 0.4646,
      "mean_token_accuracy": 0.8450733661651612,
      "num_tokens": 589123420.0,
      "step": 17585
    },
    {
      "epoch": 1.0500238777459407,
      "grad_norm": 0.5025606155395508,
      "learning_rate": 2.750222086950456e-05,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8448705673217773,
      "num_tokens": 589291100.0,
      "step": 17590
    },
    {
      "epoch": 1.0503223495702005,
      "grad_norm": 0.5384414792060852,
      "learning_rate": 2.7491116522198155e-05,
      "loss": 0.4612,
      "mean_token_accuracy": 0.846302044391632,
      "num_tokens": 589458780.0,
      "step": 17595
    },
    {
      "epoch": 1.0506208213944603,
      "grad_norm": 0.4843895435333252,
      "learning_rate": 2.7480012177055487e-05,
      "loss": 0.4364,
      "mean_token_accuracy": 0.8541870474815368,
      "num_tokens": 589626460.0,
      "step": 17600
    },
    {
      "epoch": 1.0509192932187201,
      "grad_norm": 0.5766077637672424,
      "learning_rate": 2.7468907836781222e-05,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8388882279396057,
      "num_tokens": 589794140.0,
      "step": 17605
    },
    {
      "epoch": 1.05121776504298,
      "grad_norm": 0.5743246674537659,
      "learning_rate": 2.7457803504080025e-05,
      "loss": 0.444,
      "mean_token_accuracy": 0.8532267689704895,
      "num_tokens": 589961820.0,
      "step": 17610
    },
    {
      "epoch": 1.0515162368672397,
      "grad_norm": 0.5840647220611572,
      "learning_rate": 2.7446699181656566e-05,
      "loss": 0.5097,
      "mean_token_accuracy": 0.8343552470207214,
      "num_tokens": 590129500.0,
      "step": 17615
    },
    {
      "epoch": 1.0518147086914995,
      "grad_norm": 0.5343279838562012,
      "learning_rate": 2.743559487221551e-05,
      "loss": 0.4998,
      "mean_token_accuracy": 0.8360849499702454,
      "num_tokens": 590297180.0,
      "step": 17620
    },
    {
      "epoch": 1.0521131805157593,
      "grad_norm": 0.5026448369026184,
      "learning_rate": 2.7424490578461504e-05,
      "loss": 0.4602,
      "mean_token_accuracy": 0.8483538150787353,
      "num_tokens": 590464860.0,
      "step": 17625
    },
    {
      "epoch": 1.052411652340019,
      "grad_norm": 0.48948389291763306,
      "learning_rate": 2.7413386303099203e-05,
      "loss": 0.4704,
      "mean_token_accuracy": 0.8451151132583619,
      "num_tokens": 590632540.0,
      "step": 17630
    },
    {
      "epoch": 1.052710124164279,
      "grad_norm": 0.5424418449401855,
      "learning_rate": 2.7402282048833283e-05,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8469819903373719,
      "num_tokens": 590800220.0,
      "step": 17635
    },
    {
      "epoch": 1.0530085959885387,
      "grad_norm": 0.45166853070259094,
      "learning_rate": 2.7391177818368363e-05,
      "loss": 0.4448,
      "mean_token_accuracy": 0.8528033018112182,
      "num_tokens": 590967900.0,
      "step": 17640
    },
    {
      "epoch": 1.0533070678127985,
      "grad_norm": 0.7576605081558228,
      "learning_rate": 2.7380073614409103e-05,
      "loss": 0.4892,
      "mean_token_accuracy": 0.8398209929466247,
      "num_tokens": 591134125.0,
      "step": 17645
    },
    {
      "epoch": 1.0536055396370583,
      "grad_norm": 0.5706770420074463,
      "learning_rate": 2.736896943966012e-05,
      "loss": 0.4563,
      "mean_token_accuracy": 0.850481903553009,
      "num_tokens": 591293542.0,
      "step": 17650
    },
    {
      "epoch": 1.053904011461318,
      "grad_norm": 0.5847463607788086,
      "learning_rate": 2.7357865296826053e-05,
      "loss": 0.496,
      "mean_token_accuracy": 0.8374627232551575,
      "num_tokens": 591461222.0,
      "step": 17655
    },
    {
      "epoch": 1.0542024832855779,
      "grad_norm": 0.5394697189331055,
      "learning_rate": 2.734676118861152e-05,
      "loss": 0.4372,
      "mean_token_accuracy": 0.8544315934181214,
      "num_tokens": 591628902.0,
      "step": 17660
    },
    {
      "epoch": 1.0545009551098377,
      "grad_norm": 0.5168535113334656,
      "learning_rate": 2.733565711772112e-05,
      "loss": 0.4469,
      "mean_token_accuracy": 0.8512823581695557,
      "num_tokens": 591796582.0,
      "step": 17665
    },
    {
      "epoch": 1.0547994269340975,
      "grad_norm": 0.5262973308563232,
      "learning_rate": 2.7324553086859465e-05,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8451747536659241,
      "num_tokens": 591964262.0,
      "step": 17670
    },
    {
      "epoch": 1.0550978987583572,
      "grad_norm": 0.4767979383468628,
      "learning_rate": 2.7313449098731142e-05,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8475486159324646,
      "num_tokens": 592131942.0,
      "step": 17675
    },
    {
      "epoch": 1.055396370582617,
      "grad_norm": 0.5223391056060791,
      "learning_rate": 2.7302345156040736e-05,
      "loss": 0.4856,
      "mean_token_accuracy": 0.8395502805709839,
      "num_tokens": 592299622.0,
      "step": 17680
    },
    {
      "epoch": 1.0556948424068768,
      "grad_norm": 0.5225414037704468,
      "learning_rate": 2.7291241261492823e-05,
      "loss": 0.4737,
      "mean_token_accuracy": 0.8428128242492676,
      "num_tokens": 592467302.0,
      "step": 17685
    },
    {
      "epoch": 1.0559933142311366,
      "grad_norm": 0.5096659064292908,
      "learning_rate": 2.728013741779194e-05,
      "loss": 0.4575,
      "mean_token_accuracy": 0.8494088411331177,
      "num_tokens": 592633679.0,
      "step": 17690
    },
    {
      "epoch": 1.0562917860553964,
      "grad_norm": 0.5200692415237427,
      "learning_rate": 2.7269033627642665e-05,
      "loss": 0.4871,
      "mean_token_accuracy": 0.8391387343406678,
      "num_tokens": 592801359.0,
      "step": 17695
    },
    {
      "epoch": 1.0565902578796562,
      "grad_norm": 0.5188535451889038,
      "learning_rate": 2.7257929893749496e-05,
      "loss": 0.449,
      "mean_token_accuracy": 0.8516700625419616,
      "num_tokens": 592969039.0,
      "step": 17700
    },
    {
      "epoch": 1.056888729703916,
      "grad_norm": 0.503955602645874,
      "learning_rate": 2.7246826218816985e-05,
      "loss": 0.4496,
      "mean_token_accuracy": 0.8505427718162537,
      "num_tokens": 593136719.0,
      "step": 17705
    },
    {
      "epoch": 1.0571872015281758,
      "grad_norm": 0.46807822585105896,
      "learning_rate": 2.72357226055496e-05,
      "loss": 0.4367,
      "mean_token_accuracy": 0.8553262591362,
      "num_tokens": 593304399.0,
      "step": 17710
    },
    {
      "epoch": 1.0574856733524356,
      "grad_norm": 0.5091184377670288,
      "learning_rate": 2.7224619056651866e-05,
      "loss": 0.4357,
      "mean_token_accuracy": 0.854723846912384,
      "num_tokens": 593472079.0,
      "step": 17715
    },
    {
      "epoch": 1.0577841451766954,
      "grad_norm": 0.5885462164878845,
      "learning_rate": 2.7213515574828248e-05,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8463080048561096,
      "num_tokens": 593639759.0,
      "step": 17720
    },
    {
      "epoch": 1.0580826170009552,
      "grad_norm": 0.47241973876953125,
      "learning_rate": 2.7202412162783193e-05,
      "loss": 0.4412,
      "mean_token_accuracy": 0.852731716632843,
      "num_tokens": 593807439.0,
      "step": 17725
    },
    {
      "epoch": 1.058381088825215,
      "grad_norm": 0.49074244499206543,
      "learning_rate": 2.7191308823221157e-05,
      "loss": 0.4384,
      "mean_token_accuracy": 0.8535846471786499,
      "num_tokens": 593975119.0,
      "step": 17730
    },
    {
      "epoch": 1.0586795606494748,
      "grad_norm": 0.5047051310539246,
      "learning_rate": 2.7180205558846543e-05,
      "loss": 0.4495,
      "mean_token_accuracy": 0.849988067150116,
      "num_tokens": 594142799.0,
      "step": 17735
    },
    {
      "epoch": 1.0589780324737346,
      "grad_norm": 0.5465596914291382,
      "learning_rate": 2.716910237236377e-05,
      "loss": 0.436,
      "mean_token_accuracy": 0.8557616591453552,
      "num_tokens": 594310479.0,
      "step": 17740
    },
    {
      "epoch": 1.0592765042979944,
      "grad_norm": 0.48841941356658936,
      "learning_rate": 2.715799926647723e-05,
      "loss": 0.4258,
      "mean_token_accuracy": 0.8571812033653259,
      "num_tokens": 594478159.0,
      "step": 17745
    },
    {
      "epoch": 1.0595749761222542,
      "grad_norm": 0.4936128258705139,
      "learning_rate": 2.7146896243891267e-05,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8527555823326111,
      "num_tokens": 594645839.0,
      "step": 17750
    },
    {
      "epoch": 1.059873447946514,
      "grad_norm": 0.5698809623718262,
      "learning_rate": 2.7135793307310258e-05,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8386377096176147,
      "num_tokens": 594813519.0,
      "step": 17755
    },
    {
      "epoch": 1.0601719197707737,
      "grad_norm": 0.5826289057731628,
      "learning_rate": 2.7124690459438494e-05,
      "loss": 0.4537,
      "mean_token_accuracy": 0.8490039467811584,
      "num_tokens": 594981199.0,
      "step": 17760
    },
    {
      "epoch": 1.0604703915950335,
      "grad_norm": 0.47566941380500793,
      "learning_rate": 2.711358770298031e-05,
      "loss": 0.4743,
      "mean_token_accuracy": 0.8447130441665649,
      "num_tokens": 595142018.0,
      "step": 17765
    },
    {
      "epoch": 1.0607688634192933,
      "grad_norm": 0.5589858293533325,
      "learning_rate": 2.7102485040639976e-05,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8450912356376648,
      "num_tokens": 595309698.0,
      "step": 17770
    },
    {
      "epoch": 1.061067335243553,
      "grad_norm": 0.4675428569316864,
      "learning_rate": 2.7091382475121736e-05,
      "loss": 0.4792,
      "mean_token_accuracy": 0.8422939300537109,
      "num_tokens": 595477378.0,
      "step": 17775
    },
    {
      "epoch": 1.0613658070678127,
      "grad_norm": 0.49825581908226013,
      "learning_rate": 2.7080280009129844e-05,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8545091271400451,
      "num_tokens": 595645058.0,
      "step": 17780
    },
    {
      "epoch": 1.0616642788920725,
      "grad_norm": 0.5521115660667419,
      "learning_rate": 2.70691776453685e-05,
      "loss": 0.4608,
      "mean_token_accuracy": 0.8469760179519653,
      "num_tokens": 595812738.0,
      "step": 17785
    },
    {
      "epoch": 1.0619627507163323,
      "grad_norm": 0.5025843381881714,
      "learning_rate": 2.7058075386541886e-05,
      "loss": 0.4446,
      "mean_token_accuracy": 0.8522426366806031,
      "num_tokens": 595980418.0,
      "step": 17790
    },
    {
      "epoch": 1.062261222540592,
      "grad_norm": 0.5706450343132019,
      "learning_rate": 2.704697323535418e-05,
      "loss": 0.4712,
      "mean_token_accuracy": 0.8440653681755066,
      "num_tokens": 596148098.0,
      "step": 17795
    },
    {
      "epoch": 1.0625596943648519,
      "grad_norm": 0.6271229982376099,
      "learning_rate": 2.70358711945095e-05,
      "loss": 0.4555,
      "mean_token_accuracy": 0.8511213183403015,
      "num_tokens": 596315778.0,
      "step": 17800
    },
    {
      "epoch": 1.0628581661891117,
      "grad_norm": 0.47752127051353455,
      "learning_rate": 2.7024769266711957e-05,
      "loss": 0.4231,
      "mean_token_accuracy": 0.8593940138816833,
      "num_tokens": 596483458.0,
      "step": 17805
    },
    {
      "epoch": 1.0631566380133715,
      "grad_norm": 0.501616895198822,
      "learning_rate": 2.701366745466563e-05,
      "loss": 0.4503,
      "mean_token_accuracy": 0.8496540665626526,
      "num_tokens": 596651138.0,
      "step": 17810
    },
    {
      "epoch": 1.0634551098376313,
      "grad_norm": 0.5748869180679321,
      "learning_rate": 2.700256576107458e-05,
      "loss": 0.4503,
      "mean_token_accuracy": 0.8501431465148925,
      "num_tokens": 596818818.0,
      "step": 17815
    },
    {
      "epoch": 1.063753581661891,
      "grad_norm": 0.5291325449943542,
      "learning_rate": 2.6991464188642808e-05,
      "loss": 0.4477,
      "mean_token_accuracy": 0.8504711985588074,
      "num_tokens": 596986498.0,
      "step": 17820
    },
    {
      "epoch": 1.0640520534861508,
      "grad_norm": 0.4641845226287842,
      "learning_rate": 2.6980362740074315e-05,
      "loss": 0.4153,
      "mean_token_accuracy": 0.8613503575325012,
      "num_tokens": 597154178.0,
      "step": 17825
    },
    {
      "epoch": 1.0643505253104106,
      "grad_norm": 0.514773428440094,
      "learning_rate": 2.6969261418073087e-05,
      "loss": 0.4575,
      "mean_token_accuracy": 0.8479005217552185,
      "num_tokens": 597321858.0,
      "step": 17830
    },
    {
      "epoch": 1.0646489971346704,
      "grad_norm": 0.48946326971054077,
      "learning_rate": 2.6958160225343022e-05,
      "loss": 0.4837,
      "mean_token_accuracy": 0.839836585521698,
      "num_tokens": 597489538.0,
      "step": 17835
    },
    {
      "epoch": 1.0649474689589302,
      "grad_norm": 0.5289530754089355,
      "learning_rate": 2.6947059164588045e-05,
      "loss": 0.4155,
      "mean_token_accuracy": 0.8611475467681885,
      "num_tokens": 597657218.0,
      "step": 17840
    },
    {
      "epoch": 1.06524594078319,
      "grad_norm": 0.4798479974269867,
      "learning_rate": 2.6935958238512006e-05,
      "loss": 0.4313,
      "mean_token_accuracy": 0.855827271938324,
      "num_tokens": 597824898.0,
      "step": 17845
    },
    {
      "epoch": 1.0655444126074498,
      "grad_norm": 0.4973350465297699,
      "learning_rate": 2.6924857449818754e-05,
      "loss": 0.4733,
      "mean_token_accuracy": 0.8430156230926513,
      "num_tokens": 597992578.0,
      "step": 17850
    },
    {
      "epoch": 1.0658428844317096,
      "grad_norm": 0.5921276807785034,
      "learning_rate": 2.6913756801212086e-05,
      "loss": 0.4491,
      "mean_token_accuracy": 0.8509006381034852,
      "num_tokens": 598160258.0,
      "step": 17855
    },
    {
      "epoch": 1.0661413562559694,
      "grad_norm": 0.5628989338874817,
      "learning_rate": 2.6902656295395766e-05,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8491053342819214,
      "num_tokens": 598327938.0,
      "step": 17860
    },
    {
      "epoch": 1.0664398280802292,
      "grad_norm": 0.545533299446106,
      "learning_rate": 2.6891555935073537e-05,
      "loss": 0.4468,
      "mean_token_accuracy": 0.85284503698349,
      "num_tokens": 598495618.0,
      "step": 17865
    },
    {
      "epoch": 1.066738299904489,
      "grad_norm": 0.5277979373931885,
      "learning_rate": 2.6880455722949083e-05,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8540200352668762,
      "num_tokens": 598663298.0,
      "step": 17870
    },
    {
      "epoch": 1.0670367717287488,
      "grad_norm": 0.5081989169120789,
      "learning_rate": 2.686935566172607e-05,
      "loss": 0.4442,
      "mean_token_accuracy": 0.8515626788139343,
      "num_tokens": 598830978.0,
      "step": 17875
    },
    {
      "epoch": 1.0673352435530086,
      "grad_norm": 0.4833400249481201,
      "learning_rate": 2.6858255754108148e-05,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8425324916839599,
      "num_tokens": 598998658.0,
      "step": 17880
    },
    {
      "epoch": 1.0676337153772684,
      "grad_norm": 0.5343855023384094,
      "learning_rate": 2.6847156002798857e-05,
      "loss": 0.455,
      "mean_token_accuracy": 0.8490456938743591,
      "num_tokens": 599166338.0,
      "step": 17885
    },
    {
      "epoch": 1.0679321872015282,
      "grad_norm": 0.5258563160896301,
      "learning_rate": 2.683605641050179e-05,
      "loss": 0.4827,
      "mean_token_accuracy": 0.8414052128791809,
      "num_tokens": 599334018.0,
      "step": 17890
    },
    {
      "epoch": 1.068230659025788,
      "grad_norm": 0.4840978682041168,
      "learning_rate": 2.6824956979920424e-05,
      "loss": 0.4328,
      "mean_token_accuracy": 0.8547417402267456,
      "num_tokens": 599501698.0,
      "step": 17895
    },
    {
      "epoch": 1.0685291308500477,
      "grad_norm": 0.5214992761611938,
      "learning_rate": 2.681385771375825e-05,
      "loss": 0.4832,
      "mean_token_accuracy": 0.8407670259475708,
      "num_tokens": 599669378.0,
      "step": 17900
    },
    {
      "epoch": 1.0688276026743075,
      "grad_norm": 0.5052345991134644,
      "learning_rate": 2.6802758614718703e-05,
      "loss": 0.491,
      "mean_token_accuracy": 0.8385064959526062,
      "num_tokens": 599837058.0,
      "step": 17905
    },
    {
      "epoch": 1.0691260744985673,
      "grad_norm": 0.46404924988746643,
      "learning_rate": 2.679165968550516e-05,
      "loss": 0.432,
      "mean_token_accuracy": 0.8560240983963012,
      "num_tokens": 600004738.0,
      "step": 17910
    },
    {
      "epoch": 1.0694245463228271,
      "grad_norm": 0.5729808807373047,
      "learning_rate": 2.6780560928820974e-05,
      "loss": 0.4618,
      "mean_token_accuracy": 0.8469521641731262,
      "num_tokens": 600172418.0,
      "step": 17915
    },
    {
      "epoch": 1.069723018147087,
      "grad_norm": 0.5375544428825378,
      "learning_rate": 2.6769462347369456e-05,
      "loss": 0.4788,
      "mean_token_accuracy": 0.842067289352417,
      "num_tokens": 600340098.0,
      "step": 17920
    },
    {
      "epoch": 1.0700214899713467,
      "grad_norm": 0.5067377090454102,
      "learning_rate": 2.6758363943853865e-05,
      "loss": 0.4592,
      "mean_token_accuracy": 0.8473219394683837,
      "num_tokens": 600507778.0,
      "step": 17925
    },
    {
      "epoch": 1.0703199617956065,
      "grad_norm": 0.5274377465248108,
      "learning_rate": 2.674726572097742e-05,
      "loss": 0.4669,
      "mean_token_accuracy": 0.8452940464019776,
      "num_tokens": 600675458.0,
      "step": 17930
    },
    {
      "epoch": 1.0706184336198663,
      "grad_norm": 0.5200303792953491,
      "learning_rate": 2.6736167681443308e-05,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8351008057594299,
      "num_tokens": 600843138.0,
      "step": 17935
    },
    {
      "epoch": 1.070916905444126,
      "grad_norm": 0.5001233220100403,
      "learning_rate": 2.6725069827954663e-05,
      "loss": 0.4535,
      "mean_token_accuracy": 0.8501670241355896,
      "num_tokens": 601010818.0,
      "step": 17940
    },
    {
      "epoch": 1.071215377268386,
      "grad_norm": 0.6038535237312317,
      "learning_rate": 2.671397216321455e-05,
      "loss": 0.4644,
      "mean_token_accuracy": 0.8464929103851319,
      "num_tokens": 601178498.0,
      "step": 17945
    },
    {
      "epoch": 1.0715138490926457,
      "grad_norm": 0.4926758408546448,
      "learning_rate": 2.670287468992604e-05,
      "loss": 0.4407,
      "mean_token_accuracy": 0.8539067149162293,
      "num_tokens": 601346178.0,
      "step": 17950
    },
    {
      "epoch": 1.0718123209169055,
      "grad_norm": 0.48250171542167664,
      "learning_rate": 2.6691777410792102e-05,
      "loss": 0.4453,
      "mean_token_accuracy": 0.8525050640106201,
      "num_tokens": 601513858.0,
      "step": 17955
    },
    {
      "epoch": 1.0721107927411653,
      "grad_norm": 0.4897820055484772,
      "learning_rate": 2.668068032851569e-05,
      "loss": 0.4679,
      "mean_token_accuracy": 0.8459441781044006,
      "num_tokens": 601681538.0,
      "step": 17960
    },
    {
      "epoch": 1.072409264565425,
      "grad_norm": 0.5335580110549927,
      "learning_rate": 2.666958344579972e-05,
      "loss": 0.4112,
      "mean_token_accuracy": 0.8627162218093872,
      "num_tokens": 601849218.0,
      "step": 17965
    },
    {
      "epoch": 1.0727077363896849,
      "grad_norm": 0.5549975633621216,
      "learning_rate": 2.665848676534702e-05,
      "loss": 0.4692,
      "mean_token_accuracy": 0.8447095274925231,
      "num_tokens": 602016898.0,
      "step": 17970
    },
    {
      "epoch": 1.0730062082139447,
      "grad_norm": 0.4920300543308258,
      "learning_rate": 2.66473902898604e-05,
      "loss": 0.4521,
      "mean_token_accuracy": 0.8501133322715759,
      "num_tokens": 602184578.0,
      "step": 17975
    },
    {
      "epoch": 1.0733046800382045,
      "grad_norm": 0.47080594301223755,
      "learning_rate": 2.6636294022042618e-05,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8427054762840271,
      "num_tokens": 602352258.0,
      "step": 17980
    },
    {
      "epoch": 1.0736031518624642,
      "grad_norm": 0.5335698127746582,
      "learning_rate": 2.6625197964596366e-05,
      "loss": 0.4543,
      "mean_token_accuracy": 0.8492305755615235,
      "num_tokens": 602519938.0,
      "step": 17985
    },
    {
      "epoch": 1.073901623686724,
      "grad_norm": 0.49977582693099976,
      "learning_rate": 2.6614102120224298e-05,
      "loss": 0.4838,
      "mean_token_accuracy": 0.838834536075592,
      "num_tokens": 602687618.0,
      "step": 17990
    },
    {
      "epoch": 1.0742000955109838,
      "grad_norm": 0.5056637525558472,
      "learning_rate": 2.660300649162901e-05,
      "loss": 0.4109,
      "mean_token_accuracy": 0.8633066892623902,
      "num_tokens": 602855298.0,
      "step": 17995
    },
    {
      "epoch": 1.0744985673352436,
      "grad_norm": 0.4958396553993225,
      "learning_rate": 2.659191108151305e-05,
      "loss": 0.4396,
      "mean_token_accuracy": 0.8530060887336731,
      "num_tokens": 603022978.0,
      "step": 18000
    },
    {
      "epoch": 1.0747970391595034,
      "grad_norm": 0.4855215847492218,
      "learning_rate": 2.6580815892578904e-05,
      "loss": 0.4452,
      "mean_token_accuracy": 0.8541870594024659,
      "num_tokens": 603190658.0,
      "step": 18005
    },
    {
      "epoch": 1.0750955109837632,
      "grad_norm": 0.5160524845123291,
      "learning_rate": 2.6569720927529018e-05,
      "loss": 0.4328,
      "mean_token_accuracy": 0.856065857410431,
      "num_tokens": 603358338.0,
      "step": 18010
    },
    {
      "epoch": 1.075393982808023,
      "grad_norm": 0.5940030813217163,
      "learning_rate": 2.6558626189065765e-05,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8556662321090698,
      "num_tokens": 603526018.0,
      "step": 18015
    },
    {
      "epoch": 1.0756924546322828,
      "grad_norm": 0.48193079233169556,
      "learning_rate": 2.654753167989148e-05,
      "loss": 0.4569,
      "mean_token_accuracy": 0.84797123670578,
      "num_tokens": 603692972.0,
      "step": 18020
    },
    {
      "epoch": 1.0759909264565426,
      "grad_norm": 0.5851207375526428,
      "learning_rate": 2.6536437402708442e-05,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8444053530693054,
      "num_tokens": 603860652.0,
      "step": 18025
    },
    {
      "epoch": 1.0762893982808024,
      "grad_norm": 0.5075387358665466,
      "learning_rate": 2.6525343360218853e-05,
      "loss": 0.4432,
      "mean_token_accuracy": 0.8525706887245178,
      "num_tokens": 604028332.0,
      "step": 18030
    },
    {
      "epoch": 1.0765878701050622,
      "grad_norm": 0.47441422939300537,
      "learning_rate": 2.6514249555124887e-05,
      "loss": 0.4495,
      "mean_token_accuracy": 0.8499045729637146,
      "num_tokens": 604196012.0,
      "step": 18035
    },
    {
      "epoch": 1.0768863419293218,
      "grad_norm": 0.46453920006752014,
      "learning_rate": 2.650315599012862e-05,
      "loss": 0.427,
      "mean_token_accuracy": 0.856727910041809,
      "num_tokens": 604363692.0,
      "step": 18040
    },
    {
      "epoch": 1.0771848137535818,
      "grad_norm": 0.4999887943267822,
      "learning_rate": 2.6492062667932123e-05,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8486758947372437,
      "num_tokens": 604531372.0,
      "step": 18045
    },
    {
      "epoch": 1.0774832855778413,
      "grad_norm": 0.5404815077781677,
      "learning_rate": 2.648096959123737e-05,
      "loss": 0.4626,
      "mean_token_accuracy": 0.8473219513893128,
      "num_tokens": 604699052.0,
      "step": 18050
    },
    {
      "epoch": 1.0777817574021014,
      "grad_norm": 0.5161058306694031,
      "learning_rate": 2.6469876762746277e-05,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8495407462120056,
      "num_tokens": 604866732.0,
      "step": 18055
    },
    {
      "epoch": 1.078080229226361,
      "grad_norm": 0.48695194721221924,
      "learning_rate": 2.645878418516072e-05,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8346176862716674,
      "num_tokens": 605034412.0,
      "step": 18060
    },
    {
      "epoch": 1.0783787010506207,
      "grad_norm": 0.5438864827156067,
      "learning_rate": 2.6447691861182483e-05,
      "loss": 0.473,
      "mean_token_accuracy": 0.8427651166915894,
      "num_tokens": 605202092.0,
      "step": 18065
    },
    {
      "epoch": 1.0786771728748805,
      "grad_norm": 0.5088119506835938,
      "learning_rate": 2.6436599793513317e-05,
      "loss": 0.4621,
      "mean_token_accuracy": 0.8465764045715332,
      "num_tokens": 605369772.0,
      "step": 18070
    },
    {
      "epoch": 1.0789756446991403,
      "grad_norm": 0.5458279252052307,
      "learning_rate": 2.6425507984854907e-05,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8333353161811828,
      "num_tokens": 605537452.0,
      "step": 18075
    },
    {
      "epoch": 1.0792741165234,
      "grad_norm": 0.4567376971244812,
      "learning_rate": 2.6414416437908857e-05,
      "loss": 0.4539,
      "mean_token_accuracy": 0.8499701857566834,
      "num_tokens": 605705132.0,
      "step": 18080
    },
    {
      "epoch": 1.07957258834766,
      "grad_norm": 0.5640619993209839,
      "learning_rate": 2.640332515537672e-05,
      "loss": 0.4762,
      "mean_token_accuracy": 0.842544436454773,
      "num_tokens": 605872812.0,
      "step": 18085
    },
    {
      "epoch": 1.0798710601719197,
      "grad_norm": 0.48929524421691895,
      "learning_rate": 2.6392234139959988e-05,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8481092691421509,
      "num_tokens": 606040492.0,
      "step": 18090
    },
    {
      "epoch": 1.0801695319961795,
      "grad_norm": 0.5491763353347778,
      "learning_rate": 2.6381143394360063e-05,
      "loss": 0.4443,
      "mean_token_accuracy": 0.8519145965576171,
      "num_tokens": 606208172.0,
      "step": 18095
    },
    {
      "epoch": 1.0804680038204393,
      "grad_norm": 0.5325464010238647,
      "learning_rate": 2.637005292127833e-05,
      "loss": 0.4418,
      "mean_token_accuracy": 0.8536383032798767,
      "num_tokens": 606375852.0,
      "step": 18100
    },
    {
      "epoch": 1.080766475644699,
      "grad_norm": 0.4878290593624115,
      "learning_rate": 2.6358962723416063e-05,
      "loss": 0.4244,
      "mean_token_accuracy": 0.8593820810317994,
      "num_tokens": 606543532.0,
      "step": 18105
    },
    {
      "epoch": 1.0810649474689589,
      "grad_norm": 0.46349263191223145,
      "learning_rate": 2.6347872803474487e-05,
      "loss": 0.4709,
      "mean_token_accuracy": 0.8423237442970276,
      "num_tokens": 606711212.0,
      "step": 18110
    },
    {
      "epoch": 1.0813634192932187,
      "grad_norm": 0.5191339254379272,
      "learning_rate": 2.6336783164154753e-05,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8524752497673035,
      "num_tokens": 606878892.0,
      "step": 18115
    },
    {
      "epoch": 1.0816618911174785,
      "grad_norm": 0.4806431233882904,
      "learning_rate": 2.6325693808157952e-05,
      "loss": 0.4457,
      "mean_token_accuracy": 0.8508051991462707,
      "num_tokens": 607046572.0,
      "step": 18120
    },
    {
      "epoch": 1.0819603629417383,
      "grad_norm": 0.498297780752182,
      "learning_rate": 2.6314604738185095e-05,
      "loss": 0.443,
      "mean_token_accuracy": 0.8525647163391114,
      "num_tokens": 607214252.0,
      "step": 18125
    },
    {
      "epoch": 1.082258834765998,
      "grad_norm": 0.5350227952003479,
      "learning_rate": 2.6303515956937137e-05,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8453835010528564,
      "num_tokens": 607381932.0,
      "step": 18130
    },
    {
      "epoch": 1.0825573065902578,
      "grad_norm": 0.4777717888355255,
      "learning_rate": 2.6292427467114957e-05,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8528271555900574,
      "num_tokens": 607549612.0,
      "step": 18135
    },
    {
      "epoch": 1.0828557784145176,
      "grad_norm": 0.5520328879356384,
      "learning_rate": 2.6281339271419347e-05,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8443755269050598,
      "num_tokens": 607717292.0,
      "step": 18140
    },
    {
      "epoch": 1.0831542502387774,
      "grad_norm": 0.4736959636211395,
      "learning_rate": 2.6270251372551058e-05,
      "loss": 0.4328,
      "mean_token_accuracy": 0.855189073085785,
      "num_tokens": 607884972.0,
      "step": 18145
    },
    {
      "epoch": 1.0834527220630372,
      "grad_norm": 0.6132194995880127,
      "learning_rate": 2.6259163773210743e-05,
      "loss": 0.4627,
      "mean_token_accuracy": 0.8478289365768432,
      "num_tokens": 608052652.0,
      "step": 18150
    },
    {
      "epoch": 1.083751193887297,
      "grad_norm": 0.5220558643341064,
      "learning_rate": 2.6248076476098986e-05,
      "loss": 0.4661,
      "mean_token_accuracy": 0.8455803275108338,
      "num_tokens": 608220332.0,
      "step": 18155
    },
    {
      "epoch": 1.0840496657115568,
      "grad_norm": 0.5272908210754395,
      "learning_rate": 2.623698948391633e-05,
      "loss": 0.4794,
      "mean_token_accuracy": 0.841291892528534,
      "num_tokens": 608388012.0,
      "step": 18160
    },
    {
      "epoch": 1.0843481375358166,
      "grad_norm": 0.629886269569397,
      "learning_rate": 2.6225902799363178e-05,
      "loss": 0.498,
      "mean_token_accuracy": 0.8378742575645447,
      "num_tokens": 608555692.0,
      "step": 18165
    },
    {
      "epoch": 1.0846466093600764,
      "grad_norm": 0.49738672375679016,
      "learning_rate": 2.6214816425139933e-05,
      "loss": 0.4463,
      "mean_token_accuracy": 0.8533043026924133,
      "num_tokens": 608723372.0,
      "step": 18170
    },
    {
      "epoch": 1.0849450811843362,
      "grad_norm": 0.5750638842582703,
      "learning_rate": 2.6203730363946855e-05,
      "loss": 0.4622,
      "mean_token_accuracy": 0.8458427786827087,
      "num_tokens": 608891052.0,
      "step": 18175
    },
    {
      "epoch": 1.085243553008596,
      "grad_norm": 0.5631683468818665,
      "learning_rate": 2.619264461848418e-05,
      "loss": 0.5008,
      "mean_token_accuracy": 0.834385073184967,
      "num_tokens": 609058732.0,
      "step": 18180
    },
    {
      "epoch": 1.0855420248328558,
      "grad_norm": 0.5108806490898132,
      "learning_rate": 2.6181559191452043e-05,
      "loss": 0.4356,
      "mean_token_accuracy": 0.854461419582367,
      "num_tokens": 609226412.0,
      "step": 18185
    },
    {
      "epoch": 1.0858404966571156,
      "grad_norm": 0.5062009692192078,
      "learning_rate": 2.6170474085550496e-05,
      "loss": 0.4404,
      "mean_token_accuracy": 0.8520398378372193,
      "num_tokens": 609394092.0,
      "step": 18190
    },
    {
      "epoch": 1.0861389684813754,
      "grad_norm": 0.4719405770301819,
      "learning_rate": 2.615938930347953e-05,
      "loss": 0.4274,
      "mean_token_accuracy": 0.8566384315490723,
      "num_tokens": 609561772.0,
      "step": 18195
    },
    {
      "epoch": 1.0864374403056352,
      "grad_norm": 0.555225133895874,
      "learning_rate": 2.614830484793904e-05,
      "loss": 0.4599,
      "mean_token_accuracy": 0.8467851638793945,
      "num_tokens": 609729452.0,
      "step": 18200
    },
    {
      "epoch": 1.086735912129895,
      "grad_norm": 0.5339514017105103,
      "learning_rate": 2.613722072162885e-05,
      "loss": 0.4784,
      "mean_token_accuracy": 0.8430871963500977,
      "num_tokens": 609897132.0,
      "step": 18205
    },
    {
      "epoch": 1.0870343839541547,
      "grad_norm": 0.5332335233688354,
      "learning_rate": 2.6126136927248718e-05,
      "loss": 0.4669,
      "mean_token_accuracy": 0.8456220984458923,
      "num_tokens": 610064812.0,
      "step": 18210
    },
    {
      "epoch": 1.0873328557784145,
      "grad_norm": 0.5602828860282898,
      "learning_rate": 2.6115053467498285e-05,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8415722250938416,
      "num_tokens": 610232492.0,
      "step": 18215
    },
    {
      "epoch": 1.0876313276026743,
      "grad_norm": 0.48973003029823303,
      "learning_rate": 2.6103970345077154e-05,
      "loss": 0.4307,
      "mean_token_accuracy": 0.8565191507339478,
      "num_tokens": 610400172.0,
      "step": 18220
    },
    {
      "epoch": 1.0879297994269341,
      "grad_norm": 0.5911539196968079,
      "learning_rate": 2.6092887562684803e-05,
      "loss": 0.4706,
      "mean_token_accuracy": 0.8445567727088928,
      "num_tokens": 610566566.0,
      "step": 18225
    },
    {
      "epoch": 1.088228271251194,
      "grad_norm": 0.5489808917045593,
      "learning_rate": 2.608180512302067e-05,
      "loss": 0.4705,
      "mean_token_accuracy": 0.8445007801055908,
      "num_tokens": 610734246.0,
      "step": 18230
    },
    {
      "epoch": 1.0885267430754537,
      "grad_norm": 0.5308067798614502,
      "learning_rate": 2.607072302878406e-05,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8454849004745484,
      "num_tokens": 610901926.0,
      "step": 18235
    },
    {
      "epoch": 1.0888252148997135,
      "grad_norm": 0.5792397260665894,
      "learning_rate": 2.6059641282674247e-05,
      "loss": 0.4713,
      "mean_token_accuracy": 0.8451688051223755,
      "num_tokens": 611069606.0,
      "step": 18240
    },
    {
      "epoch": 1.0891236867239733,
      "grad_norm": 0.47498780488967896,
      "learning_rate": 2.604855988739039e-05,
      "loss": 0.4616,
      "mean_token_accuracy": 0.8469521760940552,
      "num_tokens": 611237286.0,
      "step": 18245
    },
    {
      "epoch": 1.089422158548233,
      "grad_norm": 0.5143116116523743,
      "learning_rate": 2.603747884563155e-05,
      "loss": 0.4255,
      "mean_token_accuracy": 0.8578432679176331,
      "num_tokens": 611404966.0,
      "step": 18250
    },
    {
      "epoch": 1.089720630372493,
      "grad_norm": 1.8472613096237183,
      "learning_rate": 2.6026398160096742e-05,
      "loss": 0.445,
      "mean_token_accuracy": 0.8542347550392151,
      "num_tokens": 611572646.0,
      "step": 18255
    },
    {
      "epoch": 1.0900191021967527,
      "grad_norm": 0.5715947151184082,
      "learning_rate": 2.6015317833484863e-05,
      "loss": 0.4933,
      "mean_token_accuracy": 0.8368841648101807,
      "num_tokens": 611740326.0,
      "step": 18260
    },
    {
      "epoch": 1.0903175740210125,
      "grad_norm": 0.5168896317481995,
      "learning_rate": 2.6004237868494725e-05,
      "loss": 0.4454,
      "mean_token_accuracy": 0.8511213183403015,
      "num_tokens": 611908006.0,
      "step": 18265
    },
    {
      "epoch": 1.0906160458452723,
      "grad_norm": 0.5258520245552063,
      "learning_rate": 2.5993158267825068e-05,
      "loss": 0.4661,
      "mean_token_accuracy": 0.8458487391471863,
      "num_tokens": 612075686.0,
      "step": 18270
    },
    {
      "epoch": 1.090914517669532,
      "grad_norm": 0.5208800435066223,
      "learning_rate": 2.5982079034174522e-05,
      "loss": 0.4692,
      "mean_token_accuracy": 0.8447512745857239,
      "num_tokens": 612243366.0,
      "step": 18275
    },
    {
      "epoch": 1.0912129894937919,
      "grad_norm": 0.5926040410995483,
      "learning_rate": 2.5971000170241648e-05,
      "loss": 0.4338,
      "mean_token_accuracy": 0.8565012454986572,
      "num_tokens": 612411046.0,
      "step": 18280
    },
    {
      "epoch": 1.0915114613180517,
      "grad_norm": 0.5341820120811462,
      "learning_rate": 2.59599216787249e-05,
      "loss": 0.4312,
      "mean_token_accuracy": 0.8560837507247925,
      "num_tokens": 612578726.0,
      "step": 18285
    },
    {
      "epoch": 1.0918099331423115,
      "grad_norm": 0.5410572290420532,
      "learning_rate": 2.5948843562322655e-05,
      "loss": 0.429,
      "mean_token_accuracy": 0.8570141911506652,
      "num_tokens": 612746406.0,
      "step": 18290
    },
    {
      "epoch": 1.0921084049665712,
      "grad_norm": 0.5297465920448303,
      "learning_rate": 2.59377658237332e-05,
      "loss": 0.4638,
      "mean_token_accuracy": 0.8459386944770813,
      "num_tokens": 612908700.0,
      "step": 18295
    },
    {
      "epoch": 1.092406876790831,
      "grad_norm": 0.5624565482139587,
      "learning_rate": 2.5926688465654703e-05,
      "loss": 0.4409,
      "mean_token_accuracy": 0.8532744884490967,
      "num_tokens": 613076380.0,
      "step": 18300
    },
    {
      "epoch": 1.0927053486150908,
      "grad_norm": 0.5259370803833008,
      "learning_rate": 2.5915611490785286e-05,
      "loss": 0.4725,
      "mean_token_accuracy": 0.8425265431404114,
      "num_tokens": 613244060.0,
      "step": 18305
    },
    {
      "epoch": 1.0930038204393506,
      "grad_norm": 0.4987262487411499,
      "learning_rate": 2.590453490182292e-05,
      "loss": 0.434,
      "mean_token_accuracy": 0.8563044309616089,
      "num_tokens": 613411740.0,
      "step": 18310
    },
    {
      "epoch": 1.0933022922636102,
      "grad_norm": 0.4512346088886261,
      "learning_rate": 2.5893458701465528e-05,
      "loss": 0.3879,
      "mean_token_accuracy": 0.8685136556625366,
      "num_tokens": 613579420.0,
      "step": 18315
    },
    {
      "epoch": 1.0936007640878702,
      "grad_norm": 0.5192257761955261,
      "learning_rate": 2.5882382892410927e-05,
      "loss": 0.4245,
      "mean_token_accuracy": 0.8574973344802856,
      "num_tokens": 613747100.0,
      "step": 18320
    },
    {
      "epoch": 1.0938992359121298,
      "grad_norm": 0.5196990966796875,
      "learning_rate": 2.5871307477356832e-05,
      "loss": 0.4422,
      "mean_token_accuracy": 0.8525169968605042,
      "num_tokens": 613914780.0,
      "step": 18325
    },
    {
      "epoch": 1.0941977077363898,
      "grad_norm": 0.5397512912750244,
      "learning_rate": 2.586023245900086e-05,
      "loss": 0.4607,
      "mean_token_accuracy": 0.8479601621627808,
      "num_tokens": 614082460.0,
      "step": 18330
    },
    {
      "epoch": 1.0944961795606494,
      "grad_norm": 0.5651199221611023,
      "learning_rate": 2.584915784004054e-05,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8499284267425538,
      "num_tokens": 614250140.0,
      "step": 18335
    },
    {
      "epoch": 1.0947946513849092,
      "grad_norm": 0.514983057975769,
      "learning_rate": 2.5838083623173303e-05,
      "loss": 0.4822,
      "mean_token_accuracy": 0.842204463481903,
      "num_tokens": 614417820.0,
      "step": 18340
    },
    {
      "epoch": 1.095093123209169,
      "grad_norm": 0.5464887619018555,
      "learning_rate": 2.582700981109647e-05,
      "loss": 0.4337,
      "mean_token_accuracy": 0.8544435024261474,
      "num_tokens": 614585500.0,
      "step": 18345
    },
    {
      "epoch": 1.0953915950334288,
      "grad_norm": 0.5516802668571472,
      "learning_rate": 2.5815936406507273e-05,
      "loss": 0.4588,
      "mean_token_accuracy": 0.8497495055198669,
      "num_tokens": 614753180.0,
      "step": 18350
    },
    {
      "epoch": 1.0956900668576885,
      "grad_norm": 0.5500524044036865,
      "learning_rate": 2.580486341210286e-05,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8565907120704651,
      "num_tokens": 614920860.0,
      "step": 18355
    },
    {
      "epoch": 1.0959885386819483,
      "grad_norm": 0.5346205234527588,
      "learning_rate": 2.5793790830580243e-05,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8499761462211609,
      "num_tokens": 615088540.0,
      "step": 18360
    },
    {
      "epoch": 1.0962870105062081,
      "grad_norm": 0.48874184489250183,
      "learning_rate": 2.5782718664636375e-05,
      "loss": 0.4675,
      "mean_token_accuracy": 0.8460873246192933,
      "num_tokens": 615256220.0,
      "step": 18365
    },
    {
      "epoch": 1.096585482330468,
      "grad_norm": 0.5291923880577087,
      "learning_rate": 2.577164691696805e-05,
      "loss": 0.4301,
      "mean_token_accuracy": 0.8561672449111939,
      "num_tokens": 615423900.0,
      "step": 18370
    },
    {
      "epoch": 1.0968839541547277,
      "grad_norm": 0.5688678622245789,
      "learning_rate": 2.5760575590272033e-05,
      "loss": 0.4526,
      "mean_token_accuracy": 0.8502266526222229,
      "num_tokens": 615591580.0,
      "step": 18375
    },
    {
      "epoch": 1.0971824259789875,
      "grad_norm": 0.5148406624794006,
      "learning_rate": 2.574950468724494e-05,
      "loss": 0.4472,
      "mean_token_accuracy": 0.8511391878128052,
      "num_tokens": 615759260.0,
      "step": 18380
    },
    {
      "epoch": 1.0974808978032473,
      "grad_norm": 0.5478710532188416,
      "learning_rate": 2.573843421058329e-05,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8308302521705627,
      "num_tokens": 615926940.0,
      "step": 18385
    },
    {
      "epoch": 1.097779369627507,
      "grad_norm": 0.5326007008552551,
      "learning_rate": 2.5727364162983507e-05,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8551711916923523,
      "num_tokens": 616094620.0,
      "step": 18390
    },
    {
      "epoch": 1.098077841451767,
      "grad_norm": 0.502690315246582,
      "learning_rate": 2.5716294547141895e-05,
      "loss": 0.46,
      "mean_token_accuracy": 0.8459739923477173,
      "num_tokens": 616262300.0,
      "step": 18395
    },
    {
      "epoch": 1.0983763132760267,
      "grad_norm": 0.4984673857688904,
      "learning_rate": 2.5705225365754665e-05,
      "loss": 0.4694,
      "mean_token_accuracy": 0.844924247264862,
      "num_tokens": 616429980.0,
      "step": 18400
    },
    {
      "epoch": 1.0986747851002865,
      "grad_norm": 0.5066306591033936,
      "learning_rate": 2.5694156621517935e-05,
      "loss": 0.4675,
      "mean_token_accuracy": 0.8442920327186585,
      "num_tokens": 616597660.0,
      "step": 18405
    },
    {
      "epoch": 1.0989732569245463,
      "grad_norm": 0.5151629447937012,
      "learning_rate": 2.5683088317127684e-05,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8567219376564026,
      "num_tokens": 616765340.0,
      "step": 18410
    },
    {
      "epoch": 1.099271728748806,
      "grad_norm": 0.5981072187423706,
      "learning_rate": 2.5672020455279815e-05,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8464453339576721,
      "num_tokens": 616927635.0,
      "step": 18415
    },
    {
      "epoch": 1.0995702005730659,
      "grad_norm": 0.5237789154052734,
      "learning_rate": 2.5660953038670105e-05,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8574078559875489,
      "num_tokens": 617095315.0,
      "step": 18420
    },
    {
      "epoch": 1.0998686723973257,
      "grad_norm": 0.526952862739563,
      "learning_rate": 2.5649886069994217e-05,
      "loss": 0.4169,
      "mean_token_accuracy": 0.860187292098999,
      "num_tokens": 617262995.0,
      "step": 18425
    },
    {
      "epoch": 1.1001671442215855,
      "grad_norm": 0.5191293358802795,
      "learning_rate": 2.563881955194774e-05,
      "loss": 0.5046,
      "mean_token_accuracy": 0.8336156487464905,
      "num_tokens": 617430675.0,
      "step": 18430
    },
    {
      "epoch": 1.1004656160458453,
      "grad_norm": 0.4855729937553406,
      "learning_rate": 2.56277534872261e-05,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8450912594795227,
      "num_tokens": 617598355.0,
      "step": 18435
    },
    {
      "epoch": 1.100764087870105,
      "grad_norm": 0.5127715468406677,
      "learning_rate": 2.561668787852467e-05,
      "loss": 0.468,
      "mean_token_accuracy": 0.8464809656143188,
      "num_tokens": 617766035.0,
      "step": 18440
    },
    {
      "epoch": 1.1010625596943648,
      "grad_norm": 0.520465612411499,
      "learning_rate": 2.560562272853866e-05,
      "loss": 0.4621,
      "mean_token_accuracy": 0.8458785653114319,
      "num_tokens": 617933715.0,
      "step": 18445
    },
    {
      "epoch": 1.1013610315186246,
      "grad_norm": 0.5159270763397217,
      "learning_rate": 2.5594558039963207e-05,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8489025473594666,
      "num_tokens": 618101395.0,
      "step": 18450
    },
    {
      "epoch": 1.1016595033428844,
      "grad_norm": 0.5116859078407288,
      "learning_rate": 2.558349381549331e-05,
      "loss": 0.4681,
      "mean_token_accuracy": 0.8442264199256897,
      "num_tokens": 618269075.0,
      "step": 18455
    },
    {
      "epoch": 1.1019579751671442,
      "grad_norm": 0.48104482889175415,
      "learning_rate": 2.557243005782387e-05,
      "loss": 0.4372,
      "mean_token_accuracy": 0.8558511137962341,
      "num_tokens": 618436755.0,
      "step": 18460
    },
    {
      "epoch": 1.102256446991404,
      "grad_norm": 0.5392863154411316,
      "learning_rate": 2.556136676964967e-05,
      "loss": 0.4455,
      "mean_token_accuracy": 0.8509841322898865,
      "num_tokens": 618604435.0,
      "step": 18465
    },
    {
      "epoch": 1.1025549188156638,
      "grad_norm": 0.49735796451568604,
      "learning_rate": 2.5550303953665373e-05,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8451151132583619,
      "num_tokens": 618772115.0,
      "step": 18470
    },
    {
      "epoch": 1.1028533906399236,
      "grad_norm": 0.5180985927581787,
      "learning_rate": 2.5539241612565533e-05,
      "loss": 0.4707,
      "mean_token_accuracy": 0.8442562341690063,
      "num_tokens": 618939795.0,
      "step": 18475
    },
    {
      "epoch": 1.1031518624641834,
      "grad_norm": 0.6098250150680542,
      "learning_rate": 2.5528179749044595e-05,
      "loss": 0.4807,
      "mean_token_accuracy": 0.8403674244880677,
      "num_tokens": 619107475.0,
      "step": 18480
    },
    {
      "epoch": 1.1034503342884432,
      "grad_norm": 0.576880156993866,
      "learning_rate": 2.5517118365796866e-05,
      "loss": 0.4786,
      "mean_token_accuracy": 0.8427412629127502,
      "num_tokens": 619275155.0,
      "step": 18485
    },
    {
      "epoch": 1.103748806112703,
      "grad_norm": 0.5477427244186401,
      "learning_rate": 2.5506057465516574e-05,
      "loss": 0.43,
      "mean_token_accuracy": 0.8560062050819397,
      "num_tokens": 619442835.0,
      "step": 18490
    },
    {
      "epoch": 1.1040472779369628,
      "grad_norm": 0.49006766080856323,
      "learning_rate": 2.549499705089778e-05,
      "loss": 0.4444,
      "mean_token_accuracy": 0.8523619294166564,
      "num_tokens": 619610515.0,
      "step": 18495
    },
    {
      "epoch": 1.1043457497612226,
      "grad_norm": 0.5077897310256958,
      "learning_rate": 2.548393712463447e-05,
      "loss": 0.4556,
      "mean_token_accuracy": 0.8489443063735962,
      "num_tokens": 619778195.0,
      "step": 18500
    },
    {
      "epoch": 1.1046442215854824,
      "grad_norm": 0.5169222950935364,
      "learning_rate": 2.547287768942047e-05,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8335858345031738,
      "num_tokens": 619945875.0,
      "step": 18505
    },
    {
      "epoch": 1.1049426934097422,
      "grad_norm": 0.4919980466365814,
      "learning_rate": 2.5461818747949533e-05,
      "loss": 0.4366,
      "mean_token_accuracy": 0.8535548210144043,
      "num_tokens": 620113555.0,
      "step": 18510
    },
    {
      "epoch": 1.105241165234002,
      "grad_norm": 0.47955581545829773,
      "learning_rate": 2.5450760302915273e-05,
      "loss": 0.4327,
      "mean_token_accuracy": 0.855451512336731,
      "num_tokens": 620281235.0,
      "step": 18515
    },
    {
      "epoch": 1.1055396370582617,
      "grad_norm": 0.5274818539619446,
      "learning_rate": 2.5439702357011152e-05,
      "loss": 0.4077,
      "mean_token_accuracy": 0.8651019930839539,
      "num_tokens": 620448915.0,
      "step": 18520
    },
    {
      "epoch": 1.1058381088825215,
      "grad_norm": 0.5645363926887512,
      "learning_rate": 2.5428644912930555e-05,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8398842930793762,
      "num_tokens": 620616595.0,
      "step": 18525
    },
    {
      "epoch": 1.1061365807067813,
      "grad_norm": 0.5423452258110046,
      "learning_rate": 2.541758797336672e-05,
      "loss": 0.4818,
      "mean_token_accuracy": 0.8411606907844543,
      "num_tokens": 620784275.0,
      "step": 18530
    },
    {
      "epoch": 1.1064350525310411,
      "grad_norm": 0.5464590191841125,
      "learning_rate": 2.5406531541012767e-05,
      "loss": 0.4853,
      "mean_token_accuracy": 0.8405761599540711,
      "num_tokens": 620951955.0,
      "step": 18535
    },
    {
      "epoch": 1.106733524355301,
      "grad_norm": 0.5267860889434814,
      "learning_rate": 2.539547561856171e-05,
      "loss": 0.4647,
      "mean_token_accuracy": 0.844661808013916,
      "num_tokens": 621119635.0,
      "step": 18540
    },
    {
      "epoch": 1.1070319961795607,
      "grad_norm": 0.46543750166893005,
      "learning_rate": 2.5384420208706406e-05,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8505368113517762,
      "num_tokens": 621287315.0,
      "step": 18545
    },
    {
      "epoch": 1.1073304680038205,
      "grad_norm": 0.5024165511131287,
      "learning_rate": 2.5373365314139608e-05,
      "loss": 0.4767,
      "mean_token_accuracy": 0.842544436454773,
      "num_tokens": 621454995.0,
      "step": 18550
    },
    {
      "epoch": 1.1076289398280803,
      "grad_norm": 0.5313209891319275,
      "learning_rate": 2.536231093755394e-05,
      "loss": 0.453,
      "mean_token_accuracy": 0.8497912406921386,
      "num_tokens": 621622675.0,
      "step": 18555
    },
    {
      "epoch": 1.10792741165234,
      "grad_norm": 0.511211097240448,
      "learning_rate": 2.5351257081641906e-05,
      "loss": 0.4414,
      "mean_token_accuracy": 0.8542053461074829,
      "num_tokens": 621790225.0,
      "step": 18560
    },
    {
      "epoch": 1.1082258834766,
      "grad_norm": 0.5000944137573242,
      "learning_rate": 2.534020374909586e-05,
      "loss": 0.4634,
      "mean_token_accuracy": 0.8475009083747864,
      "num_tokens": 621957905.0,
      "step": 18565
    },
    {
      "epoch": 1.1085243553008597,
      "grad_norm": 0.507066547870636,
      "learning_rate": 2.532915094260806e-05,
      "loss": 0.4326,
      "mean_token_accuracy": 0.8578014969825745,
      "num_tokens": 622125585.0,
      "step": 18570
    },
    {
      "epoch": 1.1088228271251195,
      "grad_norm": 0.5120506882667542,
      "learning_rate": 2.531809866487062e-05,
      "loss": 0.4689,
      "mean_token_accuracy": 0.8474770307540893,
      "num_tokens": 622293265.0,
      "step": 18575
    },
    {
      "epoch": 1.1091212989493793,
      "grad_norm": 0.5658789277076721,
      "learning_rate": 2.530704691857552e-05,
      "loss": 0.4444,
      "mean_token_accuracy": 0.8538232207298279,
      "num_tokens": 622460945.0,
      "step": 18580
    },
    {
      "epoch": 1.109419770773639,
      "grad_norm": 0.6437373161315918,
      "learning_rate": 2.529599570641463e-05,
      "loss": 0.5148,
      "mean_token_accuracy": 0.8337647676467895,
      "num_tokens": 622628625.0,
      "step": 18585
    },
    {
      "epoch": 1.1097182425978986,
      "grad_norm": 0.4807692766189575,
      "learning_rate": 2.528494503107966e-05,
      "loss": 0.4625,
      "mean_token_accuracy": 0.8465406179428101,
      "num_tokens": 622796305.0,
      "step": 18590
    },
    {
      "epoch": 1.1100167144221587,
      "grad_norm": 0.46318143606185913,
      "learning_rate": 2.5273894895262213e-05,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8503041863441467,
      "num_tokens": 622963985.0,
      "step": 18595
    },
    {
      "epoch": 1.1103151862464182,
      "grad_norm": 0.5829127430915833,
      "learning_rate": 2.5262845301653753e-05,
      "loss": 0.473,
      "mean_token_accuracy": 0.8428963422775269,
      "num_tokens": 623131665.0,
      "step": 18600
    },
    {
      "epoch": 1.1106136580706782,
      "grad_norm": 0.5519548058509827,
      "learning_rate": 2.5251796252945614e-05,
      "loss": 0.4838,
      "mean_token_accuracy": 0.840576171875,
      "num_tokens": 623299345.0,
      "step": 18605
    },
    {
      "epoch": 1.1109121298949378,
      "grad_norm": 0.5513665080070496,
      "learning_rate": 2.5240747751828998e-05,
      "loss": 0.4553,
      "mean_token_accuracy": 0.8498568534851074,
      "num_tokens": 623467025.0,
      "step": 18610
    },
    {
      "epoch": 1.1112106017191976,
      "grad_norm": 0.5571180582046509,
      "learning_rate": 2.522969980099496e-05,
      "loss": 0.5205,
      "mean_token_accuracy": 0.8306632518768311,
      "num_tokens": 623634705.0,
      "step": 18615
    },
    {
      "epoch": 1.1115090735434574,
      "grad_norm": 0.4902142286300659,
      "learning_rate": 2.521865240313444e-05,
      "loss": 0.452,
      "mean_token_accuracy": 0.8496480941772461,
      "num_tokens": 623802385.0,
      "step": 18620
    },
    {
      "epoch": 1.1118075453677172,
      "grad_norm": 0.4946683645248413,
      "learning_rate": 2.5207605560938248e-05,
      "loss": 0.4288,
      "mean_token_accuracy": 0.8569724440574646,
      "num_tokens": 623970065.0,
      "step": 18625
    },
    {
      "epoch": 1.112106017191977,
      "grad_norm": 0.5599443316459656,
      "learning_rate": 2.519655927709702e-05,
      "loss": 0.4285,
      "mean_token_accuracy": 0.8579577803611755,
      "num_tokens": 624135457.0,
      "step": 18630
    },
    {
      "epoch": 1.1124044890162368,
      "grad_norm": 0.4648134112358093,
      "learning_rate": 2.5185513554301315e-05,
      "loss": 0.4011,
      "mean_token_accuracy": 0.865650725364685,
      "num_tokens": 624303137.0,
      "step": 18635
    },
    {
      "epoch": 1.1127029608404966,
      "grad_norm": 0.45229512453079224,
      "learning_rate": 2.5174468395241484e-05,
      "loss": 0.4977,
      "mean_token_accuracy": 0.8354169011116028,
      "num_tokens": 624470817.0,
      "step": 18640
    },
    {
      "epoch": 1.1130014326647564,
      "grad_norm": 0.513867199420929,
      "learning_rate": 2.516342380260781e-05,
      "loss": 0.4482,
      "mean_token_accuracy": 0.851741623878479,
      "num_tokens": 624638497.0,
      "step": 18645
    },
    {
      "epoch": 1.1132999044890162,
      "grad_norm": 0.5177357196807861,
      "learning_rate": 2.5152379779090397e-05,
      "loss": 0.432,
      "mean_token_accuracy": 0.8550399661064148,
      "num_tokens": 624806177.0,
      "step": 18650
    },
    {
      "epoch": 1.113598376313276,
      "grad_norm": 0.5707488656044006,
      "learning_rate": 2.5141336327379218e-05,
      "loss": 0.445,
      "mean_token_accuracy": 0.8517833709716797,
      "num_tokens": 624973857.0,
      "step": 18655
    },
    {
      "epoch": 1.1138968481375358,
      "grad_norm": 0.5320921540260315,
      "learning_rate": 2.5130293450164123e-05,
      "loss": 0.4386,
      "mean_token_accuracy": 0.8533460617065429,
      "num_tokens": 625141537.0,
      "step": 18660
    },
    {
      "epoch": 1.1141953199617955,
      "grad_norm": 0.5352445840835571,
      "learning_rate": 2.511925115013479e-05,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8527198076248169,
      "num_tokens": 625309217.0,
      "step": 18665
    },
    {
      "epoch": 1.1144937917860553,
      "grad_norm": 0.5579648017883301,
      "learning_rate": 2.5108209429980784e-05,
      "loss": 0.453,
      "mean_token_accuracy": 0.8490635871887207,
      "num_tokens": 625476897.0,
      "step": 18670
    },
    {
      "epoch": 1.1147922636103151,
      "grad_norm": 0.5246790051460266,
      "learning_rate": 2.509716829239151e-05,
      "loss": 0.4118,
      "mean_token_accuracy": 0.8606226921081543,
      "num_tokens": 625644577.0,
      "step": 18675
    },
    {
      "epoch": 1.115090735434575,
      "grad_norm": 0.5772244334220886,
      "learning_rate": 2.5086127740056258e-05,
      "loss": 0.4445,
      "mean_token_accuracy": 0.8525766491889953,
      "num_tokens": 625812257.0,
      "step": 18680
    },
    {
      "epoch": 1.1153892072588347,
      "grad_norm": 0.6147289276123047,
      "learning_rate": 2.507508777566415e-05,
      "loss": 0.4243,
      "mean_token_accuracy": 0.8570499777793884,
      "num_tokens": 625979937.0,
      "step": 18685
    },
    {
      "epoch": 1.1156876790830945,
      "grad_norm": 0.4873496890068054,
      "learning_rate": 2.5064048401904168e-05,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8500179052352905,
      "num_tokens": 626147617.0,
      "step": 18690
    },
    {
      "epoch": 1.1159861509073543,
      "grad_norm": 0.4933112561702728,
      "learning_rate": 2.5053009621465167e-05,
      "loss": 0.4774,
      "mean_token_accuracy": 0.842908251285553,
      "num_tokens": 626315297.0,
      "step": 18695
    },
    {
      "epoch": 1.116284622731614,
      "grad_norm": 0.5187444090843201,
      "learning_rate": 2.5041971437035833e-05,
      "loss": 0.5004,
      "mean_token_accuracy": 0.8352618336677551,
      "num_tokens": 626482977.0,
      "step": 18700
    },
    {
      "epoch": 1.116583094555874,
      "grad_norm": 0.4695175290107727,
      "learning_rate": 2.503093385130473e-05,
      "loss": 0.4608,
      "mean_token_accuracy": 0.8457294464111328,
      "num_tokens": 626650657.0,
      "step": 18705
    },
    {
      "epoch": 1.1168815663801337,
      "grad_norm": 0.6617633104324341,
      "learning_rate": 2.5019896866960258e-05,
      "loss": 0.4471,
      "mean_token_accuracy": 0.8519742369651795,
      "num_tokens": 626818337.0,
      "step": 18710
    },
    {
      "epoch": 1.1171800382043935,
      "grad_norm": 0.5181426405906677,
      "learning_rate": 2.5008860486690682e-05,
      "loss": 0.4139,
      "mean_token_accuracy": 0.8622509837150574,
      "num_tokens": 626986017.0,
      "step": 18715
    },
    {
      "epoch": 1.1174785100286533,
      "grad_norm": 0.5520211458206177,
      "learning_rate": 2.499782471318412e-05,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8423714637756348,
      "num_tokens": 627153697.0,
      "step": 18720
    },
    {
      "epoch": 1.117776981852913,
      "grad_norm": 0.4633947014808655,
      "learning_rate": 2.498678954912853e-05,
      "loss": 0.4499,
      "mean_token_accuracy": 0.8509954452514649,
      "num_tokens": 627316377.0,
      "step": 18725
    },
    {
      "epoch": 1.1180754536771729,
      "grad_norm": 0.566048800945282,
      "learning_rate": 2.4975754997211733e-05,
      "loss": 0.4564,
      "mean_token_accuracy": 0.8484796643257141,
      "num_tokens": 627478346.0,
      "step": 18730
    },
    {
      "epoch": 1.1183739255014327,
      "grad_norm": 0.5108193159103394,
      "learning_rate": 2.4964721060121405e-05,
      "loss": 0.4202,
      "mean_token_accuracy": 0.8592508673667908,
      "num_tokens": 627646026.0,
      "step": 18735
    },
    {
      "epoch": 1.1186723973256925,
      "grad_norm": 0.4598321318626404,
      "learning_rate": 2.4953687740545052e-05,
      "loss": 0.4388,
      "mean_token_accuracy": 0.85297030210495,
      "num_tokens": 627813706.0,
      "step": 18740
    },
    {
      "epoch": 1.1189708691499523,
      "grad_norm": 0.5469609498977661,
      "learning_rate": 2.4942655041170054e-05,
      "loss": 0.4781,
      "mean_token_accuracy": 0.844172728061676,
      "num_tokens": 627981386.0,
      "step": 18745
    },
    {
      "epoch": 1.119269340974212,
      "grad_norm": 0.4844552278518677,
      "learning_rate": 2.4931622964683622e-05,
      "loss": 0.4642,
      "mean_token_accuracy": 0.8461052179336548,
      "num_tokens": 628149066.0,
      "step": 18750
    },
    {
      "epoch": 1.1195678127984718,
      "grad_norm": 0.5120344758033752,
      "learning_rate": 2.4920591513772813e-05,
      "loss": 0.4253,
      "mean_token_accuracy": 0.8579744815826416,
      "num_tokens": 628316746.0,
      "step": 18755
    },
    {
      "epoch": 1.1198662846227316,
      "grad_norm": 0.4837286174297333,
      "learning_rate": 2.4909560691124573e-05,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8561970591545105,
      "num_tokens": 628484426.0,
      "step": 18760
    },
    {
      "epoch": 1.1201647564469914,
      "grad_norm": 0.4842304289340973,
      "learning_rate": 2.489853049942562e-05,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8488130688667297,
      "num_tokens": 628652106.0,
      "step": 18765
    },
    {
      "epoch": 1.1204632282712512,
      "grad_norm": 0.5748688578605652,
      "learning_rate": 2.4887500941362597e-05,
      "loss": 0.4872,
      "mean_token_accuracy": 0.8410055994987488,
      "num_tokens": 628819786.0,
      "step": 18770
    },
    {
      "epoch": 1.120761700095511,
      "grad_norm": 0.525775134563446,
      "learning_rate": 2.4876472019621926e-05,
      "loss": 0.4363,
      "mean_token_accuracy": 0.8562507510185242,
      "num_tokens": 628987466.0,
      "step": 18775
    },
    {
      "epoch": 1.1210601719197708,
      "grad_norm": 0.5338101983070374,
      "learning_rate": 2.486544373688993e-05,
      "loss": 0.4824,
      "mean_token_accuracy": 0.8400930523872375,
      "num_tokens": 629155146.0,
      "step": 18780
    },
    {
      "epoch": 1.1213586437440306,
      "grad_norm": 0.523887038230896,
      "learning_rate": 2.485441609585274e-05,
      "loss": 0.4439,
      "mean_token_accuracy": 0.8521949172019958,
      "num_tokens": 629322826.0,
      "step": 18785
    },
    {
      "epoch": 1.1216571155682904,
      "grad_norm": 0.5007719397544861,
      "learning_rate": 2.4843389099196333e-05,
      "loss": 0.4846,
      "mean_token_accuracy": 0.8414231300354004,
      "num_tokens": 629490506.0,
      "step": 18790
    },
    {
      "epoch": 1.1219555873925502,
      "grad_norm": 0.5108416080474854,
      "learning_rate": 2.4832362749606563e-05,
      "loss": 0.4365,
      "mean_token_accuracy": 0.8550638079643249,
      "num_tokens": 629658186.0,
      "step": 18795
    },
    {
      "epoch": 1.12225405921681,
      "grad_norm": 0.49002453684806824,
      "learning_rate": 2.482133704976907e-05,
      "loss": 0.4524,
      "mean_token_accuracy": 0.8481234669685364,
      "num_tokens": 629822897.0,
      "step": 18800
    },
    {
      "epoch": 1.1225525310410698,
      "grad_norm": 0.5626703500747681,
      "learning_rate": 2.481031200236939e-05,
      "loss": 0.4688,
      "mean_token_accuracy": 0.8455027937889099,
      "num_tokens": 629990577.0,
      "step": 18805
    },
    {
      "epoch": 1.1228510028653296,
      "grad_norm": 0.5119451284408569,
      "learning_rate": 2.4799287610092857e-05,
      "loss": 0.4761,
      "mean_token_accuracy": 0.8410875558853149,
      "num_tokens": 630150033.0,
      "step": 18810
    },
    {
      "epoch": 1.1231494746895894,
      "grad_norm": 0.48004892468452454,
      "learning_rate": 2.478826387562468e-05,
      "loss": 0.5217,
      "mean_token_accuracy": 0.8309793591499328,
      "num_tokens": 630317713.0,
      "step": 18815
    },
    {
      "epoch": 1.1234479465138492,
      "grad_norm": 0.507409393787384,
      "learning_rate": 2.4777240801649898e-05,
      "loss": 0.4183,
      "mean_token_accuracy": 0.8602648258209229,
      "num_tokens": 630485393.0,
      "step": 18820
    },
    {
      "epoch": 1.123746418338109,
      "grad_norm": 0.47925466299057007,
      "learning_rate": 2.476621839085336e-05,
      "loss": 0.4419,
      "mean_token_accuracy": 0.8538053154945373,
      "num_tokens": 630653073.0,
      "step": 18825
    },
    {
      "epoch": 1.1240448901623687,
      "grad_norm": 0.5088526010513306,
      "learning_rate": 2.4755196645919816e-05,
      "loss": 0.4178,
      "mean_token_accuracy": 0.8601216793060302,
      "num_tokens": 630820753.0,
      "step": 18830
    },
    {
      "epoch": 1.1243433619866285,
      "grad_norm": 0.5245963335037231,
      "learning_rate": 2.4744175569533767e-05,
      "loss": 0.4336,
      "mean_token_accuracy": 0.8552546858787536,
      "num_tokens": 630988433.0,
      "step": 18835
    },
    {
      "epoch": 1.1246418338108883,
      "grad_norm": 0.48288968205451965,
      "learning_rate": 2.4733155164379628e-05,
      "loss": 0.4132,
      "mean_token_accuracy": 0.8623464107513428,
      "num_tokens": 631156113.0,
      "step": 18840
    },
    {
      "epoch": 1.1249403056351481,
      "grad_norm": 0.5197480320930481,
      "learning_rate": 2.4722135433141628e-05,
      "loss": 0.4786,
      "mean_token_accuracy": 0.8421090245246887,
      "num_tokens": 631323793.0,
      "step": 18845
    },
    {
      "epoch": 1.125238777459408,
      "grad_norm": 0.8062328100204468,
      "learning_rate": 2.4711116378503816e-05,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8470416307449341,
      "num_tokens": 631491473.0,
      "step": 18850
    },
    {
      "epoch": 1.1255372492836677,
      "grad_norm": 0.5386670827865601,
      "learning_rate": 2.470009800315009e-05,
      "loss": 0.4444,
      "mean_token_accuracy": 0.8508708119392395,
      "num_tokens": 631659153.0,
      "step": 18855
    },
    {
      "epoch": 1.1258357211079275,
      "grad_norm": 0.48615118861198425,
      "learning_rate": 2.4689080309764166e-05,
      "loss": 0.4568,
      "mean_token_accuracy": 0.8480973362922668,
      "num_tokens": 631826833.0,
      "step": 18860
    },
    {
      "epoch": 1.126134192932187,
      "grad_norm": 0.557109534740448,
      "learning_rate": 2.4678063301029623e-05,
      "loss": 0.4586,
      "mean_token_accuracy": 0.8491649746894836,
      "num_tokens": 631994513.0,
      "step": 18865
    },
    {
      "epoch": 1.126432664756447,
      "grad_norm": 0.4750424027442932,
      "learning_rate": 2.466704697962985e-05,
      "loss": 0.4425,
      "mean_token_accuracy": 0.8523380756378174,
      "num_tokens": 632162193.0,
      "step": 18870
    },
    {
      "epoch": 1.1267311365807067,
      "grad_norm": 0.5284352898597717,
      "learning_rate": 2.465603134824807e-05,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8445007681846619,
      "num_tokens": 632329873.0,
      "step": 18875
    },
    {
      "epoch": 1.1270296084049667,
      "grad_norm": 0.5021812915802002,
      "learning_rate": 2.4645016409567358e-05,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8509423971176148,
      "num_tokens": 632497553.0,
      "step": 18880
    },
    {
      "epoch": 1.1273280802292263,
      "grad_norm": 0.5287967324256897,
      "learning_rate": 2.4634002166270586e-05,
      "loss": 0.4859,
      "mean_token_accuracy": 0.8388882279396057,
      "num_tokens": 632665233.0,
      "step": 18885
    },
    {
      "epoch": 1.1276265520534863,
      "grad_norm": 0.5061269998550415,
      "learning_rate": 2.46229886210405e-05,
      "loss": 0.5049,
      "mean_token_accuracy": 0.8349457263946534,
      "num_tokens": 632832913.0,
      "step": 18890
    },
    {
      "epoch": 1.1279250238777458,
      "grad_norm": 0.5035704970359802,
      "learning_rate": 2.4611975776559626e-05,
      "loss": 0.472,
      "mean_token_accuracy": 0.8462602853775024,
      "num_tokens": 633000593.0,
      "step": 18895
    },
    {
      "epoch": 1.1282234957020056,
      "grad_norm": 0.6159593462944031,
      "learning_rate": 2.4600963635510354e-05,
      "loss": 0.4773,
      "mean_token_accuracy": 0.8428009033203125,
      "num_tokens": 633168273.0,
      "step": 18900
    },
    {
      "epoch": 1.1285219675262654,
      "grad_norm": 0.5488763451576233,
      "learning_rate": 2.458995220057491e-05,
      "loss": 0.4367,
      "mean_token_accuracy": 0.8522724628448486,
      "num_tokens": 633335953.0,
      "step": 18905
    },
    {
      "epoch": 1.1288204393505252,
      "grad_norm": 0.5279453992843628,
      "learning_rate": 2.457894147443532e-05,
      "loss": 0.435,
      "mean_token_accuracy": 0.8547178864479065,
      "num_tokens": 633503633.0,
      "step": 18910
    },
    {
      "epoch": 1.129118911174785,
      "grad_norm": 0.5483424067497253,
      "learning_rate": 2.4567931459773452e-05,
      "loss": 0.4452,
      "mean_token_accuracy": 0.8512585043907166,
      "num_tokens": 633671313.0,
      "step": 18915
    },
    {
      "epoch": 1.1294173829990448,
      "grad_norm": 0.5122362375259399,
      "learning_rate": 2.455692215927099e-05,
      "loss": 0.4769,
      "mean_token_accuracy": 0.8424489974975586,
      "num_tokens": 633838993.0,
      "step": 18920
    },
    {
      "epoch": 1.1297158548233046,
      "grad_norm": 0.5410659909248352,
      "learning_rate": 2.4545913575609463e-05,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8473100304603577,
      "num_tokens": 634006673.0,
      "step": 18925
    },
    {
      "epoch": 1.1300143266475644,
      "grad_norm": 0.4490204453468323,
      "learning_rate": 2.4534905711470215e-05,
      "loss": 0.4404,
      "mean_token_accuracy": 0.8527018904685975,
      "num_tokens": 634174353.0,
      "step": 18930
    },
    {
      "epoch": 1.1303127984718242,
      "grad_norm": 0.5037288069725037,
      "learning_rate": 2.45238985695344e-05,
      "loss": 0.4582,
      "mean_token_accuracy": 0.8473219633102417,
      "num_tokens": 634342033.0,
      "step": 18935
    },
    {
      "epoch": 1.130611270296084,
      "grad_norm": 0.5119936466217041,
      "learning_rate": 2.451289215248303e-05,
      "loss": 0.454,
      "mean_token_accuracy": 0.8490933895111084,
      "num_tokens": 634509713.0,
      "step": 18940
    },
    {
      "epoch": 1.1309097421203438,
      "grad_norm": 0.4720776379108429,
      "learning_rate": 2.4501886462996908e-05,
      "loss": 0.4586,
      "mean_token_accuracy": 0.8477633237838745,
      "num_tokens": 634677393.0,
      "step": 18945
    },
    {
      "epoch": 1.1312082139446036,
      "grad_norm": 0.5458407998085022,
      "learning_rate": 2.4490881503756668e-05,
      "loss": 0.4678,
      "mean_token_accuracy": 0.8445604324340821,
      "num_tokens": 634845073.0,
      "step": 18950
    },
    {
      "epoch": 1.1315066857688634,
      "grad_norm": 0.5116104483604431,
      "learning_rate": 2.44798772774428e-05,
      "loss": 0.4191,
      "mean_token_accuracy": 0.8598532795906066,
      "num_tokens": 635012753.0,
      "step": 18955
    },
    {
      "epoch": 1.1318051575931232,
      "grad_norm": 0.4883235991001129,
      "learning_rate": 2.4468873786735553e-05,
      "loss": 0.4616,
      "mean_token_accuracy": 0.8462841510772705,
      "num_tokens": 635180433.0,
      "step": 18960
    },
    {
      "epoch": 1.132103629417383,
      "grad_norm": 0.5117110013961792,
      "learning_rate": 2.445787103431505e-05,
      "loss": 0.4354,
      "mean_token_accuracy": 0.8550399541854858,
      "num_tokens": 635348113.0,
      "step": 18965
    },
    {
      "epoch": 1.1324021012416428,
      "grad_norm": 0.49613380432128906,
      "learning_rate": 2.4446869022861195e-05,
      "loss": 0.4398,
      "mean_token_accuracy": 0.8532565832138062,
      "num_tokens": 635515793.0,
      "step": 18970
    },
    {
      "epoch": 1.1327005730659025,
      "grad_norm": 0.4743483066558838,
      "learning_rate": 2.443586775505376e-05,
      "loss": 0.4307,
      "mean_token_accuracy": 0.8560479640960693,
      "num_tokens": 635683473.0,
      "step": 18975
    },
    {
      "epoch": 1.1329990448901623,
      "grad_norm": 0.47173023223876953,
      "learning_rate": 2.442486723357228e-05,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8483001232147217,
      "num_tokens": 635851153.0,
      "step": 18980
    },
    {
      "epoch": 1.1332975167144221,
      "grad_norm": 0.5263618230819702,
      "learning_rate": 2.4413867461096147e-05,
      "loss": 0.4633,
      "mean_token_accuracy": 0.845174765586853,
      "num_tokens": 636018833.0,
      "step": 18985
    },
    {
      "epoch": 1.133595988538682,
      "grad_norm": 0.5367741584777832,
      "learning_rate": 2.440286844030456e-05,
      "loss": 0.4773,
      "mean_token_accuracy": 0.8444530606269837,
      "num_tokens": 636186513.0,
      "step": 18990
    },
    {
      "epoch": 1.1338944603629417,
      "grad_norm": 0.5594318509101868,
      "learning_rate": 2.439187017387653e-05,
      "loss": 0.4809,
      "mean_token_accuracy": 0.8401586532592773,
      "num_tokens": 636354193.0,
      "step": 18995
    },
    {
      "epoch": 1.1341929321872015,
      "grad_norm": 0.5120664834976196,
      "learning_rate": 2.438087266449089e-05,
      "loss": 0.4567,
      "mean_token_accuracy": 0.8504831314086914,
      "num_tokens": 636521873.0,
      "step": 19000
    },
    {
      "epoch": 1.1344914040114613,
      "grad_norm": 0.48474791646003723,
      "learning_rate": 2.4369875914826283e-05,
      "loss": 0.4429,
      "mean_token_accuracy": 0.8525647163391114,
      "num_tokens": 636689553.0,
      "step": 19005
    },
    {
      "epoch": 1.134789875835721,
      "grad_norm": 0.43569645285606384,
      "learning_rate": 2.4358879927561173e-05,
      "loss": 0.4175,
      "mean_token_accuracy": 0.8619527578353882,
      "num_tokens": 636857233.0,
      "step": 19010
    },
    {
      "epoch": 1.135088347659981,
      "grad_norm": 0.5542641282081604,
      "learning_rate": 2.4347884705373836e-05,
      "loss": 0.4563,
      "mean_token_accuracy": 0.8471847891807556,
      "num_tokens": 637024913.0,
      "step": 19015
    },
    {
      "epoch": 1.1353868194842407,
      "grad_norm": 0.45384833216667175,
      "learning_rate": 2.4336890250942357e-05,
      "loss": 0.4741,
      "mean_token_accuracy": 0.843409287929535,
      "num_tokens": 637192593.0,
      "step": 19020
    },
    {
      "epoch": 1.1356852913085005,
      "grad_norm": 0.5924787521362305,
      "learning_rate": 2.432589656694465e-05,
      "loss": 0.4467,
      "mean_token_accuracy": 0.8513658523559571,
      "num_tokens": 637360273.0,
      "step": 19025
    },
    {
      "epoch": 1.1359837631327603,
      "grad_norm": 0.5013798475265503,
      "learning_rate": 2.4314903656058407e-05,
      "loss": 0.4798,
      "mean_token_accuracy": 0.8411785721778869,
      "num_tokens": 637527953.0,
      "step": 19030
    },
    {
      "epoch": 1.13628223495702,
      "grad_norm": 0.5134493708610535,
      "learning_rate": 2.4303911520961176e-05,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8435822606086731,
      "num_tokens": 637695633.0,
      "step": 19035
    },
    {
      "epoch": 1.1365807067812799,
      "grad_norm": 0.5029426217079163,
      "learning_rate": 2.429292016433029e-05,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8541810870170593,
      "num_tokens": 637863313.0,
      "step": 19040
    },
    {
      "epoch": 1.1368791786055397,
      "grad_norm": 0.48797866702079773,
      "learning_rate": 2.4281929588842893e-05,
      "loss": 0.4523,
      "mean_token_accuracy": 0.8495467066764831,
      "num_tokens": 638030993.0,
      "step": 19045
    },
    {
      "epoch": 1.1371776504297995,
      "grad_norm": 0.5125130414962769,
      "learning_rate": 2.4270939797175946e-05,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8382977604866028,
      "num_tokens": 638198673.0,
      "step": 19050
    },
    {
      "epoch": 1.1374761222540593,
      "grad_norm": 0.5460643768310547,
      "learning_rate": 2.425995079200621e-05,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8456459522247315,
      "num_tokens": 638366353.0,
      "step": 19055
    },
    {
      "epoch": 1.137774594078319,
      "grad_norm": 0.4995202422142029,
      "learning_rate": 2.424896257601026e-05,
      "loss": 0.4412,
      "mean_token_accuracy": 0.8523499965667725,
      "num_tokens": 638534033.0,
      "step": 19060
    },
    {
      "epoch": 1.1380730659025788,
      "grad_norm": 0.4863225817680359,
      "learning_rate": 2.423797515186449e-05,
      "loss": 0.4184,
      "mean_token_accuracy": 0.8598890542984009,
      "num_tokens": 638701713.0,
      "step": 19065
    },
    {
      "epoch": 1.1383715377268386,
      "grad_norm": 0.597139298915863,
      "learning_rate": 2.422698852224507e-05,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8443874478340149,
      "num_tokens": 638869393.0,
      "step": 19070
    },
    {
      "epoch": 1.1386700095510984,
      "grad_norm": 0.5598382949829102,
      "learning_rate": 2.4216002689828022e-05,
      "loss": 0.4705,
      "mean_token_accuracy": 0.8461409926414489,
      "num_tokens": 639037073.0,
      "step": 19075
    },
    {
      "epoch": 1.1389684813753582,
      "grad_norm": 0.5371782183647156,
      "learning_rate": 2.4205017657289127e-05,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8455564975738525,
      "num_tokens": 639204753.0,
      "step": 19080
    },
    {
      "epoch": 1.139266953199618,
      "grad_norm": 0.5139977931976318,
      "learning_rate": 2.4194033427304e-05,
      "loss": 0.4497,
      "mean_token_accuracy": 0.8494035363197326,
      "num_tokens": 639372433.0,
      "step": 19085
    },
    {
      "epoch": 1.1395654250238778,
      "grad_norm": 0.5258270502090454,
      "learning_rate": 2.4183050002548053e-05,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8466984033584595,
      "num_tokens": 639532795.0,
      "step": 19090
    },
    {
      "epoch": 1.1398638968481376,
      "grad_norm": 0.5295782089233398,
      "learning_rate": 2.4172067385696495e-05,
      "loss": 0.4601,
      "mean_token_accuracy": 0.8459859251976013,
      "num_tokens": 639700475.0,
      "step": 19095
    },
    {
      "epoch": 1.1401623686723974,
      "grad_norm": 0.5212468504905701,
      "learning_rate": 2.416108557942437e-05,
      "loss": 0.4926,
      "mean_token_accuracy": 0.8374627232551575,
      "num_tokens": 639868155.0,
      "step": 19100
    },
    {
      "epoch": 1.1404608404966572,
      "grad_norm": 0.5109013915061951,
      "learning_rate": 2.4150104586406458e-05,
      "loss": 0.4567,
      "mean_token_accuracy": 0.8484373092651367,
      "num_tokens": 640035835.0,
      "step": 19105
    },
    {
      "epoch": 1.140759312320917,
      "grad_norm": 0.4664410650730133,
      "learning_rate": 2.4139124409317426e-05,
      "loss": 0.4414,
      "mean_token_accuracy": 0.8532983422279358,
      "num_tokens": 640203515.0,
      "step": 19110
    },
    {
      "epoch": 1.1410577841451768,
      "grad_norm": 0.49429747462272644,
      "learning_rate": 2.412814505083167e-05,
      "loss": 0.4539,
      "mean_token_accuracy": 0.8493438959121704,
      "num_tokens": 640371195.0,
      "step": 19115
    },
    {
      "epoch": 1.1413562559694366,
      "grad_norm": 0.5251138806343079,
      "learning_rate": 2.4117166513623422e-05,
      "loss": 0.4464,
      "mean_token_accuracy": 0.8530895948410034,
      "num_tokens": 640538875.0,
      "step": 19120
    },
    {
      "epoch": 1.1416547277936964,
      "grad_norm": 0.6203110218048096,
      "learning_rate": 2.4106188800366714e-05,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8427710771560669,
      "num_tokens": 640706555.0,
      "step": 19125
    },
    {
      "epoch": 1.1419531996179562,
      "grad_norm": 0.4993927478790283,
      "learning_rate": 2.409521191373536e-05,
      "loss": 0.431,
      "mean_token_accuracy": 0.8566264986991883,
      "num_tokens": 640874235.0,
      "step": 19130
    },
    {
      "epoch": 1.142251671442216,
      "grad_norm": 0.6741517186164856,
      "learning_rate": 2.4084235856402996e-05,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8510616779327392,
      "num_tokens": 641041915.0,
      "step": 19135
    },
    {
      "epoch": 1.1425501432664755,
      "grad_norm": 0.5305507779121399,
      "learning_rate": 2.4073260631043038e-05,
      "loss": 0.4983,
      "mean_token_accuracy": 0.8357807517051696,
      "num_tokens": 641209595.0,
      "step": 19140
    },
    {
      "epoch": 1.1428486150907355,
      "grad_norm": 0.5178223848342896,
      "learning_rate": 2.4062286240328702e-05,
      "loss": 0.4544,
      "mean_token_accuracy": 0.849487054347992,
      "num_tokens": 641377275.0,
      "step": 19145
    },
    {
      "epoch": 1.1431470869149951,
      "grad_norm": 0.4566723704338074,
      "learning_rate": 2.4051312686933004e-05,
      "loss": 0.4274,
      "mean_token_accuracy": 0.8563103914260864,
      "num_tokens": 641544955.0,
      "step": 19150
    },
    {
      "epoch": 1.1434455587392551,
      "grad_norm": 0.5075921416282654,
      "learning_rate": 2.4040339973528754e-05,
      "loss": 0.4559,
      "mean_token_accuracy": 0.8491172552108764,
      "num_tokens": 641712635.0,
      "step": 19155
    },
    {
      "epoch": 1.1437440305635147,
      "grad_norm": 0.4714864492416382,
      "learning_rate": 2.402936810278857e-05,
      "loss": 0.4799,
      "mean_token_accuracy": 0.8408684253692627,
      "num_tokens": 641880315.0,
      "step": 19160
    },
    {
      "epoch": 1.1440425023877747,
      "grad_norm": 0.48537150025367737,
      "learning_rate": 2.401839707738483e-05,
      "loss": 0.4327,
      "mean_token_accuracy": 0.8557616472244263,
      "num_tokens": 642047995.0,
      "step": 19165
    },
    {
      "epoch": 1.1443409742120343,
      "grad_norm": 0.5679064989089966,
      "learning_rate": 2.4007426899989753e-05,
      "loss": 0.4554,
      "mean_token_accuracy": 0.850477147102356,
      "num_tokens": 642215675.0,
      "step": 19170
    },
    {
      "epoch": 1.144639446036294,
      "grad_norm": 0.4865145981311798,
      "learning_rate": 2.3996457573275326e-05,
      "loss": 0.4414,
      "mean_token_accuracy": 0.8535130500793457,
      "num_tokens": 642383355.0,
      "step": 19175
    },
    {
      "epoch": 1.1449379178605539,
      "grad_norm": 0.5719457864761353,
      "learning_rate": 2.3985489099913322e-05,
      "loss": 0.4634,
      "mean_token_accuracy": 0.8466837644577027,
      "num_tokens": 642551035.0,
      "step": 19180
    },
    {
      "epoch": 1.1452363896848137,
      "grad_norm": 0.4855192005634308,
      "learning_rate": 2.3974521482575317e-05,
      "loss": 0.4571,
      "mean_token_accuracy": 0.8489920139312744,
      "num_tokens": 642718715.0,
      "step": 19185
    },
    {
      "epoch": 1.1455348615090735,
      "grad_norm": 0.4773414433002472,
      "learning_rate": 2.3963554723932673e-05,
      "loss": 0.4369,
      "mean_token_accuracy": 0.8547298073768616,
      "num_tokens": 642886395.0,
      "step": 19190
    },
    {
      "epoch": 1.1458333333333333,
      "grad_norm": 0.516990602016449,
      "learning_rate": 2.3952588826656562e-05,
      "loss": 0.4389,
      "mean_token_accuracy": 0.8544315814971923,
      "num_tokens": 643054075.0,
      "step": 19195
    },
    {
      "epoch": 1.146131805157593,
      "grad_norm": 0.45904049277305603,
      "learning_rate": 2.394162379341791e-05,
      "loss": 0.4616,
      "mean_token_accuracy": 0.8474591493606567,
      "num_tokens": 643221755.0,
      "step": 19200
    },
    {
      "epoch": 1.1464302769818528,
      "grad_norm": 0.5339150428771973,
      "learning_rate": 2.393065962688747e-05,
      "loss": 0.4753,
      "mean_token_accuracy": 0.840399706363678,
      "num_tokens": 643386233.0,
      "step": 19205
    },
    {
      "epoch": 1.1467287488061126,
      "grad_norm": 0.5641091465950012,
      "learning_rate": 2.3919696329735757e-05,
      "loss": 0.483,
      "mean_token_accuracy": 0.8402719855308532,
      "num_tokens": 643553913.0,
      "step": 19210
    },
    {
      "epoch": 1.1470272206303724,
      "grad_norm": 0.6663140058517456,
      "learning_rate": 2.3908733904633086e-05,
      "loss": 0.4482,
      "mean_token_accuracy": 0.8502909898757934,
      "num_tokens": 643719878.0,
      "step": 19215
    },
    {
      "epoch": 1.1473256924546322,
      "grad_norm": 0.538935661315918,
      "learning_rate": 2.3897772354249565e-05,
      "loss": 0.4574,
      "mean_token_accuracy": 0.8498151063919067,
      "num_tokens": 643887558.0,
      "step": 19220
    },
    {
      "epoch": 1.147624164278892,
      "grad_norm": 0.6172839403152466,
      "learning_rate": 2.3886811681255074e-05,
      "loss": 0.4593,
      "mean_token_accuracy": 0.8503340005874633,
      "num_tokens": 644055238.0,
      "step": 19225
    },
    {
      "epoch": 1.1479226361031518,
      "grad_norm": 0.45632052421569824,
      "learning_rate": 2.3875851888319277e-05,
      "loss": 0.4444,
      "mean_token_accuracy": 0.8506024122238159,
      "num_tokens": 644222918.0,
      "step": 19230
    },
    {
      "epoch": 1.1482211079274116,
      "grad_norm": 0.46800461411476135,
      "learning_rate": 2.386489297811167e-05,
      "loss": 0.3943,
      "mean_token_accuracy": 0.8668734431266785,
      "num_tokens": 644390598.0,
      "step": 19235
    },
    {
      "epoch": 1.1485195797516714,
      "grad_norm": 0.5249519348144531,
      "learning_rate": 2.3853934953301455e-05,
      "loss": 0.4449,
      "mean_token_accuracy": 0.8518907308578492,
      "num_tokens": 644558278.0,
      "step": 19240
    },
    {
      "epoch": 1.1488180515759312,
      "grad_norm": 0.5473239421844482,
      "learning_rate": 2.3842977816557694e-05,
      "loss": 0.4078,
      "mean_token_accuracy": 0.863545274734497,
      "num_tokens": 644725958.0,
      "step": 19245
    },
    {
      "epoch": 1.149116523400191,
      "grad_norm": 0.5105882883071899,
      "learning_rate": 2.3832021570549178e-05,
      "loss": 0.4436,
      "mean_token_accuracy": 0.8536741018295289,
      "num_tokens": 644893638.0,
      "step": 19250
    },
    {
      "epoch": 1.1494149952244508,
      "grad_norm": 0.5072927474975586,
      "learning_rate": 2.3821066217944517e-05,
      "loss": 0.476,
      "mean_token_accuracy": 0.8418585300445557,
      "num_tokens": 645061318.0,
      "step": 19255
    },
    {
      "epoch": 1.1497134670487106,
      "grad_norm": 0.5030297636985779,
      "learning_rate": 2.3810111761412086e-05,
      "loss": 0.4575,
      "mean_token_accuracy": 0.8480973362922668,
      "num_tokens": 645228998.0,
      "step": 19260
    },
    {
      "epoch": 1.1500119388729704,
      "grad_norm": 0.46399497985839844,
      "learning_rate": 2.379915820362003e-05,
      "loss": 0.4254,
      "mean_token_accuracy": 0.8578969359397888,
      "num_tokens": 645396678.0,
      "step": 19265
    },
    {
      "epoch": 1.1503104106972302,
      "grad_norm": 0.5404046177864075,
      "learning_rate": 2.378820554723632e-05,
      "loss": 0.4837,
      "mean_token_accuracy": 0.8409817457199097,
      "num_tokens": 645564358.0,
      "step": 19270
    },
    {
      "epoch": 1.15060888252149,
      "grad_norm": 0.5211895108222961,
      "learning_rate": 2.377725379492865e-05,
      "loss": 0.4405,
      "mean_token_accuracy": 0.8542765140533447,
      "num_tokens": 645732038.0,
      "step": 19275
    },
    {
      "epoch": 1.1509073543457498,
      "grad_norm": 0.5210592746734619,
      "learning_rate": 2.3766302949364528e-05,
      "loss": 0.4364,
      "mean_token_accuracy": 0.8568591117858887,
      "num_tokens": 645899718.0,
      "step": 19280
    },
    {
      "epoch": 1.1512058261700095,
      "grad_norm": 0.4576818645000458,
      "learning_rate": 2.3755353013211255e-05,
      "loss": 0.438,
      "mean_token_accuracy": 0.8534355163574219,
      "num_tokens": 646067398.0,
      "step": 19285
    },
    {
      "epoch": 1.1515042979942693,
      "grad_norm": 0.45660170912742615,
      "learning_rate": 2.374440398913586e-05,
      "loss": 0.4487,
      "mean_token_accuracy": 0.8509364128112793,
      "num_tokens": 646235078.0,
      "step": 19290
    },
    {
      "epoch": 1.1518027698185291,
      "grad_norm": 0.5278347134590149,
      "learning_rate": 2.3733455879805207e-05,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8441906213760376,
      "num_tokens": 646402758.0,
      "step": 19295
    },
    {
      "epoch": 1.152101241642789,
      "grad_norm": 0.5075232982635498,
      "learning_rate": 2.372250868788588e-05,
      "loss": 0.4281,
      "mean_token_accuracy": 0.8569843769073486,
      "num_tokens": 646570438.0,
      "step": 19300
    },
    {
      "epoch": 1.1523997134670487,
      "grad_norm": 0.4600481688976288,
      "learning_rate": 2.371156241604431e-05,
      "loss": 0.4298,
      "mean_token_accuracy": 0.857318389415741,
      "num_tokens": 646738118.0,
      "step": 19305
    },
    {
      "epoch": 1.1526981852913085,
      "grad_norm": 0.5240980386734009,
      "learning_rate": 2.370061706694662e-05,
      "loss": 0.462,
      "mean_token_accuracy": 0.8460336327552795,
      "num_tokens": 646905798.0,
      "step": 19310
    },
    {
      "epoch": 1.1529966571155683,
      "grad_norm": 0.5291630625724792,
      "learning_rate": 2.368967264325877e-05,
      "loss": 0.4288,
      "mean_token_accuracy": 0.8553501009941101,
      "num_tokens": 647073478.0,
      "step": 19315
    },
    {
      "epoch": 1.153295128939828,
      "grad_norm": 0.498884379863739,
      "learning_rate": 2.367872914764649e-05,
      "loss": 0.4495,
      "mean_token_accuracy": 0.8502147197723389,
      "num_tokens": 647241158.0,
      "step": 19320
    },
    {
      "epoch": 1.153593600764088,
      "grad_norm": 0.47125518321990967,
      "learning_rate": 2.366778658277525e-05,
      "loss": 0.4737,
      "mean_token_accuracy": 0.844172728061676,
      "num_tokens": 647408838.0,
      "step": 19325
    },
    {
      "epoch": 1.1538920725883477,
      "grad_norm": 0.5179007053375244,
      "learning_rate": 2.365684495131033e-05,
      "loss": 0.4929,
      "mean_token_accuracy": 0.8375044584274292,
      "num_tokens": 647576518.0,
      "step": 19330
    },
    {
      "epoch": 1.1541905444126075,
      "grad_norm": 0.5301775336265564,
      "learning_rate": 2.3645904255916756e-05,
      "loss": 0.4452,
      "mean_token_accuracy": 0.8518907308578492,
      "num_tokens": 647744198.0,
      "step": 19335
    },
    {
      "epoch": 1.1544890162368673,
      "grad_norm": 0.5048215985298157,
      "learning_rate": 2.3634964499259337e-05,
      "loss": 0.4425,
      "mean_token_accuracy": 0.8521770238876343,
      "num_tokens": 647911878.0,
      "step": 19340
    },
    {
      "epoch": 1.154787488061127,
      "grad_norm": 0.5531294941902161,
      "learning_rate": 2.362402568400266e-05,
      "loss": 0.458,
      "mean_token_accuracy": 0.8494453072547913,
      "num_tokens": 648079558.0,
      "step": 19345
    },
    {
      "epoch": 1.1550859598853869,
      "grad_norm": 0.5473691821098328,
      "learning_rate": 2.3613087812811065e-05,
      "loss": 0.4733,
      "mean_token_accuracy": 0.841691541671753,
      "num_tokens": 648247238.0,
      "step": 19350
    },
    {
      "epoch": 1.1553844317096467,
      "grad_norm": 0.4831838607788086,
      "learning_rate": 2.3602150888348682e-05,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8529106616973877,
      "num_tokens": 648414918.0,
      "step": 19355
    },
    {
      "epoch": 1.1556829035339065,
      "grad_norm": 0.5275028347969055,
      "learning_rate": 2.359121491327939e-05,
      "loss": 0.4662,
      "mean_token_accuracy": 0.8465465784072876,
      "num_tokens": 648582598.0,
      "step": 19360
    },
    {
      "epoch": 1.1559813753581663,
      "grad_norm": 0.47243648767471313,
      "learning_rate": 2.3580279890266848e-05,
      "loss": 0.4417,
      "mean_token_accuracy": 0.8519145846366882,
      "num_tokens": 648750278.0,
      "step": 19365
    },
    {
      "epoch": 1.156279847182426,
      "grad_norm": 0.4737529158592224,
      "learning_rate": 2.3569345821974502e-05,
      "loss": 0.469,
      "mean_token_accuracy": 0.8454610586166382,
      "num_tokens": 648917958.0,
      "step": 19370
    },
    {
      "epoch": 1.1565783190066858,
      "grad_norm": 0.47167718410491943,
      "learning_rate": 2.3558412711065525e-05,
      "loss": 0.4791,
      "mean_token_accuracy": 0.8408207058906555,
      "num_tokens": 649085638.0,
      "step": 19375
    },
    {
      "epoch": 1.1568767908309456,
      "grad_norm": 0.4872191846370697,
      "learning_rate": 2.3547480560202895e-05,
      "loss": 0.4546,
      "mean_token_accuracy": 0.8506620407104493,
      "num_tokens": 649253318.0,
      "step": 19380
    },
    {
      "epoch": 1.1571752626552054,
      "grad_norm": 0.49179843068122864,
      "learning_rate": 2.353654937204932e-05,
      "loss": 0.4719,
      "mean_token_accuracy": 0.8441787004470825,
      "num_tokens": 649420998.0,
      "step": 19385
    },
    {
      "epoch": 1.1574737344794652,
      "grad_norm": 0.5116528868675232,
      "learning_rate": 2.3525619149267303e-05,
      "loss": 0.4881,
      "mean_token_accuracy": 0.8374865889549256,
      "num_tokens": 649588678.0,
      "step": 19390
    },
    {
      "epoch": 1.157772206303725,
      "grad_norm": 0.553030788898468,
      "learning_rate": 2.3514689894519105e-05,
      "loss": 0.496,
      "mean_token_accuracy": 0.8368543386459351,
      "num_tokens": 649756358.0,
      "step": 19395
    },
    {
      "epoch": 1.1580706781279848,
      "grad_norm": 0.5112454295158386,
      "learning_rate": 2.3503761610466742e-05,
      "loss": 0.4356,
      "mean_token_accuracy": 0.854849112033844,
      "num_tokens": 649924038.0,
      "step": 19400
    },
    {
      "epoch": 1.1583691499522446,
      "grad_norm": 0.5680729746818542,
      "learning_rate": 2.3492834299772003e-05,
      "loss": 0.4452,
      "mean_token_accuracy": 0.8531552076339721,
      "num_tokens": 650091718.0,
      "step": 19405
    },
    {
      "epoch": 1.1586676217765044,
      "grad_norm": 0.5384697914123535,
      "learning_rate": 2.3481907965096427e-05,
      "loss": 0.4385,
      "mean_token_accuracy": 0.853483235836029,
      "num_tokens": 650259398.0,
      "step": 19410
    },
    {
      "epoch": 1.158966093600764,
      "grad_norm": 0.5490452647209167,
      "learning_rate": 2.347098260910134e-05,
      "loss": 0.4511,
      "mean_token_accuracy": 0.8497077584266662,
      "num_tokens": 650427078.0,
      "step": 19415
    },
    {
      "epoch": 1.159264565425024,
      "grad_norm": 0.5354728102684021,
      "learning_rate": 2.34600582344478e-05,
      "loss": 0.4774,
      "mean_token_accuracy": 0.8451926469802856,
      "num_tokens": 650594758.0,
      "step": 19420
    },
    {
      "epoch": 1.1595630372492836,
      "grad_norm": 0.571340799331665,
      "learning_rate": 2.3449134843796643e-05,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8448467016220093,
      "num_tokens": 650762438.0,
      "step": 19425
    },
    {
      "epoch": 1.1598615090735436,
      "grad_norm": 0.5150979161262512,
      "learning_rate": 2.3438212439808476e-05,
      "loss": 0.4673,
      "mean_token_accuracy": 0.8454371929168701,
      "num_tokens": 650930118.0,
      "step": 19430
    },
    {
      "epoch": 1.1601599808978031,
      "grad_norm": 0.5014667510986328,
      "learning_rate": 2.3427291025143628e-05,
      "loss": 0.447,
      "mean_token_accuracy": 0.8523857712745666,
      "num_tokens": 651097798.0,
      "step": 19435
    },
    {
      "epoch": 1.1604584527220632,
      "grad_norm": 0.5894905924797058,
      "learning_rate": 2.341637060246224e-05,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8483299493789673,
      "num_tokens": 651265478.0,
      "step": 19440
    },
    {
      "epoch": 1.1607569245463227,
      "grad_norm": 0.5215399265289307,
      "learning_rate": 2.3405451174424152e-05,
      "loss": 0.4499,
      "mean_token_accuracy": 0.8494751334190369,
      "num_tokens": 651433158.0,
      "step": 19445
    },
    {
      "epoch": 1.1610553963705825,
      "grad_norm": 0.552625834941864,
      "learning_rate": 2.3394532743689014e-05,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8379816293716431,
      "num_tokens": 651600838.0,
      "step": 19450
    },
    {
      "epoch": 1.1613538681948423,
      "grad_norm": 0.5501632690429688,
      "learning_rate": 2.3383615312916206e-05,
      "loss": 0.439,
      "mean_token_accuracy": 0.8531134486198425,
      "num_tokens": 651768518.0,
      "step": 19455
    },
    {
      "epoch": 1.1616523400191021,
      "grad_norm": 0.5946309566497803,
      "learning_rate": 2.3372698884764865e-05,
      "loss": 0.4862,
      "mean_token_accuracy": 0.8406656384468079,
      "num_tokens": 651936198.0,
      "step": 19460
    },
    {
      "epoch": 1.161950811843362,
      "grad_norm": 0.5006893277168274,
      "learning_rate": 2.3361783461893894e-05,
      "loss": 0.4093,
      "mean_token_accuracy": 0.8622271299362183,
      "num_tokens": 652103878.0,
      "step": 19465
    },
    {
      "epoch": 1.1622492836676217,
      "grad_norm": 0.49508512020111084,
      "learning_rate": 2.335086904696194e-05,
      "loss": 0.5054,
      "mean_token_accuracy": 0.8347190856933594,
      "num_tokens": 652271558.0,
      "step": 19470
    },
    {
      "epoch": 1.1625477554918815,
      "grad_norm": 0.5250930190086365,
      "learning_rate": 2.3339955642627414e-05,
      "loss": 0.4809,
      "mean_token_accuracy": 0.8424788355827332,
      "num_tokens": 652439238.0,
      "step": 19475
    },
    {
      "epoch": 1.1628462273161413,
      "grad_norm": 0.5129685401916504,
      "learning_rate": 2.3329043251548488e-05,
      "loss": 0.4829,
      "mean_token_accuracy": 0.8413634657859802,
      "num_tokens": 652606918.0,
      "step": 19480
    },
    {
      "epoch": 1.163144699140401,
      "grad_norm": 0.6268259882926941,
      "learning_rate": 2.3318131876383053e-05,
      "loss": 0.497,
      "mean_token_accuracy": 0.8373672842979432,
      "num_tokens": 652774598.0,
      "step": 19485
    },
    {
      "epoch": 1.1634431709646609,
      "grad_norm": 0.575529158115387,
      "learning_rate": 2.3307221519788795e-05,
      "loss": 0.4404,
      "mean_token_accuracy": 0.852856981754303,
      "num_tokens": 652942278.0,
      "step": 19490
    },
    {
      "epoch": 1.1637416427889207,
      "grad_norm": 0.4819105267524719,
      "learning_rate": 2.3296312184423113e-05,
      "loss": 0.4378,
      "mean_token_accuracy": 0.8558332443237304,
      "num_tokens": 653109958.0,
      "step": 19495
    },
    {
      "epoch": 1.1640401146131805,
      "grad_norm": 0.527688205242157,
      "learning_rate": 2.32854038729432e-05,
      "loss": 0.4838,
      "mean_token_accuracy": 0.8400274276733398,
      "num_tokens": 653277638.0,
      "step": 19500
    },
    {
      "epoch": 1.1643385864374403,
      "grad_norm": 0.4638756215572357,
      "learning_rate": 2.3274496588005974e-05,
      "loss": 0.3933,
      "mean_token_accuracy": 0.8679529905319214,
      "num_tokens": 653445318.0,
      "step": 19505
    },
    {
      "epoch": 1.1646370582617,
      "grad_norm": 0.566583514213562,
      "learning_rate": 2.3263590332268082e-05,
      "loss": 0.4832,
      "mean_token_accuracy": 0.8397829055786132,
      "num_tokens": 653612998.0,
      "step": 19510
    },
    {
      "epoch": 1.1649355300859598,
      "grad_norm": 0.5689113140106201,
      "learning_rate": 2.3252685108385967e-05,
      "loss": 0.4448,
      "mean_token_accuracy": 0.8518251180648804,
      "num_tokens": 653780678.0,
      "step": 19515
    },
    {
      "epoch": 1.1652340019102196,
      "grad_norm": 0.5511366724967957,
      "learning_rate": 2.3241780919015793e-05,
      "loss": 0.4742,
      "mean_token_accuracy": 0.843671715259552,
      "num_tokens": 653948358.0,
      "step": 19520
    },
    {
      "epoch": 1.1655324737344794,
      "grad_norm": 0.5243823528289795,
      "learning_rate": 2.3230877766813474e-05,
      "loss": 0.4877,
      "mean_token_accuracy": 0.8391685366630555,
      "num_tokens": 654116038.0,
      "step": 19525
    },
    {
      "epoch": 1.1658309455587392,
      "grad_norm": 0.5512502789497375,
      "learning_rate": 2.3219975654434666e-05,
      "loss": 0.4634,
      "mean_token_accuracy": 0.8468567252159118,
      "num_tokens": 654283718.0,
      "step": 19530
    },
    {
      "epoch": 1.166129417382999,
      "grad_norm": 0.499554306268692,
      "learning_rate": 2.320907458453479e-05,
      "loss": 0.4884,
      "mean_token_accuracy": 0.8369438171386718,
      "num_tokens": 654451398.0,
      "step": 19535
    },
    {
      "epoch": 1.1664278892072588,
      "grad_norm": 0.46907830238342285,
      "learning_rate": 2.319817455976901e-05,
      "loss": 0.4533,
      "mean_token_accuracy": 0.85076345205307,
      "num_tokens": 654619078.0,
      "step": 19540
    },
    {
      "epoch": 1.1667263610315186,
      "grad_norm": 0.5021779537200928,
      "learning_rate": 2.3187275582792206e-05,
      "loss": 0.4319,
      "mean_token_accuracy": 0.8558093786239624,
      "num_tokens": 654786758.0,
      "step": 19545
    },
    {
      "epoch": 1.1670248328557784,
      "grad_norm": 0.5622022747993469,
      "learning_rate": 2.317637765625905e-05,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8429380774497985,
      "num_tokens": 654954438.0,
      "step": 19550
    },
    {
      "epoch": 1.1673233046800382,
      "grad_norm": 0.5311176180839539,
      "learning_rate": 2.3165480782823912e-05,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8349815130233764,
      "num_tokens": 655122118.0,
      "step": 19555
    },
    {
      "epoch": 1.167621776504298,
      "grad_norm": 0.5961161851882935,
      "learning_rate": 2.315458496514093e-05,
      "loss": 0.485,
      "mean_token_accuracy": 0.8389657735824585,
      "num_tokens": 655289798.0,
      "step": 19560
    },
    {
      "epoch": 1.1679202483285578,
      "grad_norm": 0.48575928807258606,
      "learning_rate": 2.3143690205863998e-05,
      "loss": 0.3969,
      "mean_token_accuracy": 0.8668615102767945,
      "num_tokens": 655457478.0,
      "step": 19565
    },
    {
      "epoch": 1.1682187201528176,
      "grad_norm": 0.541350245475769,
      "learning_rate": 2.3132796507646716e-05,
      "loss": 0.4299,
      "mean_token_accuracy": 0.8556185126304626,
      "num_tokens": 655625158.0,
      "step": 19570
    },
    {
      "epoch": 1.1685171919770774,
      "grad_norm": 0.515748143196106,
      "learning_rate": 2.312190387314247e-05,
      "loss": 0.4587,
      "mean_token_accuracy": 0.8487236022949218,
      "num_tokens": 655792838.0,
      "step": 19575
    },
    {
      "epoch": 1.1688156638013372,
      "grad_norm": 0.5054634213447571,
      "learning_rate": 2.3111012305004325e-05,
      "loss": 0.4397,
      "mean_token_accuracy": 0.8530717015266418,
      "num_tokens": 655960518.0,
      "step": 19580
    },
    {
      "epoch": 1.169114135625597,
      "grad_norm": 0.49361228942871094,
      "learning_rate": 2.310012180588515e-05,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8503817200660706,
      "num_tokens": 656128198.0,
      "step": 19585
    },
    {
      "epoch": 1.1694126074498568,
      "grad_norm": 0.5183147192001343,
      "learning_rate": 2.308923237843753e-05,
      "loss": 0.4498,
      "mean_token_accuracy": 0.8505487203598022,
      "num_tokens": 656295878.0,
      "step": 19590
    },
    {
      "epoch": 1.1697110792741165,
      "grad_norm": 0.46185898780822754,
      "learning_rate": 2.3078344025313775e-05,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8518907308578492,
      "num_tokens": 656463558.0,
      "step": 19595
    },
    {
      "epoch": 1.1700095510983763,
      "grad_norm": 0.5042920112609863,
      "learning_rate": 2.3067456749165946e-05,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8442502617835999,
      "num_tokens": 656631238.0,
      "step": 19600
    },
    {
      "epoch": 1.1703080229226361,
      "grad_norm": 0.5162453651428223,
      "learning_rate": 2.305657055264584e-05,
      "loss": 0.4553,
      "mean_token_accuracy": 0.8501014113426208,
      "num_tokens": 656798918.0,
      "step": 19605
    },
    {
      "epoch": 1.170606494746896,
      "grad_norm": 0.578216552734375,
      "learning_rate": 2.304568543840499e-05,
      "loss": 0.455,
      "mean_token_accuracy": 0.8495765209197998,
      "num_tokens": 656966598.0,
      "step": 19610
    },
    {
      "epoch": 1.1709049665711557,
      "grad_norm": 0.5236524343490601,
      "learning_rate": 2.3034801409094675e-05,
      "loss": 0.4887,
      "mean_token_accuracy": 0.8392257809638977,
      "num_tokens": 657133177.0,
      "step": 19615
    },
    {
      "epoch": 1.1712034383954155,
      "grad_norm": 0.5342540144920349,
      "learning_rate": 2.3023918467365886e-05,
      "loss": 0.4308,
      "mean_token_accuracy": 0.8558153510093689,
      "num_tokens": 657300857.0,
      "step": 19620
    },
    {
      "epoch": 1.1715019102196753,
      "grad_norm": 0.5280064344406128,
      "learning_rate": 2.301303661586939e-05,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8483418822288513,
      "num_tokens": 657468537.0,
      "step": 19625
    },
    {
      "epoch": 1.171800382043935,
      "grad_norm": 0.5262241363525391,
      "learning_rate": 2.3002155857255624e-05,
      "loss": 0.4837,
      "mean_token_accuracy": 0.8410354375839233,
      "num_tokens": 657636217.0,
      "step": 19630
    },
    {
      "epoch": 1.172098853868195,
      "grad_norm": 0.4653843343257904,
      "learning_rate": 2.2991276194174838e-05,
      "loss": 0.4332,
      "mean_token_accuracy": 0.8568114042282104,
      "num_tokens": 657803897.0,
      "step": 19635
    },
    {
      "epoch": 1.1723973256924547,
      "grad_norm": 0.5430920124053955,
      "learning_rate": 2.2980397629276944e-05,
      "loss": 0.4203,
      "mean_token_accuracy": 0.8596325993537903,
      "num_tokens": 657971577.0,
      "step": 19640
    },
    {
      "epoch": 1.1726957975167145,
      "grad_norm": 0.5130705833435059,
      "learning_rate": 2.2969520165211634e-05,
      "loss": 0.4327,
      "mean_token_accuracy": 0.8546701669692993,
      "num_tokens": 658139257.0,
      "step": 19645
    },
    {
      "epoch": 1.1729942693409743,
      "grad_norm": 0.547199547290802,
      "learning_rate": 2.2958643804628315e-05,
      "loss": 0.4563,
      "mean_token_accuracy": 0.8479959487915039,
      "num_tokens": 658306937.0,
      "step": 19650
    },
    {
      "epoch": 1.173292741165234,
      "grad_norm": 0.49585190415382385,
      "learning_rate": 2.2947768550176112e-05,
      "loss": 0.4392,
      "mean_token_accuracy": 0.8531909823417664,
      "num_tokens": 658474617.0,
      "step": 19655
    },
    {
      "epoch": 1.1735912129894939,
      "grad_norm": 0.49096888303756714,
      "learning_rate": 2.2936894404503912e-05,
      "loss": 0.4196,
      "mean_token_accuracy": 0.8608910799026489,
      "num_tokens": 658642297.0,
      "step": 19660
    },
    {
      "epoch": 1.1738896848137537,
      "grad_norm": 0.505174994468689,
      "learning_rate": 2.2926021370260305e-05,
      "loss": 0.4436,
      "mean_token_accuracy": 0.8531253814697266,
      "num_tokens": 658809977.0,
      "step": 19665
    },
    {
      "epoch": 1.1741881566380135,
      "grad_norm": 0.4876374304294586,
      "learning_rate": 2.291514945009361e-05,
      "loss": 0.4292,
      "mean_token_accuracy": 0.8567875385284424,
      "num_tokens": 658977657.0,
      "step": 19670
    },
    {
      "epoch": 1.1744866284622733,
      "grad_norm": 0.5102152228355408,
      "learning_rate": 2.2904278646651905e-05,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8490337610244751,
      "num_tokens": 659145337.0,
      "step": 19675
    },
    {
      "epoch": 1.174785100286533,
      "grad_norm": 0.4950414001941681,
      "learning_rate": 2.2893408962582958e-05,
      "loss": 0.4423,
      "mean_token_accuracy": 0.8534116744995117,
      "num_tokens": 659313017.0,
      "step": 19680
    },
    {
      "epoch": 1.1750835721107928,
      "grad_norm": 0.4529719054698944,
      "learning_rate": 2.2882540400534287e-05,
      "loss": 0.4342,
      "mean_token_accuracy": 0.8553143262863159,
      "num_tokens": 659480697.0,
      "step": 19685
    },
    {
      "epoch": 1.1753820439350524,
      "grad_norm": 0.5090801119804382,
      "learning_rate": 2.2871672963153124e-05,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8520517706871032,
      "num_tokens": 659648377.0,
      "step": 19690
    },
    {
      "epoch": 1.1756805157593124,
      "grad_norm": 0.4853980243206024,
      "learning_rate": 2.2860806653086437e-05,
      "loss": 0.4248,
      "mean_token_accuracy": 0.857819402217865,
      "num_tokens": 659816057.0,
      "step": 19695
    },
    {
      "epoch": 1.175978987583572,
      "grad_norm": 0.585038959980011,
      "learning_rate": 2.2849941472980936e-05,
      "loss": 0.4683,
      "mean_token_accuracy": 0.8449839115142822,
      "num_tokens": 659983737.0,
      "step": 19700
    },
    {
      "epoch": 1.176277459407832,
      "grad_norm": 0.48393943905830383,
      "learning_rate": 2.2839077425483003e-05,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8494393467903137,
      "num_tokens": 660151417.0,
      "step": 19705
    },
    {
      "epoch": 1.1765759312320916,
      "grad_norm": 0.49411025643348694,
      "learning_rate": 2.2828214513238806e-05,
      "loss": 0.4392,
      "mean_token_accuracy": 0.8544196486473083,
      "num_tokens": 660319097.0,
      "step": 19710
    },
    {
      "epoch": 1.1768744030563516,
      "grad_norm": 0.5259071588516235,
      "learning_rate": 2.2817352738894187e-05,
      "loss": 0.4819,
      "mean_token_accuracy": 0.8411249041557312,
      "num_tokens": 660486777.0,
      "step": 19715
    },
    {
      "epoch": 1.1771728748806112,
      "grad_norm": 0.515748143196106,
      "learning_rate": 2.280649210509474e-05,
      "loss": 0.4365,
      "mean_token_accuracy": 0.8544196486473083,
      "num_tokens": 660654457.0,
      "step": 19720
    },
    {
      "epoch": 1.177471346704871,
      "grad_norm": 0.5063372254371643,
      "learning_rate": 2.2795632614485777e-05,
      "loss": 0.4715,
      "mean_token_accuracy": 0.842180609703064,
      "num_tokens": 660822137.0,
      "step": 19725
    },
    {
      "epoch": 1.1777698185291308,
      "grad_norm": 0.5847283005714417,
      "learning_rate": 2.2784774269712324e-05,
      "loss": 0.4647,
      "mean_token_accuracy": 0.845765233039856,
      "num_tokens": 660989817.0,
      "step": 19730
    },
    {
      "epoch": 1.1780682903533906,
      "grad_norm": 0.572058379650116,
      "learning_rate": 2.277391707341913e-05,
      "loss": 0.5071,
      "mean_token_accuracy": 0.8348562598228455,
      "num_tokens": 661157497.0,
      "step": 19735
    },
    {
      "epoch": 1.1783667621776504,
      "grad_norm": 0.5968338847160339,
      "learning_rate": 2.2763061028250665e-05,
      "loss": 0.4733,
      "mean_token_accuracy": 0.8430752754211426,
      "num_tokens": 661325177.0,
      "step": 19740
    },
    {
      "epoch": 1.1786652340019101,
      "grad_norm": 0.5082035064697266,
      "learning_rate": 2.2752206136851124e-05,
      "loss": 0.5075,
      "mean_token_accuracy": 0.8324883580207825,
      "num_tokens": 661492857.0,
      "step": 19745
    },
    {
      "epoch": 1.17896370582617,
      "grad_norm": 0.5289940237998962,
      "learning_rate": 2.2741352401864408e-05,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8542049288749695,
      "num_tokens": 661660537.0,
      "step": 19750
    },
    {
      "epoch": 1.1792621776504297,
      "grad_norm": 0.47402361035346985,
      "learning_rate": 2.273049982593415e-05,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8553620457649231,
      "num_tokens": 661828217.0,
      "step": 19755
    },
    {
      "epoch": 1.1795606494746895,
      "grad_norm": 0.5124250650405884,
      "learning_rate": 2.2719648411703698e-05,
      "loss": 0.4506,
      "mean_token_accuracy": 0.8496361613273621,
      "num_tokens": 661995897.0,
      "step": 19760
    },
    {
      "epoch": 1.1798591212989493,
      "grad_norm": 0.5215466022491455,
      "learning_rate": 2.2708798161816103e-05,
      "loss": 0.4312,
      "mean_token_accuracy": 0.8568650841712951,
      "num_tokens": 662163577.0,
      "step": 19765
    },
    {
      "epoch": 1.1801575931232091,
      "grad_norm": 0.518336296081543,
      "learning_rate": 2.2697949078914166e-05,
      "loss": 0.4436,
      "mean_token_accuracy": 0.8518847703933716,
      "num_tokens": 662331257.0,
      "step": 19770
    },
    {
      "epoch": 1.180456064947469,
      "grad_norm": 0.8242446184158325,
      "learning_rate": 2.2687101165640347e-05,
      "loss": 0.4625,
      "mean_token_accuracy": 0.8481629490852356,
      "num_tokens": 662498937.0,
      "step": 19775
    },
    {
      "epoch": 1.1807545367717287,
      "grad_norm": 0.521611213684082,
      "learning_rate": 2.2676254424636884e-05,
      "loss": 0.4297,
      "mean_token_accuracy": 0.8576404571533203,
      "num_tokens": 662666617.0,
      "step": 19780
    },
    {
      "epoch": 1.1810530085959885,
      "grad_norm": 0.5193411707878113,
      "learning_rate": 2.2665408858545694e-05,
      "loss": 0.441,
      "mean_token_accuracy": 0.8525885820388794,
      "num_tokens": 662834297.0,
      "step": 19785
    },
    {
      "epoch": 1.1813514804202483,
      "grad_norm": 0.5576213002204895,
      "learning_rate": 2.265456447000841e-05,
      "loss": 0.4523,
      "mean_token_accuracy": 0.8498926401138306,
      "num_tokens": 663001977.0,
      "step": 19790
    },
    {
      "epoch": 1.181649952244508,
      "grad_norm": 0.47739002108573914,
      "learning_rate": 2.264372126166639e-05,
      "loss": 0.4654,
      "mean_token_accuracy": 0.8471072316169739,
      "num_tokens": 663169657.0,
      "step": 19795
    },
    {
      "epoch": 1.1819484240687679,
      "grad_norm": 0.4830189645290375,
      "learning_rate": 2.263287923616069e-05,
      "loss": 0.4629,
      "mean_token_accuracy": 0.8477036833763123,
      "num_tokens": 663337337.0,
      "step": 19800
    },
    {
      "epoch": 1.1822468958930277,
      "grad_norm": 0.5114553570747375,
      "learning_rate": 2.262203839613209e-05,
      "loss": 0.4389,
      "mean_token_accuracy": 0.8536621570587158,
      "num_tokens": 663505017.0,
      "step": 19805
    },
    {
      "epoch": 1.1825453677172875,
      "grad_norm": 0.46853530406951904,
      "learning_rate": 2.261119874422108e-05,
      "loss": 0.4766,
      "mean_token_accuracy": 0.8436239957809448,
      "num_tokens": 663672697.0,
      "step": 19810
    },
    {
      "epoch": 1.1828438395415473,
      "grad_norm": 0.5454878807067871,
      "learning_rate": 2.260036028306784e-05,
      "loss": 0.4783,
      "mean_token_accuracy": 0.841315758228302,
      "num_tokens": 663840377.0,
      "step": 19815
    },
    {
      "epoch": 1.183142311365807,
      "grad_norm": 0.5502786636352539,
      "learning_rate": 2.2589523015312303e-05,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8417929172515869,
      "num_tokens": 664008057.0,
      "step": 19820
    },
    {
      "epoch": 1.1834407831900668,
      "grad_norm": 0.5025622844696045,
      "learning_rate": 2.257868694359407e-05,
      "loss": 0.4694,
      "mean_token_accuracy": 0.8456996321678162,
      "num_tokens": 664175737.0,
      "step": 19825
    },
    {
      "epoch": 1.1837392550143266,
      "grad_norm": 0.4925752580165863,
      "learning_rate": 2.2567852070552457e-05,
      "loss": 0.455,
      "mean_token_accuracy": 0.8487653613090516,
      "num_tokens": 664343417.0,
      "step": 19830
    },
    {
      "epoch": 1.1840377268385864,
      "grad_norm": 0.5200989246368408,
      "learning_rate": 2.2557018398826517e-05,
      "loss": 0.4664,
      "mean_token_accuracy": 0.8442025542259216,
      "num_tokens": 664511097.0,
      "step": 19835
    },
    {
      "epoch": 1.1843361986628462,
      "grad_norm": 0.4849730134010315,
      "learning_rate": 2.2546185931054987e-05,
      "loss": 0.423,
      "mean_token_accuracy": 0.8579267621040344,
      "num_tokens": 664678777.0,
      "step": 19840
    },
    {
      "epoch": 1.184634670487106,
      "grad_norm": 0.45852211117744446,
      "learning_rate": 2.2535354669876306e-05,
      "loss": 0.4164,
      "mean_token_accuracy": 0.8617380380630493,
      "num_tokens": 664846457.0,
      "step": 19845
    },
    {
      "epoch": 1.1849331423113658,
      "grad_norm": 0.5718573927879333,
      "learning_rate": 2.2524524617928638e-05,
      "loss": 0.4885,
      "mean_token_accuracy": 0.8376297235488892,
      "num_tokens": 665014137.0,
      "step": 19850
    },
    {
      "epoch": 1.1852316141356256,
      "grad_norm": 0.5277978777885437,
      "learning_rate": 2.2513695777849834e-05,
      "loss": 0.4728,
      "mean_token_accuracy": 0.8458845257759094,
      "num_tokens": 665181817.0,
      "step": 19855
    },
    {
      "epoch": 1.1855300859598854,
      "grad_norm": 0.4968494772911072,
      "learning_rate": 2.2502868152277456e-05,
      "loss": 0.4092,
      "mean_token_accuracy": 0.8622211813926697,
      "num_tokens": 665349497.0,
      "step": 19860
    },
    {
      "epoch": 1.1858285577841452,
      "grad_norm": 0.5447377562522888,
      "learning_rate": 2.2492041743848776e-05,
      "loss": 0.4588,
      "mean_token_accuracy": 0.8484253764152527,
      "num_tokens": 665517177.0,
      "step": 19865
    },
    {
      "epoch": 1.186127029608405,
      "grad_norm": 0.5029953122138977,
      "learning_rate": 2.2481216555200774e-05,
      "loss": 0.4831,
      "mean_token_accuracy": 0.8423774242401123,
      "num_tokens": 665684857.0,
      "step": 19870
    },
    {
      "epoch": 1.1864255014326648,
      "grad_norm": 0.5610088109970093,
      "learning_rate": 2.247039258897011e-05,
      "loss": 0.4364,
      "mean_token_accuracy": 0.853972327709198,
      "num_tokens": 665852537.0,
      "step": 19875
    },
    {
      "epoch": 1.1867239732569246,
      "grad_norm": 0.5005916953086853,
      "learning_rate": 2.2459569847793166e-05,
      "loss": 0.4308,
      "mean_token_accuracy": 0.8550638198852539,
      "num_tokens": 666020217.0,
      "step": 19880
    },
    {
      "epoch": 1.1870224450811844,
      "grad_norm": 0.5791324973106384,
      "learning_rate": 2.244874833430602e-05,
      "loss": 0.4541,
      "mean_token_accuracy": 0.8501073598861695,
      "num_tokens": 666187897.0,
      "step": 19885
    },
    {
      "epoch": 1.1873209169054442,
      "grad_norm": 0.4910011291503906,
      "learning_rate": 2.2437928051144446e-05,
      "loss": 0.4091,
      "mean_token_accuracy": 0.8643385529518127,
      "num_tokens": 666355577.0,
      "step": 19890
    },
    {
      "epoch": 1.187619388729704,
      "grad_norm": 0.49433910846710205,
      "learning_rate": 2.2427109000943946e-05,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8510557055473328,
      "num_tokens": 666523257.0,
      "step": 19895
    },
    {
      "epoch": 1.1879178605539638,
      "grad_norm": 0.500910758972168,
      "learning_rate": 2.2416291186339672e-05,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8530776619911193,
      "num_tokens": 666690937.0,
      "step": 19900
    },
    {
      "epoch": 1.1882163323782235,
      "grad_norm": 0.48798131942749023,
      "learning_rate": 2.2405474609966524e-05,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8563342452049255,
      "num_tokens": 666858617.0,
      "step": 19905
    },
    {
      "epoch": 1.1885148042024833,
      "grad_norm": 0.46963825821876526,
      "learning_rate": 2.2394659274459057e-05,
      "loss": 0.437,
      "mean_token_accuracy": 0.8557378053665161,
      "num_tokens": 667026297.0,
      "step": 19910
    },
    {
      "epoch": 1.1888132760267431,
      "grad_norm": 0.5557814240455627,
      "learning_rate": 2.238384518245156e-05,
      "loss": 0.4292,
      "mean_token_accuracy": 0.8553560853004456,
      "num_tokens": 667193977.0,
      "step": 19915
    },
    {
      "epoch": 1.189111747851003,
      "grad_norm": 0.4949130415916443,
      "learning_rate": 2.237303233657801e-05,
      "loss": 0.4284,
      "mean_token_accuracy": 0.8582070946693421,
      "num_tokens": 667361657.0,
      "step": 19920
    },
    {
      "epoch": 1.1894102196752627,
      "grad_norm": 0.494113028049469,
      "learning_rate": 2.236222073947206e-05,
      "loss": 0.4411,
      "mean_token_accuracy": 0.8533460617065429,
      "num_tokens": 667529337.0,
      "step": 19925
    },
    {
      "epoch": 1.1897086914995225,
      "grad_norm": 0.47995102405548096,
      "learning_rate": 2.2351410393767092e-05,
      "loss": 0.4214,
      "mean_token_accuracy": 0.8601097345352173,
      "num_tokens": 667697017.0,
      "step": 19930
    },
    {
      "epoch": 1.1900071633237823,
      "grad_norm": 0.5150399804115295,
      "learning_rate": 2.2340601302096147e-05,
      "loss": 0.4693,
      "mean_token_accuracy": 0.8455803394317627,
      "num_tokens": 667864697.0,
      "step": 19935
    },
    {
      "epoch": 1.190305635148042,
      "grad_norm": 0.5533705949783325,
      "learning_rate": 2.2329793467091987e-05,
      "loss": 0.4451,
      "mean_token_accuracy": 0.8515030384063721,
      "num_tokens": 668032377.0,
      "step": 19940
    },
    {
      "epoch": 1.190604106972302,
      "grad_norm": 0.5789175033569336,
      "learning_rate": 2.2318986891387054e-05,
      "loss": 0.4662,
      "mean_token_accuracy": 0.8456566810607911,
      "num_tokens": 668192213.0,
      "step": 19945
    },
    {
      "epoch": 1.1909025787965617,
      "grad_norm": 0.5290037989616394,
      "learning_rate": 2.23081815776135e-05,
      "loss": 0.4766,
      "mean_token_accuracy": 0.8415423989295959,
      "num_tokens": 668359893.0,
      "step": 19950
    },
    {
      "epoch": 1.1912010506208215,
      "grad_norm": 0.500445544719696,
      "learning_rate": 2.229737752840316e-05,
      "loss": 0.4528,
      "mean_token_accuracy": 0.8498270153999329,
      "num_tokens": 668527573.0,
      "step": 19955
    },
    {
      "epoch": 1.1914995224450813,
      "grad_norm": 0.5138988494873047,
      "learning_rate": 2.2286574746387534e-05,
      "loss": 0.4824,
      "mean_token_accuracy": 0.8410235047340393,
      "num_tokens": 668695253.0,
      "step": 19960
    },
    {
      "epoch": 1.1917979942693409,
      "grad_norm": 0.5639352202415466,
      "learning_rate": 2.2275773234197878e-05,
      "loss": 0.4822,
      "mean_token_accuracy": 0.8398067593574524,
      "num_tokens": 668862933.0,
      "step": 19965
    },
    {
      "epoch": 1.1920964660936009,
      "grad_norm": 0.5014837384223938,
      "learning_rate": 2.226497299446506e-05,
      "loss": 0.4456,
      "mean_token_accuracy": 0.8509483456611633,
      "num_tokens": 669030613.0,
      "step": 19970
    },
    {
      "epoch": 1.1923949379178604,
      "grad_norm": 0.5052182674407959,
      "learning_rate": 2.2254174029819703e-05,
      "loss": 0.4733,
      "mean_token_accuracy": 0.8446021676063538,
      "num_tokens": 669198293.0,
      "step": 19975
    },
    {
      "epoch": 1.1926934097421205,
      "grad_norm": 0.5078866481781006,
      "learning_rate": 2.22433763428921e-05,
      "loss": 0.4802,
      "mean_token_accuracy": 0.8433019280433655,
      "num_tokens": 669365973.0,
      "step": 19980
    },
    {
      "epoch": 1.19299188156638,
      "grad_norm": 0.5302884578704834,
      "learning_rate": 2.22325799363122e-05,
      "loss": 0.4328,
      "mean_token_accuracy": 0.8557437539100647,
      "num_tokens": 669533653.0,
      "step": 19985
    },
    {
      "epoch": 1.19329035339064,
      "grad_norm": 0.5151789784431458,
      "learning_rate": 2.222178481270969e-05,
      "loss": 0.467,
      "mean_token_accuracy": 0.8452582597732544,
      "num_tokens": 669701333.0,
      "step": 19990
    },
    {
      "epoch": 1.1935888252148996,
      "grad_norm": 0.5370619893074036,
      "learning_rate": 2.2210990974713908e-05,
      "loss": 0.4785,
      "mean_token_accuracy": 0.8420016765594482,
      "num_tokens": 669869013.0,
      "step": 19995
    },
    {
      "epoch": 1.1938872970391594,
      "grad_norm": 0.4923109710216522,
      "learning_rate": 2.2200198424953894e-05,
      "loss": 0.4457,
      "mean_token_accuracy": 0.8509841322898865,
      "num_tokens": 670036693.0,
      "step": 20000
    },
    {
      "epoch": 1.1941857688634192,
      "grad_norm": 0.47648292779922485,
      "learning_rate": 2.218940716605839e-05,
      "loss": 0.4245,
      "mean_token_accuracy": 0.8580043077468872,
      "num_tokens": 670204373.0,
      "step": 20005
    },
    {
      "epoch": 1.194484240687679,
      "grad_norm": 0.4782179892063141,
      "learning_rate": 2.217861720065578e-05,
      "loss": 0.465,
      "mean_token_accuracy": 0.8462722182273865,
      "num_tokens": 670372053.0,
      "step": 20010
    },
    {
      "epoch": 1.1947827125119388,
      "grad_norm": 0.5209377408027649,
      "learning_rate": 2.2167828531374184e-05,
      "loss": 0.4291,
      "mean_token_accuracy": 0.8558869123458862,
      "num_tokens": 670539733.0,
      "step": 20015
    },
    {
      "epoch": 1.1950811843361986,
      "grad_norm": 0.5709136128425598,
      "learning_rate": 2.2157041160841362e-05,
      "loss": 0.4818,
      "mean_token_accuracy": 0.8399618268013,
      "num_tokens": 670707413.0,
      "step": 20020
    },
    {
      "epoch": 1.1953796561604584,
      "grad_norm": 0.5084706544876099,
      "learning_rate": 2.2146255091684786e-05,
      "loss": 0.4926,
      "mean_token_accuracy": 0.8385482549667358,
      "num_tokens": 670875093.0,
      "step": 20025
    },
    {
      "epoch": 1.1956781279847182,
      "grad_norm": 0.4596470296382904,
      "learning_rate": 2.213547032653161e-05,
      "loss": 0.4384,
      "mean_token_accuracy": 0.8529046893119812,
      "num_tokens": 671042773.0,
      "step": 20030
    },
    {
      "epoch": 1.195976599808978,
      "grad_norm": 0.49420779943466187,
      "learning_rate": 2.2124686868008638e-05,
      "loss": 0.4143,
      "mean_token_accuracy": 0.8603900790214538,
      "num_tokens": 671210453.0,
      "step": 20035
    },
    {
      "epoch": 1.1962750716332378,
      "grad_norm": 0.5122491121292114,
      "learning_rate": 2.211390471874241e-05,
      "loss": 0.4536,
      "mean_token_accuracy": 0.8490934014320374,
      "num_tokens": 671378133.0,
      "step": 20040
    },
    {
      "epoch": 1.1965735434574976,
      "grad_norm": 0.4651281237602234,
      "learning_rate": 2.2103123881359094e-05,
      "loss": 0.4694,
      "mean_token_accuracy": 0.8450912594795227,
      "num_tokens": 671545813.0,
      "step": 20045
    },
    {
      "epoch": 1.1968720152817574,
      "grad_norm": 0.5637299418449402,
      "learning_rate": 2.2092344358484586e-05,
      "loss": 0.4944,
      "mean_token_accuracy": 0.8379220008850098,
      "num_tokens": 671713493.0,
      "step": 20050
    },
    {
      "epoch": 1.1971704871060171,
      "grad_norm": 0.5713850259780884,
      "learning_rate": 2.2081566152744404e-05,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8296731472015381,
      "num_tokens": 671881173.0,
      "step": 20055
    },
    {
      "epoch": 1.197468958930277,
      "grad_norm": 0.5393882989883423,
      "learning_rate": 2.207078926676381e-05,
      "loss": 0.5106,
      "mean_token_accuracy": 0.8318620920181274,
      "num_tokens": 672048853.0,
      "step": 20060
    },
    {
      "epoch": 1.1977674307545367,
      "grad_norm": 0.4390849173069,
      "learning_rate": 2.20600137031677e-05,
      "loss": 0.4228,
      "mean_token_accuracy": 0.8580997228622437,
      "num_tokens": 672216533.0,
      "step": 20065
    },
    {
      "epoch": 1.1980659025787965,
      "grad_norm": 0.49061521887779236,
      "learning_rate": 2.2049239464580664e-05,
      "loss": 0.4336,
      "mean_token_accuracy": 0.8548789143562316,
      "num_tokens": 672384213.0,
      "step": 20070
    },
    {
      "epoch": 1.1983643744030563,
      "grad_norm": 0.525015115737915,
      "learning_rate": 2.2038466553626965e-05,
      "loss": 0.4541,
      "mean_token_accuracy": 0.848759388923645,
      "num_tokens": 672551893.0,
      "step": 20075
    },
    {
      "epoch": 1.1986628462273161,
      "grad_norm": 0.5270291566848755,
      "learning_rate": 2.202769497293054e-05,
      "loss": 0.4355,
      "mean_token_accuracy": 0.854109525680542,
      "num_tokens": 672719573.0,
      "step": 20080
    },
    {
      "epoch": 1.198961318051576,
      "grad_norm": 0.5062497854232788,
      "learning_rate": 2.201692472511501e-05,
      "loss": 0.4594,
      "mean_token_accuracy": 0.8487057089805603,
      "num_tokens": 672887253.0,
      "step": 20085
    },
    {
      "epoch": 1.1992597898758357,
      "grad_norm": 0.6046461462974548,
      "learning_rate": 2.2006155812803685e-05,
      "loss": 0.4172,
      "mean_token_accuracy": 0.8603602647781372,
      "num_tokens": 673054933.0,
      "step": 20090
    },
    {
      "epoch": 1.1995582617000955,
      "grad_norm": 0.49185025691986084,
      "learning_rate": 2.19953882386195e-05,
      "loss": 0.4564,
      "mean_token_accuracy": 0.8482404947280884,
      "num_tokens": 673222613.0,
      "step": 20095
    },
    {
      "epoch": 1.1998567335243553,
      "grad_norm": 0.576162576675415,
      "learning_rate": 2.198462200518513e-05,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8519980907440186,
      "num_tokens": 673390293.0,
      "step": 20100
    },
    {
      "epoch": 1.200155205348615,
      "grad_norm": 0.4832688570022583,
      "learning_rate": 2.1973857115122852e-05,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8535011410713196,
      "num_tokens": 673557973.0,
      "step": 20105
    },
    {
      "epoch": 1.2004536771728749,
      "grad_norm": 0.5112908482551575,
      "learning_rate": 2.196309357105469e-05,
      "loss": 0.4356,
      "mean_token_accuracy": 0.8550996065139771,
      "num_tokens": 673725653.0,
      "step": 20110
    },
    {
      "epoch": 1.2007521489971347,
      "grad_norm": 0.5340437293052673,
      "learning_rate": 2.1952331375602288e-05,
      "loss": 0.4698,
      "mean_token_accuracy": 0.844399380683899,
      "num_tokens": 673893333.0,
      "step": 20115
    },
    {
      "epoch": 1.2010506208213945,
      "grad_norm": 0.4853335916996002,
      "learning_rate": 2.1941570531386968e-05,
      "loss": 0.4379,
      "mean_token_accuracy": 0.8539126873016357,
      "num_tokens": 674061013.0,
      "step": 20120
    },
    {
      "epoch": 1.2013490926456543,
      "grad_norm": 0.5637503266334534,
      "learning_rate": 2.1930811041029754e-05,
      "loss": 0.4887,
      "mean_token_accuracy": 0.8378205895423889,
      "num_tokens": 674228693.0,
      "step": 20125
    },
    {
      "epoch": 1.201647564469914,
      "grad_norm": 0.500538170337677,
      "learning_rate": 2.1920052907151296e-05,
      "loss": 0.4452,
      "mean_token_accuracy": 0.852493143081665,
      "num_tokens": 674396373.0,
      "step": 20130
    },
    {
      "epoch": 1.2019460362941738,
      "grad_norm": 0.48072803020477295,
      "learning_rate": 2.1909296132371947e-05,
      "loss": 0.4306,
      "mean_token_accuracy": 0.8552845120429993,
      "num_tokens": 674564053.0,
      "step": 20135
    },
    {
      "epoch": 1.2022445081184336,
      "grad_norm": 0.5291653275489807,
      "learning_rate": 2.1898540719311722e-05,
      "loss": 0.4813,
      "mean_token_accuracy": 0.840450918674469,
      "num_tokens": 674731733.0,
      "step": 20140
    },
    {
      "epoch": 1.2025429799426934,
      "grad_norm": 0.5230559706687927,
      "learning_rate": 2.1887786670590295e-05,
      "loss": 0.4519,
      "mean_token_accuracy": 0.8490635871887207,
      "num_tokens": 674899413.0,
      "step": 20145
    },
    {
      "epoch": 1.2028414517669532,
      "grad_norm": 0.5714749097824097,
      "learning_rate": 2.1877033988827016e-05,
      "loss": 0.4559,
      "mean_token_accuracy": 0.8475366830825806,
      "num_tokens": 675067093.0,
      "step": 20150
    },
    {
      "epoch": 1.203139923591213,
      "grad_norm": 0.566028356552124,
      "learning_rate": 2.186628267664089e-05,
      "loss": 0.4773,
      "mean_token_accuracy": 0.842067277431488,
      "num_tokens": 675234773.0,
      "step": 20155
    },
    {
      "epoch": 1.2034383954154728,
      "grad_norm": 0.5234898924827576,
      "learning_rate": 2.1855532736650617e-05,
      "loss": 0.4304,
      "mean_token_accuracy": 0.8562745928764344,
      "num_tokens": 675402453.0,
      "step": 20160
    },
    {
      "epoch": 1.2037368672397326,
      "grad_norm": 0.5434269905090332,
      "learning_rate": 2.1844784171474526e-05,
      "loss": 0.4243,
      "mean_token_accuracy": 0.8572945237159729,
      "num_tokens": 675570133.0,
      "step": 20165
    },
    {
      "epoch": 1.2040353390639924,
      "grad_norm": 0.5061639547348022,
      "learning_rate": 2.183403698373062e-05,
      "loss": 0.4515,
      "mean_token_accuracy": 0.8479184031486511,
      "num_tokens": 675737813.0,
      "step": 20170
    },
    {
      "epoch": 1.2043338108882522,
      "grad_norm": 0.5039093494415283,
      "learning_rate": 2.1823291176036608e-05,
      "loss": 0.4717,
      "mean_token_accuracy": 0.8431229829788208,
      "num_tokens": 675905493.0,
      "step": 20175
    },
    {
      "epoch": 1.204632282712512,
      "grad_norm": 0.4952000081539154,
      "learning_rate": 2.1812546751009793e-05,
      "loss": 0.4272,
      "mean_token_accuracy": 0.8578253746032715,
      "num_tokens": 676073173.0,
      "step": 20180
    },
    {
      "epoch": 1.2049307545367718,
      "grad_norm": 0.5139133930206299,
      "learning_rate": 2.1801803711267203e-05,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8558630585670471,
      "num_tokens": 676240853.0,
      "step": 20185
    },
    {
      "epoch": 1.2052292263610316,
      "grad_norm": 0.4910508692264557,
      "learning_rate": 2.17910620594255e-05,
      "loss": 0.4216,
      "mean_token_accuracy": 0.8602051734924316,
      "num_tokens": 676408533.0,
      "step": 20190
    },
    {
      "epoch": 1.2055276981852914,
      "grad_norm": 0.4882889986038208,
      "learning_rate": 2.1780321798101e-05,
      "loss": 0.4192,
      "mean_token_accuracy": 0.8586782813072205,
      "num_tokens": 676576213.0,
      "step": 20195
    },
    {
      "epoch": 1.2058261700095512,
      "grad_norm": 0.47544705867767334,
      "learning_rate": 2.1769582929909703e-05,
      "loss": 0.4643,
      "mean_token_accuracy": 0.8458606719970703,
      "num_tokens": 676743893.0,
      "step": 20200
    },
    {
      "epoch": 1.206124641833811,
      "grad_norm": 0.5053207874298096,
      "learning_rate": 2.1758845457467254e-05,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8504055857658386,
      "num_tokens": 676911573.0,
      "step": 20205
    },
    {
      "epoch": 1.2064231136580708,
      "grad_norm": 0.5221449136734009,
      "learning_rate": 2.174810938338897e-05,
      "loss": 0.4339,
      "mean_token_accuracy": 0.8546045422554016,
      "num_tokens": 677079253.0,
      "step": 20210
    },
    {
      "epoch": 1.2067215854823305,
      "grad_norm": 0.4730874300003052,
      "learning_rate": 2.1737374710289803e-05,
      "loss": 0.4402,
      "mean_token_accuracy": 0.8543063282966614,
      "num_tokens": 677246933.0,
      "step": 20215
    },
    {
      "epoch": 1.2070200573065903,
      "grad_norm": 0.531874418258667,
      "learning_rate": 2.172664144078439e-05,
      "loss": 0.4346,
      "mean_token_accuracy": 0.8557795524597168,
      "num_tokens": 677414613.0,
      "step": 20220
    },
    {
      "epoch": 1.2073185291308501,
      "grad_norm": 0.4817505478858948,
      "learning_rate": 2.171590957748703e-05,
      "loss": 0.4351,
      "mean_token_accuracy": 0.855964457988739,
      "num_tokens": 677582293.0,
      "step": 20225
    },
    {
      "epoch": 1.20761700095511,
      "grad_norm": 0.5840501189231873,
      "learning_rate": 2.1705179123011643e-05,
      "loss": 0.5032,
      "mean_token_accuracy": 0.8334307670593262,
      "num_tokens": 677749973.0,
      "step": 20230
    },
    {
      "epoch": 1.2079154727793697,
      "grad_norm": 0.6456384658813477,
      "learning_rate": 2.169445007997185e-05,
      "loss": 0.4287,
      "mean_token_accuracy": 0.8555767655372619,
      "num_tokens": 677917653.0,
      "step": 20235
    },
    {
      "epoch": 1.2082139446036293,
      "grad_norm": 0.5162643790245056,
      "learning_rate": 2.168372245098089e-05,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8411666512489319,
      "num_tokens": 678085333.0,
      "step": 20240
    },
    {
      "epoch": 1.2085124164278893,
      "grad_norm": 0.5398520231246948,
      "learning_rate": 2.1672996238651682e-05,
      "loss": 0.4724,
      "mean_token_accuracy": 0.8448705792427063,
      "num_tokens": 678253013.0,
      "step": 20245
    },
    {
      "epoch": 1.2088108882521489,
      "grad_norm": 0.5212766528129578,
      "learning_rate": 2.1662271445596805e-05,
      "loss": 0.433,
      "mean_token_accuracy": 0.8571752548217774,
      "num_tokens": 678420693.0,
      "step": 20250
    },
    {
      "epoch": 1.209109360076409,
      "grad_norm": 0.535269558429718,
      "learning_rate": 2.1651548074428462e-05,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8452463388442993,
      "num_tokens": 678588373.0,
      "step": 20255
    },
    {
      "epoch": 1.2094078319006685,
      "grad_norm": 0.49153932929039,
      "learning_rate": 2.1640826127758544e-05,
      "loss": 0.4459,
      "mean_token_accuracy": 0.8492484807968139,
      "num_tokens": 678756053.0,
      "step": 20260
    },
    {
      "epoch": 1.2097063037249285,
      "grad_norm": 0.5382280349731445,
      "learning_rate": 2.1630105608198566e-05,
      "loss": 0.483,
      "mean_token_accuracy": 0.8415304780006408,
      "num_tokens": 678923733.0,
      "step": 20265
    },
    {
      "epoch": 1.210004775549188,
      "grad_norm": 0.5323084592819214,
      "learning_rate": 2.1619386518359708e-05,
      "loss": 0.4817,
      "mean_token_accuracy": 0.8401944398880005,
      "num_tokens": 679091413.0,
      "step": 20270
    },
    {
      "epoch": 1.2103032473734479,
      "grad_norm": 0.46458926796913147,
      "learning_rate": 2.160866886085281e-05,
      "loss": 0.4369,
      "mean_token_accuracy": 0.8556722044944763,
      "num_tokens": 679259093.0,
      "step": 20275
    },
    {
      "epoch": 1.2106017191977076,
      "grad_norm": 0.5003271698951721,
      "learning_rate": 2.159795263828835e-05,
      "loss": 0.4318,
      "mean_token_accuracy": 0.8547119259834289,
      "num_tokens": 679426773.0,
      "step": 20280
    },
    {
      "epoch": 1.2109001910219674,
      "grad_norm": 0.5180540680885315,
      "learning_rate": 2.158723785327647e-05,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8484253883361816,
      "num_tokens": 679594453.0,
      "step": 20285
    },
    {
      "epoch": 1.2111986628462272,
      "grad_norm": 0.5006572604179382,
      "learning_rate": 2.1576524508426944e-05,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8442502737045288,
      "num_tokens": 679762133.0,
      "step": 20290
    },
    {
      "epoch": 1.211497134670487,
      "grad_norm": 0.4925740361213684,
      "learning_rate": 2.1565812606349212e-05,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8501490950584412,
      "num_tokens": 679929813.0,
      "step": 20295
    },
    {
      "epoch": 1.2117956064947468,
      "grad_norm": 0.5071954131126404,
      "learning_rate": 2.1555102149652346e-05,
      "loss": 0.4499,
      "mean_token_accuracy": 0.8497733473777771,
      "num_tokens": 680097493.0,
      "step": 20300
    },
    {
      "epoch": 1.2120940783190066,
      "grad_norm": 0.5857788920402527,
      "learning_rate": 2.154439314094508e-05,
      "loss": 0.4644,
      "mean_token_accuracy": 0.8469462037086487,
      "num_tokens": 680265173.0,
      "step": 20305
    },
    {
      "epoch": 1.2123925501432664,
      "grad_norm": 0.5620033144950867,
      "learning_rate": 2.15336855828358e-05,
      "loss": 0.4646,
      "mean_token_accuracy": 0.8460574984550476,
      "num_tokens": 680432853.0,
      "step": 20310
    },
    {
      "epoch": 1.2126910219675262,
      "grad_norm": 0.4749830961227417,
      "learning_rate": 2.152297947793252e-05,
      "loss": 0.4384,
      "mean_token_accuracy": 0.8538947939872742,
      "num_tokens": 680600533.0,
      "step": 20315
    },
    {
      "epoch": 1.212989493791786,
      "grad_norm": 0.4631313383579254,
      "learning_rate": 2.1512274828842914e-05,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8555290341377259,
      "num_tokens": 680768213.0,
      "step": 20320
    },
    {
      "epoch": 1.2132879656160458,
      "grad_norm": 0.5602476000785828,
      "learning_rate": 2.150157163817429e-05,
      "loss": 0.454,
      "mean_token_accuracy": 0.8492067337036133,
      "num_tokens": 680935893.0,
      "step": 20325
    },
    {
      "epoch": 1.2135864374403056,
      "grad_norm": 0.5524552464485168,
      "learning_rate": 2.149086990853361e-05,
      "loss": 0.4615,
      "mean_token_accuracy": 0.8464928984642028,
      "num_tokens": 681103573.0,
      "step": 20330
    },
    {
      "epoch": 1.2138849092645654,
      "grad_norm": 0.5774242281913757,
      "learning_rate": 2.1480169642527488e-05,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8492007613182068,
      "num_tokens": 681271253.0,
      "step": 20335
    },
    {
      "epoch": 1.2141833810888252,
      "grad_norm": 0.5189878940582275,
      "learning_rate": 2.1469470842762153e-05,
      "loss": 0.4459,
      "mean_token_accuracy": 0.8516998648643493,
      "num_tokens": 681438933.0,
      "step": 20340
    },
    {
      "epoch": 1.214481852913085,
      "grad_norm": 0.48288533091545105,
      "learning_rate": 2.1458773511843516e-05,
      "loss": 0.4893,
      "mean_token_accuracy": 0.8384528279304504,
      "num_tokens": 681606613.0,
      "step": 20345
    },
    {
      "epoch": 1.2147803247373448,
      "grad_norm": 0.4914621412754059,
      "learning_rate": 2.144807765237709e-05,
      "loss": 0.4533,
      "mean_token_accuracy": 0.848133134841919,
      "num_tokens": 681774293.0,
      "step": 20350
    },
    {
      "epoch": 1.2150787965616046,
      "grad_norm": 0.5602121949195862,
      "learning_rate": 2.1437383266968054e-05,
      "loss": 0.4552,
      "mean_token_accuracy": 0.8499940514564515,
      "num_tokens": 681941973.0,
      "step": 20355
    },
    {
      "epoch": 1.2153772683858644,
      "grad_norm": 0.5511289238929749,
      "learning_rate": 2.1426690358221237e-05,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8555827379226685,
      "num_tokens": 682109653.0,
      "step": 20360
    },
    {
      "epoch": 1.2156757402101241,
      "grad_norm": 0.5459198951721191,
      "learning_rate": 2.141599892874107e-05,
      "loss": 0.4412,
      "mean_token_accuracy": 0.8522485852241516,
      "num_tokens": 682277333.0,
      "step": 20365
    },
    {
      "epoch": 1.215974212034384,
      "grad_norm": 0.5703451633453369,
      "learning_rate": 2.140530898113168e-05,
      "loss": 0.4896,
      "mean_token_accuracy": 0.8384468674659729,
      "num_tokens": 682445013.0,
      "step": 20370
    },
    {
      "epoch": 1.2162726838586437,
      "grad_norm": 0.5362686514854431,
      "learning_rate": 2.1394620517996766e-05,
      "loss": 0.4403,
      "mean_token_accuracy": 0.8533162355422974,
      "num_tokens": 682612693.0,
      "step": 20375
    },
    {
      "epoch": 1.2165711556829035,
      "grad_norm": 0.5646585822105408,
      "learning_rate": 2.1383933541939734e-05,
      "loss": 0.4701,
      "mean_token_accuracy": 0.8428307294845581,
      "num_tokens": 682780373.0,
      "step": 20380
    },
    {
      "epoch": 1.2168696275071633,
      "grad_norm": 0.5130622386932373,
      "learning_rate": 2.137324805556356e-05,
      "loss": 0.4715,
      "mean_token_accuracy": 0.8437194347381591,
      "num_tokens": 682948053.0,
      "step": 20385
    },
    {
      "epoch": 1.2171680993314231,
      "grad_norm": 0.49547046422958374,
      "learning_rate": 2.136256406147092e-05,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8436359167098999,
      "num_tokens": 683115733.0,
      "step": 20390
    },
    {
      "epoch": 1.217466571155683,
      "grad_norm": 0.5067601799964905,
      "learning_rate": 2.1351881562264088e-05,
      "loss": 0.441,
      "mean_token_accuracy": 0.8517953038215638,
      "num_tokens": 683283413.0,
      "step": 20395
    },
    {
      "epoch": 1.2177650429799427,
      "grad_norm": 0.4947945177555084,
      "learning_rate": 2.1341200560544973e-05,
      "loss": 0.448,
      "mean_token_accuracy": 0.8514135599136352,
      "num_tokens": 683451093.0,
      "step": 20400
    },
    {
      "epoch": 1.2180635148042025,
      "grad_norm": 0.5381024479866028,
      "learning_rate": 2.133052105891515e-05,
      "loss": 0.4819,
      "mean_token_accuracy": 0.8396755218505859,
      "num_tokens": 683618773.0,
      "step": 20405
    },
    {
      "epoch": 1.2183619866284623,
      "grad_norm": 0.5483804941177368,
      "learning_rate": 2.131984305997579e-05,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8510020256042481,
      "num_tokens": 683786453.0,
      "step": 20410
    },
    {
      "epoch": 1.218660458452722,
      "grad_norm": 0.506986677646637,
      "learning_rate": 2.1309166566327726e-05,
      "loss": 0.4516,
      "mean_token_accuracy": 0.8493916153907776,
      "num_tokens": 683954133.0,
      "step": 20415
    },
    {
      "epoch": 1.2189589302769819,
      "grad_norm": 0.45554739236831665,
      "learning_rate": 2.1298491580571423e-05,
      "loss": 0.379,
      "mean_token_accuracy": 0.8714839458465576,
      "num_tokens": 684121813.0,
      "step": 20420
    },
    {
      "epoch": 1.2192574021012417,
      "grad_norm": 0.5481733083724976,
      "learning_rate": 2.128781810530695e-05,
      "loss": 0.4582,
      "mean_token_accuracy": 0.8475664973258972,
      "num_tokens": 684289493.0,
      "step": 20425
    },
    {
      "epoch": 1.2195558739255015,
      "grad_norm": 0.49845367670059204,
      "learning_rate": 2.127714614313406e-05,
      "loss": 0.4688,
      "mean_token_accuracy": 0.8436497330665589,
      "num_tokens": 684456105.0,
      "step": 20430
    },
    {
      "epoch": 1.2198543457497613,
      "grad_norm": 0.5329987406730652,
      "learning_rate": 2.1266475696652065e-05,
      "loss": 0.4691,
      "mean_token_accuracy": 0.8459978461265564,
      "num_tokens": 684623785.0,
      "step": 20435
    },
    {
      "epoch": 1.220152817574021,
      "grad_norm": 0.514126718044281,
      "learning_rate": 2.125580676845998e-05,
      "loss": 0.4594,
      "mean_token_accuracy": 0.8477872014045715,
      "num_tokens": 684791465.0,
      "step": 20440
    },
    {
      "epoch": 1.2204512893982808,
      "grad_norm": 0.6443539261817932,
      "learning_rate": 2.1245139361156418e-05,
      "loss": 0.458,
      "mean_token_accuracy": 0.8491709470748902,
      "num_tokens": 684959145.0,
      "step": 20445
    },
    {
      "epoch": 1.2207497612225406,
      "grad_norm": 0.48586156964302063,
      "learning_rate": 2.123447347733961e-05,
      "loss": 0.4365,
      "mean_token_accuracy": 0.8554720878601074,
      "num_tokens": 685119867.0,
      "step": 20450
    },
    {
      "epoch": 1.2210482330468004,
      "grad_norm": 0.4845660924911499,
      "learning_rate": 2.1223809119607442e-05,
      "loss": 0.4949,
      "mean_token_accuracy": 0.8369378447532654,
      "num_tokens": 685287547.0,
      "step": 20455
    },
    {
      "epoch": 1.2213467048710602,
      "grad_norm": 0.49853649735450745,
      "learning_rate": 2.12131462905574e-05,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8490635752677917,
      "num_tokens": 685455227.0,
      "step": 20460
    },
    {
      "epoch": 1.22164517669532,
      "grad_norm": 0.5116775631904602,
      "learning_rate": 2.1202484992786625e-05,
      "loss": 0.4622,
      "mean_token_accuracy": 0.8474949359893799,
      "num_tokens": 685622907.0,
      "step": 20465
    },
    {
      "epoch": 1.2219436485195798,
      "grad_norm": 0.5472502112388611,
      "learning_rate": 2.119182522889187e-05,
      "loss": 0.4004,
      "mean_token_accuracy": 0.8642333388328552,
      "num_tokens": 685776657.0,
      "step": 20470
    },
    {
      "epoch": 1.2222421203438396,
      "grad_norm": 0.4789251983165741,
      "learning_rate": 2.118116700146951e-05,
      "loss": 0.4308,
      "mean_token_accuracy": 0.8540617942810058,
      "num_tokens": 685944337.0,
      "step": 20475
    },
    {
      "epoch": 1.2225405921680994,
      "grad_norm": 10.693766593933105,
      "learning_rate": 2.1170510313115565e-05,
      "loss": 0.4689,
      "mean_token_accuracy": 0.8528689026832581,
      "num_tokens": 686112017.0,
      "step": 20480
    },
    {
      "epoch": 1.2228390639923592,
      "grad_norm": 0.5329935550689697,
      "learning_rate": 2.1159855166425664e-05,
      "loss": 0.4382,
      "mean_token_accuracy": 0.8534355282783508,
      "num_tokens": 686279697.0,
      "step": 20485
    },
    {
      "epoch": 1.223137535816619,
      "grad_norm": 0.49704545736312866,
      "learning_rate": 2.114920156399506e-05,
      "loss": 0.4327,
      "mean_token_accuracy": 0.8557318449020386,
      "num_tokens": 686447377.0,
      "step": 20490
    },
    {
      "epoch": 1.2234360076408788,
      "grad_norm": 0.5286566615104675,
      "learning_rate": 2.113854950841863e-05,
      "loss": 0.438,
      "mean_token_accuracy": 0.8551473259925843,
      "num_tokens": 686615057.0,
      "step": 20495
    },
    {
      "epoch": 1.2237344794651386,
      "grad_norm": 0.4883231222629547,
      "learning_rate": 2.112789900229089e-05,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8521352767944336,
      "num_tokens": 686782737.0,
      "step": 20500
    },
    {
      "epoch": 1.2240329512893984,
      "grad_norm": 0.48276063799858093,
      "learning_rate": 2.111725004820597e-05,
      "loss": 0.4198,
      "mean_token_accuracy": 0.859071922302246,
      "num_tokens": 686950417.0,
      "step": 20505
    },
    {
      "epoch": 1.2243314231136582,
      "grad_norm": 0.5543134808540344,
      "learning_rate": 2.1106602648757606e-05,
      "loss": 0.4647,
      "mean_token_accuracy": 0.8451687932014466,
      "num_tokens": 687118097.0,
      "step": 20510
    },
    {
      "epoch": 1.2246298949379177,
      "grad_norm": 0.461599200963974,
      "learning_rate": 2.1095956806539184e-05,
      "loss": 0.3925,
      "mean_token_accuracy": 0.8689788818359375,
      "num_tokens": 687285777.0,
      "step": 20515
    },
    {
      "epoch": 1.2249283667621778,
      "grad_norm": 0.5091221332550049,
      "learning_rate": 2.1085312524143675e-05,
      "loss": 0.4722,
      "mean_token_accuracy": 0.8436001420021058,
      "num_tokens": 687453457.0,
      "step": 20520
    },
    {
      "epoch": 1.2252268385864373,
      "grad_norm": 0.5655896067619324,
      "learning_rate": 2.107466980416371e-05,
      "loss": 0.4435,
      "mean_token_accuracy": 0.852606475353241,
      "num_tokens": 687621137.0,
      "step": 20525
    },
    {
      "epoch": 1.2255253104106973,
      "grad_norm": 0.5249313116073608,
      "learning_rate": 2.106402864919151e-05,
      "loss": 0.4217,
      "mean_token_accuracy": 0.8591136813163758,
      "num_tokens": 687788817.0,
      "step": 20530
    },
    {
      "epoch": 1.225823782234957,
      "grad_norm": 0.5126509666442871,
      "learning_rate": 2.1053389061818928e-05,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8487355351448059,
      "num_tokens": 687956497.0,
      "step": 20535
    },
    {
      "epoch": 1.226122254059217,
      "grad_norm": 0.4747845232486725,
      "learning_rate": 2.104275104463743e-05,
      "loss": 0.4363,
      "mean_token_accuracy": 0.853984260559082,
      "num_tokens": 688124177.0,
      "step": 20540
    },
    {
      "epoch": 1.2264207258834765,
      "grad_norm": 0.5217232704162598,
      "learning_rate": 2.1032114600238107e-05,
      "loss": 0.4543,
      "mean_token_accuracy": 0.8476738572120667,
      "num_tokens": 688291857.0,
      "step": 20545
    },
    {
      "epoch": 1.2267191977077363,
      "grad_norm": 0.4815486967563629,
      "learning_rate": 2.1021479731211642e-05,
      "loss": 0.4398,
      "mean_token_accuracy": 0.8530478358268738,
      "num_tokens": 688459537.0,
      "step": 20550
    },
    {
      "epoch": 1.227017669531996,
      "grad_norm": 0.5050053596496582,
      "learning_rate": 2.1010846440148384e-05,
      "loss": 0.4315,
      "mean_token_accuracy": 0.8557318568229675,
      "num_tokens": 688627217.0,
      "step": 20555
    },
    {
      "epoch": 1.2273161413562559,
      "grad_norm": 0.5565704107284546,
      "learning_rate": 2.100021472963824e-05,
      "loss": 0.47,
      "mean_token_accuracy": 0.8439401149749756,
      "num_tokens": 688794897.0,
      "step": 20560
    },
    {
      "epoch": 1.2276146131805157,
      "grad_norm": 0.5366640090942383,
      "learning_rate": 2.098958460227079e-05,
      "loss": 0.4457,
      "mean_token_accuracy": 0.8527023315429687,
      "num_tokens": 688960297.0,
      "step": 20565
    },
    {
      "epoch": 1.2279130850047755,
      "grad_norm": 0.46611812710762024,
      "learning_rate": 2.097895606063516e-05,
      "loss": 0.4257,
      "mean_token_accuracy": 0.8582309484481812,
      "num_tokens": 689127977.0,
      "step": 20570
    },
    {
      "epoch": 1.2282115568290353,
      "grad_norm": 0.49351072311401367,
      "learning_rate": 2.0968329107320145e-05,
      "loss": 0.456,
      "mean_token_accuracy": 0.8475366950035095,
      "num_tokens": 689295657.0,
      "step": 20575
    },
    {
      "epoch": 1.228510028653295,
      "grad_norm": 0.5155776143074036,
      "learning_rate": 2.095770374491415e-05,
      "loss": 0.4549,
      "mean_token_accuracy": 0.8492707490921021,
      "num_tokens": 689461982.0,
      "step": 20580
    },
    {
      "epoch": 1.2288085004775549,
      "grad_norm": 0.5490033030509949,
      "learning_rate": 2.0947079976005156e-05,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8493021607398987,
      "num_tokens": 689629662.0,
      "step": 20585
    },
    {
      "epoch": 1.2291069723018146,
      "grad_norm": 0.5510979890823364,
      "learning_rate": 2.0936457803180792e-05,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8399916291236877,
      "num_tokens": 689797342.0,
      "step": 20590
    },
    {
      "epoch": 1.2294054441260744,
      "grad_norm": 0.4988440275192261,
      "learning_rate": 2.0925837229028273e-05,
      "loss": 0.4362,
      "mean_token_accuracy": 0.853835153579712,
      "num_tokens": 689965022.0,
      "step": 20595
    },
    {
      "epoch": 1.2297039159503342,
      "grad_norm": 0.6232066750526428,
      "learning_rate": 2.091521825613445e-05,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8367827653884887,
      "num_tokens": 690132702.0,
      "step": 20600
    },
    {
      "epoch": 1.230002387774594,
      "grad_norm": 0.5446926355361938,
      "learning_rate": 2.0904600887085757e-05,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8578074693679809,
      "num_tokens": 690300382.0,
      "step": 20605
    },
    {
      "epoch": 1.2303008595988538,
      "grad_norm": 0.5652893781661987,
      "learning_rate": 2.089398512446825e-05,
      "loss": 0.476,
      "mean_token_accuracy": 0.8420493721961975,
      "num_tokens": 690468062.0,
      "step": 20610
    },
    {
      "epoch": 1.2305993314231136,
      "grad_norm": 0.44775089621543884,
      "learning_rate": 2.0883370970867604e-05,
      "loss": 0.4047,
      "mean_token_accuracy": 0.8656089663505554,
      "num_tokens": 690635742.0,
      "step": 20615
    },
    {
      "epoch": 1.2308978032473734,
      "grad_norm": 0.4957120418548584,
      "learning_rate": 2.0872758428869088e-05,
      "loss": 0.4462,
      "mean_token_accuracy": 0.8523738503456115,
      "num_tokens": 690803422.0,
      "step": 20620
    },
    {
      "epoch": 1.2311962750716332,
      "grad_norm": 0.522113025188446,
      "learning_rate": 2.0862147501057583e-05,
      "loss": 0.4773,
      "mean_token_accuracy": 0.8456698060035706,
      "num_tokens": 690971102.0,
      "step": 20625
    },
    {
      "epoch": 1.231494746895893,
      "grad_norm": 0.5285800695419312,
      "learning_rate": 2.0851538190017573e-05,
      "loss": 0.4488,
      "mean_token_accuracy": 0.8500119209289551,
      "num_tokens": 691138782.0,
      "step": 20630
    },
    {
      "epoch": 1.2317932187201528,
      "grad_norm": 0.5293931365013123,
      "learning_rate": 2.084093049833315e-05,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8391029477119446,
      "num_tokens": 691306462.0,
      "step": 20635
    },
    {
      "epoch": 1.2320916905444126,
      "grad_norm": 0.4784390926361084,
      "learning_rate": 2.0830324428588028e-05,
      "loss": 0.4398,
      "mean_token_accuracy": 0.854085648059845,
      "num_tokens": 691474142.0,
      "step": 20640
    },
    {
      "epoch": 1.2323901623686724,
      "grad_norm": 0.4853532314300537,
      "learning_rate": 2.0819719983365484e-05,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8484850406646729,
      "num_tokens": 691641822.0,
      "step": 20645
    },
    {
      "epoch": 1.2326886341929322,
      "grad_norm": 0.5179458260536194,
      "learning_rate": 2.0809117165248447e-05,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8475903511047364,
      "num_tokens": 691809502.0,
      "step": 20650
    },
    {
      "epoch": 1.232987106017192,
      "grad_norm": 0.5093544125556946,
      "learning_rate": 2.0798515976819427e-05,
      "loss": 0.494,
      "mean_token_accuracy": 0.8383752822875976,
      "num_tokens": 691977182.0,
      "step": 20655
    },
    {
      "epoch": 1.2332855778414518,
      "grad_norm": 0.5244585275650024,
      "learning_rate": 2.078791642066053e-05,
      "loss": 0.4677,
      "mean_token_accuracy": 0.8452880859375,
      "num_tokens": 692144862.0,
      "step": 20660
    },
    {
      "epoch": 1.2335840496657116,
      "grad_norm": 0.47669073939323425,
      "learning_rate": 2.0777318499353486e-05,
      "loss": 0.3941,
      "mean_token_accuracy": 0.8674281358718872,
      "num_tokens": 692312542.0,
      "step": 20665
    },
    {
      "epoch": 1.2338825214899714,
      "grad_norm": 0.4995233416557312,
      "learning_rate": 2.0766722215479607e-05,
      "loss": 0.4621,
      "mean_token_accuracy": 0.8483538150787353,
      "num_tokens": 692480222.0,
      "step": 20670
    },
    {
      "epoch": 1.2341809933142311,
      "grad_norm": 0.5297977924346924,
      "learning_rate": 2.0756127571619815e-05,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8449659943580627,
      "num_tokens": 692647902.0,
      "step": 20675
    },
    {
      "epoch": 1.234479465138491,
      "grad_norm": 0.5540143251419067,
      "learning_rate": 2.074553457035463e-05,
      "loss": 0.4837,
      "mean_token_accuracy": 0.839085054397583,
      "num_tokens": 692815582.0,
      "step": 20680
    },
    {
      "epoch": 1.2347779369627507,
      "grad_norm": 0.5271779298782349,
      "learning_rate": 2.0734943214264168e-05,
      "loss": 0.4724,
      "mean_token_accuracy": 0.8444709539413452,
      "num_tokens": 692983262.0,
      "step": 20685
    },
    {
      "epoch": 1.2350764087870105,
      "grad_norm": 0.4945247173309326,
      "learning_rate": 2.0724353505928167e-05,
      "loss": 0.419,
      "mean_token_accuracy": 0.8598055601119995,
      "num_tokens": 693150942.0,
      "step": 20690
    },
    {
      "epoch": 1.2353748806112703,
      "grad_norm": 0.45129886269569397,
      "learning_rate": 2.071376544792592e-05,
      "loss": 0.3999,
      "mean_token_accuracy": 0.8669092297554016,
      "num_tokens": 693318622.0,
      "step": 20695
    },
    {
      "epoch": 1.2356733524355301,
      "grad_norm": 0.4860512912273407,
      "learning_rate": 2.0703179042836375e-05,
      "loss": 0.4071,
      "mean_token_accuracy": 0.8625849962234498,
      "num_tokens": 693486302.0,
      "step": 20700
    },
    {
      "epoch": 1.23597182425979,
      "grad_norm": 0.5046267509460449,
      "learning_rate": 2.069259429323801e-05,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8464213252067566,
      "num_tokens": 693653982.0,
      "step": 20705
    },
    {
      "epoch": 1.2362702960840497,
      "grad_norm": 0.5387168526649475,
      "learning_rate": 2.0682011201708973e-05,
      "loss": 0.445,
      "mean_token_accuracy": 0.8519145846366882,
      "num_tokens": 693821662.0,
      "step": 20710
    },
    {
      "epoch": 1.2365687679083095,
      "grad_norm": 0.45135805010795593,
      "learning_rate": 2.0671429770826937e-05,
      "loss": 0.4226,
      "mean_token_accuracy": 0.8580281496047973,
      "num_tokens": 693989342.0,
      "step": 20715
    },
    {
      "epoch": 1.2368672397325693,
      "grad_norm": 0.5135748386383057,
      "learning_rate": 2.0660850003169225e-05,
      "loss": 0.4464,
      "mean_token_accuracy": 0.8520040512084961,
      "num_tokens": 694157022.0,
      "step": 20720
    },
    {
      "epoch": 1.237165711556829,
      "grad_norm": 0.5206548571586609,
      "learning_rate": 2.065027190131274e-05,
      "loss": 0.4733,
      "mean_token_accuracy": 0.8424489974975586,
      "num_tokens": 694324702.0,
      "step": 20725
    },
    {
      "epoch": 1.2374641833810889,
      "grad_norm": 0.5115079879760742,
      "learning_rate": 2.0639695467833954e-05,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8463378310203552,
      "num_tokens": 694492382.0,
      "step": 20730
    },
    {
      "epoch": 1.2377626552053487,
      "grad_norm": 0.46827027201652527,
      "learning_rate": 2.0629120705308974e-05,
      "loss": 0.412,
      "mean_token_accuracy": 0.8624477982521057,
      "num_tokens": 694660062.0,
      "step": 20735
    },
    {
      "epoch": 1.2380611270296085,
      "grad_norm": 0.5105363130569458,
      "learning_rate": 2.061854761631346e-05,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8570320844650269,
      "num_tokens": 694827742.0,
      "step": 20740
    },
    {
      "epoch": 1.2383595988538683,
      "grad_norm": 0.6181185245513916,
      "learning_rate": 2.060797620342269e-05,
      "loss": 0.4526,
      "mean_token_accuracy": 0.848997962474823,
      "num_tokens": 694995422.0,
      "step": 20745
    },
    {
      "epoch": 1.238658070678128,
      "grad_norm": 0.535239577293396,
      "learning_rate": 2.059740646921154e-05,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8479243636131286,
      "num_tokens": 695163102.0,
      "step": 20750
    },
    {
      "epoch": 1.2389565425023878,
      "grad_norm": 0.4934963583946228,
      "learning_rate": 2.0586838416254438e-05,
      "loss": 0.4619,
      "mean_token_accuracy": 0.8474710822105408,
      "num_tokens": 695330782.0,
      "step": 20755
    },
    {
      "epoch": 1.2392550143266476,
      "grad_norm": 0.541788637638092,
      "learning_rate": 2.057627204712545e-05,
      "loss": 0.4779,
      "mean_token_accuracy": 0.842794942855835,
      "num_tokens": 695498462.0,
      "step": 20760
    },
    {
      "epoch": 1.2395534861509074,
      "grad_norm": 0.4743037819862366,
      "learning_rate": 2.05657073643982e-05,
      "loss": 0.436,
      "mean_token_accuracy": 0.8543301939964294,
      "num_tokens": 695666142.0,
      "step": 20765
    },
    {
      "epoch": 1.2398519579751672,
      "grad_norm": 0.5119059681892395,
      "learning_rate": 2.0555144370645906e-05,
      "loss": 0.4212,
      "mean_token_accuracy": 0.8579625368118287,
      "num_tokens": 695833822.0,
      "step": 20770
    },
    {
      "epoch": 1.240150429799427,
      "grad_norm": 0.5168021321296692,
      "learning_rate": 2.0544583068441397e-05,
      "loss": 0.4612,
      "mean_token_accuracy": 0.8471311092376709,
      "num_tokens": 696001502.0,
      "step": 20775
    },
    {
      "epoch": 1.2404489016236868,
      "grad_norm": 0.49903252720832825,
      "learning_rate": 2.0534023460357067e-05,
      "loss": 0.4371,
      "mean_token_accuracy": 0.8529404759407043,
      "num_tokens": 696169182.0,
      "step": 20780
    },
    {
      "epoch": 1.2407473734479466,
      "grad_norm": 0.5115178823471069,
      "learning_rate": 2.0523465548964898e-05,
      "loss": 0.4563,
      "mean_token_accuracy": 0.8501013994216919,
      "num_tokens": 696336862.0,
      "step": 20785
    },
    {
      "epoch": 1.2410458452722062,
      "grad_norm": 0.5184021592140198,
      "learning_rate": 2.051290933683646e-05,
      "loss": 0.4402,
      "mean_token_accuracy": 0.8541095018386841,
      "num_tokens": 696504542.0,
      "step": 20790
    },
    {
      "epoch": 1.2413443170964662,
      "grad_norm": 0.48175662755966187,
      "learning_rate": 2.0502354826542925e-05,
      "loss": 0.4681,
      "mean_token_accuracy": 0.84702969789505,
      "num_tokens": 696672222.0,
      "step": 20795
    },
    {
      "epoch": 1.2416427889207258,
      "grad_norm": 0.4659521281719208,
      "learning_rate": 2.049180202065504e-05,
      "loss": 0.437,
      "mean_token_accuracy": 0.8540677547454834,
      "num_tokens": 696839902.0,
      "step": 20800
    },
    {
      "epoch": 1.2419412607449858,
      "grad_norm": 0.5196808576583862,
      "learning_rate": 2.0481250921743116e-05,
      "loss": 0.4808,
      "mean_token_accuracy": 0.8409877061843872,
      "num_tokens": 697007582.0,
      "step": 20805
    },
    {
      "epoch": 1.2422397325692454,
      "grad_norm": 0.5354881882667542,
      "learning_rate": 2.047070153237709e-05,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8516044378280639,
      "num_tokens": 697175262.0,
      "step": 20810
    },
    {
      "epoch": 1.2425382043935054,
      "grad_norm": 0.5706625580787659,
      "learning_rate": 2.0460153855126444e-05,
      "loss": 0.4707,
      "mean_token_accuracy": 0.8445604085922241,
      "num_tokens": 697342942.0,
      "step": 20815
    },
    {
      "epoch": 1.242836676217765,
      "grad_norm": 0.54524827003479,
      "learning_rate": 2.044960789256028e-05,
      "loss": 0.4442,
      "mean_token_accuracy": 0.8526422500610351,
      "num_tokens": 697510622.0,
      "step": 20820
    },
    {
      "epoch": 1.2431351480420247,
      "grad_norm": 0.49509236216545105,
      "learning_rate": 2.043906364724724e-05,
      "loss": 0.4434,
      "mean_token_accuracy": 0.8518549561500549,
      "num_tokens": 697678302.0,
      "step": 20825
    },
    {
      "epoch": 1.2434336198662845,
      "grad_norm": 0.5055444836616516,
      "learning_rate": 2.0428521121755572e-05,
      "loss": 0.4636,
      "mean_token_accuracy": 0.8464571237564087,
      "num_tokens": 697845982.0,
      "step": 20830
    },
    {
      "epoch": 1.2437320916905443,
      "grad_norm": 0.504357635974884,
      "learning_rate": 2.041798031865313e-05,
      "loss": 0.4616,
      "mean_token_accuracy": 0.846540629863739,
      "num_tokens": 698013662.0,
      "step": 20835
    },
    {
      "epoch": 1.2440305635148041,
      "grad_norm": 0.4978924095630646,
      "learning_rate": 2.040744124050728e-05,
      "loss": 0.3966,
      "mean_token_accuracy": 0.8676249504089355,
      "num_tokens": 698181342.0,
      "step": 20840
    },
    {
      "epoch": 1.244329035339064,
      "grad_norm": 0.5274893045425415,
      "learning_rate": 2.0396903889885045e-05,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8450196743011474,
      "num_tokens": 698349022.0,
      "step": 20845
    },
    {
      "epoch": 1.2446275071633237,
      "grad_norm": 0.5136992931365967,
      "learning_rate": 2.038636826935296e-05,
      "loss": 0.4463,
      "mean_token_accuracy": 0.849977707862854,
      "num_tokens": 698508905.0,
      "step": 20850
    },
    {
      "epoch": 1.2449259789875835,
      "grad_norm": 0.5866198539733887,
      "learning_rate": 2.03758343814772e-05,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8517177700996399,
      "num_tokens": 698676585.0,
      "step": 20855
    },
    {
      "epoch": 1.2452244508118433,
      "grad_norm": 0.4704689383506775,
      "learning_rate": 2.036530222882347e-05,
      "loss": 0.4042,
      "mean_token_accuracy": 0.8641178607940674,
      "num_tokens": 698844265.0,
      "step": 20860
    },
    {
      "epoch": 1.245522922636103,
      "grad_norm": 0.510746955871582,
      "learning_rate": 2.0354771813957064e-05,
      "loss": 0.433,
      "mean_token_accuracy": 0.8544494867324829,
      "num_tokens": 699011945.0,
      "step": 20865
    },
    {
      "epoch": 1.2458213944603629,
      "grad_norm": 0.49832162261009216,
      "learning_rate": 2.0344243139442882e-05,
      "loss": 0.4382,
      "mean_token_accuracy": 0.8540260076522828,
      "num_tokens": 699179625.0,
      "step": 20870
    },
    {
      "epoch": 1.2461198662846227,
      "grad_norm": 0.49919718503952026,
      "learning_rate": 2.0333716207845353e-05,
      "loss": 0.4314,
      "mean_token_accuracy": 0.8555111408233642,
      "num_tokens": 699347305.0,
      "step": 20875
    },
    {
      "epoch": 1.2464183381088825,
      "grad_norm": 0.4491886496543884,
      "learning_rate": 2.032319102172852e-05,
      "loss": 0.4263,
      "mean_token_accuracy": 0.857318389415741,
      "num_tokens": 699514985.0,
      "step": 20880
    },
    {
      "epoch": 1.2467168099331423,
      "grad_norm": 0.5185744762420654,
      "learning_rate": 2.0312667583655987e-05,
      "loss": 0.3924,
      "mean_token_accuracy": 0.8676130414009094,
      "num_tokens": 699682665.0,
      "step": 20885
    },
    {
      "epoch": 1.247015281757402,
      "grad_norm": 0.5070383548736572,
      "learning_rate": 2.030214589619092e-05,
      "loss": 0.4611,
      "mean_token_accuracy": 0.8475545763969421,
      "num_tokens": 699850345.0,
      "step": 20890
    },
    {
      "epoch": 1.2473137535816619,
      "grad_norm": 0.5295190215110779,
      "learning_rate": 2.0291625961896082e-05,
      "loss": 0.4478,
      "mean_token_accuracy": 0.8517117738723755,
      "num_tokens": 700018025.0,
      "step": 20895
    },
    {
      "epoch": 1.2476122254059216,
      "grad_norm": 0.5279406309127808,
      "learning_rate": 2.0281107783333787e-05,
      "loss": 0.4519,
      "mean_token_accuracy": 0.8491053342819214,
      "num_tokens": 700185705.0,
      "step": 20900
    },
    {
      "epoch": 1.2479106972301814,
      "grad_norm": 0.5204823017120361,
      "learning_rate": 2.0270591363065945e-05,
      "loss": 0.4487,
      "mean_token_accuracy": 0.8485327363014221,
      "num_tokens": 700353385.0,
      "step": 20905
    },
    {
      "epoch": 1.2482091690544412,
      "grad_norm": 0.4830276668071747,
      "learning_rate": 2.0260076703654007e-05,
      "loss": 0.4311,
      "mean_token_accuracy": 0.8544077277183533,
      "num_tokens": 700521065.0,
      "step": 20910
    },
    {
      "epoch": 1.248507640878701,
      "grad_norm": 0.5601676106452942,
      "learning_rate": 2.0249563807659023e-05,
      "loss": 0.4761,
      "mean_token_accuracy": 0.8422819972038269,
      "num_tokens": 700688745.0,
      "step": 20915
    },
    {
      "epoch": 1.2488061127029608,
      "grad_norm": 0.4411035180091858,
      "learning_rate": 2.0239052677641607e-05,
      "loss": 0.4246,
      "mean_token_accuracy": 0.8582548141479492,
      "num_tokens": 700856425.0,
      "step": 20920
    },
    {
      "epoch": 1.2491045845272206,
      "grad_norm": 0.5106505751609802,
      "learning_rate": 2.022854331616193e-05,
      "loss": 0.4429,
      "mean_token_accuracy": 0.8539782881736755,
      "num_tokens": 701024105.0,
      "step": 20925
    },
    {
      "epoch": 1.2494030563514804,
      "grad_norm": 0.4430442750453949,
      "learning_rate": 2.0218035725779745e-05,
      "loss": 0.4174,
      "mean_token_accuracy": 0.8587796807289123,
      "num_tokens": 701191785.0,
      "step": 20930
    },
    {
      "epoch": 1.2497015281757402,
      "grad_norm": 0.5601619482040405,
      "learning_rate": 2.020752990905437e-05,
      "loss": 0.476,
      "mean_token_accuracy": 0.843546462059021,
      "num_tokens": 701359465.0,
      "step": 20935
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.46632280945777893,
      "learning_rate": 2.0197025868544682e-05,
      "loss": 0.4539,
      "mean_token_accuracy": 0.8478050827980042,
      "num_tokens": 701527145.0,
      "step": 20940
    },
    {
      "epoch": 1.2502984718242598,
      "grad_norm": 0.5367595553398132,
      "learning_rate": 2.018652360680915e-05,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8531731128692627,
      "num_tokens": 701694825.0,
      "step": 20945
    },
    {
      "epoch": 1.2505969436485196,
      "grad_norm": 0.499789834022522,
      "learning_rate": 2.017602312640578e-05,
      "loss": 0.4521,
      "mean_token_accuracy": 0.8490934014320374,
      "num_tokens": 701862505.0,
      "step": 20950
    },
    {
      "epoch": 1.2508954154727794,
      "grad_norm": 0.5380803942680359,
      "learning_rate": 2.0165524429892163e-05,
      "loss": 0.4606,
      "mean_token_accuracy": 0.8480734825134277,
      "num_tokens": 702030185.0,
      "step": 20955
    },
    {
      "epoch": 1.2511938872970392,
      "grad_norm": 0.5221821069717407,
      "learning_rate": 2.0155027519825448e-05,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8515090107917785,
      "num_tokens": 702197865.0,
      "step": 20960
    },
    {
      "epoch": 1.251492359121299,
      "grad_norm": 0.508967936038971,
      "learning_rate": 2.0144532398762357e-05,
      "loss": 0.4327,
      "mean_token_accuracy": 0.8550340056419372,
      "num_tokens": 702365545.0,
      "step": 20965
    },
    {
      "epoch": 1.2517908309455588,
      "grad_norm": 0.514724850654602,
      "learning_rate": 2.013403906925917e-05,
      "loss": 0.4381,
      "mean_token_accuracy": 0.8526661038398743,
      "num_tokens": 702533225.0,
      "step": 20970
    },
    {
      "epoch": 1.2520893027698186,
      "grad_norm": 0.5254999399185181,
      "learning_rate": 2.0123547533871717e-05,
      "loss": 0.4195,
      "mean_token_accuracy": 0.858547055721283,
      "num_tokens": 702700905.0,
      "step": 20975
    },
    {
      "epoch": 1.2523877745940784,
      "grad_norm": 0.565106987953186,
      "learning_rate": 2.0113057795155426e-05,
      "loss": 0.457,
      "mean_token_accuracy": 0.8483597755432128,
      "num_tokens": 702868585.0,
      "step": 20980
    },
    {
      "epoch": 1.2526862464183381,
      "grad_norm": 0.5399832129478455,
      "learning_rate": 2.0102569855665248e-05,
      "loss": 0.4664,
      "mean_token_accuracy": 0.8447751522064209,
      "num_tokens": 703036265.0,
      "step": 20985
    },
    {
      "epoch": 1.252984718242598,
      "grad_norm": 0.5129609704017639,
      "learning_rate": 2.0092083717955728e-05,
      "loss": 0.4352,
      "mean_token_accuracy": 0.8541810750961304,
      "num_tokens": 703203945.0,
      "step": 20990
    },
    {
      "epoch": 1.2532831900668577,
      "grad_norm": 0.5159810185432434,
      "learning_rate": 2.0081599384580952e-05,
      "loss": 0.4441,
      "mean_token_accuracy": 0.8502206802368164,
      "num_tokens": 703371625.0,
      "step": 20995
    },
    {
      "epoch": 1.2535816618911175,
      "grad_norm": 0.5023723244667053,
      "learning_rate": 2.0071116858094573e-05,
      "loss": 0.4267,
      "mean_token_accuracy": 0.85719313621521,
      "num_tokens": 703539305.0,
      "step": 21000
    },
    {
      "epoch": 1.2538801337153773,
      "grad_norm": 0.5119024515151978,
      "learning_rate": 2.006063614104981e-05,
      "loss": 0.4386,
      "mean_token_accuracy": 0.8528092622756958,
      "num_tokens": 703706985.0,
      "step": 21005
    },
    {
      "epoch": 1.2541786055396371,
      "grad_norm": 0.4861049950122833,
      "learning_rate": 2.0050157235999424e-05,
      "loss": 0.455,
      "mean_token_accuracy": 0.848872721195221,
      "num_tokens": 703874665.0,
      "step": 21010
    },
    {
      "epoch": 1.254477077363897,
      "grad_norm": 0.48757025599479675,
      "learning_rate": 2.0039680145495753e-05,
      "loss": 0.4543,
      "mean_token_accuracy": 0.849385666847229,
      "num_tokens": 704042345.0,
      "step": 21015
    },
    {
      "epoch": 1.2547755491881567,
      "grad_norm": 0.513938307762146,
      "learning_rate": 2.0029204872090685e-05,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8479482412338257,
      "num_tokens": 704210025.0,
      "step": 21020
    },
    {
      "epoch": 1.2550740210124165,
      "grad_norm": 0.5094524025917053,
      "learning_rate": 2.0018731418335658e-05,
      "loss": 0.4601,
      "mean_token_accuracy": 0.8463258981704712,
      "num_tokens": 704377705.0,
      "step": 21025
    },
    {
      "epoch": 1.2553724928366763,
      "grad_norm": 0.5137123465538025,
      "learning_rate": 2.0008259786781695e-05,
      "loss": 0.4176,
      "mean_token_accuracy": 0.8611833572387695,
      "num_tokens": 704545385.0,
      "step": 21030
    },
    {
      "epoch": 1.255670964660936,
      "grad_norm": 0.4705581068992615,
      "learning_rate": 1.9997789979979332e-05,
      "loss": 0.4448,
      "mean_token_accuracy": 0.8523977041244507,
      "num_tokens": 704713065.0,
      "step": 21035
    },
    {
      "epoch": 1.2559694364851959,
      "grad_norm": 0.5636093020439148,
      "learning_rate": 1.998732200047871e-05,
      "loss": 0.4859,
      "mean_token_accuracy": 0.839198362827301,
      "num_tokens": 704880745.0,
      "step": 21040
    },
    {
      "epoch": 1.2562679083094554,
      "grad_norm": 0.6135547161102295,
      "learning_rate": 1.9976855850829463e-05,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8450972080230713,
      "num_tokens": 705048425.0,
      "step": 21045
    },
    {
      "epoch": 1.2565663801337155,
      "grad_norm": 0.491524875164032,
      "learning_rate": 1.996639153358084e-05,
      "loss": 0.4376,
      "mean_token_accuracy": 0.8533401012420654,
      "num_tokens": 705216105.0,
      "step": 21050
    },
    {
      "epoch": 1.256864851957975,
      "grad_norm": 0.5055975317955017,
      "learning_rate": 1.995592905128162e-05,
      "loss": 0.4174,
      "mean_token_accuracy": 0.859686267375946,
      "num_tokens": 705383785.0,
      "step": 21055
    },
    {
      "epoch": 1.257163323782235,
      "grad_norm": 0.47449231147766113,
      "learning_rate": 1.9945468406480127e-05,
      "loss": 0.4309,
      "mean_token_accuracy": 0.8571374177932739,
      "num_tokens": 705548788.0,
      "step": 21060
    },
    {
      "epoch": 1.2574617956064946,
      "grad_norm": 0.5250605940818787,
      "learning_rate": 1.9935009601724243e-05,
      "loss": 0.4512,
      "mean_token_accuracy": 0.848520827293396,
      "num_tokens": 705716468.0,
      "step": 21065
    },
    {
      "epoch": 1.2577602674307546,
      "grad_norm": 0.5314686298370361,
      "learning_rate": 1.99245526395614e-05,
      "loss": 0.4338,
      "mean_token_accuracy": 0.85520099401474,
      "num_tokens": 705884148.0,
      "step": 21070
    },
    {
      "epoch": 1.2580587392550142,
      "grad_norm": 0.5164315104484558,
      "learning_rate": 1.9914097522538587e-05,
      "loss": 0.4721,
      "mean_token_accuracy": 0.8424549698829651,
      "num_tokens": 706051828.0,
      "step": 21075
    },
    {
      "epoch": 1.2583572110792742,
      "grad_norm": 0.5837658047676086,
      "learning_rate": 1.9903644253202353e-05,
      "loss": 0.4772,
      "mean_token_accuracy": 0.8415066242218018,
      "num_tokens": 706219508.0,
      "step": 21080
    },
    {
      "epoch": 1.2586556829035338,
      "grad_norm": 0.483070969581604,
      "learning_rate": 1.9893192834098767e-05,
      "loss": 0.4164,
      "mean_token_accuracy": 0.8601574659347534,
      "num_tokens": 706387188.0,
      "step": 21085
    },
    {
      "epoch": 1.2589541547277938,
      "grad_norm": 0.5082573890686035,
      "learning_rate": 1.9882743267773474e-05,
      "loss": 0.4692,
      "mean_token_accuracy": 0.8421984910964966,
      "num_tokens": 706554868.0,
      "step": 21090
    },
    {
      "epoch": 1.2592526265520534,
      "grad_norm": 0.5246371030807495,
      "learning_rate": 1.9872295556771652e-05,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8465238451957703,
      "num_tokens": 706722474.0,
      "step": 21095
    },
    {
      "epoch": 1.2595510983763134,
      "grad_norm": 0.48530304431915283,
      "learning_rate": 1.986184970363803e-05,
      "loss": 0.432,
      "mean_token_accuracy": 0.854861032962799,
      "num_tokens": 706890154.0,
      "step": 21100
    },
    {
      "epoch": 1.259849570200573,
      "grad_norm": 0.45325836539268494,
      "learning_rate": 1.985140571091692e-05,
      "loss": 0.4395,
      "mean_token_accuracy": 0.8546761393547058,
      "num_tokens": 707057834.0,
      "step": 21105
    },
    {
      "epoch": 1.260148042024833,
      "grad_norm": 0.4651221036911011,
      "learning_rate": 1.984096358115211e-05,
      "loss": 0.4472,
      "mean_token_accuracy": 0.8494453191757202,
      "num_tokens": 707225514.0,
      "step": 21110
    },
    {
      "epoch": 1.2604465138490926,
      "grad_norm": 0.48300737142562866,
      "learning_rate": 1.9830523316886995e-05,
      "loss": 0.4439,
      "mean_token_accuracy": 0.8510388016700745,
      "num_tokens": 707386234.0,
      "step": 21115
    },
    {
      "epoch": 1.2607449856733524,
      "grad_norm": 0.4908396303653717,
      "learning_rate": 1.9820084920664484e-05,
      "loss": 0.4403,
      "mean_token_accuracy": 0.8526840090751648,
      "num_tokens": 707553914.0,
      "step": 21120
    },
    {
      "epoch": 1.2610434574976122,
      "grad_norm": 0.47433868050575256,
      "learning_rate": 1.980964839502705e-05,
      "loss": 0.4126,
      "mean_token_accuracy": 0.8613742113113403,
      "num_tokens": 707721594.0,
      "step": 21125
    },
    {
      "epoch": 1.261341929321872,
      "grad_norm": 0.5637901425361633,
      "learning_rate": 1.9799213742516692e-05,
      "loss": 0.4685,
      "mean_token_accuracy": 0.845753300189972,
      "num_tokens": 707889274.0,
      "step": 21130
    },
    {
      "epoch": 1.2616404011461317,
      "grad_norm": 0.48012587428092957,
      "learning_rate": 1.9788780965674973e-05,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8419658780097962,
      "num_tokens": 708056954.0,
      "step": 21135
    },
    {
      "epoch": 1.2619388729703915,
      "grad_norm": 0.47778651118278503,
      "learning_rate": 1.9778350067042983e-05,
      "loss": 0.4067,
      "mean_token_accuracy": 0.8626267433166503,
      "num_tokens": 708224634.0,
      "step": 21140
    },
    {
      "epoch": 1.2622373447946513,
      "grad_norm": 0.5273913741111755,
      "learning_rate": 1.9767921049161354e-05,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8585768699645996,
      "num_tokens": 708392314.0,
      "step": 21145
    },
    {
      "epoch": 1.2625358166189111,
      "grad_norm": 0.47344717383384705,
      "learning_rate": 1.9757493914570275e-05,
      "loss": 0.4309,
      "mean_token_accuracy": 0.8556125640869141,
      "num_tokens": 708559994.0,
      "step": 21150
    },
    {
      "epoch": 1.262834288443171,
      "grad_norm": 0.44999417662620544,
      "learning_rate": 1.9747068665809464e-05,
      "loss": 0.4184,
      "mean_token_accuracy": 0.860151493549347,
      "num_tokens": 708727674.0,
      "step": 21155
    },
    {
      "epoch": 1.2631327602674307,
      "grad_norm": 0.5429452657699585,
      "learning_rate": 1.9736645305418173e-05,
      "loss": 0.4269,
      "mean_token_accuracy": 0.8565700888633728,
      "num_tokens": 708890569.0,
      "step": 21160
    },
    {
      "epoch": 1.2634312320916905,
      "grad_norm": 0.4813825786113739,
      "learning_rate": 1.972622383593523e-05,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8496003985404968,
      "num_tokens": 709058249.0,
      "step": 21165
    },
    {
      "epoch": 1.2637297039159503,
      "grad_norm": 0.5351575613021851,
      "learning_rate": 1.971580425989894e-05,
      "loss": 0.4544,
      "mean_token_accuracy": 0.8501670122146606,
      "num_tokens": 709225929.0,
      "step": 21170
    },
    {
      "epoch": 1.26402817574021,
      "grad_norm": 0.49393555521965027,
      "learning_rate": 1.9705386579847217e-05,
      "loss": 0.4573,
      "mean_token_accuracy": 0.8498052358627319,
      "num_tokens": 709388339.0,
      "step": 21175
    },
    {
      "epoch": 1.2643266475644699,
      "grad_norm": 0.5100563764572144,
      "learning_rate": 1.9694970798317458e-05,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8513360261917114,
      "num_tokens": 709556019.0,
      "step": 21180
    },
    {
      "epoch": 1.2646251193887297,
      "grad_norm": 0.48210608959198,
      "learning_rate": 1.9684556917846624e-05,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8570201635360718,
      "num_tokens": 709723699.0,
      "step": 21185
    },
    {
      "epoch": 1.2649235912129895,
      "grad_norm": 0.4726911187171936,
      "learning_rate": 1.9674144940971217e-05,
      "loss": 0.439,
      "mean_token_accuracy": 0.8528653502464294,
      "num_tokens": 709885029.0,
      "step": 21190
    },
    {
      "epoch": 1.2652220630372493,
      "grad_norm": 0.4879405200481415,
      "learning_rate": 1.9663734870227253e-05,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8567935228347778,
      "num_tokens": 710052709.0,
      "step": 21195
    },
    {
      "epoch": 1.265520534861509,
      "grad_norm": 0.49307912588119507,
      "learning_rate": 1.96533267081503e-05,
      "loss": 0.4238,
      "mean_token_accuracy": 0.8584993600845336,
      "num_tokens": 710220389.0,
      "step": 21200
    },
    {
      "epoch": 1.2658190066857689,
      "grad_norm": 0.46078819036483765,
      "learning_rate": 1.9642920457275466e-05,
      "loss": 0.382,
      "mean_token_accuracy": 0.8713944911956787,
      "num_tokens": 710388069.0,
      "step": 21205
    },
    {
      "epoch": 1.2661174785100286,
      "grad_norm": 0.5175058841705322,
      "learning_rate": 1.9632516120137372e-05,
      "loss": 0.4623,
      "mean_token_accuracy": 0.8475844025611877,
      "num_tokens": 710555749.0,
      "step": 21210
    },
    {
      "epoch": 1.2664159503342884,
      "grad_norm": 0.5143455266952515,
      "learning_rate": 1.96221136992702e-05,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8443755149841309,
      "num_tokens": 710723429.0,
      "step": 21215
    },
    {
      "epoch": 1.2667144221585482,
      "grad_norm": 0.5266312956809998,
      "learning_rate": 1.9611713197207643e-05,
      "loss": 0.4307,
      "mean_token_accuracy": 0.8573720693588257,
      "num_tokens": 710891109.0,
      "step": 21220
    },
    {
      "epoch": 1.267012893982808,
      "grad_norm": 0.4937094748020172,
      "learning_rate": 1.9601314616482942e-05,
      "loss": 0.4584,
      "mean_token_accuracy": 0.8467195510864258,
      "num_tokens": 711058789.0,
      "step": 21225
    },
    {
      "epoch": 1.2673113658070678,
      "grad_norm": 0.562179446220398,
      "learning_rate": 1.9590917959628856e-05,
      "loss": 0.4469,
      "mean_token_accuracy": 0.8507336378097534,
      "num_tokens": 711226469.0,
      "step": 21230
    },
    {
      "epoch": 1.2676098376313276,
      "grad_norm": 0.5001862049102783,
      "learning_rate": 1.958052322917769e-05,
      "loss": 0.4076,
      "mean_token_accuracy": 0.8633365273475647,
      "num_tokens": 711394149.0,
      "step": 21235
    },
    {
      "epoch": 1.2679083094555874,
      "grad_norm": 0.5662214756011963,
      "learning_rate": 1.957013042766126e-05,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8486818432807922,
      "num_tokens": 711561829.0,
      "step": 21240
    },
    {
      "epoch": 1.2682067812798472,
      "grad_norm": 0.4831666648387909,
      "learning_rate": 1.9559739557610936e-05,
      "loss": 0.4079,
      "mean_token_accuracy": 0.863873302936554,
      "num_tokens": 711729509.0,
      "step": 21245
    },
    {
      "epoch": 1.268505253104107,
      "grad_norm": 0.5003403425216675,
      "learning_rate": 1.9549350621557604e-05,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8470654964447022,
      "num_tokens": 711897189.0,
      "step": 21250
    },
    {
      "epoch": 1.2688037249283668,
      "grad_norm": 0.5404338836669922,
      "learning_rate": 1.953896362203168e-05,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8459620714187622,
      "num_tokens": 712064869.0,
      "step": 21255
    },
    {
      "epoch": 1.2691021967526266,
      "grad_norm": 0.4927692711353302,
      "learning_rate": 1.9528578561563116e-05,
      "loss": 0.4459,
      "mean_token_accuracy": 0.8504831314086914,
      "num_tokens": 712232549.0,
      "step": 21260
    },
    {
      "epoch": 1.2694006685768864,
      "grad_norm": 0.538818359375,
      "learning_rate": 1.951819544268137e-05,
      "loss": 0.4919,
      "mean_token_accuracy": 0.837242043018341,
      "num_tokens": 712400229.0,
      "step": 21265
    },
    {
      "epoch": 1.2696991404011462,
      "grad_norm": 0.47741541266441345,
      "learning_rate": 1.9507814267915452e-05,
      "loss": 0.4078,
      "mean_token_accuracy": 0.8616187453269959,
      "num_tokens": 712567909.0,
      "step": 21270
    },
    {
      "epoch": 1.269997612225406,
      "grad_norm": 0.5630694627761841,
      "learning_rate": 1.9497435039793892e-05,
      "loss": 0.4403,
      "mean_token_accuracy": 0.854849112033844,
      "num_tokens": 712735589.0,
      "step": 21275
    },
    {
      "epoch": 1.2702960840496658,
      "grad_norm": 0.4834173619747162,
      "learning_rate": 1.948705776084474e-05,
      "loss": 0.4271,
      "mean_token_accuracy": 0.8579267501831055,
      "num_tokens": 712903269.0,
      "step": 21280
    },
    {
      "epoch": 1.2705945558739256,
      "grad_norm": 0.7313836216926575,
      "learning_rate": 1.9476682433595573e-05,
      "loss": 0.4571,
      "mean_token_accuracy": 0.8482643604278565,
      "num_tokens": 713070949.0,
      "step": 21285
    },
    {
      "epoch": 1.2708930276981854,
      "grad_norm": 0.5565345883369446,
      "learning_rate": 1.9466309060573486e-05,
      "loss": 0.4797,
      "mean_token_accuracy": 0.8417273163795471,
      "num_tokens": 713238629.0,
      "step": 21290
    },
    {
      "epoch": 1.2711914995224451,
      "grad_norm": 0.5181086659431458,
      "learning_rate": 1.945593764430511e-05,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8472981095314026,
      "num_tokens": 713406309.0,
      "step": 21295
    },
    {
      "epoch": 1.271489971346705,
      "grad_norm": 0.5557349324226379,
      "learning_rate": 1.9445568187316608e-05,
      "loss": 0.4643,
      "mean_token_accuracy": 0.8465346574783326,
      "num_tokens": 713573989.0,
      "step": 21300
    },
    {
      "epoch": 1.2717884431709647,
      "grad_norm": 0.5677387714385986,
      "learning_rate": 1.9435200692133633e-05,
      "loss": 0.4624,
      "mean_token_accuracy": 0.8455564856529236,
      "num_tokens": 713741669.0,
      "step": 21305
    },
    {
      "epoch": 1.2720869149952245,
      "grad_norm": 0.5147316455841064,
      "learning_rate": 1.9424835161281395e-05,
      "loss": 0.4841,
      "mean_token_accuracy": 0.8406000256538391,
      "num_tokens": 713909349.0,
      "step": 21310
    },
    {
      "epoch": 1.2723853868194843,
      "grad_norm": 0.5038878917694092,
      "learning_rate": 1.94144715972846e-05,
      "loss": 0.4411,
      "mean_token_accuracy": 0.85284503698349,
      "num_tokens": 714077029.0,
      "step": 21315
    },
    {
      "epoch": 1.272683858643744,
      "grad_norm": 0.5570834875106812,
      "learning_rate": 1.9404110002667486e-05,
      "loss": 0.4488,
      "mean_token_accuracy": 0.8506739735603333,
      "num_tokens": 714244709.0,
      "step": 21320
    },
    {
      "epoch": 1.272982330468004,
      "grad_norm": 0.4947250485420227,
      "learning_rate": 1.9393750379953824e-05,
      "loss": 0.452,
      "mean_token_accuracy": 0.8522903442382812,
      "num_tokens": 714412389.0,
      "step": 21325
    },
    {
      "epoch": 1.2732808022922635,
      "grad_norm": 0.5436123013496399,
      "learning_rate": 1.9383392731666873e-05,
      "loss": 0.482,
      "mean_token_accuracy": 0.8396278262138367,
      "num_tokens": 714580069.0,
      "step": 21330
    },
    {
      "epoch": 1.2735792741165235,
      "grad_norm": 0.49210286140441895,
      "learning_rate": 1.937303706032944e-05,
      "loss": 0.4496,
      "mean_token_accuracy": 0.8511392116546631,
      "num_tokens": 714747749.0,
      "step": 21335
    },
    {
      "epoch": 1.273877745940783,
      "grad_norm": 0.4916885197162628,
      "learning_rate": 1.9362683368463835e-05,
      "loss": 0.4069,
      "mean_token_accuracy": 0.8642789006233216,
      "num_tokens": 714915429.0,
      "step": 21340
    },
    {
      "epoch": 1.274176217765043,
      "grad_norm": 0.48606356978416443,
      "learning_rate": 1.93523316585919e-05,
      "loss": 0.3987,
      "mean_token_accuracy": 0.865638792514801,
      "num_tokens": 715083109.0,
      "step": 21345
    },
    {
      "epoch": 1.2744746895893027,
      "grad_norm": 0.4798738360404968,
      "learning_rate": 1.9341981933234975e-05,
      "loss": 0.4908,
      "mean_token_accuracy": 0.8389955759048462,
      "num_tokens": 715250789.0,
      "step": 21350
    },
    {
      "epoch": 1.2747731614135627,
      "grad_norm": 0.47659364342689514,
      "learning_rate": 1.933163419491393e-05,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8537814617156982,
      "num_tokens": 715418469.0,
      "step": 21355
    },
    {
      "epoch": 1.2750716332378222,
      "grad_norm": 0.530511200428009,
      "learning_rate": 1.9321288446149145e-05,
      "loss": 0.4304,
      "mean_token_accuracy": 0.8576046705245972,
      "num_tokens": 715586149.0,
      "step": 21360
    },
    {
      "epoch": 1.2753701050620823,
      "grad_norm": 0.5185045599937439,
      "learning_rate": 1.9310944689460515e-05,
      "loss": 0.4503,
      "mean_token_accuracy": 0.8511928796768189,
      "num_tokens": 715753829.0,
      "step": 21365
    },
    {
      "epoch": 1.2756685768863418,
      "grad_norm": 0.4745495021343231,
      "learning_rate": 1.930060292736747e-05,
      "loss": 0.4214,
      "mean_token_accuracy": 0.8583382964134216,
      "num_tokens": 715921509.0,
      "step": 21370
    },
    {
      "epoch": 1.2759670487106018,
      "grad_norm": 0.5337730646133423,
      "learning_rate": 1.9290263162388917e-05,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8472265362739563,
      "num_tokens": 716089189.0,
      "step": 21375
    },
    {
      "epoch": 1.2762655205348614,
      "grad_norm": 0.5391858220100403,
      "learning_rate": 1.927992539704331e-05,
      "loss": 0.4677,
      "mean_token_accuracy": 0.8453477263450623,
      "num_tokens": 716256869.0,
      "step": 21380
    },
    {
      "epoch": 1.2765639923591214,
      "grad_norm": 0.5120289325714111,
      "learning_rate": 1.92695896338486e-05,
      "loss": 0.4166,
      "mean_token_accuracy": 0.8610879063606263,
      "num_tokens": 716424549.0,
      "step": 21385
    },
    {
      "epoch": 1.276862464183381,
      "grad_norm": 0.5688652992248535,
      "learning_rate": 1.9259255875322246e-05,
      "loss": 0.4832,
      "mean_token_accuracy": 0.8419386863708496,
      "num_tokens": 716589499.0,
      "step": 21390
    },
    {
      "epoch": 1.2771609360076408,
      "grad_norm": 0.5396602153778076,
      "learning_rate": 1.9248924123981233e-05,
      "loss": 0.4598,
      "mean_token_accuracy": 0.8455326437950135,
      "num_tokens": 716757179.0,
      "step": 21395
    },
    {
      "epoch": 1.2774594078319006,
      "grad_norm": 0.5797765851020813,
      "learning_rate": 1.9238594382342046e-05,
      "loss": 0.4571,
      "mean_token_accuracy": 0.8484671473503113,
      "num_tokens": 716924859.0,
      "step": 21400
    },
    {
      "epoch": 1.2777578796561604,
      "grad_norm": 0.5455407500267029,
      "learning_rate": 1.9228266652920684e-05,
      "loss": 0.449,
      "mean_token_accuracy": 0.8498537302017212,
      "num_tokens": 717087609.0,
      "step": 21405
    },
    {
      "epoch": 1.2780563514804202,
      "grad_norm": 0.522078812122345,
      "learning_rate": 1.921794093823266e-05,
      "loss": 0.4482,
      "mean_token_accuracy": 0.8496779203414917,
      "num_tokens": 717255289.0,
      "step": 21410
    },
    {
      "epoch": 1.27835482330468,
      "grad_norm": 0.4904731214046478,
      "learning_rate": 1.9207617240792987e-05,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8468030571937561,
      "num_tokens": 717422969.0,
      "step": 21415
    },
    {
      "epoch": 1.2786532951289398,
      "grad_norm": 0.5107303857803345,
      "learning_rate": 1.9197295563116204e-05,
      "loss": 0.4187,
      "mean_token_accuracy": 0.8584695219993591,
      "num_tokens": 717590649.0,
      "step": 21420
    },
    {
      "epoch": 1.2789517669531996,
      "grad_norm": 0.5600069761276245,
      "learning_rate": 1.9186975907716325e-05,
      "loss": 0.4638,
      "mean_token_accuracy": 0.8463199257850647,
      "num_tokens": 717758329.0,
      "step": 21425
    },
    {
      "epoch": 1.2792502387774594,
      "grad_norm": 0.5263044834136963,
      "learning_rate": 1.91766582771069e-05,
      "loss": 0.4256,
      "mean_token_accuracy": 0.8587987780570984,
      "num_tokens": 717923781.0,
      "step": 21430
    },
    {
      "epoch": 1.2795487106017192,
      "grad_norm": 0.5339877605438232,
      "learning_rate": 1.9166342673801003e-05,
      "loss": 0.4151,
      "mean_token_accuracy": 0.859775722026825,
      "num_tokens": 718091461.0,
      "step": 21435
    },
    {
      "epoch": 1.279847182425979,
      "grad_norm": 0.48792290687561035,
      "learning_rate": 1.9156029100311144e-05,
      "loss": 0.3939,
      "mean_token_accuracy": 0.8680245757102967,
      "num_tokens": 718259141.0,
      "step": 21440
    },
    {
      "epoch": 1.2801456542502387,
      "grad_norm": 0.5102375149726868,
      "learning_rate": 1.9145717559149435e-05,
      "loss": 0.5101,
      "mean_token_accuracy": 0.8314744234085083,
      "num_tokens": 718426821.0,
      "step": 21445
    },
    {
      "epoch": 1.2804441260744985,
      "grad_norm": 0.5072603225708008,
      "learning_rate": 1.9135408052827397e-05,
      "loss": 0.4219,
      "mean_token_accuracy": 0.8588930010795593,
      "num_tokens": 718594501.0,
      "step": 21450
    },
    {
      "epoch": 1.2807425978987583,
      "grad_norm": 0.4996205270290375,
      "learning_rate": 1.9125100583856135e-05,
      "loss": 0.4544,
      "mean_token_accuracy": 0.8499343991279602,
      "num_tokens": 718762181.0,
      "step": 21455
    },
    {
      "epoch": 1.2810410697230181,
      "grad_norm": 0.473214328289032,
      "learning_rate": 1.9114795154746202e-05,
      "loss": 0.4312,
      "mean_token_accuracy": 0.855075740814209,
      "num_tokens": 718929861.0,
      "step": 21460
    },
    {
      "epoch": 1.281339541547278,
      "grad_norm": 0.505867600440979,
      "learning_rate": 1.9104491768007682e-05,
      "loss": 0.449,
      "mean_token_accuracy": 0.8504175186157227,
      "num_tokens": 719097541.0,
      "step": 21465
    },
    {
      "epoch": 1.2816380133715377,
      "grad_norm": 0.4916333854198456,
      "learning_rate": 1.9094190426150153e-05,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8483060956001282,
      "num_tokens": 719265221.0,
      "step": 21470
    },
    {
      "epoch": 1.2819364851957975,
      "grad_norm": 0.5311673879623413,
      "learning_rate": 1.9083891131682707e-05,
      "loss": 0.4435,
      "mean_token_accuracy": 0.853071677684784,
      "num_tokens": 719432901.0,
      "step": 21475
    },
    {
      "epoch": 1.2822349570200573,
      "grad_norm": 0.5119551420211792,
      "learning_rate": 1.9073593887113917e-05,
      "loss": 0.4425,
      "mean_token_accuracy": 0.852493143081665,
      "num_tokens": 719600581.0,
      "step": 21480
    },
    {
      "epoch": 1.282533428844317,
      "grad_norm": 0.5507540702819824,
      "learning_rate": 1.9063298694951863e-05,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8495824813842774,
      "num_tokens": 719768261.0,
      "step": 21485
    },
    {
      "epoch": 1.2828319006685769,
      "grad_norm": 0.47302794456481934,
      "learning_rate": 1.9053005557704133e-05,
      "loss": 0.4455,
      "mean_token_accuracy": 0.8505964517593384,
      "num_tokens": 719935941.0,
      "step": 21490
    },
    {
      "epoch": 1.2831303724928367,
      "grad_norm": 0.5300539135932922,
      "learning_rate": 1.9042714477877827e-05,
      "loss": 0.4364,
      "mean_token_accuracy": 0.8546403408050537,
      "num_tokens": 720103621.0,
      "step": 21495
    },
    {
      "epoch": 1.2834288443170965,
      "grad_norm": 0.47543445229530334,
      "learning_rate": 1.9032425457979503e-05,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8566324710845947,
      "num_tokens": 720271301.0,
      "step": 21500
    },
    {
      "epoch": 1.2837273161413563,
      "grad_norm": 0.5631353855133057,
      "learning_rate": 1.9022138500515264e-05,
      "loss": 0.4496,
      "mean_token_accuracy": 0.8516819715499878,
      "num_tokens": 720438981.0,
      "step": 21505
    },
    {
      "epoch": 1.284025787965616,
      "grad_norm": 0.46531927585601807,
      "learning_rate": 1.901185360799066e-05,
      "loss": 0.4203,
      "mean_token_accuracy": 0.859060001373291,
      "num_tokens": 720606661.0,
      "step": 21510
    },
    {
      "epoch": 1.2843242597898759,
      "grad_norm": 0.4696912467479706,
      "learning_rate": 1.9001570782910787e-05,
      "loss": 0.4612,
      "mean_token_accuracy": 0.8484373211860656,
      "num_tokens": 720774341.0,
      "step": 21515
    },
    {
      "epoch": 1.2846227316141356,
      "grad_norm": 0.4956842362880707,
      "learning_rate": 1.8991290027780222e-05,
      "loss": 0.457,
      "mean_token_accuracy": 0.8481569766998291,
      "num_tokens": 720942021.0,
      "step": 21520
    },
    {
      "epoch": 1.2849212034383954,
      "grad_norm": 0.4649125933647156,
      "learning_rate": 1.8981011345103015e-05,
      "loss": 0.4805,
      "mean_token_accuracy": 0.841727864742279,
      "num_tokens": 721102300.0,
      "step": 21525
    },
    {
      "epoch": 1.2852196752626552,
      "grad_norm": 0.47681552171707153,
      "learning_rate": 1.8970734737382744e-05,
      "loss": 0.4417,
      "mean_token_accuracy": 0.8522963285446167,
      "num_tokens": 721269980.0,
      "step": 21530
    },
    {
      "epoch": 1.285518147086915,
      "grad_norm": 0.4758831858634949,
      "learning_rate": 1.896046020712246e-05,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8505964398384094,
      "num_tokens": 721437660.0,
      "step": 21535
    },
    {
      "epoch": 1.2858166189111748,
      "grad_norm": 0.595343828201294,
      "learning_rate": 1.8950187756824705e-05,
      "loss": 0.5001,
      "mean_token_accuracy": 0.8331444501876831,
      "num_tokens": 721605340.0,
      "step": 21540
    },
    {
      "epoch": 1.2861150907354346,
      "grad_norm": 0.6689709424972534,
      "learning_rate": 1.8939917388991547e-05,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8417451977729797,
      "num_tokens": 721773020.0,
      "step": 21545
    },
    {
      "epoch": 1.2864135625596944,
      "grad_norm": 0.5414208173751831,
      "learning_rate": 1.8929649106124497e-05,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8577716827392579,
      "num_tokens": 721940700.0,
      "step": 21550
    },
    {
      "epoch": 1.2867120343839542,
      "grad_norm": 0.48877057433128357,
      "learning_rate": 1.891938291072461e-05,
      "loss": 0.4317,
      "mean_token_accuracy": 0.8560479402542114,
      "num_tokens": 722108380.0,
      "step": 21555
    },
    {
      "epoch": 1.287010506208214,
      "grad_norm": 0.49305394291877747,
      "learning_rate": 1.8909118805292388e-05,
      "loss": 0.4808,
      "mean_token_accuracy": 0.8400811076164245,
      "num_tokens": 722276060.0,
      "step": 21560
    },
    {
      "epoch": 1.2873089780324738,
      "grad_norm": 0.482348769903183,
      "learning_rate": 1.8898856792327852e-05,
      "loss": 0.4667,
      "mean_token_accuracy": 0.8455624461174012,
      "num_tokens": 722443740.0,
      "step": 21565
    },
    {
      "epoch": 1.2876074498567336,
      "grad_norm": 0.5003317594528198,
      "learning_rate": 1.88885968743305e-05,
      "loss": 0.3979,
      "mean_token_accuracy": 0.8663604855537415,
      "num_tokens": 722611420.0,
      "step": 21570
    },
    {
      "epoch": 1.2879059216809934,
      "grad_norm": 0.5651480555534363,
      "learning_rate": 1.887833905379932e-05,
      "loss": 0.477,
      "mean_token_accuracy": 0.8425503849983216,
      "num_tokens": 722779100.0,
      "step": 21575
    },
    {
      "epoch": 1.2882043935052532,
      "grad_norm": 0.4775930345058441,
      "learning_rate": 1.8868083333232817e-05,
      "loss": 0.4061,
      "mean_token_accuracy": 0.8624179840087891,
      "num_tokens": 722946780.0,
      "step": 21580
    },
    {
      "epoch": 1.288502865329513,
      "grad_norm": 0.5129034519195557,
      "learning_rate": 1.885782971512894e-05,
      "loss": 0.4584,
      "mean_token_accuracy": 0.8483120560646057,
      "num_tokens": 723114460.0,
      "step": 21585
    },
    {
      "epoch": 1.2888013371537728,
      "grad_norm": 0.5143440961837769,
      "learning_rate": 1.884757820198515e-05,
      "loss": 0.4386,
      "mean_token_accuracy": 0.8527138233184814,
      "num_tokens": 723282140.0,
      "step": 21590
    },
    {
      "epoch": 1.2890998089780323,
      "grad_norm": 0.4904709756374359,
      "learning_rate": 1.883732879629839e-05,
      "loss": 0.4381,
      "mean_token_accuracy": 0.8531909823417664,
      "num_tokens": 723449820.0,
      "step": 21595
    },
    {
      "epoch": 1.2893982808022924,
      "grad_norm": 0.5331194996833801,
      "learning_rate": 1.8827081500565104e-05,
      "loss": 0.4399,
      "mean_token_accuracy": 0.8531492352485657,
      "num_tokens": 723617500.0,
      "step": 21600
    },
    {
      "epoch": 1.289696752626552,
      "grad_norm": 0.5755446553230286,
      "learning_rate": 1.8816836317281206e-05,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8476082563400269,
      "num_tokens": 723785180.0,
      "step": 21605
    },
    {
      "epoch": 1.289995224450812,
      "grad_norm": 0.5003317594528198,
      "learning_rate": 1.8806593248942083e-05,
      "loss": 0.417,
      "mean_token_accuracy": 0.8610342383384705,
      "num_tokens": 723952860.0,
      "step": 21610
    },
    {
      "epoch": 1.2902936962750715,
      "grad_norm": 0.5023718476295471,
      "learning_rate": 1.8796352298042648e-05,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8536502480506897,
      "num_tokens": 724120540.0,
      "step": 21615
    },
    {
      "epoch": 1.2905921680993315,
      "grad_norm": 0.5326288938522339,
      "learning_rate": 1.8786113467077256e-05,
      "loss": 0.4512,
      "mean_token_accuracy": 0.8493319749832153,
      "num_tokens": 724288220.0,
      "step": 21620
    },
    {
      "epoch": 1.290890639923591,
      "grad_norm": 0.7071923017501831,
      "learning_rate": 1.877587675853976e-05,
      "loss": 0.4514,
      "mean_token_accuracy": 0.8503578543663025,
      "num_tokens": 724455900.0,
      "step": 21625
    },
    {
      "epoch": 1.2911891117478511,
      "grad_norm": 0.7134477496147156,
      "learning_rate": 1.8765642174923525e-05,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8470975160598755,
      "num_tokens": 724619592.0,
      "step": 21630
    },
    {
      "epoch": 1.2914875835721107,
      "grad_norm": 0.4384186267852783,
      "learning_rate": 1.8755409718721338e-05,
      "loss": 0.4575,
      "mean_token_accuracy": 0.8496779203414917,
      "num_tokens": 724787272.0,
      "step": 21635
    },
    {
      "epoch": 1.2917860553963707,
      "grad_norm": 0.44683679938316345,
      "learning_rate": 1.8745179392425544e-05,
      "loss": 0.4262,
      "mean_token_accuracy": 0.8579386830329895,
      "num_tokens": 724954952.0,
      "step": 21640
    },
    {
      "epoch": 1.2920845272206303,
      "grad_norm": 0.5542892217636108,
      "learning_rate": 1.8734951198527882e-05,
      "loss": 0.4468,
      "mean_token_accuracy": 0.8526363015174866,
      "num_tokens": 725122632.0,
      "step": 21645
    },
    {
      "epoch": 1.2923829990448903,
      "grad_norm": 0.5385172367095947,
      "learning_rate": 1.872472513951965e-05,
      "loss": 0.464,
      "mean_token_accuracy": 0.846528697013855,
      "num_tokens": 725290312.0,
      "step": 21650
    },
    {
      "epoch": 1.2926814708691499,
      "grad_norm": 0.5000436902046204,
      "learning_rate": 1.871450121789159e-05,
      "loss": 0.445,
      "mean_token_accuracy": 0.8502326250076294,
      "num_tokens": 725457992.0,
      "step": 21655
    },
    {
      "epoch": 1.2929799426934099,
      "grad_norm": 0.4647788405418396,
      "learning_rate": 1.870427943613391e-05,
      "loss": 0.3889,
      "mean_token_accuracy": 0.8675832033157349,
      "num_tokens": 725625672.0,
      "step": 21660
    },
    {
      "epoch": 1.2932784145176695,
      "grad_norm": 0.5435347557067871,
      "learning_rate": 1.869405979673633e-05,
      "loss": 0.4466,
      "mean_token_accuracy": 0.8517237305641174,
      "num_tokens": 725793352.0,
      "step": 21665
    },
    {
      "epoch": 1.2935768863419292,
      "grad_norm": 0.5936532020568848,
      "learning_rate": 1.8683842302188013e-05,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8409936785697937,
      "num_tokens": 725961032.0,
      "step": 21670
    },
    {
      "epoch": 1.293875358166189,
      "grad_norm": 0.4887048006057739,
      "learning_rate": 1.8673626954977645e-05,
      "loss": 0.4423,
      "mean_token_accuracy": 0.8525050640106201,
      "num_tokens": 726128712.0,
      "step": 21675
    },
    {
      "epoch": 1.2941738299904488,
      "grad_norm": 0.5274898409843445,
      "learning_rate": 1.8663413757593346e-05,
      "loss": 0.4511,
      "mean_token_accuracy": 0.8489383339881897,
      "num_tokens": 726296392.0,
      "step": 21680
    },
    {
      "epoch": 1.2944723018147086,
      "grad_norm": 0.4889707863330841,
      "learning_rate": 1.8653202712522728e-05,
      "loss": 0.4133,
      "mean_token_accuracy": 0.8612608909606934,
      "num_tokens": 726464072.0,
      "step": 21685
    },
    {
      "epoch": 1.2947707736389684,
      "grad_norm": 0.48489663004875183,
      "learning_rate": 1.864299382225288e-05,
      "loss": 0.4464,
      "mean_token_accuracy": 0.852982223033905,
      "num_tokens": 726631752.0,
      "step": 21690
    },
    {
      "epoch": 1.2950692454632282,
      "grad_norm": 0.5408115386962891,
      "learning_rate": 1.8632787089270374e-05,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8479840159416199,
      "num_tokens": 726799432.0,
      "step": 21695
    },
    {
      "epoch": 1.295367717287488,
      "grad_norm": 0.48788145184516907,
      "learning_rate": 1.8622582516061245e-05,
      "loss": 0.4116,
      "mean_token_accuracy": 0.8613921046257019,
      "num_tokens": 726967112.0,
      "step": 21700
    },
    {
      "epoch": 1.2956661891117478,
      "grad_norm": 0.5170052647590637,
      "learning_rate": 1.861238010511099e-05,
      "loss": 0.4405,
      "mean_token_accuracy": 0.8541631817817688,
      "num_tokens": 727134792.0,
      "step": 21705
    },
    {
      "epoch": 1.2959646609360076,
      "grad_norm": 0.5319960117340088,
      "learning_rate": 1.860217985890461e-05,
      "loss": 0.4656,
      "mean_token_accuracy": 0.8464809775352478,
      "num_tokens": 727302472.0,
      "step": 21710
    },
    {
      "epoch": 1.2962631327602674,
      "grad_norm": 0.6124706268310547,
      "learning_rate": 1.8591981779926564e-05,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8469462156295776,
      "num_tokens": 727470152.0,
      "step": 21715
    },
    {
      "epoch": 1.2965616045845272,
      "grad_norm": 0.5003710985183716,
      "learning_rate": 1.8581785870660783e-05,
      "loss": 0.4088,
      "mean_token_accuracy": 0.863145649433136,
      "num_tokens": 727637832.0,
      "step": 21720
    },
    {
      "epoch": 1.296860076408787,
      "grad_norm": 0.5208696126937866,
      "learning_rate": 1.857159213359066e-05,
      "loss": 0.4482,
      "mean_token_accuracy": 0.849761426448822,
      "num_tokens": 727805512.0,
      "step": 21725
    },
    {
      "epoch": 1.2971585482330468,
      "grad_norm": 0.5592601299285889,
      "learning_rate": 1.8561400571199072e-05,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8479124426841735,
      "num_tokens": 727973192.0,
      "step": 21730
    },
    {
      "epoch": 1.2974570200573066,
      "grad_norm": 0.4732402265071869,
      "learning_rate": 1.8551211185968363e-05,
      "loss": 0.4341,
      "mean_token_accuracy": 0.8565788030624389,
      "num_tokens": 728140872.0,
      "step": 21735
    },
    {
      "epoch": 1.2977554918815664,
      "grad_norm": 0.5079814195632935,
      "learning_rate": 1.8541023980380352e-05,
      "loss": 0.4462,
      "mean_token_accuracy": 0.8484015226364136,
      "num_tokens": 728308552.0,
      "step": 21740
    },
    {
      "epoch": 1.2980539637058262,
      "grad_norm": 0.47837844491004944,
      "learning_rate": 1.85308389569163e-05,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8500894784927369,
      "num_tokens": 728476232.0,
      "step": 21745
    },
    {
      "epoch": 1.298352435530086,
      "grad_norm": 0.512665331363678,
      "learning_rate": 1.852065611805699e-05,
      "loss": 0.4669,
      "mean_token_accuracy": 0.843710470199585,
      "num_tokens": 728642001.0,
      "step": 21750
    },
    {
      "epoch": 1.2986509073543457,
      "grad_norm": 0.4837343096733093,
      "learning_rate": 1.851047546628261e-05,
      "loss": 0.4301,
      "mean_token_accuracy": 0.8560479521751404,
      "num_tokens": 728809681.0,
      "step": 21755
    },
    {
      "epoch": 1.2989493791786055,
      "grad_norm": 0.478515088558197,
      "learning_rate": 1.8500297004072854e-05,
      "loss": 0.4292,
      "mean_token_accuracy": 0.8570082187652588,
      "num_tokens": 728977361.0,
      "step": 21760
    },
    {
      "epoch": 1.2992478510028653,
      "grad_norm": 0.5688692927360535,
      "learning_rate": 1.8490120733906892e-05,
      "loss": 0.449,
      "mean_token_accuracy": 0.8524573683738709,
      "num_tokens": 729145041.0,
      "step": 21765
    },
    {
      "epoch": 1.2995463228271251,
      "grad_norm": 0.5579441785812378,
      "learning_rate": 1.8479946658263315e-05,
      "loss": 0.4667,
      "mean_token_accuracy": 0.8456101655960083,
      "num_tokens": 729312721.0,
      "step": 21770
    },
    {
      "epoch": 1.299844794651385,
      "grad_norm": 0.44270285964012146,
      "learning_rate": 1.846977477962023e-05,
      "loss": 0.4389,
      "mean_token_accuracy": 0.8549266338348389,
      "num_tokens": 729480401.0,
      "step": 21775
    },
    {
      "epoch": 1.3001432664756447,
      "grad_norm": 0.5454878211021423,
      "learning_rate": 1.845960510045517e-05,
      "loss": 0.4245,
      "mean_token_accuracy": 0.8591613888740539,
      "num_tokens": 729648081.0,
      "step": 21780
    },
    {
      "epoch": 1.3004417382999045,
      "grad_norm": 0.46221116185188293,
      "learning_rate": 1.8449437623245163e-05,
      "loss": 0.4448,
      "mean_token_accuracy": 0.852117383480072,
      "num_tokens": 729815761.0,
      "step": 21785
    },
    {
      "epoch": 1.3007402101241643,
      "grad_norm": 0.47009992599487305,
      "learning_rate": 1.8439272350466663e-05,
      "loss": 0.4326,
      "mean_token_accuracy": 0.8562328577041626,
      "num_tokens": 729983441.0,
      "step": 21790
    },
    {
      "epoch": 1.301038681948424,
      "grad_norm": 0.6485724449157715,
      "learning_rate": 1.842910928459563e-05,
      "loss": 0.5059,
      "mean_token_accuracy": 0.8340272068977356,
      "num_tokens": 730151121.0,
      "step": 21795
    },
    {
      "epoch": 1.3013371537726839,
      "grad_norm": 0.508536159992218,
      "learning_rate": 1.841894842810747e-05,
      "loss": 0.4614,
      "mean_token_accuracy": 0.847005832195282,
      "num_tokens": 730318801.0,
      "step": 21800
    },
    {
      "epoch": 1.3016356255969437,
      "grad_norm": 0.5798758268356323,
      "learning_rate": 1.840878978347702e-05,
      "loss": 0.4581,
      "mean_token_accuracy": 0.847757351398468,
      "num_tokens": 730486481.0,
      "step": 21805
    },
    {
      "epoch": 1.3019340974212035,
      "grad_norm": 0.48340994119644165,
      "learning_rate": 1.839863335317864e-05,
      "loss": 0.403,
      "mean_token_accuracy": 0.8640403032302857,
      "num_tokens": 730654161.0,
      "step": 21810
    },
    {
      "epoch": 1.3022325692454633,
      "grad_norm": 0.5189979672431946,
      "learning_rate": 1.8388479139686088e-05,
      "loss": 0.4575,
      "mean_token_accuracy": 0.8482285499572754,
      "num_tokens": 730821841.0,
      "step": 21815
    },
    {
      "epoch": 1.302531041069723,
      "grad_norm": 0.49118292331695557,
      "learning_rate": 1.8378327145472624e-05,
      "loss": 0.4336,
      "mean_token_accuracy": 0.8549564599990844,
      "num_tokens": 730989521.0,
      "step": 21820
    },
    {
      "epoch": 1.3028295128939829,
      "grad_norm": 0.5009886026382446,
      "learning_rate": 1.8368177373010954e-05,
      "loss": 0.438,
      "mean_token_accuracy": 0.8536144495010376,
      "num_tokens": 731157201.0,
      "step": 21825
    },
    {
      "epoch": 1.3031279847182426,
      "grad_norm": 0.5465167760848999,
      "learning_rate": 1.8358029824773225e-05,
      "loss": 0.4667,
      "mean_token_accuracy": 0.8434390902519227,
      "num_tokens": 731324881.0,
      "step": 21830
    },
    {
      "epoch": 1.3034264565425024,
      "grad_norm": 0.5582911372184753,
      "learning_rate": 1.8347884503231094e-05,
      "loss": 0.4551,
      "mean_token_accuracy": 0.848270308971405,
      "num_tokens": 731492561.0,
      "step": 21835
    },
    {
      "epoch": 1.3037249283667622,
      "grad_norm": 0.4791560173034668,
      "learning_rate": 1.8337741410855602e-05,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8467970967292786,
      "num_tokens": 731660241.0,
      "step": 21840
    },
    {
      "epoch": 1.304023400191022,
      "grad_norm": 0.4999723434448242,
      "learning_rate": 1.832760055011731e-05,
      "loss": 0.4595,
      "mean_token_accuracy": 0.8481748700141907,
      "num_tokens": 731827921.0,
      "step": 21845
    },
    {
      "epoch": 1.3043218720152818,
      "grad_norm": 0.5018539428710938,
      "learning_rate": 1.8317461923486218e-05,
      "loss": 0.4263,
      "mean_token_accuracy": 0.8575450301170349,
      "num_tokens": 731995601.0,
      "step": 21850
    },
    {
      "epoch": 1.3046203438395416,
      "grad_norm": 0.4244856834411621,
      "learning_rate": 1.8307325533431757e-05,
      "loss": 0.4258,
      "mean_token_accuracy": 0.8578730702400208,
      "num_tokens": 732163281.0,
      "step": 21855
    },
    {
      "epoch": 1.3049188156638014,
      "grad_norm": 0.5131832361221313,
      "learning_rate": 1.8297191382422847e-05,
      "loss": 0.4405,
      "mean_token_accuracy": 0.8525169968605042,
      "num_tokens": 732330961.0,
      "step": 21860
    },
    {
      "epoch": 1.3052172874880612,
      "grad_norm": 0.4526806175708771,
      "learning_rate": 1.828705947292783e-05,
      "loss": 0.4192,
      "mean_token_accuracy": 0.8599367737770081,
      "num_tokens": 732498641.0,
      "step": 21865
    },
    {
      "epoch": 1.3055157593123208,
      "grad_norm": 0.4389190375804901,
      "learning_rate": 1.827692980741454e-05,
      "loss": 0.4107,
      "mean_token_accuracy": 0.8621078372001648,
      "num_tokens": 732666321.0,
      "step": 21870
    },
    {
      "epoch": 1.3058142311365808,
      "grad_norm": 0.5271827578544617,
      "learning_rate": 1.8266802388350225e-05,
      "loss": 0.4548,
      "mean_token_accuracy": 0.8494453072547913,
      "num_tokens": 732834001.0,
      "step": 21875
    },
    {
      "epoch": 1.3061127029608404,
      "grad_norm": 0.49042731523513794,
      "learning_rate": 1.825667721820161e-05,
      "loss": 0.4332,
      "mean_token_accuracy": 0.8544375658035278,
      "num_tokens": 733001681.0,
      "step": 21880
    },
    {
      "epoch": 1.3064111747851004,
      "grad_norm": 0.5099753141403198,
      "learning_rate": 1.8246554299434888e-05,
      "loss": 0.4562,
      "mean_token_accuracy": 0.8488906025886536,
      "num_tokens": 733169361.0,
      "step": 21885
    },
    {
      "epoch": 1.30670964660936,
      "grad_norm": 0.5492095351219177,
      "learning_rate": 1.8236433634515648e-05,
      "loss": 0.4438,
      "mean_token_accuracy": 0.8514851570129395,
      "num_tokens": 733337041.0,
      "step": 21890
    },
    {
      "epoch": 1.30700811843362,
      "grad_norm": 0.4917539358139038,
      "learning_rate": 1.8226315225908993e-05,
      "loss": 0.4481,
      "mean_token_accuracy": 0.8518966913223267,
      "num_tokens": 733504721.0,
      "step": 21895
    },
    {
      "epoch": 1.3073065902578795,
      "grad_norm": 0.6124330759048462,
      "learning_rate": 1.8216199076079426e-05,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8449481129646301,
      "num_tokens": 733672401.0,
      "step": 21900
    },
    {
      "epoch": 1.3076050620821396,
      "grad_norm": 0.496549129486084,
      "learning_rate": 1.820608518749094e-05,
      "loss": 0.4233,
      "mean_token_accuracy": 0.8578551769256592,
      "num_tokens": 733840081.0,
      "step": 21905
    },
    {
      "epoch": 1.3079035339063991,
      "grad_norm": 0.5200099945068359,
      "learning_rate": 1.819597356260696e-05,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8500178933143616,
      "num_tokens": 734007761.0,
      "step": 21910
    },
    {
      "epoch": 1.3082020057306591,
      "grad_norm": 0.48620837926864624,
      "learning_rate": 1.8185864203890334e-05,
      "loss": 0.4314,
      "mean_token_accuracy": 0.8548789262771607,
      "num_tokens": 734175441.0,
      "step": 21915
    },
    {
      "epoch": 1.3085004775549187,
      "grad_norm": 0.4823870360851288,
      "learning_rate": 1.8175757113803415e-05,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8423535823822021,
      "num_tokens": 734343121.0,
      "step": 21920
    },
    {
      "epoch": 1.3087989493791787,
      "grad_norm": 0.4812258780002594,
      "learning_rate": 1.816565229480795e-05,
      "loss": 0.4381,
      "mean_token_accuracy": 0.8525587439537048,
      "num_tokens": 734510801.0,
      "step": 21925
    },
    {
      "epoch": 1.3090974212034383,
      "grad_norm": 0.5481657385826111,
      "learning_rate": 1.8155549749365163e-05,
      "loss": 0.4539,
      "mean_token_accuracy": 0.8486699342727662,
      "num_tokens": 734678481.0,
      "step": 21930
    },
    {
      "epoch": 1.3093958930276983,
      "grad_norm": 0.4960908889770508,
      "learning_rate": 1.8145449479935717e-05,
      "loss": 0.4391,
      "mean_token_accuracy": 0.8538112878799439,
      "num_tokens": 734846161.0,
      "step": 21935
    },
    {
      "epoch": 1.309694364851958,
      "grad_norm": 0.5120190382003784,
      "learning_rate": 1.8135351488979708e-05,
      "loss": 0.4611,
      "mean_token_accuracy": 0.8461946845054626,
      "num_tokens": 735013841.0,
      "step": 21940
    },
    {
      "epoch": 1.3099928366762177,
      "grad_norm": 0.5206858515739441,
      "learning_rate": 1.8125255778956714e-05,
      "loss": 0.4319,
      "mean_token_accuracy": 0.8559882998466491,
      "num_tokens": 735181521.0,
      "step": 21945
    },
    {
      "epoch": 1.3102913085004775,
      "grad_norm": 0.5164458155632019,
      "learning_rate": 1.811516235232568e-05,
      "loss": 0.4381,
      "mean_token_accuracy": 0.8545511484146118,
      "num_tokens": 735344176.0,
      "step": 21950
    },
    {
      "epoch": 1.3105897803247373,
      "grad_norm": 0.48718783259391785,
      "learning_rate": 1.81050712115451e-05,
      "loss": 0.4282,
      "mean_token_accuracy": 0.8561851382255554,
      "num_tokens": 735511856.0,
      "step": 21955
    },
    {
      "epoch": 1.310888252148997,
      "grad_norm": 0.5037433505058289,
      "learning_rate": 1.8094982359072838e-05,
      "loss": 0.4649,
      "mean_token_accuracy": 0.8462185382843017,
      "num_tokens": 735679536.0,
      "step": 21960
    },
    {
      "epoch": 1.3111867239732569,
      "grad_norm": 0.5409345030784607,
      "learning_rate": 1.8084895797366215e-05,
      "loss": 0.4516,
      "mean_token_accuracy": 0.8509543061256408,
      "num_tokens": 735847216.0,
      "step": 21965
    },
    {
      "epoch": 1.3114851957975167,
      "grad_norm": 0.4389113485813141,
      "learning_rate": 1.8074811528882007e-05,
      "loss": 0.4206,
      "mean_token_accuracy": 0.8593583345413208,
      "num_tokens": 736010296.0,
      "step": 21970
    },
    {
      "epoch": 1.3117836676217765,
      "grad_norm": 0.48540887236595154,
      "learning_rate": 1.8064729556076414e-05,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8471907615661621,
      "num_tokens": 736177976.0,
      "step": 21975
    },
    {
      "epoch": 1.3120821394460362,
      "grad_norm": 0.5117364525794983,
      "learning_rate": 1.80546498814051e-05,
      "loss": 0.458,
      "mean_token_accuracy": 0.8476500034332275,
      "num_tokens": 736345656.0,
      "step": 21980
    },
    {
      "epoch": 1.312380611270296,
      "grad_norm": 0.5079487562179565,
      "learning_rate": 1.804457250732315e-05,
      "loss": 0.4714,
      "mean_token_accuracy": 0.8450912475585938,
      "num_tokens": 736513336.0,
      "step": 21985
    },
    {
      "epoch": 1.3126790830945558,
      "grad_norm": 0.47358182072639465,
      "learning_rate": 1.8034497436285082e-05,
      "loss": 0.4245,
      "mean_token_accuracy": 0.8578492164611816,
      "num_tokens": 736681016.0,
      "step": 21990
    },
    {
      "epoch": 1.3129775549188156,
      "grad_norm": 0.4516245126724243,
      "learning_rate": 1.802442467074488e-05,
      "loss": 0.4244,
      "mean_token_accuracy": 0.8607658386230469,
      "num_tokens": 736848696.0,
      "step": 21995
    },
    {
      "epoch": 1.3132760267430754,
      "grad_norm": 0.4859462082386017,
      "learning_rate": 1.801435421315595e-05,
      "loss": 0.4249,
      "mean_token_accuracy": 0.8593283891677856,
      "num_tokens": 737016376.0,
      "step": 22000
    },
    {
      "epoch": 1.3135744985673352,
      "grad_norm": 0.5580401420593262,
      "learning_rate": 1.8004286065971142e-05,
      "loss": 0.4215,
      "mean_token_accuracy": 0.858069896697998,
      "num_tokens": 737184056.0,
      "step": 22005
    },
    {
      "epoch": 1.313872970391595,
      "grad_norm": 0.5447243452072144,
      "learning_rate": 1.799422023164272e-05,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8559942722320557,
      "num_tokens": 737351736.0,
      "step": 22010
    },
    {
      "epoch": 1.3141714422158548,
      "grad_norm": 0.5117738246917725,
      "learning_rate": 1.7984156712622425e-05,
      "loss": 0.4404,
      "mean_token_accuracy": 0.8534892082214356,
      "num_tokens": 737519416.0,
      "step": 22015
    },
    {
      "epoch": 1.3144699140401146,
      "grad_norm": 0.5081394910812378,
      "learning_rate": 1.7974095511361418e-05,
      "loss": 0.4452,
      "mean_token_accuracy": 0.8518430233001709,
      "num_tokens": 737687096.0,
      "step": 22020
    },
    {
      "epoch": 1.3147683858643744,
      "grad_norm": 0.537953794002533,
      "learning_rate": 1.7964036630310262e-05,
      "loss": 0.4611,
      "mean_token_accuracy": 0.8462543249130249,
      "num_tokens": 737854776.0,
      "step": 22025
    },
    {
      "epoch": 1.3150668576886342,
      "grad_norm": 0.4828946590423584,
      "learning_rate": 1.7953980071918996e-05,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8448586463928223,
      "num_tokens": 738022456.0,
      "step": 22030
    },
    {
      "epoch": 1.315365329512894,
      "grad_norm": 0.465561181306839,
      "learning_rate": 1.7943925838637087e-05,
      "loss": 0.4255,
      "mean_token_accuracy": 0.8582488536834717,
      "num_tokens": 738190136.0,
      "step": 22035
    },
    {
      "epoch": 1.3156638013371538,
      "grad_norm": 0.5207686424255371,
      "learning_rate": 1.7933873932913425e-05,
      "loss": 0.4227,
      "mean_token_accuracy": 0.8588572025299073,
      "num_tokens": 738357816.0,
      "step": 22040
    },
    {
      "epoch": 1.3159622731614136,
      "grad_norm": 0.512942910194397,
      "learning_rate": 1.7923824357196343e-05,
      "loss": 0.4669,
      "mean_token_accuracy": 0.8468805909156799,
      "num_tokens": 738525496.0,
      "step": 22045
    },
    {
      "epoch": 1.3162607449856734,
      "grad_norm": 0.5877902507781982,
      "learning_rate": 1.7913777113933594e-05,
      "loss": 0.4335,
      "mean_token_accuracy": 0.8556602597236633,
      "num_tokens": 738693176.0,
      "step": 22050
    },
    {
      "epoch": 1.3165592168099332,
      "grad_norm": 0.5309934616088867,
      "learning_rate": 1.7903732205572366e-05,
      "loss": 0.4474,
      "mean_token_accuracy": 0.8513300776481628,
      "num_tokens": 738860856.0,
      "step": 22055
    },
    {
      "epoch": 1.316857688634193,
      "grad_norm": 0.5761898159980774,
      "learning_rate": 1.7893689634559284e-05,
      "loss": 0.4621,
      "mean_token_accuracy": 0.8478170156478881,
      "num_tokens": 739028536.0,
      "step": 22060
    },
    {
      "epoch": 1.3171561604584527,
      "grad_norm": 0.4726199805736542,
      "learning_rate": 1.7883649403340403e-05,
      "loss": 0.4165,
      "mean_token_accuracy": 0.8606584787368774,
      "num_tokens": 739196216.0,
      "step": 22065
    },
    {
      "epoch": 1.3174546322827125,
      "grad_norm": 0.5314990282058716,
      "learning_rate": 1.7873611514361216e-05,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8551055788993835,
      "num_tokens": 739363896.0,
      "step": 22070
    },
    {
      "epoch": 1.3177531041069723,
      "grad_norm": 0.5096940398216248,
      "learning_rate": 1.7863575970066615e-05,
      "loss": 0.4407,
      "mean_token_accuracy": 0.854097580909729,
      "num_tokens": 739531576.0,
      "step": 22075
    },
    {
      "epoch": 1.3180515759312321,
      "grad_norm": 0.5359295606613159,
      "learning_rate": 1.7853542772900965e-05,
      "loss": 0.4504,
      "mean_token_accuracy": 0.8497316002845764,
      "num_tokens": 739699256.0,
      "step": 22080
    },
    {
      "epoch": 1.318350047755492,
      "grad_norm": 0.49495822191238403,
      "learning_rate": 1.784351192530802e-05,
      "loss": 0.449,
      "mean_token_accuracy": 0.8510258793830872,
      "num_tokens": 739866936.0,
      "step": 22085
    },
    {
      "epoch": 1.3186485195797517,
      "grad_norm": 0.472333163022995,
      "learning_rate": 1.7833483429730986e-05,
      "loss": 0.4204,
      "mean_token_accuracy": 0.8596803069114685,
      "num_tokens": 740034616.0,
      "step": 22090
    },
    {
      "epoch": 1.3189469914040115,
      "grad_norm": 0.4684799015522003,
      "learning_rate": 1.782345728861249e-05,
      "loss": 0.4251,
      "mean_token_accuracy": 0.8584158420562744,
      "num_tokens": 740202296.0,
      "step": 22095
    },
    {
      "epoch": 1.3192454632282713,
      "grad_norm": 0.4827628433704376,
      "learning_rate": 1.7813433504394572e-05,
      "loss": 0.4281,
      "mean_token_accuracy": 0.8565489649772644,
      "num_tokens": 740369976.0,
      "step": 22100
    },
    {
      "epoch": 1.319543935052531,
      "grad_norm": 0.5313222408294678,
      "learning_rate": 1.7803412079518727e-05,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8430335283279419,
      "num_tokens": 740537656.0,
      "step": 22105
    },
    {
      "epoch": 1.3198424068767909,
      "grad_norm": 0.5034235715866089,
      "learning_rate": 1.779339301642584e-05,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8488369345664978,
      "num_tokens": 740705336.0,
      "step": 22110
    },
    {
      "epoch": 1.3201408787010507,
      "grad_norm": 0.5767083168029785,
      "learning_rate": 1.778337631755625e-05,
      "loss": 0.4408,
      "mean_token_accuracy": 0.8530180096626282,
      "num_tokens": 740873016.0,
      "step": 22115
    },
    {
      "epoch": 1.3204393505253105,
      "grad_norm": 0.5320366621017456,
      "learning_rate": 1.7773361985349706e-05,
      "loss": 0.4366,
      "mean_token_accuracy": 0.8537098884582519,
      "num_tokens": 741040696.0,
      "step": 22120
    },
    {
      "epoch": 1.3207378223495703,
      "grad_norm": 0.5269381403923035,
      "learning_rate": 1.7763350022245386e-05,
      "loss": 0.4165,
      "mean_token_accuracy": 0.8591792941093445,
      "num_tokens": 741208376.0,
      "step": 22125
    },
    {
      "epoch": 1.32103629417383,
      "grad_norm": 0.54883873462677,
      "learning_rate": 1.7753340430681892e-05,
      "loss": 0.4405,
      "mean_token_accuracy": 0.8532983422279358,
      "num_tokens": 741376056.0,
      "step": 22130
    },
    {
      "epoch": 1.3213347659980899,
      "grad_norm": 0.4980525076389313,
      "learning_rate": 1.774333321309722e-05,
      "loss": 0.4314,
      "mean_token_accuracy": 0.8546761274337769,
      "num_tokens": 741543736.0,
      "step": 22135
    },
    {
      "epoch": 1.3216332378223496,
      "grad_norm": 0.5273322463035583,
      "learning_rate": 1.773332837192885e-05,
      "loss": 0.4467,
      "mean_token_accuracy": 0.8511451840400696,
      "num_tokens": 741711416.0,
      "step": 22140
    },
    {
      "epoch": 1.3219317096466094,
      "grad_norm": 0.4954814612865448,
      "learning_rate": 1.7723325909613614e-05,
      "loss": 0.4465,
      "mean_token_accuracy": 0.8516461968421936,
      "num_tokens": 741879096.0,
      "step": 22145
    },
    {
      "epoch": 1.3222301814708692,
      "grad_norm": 0.4901604652404785,
      "learning_rate": 1.771332582858782e-05,
      "loss": 0.4548,
      "mean_token_accuracy": 0.8492007613182068,
      "num_tokens": 742046776.0,
      "step": 22150
    },
    {
      "epoch": 1.3225286532951288,
      "grad_norm": 0.5273316502571106,
      "learning_rate": 1.770332813128715e-05,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8497792959213257,
      "num_tokens": 742214456.0,
      "step": 22155
    },
    {
      "epoch": 1.3228271251193888,
      "grad_norm": 0.5013324618339539,
      "learning_rate": 1.7693332820146746e-05,
      "loss": 0.4285,
      "mean_token_accuracy": 0.857055950164795,
      "num_tokens": 742382136.0,
      "step": 22160
    },
    {
      "epoch": 1.3231255969436484,
      "grad_norm": 0.4954442083835602,
      "learning_rate": 1.7683339897601154e-05,
      "loss": 0.4185,
      "mean_token_accuracy": 0.8604020118713379,
      "num_tokens": 742549816.0,
      "step": 22165
    },
    {
      "epoch": 1.3234240687679084,
      "grad_norm": 0.5129901170730591,
      "learning_rate": 1.7673349366084312e-05,
      "loss": 0.4497,
      "mean_token_accuracy": 0.8501550674438476,
      "num_tokens": 742717496.0,
      "step": 22170
    },
    {
      "epoch": 1.323722540592168,
      "grad_norm": 0.5030074119567871,
      "learning_rate": 1.7663361228029612e-05,
      "loss": 0.4444,
      "mean_token_accuracy": 0.8514970779418946,
      "num_tokens": 742885176.0,
      "step": 22175
    },
    {
      "epoch": 1.324021012416428,
      "grad_norm": 0.5100350379943848,
      "learning_rate": 1.765337548586985e-05,
      "loss": 0.4335,
      "mean_token_accuracy": 0.8553143262863159,
      "num_tokens": 743052856.0,
      "step": 22180
    },
    {
      "epoch": 1.3243194842406876,
      "grad_norm": 0.5419145822525024,
      "learning_rate": 1.7643392142037236e-05,
      "loss": 0.4553,
      "mean_token_accuracy": 0.8496600151062011,
      "num_tokens": 743220536.0,
      "step": 22185
    },
    {
      "epoch": 1.3246179560649476,
      "grad_norm": 0.4562287926673889,
      "learning_rate": 1.7633411198963407e-05,
      "loss": 0.4078,
      "mean_token_accuracy": 0.8644419193267823,
      "num_tokens": 743382694.0,
      "step": 22190
    },
    {
      "epoch": 1.3249164278892072,
      "grad_norm": 0.47411105036735535,
      "learning_rate": 1.7623432659079388e-05,
      "loss": 0.4758,
      "mean_token_accuracy": 0.8425563693046569,
      "num_tokens": 743550374.0,
      "step": 22195
    },
    {
      "epoch": 1.3252148997134672,
      "grad_norm": 0.4742583930492401,
      "learning_rate": 1.761345652481564e-05,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8504473328590393,
      "num_tokens": 743718054.0,
      "step": 22200
    },
    {
      "epoch": 1.3255133715377267,
      "grad_norm": 0.5340882539749146,
      "learning_rate": 1.7603482798602045e-05,
      "loss": 0.4438,
      "mean_token_accuracy": 0.8522545576095581,
      "num_tokens": 743885734.0,
      "step": 22205
    },
    {
      "epoch": 1.3258118433619868,
      "grad_norm": 0.5445700287818909,
      "learning_rate": 1.7593511482867886e-05,
      "loss": 0.4315,
      "mean_token_accuracy": 0.855463445186615,
      "num_tokens": 744053414.0,
      "step": 22210
    },
    {
      "epoch": 1.3261103151862463,
      "grad_norm": 0.7758572101593018,
      "learning_rate": 1.7583542580041857e-05,
      "loss": 0.4618,
      "mean_token_accuracy": 0.8486937880516052,
      "num_tokens": 744221094.0,
      "step": 22215
    },
    {
      "epoch": 1.3264087870105061,
      "grad_norm": 0.5776500701904297,
      "learning_rate": 1.7573576092552063e-05,
      "loss": 0.4541,
      "mean_token_accuracy": 0.8487951874732971,
      "num_tokens": 744388774.0,
      "step": 22220
    },
    {
      "epoch": 1.326707258834766,
      "grad_norm": 0.5077141523361206,
      "learning_rate": 1.756361202282605e-05,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8538888216018676,
      "num_tokens": 744556454.0,
      "step": 22225
    },
    {
      "epoch": 1.3270057306590257,
      "grad_norm": 0.474239319562912,
      "learning_rate": 1.755365037329072e-05,
      "loss": 0.4394,
      "mean_token_accuracy": 0.8527615427970886,
      "num_tokens": 744724134.0,
      "step": 22230
    },
    {
      "epoch": 1.3273042024832855,
      "grad_norm": 0.5360094308853149,
      "learning_rate": 1.7543691146372428e-05,
      "loss": 0.4343,
      "mean_token_accuracy": 0.8550459265708923,
      "num_tokens": 744891814.0,
      "step": 22235
    },
    {
      "epoch": 1.3276026743075453,
      "grad_norm": 0.5197581052780151,
      "learning_rate": 1.7533734344496933e-05,
      "loss": 0.427,
      "mean_token_accuracy": 0.8574973106384277,
      "num_tokens": 745059494.0,
      "step": 22240
    },
    {
      "epoch": 1.327901146131805,
      "grad_norm": 0.5478337407112122,
      "learning_rate": 1.7523779970089395e-05,
      "loss": 0.4556,
      "mean_token_accuracy": 0.8478110551834106,
      "num_tokens": 745227174.0,
      "step": 22245
    },
    {
      "epoch": 1.328199617956065,
      "grad_norm": 0.5398106575012207,
      "learning_rate": 1.7513828025574393e-05,
      "loss": 0.4642,
      "mean_token_accuracy": 0.8458069920539856,
      "num_tokens": 745394854.0,
      "step": 22250
    },
    {
      "epoch": 1.3284980897803247,
      "grad_norm": 0.5628871917724609,
      "learning_rate": 1.750387851337589e-05,
      "loss": 0.4775,
      "mean_token_accuracy": 0.8413515448570251,
      "num_tokens": 745562534.0,
      "step": 22255
    },
    {
      "epoch": 1.3287965616045845,
      "grad_norm": 0.5625954270362854,
      "learning_rate": 1.749393143591727e-05,
      "loss": 0.4382,
      "mean_token_accuracy": 0.8533102631568908,
      "num_tokens": 745730214.0,
      "step": 22260
    },
    {
      "epoch": 1.3290950334288443,
      "grad_norm": 0.4975132644176483,
      "learning_rate": 1.7483986795621367e-05,
      "loss": 0.4541,
      "mean_token_accuracy": 0.848634135723114,
      "num_tokens": 745897894.0,
      "step": 22265
    },
    {
      "epoch": 1.329393505253104,
      "grad_norm": 0.4786888062953949,
      "learning_rate": 1.7474044594910334e-05,
      "loss": 0.4066,
      "mean_token_accuracy": 0.8636228084564209,
      "num_tokens": 746065574.0,
      "step": 22270
    },
    {
      "epoch": 1.3296919770773639,
      "grad_norm": 0.5369954109191895,
      "learning_rate": 1.7464104836205815e-05,
      "loss": 0.4602,
      "mean_token_accuracy": 0.8474531769752502,
      "num_tokens": 746233254.0,
      "step": 22275
    },
    {
      "epoch": 1.3299904489016237,
      "grad_norm": 0.46432721614837646,
      "learning_rate": 1.7454167521928783e-05,
      "loss": 0.484,
      "mean_token_accuracy": 0.8396337747573852,
      "num_tokens": 746400934.0,
      "step": 22280
    },
    {
      "epoch": 1.3302889207258835,
      "grad_norm": 0.4735514223575592,
      "learning_rate": 1.7444232654499687e-05,
      "loss": 0.4342,
      "mean_token_accuracy": 0.8554694056510925,
      "num_tokens": 746568614.0,
      "step": 22285
    },
    {
      "epoch": 1.3305873925501432,
      "grad_norm": 0.5196797251701355,
      "learning_rate": 1.743430023633834e-05,
      "loss": 0.451,
      "mean_token_accuracy": 0.8498270273208618,
      "num_tokens": 746736294.0,
      "step": 22290
    },
    {
      "epoch": 1.330885864374403,
      "grad_norm": 0.48476192355155945,
      "learning_rate": 1.742437026986396e-05,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8505964398384094,
      "num_tokens": 746903974.0,
      "step": 22295
    },
    {
      "epoch": 1.3311843361986628,
      "grad_norm": 0.5303208827972412,
      "learning_rate": 1.7414442757495173e-05,
      "loss": 0.4415,
      "mean_token_accuracy": 0.8522784352302551,
      "num_tokens": 747071654.0,
      "step": 22300
    },
    {
      "epoch": 1.3314828080229226,
      "grad_norm": 0.48505687713623047,
      "learning_rate": 1.7404517701650007e-05,
      "loss": 0.4322,
      "mean_token_accuracy": 0.85434809923172,
      "num_tokens": 747239334.0,
      "step": 22305
    },
    {
      "epoch": 1.3317812798471824,
      "grad_norm": 0.5132880806922913,
      "learning_rate": 1.7394595104745908e-05,
      "loss": 0.4431,
      "mean_token_accuracy": 0.8535667419433594,
      "num_tokens": 747407014.0,
      "step": 22310
    },
    {
      "epoch": 1.3320797516714422,
      "grad_norm": 0.5003713369369507,
      "learning_rate": 1.7384674969199688e-05,
      "loss": 0.4414,
      "mean_token_accuracy": 0.8522605419158935,
      "num_tokens": 747574694.0,
      "step": 22315
    },
    {
      "epoch": 1.332378223495702,
      "grad_norm": 0.4654309153556824,
      "learning_rate": 1.7374757297427584e-05,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8427293300628662,
      "num_tokens": 747742374.0,
      "step": 22320
    },
    {
      "epoch": 1.3326766953199618,
      "grad_norm": 0.5223366022109985,
      "learning_rate": 1.7364842091845234e-05,
      "loss": 0.4046,
      "mean_token_accuracy": 0.8646665811538696,
      "num_tokens": 747910054.0,
      "step": 22325
    },
    {
      "epoch": 1.3329751671442216,
      "grad_norm": 0.47959938645362854,
      "learning_rate": 1.7354929354867667e-05,
      "loss": 0.455,
      "mean_token_accuracy": 0.8486102819442749,
      "num_tokens": 748077734.0,
      "step": 22330
    },
    {
      "epoch": 1.3332736389684814,
      "grad_norm": 0.5158442258834839,
      "learning_rate": 1.7345019088909314e-05,
      "loss": 0.4638,
      "mean_token_accuracy": 0.8464021801948547,
      "num_tokens": 748237116.0,
      "step": 22335
    },
    {
      "epoch": 1.3335721107927412,
      "grad_norm": 0.5109057426452637,
      "learning_rate": 1.7335111296383987e-05,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8424370765686036,
      "num_tokens": 748404796.0,
      "step": 22340
    },
    {
      "epoch": 1.333870582617001,
      "grad_norm": 0.5325325727462769,
      "learning_rate": 1.7325205979704936e-05,
      "loss": 0.4558,
      "mean_token_accuracy": 0.8487951874732971,
      "num_tokens": 748572476.0,
      "step": 22345
    },
    {
      "epoch": 1.3341690544412608,
      "grad_norm": 0.5609932541847229,
      "learning_rate": 1.7315303141284782e-05,
      "loss": 0.4718,
      "mean_token_accuracy": 0.8445723533630372,
      "num_tokens": 748740156.0,
      "step": 22350
    },
    {
      "epoch": 1.3344675262655206,
      "grad_norm": 0.503879189491272,
      "learning_rate": 1.730540278353553e-05,
      "loss": 0.4839,
      "mean_token_accuracy": 0.8394071459770203,
      "num_tokens": 748907836.0,
      "step": 22355
    },
    {
      "epoch": 1.3347659980897804,
      "grad_norm": 0.4853704869747162,
      "learning_rate": 1.7295504908868597e-05,
      "loss": 0.4301,
      "mean_token_accuracy": 0.8537755012512207,
      "num_tokens": 749075516.0,
      "step": 22360
    },
    {
      "epoch": 1.3350644699140402,
      "grad_norm": 0.45328521728515625,
      "learning_rate": 1.72856095196948e-05,
      "loss": 0.4504,
      "mean_token_accuracy": 0.8498270273208618,
      "num_tokens": 749243196.0,
      "step": 22365
    },
    {
      "epoch": 1.3353629417383,
      "grad_norm": 0.4905661642551422,
      "learning_rate": 1.7275716618424337e-05,
      "loss": 0.4571,
      "mean_token_accuracy": 0.84889657497406,
      "num_tokens": 749410876.0,
      "step": 22370
    },
    {
      "epoch": 1.3356614135625597,
      "grad_norm": 0.4668544828891754,
      "learning_rate": 1.7265826207466823e-05,
      "loss": 0.4246,
      "mean_token_accuracy": 0.8590003609657287,
      "num_tokens": 749578556.0,
      "step": 22375
    },
    {
      "epoch": 1.3359598853868195,
      "grad_norm": 0.4947129487991333,
      "learning_rate": 1.7255938289231227e-05,
      "loss": 0.4175,
      "mean_token_accuracy": 0.8594894528388977,
      "num_tokens": 749746236.0,
      "step": 22380
    },
    {
      "epoch": 1.3362583572110793,
      "grad_norm": 0.49391013383865356,
      "learning_rate": 1.7246052866125943e-05,
      "loss": 0.4471,
      "mean_token_accuracy": 0.8513598799705505,
      "num_tokens": 749913916.0,
      "step": 22385
    },
    {
      "epoch": 1.3365568290353391,
      "grad_norm": 0.5421164631843567,
      "learning_rate": 1.723616994055875e-05,
      "loss": 0.4498,
      "mean_token_accuracy": 0.8504354000091553,
      "num_tokens": 750081596.0,
      "step": 22390
    },
    {
      "epoch": 1.336855300859599,
      "grad_norm": 0.4698849618434906,
      "learning_rate": 1.7226289514936815e-05,
      "loss": 0.4279,
      "mean_token_accuracy": 0.8579625487327576,
      "num_tokens": 750249276.0,
      "step": 22395
    },
    {
      "epoch": 1.3371537726838587,
      "grad_norm": 0.47517839074134827,
      "learning_rate": 1.7216411591666713e-05,
      "loss": 0.4185,
      "mean_token_accuracy": 0.8602886915206909,
      "num_tokens": 750416956.0,
      "step": 22400
    },
    {
      "epoch": 1.3374522445081185,
      "grad_norm": 0.5229873061180115,
      "learning_rate": 1.720653617315436e-05,
      "loss": 0.4241,
      "mean_token_accuracy": 0.8571513772010804,
      "num_tokens": 750584636.0,
      "step": 22405
    },
    {
      "epoch": 1.3377507163323783,
      "grad_norm": 0.5081853270530701,
      "learning_rate": 1.7196663261805137e-05,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8521591305732727,
      "num_tokens": 750752316.0,
      "step": 22410
    },
    {
      "epoch": 1.338049188156638,
      "grad_norm": 0.6048567891120911,
      "learning_rate": 1.7186792860023743e-05,
      "loss": 0.445,
      "mean_token_accuracy": 0.8530359148979187,
      "num_tokens": 750919996.0,
      "step": 22415
    },
    {
      "epoch": 1.3383476599808979,
      "grad_norm": 0.500054121017456,
      "learning_rate": 1.717692497021432e-05,
      "loss": 0.4715,
      "mean_token_accuracy": 0.8435524344444275,
      "num_tokens": 751087676.0,
      "step": 22420
    },
    {
      "epoch": 1.3386461318051577,
      "grad_norm": 0.5501048564910889,
      "learning_rate": 1.7167059594780344e-05,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8502564787864685,
      "num_tokens": 751255356.0,
      "step": 22425
    },
    {
      "epoch": 1.3389446036294173,
      "grad_norm": 0.48309409618377686,
      "learning_rate": 1.7157196736124735e-05,
      "loss": 0.4228,
      "mean_token_accuracy": 0.857455563545227,
      "num_tokens": 751423036.0,
      "step": 22430
    },
    {
      "epoch": 1.3392430754536773,
      "grad_norm": 0.47978973388671875,
      "learning_rate": 1.7147336396649777e-05,
      "loss": 0.4212,
      "mean_token_accuracy": 0.8584874153137207,
      "num_tokens": 751590716.0,
      "step": 22435
    },
    {
      "epoch": 1.3395415472779368,
      "grad_norm": 0.5024915933609009,
      "learning_rate": 1.713747857875712e-05,
      "loss": 0.4249,
      "mean_token_accuracy": 0.8580553293228149,
      "num_tokens": 751756682.0,
      "step": 22440
    },
    {
      "epoch": 1.3398400191021969,
      "grad_norm": 0.5611995458602905,
      "learning_rate": 1.7127623284847827e-05,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8371108174324036,
      "num_tokens": 751924362.0,
      "step": 22445
    },
    {
      "epoch": 1.3401384909264564,
      "grad_norm": 0.5121062397956848,
      "learning_rate": 1.711777051732233e-05,
      "loss": 0.4299,
      "mean_token_accuracy": 0.8562328577041626,
      "num_tokens": 752092042.0,
      "step": 22450
    },
    {
      "epoch": 1.3404369627507164,
      "grad_norm": 0.6843779683113098,
      "learning_rate": 1.710792027858046e-05,
      "loss": 0.437,
      "mean_token_accuracy": 0.8552904844284057,
      "num_tokens": 752259722.0,
      "step": 22455
    },
    {
      "epoch": 1.340735434574976,
      "grad_norm": 0.5093886852264404,
      "learning_rate": 1.7098072571021432e-05,
      "loss": 0.473,
      "mean_token_accuracy": 0.8442622065544129,
      "num_tokens": 752427402.0,
      "step": 22460
    },
    {
      "epoch": 1.341033906399236,
      "grad_norm": 0.48257824778556824,
      "learning_rate": 1.7088227397043805e-05,
      "loss": 0.4188,
      "mean_token_accuracy": 0.8596325874328613,
      "num_tokens": 752595082.0,
      "step": 22465
    },
    {
      "epoch": 1.3413323782234956,
      "grad_norm": 0.5374806523323059,
      "learning_rate": 1.70783847590456e-05,
      "loss": 0.4274,
      "mean_token_accuracy": 0.8561076045036315,
      "num_tokens": 752762762.0,
      "step": 22470
    },
    {
      "epoch": 1.3416308500477556,
      "grad_norm": 0.5240042209625244,
      "learning_rate": 1.7068544659424133e-05,
      "loss": 0.4326,
      "mean_token_accuracy": 0.8562090039253235,
      "num_tokens": 752930442.0,
      "step": 22475
    },
    {
      "epoch": 1.3419293218720152,
      "grad_norm": 0.5293153524398804,
      "learning_rate": 1.705870710057616e-05,
      "loss": 0.4258,
      "mean_token_accuracy": 0.8556304454803467,
      "num_tokens": 753098122.0,
      "step": 22480
    },
    {
      "epoch": 1.3422277936962752,
      "grad_norm": 0.5044662952423096,
      "learning_rate": 1.7048872084897793e-05,
      "loss": 0.4608,
      "mean_token_accuracy": 0.8467255115509034,
      "num_tokens": 753265802.0,
      "step": 22485
    },
    {
      "epoch": 1.3425262655205348,
      "grad_norm": 0.5189728736877441,
      "learning_rate": 1.7039039614784537e-05,
      "loss": 0.4435,
      "mean_token_accuracy": 0.852606475353241,
      "num_tokens": 753433482.0,
      "step": 22490
    },
    {
      "epoch": 1.3428247373447946,
      "grad_norm": 0.5683044791221619,
      "learning_rate": 1.7029209692631283e-05,
      "loss": 0.4554,
      "mean_token_accuracy": 0.8494452953338623,
      "num_tokens": 753601162.0,
      "step": 22495
    },
    {
      "epoch": 1.3431232091690544,
      "grad_norm": 0.5045459270477295,
      "learning_rate": 1.701938232083226e-05,
      "loss": 0.4062,
      "mean_token_accuracy": 0.8633782625198364,
      "num_tokens": 753768842.0,
      "step": 22500
    },
    {
      "epoch": 1.3434216809933142,
      "grad_norm": 0.46433112025260925,
      "learning_rate": 1.7009557501781132e-05,
      "loss": 0.453,
      "mean_token_accuracy": 0.8479422688484192,
      "num_tokens": 753936522.0,
      "step": 22505
    },
    {
      "epoch": 1.343720152817574,
      "grad_norm": 0.4946271777153015,
      "learning_rate": 1.69997352378709e-05,
      "loss": 0.4574,
      "mean_token_accuracy": 0.8473040580749511,
      "num_tokens": 754104202.0,
      "step": 22510
    },
    {
      "epoch": 1.3440186246418337,
      "grad_norm": 0.5034542083740234,
      "learning_rate": 1.6989915531493963e-05,
      "loss": 0.4235,
      "mean_token_accuracy": 0.8585232019424438,
      "num_tokens": 754271882.0,
      "step": 22515
    },
    {
      "epoch": 1.3443170964660935,
      "grad_norm": 0.5567042827606201,
      "learning_rate": 1.6980098385042108e-05,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8463497638702393,
      "num_tokens": 754439562.0,
      "step": 22520
    },
    {
      "epoch": 1.3446155682903533,
      "grad_norm": 0.4457235634326935,
      "learning_rate": 1.6970283800906446e-05,
      "loss": 0.4504,
      "mean_token_accuracy": 0.8513956665992737,
      "num_tokens": 754607242.0,
      "step": 22525
    },
    {
      "epoch": 1.3449140401146131,
      "grad_norm": 0.5092954039573669,
      "learning_rate": 1.6960471781477532e-05,
      "loss": 0.4536,
      "mean_token_accuracy": 0.8478408694267273,
      "num_tokens": 754774922.0,
      "step": 22530
    },
    {
      "epoch": 1.345212511938873,
      "grad_norm": 0.5446640253067017,
      "learning_rate": 1.6950662329145246e-05,
      "loss": 0.4523,
      "mean_token_accuracy": 0.8477275490760803,
      "num_tokens": 754942602.0,
      "step": 22535
    },
    {
      "epoch": 1.3455109837631327,
      "grad_norm": 0.5287734270095825,
      "learning_rate": 1.6940855446298874e-05,
      "loss": 0.4668,
      "mean_token_accuracy": 0.8447274208068848,
      "num_tokens": 755110282.0,
      "step": 22540
    },
    {
      "epoch": 1.3458094555873925,
      "grad_norm": 0.5278980731964111,
      "learning_rate": 1.693105113532706e-05,
      "loss": 0.4604,
      "mean_token_accuracy": 0.849773359298706,
      "num_tokens": 755277962.0,
      "step": 22545
    },
    {
      "epoch": 1.3461079274116523,
      "grad_norm": 0.5189603567123413,
      "learning_rate": 1.692124939861781e-05,
      "loss": 0.4353,
      "mean_token_accuracy": 0.8555469393730164,
      "num_tokens": 755445642.0,
      "step": 22550
    },
    {
      "epoch": 1.346406399235912,
      "grad_norm": 0.5070314407348633,
      "learning_rate": 1.6911450238558544e-05,
      "loss": 0.4381,
      "mean_token_accuracy": 0.8529345035552979,
      "num_tokens": 755613322.0,
      "step": 22555
    },
    {
      "epoch": 1.346704871060172,
      "grad_norm": 0.4751354455947876,
      "learning_rate": 1.690165365753601e-05,
      "loss": 0.3935,
      "mean_token_accuracy": 0.8694918274879455,
      "num_tokens": 755781002.0,
      "step": 22560
    },
    {
      "epoch": 1.3470033428844317,
      "grad_norm": 0.5475949645042419,
      "learning_rate": 1.6891859657936346e-05,
      "loss": 0.4479,
      "mean_token_accuracy": 0.8498329877853393,
      "num_tokens": 755948682.0,
      "step": 22565
    },
    {
      "epoch": 1.3473018147086915,
      "grad_norm": 0.48246076703071594,
      "learning_rate": 1.6882068242145067e-05,
      "loss": 0.438,
      "mean_token_accuracy": 0.8531313538551331,
      "num_tokens": 756116362.0,
      "step": 22570
    },
    {
      "epoch": 1.3476002865329513,
      "grad_norm": 0.5266431570053101,
      "learning_rate": 1.6872279412547055e-05,
      "loss": 0.4304,
      "mean_token_accuracy": 0.8557318329811097,
      "num_tokens": 756284042.0,
      "step": 22575
    },
    {
      "epoch": 1.347898758357211,
      "grad_norm": 0.5577112436294556,
      "learning_rate": 1.686249317152656e-05,
      "loss": 0.4453,
      "mean_token_accuracy": 0.8521591424942017,
      "num_tokens": 756451722.0,
      "step": 22580
    },
    {
      "epoch": 1.3481972301814709,
      "grad_norm": 0.50678950548172,
      "learning_rate": 1.6852709521467198e-05,
      "loss": 0.4117,
      "mean_token_accuracy": 0.8616903305053711,
      "num_tokens": 756619402.0,
      "step": 22585
    },
    {
      "epoch": 1.3484957020057307,
      "grad_norm": 0.5217883586883545,
      "learning_rate": 1.6842928464751946e-05,
      "loss": 0.4564,
      "mean_token_accuracy": 0.84849693775177,
      "num_tokens": 756787082.0,
      "step": 22590
    },
    {
      "epoch": 1.3487941738299905,
      "grad_norm": 0.47241899371147156,
      "learning_rate": 1.6833150003763195e-05,
      "loss": 0.4282,
      "mean_token_accuracy": 0.856179165840149,
      "num_tokens": 756954762.0,
      "step": 22595
    },
    {
      "epoch": 1.3490926456542502,
      "grad_norm": 0.4343146085739136,
      "learning_rate": 1.6823374140882634e-05,
      "loss": 0.4232,
      "mean_token_accuracy": 0.8582846283912658,
      "num_tokens": 757122442.0,
      "step": 22600
    },
    {
      "epoch": 1.34939111747851,
      "grad_norm": 0.5446040034294128,
      "learning_rate": 1.6813600878491376e-05,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8460023999214172,
      "num_tokens": 757288541.0,
      "step": 22605
    },
    {
      "epoch": 1.3496895893027698,
      "grad_norm": 0.5250765085220337,
      "learning_rate": 1.6803830218969857e-05,
      "loss": 0.4535,
      "mean_token_accuracy": 0.8486818552017212,
      "num_tokens": 757456221.0,
      "step": 22610
    },
    {
      "epoch": 1.3499880611270296,
      "grad_norm": 0.5114559531211853,
      "learning_rate": 1.6794062164697928e-05,
      "loss": 0.4254,
      "mean_token_accuracy": 0.8572885751724243,
      "num_tokens": 757623901.0,
      "step": 22615
    },
    {
      "epoch": 1.3502865329512894,
      "grad_norm": 0.511920690536499,
      "learning_rate": 1.678429671805478e-05,
      "loss": 0.455,
      "mean_token_accuracy": 0.8480973362922668,
      "num_tokens": 757791581.0,
      "step": 22620
    },
    {
      "epoch": 1.3505850047755492,
      "grad_norm": 0.5332680940628052,
      "learning_rate": 1.677453388141894e-05,
      "loss": 0.3948,
      "mean_token_accuracy": 0.8664439916610718,
      "num_tokens": 757959261.0,
      "step": 22625
    },
    {
      "epoch": 1.350883476599809,
      "grad_norm": 0.5050813555717468,
      "learning_rate": 1.6764773657168346e-05,
      "loss": 0.4368,
      "mean_token_accuracy": 0.8537695288658143,
      "num_tokens": 758126941.0,
      "step": 22630
    },
    {
      "epoch": 1.3511819484240688,
      "grad_norm": 0.5128730535507202,
      "learning_rate": 1.6755016047680276e-05,
      "loss": 0.4593,
      "mean_token_accuracy": 0.845890486240387,
      "num_tokens": 758294621.0,
      "step": 22635
    },
    {
      "epoch": 1.3514804202483286,
      "grad_norm": 0.5740980505943298,
      "learning_rate": 1.6745261055331384e-05,
      "loss": 0.4493,
      "mean_token_accuracy": 0.8507873177528381,
      "num_tokens": 758462301.0,
      "step": 22640
    },
    {
      "epoch": 1.3517788920725884,
      "grad_norm": 0.4692186117172241,
      "learning_rate": 1.6735508682497662e-05,
      "loss": 0.3948,
      "mean_token_accuracy": 0.8667660713195801,
      "num_tokens": 758629981.0,
      "step": 22645
    },
    {
      "epoch": 1.3520773638968482,
      "grad_norm": 0.5040562748908997,
      "learning_rate": 1.6725758931554496e-05,
      "loss": 0.4582,
      "mean_token_accuracy": 0.8474650979042053,
      "num_tokens": 758797661.0,
      "step": 22650
    },
    {
      "epoch": 1.352375835721108,
      "grad_norm": 0.5797613263130188,
      "learning_rate": 1.671601180487661e-05,
      "loss": 0.4563,
      "mean_token_accuracy": 0.847769296169281,
      "num_tokens": 758965341.0,
      "step": 22655
    },
    {
      "epoch": 1.3526743075453678,
      "grad_norm": 0.5466217398643494,
      "learning_rate": 1.6706267304838097e-05,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8430215954780579,
      "num_tokens": 759133021.0,
      "step": 22660
    },
    {
      "epoch": 1.3529727793696276,
      "grad_norm": 0.5243255496025085,
      "learning_rate": 1.669652543381242e-05,
      "loss": 0.4604,
      "mean_token_accuracy": 0.8472503900527955,
      "num_tokens": 759300701.0,
      "step": 22665
    },
    {
      "epoch": 1.3532712511938874,
      "grad_norm": 0.5923428535461426,
      "learning_rate": 1.668678619417238e-05,
      "loss": 0.4488,
      "mean_token_accuracy": 0.8497972130775452,
      "num_tokens": 759468381.0,
      "step": 22670
    },
    {
      "epoch": 1.3535697230181472,
      "grad_norm": 0.5448712706565857,
      "learning_rate": 1.6677049588290132e-05,
      "loss": 0.4754,
      "mean_token_accuracy": 0.8438029527664185,
      "num_tokens": 759636061.0,
      "step": 22675
    },
    {
      "epoch": 1.353868194842407,
      "grad_norm": 0.48216933012008667,
      "learning_rate": 1.6667315618537248e-05,
      "loss": 0.4344,
      "mean_token_accuracy": 0.8535548210144043,
      "num_tokens": 759803741.0,
      "step": 22680
    },
    {
      "epoch": 1.3541666666666667,
      "grad_norm": 0.47375988960266113,
      "learning_rate": 1.6657584287284588e-05,
      "loss": 0.4325,
      "mean_token_accuracy": 0.8546522736549378,
      "num_tokens": 759971421.0,
      "step": 22685
    },
    {
      "epoch": 1.3544651384909265,
      "grad_norm": 0.500453770160675,
      "learning_rate": 1.66478555969024e-05,
      "loss": 0.4166,
      "mean_token_accuracy": 0.8598115086555481,
      "num_tokens": 760139101.0,
      "step": 22690
    },
    {
      "epoch": 1.3547636103151863,
      "grad_norm": 0.5385774374008179,
      "learning_rate": 1.663812954976029e-05,
      "loss": 0.4667,
      "mean_token_accuracy": 0.845162832736969,
      "num_tokens": 760306781.0,
      "step": 22695
    },
    {
      "epoch": 1.3550620821394461,
      "grad_norm": 0.5080847144126892,
      "learning_rate": 1.6628406148227216e-05,
      "loss": 0.4656,
      "mean_token_accuracy": 0.8456340312957764,
      "num_tokens": 760474461.0,
      "step": 22700
    },
    {
      "epoch": 1.3553605539637057,
      "grad_norm": 0.521116316318512,
      "learning_rate": 1.6618685394671495e-05,
      "loss": 0.4402,
      "mean_token_accuracy": 0.8540498614311218,
      "num_tokens": 760642141.0,
      "step": 22705
    },
    {
      "epoch": 1.3556590257879657,
      "grad_norm": 0.5268535017967224,
      "learning_rate": 1.6608967291460787e-05,
      "loss": 0.4284,
      "mean_token_accuracy": 0.8576523900032044,
      "num_tokens": 760809821.0,
      "step": 22710
    },
    {
      "epoch": 1.3559574976122253,
      "grad_norm": 0.46757519245147705,
      "learning_rate": 1.6599251840962125e-05,
      "loss": 0.4173,
      "mean_token_accuracy": 0.8612072110176087,
      "num_tokens": 760977501.0,
      "step": 22715
    },
    {
      "epoch": 1.3562559694364853,
      "grad_norm": 0.5216953158378601,
      "learning_rate": 1.6589539045541876e-05,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8494214415550232,
      "num_tokens": 761145181.0,
      "step": 22720
    },
    {
      "epoch": 1.3565544412607449,
      "grad_norm": 0.48437798023223877,
      "learning_rate": 1.6579828907565774e-05,
      "loss": 0.4189,
      "mean_token_accuracy": 0.8591196417808533,
      "num_tokens": 761312861.0,
      "step": 22725
    },
    {
      "epoch": 1.3568529130850049,
      "grad_norm": 0.5425718426704407,
      "learning_rate": 1.657012142939892e-05,
      "loss": 0.482,
      "mean_token_accuracy": 0.8402302145957947,
      "num_tokens": 761480541.0,
      "step": 22730
    },
    {
      "epoch": 1.3571513849092645,
      "grad_norm": 0.46816858649253845,
      "learning_rate": 1.6560416613405714e-05,
      "loss": 0.4567,
      "mean_token_accuracy": 0.848496961593628,
      "num_tokens": 761648221.0,
      "step": 22735
    },
    {
      "epoch": 1.3574498567335245,
      "grad_norm": 0.50034499168396,
      "learning_rate": 1.6550714461949983e-05,
      "loss": 0.4856,
      "mean_token_accuracy": 0.8382380962371826,
      "num_tokens": 761815901.0,
      "step": 22740
    },
    {
      "epoch": 1.357748328557784,
      "grad_norm": 0.5078026056289673,
      "learning_rate": 1.6541014977394833e-05,
      "loss": 0.437,
      "mean_token_accuracy": 0.8542407155036926,
      "num_tokens": 761983581.0,
      "step": 22745
    },
    {
      "epoch": 1.358046800382044,
      "grad_norm": 0.49565309286117554,
      "learning_rate": 1.6531318162102778e-05,
      "loss": 0.4536,
      "mean_token_accuracy": 0.8489264011383056,
      "num_tokens": 762151261.0,
      "step": 22750
    },
    {
      "epoch": 1.3583452722063036,
      "grad_norm": 0.43293872475624084,
      "learning_rate": 1.652162401843562e-05,
      "loss": 0.4114,
      "mean_token_accuracy": 0.8619587302207947,
      "num_tokens": 762318941.0,
      "step": 22755
    },
    {
      "epoch": 1.3586437440305636,
      "grad_norm": 0.4872153401374817,
      "learning_rate": 1.6511932548754585e-05,
      "loss": 0.4158,
      "mean_token_accuracy": 0.8595908403396606,
      "num_tokens": 762486621.0,
      "step": 22760
    },
    {
      "epoch": 1.3589422158548232,
      "grad_norm": 0.4540598392486572,
      "learning_rate": 1.6502243755420202e-05,
      "loss": 0.4115,
      "mean_token_accuracy": 0.8612668514251709,
      "num_tokens": 762654301.0,
      "step": 22765
    },
    {
      "epoch": 1.359240687679083,
      "grad_norm": 0.5416213274002075,
      "learning_rate": 1.6492557640792335e-05,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8563521385192872,
      "num_tokens": 762821981.0,
      "step": 22770
    },
    {
      "epoch": 1.3595391595033428,
      "grad_norm": 0.5540740489959717,
      "learning_rate": 1.6482874207230232e-05,
      "loss": 0.4339,
      "mean_token_accuracy": 0.8556304454803467,
      "num_tokens": 762989661.0,
      "step": 22775
    },
    {
      "epoch": 1.3598376313276026,
      "grad_norm": 0.5466917753219604,
      "learning_rate": 1.6473193457092468e-05,
      "loss": 0.4316,
      "mean_token_accuracy": 0.8545926213264465,
      "num_tokens": 763157341.0,
      "step": 22780
    },
    {
      "epoch": 1.3601361031518624,
      "grad_norm": 0.53810715675354,
      "learning_rate": 1.6463515392736967e-05,
      "loss": 0.4579,
      "mean_token_accuracy": 0.8482643485069274,
      "num_tokens": 763325021.0,
      "step": 22785
    },
    {
      "epoch": 1.3604345749761222,
      "grad_norm": 0.4831179976463318,
      "learning_rate": 1.6453840016521016e-05,
      "loss": 0.4552,
      "mean_token_accuracy": 0.8488429069519043,
      "num_tokens": 763492701.0,
      "step": 22790
    },
    {
      "epoch": 1.360733046800382,
      "grad_norm": 0.4876408278942108,
      "learning_rate": 1.6444167330801203e-05,
      "loss": 0.4372,
      "mean_token_accuracy": 0.854485273361206,
      "num_tokens": 763660381.0,
      "step": 22795
    },
    {
      "epoch": 1.3610315186246418,
      "grad_norm": 0.5168662071228027,
      "learning_rate": 1.6434497337933506e-05,
      "loss": 0.4232,
      "mean_token_accuracy": 0.8589884281158447,
      "num_tokens": 763828061.0,
      "step": 22800
    },
    {
      "epoch": 1.3613299904489016,
      "grad_norm": 0.5481839179992676,
      "learning_rate": 1.642483004027323e-05,
      "loss": 0.4506,
      "mean_token_accuracy": 0.8492007493972779,
      "num_tokens": 763995741.0,
      "step": 22805
    },
    {
      "epoch": 1.3616284622731614,
      "grad_norm": 0.5525104403495789,
      "learning_rate": 1.641516544017502e-05,
      "loss": 0.4724,
      "mean_token_accuracy": 0.8434033036231995,
      "num_tokens": 764163421.0,
      "step": 22810
    },
    {
      "epoch": 1.3619269340974212,
      "grad_norm": 0.5119472146034241,
      "learning_rate": 1.640550353999287e-05,
      "loss": 0.4888,
      "mean_token_accuracy": 0.838446855545044,
      "num_tokens": 764331101.0,
      "step": 22815
    },
    {
      "epoch": 1.362225405921681,
      "grad_norm": 0.46866685152053833,
      "learning_rate": 1.6395844342080114e-05,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8503459334373474,
      "num_tokens": 764498781.0,
      "step": 22820
    },
    {
      "epoch": 1.3625238777459407,
      "grad_norm": 0.5765346884727478,
      "learning_rate": 1.6386187848789436e-05,
      "loss": 0.4822,
      "mean_token_accuracy": 0.8402532815933228,
      "num_tokens": 764660775.0,
      "step": 22825
    },
    {
      "epoch": 1.3628223495702005,
      "grad_norm": 0.6466535329818726,
      "learning_rate": 1.6376534062472838e-05,
      "loss": 0.4577,
      "mean_token_accuracy": 0.8486579895019531,
      "num_tokens": 764828455.0,
      "step": 22830
    },
    {
      "epoch": 1.3631208213944603,
      "grad_norm": 0.5960538387298584,
      "learning_rate": 1.636688298548168e-05,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8417690515518188,
      "num_tokens": 764996135.0,
      "step": 22835
    },
    {
      "epoch": 1.3634192932187201,
      "grad_norm": 0.5411390662193298,
      "learning_rate": 1.6357234620166668e-05,
      "loss": 0.4798,
      "mean_token_accuracy": 0.8389538407325745,
      "num_tokens": 765163815.0,
      "step": 22840
    },
    {
      "epoch": 1.36371776504298,
      "grad_norm": 0.5571218132972717,
      "learning_rate": 1.6347588968877837e-05,
      "loss": 0.4251,
      "mean_token_accuracy": 0.8586663484573365,
      "num_tokens": 765331495.0,
      "step": 22845
    },
    {
      "epoch": 1.3640162368672397,
      "grad_norm": 0.510390043258667,
      "learning_rate": 1.633794603396457e-05,
      "loss": 0.4003,
      "mean_token_accuracy": 0.8647376418113708,
      "num_tokens": 765491234.0,
      "step": 22850
    },
    {
      "epoch": 1.3643147086914995,
      "grad_norm": 0.46171265840530396,
      "learning_rate": 1.6328305817775566e-05,
      "loss": 0.4223,
      "mean_token_accuracy": 0.8602051854133606,
      "num_tokens": 765658914.0,
      "step": 22855
    },
    {
      "epoch": 1.3646131805157593,
      "grad_norm": 0.5314909219741821,
      "learning_rate": 1.6318668322658888e-05,
      "loss": 0.464,
      "mean_token_accuracy": 0.844560420513153,
      "num_tokens": 765826594.0,
      "step": 22860
    },
    {
      "epoch": 1.364911652340019,
      "grad_norm": 0.525103747844696,
      "learning_rate": 1.6309033550961918e-05,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8451986193656922,
      "num_tokens": 765994274.0,
      "step": 22865
    },
    {
      "epoch": 1.365210124164279,
      "grad_norm": 0.5250957608222961,
      "learning_rate": 1.6299401505031395e-05,
      "loss": 0.4322,
      "mean_token_accuracy": 0.8557735919952393,
      "num_tokens": 766161954.0,
      "step": 22870
    },
    {
      "epoch": 1.3655085959885387,
      "grad_norm": 0.48402971029281616,
      "learning_rate": 1.6289772187213378e-05,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8474830031394959,
      "num_tokens": 766329634.0,
      "step": 22875
    },
    {
      "epoch": 1.3658070678127985,
      "grad_norm": 0.4996907413005829,
      "learning_rate": 1.6280145599853246e-05,
      "loss": 0.4344,
      "mean_token_accuracy": 0.8563044190406799,
      "num_tokens": 766497314.0,
      "step": 22880
    },
    {
      "epoch": 1.3661055396370583,
      "grad_norm": 0.5941933989524841,
      "learning_rate": 1.627052174529577e-05,
      "loss": 0.4488,
      "mean_token_accuracy": 0.8501192927360535,
      "num_tokens": 766664994.0,
      "step": 22885
    },
    {
      "epoch": 1.366404011461318,
      "grad_norm": 0.5185617804527283,
      "learning_rate": 1.626090062588498e-05,
      "loss": 0.445,
      "mean_token_accuracy": 0.851240611076355,
      "num_tokens": 766832674.0,
      "step": 22890
    },
    {
      "epoch": 1.3667024832855779,
      "grad_norm": 0.5115506052970886,
      "learning_rate": 1.6251282243964295e-05,
      "loss": 0.481,
      "mean_token_accuracy": 0.8410831451416015,
      "num_tokens": 767000354.0,
      "step": 22895
    },
    {
      "epoch": 1.3670009551098377,
      "grad_norm": 0.4756830334663391,
      "learning_rate": 1.624166660187645e-05,
      "loss": 0.4195,
      "mean_token_accuracy": 0.8597459197044373,
      "num_tokens": 767168034.0,
      "step": 22900
    },
    {
      "epoch": 1.3672994269340975,
      "grad_norm": 0.492611825466156,
      "learning_rate": 1.6232053701963503e-05,
      "loss": 0.4567,
      "mean_token_accuracy": 0.8500656247138977,
      "num_tokens": 767335714.0,
      "step": 22905
    },
    {
      "epoch": 1.3675978987583572,
      "grad_norm": 0.5117066502571106,
      "learning_rate": 1.6222443546566874e-05,
      "loss": 0.42,
      "mean_token_accuracy": 0.8589168548583984,
      "num_tokens": 767503394.0,
      "step": 22910
    },
    {
      "epoch": 1.367896370582617,
      "grad_norm": 0.5237162113189697,
      "learning_rate": 1.6212836138027268e-05,
      "loss": 0.4135,
      "mean_token_accuracy": 0.8620004773139953,
      "num_tokens": 767671074.0,
      "step": 22915
    },
    {
      "epoch": 1.3681948424068768,
      "grad_norm": 0.4933103024959564,
      "learning_rate": 1.6203231478684755e-05,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8567100048065186,
      "num_tokens": 767838754.0,
      "step": 22920
    },
    {
      "epoch": 1.3684933142311366,
      "grad_norm": 0.5579466223716736,
      "learning_rate": 1.6193629570878735e-05,
      "loss": 0.441,
      "mean_token_accuracy": 0.8531969547271728,
      "num_tokens": 768006434.0,
      "step": 22925
    },
    {
      "epoch": 1.3687917860553964,
      "grad_norm": 0.5770542621612549,
      "learning_rate": 1.618403041694792e-05,
      "loss": 0.4476,
      "mean_token_accuracy": 0.8525587439537048,
      "num_tokens": 768174114.0,
      "step": 22930
    },
    {
      "epoch": 1.3690902578796562,
      "grad_norm": 0.5308911800384521,
      "learning_rate": 1.6174434019230377e-05,
      "loss": 0.4437,
      "mean_token_accuracy": 0.8527257561683654,
      "num_tokens": 768341794.0,
      "step": 22935
    },
    {
      "epoch": 1.369388729703916,
      "grad_norm": 0.575415313243866,
      "learning_rate": 1.6164840380063457e-05,
      "loss": 0.4367,
      "mean_token_accuracy": 0.8536621809005738,
      "num_tokens": 768509474.0,
      "step": 22940
    },
    {
      "epoch": 1.3696872015281758,
      "grad_norm": 0.546991229057312,
      "learning_rate": 1.6155249501783905e-05,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8393296003341675,
      "num_tokens": 768677154.0,
      "step": 22945
    },
    {
      "epoch": 1.3699856733524356,
      "grad_norm": 0.5168837904930115,
      "learning_rate": 1.6145661386727726e-05,
      "loss": 0.452,
      "mean_token_accuracy": 0.8490039348602295,
      "num_tokens": 768844834.0,
      "step": 22950
    },
    {
      "epoch": 1.3702841451766954,
      "grad_norm": 0.46434494853019714,
      "learning_rate": 1.6136076037230294e-05,
      "loss": 0.4432,
      "mean_token_accuracy": 0.853576409816742,
      "num_tokens": 769010588.0,
      "step": 22955
    },
    {
      "epoch": 1.3705826170009552,
      "grad_norm": 0.8401955962181091,
      "learning_rate": 1.6126493455626296e-05,
      "loss": 0.4202,
      "mean_token_accuracy": 0.8608910799026489,
      "num_tokens": 769178268.0,
      "step": 22960
    },
    {
      "epoch": 1.370881088825215,
      "grad_norm": 0.5199339985847473,
      "learning_rate": 1.6116913644249747e-05,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8502803325653077,
      "num_tokens": 769345948.0,
      "step": 22965
    },
    {
      "epoch": 1.3711795606494748,
      "grad_norm": 0.48886939883232117,
      "learning_rate": 1.6107336605433996e-05,
      "loss": 0.4088,
      "mean_token_accuracy": 0.8638792872428894,
      "num_tokens": 769513628.0,
      "step": 22970
    },
    {
      "epoch": 1.3714780324737346,
      "grad_norm": 0.5605626106262207,
      "learning_rate": 1.6097762341511693e-05,
      "loss": 0.451,
      "mean_token_accuracy": 0.8501133322715759,
      "num_tokens": 769681308.0,
      "step": 22975
    },
    {
      "epoch": 1.3717765042979941,
      "grad_norm": 0.5340942144393921,
      "learning_rate": 1.6088190854814834e-05,
      "loss": 0.4172,
      "mean_token_accuracy": 0.8595431208610534,
      "num_tokens": 769848988.0,
      "step": 22980
    },
    {
      "epoch": 1.3720749761222542,
      "grad_norm": 0.4701322615146637,
      "learning_rate": 1.607862214767473e-05,
      "loss": 0.4242,
      "mean_token_accuracy": 0.8595848798751831,
      "num_tokens": 770016668.0,
      "step": 22985
    },
    {
      "epoch": 1.3723734479465137,
      "grad_norm": 0.507288932800293,
      "learning_rate": 1.6069056222422014e-05,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8525514006614685,
      "num_tokens": 770178269.0,
      "step": 22990
    },
    {
      "epoch": 1.3726719197707737,
      "grad_norm": 0.4795933961868286,
      "learning_rate": 1.605949308138665e-05,
      "loss": 0.4298,
      "mean_token_accuracy": 0.8552904725074768,
      "num_tokens": 770345949.0,
      "step": 22995
    },
    {
      "epoch": 1.3729703915950333,
      "grad_norm": 0.5413479208946228,
      "learning_rate": 1.6049932726897917e-05,
      "loss": 0.4281,
      "mean_token_accuracy": 0.857920789718628,
      "num_tokens": 770513629.0,
      "step": 23000
    },
    {
      "epoch": 1.3732688634192933,
      "grad_norm": 0.49817144870758057,
      "learning_rate": 1.6040375161284392e-05,
      "loss": 0.4392,
      "mean_token_accuracy": 0.8535249948501586,
      "num_tokens": 770681309.0,
      "step": 23005
    },
    {
      "epoch": 1.373567335243553,
      "grad_norm": 0.5285564661026001,
      "learning_rate": 1.603082038687404e-05,
      "loss": 0.4674,
      "mean_token_accuracy": 0.8443695664405823,
      "num_tokens": 770848989.0,
      "step": 23010
    },
    {
      "epoch": 1.373865807067813,
      "grad_norm": 0.5043045282363892,
      "learning_rate": 1.6021268405994074e-05,
      "loss": 0.4677,
      "mean_token_accuracy": 0.843898355960846,
      "num_tokens": 771016669.0,
      "step": 23015
    },
    {
      "epoch": 1.3741642788920725,
      "grad_norm": 0.531291127204895,
      "learning_rate": 1.6011719220971065e-05,
      "loss": 0.4464,
      "mean_token_accuracy": 0.8496659994125366,
      "num_tokens": 771184349.0,
      "step": 23020
    },
    {
      "epoch": 1.3744627507163325,
      "grad_norm": 0.5122756361961365,
      "learning_rate": 1.6002172834130868e-05,
      "loss": 0.4535,
      "mean_token_accuracy": 0.8468448042869567,
      "num_tokens": 771352029.0,
      "step": 23025
    },
    {
      "epoch": 1.374761222540592,
      "grad_norm": 0.515861988067627,
      "learning_rate": 1.599262924779872e-05,
      "loss": 0.4242,
      "mean_token_accuracy": 0.8582488417625427,
      "num_tokens": 771519709.0,
      "step": 23030
    },
    {
      "epoch": 1.375059694364852,
      "grad_norm": 0.48412561416625977,
      "learning_rate": 1.5983088464299125e-05,
      "loss": 0.3985,
      "mean_token_accuracy": 0.8648693799972534,
      "num_tokens": 771687389.0,
      "step": 23035
    },
    {
      "epoch": 1.3753581661891117,
      "grad_norm": 0.5374919772148132,
      "learning_rate": 1.5973550485955903e-05,
      "loss": 0.4561,
      "mean_token_accuracy": 0.8482285499572754,
      "num_tokens": 771855069.0,
      "step": 23040
    },
    {
      "epoch": 1.3756566380133715,
      "grad_norm": 0.5241804718971252,
      "learning_rate": 1.5964015315092214e-05,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8494810938835144,
      "num_tokens": 772022749.0,
      "step": 23045
    },
    {
      "epoch": 1.3759551098376313,
      "grad_norm": 0.5024101734161377,
      "learning_rate": 1.5954482954030524e-05,
      "loss": 0.4347,
      "mean_token_accuracy": 0.8542466998100281,
      "num_tokens": 772190429.0,
      "step": 23050
    },
    {
      "epoch": 1.376253581661891,
      "grad_norm": 0.5139269232749939,
      "learning_rate": 1.5944953405092628e-05,
      "loss": 0.4243,
      "mean_token_accuracy": 0.8594954013824463,
      "num_tokens": 772358109.0,
      "step": 23055
    },
    {
      "epoch": 1.3765520534861508,
      "grad_norm": 0.5611736178398132,
      "learning_rate": 1.59354266705996e-05,
      "loss": 0.4837,
      "mean_token_accuracy": 0.8412263035774231,
      "num_tokens": 772525789.0,
      "step": 23060
    },
    {
      "epoch": 1.3768505253104106,
      "grad_norm": 0.5159117579460144,
      "learning_rate": 1.5925902752871857e-05,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8505964398384094,
      "num_tokens": 772693469.0,
      "step": 23065
    },
    {
      "epoch": 1.3771489971346704,
      "grad_norm": 0.4799140393733978,
      "learning_rate": 1.591638165422915e-05,
      "loss": 0.4281,
      "mean_token_accuracy": 0.856453537940979,
      "num_tokens": 772861149.0,
      "step": 23070
    },
    {
      "epoch": 1.3774474689589302,
      "grad_norm": 0.4879801869392395,
      "learning_rate": 1.5906863376990493e-05,
      "loss": 0.4619,
      "mean_token_accuracy": 0.8494154930114746,
      "num_tokens": 773028829.0,
      "step": 23075
    },
    {
      "epoch": 1.37774594078319,
      "grad_norm": 0.5032755732536316,
      "learning_rate": 1.5897347923474255e-05,
      "loss": 0.442,
      "mean_token_accuracy": 0.8528172254562378,
      "num_tokens": 773189390.0,
      "step": 23080
    },
    {
      "epoch": 1.3780444126074498,
      "grad_norm": 0.5552733540534973,
      "learning_rate": 1.588783529599807e-05,
      "loss": 0.4472,
      "mean_token_accuracy": 0.8517237305641174,
      "num_tokens": 773357070.0,
      "step": 23085
    },
    {
      "epoch": 1.3783428844317096,
      "grad_norm": 0.5376341342926025,
      "learning_rate": 1.587832549687894e-05,
      "loss": 0.4707,
      "mean_token_accuracy": 0.8450368404388428,
      "num_tokens": 773516785.0,
      "step": 23090
    },
    {
      "epoch": 1.3786413562559694,
      "grad_norm": 0.5441288948059082,
      "learning_rate": 1.586881852843316e-05,
      "loss": 0.4409,
      "mean_token_accuracy": 0.8532983541488648,
      "num_tokens": 773684465.0,
      "step": 23095
    },
    {
      "epoch": 1.3789398280802292,
      "grad_norm": 0.5238974094390869,
      "learning_rate": 1.5859314392976305e-05,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8497284531593323,
      "num_tokens": 773852062.0,
      "step": 23100
    },
    {
      "epoch": 1.379238299904489,
      "grad_norm": 0.44513800740242004,
      "learning_rate": 1.584981309282329e-05,
      "loss": 0.4177,
      "mean_token_accuracy": 0.8594119071960449,
      "num_tokens": 774019742.0,
      "step": 23105
    },
    {
      "epoch": 1.3795367717287488,
      "grad_norm": 0.545132040977478,
      "learning_rate": 1.584031463028834e-05,
      "loss": 0.4639,
      "mean_token_accuracy": 0.8463318586349488,
      "num_tokens": 774187422.0,
      "step": 23110
    },
    {
      "epoch": 1.3798352435530086,
      "grad_norm": 0.44364869594573975,
      "learning_rate": 1.583081900768497e-05,
      "loss": 0.4606,
      "mean_token_accuracy": 0.8487355351448059,
      "num_tokens": 774355102.0,
      "step": 23115
    },
    {
      "epoch": 1.3801337153772684,
      "grad_norm": 0.5100110769271851,
      "learning_rate": 1.5821326227326026e-05,
      "loss": 0.4262,
      "mean_token_accuracy": 0.8563760042190551,
      "num_tokens": 774522782.0,
      "step": 23120
    },
    {
      "epoch": 1.3804321872015282,
      "grad_norm": 0.4937985837459564,
      "learning_rate": 1.5811836291523636e-05,
      "loss": 0.432,
      "mean_token_accuracy": 0.8567935109138489,
      "num_tokens": 774690462.0,
      "step": 23125
    },
    {
      "epoch": 1.380730659025788,
      "grad_norm": 0.5006693005561829,
      "learning_rate": 1.5802349202589258e-05,
      "loss": 0.4284,
      "mean_token_accuracy": 0.856715989112854,
      "num_tokens": 774858142.0,
      "step": 23130
    },
    {
      "epoch": 1.3810291308500477,
      "grad_norm": 0.5181460976600647,
      "learning_rate": 1.579286496283364e-05,
      "loss": 0.4271,
      "mean_token_accuracy": 0.8563044190406799,
      "num_tokens": 775025822.0,
      "step": 23135
    },
    {
      "epoch": 1.3813276026743075,
      "grad_norm": 0.5043380260467529,
      "learning_rate": 1.578338357456686e-05,
      "loss": 0.4076,
      "mean_token_accuracy": 0.8640761017799378,
      "num_tokens": 775193502.0,
      "step": 23140
    },
    {
      "epoch": 1.3816260744985673,
      "grad_norm": 0.5079756379127502,
      "learning_rate": 1.5773905040098274e-05,
      "loss": 0.4093,
      "mean_token_accuracy": 0.8617261052131653,
      "num_tokens": 775361182.0,
      "step": 23145
    },
    {
      "epoch": 1.3819245463228271,
      "grad_norm": 0.4899621605873108,
      "learning_rate": 1.5764429361736537e-05,
      "loss": 0.392,
      "mean_token_accuracy": 0.867255175113678,
      "num_tokens": 775528862.0,
      "step": 23150
    },
    {
      "epoch": 1.382223018147087,
      "grad_norm": 0.523612916469574,
      "learning_rate": 1.5754956541789667e-05,
      "loss": 0.4167,
      "mean_token_accuracy": 0.8596325874328613,
      "num_tokens": 775696542.0,
      "step": 23155
    },
    {
      "epoch": 1.3825214899713467,
      "grad_norm": 0.45955783128738403,
      "learning_rate": 1.5745486582564906e-05,
      "loss": 0.4158,
      "mean_token_accuracy": 0.860676383972168,
      "num_tokens": 775864222.0,
      "step": 23160
    },
    {
      "epoch": 1.3828199617956065,
      "grad_norm": 0.5167980194091797,
      "learning_rate": 1.573601948636885e-05,
      "loss": 0.4078,
      "mean_token_accuracy": 0.8635273814201355,
      "num_tokens": 776031902.0,
      "step": 23165
    },
    {
      "epoch": 1.3831184336198663,
      "grad_norm": 0.46761518716812134,
      "learning_rate": 1.572655525550739e-05,
      "loss": 0.4237,
      "mean_token_accuracy": 0.8580341100692749,
      "num_tokens": 776199582.0,
      "step": 23170
    },
    {
      "epoch": 1.383416905444126,
      "grad_norm": 0.5134435892105103,
      "learning_rate": 1.5717093892285713e-05,
      "loss": 0.462,
      "mean_token_accuracy": 0.8456220865249634,
      "num_tokens": 776367262.0,
      "step": 23175
    },
    {
      "epoch": 1.383715377268386,
      "grad_norm": 0.47994717955589294,
      "learning_rate": 1.5707635399008308e-05,
      "loss": 0.3875,
      "mean_token_accuracy": 0.8687761068344116,
      "num_tokens": 776534942.0,
      "step": 23180
    },
    {
      "epoch": 1.3840138490926457,
      "grad_norm": 0.50049889087677,
      "learning_rate": 1.5698179777978956e-05,
      "loss": 0.4275,
      "mean_token_accuracy": 0.8575629353523254,
      "num_tokens": 776702622.0,
      "step": 23185
    },
    {
      "epoch": 1.3843123209169055,
      "grad_norm": 0.5402145385742188,
      "learning_rate": 1.568872703150076e-05,
      "loss": 0.4283,
      "mean_token_accuracy": 0.855713951587677,
      "num_tokens": 776870302.0,
      "step": 23190
    },
    {
      "epoch": 1.3846107927411653,
      "grad_norm": 0.5122275352478027,
      "learning_rate": 1.5679277161876105e-05,
      "loss": 0.4302,
      "mean_token_accuracy": 0.8575152039527894,
      "num_tokens": 777037982.0,
      "step": 23195
    },
    {
      "epoch": 1.384909264565425,
      "grad_norm": 0.502173662185669,
      "learning_rate": 1.566983017140668e-05,
      "loss": 0.4257,
      "mean_token_accuracy": 0.8584754824638366,
      "num_tokens": 777205662.0,
      "step": 23200
    },
    {
      "epoch": 1.3852077363896849,
      "grad_norm": 0.492482990026474,
      "learning_rate": 1.5660386062393485e-05,
      "loss": 0.4323,
      "mean_token_accuracy": 0.8531730771064758,
      "num_tokens": 777373342.0,
      "step": 23205
    },
    {
      "epoch": 1.3855062082139447,
      "grad_norm": 0.4919479489326477,
      "learning_rate": 1.5650944837136787e-05,
      "loss": 0.445,
      "mean_token_accuracy": 0.8522784233093261,
      "num_tokens": 777541022.0,
      "step": 23210
    },
    {
      "epoch": 1.3858046800382045,
      "grad_norm": 0.48758816719055176,
      "learning_rate": 1.5641506497936194e-05,
      "loss": 0.4312,
      "mean_token_accuracy": 0.8559107780456543,
      "num_tokens": 777708702.0,
      "step": 23215
    },
    {
      "epoch": 1.3861031518624642,
      "grad_norm": 0.5445257425308228,
      "learning_rate": 1.5632071047090573e-05,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8381703734397888,
      "num_tokens": 777870190.0,
      "step": 23220
    },
    {
      "epoch": 1.386401623686724,
      "grad_norm": 0.5068190097808838,
      "learning_rate": 1.56226384868981e-05,
      "loss": 0.4363,
      "mean_token_accuracy": 0.8535488486289978,
      "num_tokens": 778037870.0,
      "step": 23225
    },
    {
      "epoch": 1.3867000955109838,
      "grad_norm": 0.47730064392089844,
      "learning_rate": 1.5613208819656256e-05,
      "loss": 0.4656,
      "mean_token_accuracy": 0.8452463388442993,
      "num_tokens": 778205550.0,
      "step": 23230
    },
    {
      "epoch": 1.3869985673352436,
      "grad_norm": 0.5039456486701965,
      "learning_rate": 1.560378204766181e-05,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8557914853096008,
      "num_tokens": 778373230.0,
      "step": 23235
    },
    {
      "epoch": 1.3872970391595034,
      "grad_norm": 0.5148113369941711,
      "learning_rate": 1.5594358173210834e-05,
      "loss": 0.4448,
      "mean_token_accuracy": 0.85208158493042,
      "num_tokens": 778540910.0,
      "step": 23240
    },
    {
      "epoch": 1.3875955109837632,
      "grad_norm": 0.5091186761856079,
      "learning_rate": 1.558493719859867e-05,
      "loss": 0.4097,
      "mean_token_accuracy": 0.8638256072998047,
      "num_tokens": 778708590.0,
      "step": 23245
    },
    {
      "epoch": 1.387893982808023,
      "grad_norm": 0.49849849939346313,
      "learning_rate": 1.5575519126119976e-05,
      "loss": 0.4437,
      "mean_token_accuracy": 0.8517535328865051,
      "num_tokens": 778876270.0,
      "step": 23250
    },
    {
      "epoch": 1.3881924546322826,
      "grad_norm": 0.5224447250366211,
      "learning_rate": 1.5566103958068702e-05,
      "loss": 0.4487,
      "mean_token_accuracy": 0.8500298142433167,
      "num_tokens": 779043950.0,
      "step": 23255
    },
    {
      "epoch": 1.3884909264565426,
      "grad_norm": 0.5036640763282776,
      "learning_rate": 1.555669169673808e-05,
      "loss": 0.4273,
      "mean_token_accuracy": 0.8570022702217102,
      "num_tokens": 779211630.0,
      "step": 23260
    },
    {
      "epoch": 1.3887893982808022,
      "grad_norm": 0.46421748399734497,
      "learning_rate": 1.5547282344420656e-05,
      "loss": 0.4353,
      "mean_token_accuracy": 0.8537695288658143,
      "num_tokens": 779379310.0,
      "step": 23265
    },
    {
      "epoch": 1.3890878701050622,
      "grad_norm": 0.5558269023895264,
      "learning_rate": 1.553787590340823e-05,
      "loss": 0.4693,
      "mean_token_accuracy": 0.8448586463928223,
      "num_tokens": 779546990.0,
      "step": 23270
    },
    {
      "epoch": 1.3893863419293218,
      "grad_norm": 0.4932655394077301,
      "learning_rate": 1.552847237599192e-05,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8518549561500549,
      "num_tokens": 779714670.0,
      "step": 23275
    },
    {
      "epoch": 1.3896848137535818,
      "grad_norm": 0.5218916535377502,
      "learning_rate": 1.5519071764462135e-05,
      "loss": 0.4493,
      "mean_token_accuracy": 0.8492663741111756,
      "num_tokens": 779882350.0,
      "step": 23280
    },
    {
      "epoch": 1.3899832855778413,
      "grad_norm": 0.5360414385795593,
      "learning_rate": 1.550967407110856e-05,
      "loss": 0.4125,
      "mean_token_accuracy": 0.8623464107513428,
      "num_tokens": 780050030.0,
      "step": 23285
    },
    {
      "epoch": 1.3902817574021014,
      "grad_norm": 0.4707147181034088,
      "learning_rate": 1.5500279298220186e-05,
      "loss": 0.4552,
      "mean_token_accuracy": 0.8484313607215881,
      "num_tokens": 780217710.0,
      "step": 23290
    },
    {
      "epoch": 1.390580229226361,
      "grad_norm": 0.4926894009113312,
      "learning_rate": 1.5490887448085273e-05,
      "loss": 0.4322,
      "mean_token_accuracy": 0.8560718059539795,
      "num_tokens": 780385390.0,
      "step": 23295
    },
    {
      "epoch": 1.390878701050621,
      "grad_norm": 0.5696619153022766,
      "learning_rate": 1.54814985229914e-05,
      "loss": 0.4376,
      "mean_token_accuracy": 0.8527303457260131,
      "num_tokens": 780550807.0,
      "step": 23300
    },
    {
      "epoch": 1.3911771728748805,
      "grad_norm": 0.49192309379577637,
      "learning_rate": 1.5472112525225385e-05,
      "loss": 0.4185,
      "mean_token_accuracy": 0.8599188804626465,
      "num_tokens": 780718487.0,
      "step": 23305
    },
    {
      "epoch": 1.3914756446991405,
      "grad_norm": 0.4632330536842346,
      "learning_rate": 1.5462729457073374e-05,
      "loss": 0.404,
      "mean_token_accuracy": 0.8642908215522767,
      "num_tokens": 780886167.0,
      "step": 23310
    },
    {
      "epoch": 1.3917741165234,
      "grad_norm": 0.4984997510910034,
      "learning_rate": 1.5453349320820787e-05,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8462066054344177,
      "num_tokens": 781053847.0,
      "step": 23315
    },
    {
      "epoch": 1.39207258834766,
      "grad_norm": 0.5803594589233398,
      "learning_rate": 1.544397211875233e-05,
      "loss": 0.434,
      "mean_token_accuracy": 0.8580520153045654,
      "num_tokens": 781221527.0,
      "step": 23320
    },
    {
      "epoch": 1.3923710601719197,
      "grad_norm": 0.5138877034187317,
      "learning_rate": 1.5434597853152e-05,
      "loss": 0.4347,
      "mean_token_accuracy": 0.8547119140625,
      "num_tokens": 781389207.0,
      "step": 23325
    },
    {
      "epoch": 1.3926695319961795,
      "grad_norm": 0.47389793395996094,
      "learning_rate": 1.5425226526303055e-05,
      "loss": 0.4426,
      "mean_token_accuracy": 0.8544196724891663,
      "num_tokens": 781556887.0,
      "step": 23330
    },
    {
      "epoch": 1.3929680038204393,
      "grad_norm": 0.5190534591674805,
      "learning_rate": 1.5415858140488056e-05,
      "loss": 0.4936,
      "mean_token_accuracy": 0.8371704578399658,
      "num_tokens": 781724567.0,
      "step": 23335
    },
    {
      "epoch": 1.393266475644699,
      "grad_norm": 0.4836447536945343,
      "learning_rate": 1.540649269798887e-05,
      "loss": 0.439,
      "mean_token_accuracy": 0.8533937692642212,
      "num_tokens": 781892247.0,
      "step": 23340
    },
    {
      "epoch": 1.3935649474689589,
      "grad_norm": 0.5116270780563354,
      "learning_rate": 1.53971302010866e-05,
      "loss": 0.4405,
      "mean_token_accuracy": 0.8543540477752686,
      "num_tokens": 782059927.0,
      "step": 23345
    },
    {
      "epoch": 1.3938634192932187,
      "grad_norm": 0.5673056840896606,
      "learning_rate": 1.5387770652061677e-05,
      "loss": 0.4507,
      "mean_token_accuracy": 0.8505785584449768,
      "num_tokens": 782227607.0,
      "step": 23350
    },
    {
      "epoch": 1.3941618911174785,
      "grad_norm": 0.5157475471496582,
      "learning_rate": 1.5378414053193757e-05,
      "loss": 0.4351,
      "mean_token_accuracy": 0.8551354050636292,
      "num_tokens": 782395287.0,
      "step": 23355
    },
    {
      "epoch": 1.3944603629417383,
      "grad_norm": 0.5795453190803528,
      "learning_rate": 1.536906040676184e-05,
      "loss": 0.4537,
      "mean_token_accuracy": 0.8510717868804931,
      "num_tokens": 782557019.0,
      "step": 23360
    },
    {
      "epoch": 1.394758834765998,
      "grad_norm": 0.6179571151733398,
      "learning_rate": 1.5359709715044183e-05,
      "loss": 0.4141,
      "mean_token_accuracy": 0.8607181191444397,
      "num_tokens": 782724699.0,
      "step": 23365
    },
    {
      "epoch": 1.3950573065902578,
      "grad_norm": 0.4667505621910095,
      "learning_rate": 1.53503619803183e-05,
      "loss": 0.4475,
      "mean_token_accuracy": 0.8524215698242188,
      "num_tokens": 782892379.0,
      "step": 23370
    },
    {
      "epoch": 1.3953557784145176,
      "grad_norm": 0.6107589602470398,
      "learning_rate": 1.5341017204861014e-05,
      "loss": 0.4788,
      "mean_token_accuracy": 0.841703450679779,
      "num_tokens": 783060059.0,
      "step": 23375
    },
    {
      "epoch": 1.3956542502387774,
      "grad_norm": 0.4700370728969574,
      "learning_rate": 1.5331675390948415e-05,
      "loss": 0.4038,
      "mean_token_accuracy": 0.8635751008987427,
      "num_tokens": 783227739.0,
      "step": 23380
    },
    {
      "epoch": 1.3959527220630372,
      "grad_norm": 0.5494814515113831,
      "learning_rate": 1.5322336540855886e-05,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8456399798393249,
      "num_tokens": 783395419.0,
      "step": 23385
    },
    {
      "epoch": 1.396251193887297,
      "grad_norm": 0.5478232502937317,
      "learning_rate": 1.5313000656858056e-05,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8424668908119202,
      "num_tokens": 783563099.0,
      "step": 23390
    },
    {
      "epoch": 1.3965496657115568,
      "grad_norm": 0.5142637491226196,
      "learning_rate": 1.5303667741228866e-05,
      "loss": 0.4536,
      "mean_token_accuracy": 0.8494333744049072,
      "num_tokens": 783730779.0,
      "step": 23395
    },
    {
      "epoch": 1.3968481375358166,
      "grad_norm": 0.5252295732498169,
      "learning_rate": 1.5294337796241513e-05,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8457652449607849,
      "num_tokens": 783898459.0,
      "step": 23400
    },
    {
      "epoch": 1.3971466093600764,
      "grad_norm": 0.5872812271118164,
      "learning_rate": 1.528501082416848e-05,
      "loss": 0.4876,
      "mean_token_accuracy": 0.8384528040885926,
      "num_tokens": 784066139.0,
      "step": 23405
    },
    {
      "epoch": 1.3974450811843362,
      "grad_norm": 0.5298374891281128,
      "learning_rate": 1.5275686827281537e-05,
      "loss": 0.4451,
      "mean_token_accuracy": 0.8518012642860413,
      "num_tokens": 784233819.0,
      "step": 23410
    },
    {
      "epoch": 1.397743553008596,
      "grad_norm": 0.4962978959083557,
      "learning_rate": 1.5266365807851677e-05,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8532446622848511,
      "num_tokens": 784401499.0,
      "step": 23415
    },
    {
      "epoch": 1.3980420248328558,
      "grad_norm": 0.501723051071167,
      "learning_rate": 1.5257047768149247e-05,
      "loss": 0.4203,
      "mean_token_accuracy": 0.8596504926681519,
      "num_tokens": 784569179.0,
      "step": 23420
    },
    {
      "epoch": 1.3983404966571156,
      "grad_norm": 0.5250449776649475,
      "learning_rate": 1.5247732710443818e-05,
      "loss": 0.4297,
      "mean_token_accuracy": 0.8572885632514954,
      "num_tokens": 784736859.0,
      "step": 23425
    },
    {
      "epoch": 1.3986389684813754,
      "grad_norm": 0.5277840495109558,
      "learning_rate": 1.523842063700423e-05,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8527496099472046,
      "num_tokens": 784904539.0,
      "step": 23430
    },
    {
      "epoch": 1.3989374403056352,
      "grad_norm": 0.537818431854248,
      "learning_rate": 1.5229111550098619e-05,
      "loss": 0.4736,
      "mean_token_accuracy": 0.8428009033203125,
      "num_tokens": 785072219.0,
      "step": 23435
    },
    {
      "epoch": 1.399235912129895,
      "grad_norm": 0.5120958685874939,
      "learning_rate": 1.5219805451994385e-05,
      "loss": 0.4111,
      "mean_token_accuracy": 0.862406063079834,
      "num_tokens": 785239899.0,
      "step": 23440
    },
    {
      "epoch": 1.3995343839541547,
      "grad_norm": 0.5241063237190247,
      "learning_rate": 1.52105023449582e-05,
      "loss": 0.4359,
      "mean_token_accuracy": 0.8544435024261474,
      "num_tokens": 785407579.0,
      "step": 23445
    },
    {
      "epoch": 1.3998328557784145,
      "grad_norm": 0.5484216809272766,
      "learning_rate": 1.5201202231256012e-05,
      "loss": 0.4457,
      "mean_token_accuracy": 0.8524036765098572,
      "num_tokens": 785575259.0,
      "step": 23450
    },
    {
      "epoch": 1.4001313276026743,
      "grad_norm": 0.4641386866569519,
      "learning_rate": 1.5191905113153027e-05,
      "loss": 0.453,
      "mean_token_accuracy": 0.8485923886299134,
      "num_tokens": 785742939.0,
      "step": 23455
    },
    {
      "epoch": 1.4004297994269341,
      "grad_norm": 0.4222840964794159,
      "learning_rate": 1.5182610992913727e-05,
      "loss": 0.4243,
      "mean_token_accuracy": 0.8578671097755433,
      "num_tokens": 785910619.0,
      "step": 23460
    },
    {
      "epoch": 1.400728271251194,
      "grad_norm": 0.45363545417785645,
      "learning_rate": 1.5173319872801878e-05,
      "loss": 0.459,
      "mean_token_accuracy": 0.8480078697204589,
      "num_tokens": 786078299.0,
      "step": 23465
    },
    {
      "epoch": 1.4010267430754537,
      "grad_norm": 0.49216949939727783,
      "learning_rate": 1.5164031755080496e-05,
      "loss": 0.4316,
      "mean_token_accuracy": 0.8564237117767334,
      "num_tokens": 786245979.0,
      "step": 23470
    },
    {
      "epoch": 1.4013252148997135,
      "grad_norm": 0.5190386772155762,
      "learning_rate": 1.5154746642011885e-05,
      "loss": 0.4278,
      "mean_token_accuracy": 0.8559704184532165,
      "num_tokens": 786413659.0,
      "step": 23475
    },
    {
      "epoch": 1.4016236867239733,
      "grad_norm": 0.5304282903671265,
      "learning_rate": 1.5145464535857578e-05,
      "loss": 0.4528,
      "mean_token_accuracy": 0.8485566139221191,
      "num_tokens": 786581339.0,
      "step": 23480
    },
    {
      "epoch": 1.401922158548233,
      "grad_norm": 0.5135473608970642,
      "learning_rate": 1.5136185438878443e-05,
      "loss": 0.4339,
      "mean_token_accuracy": 0.8545031547546387,
      "num_tokens": 786749019.0,
      "step": 23485
    },
    {
      "epoch": 1.402220630372493,
      "grad_norm": 0.5437111258506775,
      "learning_rate": 1.5126909353334543e-05,
      "loss": 0.4232,
      "mean_token_accuracy": 0.8568471908569336,
      "num_tokens": 786916699.0,
      "step": 23490
    },
    {
      "epoch": 1.4025191021967527,
      "grad_norm": 0.4763391315937042,
      "learning_rate": 1.5117636281485251e-05,
      "loss": 0.4258,
      "mean_token_accuracy": 0.8564177393913269,
      "num_tokens": 787084379.0,
      "step": 23495
    },
    {
      "epoch": 1.4028175740210125,
      "grad_norm": 0.5312849879264832,
      "learning_rate": 1.5108366225589193e-05,
      "loss": 0.4682,
      "mean_token_accuracy": 0.8442920088768006,
      "num_tokens": 787252059.0,
      "step": 23500
    },
    {
      "epoch": 1.4031160458452723,
      "grad_norm": 0.5117261409759521,
      "learning_rate": 1.5099099187904267e-05,
      "loss": 0.4543,
      "mean_token_accuracy": 0.8490874290466308,
      "num_tokens": 787419739.0,
      "step": 23505
    },
    {
      "epoch": 1.403414517669532,
      "grad_norm": 0.5443363785743713,
      "learning_rate": 1.5089835170687633e-05,
      "loss": 0.4474,
      "mean_token_accuracy": 0.851604425907135,
      "num_tokens": 787587419.0,
      "step": 23510
    },
    {
      "epoch": 1.4037129894937919,
      "grad_norm": 0.5582711696624756,
      "learning_rate": 1.5080574176195706e-05,
      "loss": 0.4421,
      "mean_token_accuracy": 0.8532864093780518,
      "num_tokens": 787755099.0,
      "step": 23515
    },
    {
      "epoch": 1.4040114613180517,
      "grad_norm": 0.5083257555961609,
      "learning_rate": 1.5071316206684172e-05,
      "loss": 0.4379,
      "mean_token_accuracy": 0.8533639430999755,
      "num_tokens": 787922779.0,
      "step": 23520
    },
    {
      "epoch": 1.4043099331423115,
      "grad_norm": 0.6606442928314209,
      "learning_rate": 1.5062061264407981e-05,
      "loss": 0.4681,
      "mean_token_accuracy": 0.8441488742828369,
      "num_tokens": 788090459.0,
      "step": 23525
    },
    {
      "epoch": 1.404608404966571,
      "grad_norm": 0.46060633659362793,
      "learning_rate": 1.505280935162135e-05,
      "loss": 0.3909,
      "mean_token_accuracy": 0.8679052948951721,
      "num_tokens": 788258139.0,
      "step": 23530
    },
    {
      "epoch": 1.404906876790831,
      "grad_norm": 0.48085835576057434,
      "learning_rate": 1.5043560470577756e-05,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8374567747116088,
      "num_tokens": 788425819.0,
      "step": 23535
    },
    {
      "epoch": 1.4052053486150906,
      "grad_norm": 0.4876287877559662,
      "learning_rate": 1.5034314623529915e-05,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8546999931335449,
      "num_tokens": 788593499.0,
      "step": 23540
    },
    {
      "epoch": 1.4055038204393506,
      "grad_norm": 0.5198888182640076,
      "learning_rate": 1.5025071812729851e-05,
      "loss": 0.4528,
      "mean_token_accuracy": 0.849361801147461,
      "num_tokens": 788761179.0,
      "step": 23545
    },
    {
      "epoch": 1.4058022922636102,
      "grad_norm": 0.48025810718536377,
      "learning_rate": 1.5015832040428802e-05,
      "loss": 0.3951,
      "mean_token_accuracy": 0.8678508639335633,
      "num_tokens": 788924345.0,
      "step": 23550
    },
    {
      "epoch": 1.4061007640878702,
      "grad_norm": 0.5333512425422668,
      "learning_rate": 1.5006595308877292e-05,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8393355727195739,
      "num_tokens": 789092025.0,
      "step": 23555
    },
    {
      "epoch": 1.4063992359121298,
      "grad_norm": 0.5317134857177734,
      "learning_rate": 1.4997361620325092e-05,
      "loss": 0.4753,
      "mean_token_accuracy": 0.8442848682403564,
      "num_tokens": 789252200.0,
      "step": 23560
    },
    {
      "epoch": 1.4066977077363898,
      "grad_norm": 0.4926573932170868,
      "learning_rate": 1.4988130977021248e-05,
      "loss": 0.4196,
      "mean_token_accuracy": 0.8609658360481263,
      "num_tokens": 789417168.0,
      "step": 23565
    },
    {
      "epoch": 1.4069961795606494,
      "grad_norm": 0.5432858467102051,
      "learning_rate": 1.4978903381214052e-05,
      "loss": 0.47,
      "mean_token_accuracy": 0.8432363271713257,
      "num_tokens": 789584848.0,
      "step": 23570
    },
    {
      "epoch": 1.4072946513849094,
      "grad_norm": 0.5307675004005432,
      "learning_rate": 1.4969678835151038e-05,
      "loss": 0.4732,
      "mean_token_accuracy": 0.8439281940460205,
      "num_tokens": 789752528.0,
      "step": 23575
    },
    {
      "epoch": 1.407593123209169,
      "grad_norm": 0.5629969239234924,
      "learning_rate": 1.4960457341079032e-05,
      "loss": 0.4906,
      "mean_token_accuracy": 0.8383394956588746,
      "num_tokens": 789920208.0,
      "step": 23580
    },
    {
      "epoch": 1.407891595033429,
      "grad_norm": 0.5573857426643372,
      "learning_rate": 1.4951238901244093e-05,
      "loss": 0.4419,
      "mean_token_accuracy": 0.8512644648551941,
      "num_tokens": 790087888.0,
      "step": 23585
    },
    {
      "epoch": 1.4081900668576885,
      "grad_norm": 0.5876045227050781,
      "learning_rate": 1.4942023517891546e-05,
      "loss": 0.4536,
      "mean_token_accuracy": 0.8486162543296814,
      "num_tokens": 790255568.0,
      "step": 23590
    },
    {
      "epoch": 1.4084885386819483,
      "grad_norm": 0.5804503560066223,
      "learning_rate": 1.4932811193265966e-05,
      "loss": 0.4401,
      "mean_token_accuracy": 0.8550161004066468,
      "num_tokens": 790423248.0,
      "step": 23595
    },
    {
      "epoch": 1.4087870105062081,
      "grad_norm": 0.5259001851081848,
      "learning_rate": 1.492360192961118e-05,
      "loss": 0.4566,
      "mean_token_accuracy": 0.8480496406555176,
      "num_tokens": 790590928.0,
      "step": 23600
    },
    {
      "epoch": 1.409085482330468,
      "grad_norm": 0.5038285851478577,
      "learning_rate": 1.491439572917028e-05,
      "loss": 0.4463,
      "mean_token_accuracy": 0.8517774105072021,
      "num_tokens": 790758608.0,
      "step": 23605
    },
    {
      "epoch": 1.4093839541547277,
      "grad_norm": 0.5120645761489868,
      "learning_rate": 1.4905192594185602e-05,
      "loss": 0.4396,
      "mean_token_accuracy": 0.8551830649375916,
      "num_tokens": 790920496.0,
      "step": 23610
    },
    {
      "epoch": 1.4096824259789875,
      "grad_norm": 0.5468581914901733,
      "learning_rate": 1.489599252689874e-05,
      "loss": 0.4888,
      "mean_token_accuracy": 0.8406358003616333,
      "num_tokens": 791088176.0,
      "step": 23615
    },
    {
      "epoch": 1.4099808978032473,
      "grad_norm": 0.47131621837615967,
      "learning_rate": 1.488679552955054e-05,
      "loss": 0.4192,
      "mean_token_accuracy": 0.8598771333694458,
      "num_tokens": 791255856.0,
      "step": 23620
    },
    {
      "epoch": 1.410279369627507,
      "grad_norm": 0.49126923084259033,
      "learning_rate": 1.4877601604381102e-05,
      "loss": 0.4438,
      "mean_token_accuracy": 0.8514076113700867,
      "num_tokens": 791423536.0,
      "step": 23625
    },
    {
      "epoch": 1.410577841451767,
      "grad_norm": 0.49482589960098267,
      "learning_rate": 1.4868410753629786e-05,
      "loss": 0.4109,
      "mean_token_accuracy": 0.8625730752944947,
      "num_tokens": 791591216.0,
      "step": 23630
    },
    {
      "epoch": 1.4108763132760267,
      "grad_norm": 0.5141450762748718,
      "learning_rate": 1.4859222979535175e-05,
      "loss": 0.4162,
      "mean_token_accuracy": 0.8585052967071534,
      "num_tokens": 791758896.0,
      "step": 23635
    },
    {
      "epoch": 1.4111747851002865,
      "grad_norm": 0.5500440001487732,
      "learning_rate": 1.4850038284335132e-05,
      "loss": 0.461,
      "mean_token_accuracy": 0.8475366830825806,
      "num_tokens": 791926576.0,
      "step": 23640
    },
    {
      "epoch": 1.4114732569245463,
      "grad_norm": 0.5590739846229553,
      "learning_rate": 1.4840856670266751e-05,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8436061024665833,
      "num_tokens": 792094256.0,
      "step": 23645
    },
    {
      "epoch": 1.411771728748806,
      "grad_norm": 0.4894876480102539,
      "learning_rate": 1.4831678139566393e-05,
      "loss": 0.4106,
      "mean_token_accuracy": 0.8630752325057983,
      "num_tokens": 792254732.0,
      "step": 23650
    },
    {
      "epoch": 1.4120702005730659,
      "grad_norm": 0.49888819456100464,
      "learning_rate": 1.4822502694469662e-05,
      "loss": 0.4403,
      "mean_token_accuracy": 0.8525647163391114,
      "num_tokens": 792422412.0,
      "step": 23655
    },
    {
      "epoch": 1.4123686723973257,
      "grad_norm": 0.5480368733406067,
      "learning_rate": 1.481333033721139e-05,
      "loss": 0.4754,
      "mean_token_accuracy": 0.8414350390434265,
      "num_tokens": 792590092.0,
      "step": 23660
    },
    {
      "epoch": 1.4126671442215855,
      "grad_norm": 0.5386296510696411,
      "learning_rate": 1.4804161070025674e-05,
      "loss": 0.4528,
      "mean_token_accuracy": 0.8493796944618225,
      "num_tokens": 792757772.0,
      "step": 23665
    },
    {
      "epoch": 1.4129656160458453,
      "grad_norm": 0.5429210662841797,
      "learning_rate": 1.4794994895145886e-05,
      "loss": 0.4352,
      "mean_token_accuracy": 0.8546999931335449,
      "num_tokens": 792925452.0,
      "step": 23670
    },
    {
      "epoch": 1.413264087870105,
      "grad_norm": 0.49123069643974304,
      "learning_rate": 1.478583181480459e-05,
      "loss": 0.4359,
      "mean_token_accuracy": 0.8555648326873779,
      "num_tokens": 793093132.0,
      "step": 23675
    },
    {
      "epoch": 1.4135625596943648,
      "grad_norm": 0.5216329097747803,
      "learning_rate": 1.4776671831233646e-05,
      "loss": 0.4558,
      "mean_token_accuracy": 0.8480555772781372,
      "num_tokens": 793260812.0,
      "step": 23680
    },
    {
      "epoch": 1.4138610315186246,
      "grad_norm": 0.5336837768554688,
      "learning_rate": 1.4767514946664102e-05,
      "loss": 0.4297,
      "mean_token_accuracy": 0.8548789143562316,
      "num_tokens": 793428492.0,
      "step": 23685
    },
    {
      "epoch": 1.4141595033428844,
      "grad_norm": 0.464883953332901,
      "learning_rate": 1.4758361163326315e-05,
      "loss": 0.4197,
      "mean_token_accuracy": 0.859036135673523,
      "num_tokens": 793596172.0,
      "step": 23690
    },
    {
      "epoch": 1.4144579751671442,
      "grad_norm": 0.470636785030365,
      "learning_rate": 1.4749210483449857e-05,
      "loss": 0.4199,
      "mean_token_accuracy": 0.8599009871482849,
      "num_tokens": 793763852.0,
      "step": 23695
    },
    {
      "epoch": 1.414756446991404,
      "grad_norm": 0.49245259165763855,
      "learning_rate": 1.4740062909263535e-05,
      "loss": 0.4092,
      "mean_token_accuracy": 0.8628534078598022,
      "num_tokens": 793931532.0,
      "step": 23700
    },
    {
      "epoch": 1.4150549188156638,
      "grad_norm": 0.4722106456756592,
      "learning_rate": 1.4730918442995412e-05,
      "loss": 0.3979,
      "mean_token_accuracy": 0.866378390789032,
      "num_tokens": 794099212.0,
      "step": 23705
    },
    {
      "epoch": 1.4153533906399236,
      "grad_norm": 0.566887617111206,
      "learning_rate": 1.472177708687279e-05,
      "loss": 0.4541,
      "mean_token_accuracy": 0.8495586276054382,
      "num_tokens": 794266892.0,
      "step": 23710
    },
    {
      "epoch": 1.4156518624641834,
      "grad_norm": 0.4816938638687134,
      "learning_rate": 1.4712638843122223e-05,
      "loss": 0.4414,
      "mean_token_accuracy": 0.8511391997337341,
      "num_tokens": 794434572.0,
      "step": 23715
    },
    {
      "epoch": 1.4159503342884432,
      "grad_norm": 0.44670966267585754,
      "learning_rate": 1.4703503713969488e-05,
      "loss": 0.4172,
      "mean_token_accuracy": 0.8612549185752869,
      "num_tokens": 794602252.0,
      "step": 23720
    },
    {
      "epoch": 1.416248806112703,
      "grad_norm": 0.5175304412841797,
      "learning_rate": 1.4694371701639623e-05,
      "loss": 0.426,
      "mean_token_accuracy": 0.8579744815826416,
      "num_tokens": 794769932.0,
      "step": 23725
    },
    {
      "epoch": 1.4165472779369628,
      "grad_norm": 0.48323628306388855,
      "learning_rate": 1.4685242808356892e-05,
      "loss": 0.4631,
      "mean_token_accuracy": 0.8474412322044372,
      "num_tokens": 794937612.0,
      "step": 23730
    },
    {
      "epoch": 1.4168457497612226,
      "grad_norm": 0.49873974919319153,
      "learning_rate": 1.4676117036344805e-05,
      "loss": 0.4103,
      "mean_token_accuracy": 0.8622569441795349,
      "num_tokens": 795105292.0,
      "step": 23735
    },
    {
      "epoch": 1.4171442215854824,
      "grad_norm": 0.5509675145149231,
      "learning_rate": 1.4666994387826128e-05,
      "loss": 0.4399,
      "mean_token_accuracy": 0.8523977041244507,
      "num_tokens": 795272972.0,
      "step": 23740
    },
    {
      "epoch": 1.4174426934097422,
      "grad_norm": 0.49592530727386475,
      "learning_rate": 1.4657874865022825e-05,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8547954201698303,
      "num_tokens": 795440652.0,
      "step": 23745
    },
    {
      "epoch": 1.417741165234002,
      "grad_norm": 0.5493110418319702,
      "learning_rate": 1.4648758470156138e-05,
      "loss": 0.4459,
      "mean_token_accuracy": 0.8516700506210327,
      "num_tokens": 795608332.0,
      "step": 23750
    },
    {
      "epoch": 1.4180396370582617,
      "grad_norm": 0.47795289754867554,
      "learning_rate": 1.4639645205446545e-05,
      "loss": 0.4151,
      "mean_token_accuracy": 0.8609924912452698,
      "num_tokens": 795776012.0,
      "step": 23755
    },
    {
      "epoch": 1.4183381088825215,
      "grad_norm": 0.4751960337162018,
      "learning_rate": 1.4630535073113726e-05,
      "loss": 0.437,
      "mean_token_accuracy": 0.8548073530197143,
      "num_tokens": 795943692.0,
      "step": 23760
    },
    {
      "epoch": 1.4186365807067813,
      "grad_norm": 0.4811757802963257,
      "learning_rate": 1.4621428075376637e-05,
      "loss": 0.412,
      "mean_token_accuracy": 0.8597280263900757,
      "num_tokens": 796111372.0,
      "step": 23765
    },
    {
      "epoch": 1.4189350525310411,
      "grad_norm": 0.5161427855491638,
      "learning_rate": 1.461232421445345e-05,
      "loss": 0.452,
      "mean_token_accuracy": 0.8475486040115356,
      "num_tokens": 796279052.0,
      "step": 23770
    },
    {
      "epoch": 1.419233524355301,
      "grad_norm": 0.5029381513595581,
      "learning_rate": 1.4603223492561579e-05,
      "loss": 0.449,
      "mean_token_accuracy": 0.8512823581695557,
      "num_tokens": 796446732.0,
      "step": 23775
    },
    {
      "epoch": 1.4195319961795607,
      "grad_norm": 0.5676060318946838,
      "learning_rate": 1.4594125911917678e-05,
      "loss": 0.4636,
      "mean_token_accuracy": 0.8461529374122619,
      "num_tokens": 796614412.0,
      "step": 23780
    },
    {
      "epoch": 1.4198304680038205,
      "grad_norm": 0.5017569065093994,
      "learning_rate": 1.4585031474737623e-05,
      "loss": 0.4364,
      "mean_token_accuracy": 0.8541810750961304,
      "num_tokens": 796782092.0,
      "step": 23785
    },
    {
      "epoch": 1.4201289398280803,
      "grad_norm": 0.5884302854537964,
      "learning_rate": 1.4575940183236533e-05,
      "loss": 0.4895,
      "mean_token_accuracy": 0.8387987613677979,
      "num_tokens": 796949772.0,
      "step": 23790
    },
    {
      "epoch": 1.42042741165234,
      "grad_norm": 0.5938997268676758,
      "learning_rate": 1.4566852039628765e-05,
      "loss": 0.4497,
      "mean_token_accuracy": 0.8499463200569153,
      "num_tokens": 797117452.0,
      "step": 23795
    },
    {
      "epoch": 1.4207258834766,
      "grad_norm": 0.5110284686088562,
      "learning_rate": 1.4557767046127901e-05,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8543958067893982,
      "num_tokens": 797285132.0,
      "step": 23800
    },
    {
      "epoch": 1.4210243553008595,
      "grad_norm": 0.5652061104774475,
      "learning_rate": 1.4548685204946766e-05,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8398604154586792,
      "num_tokens": 797452812.0,
      "step": 23805
    },
    {
      "epoch": 1.4213228271251195,
      "grad_norm": 0.5034496784210205,
      "learning_rate": 1.4539606518297389e-05,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8438506364822388,
      "num_tokens": 797620492.0,
      "step": 23810
    },
    {
      "epoch": 1.421621298949379,
      "grad_norm": 0.491308331489563,
      "learning_rate": 1.4530530988391075e-05,
      "loss": 0.4287,
      "mean_token_accuracy": 0.8562268853187561,
      "num_tokens": 797788172.0,
      "step": 23815
    },
    {
      "epoch": 1.421919770773639,
      "grad_norm": 0.4973183274269104,
      "learning_rate": 1.4521458617438328e-05,
      "loss": 0.4389,
      "mean_token_accuracy": 0.8530537962913514,
      "num_tokens": 797955852.0,
      "step": 23820
    },
    {
      "epoch": 1.4222182425978986,
      "grad_norm": 0.513624906539917,
      "learning_rate": 1.4512389407648897e-05,
      "loss": 0.4674,
      "mean_token_accuracy": 0.844411301612854,
      "num_tokens": 798123532.0,
      "step": 23825
    },
    {
      "epoch": 1.4225167144221587,
      "grad_norm": 0.484712690114975,
      "learning_rate": 1.450332336123173e-05,
      "loss": 0.4449,
      "mean_token_accuracy": 0.8539663553237915,
      "num_tokens": 798291212.0,
      "step": 23830
    },
    {
      "epoch": 1.4228151862464182,
      "grad_norm": 0.5327746868133545,
      "learning_rate": 1.4494260480395061e-05,
      "loss": 0.4185,
      "mean_token_accuracy": 0.8599427461624145,
      "num_tokens": 798458892.0,
      "step": 23835
    },
    {
      "epoch": 1.4231136580706782,
      "grad_norm": 0.4941696226596832,
      "learning_rate": 1.4485200767346318e-05,
      "loss": 0.4065,
      "mean_token_accuracy": 0.8639627695083618,
      "num_tokens": 798626572.0,
      "step": 23840
    },
    {
      "epoch": 1.4234121298949378,
      "grad_norm": 0.5326698422431946,
      "learning_rate": 1.4476144224292143e-05,
      "loss": 0.4656,
      "mean_token_accuracy": 0.8460157513618469,
      "num_tokens": 798794252.0,
      "step": 23845
    },
    {
      "epoch": 1.4237106017191978,
      "grad_norm": 0.5434609651565552,
      "learning_rate": 1.4467090853438436e-05,
      "loss": 0.4242,
      "mean_token_accuracy": 0.8582011222839355,
      "num_tokens": 798961932.0,
      "step": 23850
    },
    {
      "epoch": 1.4240090735434574,
      "grad_norm": 0.5191842913627625,
      "learning_rate": 1.4458040656990313e-05,
      "loss": 0.4477,
      "mean_token_accuracy": 0.85235595703125,
      "num_tokens": 799129612.0,
      "step": 23855
    },
    {
      "epoch": 1.4243075453677174,
      "grad_norm": 0.5037443041801453,
      "learning_rate": 1.4448993637152118e-05,
      "loss": 0.4313,
      "mean_token_accuracy": 0.8549206852912903,
      "num_tokens": 799297292.0,
      "step": 23860
    },
    {
      "epoch": 1.424606017191977,
      "grad_norm": 0.5144136548042297,
      "learning_rate": 1.4439949796127419e-05,
      "loss": 0.4616,
      "mean_token_accuracy": 0.846630084514618,
      "num_tokens": 799464972.0,
      "step": 23865
    },
    {
      "epoch": 1.4249044890162368,
      "grad_norm": 0.5368221402168274,
      "learning_rate": 1.4430909136118998e-05,
      "loss": 0.4251,
      "mean_token_accuracy": 0.8582726955413819,
      "num_tokens": 799632652.0,
      "step": 23870
    },
    {
      "epoch": 1.4252029608404966,
      "grad_norm": 0.5508019924163818,
      "learning_rate": 1.4421871659328895e-05,
      "loss": 0.4512,
      "mean_token_accuracy": 0.8493081092834472,
      "num_tokens": 799800332.0,
      "step": 23875
    },
    {
      "epoch": 1.4255014326647564,
      "grad_norm": 0.4998466968536377,
      "learning_rate": 1.4412837367958337e-05,
      "loss": 0.465,
      "mean_token_accuracy": 0.844071340560913,
      "num_tokens": 799968012.0,
      "step": 23880
    },
    {
      "epoch": 1.4257999044890162,
      "grad_norm": 0.4987039268016815,
      "learning_rate": 1.4403806264207803e-05,
      "loss": 0.4595,
      "mean_token_accuracy": 0.8475903749465943,
      "num_tokens": 800135692.0,
      "step": 23885
    },
    {
      "epoch": 1.426098376313276,
      "grad_norm": 0.5814368724822998,
      "learning_rate": 1.4394778350276977e-05,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8407108426094055,
      "num_tokens": 800298660.0,
      "step": 23890
    },
    {
      "epoch": 1.4263968481375358,
      "grad_norm": 0.5098195672035217,
      "learning_rate": 1.4385753628364778e-05,
      "loss": 0.477,
      "mean_token_accuracy": 0.8422462105751037,
      "num_tokens": 800466340.0,
      "step": 23895
    },
    {
      "epoch": 1.4266953199617955,
      "grad_norm": 0.46268561482429504,
      "learning_rate": 1.4376732100669358e-05,
      "loss": 0.4119,
      "mean_token_accuracy": 0.8615889191627503,
      "num_tokens": 800634020.0,
      "step": 23900
    },
    {
      "epoch": 1.4269937917860553,
      "grad_norm": 0.5035937428474426,
      "learning_rate": 1.436771376938805e-05,
      "loss": 0.394,
      "mean_token_accuracy": 0.8673386573791504,
      "num_tokens": 800801700.0,
      "step": 23905
    },
    {
      "epoch": 1.4272922636103151,
      "grad_norm": 0.5187295079231262,
      "learning_rate": 1.435869863671745e-05,
      "loss": 0.4426,
      "mean_token_accuracy": 0.851503050327301,
      "num_tokens": 800969380.0,
      "step": 23910
    },
    {
      "epoch": 1.427590735434575,
      "grad_norm": 0.50020432472229,
      "learning_rate": 1.4349686704853355e-05,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8506381869316101,
      "num_tokens": 801137060.0,
      "step": 23915
    },
    {
      "epoch": 1.4278892072588347,
      "grad_norm": 0.4603779911994934,
      "learning_rate": 1.434067797599079e-05,
      "loss": 0.4087,
      "mean_token_accuracy": 0.8621853709220886,
      "num_tokens": 801304740.0,
      "step": 23920
    },
    {
      "epoch": 1.4281876790830945,
      "grad_norm": 0.48090657591819763,
      "learning_rate": 1.4331672452324007e-05,
      "loss": 0.4581,
      "mean_token_accuracy": 0.8485804677009583,
      "num_tokens": 801472420.0,
      "step": 23925
    },
    {
      "epoch": 1.4284861509073543,
      "grad_norm": 0.45343217253685,
      "learning_rate": 1.4322670136046451e-05,
      "loss": 0.4555,
      "mean_token_accuracy": 0.8493558406829834,
      "num_tokens": 801640100.0,
      "step": 23930
    },
    {
      "epoch": 1.428784622731614,
      "grad_norm": 0.5139942169189453,
      "learning_rate": 1.4313671029350805e-05,
      "loss": 0.4535,
      "mean_token_accuracy": 0.8499522805213928,
      "num_tokens": 801807780.0,
      "step": 23935
    },
    {
      "epoch": 1.429083094555874,
      "grad_norm": 0.4862551689147949,
      "learning_rate": 1.4304675134428974e-05,
      "loss": 0.4312,
      "mean_token_accuracy": 0.8545210480690002,
      "num_tokens": 801975460.0,
      "step": 23940
    },
    {
      "epoch": 1.4293815663801337,
      "grad_norm": 0.5076653361320496,
      "learning_rate": 1.4295682453472073e-05,
      "loss": 0.4399,
      "mean_token_accuracy": 0.8531074762344361,
      "num_tokens": 802143140.0,
      "step": 23945
    },
    {
      "epoch": 1.4296800382043935,
      "grad_norm": 0.49128422141075134,
      "learning_rate": 1.4286692988670442e-05,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8561255097389221,
      "num_tokens": 802310820.0,
      "step": 23950
    },
    {
      "epoch": 1.4299785100286533,
      "grad_norm": 0.6112308502197266,
      "learning_rate": 1.4277706742213603e-05,
      "loss": 0.4292,
      "mean_token_accuracy": 0.8564356327056885,
      "num_tokens": 802478500.0,
      "step": 23955
    },
    {
      "epoch": 1.430276981852913,
      "grad_norm": 0.4948352575302124,
      "learning_rate": 1.4268723716290362e-05,
      "loss": 0.4196,
      "mean_token_accuracy": 0.8581653356552124,
      "num_tokens": 802646180.0,
      "step": 23960
    },
    {
      "epoch": 1.4305754536771729,
      "grad_norm": 0.5185667872428894,
      "learning_rate": 1.4259743913088674e-05,
      "loss": 0.4348,
      "mean_token_accuracy": 0.8555648446083068,
      "num_tokens": 802813860.0,
      "step": 23965
    },
    {
      "epoch": 1.4308739255014327,
      "grad_norm": 0.5161097049713135,
      "learning_rate": 1.4250767334795743e-05,
      "loss": 0.4615,
      "mean_token_accuracy": 0.8464931726455689,
      "num_tokens": 802974882.0,
      "step": 23970
    },
    {
      "epoch": 1.4311723973256925,
      "grad_norm": 0.5076270699501038,
      "learning_rate": 1.4241793983597978e-05,
      "loss": 0.4539,
      "mean_token_accuracy": 0.849385678768158,
      "num_tokens": 803142562.0,
      "step": 23975
    },
    {
      "epoch": 1.4314708691499523,
      "grad_norm": 0.494201123714447,
      "learning_rate": 1.423282386168101e-05,
      "loss": 0.4179,
      "mean_token_accuracy": 0.8607717871665954,
      "num_tokens": 803310242.0,
      "step": 23980
    },
    {
      "epoch": 1.431769340974212,
      "grad_norm": 0.4579995572566986,
      "learning_rate": 1.4223856971229682e-05,
      "loss": 0.408,
      "mean_token_accuracy": 0.8639150619506836,
      "num_tokens": 803477922.0,
      "step": 23985
    },
    {
      "epoch": 1.4320678127984718,
      "grad_norm": 0.45281463861465454,
      "learning_rate": 1.4214893314428026e-05,
      "loss": 0.4142,
      "mean_token_accuracy": 0.8614099979400635,
      "num_tokens": 803645602.0,
      "step": 23990
    },
    {
      "epoch": 1.4323662846227316,
      "grad_norm": 0.5281286835670471,
      "learning_rate": 1.420593289345931e-05,
      "loss": 0.4645,
      "mean_token_accuracy": 0.8459799647331238,
      "num_tokens": 803813282.0,
      "step": 23995
    },
    {
      "epoch": 1.4326647564469914,
      "grad_norm": 0.516741156578064,
      "learning_rate": 1.4196975710506044e-05,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8507038116455078,
      "num_tokens": 803980962.0,
      "step": 24000
    },
    {
      "epoch": 1.4329632282712512,
      "grad_norm": 0.4869138300418854,
      "learning_rate": 1.418802176774987e-05,
      "loss": 0.4414,
      "mean_token_accuracy": 0.8525348901748657,
      "num_tokens": 804148642.0,
      "step": 24005
    },
    {
      "epoch": 1.433261700095511,
      "grad_norm": 0.48890477418899536,
      "learning_rate": 1.4179071067371718e-05,
      "loss": 0.4346,
      "mean_token_accuracy": 0.854234766960144,
      "num_tokens": 804316322.0,
      "step": 24010
    },
    {
      "epoch": 1.4335601719197708,
      "grad_norm": 0.4909629821777344,
      "learning_rate": 1.4170123611551672e-05,
      "loss": 0.442,
      "mean_token_accuracy": 0.8507872939109802,
      "num_tokens": 804484002.0,
      "step": 24015
    },
    {
      "epoch": 1.4338586437440306,
      "grad_norm": 0.4792387783527374,
      "learning_rate": 1.4161179402469072e-05,
      "loss": 0.4363,
      "mean_token_accuracy": 0.856441605091095,
      "num_tokens": 804651682.0,
      "step": 24020
    },
    {
      "epoch": 1.4341571155682904,
      "grad_norm": 0.5897955298423767,
      "learning_rate": 1.415223844230243e-05,
      "loss": 0.4915,
      "mean_token_accuracy": 0.8374150156974792,
      "num_tokens": 804819362.0,
      "step": 24025
    },
    {
      "epoch": 1.4344555873925502,
      "grad_norm": 0.4968867897987366,
      "learning_rate": 1.4143300733229484e-05,
      "loss": 0.4306,
      "mean_token_accuracy": 0.8565012574195862,
      "num_tokens": 804987042.0,
      "step": 24030
    },
    {
      "epoch": 1.43475405921681,
      "grad_norm": 0.47743961215019226,
      "learning_rate": 1.413436627742718e-05,
      "loss": 0.4195,
      "mean_token_accuracy": 0.8582786560058594,
      "num_tokens": 805154722.0,
      "step": 24035
    },
    {
      "epoch": 1.4350525310410698,
      "grad_norm": 0.4852258861064911,
      "learning_rate": 1.4125435077071675e-05,
      "loss": 0.4175,
      "mean_token_accuracy": 0.8602051734924316,
      "num_tokens": 805322402.0,
      "step": 24040
    },
    {
      "epoch": 1.4353510028653296,
      "grad_norm": 0.5231012105941772,
      "learning_rate": 1.4116507134338328e-05,
      "loss": 0.4346,
      "mean_token_accuracy": 0.8538172602653503,
      "num_tokens": 805490082.0,
      "step": 24045
    },
    {
      "epoch": 1.4356494746895894,
      "grad_norm": 0.5689031481742859,
      "learning_rate": 1.410758245140169e-05,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8490695357322693,
      "num_tokens": 805657762.0,
      "step": 24050
    },
    {
      "epoch": 1.4359479465138492,
      "grad_norm": 0.50330650806427,
      "learning_rate": 1.4098661030435545e-05,
      "loss": 0.4139,
      "mean_token_accuracy": 0.8612727999687195,
      "num_tokens": 805825442.0,
      "step": 24055
    },
    {
      "epoch": 1.436246418338109,
      "grad_norm": 0.5357939004898071,
      "learning_rate": 1.4089742873612866e-05,
      "loss": 0.4675,
      "mean_token_accuracy": 0.8457731962203979,
      "num_tokens": 805990390.0,
      "step": 24060
    },
    {
      "epoch": 1.4365448901623687,
      "grad_norm": 0.49177396297454834,
      "learning_rate": 1.4080827983105832e-05,
      "loss": 0.4519,
      "mean_token_accuracy": 0.8493379473686218,
      "num_tokens": 806158070.0,
      "step": 24065
    },
    {
      "epoch": 1.4368433619866285,
      "grad_norm": 0.4820666015148163,
      "learning_rate": 1.4071916361085846e-05,
      "loss": 0.4299,
      "mean_token_accuracy": 0.8576046705245972,
      "num_tokens": 806325750.0,
      "step": 24070
    },
    {
      "epoch": 1.4371418338108883,
      "grad_norm": 0.459162175655365,
      "learning_rate": 1.4063008009723472e-05,
      "loss": 0.4389,
      "mean_token_accuracy": 0.8532625555992126,
      "num_tokens": 806493430.0,
      "step": 24075
    },
    {
      "epoch": 1.437440305635148,
      "grad_norm": 0.5223341584205627,
      "learning_rate": 1.4054102931188507e-05,
      "loss": 0.4511,
      "mean_token_accuracy": 0.8505904793739318,
      "num_tokens": 806661110.0,
      "step": 24080
    },
    {
      "epoch": 1.437738777459408,
      "grad_norm": 0.4740830957889557,
      "learning_rate": 1.4045201127649976e-05,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8490993618965149,
      "num_tokens": 806828790.0,
      "step": 24085
    },
    {
      "epoch": 1.4380372492836675,
      "grad_norm": 0.5166698694229126,
      "learning_rate": 1.4036302601276047e-05,
      "loss": 0.4724,
      "mean_token_accuracy": 0.8437253952026367,
      "num_tokens": 806996470.0,
      "step": 24090
    },
    {
      "epoch": 1.4383357211079275,
      "grad_norm": 0.5253591537475586,
      "learning_rate": 1.4027407354234128e-05,
      "loss": 0.4581,
      "mean_token_accuracy": 0.8484790682792663,
      "num_tokens": 807164150.0,
      "step": 24095
    },
    {
      "epoch": 1.438634192932187,
      "grad_norm": 0.4748472273349762,
      "learning_rate": 1.4018515388690828e-05,
      "loss": 0.4738,
      "mean_token_accuracy": 0.8433436632156373,
      "num_tokens": 807331830.0,
      "step": 24100
    },
    {
      "epoch": 1.438932664756447,
      "grad_norm": 0.4808465242385864,
      "learning_rate": 1.4009626706811945e-05,
      "loss": 0.4875,
      "mean_token_accuracy": 0.839824640750885,
      "num_tokens": 807499510.0,
      "step": 24105
    },
    {
      "epoch": 1.4392311365807067,
      "grad_norm": 0.48362985253334045,
      "learning_rate": 1.4000741310762489e-05,
      "loss": 0.4401,
      "mean_token_accuracy": 0.8532029032707215,
      "num_tokens": 807667190.0,
      "step": 24110
    },
    {
      "epoch": 1.4395296084049667,
      "grad_norm": 0.585921049118042,
      "learning_rate": 1.3991859202706641e-05,
      "loss": 0.4538,
      "mean_token_accuracy": 0.849123227596283,
      "num_tokens": 807834870.0,
      "step": 24115
    },
    {
      "epoch": 1.4398280802292263,
      "grad_norm": 0.5421547293663025,
      "learning_rate": 1.3982980384807824e-05,
      "loss": 0.4277,
      "mean_token_accuracy": 0.8572050571441651,
      "num_tokens": 808002550.0,
      "step": 24120
    },
    {
      "epoch": 1.4401265520534863,
      "grad_norm": 0.540352463722229,
      "learning_rate": 1.3974104859228624e-05,
      "loss": 0.436,
      "mean_token_accuracy": 0.8536025285720825,
      "num_tokens": 808170230.0,
      "step": 24125
    },
    {
      "epoch": 1.4404250238777458,
      "grad_norm": 0.5015534162521362,
      "learning_rate": 1.3965232628130854e-05,
      "loss": 0.4348,
      "mean_token_accuracy": 0.8538709282875061,
      "num_tokens": 808337910.0,
      "step": 24130
    },
    {
      "epoch": 1.4407234957020059,
      "grad_norm": 0.5280135273933411,
      "learning_rate": 1.3956363693675495e-05,
      "loss": 0.4403,
      "mean_token_accuracy": 0.8545210480690002,
      "num_tokens": 808505590.0,
      "step": 24135
    },
    {
      "epoch": 1.4410219675262654,
      "grad_norm": 0.5273571610450745,
      "learning_rate": 1.3947498058022734e-05,
      "loss": 0.4306,
      "mean_token_accuracy": 0.8572229385375977,
      "num_tokens": 808673270.0,
      "step": 24140
    },
    {
      "epoch": 1.4413204393505252,
      "grad_norm": 0.44331005215644836,
      "learning_rate": 1.393863572333199e-05,
      "loss": 0.3998,
      "mean_token_accuracy": 0.8642669677734375,
      "num_tokens": 808840950.0,
      "step": 24145
    },
    {
      "epoch": 1.441618911174785,
      "grad_norm": 0.4782272279262543,
      "learning_rate": 1.392977669176182e-05,
      "loss": 0.4611,
      "mean_token_accuracy": 0.8476082444190979,
      "num_tokens": 809008630.0,
      "step": 24150
    },
    {
      "epoch": 1.4419173829990448,
      "grad_norm": 0.542181134223938,
      "learning_rate": 1.3920920965470024e-05,
      "loss": 0.431,
      "mean_token_accuracy": 0.857670271396637,
      "num_tokens": 809176310.0,
      "step": 24155
    },
    {
      "epoch": 1.4422158548233046,
      "grad_norm": 0.4955284297466278,
      "learning_rate": 1.3912068546613549e-05,
      "loss": 0.4209,
      "mean_token_accuracy": 0.8584874153137207,
      "num_tokens": 809343990.0,
      "step": 24160
    },
    {
      "epoch": 1.4425143266475644,
      "grad_norm": 0.5144757628440857,
      "learning_rate": 1.3903219437348597e-05,
      "loss": 0.4345,
      "mean_token_accuracy": 0.8535727143287659,
      "num_tokens": 809511670.0,
      "step": 24165
    },
    {
      "epoch": 1.4428127984718242,
      "grad_norm": 0.5090042948722839,
      "learning_rate": 1.3894373639830521e-05,
      "loss": 0.4337,
      "mean_token_accuracy": 0.8532446622848511,
      "num_tokens": 809679350.0,
      "step": 24170
    },
    {
      "epoch": 1.443111270296084,
      "grad_norm": 0.510388970375061,
      "learning_rate": 1.3885531156213871e-05,
      "loss": 0.435,
      "mean_token_accuracy": 0.8543182611465454,
      "num_tokens": 809847030.0,
      "step": 24175
    },
    {
      "epoch": 1.4434097421203438,
      "grad_norm": 0.5230668187141418,
      "learning_rate": 1.3876691988652402e-05,
      "loss": 0.4288,
      "mean_token_accuracy": 0.8576643228530884,
      "num_tokens": 810014710.0,
      "step": 24180
    },
    {
      "epoch": 1.4437082139446036,
      "grad_norm": 0.5183838605880737,
      "learning_rate": 1.3867856139299057e-05,
      "loss": 0.4627,
      "mean_token_accuracy": 0.8450316309928894,
      "num_tokens": 810182390.0,
      "step": 24185
    },
    {
      "epoch": 1.4440066857688634,
      "grad_norm": 0.49484556913375854,
      "learning_rate": 1.3859023610305972e-05,
      "loss": 0.4082,
      "mean_token_accuracy": 0.862143623828888,
      "num_tokens": 810350070.0,
      "step": 24190
    },
    {
      "epoch": 1.4443051575931232,
      "grad_norm": 0.5739304423332214,
      "learning_rate": 1.3850194403824474e-05,
      "loss": 0.4923,
      "mean_token_accuracy": 0.8381963491439819,
      "num_tokens": 810517750.0,
      "step": 24195
    },
    {
      "epoch": 1.444603629417383,
      "grad_norm": 0.5037735104560852,
      "learning_rate": 1.3841368522005069e-05,
      "loss": 0.4144,
      "mean_token_accuracy": 0.8597638130187988,
      "num_tokens": 810685430.0,
      "step": 24200
    },
    {
      "epoch": 1.4449021012416428,
      "grad_norm": 0.4882184863090515,
      "learning_rate": 1.3832545966997468e-05,
      "loss": 0.4281,
      "mean_token_accuracy": 0.8564177513122558,
      "num_tokens": 810853110.0,
      "step": 24205
    },
    {
      "epoch": 1.4452005730659025,
      "grad_norm": 0.5717160105705261,
      "learning_rate": 1.382372674095057e-05,
      "loss": 0.4266,
      "mean_token_accuracy": 0.856692111492157,
      "num_tokens": 811020790.0,
      "step": 24210
    },
    {
      "epoch": 1.4454990448901623,
      "grad_norm": 0.5074063539505005,
      "learning_rate": 1.3814910846012463e-05,
      "loss": 0.4328,
      "mean_token_accuracy": 0.8543566107749939,
      "num_tokens": 811180279.0,
      "step": 24215
    },
    {
      "epoch": 1.4457975167144221,
      "grad_norm": 0.48674798011779785,
      "learning_rate": 1.3806098284330415e-05,
      "loss": 0.441,
      "mean_token_accuracy": 0.8518430113792419,
      "num_tokens": 811347959.0,
      "step": 24220
    },
    {
      "epoch": 1.446095988538682,
      "grad_norm": 0.5206977725028992,
      "learning_rate": 1.379728905805089e-05,
      "loss": 0.4394,
      "mean_token_accuracy": 0.8528152227401733,
      "num_tokens": 811515639.0,
      "step": 24225
    },
    {
      "epoch": 1.4463944603629417,
      "grad_norm": 0.4975236654281616,
      "learning_rate": 1.3788483169319544e-05,
      "loss": 0.4721,
      "mean_token_accuracy": 0.8434629559516906,
      "num_tokens": 811683319.0,
      "step": 24230
    },
    {
      "epoch": 1.4466929321872015,
      "grad_norm": 0.502943217754364,
      "learning_rate": 1.3779680620281204e-05,
      "loss": 0.4168,
      "mean_token_accuracy": 0.8585530281066894,
      "num_tokens": 811850999.0,
      "step": 24235
    },
    {
      "epoch": 1.4469914040114613,
      "grad_norm": 0.5049415230751038,
      "learning_rate": 1.3770881413079895e-05,
      "loss": 0.4657,
      "mean_token_accuracy": 0.845765233039856,
      "num_tokens": 812018679.0,
      "step": 24240
    },
    {
      "epoch": 1.447289875835721,
      "grad_norm": 0.49742600321769714,
      "learning_rate": 1.3762085549858828e-05,
      "loss": 0.4042,
      "mean_token_accuracy": 0.8638136625289917,
      "num_tokens": 812186359.0,
      "step": 24245
    },
    {
      "epoch": 1.447588347659981,
      "grad_norm": 0.45932161808013916,
      "learning_rate": 1.37532930327604e-05,
      "loss": 0.4286,
      "mean_token_accuracy": 0.8566980838775635,
      "num_tokens": 812354039.0,
      "step": 24250
    },
    {
      "epoch": 1.4478868194842407,
      "grad_norm": 0.5616112947463989,
      "learning_rate": 1.3744503863926195e-05,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8550817131996155,
      "num_tokens": 812521719.0,
      "step": 24255
    },
    {
      "epoch": 1.4481852913085005,
      "grad_norm": 0.5444290637969971,
      "learning_rate": 1.3735718045496964e-05,
      "loss": 0.4383,
      "mean_token_accuracy": 0.852868914604187,
      "num_tokens": 812689399.0,
      "step": 24260
    },
    {
      "epoch": 1.4484837631327603,
      "grad_norm": 0.47080856561660767,
      "learning_rate": 1.3726935579612662e-05,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8562447905540467,
      "num_tokens": 812857079.0,
      "step": 24265
    },
    {
      "epoch": 1.44878223495702,
      "grad_norm": 0.47317975759506226,
      "learning_rate": 1.3718156468412418e-05,
      "loss": 0.446,
      "mean_token_accuracy": 0.8499165058135987,
      "num_tokens": 813024759.0,
      "step": 24270
    },
    {
      "epoch": 1.4490807067812799,
      "grad_norm": 0.4811856746673584,
      "learning_rate": 1.370938071403456e-05,
      "loss": 0.4745,
      "mean_token_accuracy": 0.8426637291908264,
      "num_tokens": 813192439.0,
      "step": 24275
    },
    {
      "epoch": 1.4493791786055397,
      "grad_norm": 0.4584912955760956,
      "learning_rate": 1.370060831861657e-05,
      "loss": 0.4451,
      "mean_token_accuracy": 0.8509421825408936,
      "num_tokens": 813355594.0,
      "step": 24280
    },
    {
      "epoch": 1.4496776504297995,
      "grad_norm": 0.5099167227745056,
      "learning_rate": 1.3691839284295121e-05,
      "loss": 0.4181,
      "mean_token_accuracy": 0.8598353862762451,
      "num_tokens": 813523274.0,
      "step": 24285
    },
    {
      "epoch": 1.4499761222540593,
      "grad_norm": 0.5842281579971313,
      "learning_rate": 1.3683073613206101e-05,
      "loss": 0.4558,
      "mean_token_accuracy": 0.8484492540359497,
      "num_tokens": 813690954.0,
      "step": 24290
    },
    {
      "epoch": 1.450274594078319,
      "grad_norm": 0.5133460164070129,
      "learning_rate": 1.3674311307484528e-05,
      "loss": 0.4506,
      "mean_token_accuracy": 0.8483180284500123,
      "num_tokens": 813858634.0,
      "step": 24295
    },
    {
      "epoch": 1.4505730659025788,
      "grad_norm": 0.46325740218162537,
      "learning_rate": 1.3665552369264625e-05,
      "loss": 0.5106,
      "mean_token_accuracy": 0.831742811203003,
      "num_tokens": 814026314.0,
      "step": 24300
    },
    {
      "epoch": 1.4508715377268386,
      "grad_norm": 0.5090318322181702,
      "learning_rate": 1.36567968006798e-05,
      "loss": 0.4332,
      "mean_token_accuracy": 0.8538888096809387,
      "num_tokens": 814193994.0,
      "step": 24305
    },
    {
      "epoch": 1.4511700095510984,
      "grad_norm": 0.5046600103378296,
      "learning_rate": 1.3648044603862625e-05,
      "loss": 0.4608,
      "mean_token_accuracy": 0.8458785653114319,
      "num_tokens": 814361674.0,
      "step": 24310
    },
    {
      "epoch": 1.4514684813753582,
      "grad_norm": 0.48981788754463196,
      "learning_rate": 1.3639295780944877e-05,
      "loss": 0.4431,
      "mean_token_accuracy": 0.8519622921943665,
      "num_tokens": 814529354.0,
      "step": 24315
    },
    {
      "epoch": 1.451766953199618,
      "grad_norm": 0.5356281995773315,
      "learning_rate": 1.3630550334057473e-05,
      "loss": 0.4579,
      "mean_token_accuracy": 0.847280216217041,
      "num_tokens": 814697034.0,
      "step": 24320
    },
    {
      "epoch": 1.4520654250238778,
      "grad_norm": 0.4935963451862335,
      "learning_rate": 1.3621808265330536e-05,
      "loss": 0.4256,
      "mean_token_accuracy": 0.8570320844650269,
      "num_tokens": 814864714.0,
      "step": 24325
    },
    {
      "epoch": 1.4523638968481376,
      "grad_norm": 0.4961432218551636,
      "learning_rate": 1.3613069576893356e-05,
      "loss": 0.441,
      "mean_token_accuracy": 0.8521710634231567,
      "num_tokens": 815032394.0,
      "step": 24330
    },
    {
      "epoch": 1.4526623686723974,
      "grad_norm": 0.5234033465385437,
      "learning_rate": 1.3604334270874405e-05,
      "loss": 0.4401,
      "mean_token_accuracy": 0.852731716632843,
      "num_tokens": 815200074.0,
      "step": 24335
    },
    {
      "epoch": 1.4529608404966572,
      "grad_norm": 0.4845702052116394,
      "learning_rate": 1.3595602349401333e-05,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8402958273887634,
      "num_tokens": 815367754.0,
      "step": 24340
    },
    {
      "epoch": 1.453259312320917,
      "grad_norm": 0.6409403085708618,
      "learning_rate": 1.3586873814600941e-05,
      "loss": 0.4563,
      "mean_token_accuracy": 0.8497852921485901,
      "num_tokens": 815535434.0,
      "step": 24345
    },
    {
      "epoch": 1.4535577841451768,
      "grad_norm": 0.5014445185661316,
      "learning_rate": 1.3578148668599256e-05,
      "loss": 0.4535,
      "mean_token_accuracy": 0.8493021726608276,
      "num_tokens": 815703114.0,
      "step": 24350
    },
    {
      "epoch": 1.4538562559694364,
      "grad_norm": 0.529164731502533,
      "learning_rate": 1.3569426913521422e-05,
      "loss": 0.4419,
      "mean_token_accuracy": 0.851264476776123,
      "num_tokens": 815870794.0,
      "step": 24355
    },
    {
      "epoch": 1.4541547277936964,
      "grad_norm": 0.5152989029884338,
      "learning_rate": 1.3560708551491785e-05,
      "loss": 0.4573,
      "mean_token_accuracy": 0.8457413792610169,
      "num_tokens": 816038474.0,
      "step": 24360
    },
    {
      "epoch": 1.454453199617956,
      "grad_norm": 0.4938627779483795,
      "learning_rate": 1.355199358463388e-05,
      "loss": 0.4583,
      "mean_token_accuracy": 0.847894549369812,
      "num_tokens": 816206154.0,
      "step": 24365
    },
    {
      "epoch": 1.454751671442216,
      "grad_norm": 0.5390340089797974,
      "learning_rate": 1.354328201507038e-05,
      "loss": 0.4548,
      "mean_token_accuracy": 0.8480436563491821,
      "num_tokens": 816373834.0,
      "step": 24370
    },
    {
      "epoch": 1.4550501432664755,
      "grad_norm": 0.4630414545536041,
      "learning_rate": 1.3534573844923171e-05,
      "loss": 0.381,
      "mean_token_accuracy": 0.8706250667572022,
      "num_tokens": 816541514.0,
      "step": 24375
    },
    {
      "epoch": 1.4553486150907355,
      "grad_norm": 0.5235395431518555,
      "learning_rate": 1.3525869076313263e-05,
      "loss": 0.4473,
      "mean_token_accuracy": 0.8517833590507508,
      "num_tokens": 816709194.0,
      "step": 24380
    },
    {
      "epoch": 1.4556470869149951,
      "grad_norm": 0.48764026165008545,
      "learning_rate": 1.3517167711360873e-05,
      "loss": 0.4042,
      "mean_token_accuracy": 0.8617141723632813,
      "num_tokens": 816876874.0,
      "step": 24385
    },
    {
      "epoch": 1.4559455587392551,
      "grad_norm": 0.5036116242408752,
      "learning_rate": 1.3508469752185382e-05,
      "loss": 0.458,
      "mean_token_accuracy": 0.8485387086868286,
      "num_tokens": 817044554.0,
      "step": 24390
    },
    {
      "epoch": 1.4562440305635147,
      "grad_norm": 0.5442743897438049,
      "learning_rate": 1.349977520090534e-05,
      "loss": 0.4369,
      "mean_token_accuracy": 0.8539186358451843,
      "num_tokens": 817212234.0,
      "step": 24395
    },
    {
      "epoch": 1.4565425023877747,
      "grad_norm": 0.5036551356315613,
      "learning_rate": 1.3491084059638468e-05,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8475008845329285,
      "num_tokens": 817379914.0,
      "step": 24400
    },
    {
      "epoch": 1.4568409742120343,
      "grad_norm": 0.511569082736969,
      "learning_rate": 1.348239633050164e-05,
      "loss": 0.4501,
      "mean_token_accuracy": 0.8490695476531982,
      "num_tokens": 817547594.0,
      "step": 24405
    },
    {
      "epoch": 1.4571394460362943,
      "grad_norm": 0.47128644585609436,
      "learning_rate": 1.3473712015610913e-05,
      "loss": 0.4643,
      "mean_token_accuracy": 0.8441130876541137,
      "num_tokens": 817715274.0,
      "step": 24410
    },
    {
      "epoch": 1.4574379178605539,
      "grad_norm": 0.5112484097480774,
      "learning_rate": 1.3465031117081533e-05,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8451151132583619,
      "num_tokens": 817882954.0,
      "step": 24415
    },
    {
      "epoch": 1.4577363896848137,
      "grad_norm": 0.523012638092041,
      "learning_rate": 1.3456353637027879e-05,
      "loss": 0.4033,
      "mean_token_accuracy": 0.8637540340423584,
      "num_tokens": 818050634.0,
      "step": 24420
    },
    {
      "epoch": 1.4580348615090735,
      "grad_norm": 0.5245733857154846,
      "learning_rate": 1.344767957756351e-05,
      "loss": 0.4593,
      "mean_token_accuracy": 0.8489562273025513,
      "num_tokens": 818218314.0,
      "step": 24425
    },
    {
      "epoch": 1.4583333333333333,
      "grad_norm": 0.4996326267719269,
      "learning_rate": 1.3439008940801156e-05,
      "loss": 0.427,
      "mean_token_accuracy": 0.8570380568504333,
      "num_tokens": 818385994.0,
      "step": 24430
    },
    {
      "epoch": 1.458631805157593,
      "grad_norm": 0.4786929786205292,
      "learning_rate": 1.3430341728852708e-05,
      "loss": 0.4061,
      "mean_token_accuracy": 0.8649528861045838,
      "num_tokens": 818553674.0,
      "step": 24435
    },
    {
      "epoch": 1.4589302769818528,
      "grad_norm": 0.552032470703125,
      "learning_rate": 1.3421677943829236e-05,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8496182799339295,
      "num_tokens": 818721354.0,
      "step": 24440
    },
    {
      "epoch": 1.4592287488061126,
      "grad_norm": 0.4969017803668976,
      "learning_rate": 1.3413017587840953e-05,
      "loss": 0.4282,
      "mean_token_accuracy": 0.8564296722412109,
      "num_tokens": 818889034.0,
      "step": 24445
    },
    {
      "epoch": 1.4595272206303724,
      "grad_norm": 0.5180125832557678,
      "learning_rate": 1.340436066299725e-05,
      "loss": 0.4584,
      "mean_token_accuracy": 0.8482822299003601,
      "num_tokens": 819056714.0,
      "step": 24450
    },
    {
      "epoch": 1.4598256924546322,
      "grad_norm": 0.5134487748146057,
      "learning_rate": 1.3395707171406685e-05,
      "loss": 0.4556,
      "mean_token_accuracy": 0.8473577499389648,
      "num_tokens": 819224394.0,
      "step": 24455
    },
    {
      "epoch": 1.460124164278892,
      "grad_norm": 0.490887314081192,
      "learning_rate": 1.3387057115176977e-05,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8564773797988892,
      "num_tokens": 819392074.0,
      "step": 24460
    },
    {
      "epoch": 1.4604226361031518,
      "grad_norm": 0.5028799176216125,
      "learning_rate": 1.3378410496415e-05,
      "loss": 0.4423,
      "mean_token_accuracy": 0.8521054506301879,
      "num_tokens": 819559754.0,
      "step": 24465
    },
    {
      "epoch": 1.4607211079274116,
      "grad_norm": 0.48976489901542664,
      "learning_rate": 1.3369767317226795e-05,
      "loss": 0.4399,
      "mean_token_accuracy": 0.8515448093414306,
      "num_tokens": 819727434.0,
      "step": 24470
    },
    {
      "epoch": 1.4610195797516714,
      "grad_norm": 0.5112046003341675,
      "learning_rate": 1.3361127579717586e-05,
      "loss": 0.4033,
      "mean_token_accuracy": 0.8626385688781738,
      "num_tokens": 819890728.0,
      "step": 24475
    },
    {
      "epoch": 1.4613180515759312,
      "grad_norm": 0.47551229596138,
      "learning_rate": 1.335249128599172e-05,
      "loss": 0.4397,
      "mean_token_accuracy": 0.8535846471786499,
      "num_tokens": 820058408.0,
      "step": 24480
    },
    {
      "epoch": 1.461616523400191,
      "grad_norm": 0.6014630198478699,
      "learning_rate": 1.3343858438152745e-05,
      "loss": 0.4642,
      "mean_token_accuracy": 0.8459501385688781,
      "num_tokens": 820226088.0,
      "step": 24485
    },
    {
      "epoch": 1.4619149952244508,
      "grad_norm": 0.5242108106613159,
      "learning_rate": 1.3335229038303326e-05,
      "loss": 0.4341,
      "mean_token_accuracy": 0.8538053154945373,
      "num_tokens": 820393768.0,
      "step": 24490
    },
    {
      "epoch": 1.4622134670487106,
      "grad_norm": 0.5242660045623779,
      "learning_rate": 1.3326603088545328e-05,
      "loss": 0.4188,
      "mean_token_accuracy": 0.8590838551521301,
      "num_tokens": 820561448.0,
      "step": 24495
    },
    {
      "epoch": 1.4625119388729704,
      "grad_norm": 0.457776814699173,
      "learning_rate": 1.3317980590979774e-05,
      "loss": 0.4054,
      "mean_token_accuracy": 0.8620660781860352,
      "num_tokens": 820729128.0,
      "step": 24500
    },
    {
      "epoch": 1.4628104106972302,
      "grad_norm": 0.5410901308059692,
      "learning_rate": 1.3309361547706806e-05,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8497972011566162,
      "num_tokens": 820896808.0,
      "step": 24505
    },
    {
      "epoch": 1.46310888252149,
      "grad_norm": 0.5776258707046509,
      "learning_rate": 1.3300745960825762e-05,
      "loss": 0.441,
      "mean_token_accuracy": 0.8529523968696594,
      "num_tokens": 821064488.0,
      "step": 24510
    },
    {
      "epoch": 1.4634073543457498,
      "grad_norm": 0.4791201949119568,
      "learning_rate": 1.3292133832435122e-05,
      "loss": 0.3998,
      "mean_token_accuracy": 0.8648276209831238,
      "num_tokens": 821232168.0,
      "step": 24515
    },
    {
      "epoch": 1.4637058261700095,
      "grad_norm": 0.5130822062492371,
      "learning_rate": 1.3283525164632543e-05,
      "loss": 0.4258,
      "mean_token_accuracy": 0.8583919882774353,
      "num_tokens": 821399848.0,
      "step": 24520
    },
    {
      "epoch": 1.4640042979942693,
      "grad_norm": 0.5081445574760437,
      "learning_rate": 1.327491995951482e-05,
      "loss": 0.4291,
      "mean_token_accuracy": 0.8570738315582276,
      "num_tokens": 821567528.0,
      "step": 24525
    },
    {
      "epoch": 1.4643027698185291,
      "grad_norm": 0.4760996103286743,
      "learning_rate": 1.3266318219177898e-05,
      "loss": 0.3947,
      "mean_token_accuracy": 0.8671120166778564,
      "num_tokens": 821735208.0,
      "step": 24530
    },
    {
      "epoch": 1.464601241642789,
      "grad_norm": 0.4987267851829529,
      "learning_rate": 1.3257719945716896e-05,
      "loss": 0.4242,
      "mean_token_accuracy": 0.8592091083526612,
      "num_tokens": 821902888.0,
      "step": 24535
    },
    {
      "epoch": 1.4648997134670487,
      "grad_norm": 0.5035992860794067,
      "learning_rate": 1.3249125141226088e-05,
      "loss": 0.4473,
      "mean_token_accuracy": 0.850876784324646,
      "num_tokens": 822070568.0,
      "step": 24540
    },
    {
      "epoch": 1.4651981852913085,
      "grad_norm": 0.4848971962928772,
      "learning_rate": 1.3240533807798888e-05,
      "loss": 0.4039,
      "mean_token_accuracy": 0.8631754755973816,
      "num_tokens": 822238248.0,
      "step": 24545
    },
    {
      "epoch": 1.4654966571155683,
      "grad_norm": 0.5123375654220581,
      "learning_rate": 1.323194594752788e-05,
      "loss": 0.4343,
      "mean_token_accuracy": 0.8553023934364319,
      "num_tokens": 822405928.0,
      "step": 24550
    },
    {
      "epoch": 1.465795128939828,
      "grad_norm": 0.5385374426841736,
      "learning_rate": 1.3223361562504796e-05,
      "loss": 0.4049,
      "mean_token_accuracy": 0.8639747023582458,
      "num_tokens": 822573608.0,
      "step": 24555
    },
    {
      "epoch": 1.466093600764088,
      "grad_norm": 0.5345389246940613,
      "learning_rate": 1.3214780654820527e-05,
      "loss": 0.4636,
      "mean_token_accuracy": 0.8473994851112365,
      "num_tokens": 822741288.0,
      "step": 24560
    },
    {
      "epoch": 1.4663920725883477,
      "grad_norm": 0.49947449564933777,
      "learning_rate": 1.3206203226565098e-05,
      "loss": 0.4104,
      "mean_token_accuracy": 0.8613324522972107,
      "num_tokens": 822908968.0,
      "step": 24565
    },
    {
      "epoch": 1.4666905444126075,
      "grad_norm": 0.5049331188201904,
      "learning_rate": 1.3197629279827705e-05,
      "loss": 0.4296,
      "mean_token_accuracy": 0.8557437658309937,
      "num_tokens": 823076648.0,
      "step": 24570
    },
    {
      "epoch": 1.4669890162368673,
      "grad_norm": 0.5047481656074524,
      "learning_rate": 1.3189058816696692e-05,
      "loss": 0.4694,
      "mean_token_accuracy": 0.8450554609298706,
      "num_tokens": 823244328.0,
      "step": 24575
    },
    {
      "epoch": 1.467287488061127,
      "grad_norm": 0.52135169506073,
      "learning_rate": 1.3180491839259557e-05,
      "loss": 0.4115,
      "mean_token_accuracy": 0.8619288921356201,
      "num_tokens": 823412008.0,
      "step": 24580
    },
    {
      "epoch": 1.4675859598853869,
      "grad_norm": 0.4863163232803345,
      "learning_rate": 1.3171928349602947e-05,
      "loss": 0.4352,
      "mean_token_accuracy": 0.8533818364143372,
      "num_tokens": 823579688.0,
      "step": 24585
    },
    {
      "epoch": 1.4678844317096467,
      "grad_norm": 0.5247222185134888,
      "learning_rate": 1.3163368349812643e-05,
      "loss": 0.4307,
      "mean_token_accuracy": 0.8565191507339478,
      "num_tokens": 823747368.0,
      "step": 24590
    },
    {
      "epoch": 1.4681829035339065,
      "grad_norm": 0.499421626329422,
      "learning_rate": 1.3154811841973608e-05,
      "loss": 0.4428,
      "mean_token_accuracy": 0.8511213183403015,
      "num_tokens": 823915048.0,
      "step": 24595
    },
    {
      "epoch": 1.4684813753581663,
      "grad_norm": 0.5525295734405518,
      "learning_rate": 1.3146258828169927e-05,
      "loss": 0.4732,
      "mean_token_accuracy": 0.8406091332435608,
      "num_tokens": 824075951.0,
      "step": 24600
    },
    {
      "epoch": 1.468779847182426,
      "grad_norm": 0.5224230289459229,
      "learning_rate": 1.3137709310484847e-05,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8445663809776306,
      "num_tokens": 824243631.0,
      "step": 24605
    },
    {
      "epoch": 1.4690783190066858,
      "grad_norm": 0.5163947343826294,
      "learning_rate": 1.3129163291000773e-05,
      "loss": 0.4537,
      "mean_token_accuracy": 0.8477394700050354,
      "num_tokens": 824411311.0,
      "step": 24610
    },
    {
      "epoch": 1.4693767908309456,
      "grad_norm": 0.5577158331871033,
      "learning_rate": 1.3120620771799216e-05,
      "loss": 0.4981,
      "mean_token_accuracy": 0.835864245891571,
      "num_tokens": 824578991.0,
      "step": 24615
    },
    {
      "epoch": 1.4696752626552054,
      "grad_norm": 0.4999620020389557,
      "learning_rate": 1.31120817549609e-05,
      "loss": 0.4491,
      "mean_token_accuracy": 0.8503220677375793,
      "num_tokens": 824746671.0,
      "step": 24620
    },
    {
      "epoch": 1.4699737344794652,
      "grad_norm": 0.46030867099761963,
      "learning_rate": 1.3103546242565635e-05,
      "loss": 0.4297,
      "mean_token_accuracy": 0.8575867772102356,
      "num_tokens": 824914351.0,
      "step": 24625
    },
    {
      "epoch": 1.4702722063037248,
      "grad_norm": 0.5278317928314209,
      "learning_rate": 1.3095014236692413e-05,
      "loss": 0.437,
      "mean_token_accuracy": 0.853095543384552,
      "num_tokens": 825082031.0,
      "step": 24630
    },
    {
      "epoch": 1.4705706781279848,
      "grad_norm": 0.5900847911834717,
      "learning_rate": 1.3086485739419358e-05,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8494751334190369,
      "num_tokens": 825249711.0,
      "step": 24635
    },
    {
      "epoch": 1.4708691499522444,
      "grad_norm": 0.5356927514076233,
      "learning_rate": 1.3077960752823742e-05,
      "loss": 0.4301,
      "mean_token_accuracy": 0.8565250992774963,
      "num_tokens": 825417391.0,
      "step": 24640
    },
    {
      "epoch": 1.4711676217765044,
      "grad_norm": 0.5412192940711975,
      "learning_rate": 1.3069439278981993e-05,
      "loss": 0.4273,
      "mean_token_accuracy": 0.8574078559875489,
      "num_tokens": 825585071.0,
      "step": 24645
    },
    {
      "epoch": 1.471466093600764,
      "grad_norm": 0.501054048538208,
      "learning_rate": 1.306092131996966e-05,
      "loss": 0.441,
      "mean_token_accuracy": 0.851968276500702,
      "num_tokens": 825752751.0,
      "step": 24650
    },
    {
      "epoch": 1.471764565425024,
      "grad_norm": 0.5443149209022522,
      "learning_rate": 1.305240687786145e-05,
      "loss": 0.4348,
      "mean_token_accuracy": 0.8554694056510925,
      "num_tokens": 825920431.0,
      "step": 24655
    },
    {
      "epoch": 1.4720630372492836,
      "grad_norm": 0.48157799243927,
      "learning_rate": 1.3043895954731222e-05,
      "loss": 0.4354,
      "mean_token_accuracy": 0.8546224474906922,
      "num_tokens": 826088111.0,
      "step": 24660
    },
    {
      "epoch": 1.4723615090735436,
      "grad_norm": 0.5196056365966797,
      "learning_rate": 1.3035388552651962e-05,
      "loss": 0.4246,
      "mean_token_accuracy": 0.8593096733093262,
      "num_tokens": 826251237.0,
      "step": 24665
    },
    {
      "epoch": 1.4726599808978031,
      "grad_norm": 0.48469680547714233,
      "learning_rate": 1.3026884673695813e-05,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8536979556083679,
      "num_tokens": 826418917.0,
      "step": 24670
    },
    {
      "epoch": 1.4729584527220632,
      "grad_norm": 0.5830373764038086,
      "learning_rate": 1.301838431993403e-05,
      "loss": 0.4656,
      "mean_token_accuracy": 0.8446558356285095,
      "num_tokens": 826586597.0,
      "step": 24675
    },
    {
      "epoch": 1.4732569245463227,
      "grad_norm": 0.45847535133361816,
      "learning_rate": 1.3009887493437061e-05,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8450017929077148,
      "num_tokens": 826754277.0,
      "step": 24680
    },
    {
      "epoch": 1.4735553963705827,
      "grad_norm": 0.5602812170982361,
      "learning_rate": 1.3001394196274441e-05,
      "loss": 0.4656,
      "mean_token_accuracy": 0.8452344059944152,
      "num_tokens": 826921957.0,
      "step": 24685
    },
    {
      "epoch": 1.4738538681948423,
      "grad_norm": 0.5139380097389221,
      "learning_rate": 1.299290443051488e-05,
      "loss": 0.4183,
      "mean_token_accuracy": 0.8604556798934937,
      "num_tokens": 827089637.0,
      "step": 24690
    },
    {
      "epoch": 1.4741523400191021,
      "grad_norm": 0.5584856271743774,
      "learning_rate": 1.2984418198226211e-05,
      "loss": 0.4624,
      "mean_token_accuracy": 0.8476619482040405,
      "num_tokens": 827257317.0,
      "step": 24695
    },
    {
      "epoch": 1.474450811843362,
      "grad_norm": 0.47130510210990906,
      "learning_rate": 1.2975935501475422e-05,
      "loss": 0.4023,
      "mean_token_accuracy": 0.864010500907898,
      "num_tokens": 827424997.0,
      "step": 24700
    },
    {
      "epoch": 1.4747492836676217,
      "grad_norm": 0.4731478989124298,
      "learning_rate": 1.2967456342328629e-05,
      "loss": 0.4778,
      "mean_token_accuracy": 0.8413694381713868,
      "num_tokens": 827592677.0,
      "step": 24705
    },
    {
      "epoch": 1.4750477554918815,
      "grad_norm": 0.5441104769706726,
      "learning_rate": 1.2958980722851078e-05,
      "loss": 0.4685,
      "mean_token_accuracy": 0.8445842742919922,
      "num_tokens": 827760357.0,
      "step": 24710
    },
    {
      "epoch": 1.4753462273161413,
      "grad_norm": 0.5230847001075745,
      "learning_rate": 1.2950508645107168e-05,
      "loss": 0.4395,
      "mean_token_accuracy": 0.851980197429657,
      "num_tokens": 827928037.0,
      "step": 24715
    },
    {
      "epoch": 1.475644699140401,
      "grad_norm": 0.5109076499938965,
      "learning_rate": 1.2942040111160433e-05,
      "loss": 0.4625,
      "mean_token_accuracy": 0.8463080048561096,
      "num_tokens": 828095717.0,
      "step": 24720
    },
    {
      "epoch": 1.4759431709646609,
      "grad_norm": 0.4509248733520508,
      "learning_rate": 1.2933575123073532e-05,
      "loss": 0.4547,
      "mean_token_accuracy": 0.8495705604553223,
      "num_tokens": 828263397.0,
      "step": 24725
    },
    {
      "epoch": 1.4762416427889207,
      "grad_norm": 0.4846537709236145,
      "learning_rate": 1.2925113682908287e-05,
      "loss": 0.4272,
      "mean_token_accuracy": 0.8562626719474793,
      "num_tokens": 828431077.0,
      "step": 24730
    },
    {
      "epoch": 1.4765401146131805,
      "grad_norm": 0.5724625587463379,
      "learning_rate": 1.2916655792725621e-05,
      "loss": 0.4549,
      "mean_token_accuracy": 0.8486401081085205,
      "num_tokens": 828598757.0,
      "step": 24735
    },
    {
      "epoch": 1.4768385864374403,
      "grad_norm": 0.5184370875358582,
      "learning_rate": 1.2908201454585602e-05,
      "loss": 0.4462,
      "mean_token_accuracy": 0.8523917555809021,
      "num_tokens": 828766437.0,
      "step": 24740
    },
    {
      "epoch": 1.4771370582617,
      "grad_norm": 0.5459449291229248,
      "learning_rate": 1.2899750670547473e-05,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8369617104530335,
      "num_tokens": 828934117.0,
      "step": 24745
    },
    {
      "epoch": 1.4774355300859598,
      "grad_norm": 0.49026742577552795,
      "learning_rate": 1.2891303442669545e-05,
      "loss": 0.4311,
      "mean_token_accuracy": 0.8568054437637329,
      "num_tokens": 829101797.0,
      "step": 24750
    },
    {
      "epoch": 1.4777340019102196,
      "grad_norm": 0.4901483654975891,
      "learning_rate": 1.2882859773009314e-05,
      "loss": 0.441,
      "mean_token_accuracy": 0.8519622921943665,
      "num_tokens": 829269477.0,
      "step": 24755
    },
    {
      "epoch": 1.4780324737344794,
      "grad_norm": 0.5515833497047424,
      "learning_rate": 1.2874419663623377e-05,
      "loss": 0.4712,
      "mean_token_accuracy": 0.8429500222206116,
      "num_tokens": 829437157.0,
      "step": 24760
    },
    {
      "epoch": 1.4783309455587392,
      "grad_norm": 0.5111850500106812,
      "learning_rate": 1.2865983116567499e-05,
      "loss": 0.478,
      "mean_token_accuracy": 0.8412501335144043,
      "num_tokens": 829604837.0,
      "step": 24765
    },
    {
      "epoch": 1.478629417382999,
      "grad_norm": 0.5199674963951111,
      "learning_rate": 1.2857550133896548e-05,
      "loss": 0.4415,
      "mean_token_accuracy": 0.8518072366714478,
      "num_tokens": 829772517.0,
      "step": 24770
    },
    {
      "epoch": 1.4789278892072588,
      "grad_norm": 0.5170202851295471,
      "learning_rate": 1.2849120717664526e-05,
      "loss": 0.4446,
      "mean_token_accuracy": 0.8533937931060791,
      "num_tokens": 829940197.0,
      "step": 24775
    },
    {
      "epoch": 1.4792263610315186,
      "grad_norm": 0.48555731773376465,
      "learning_rate": 1.2840694869924574e-05,
      "loss": 0.4068,
      "mean_token_accuracy": 0.8628414750099183,
      "num_tokens": 830107877.0,
      "step": 24780
    },
    {
      "epoch": 1.4795248328557784,
      "grad_norm": 0.5514911413192749,
      "learning_rate": 1.2832272592728966e-05,
      "loss": 0.4159,
      "mean_token_accuracy": 0.8612310647964477,
      "num_tokens": 830275557.0,
      "step": 24785
    },
    {
      "epoch": 1.4798233046800382,
      "grad_norm": 0.4827394485473633,
      "learning_rate": 1.2823853888129116e-05,
      "loss": 0.4425,
      "mean_token_accuracy": 0.8533102750778199,
      "num_tokens": 830443237.0,
      "step": 24790
    },
    {
      "epoch": 1.480121776504298,
      "grad_norm": 0.44826555252075195,
      "learning_rate": 1.281543875817553e-05,
      "loss": 0.4234,
      "mean_token_accuracy": 0.8591554403305054,
      "num_tokens": 830610917.0,
      "step": 24795
    },
    {
      "epoch": 1.4804202483285578,
      "grad_norm": 0.49727168679237366,
      "learning_rate": 1.2807027204917877e-05,
      "loss": 0.4273,
      "mean_token_accuracy": 0.8564177513122558,
      "num_tokens": 830778597.0,
      "step": 24800
    },
    {
      "epoch": 1.4807187201528176,
      "grad_norm": 0.5396794080734253,
      "learning_rate": 1.2798619230404962e-05,
      "loss": 0.4204,
      "mean_token_accuracy": 0.8594834923744201,
      "num_tokens": 830946277.0,
      "step": 24805
    },
    {
      "epoch": 1.4810171919770774,
      "grad_norm": 0.5203003287315369,
      "learning_rate": 1.279021483668468e-05,
      "loss": 0.4344,
      "mean_token_accuracy": 0.8545449018478394,
      "num_tokens": 831113957.0,
      "step": 24810
    },
    {
      "epoch": 1.4813156638013372,
      "grad_norm": 0.5016398429870605,
      "learning_rate": 1.2781814025804095e-05,
      "loss": 0.4268,
      "mean_token_accuracy": 0.8572229504585266,
      "num_tokens": 831281637.0,
      "step": 24815
    },
    {
      "epoch": 1.481614135625597,
      "grad_norm": 0.5529849529266357,
      "learning_rate": 1.2773416799809357e-05,
      "loss": 0.4443,
      "mean_token_accuracy": 0.8507992386817932,
      "num_tokens": 831449317.0,
      "step": 24820
    },
    {
      "epoch": 1.4819126074498568,
      "grad_norm": 0.8989988565444946,
      "learning_rate": 1.2765023160745788e-05,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8524036765098572,
      "num_tokens": 831616997.0,
      "step": 24825
    },
    {
      "epoch": 1.4822110792741165,
      "grad_norm": 0.5164071917533875,
      "learning_rate": 1.2756633110657807e-05,
      "loss": 0.4271,
      "mean_token_accuracy": 0.8566086292266846,
      "num_tokens": 831784677.0,
      "step": 24830
    },
    {
      "epoch": 1.4825095510983763,
      "grad_norm": 0.4770441949367523,
      "learning_rate": 1.2748246651588963e-05,
      "loss": 0.405,
      "mean_token_accuracy": 0.8625909566879273,
      "num_tokens": 831952357.0,
      "step": 24835
    },
    {
      "epoch": 1.4828080229226361,
      "grad_norm": 0.547262966632843,
      "learning_rate": 1.2739863785581924e-05,
      "loss": 0.4602,
      "mean_token_accuracy": 0.846051549911499,
      "num_tokens": 832120037.0,
      "step": 24840
    },
    {
      "epoch": 1.483106494746896,
      "grad_norm": 0.5147694945335388,
      "learning_rate": 1.2731484514678508e-05,
      "loss": 0.4019,
      "mean_token_accuracy": 0.8648813009262085,
      "num_tokens": 832287717.0,
      "step": 24845
    },
    {
      "epoch": 1.4834049665711557,
      "grad_norm": 0.47269514203071594,
      "learning_rate": 1.2723108840919631e-05,
      "loss": 0.4369,
      "mean_token_accuracy": 0.853847074508667,
      "num_tokens": 832455397.0,
      "step": 24850
    },
    {
      "epoch": 1.4837034383954155,
      "grad_norm": 0.4831273555755615,
      "learning_rate": 1.2714736766345357e-05,
      "loss": 0.4261,
      "mean_token_accuracy": 0.8571155905723572,
      "num_tokens": 832623077.0,
      "step": 24855
    },
    {
      "epoch": 1.4840019102196753,
      "grad_norm": 0.5041881799697876,
      "learning_rate": 1.2706368292994836e-05,
      "loss": 0.4378,
      "mean_token_accuracy": 0.8522306919097901,
      "num_tokens": 832790757.0,
      "step": 24860
    },
    {
      "epoch": 1.484300382043935,
      "grad_norm": 0.5632654428482056,
      "learning_rate": 1.2698003422906379e-05,
      "loss": 0.4302,
      "mean_token_accuracy": 0.8571871757507324,
      "num_tokens": 832958437.0,
      "step": 24865
    },
    {
      "epoch": 1.484598853868195,
      "grad_norm": 0.49671268463134766,
      "learning_rate": 1.2689642158117403e-05,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8494810938835144,
      "num_tokens": 833126117.0,
      "step": 24870
    },
    {
      "epoch": 1.4848973256924547,
      "grad_norm": 0.5024557709693909,
      "learning_rate": 1.2681284500664451e-05,
      "loss": 0.4761,
      "mean_token_accuracy": 0.8410294532775879,
      "num_tokens": 833293797.0,
      "step": 24875
    },
    {
      "epoch": 1.4851957975167145,
      "grad_norm": 0.4708639681339264,
      "learning_rate": 1.2672930452583176e-05,
      "loss": 0.4461,
      "mean_token_accuracy": 0.851127278804779,
      "num_tokens": 833461477.0,
      "step": 24880
    },
    {
      "epoch": 1.4854942693409743,
      "grad_norm": 0.500773549079895,
      "learning_rate": 1.2664580015908355e-05,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8556781649589539,
      "num_tokens": 833629157.0,
      "step": 24885
    },
    {
      "epoch": 1.485792741165234,
      "grad_norm": 0.5218284130096436,
      "learning_rate": 1.2656233192673922e-05,
      "loss": 0.3952,
      "mean_token_accuracy": 0.8665692567825317,
      "num_tokens": 833796837.0,
      "step": 24890
    },
    {
      "epoch": 1.4860912129894939,
      "grad_norm": 0.515834629535675,
      "learning_rate": 1.264788998491287e-05,
      "loss": 0.4582,
      "mean_token_accuracy": 0.8470476031303406,
      "num_tokens": 833964517.0,
      "step": 24895
    },
    {
      "epoch": 1.4863896848137537,
      "grad_norm": 0.5145506858825684,
      "learning_rate": 1.2639550394657348e-05,
      "loss": 0.4147,
      "mean_token_accuracy": 0.8619706630706787,
      "num_tokens": 834132197.0,
      "step": 24900
    },
    {
      "epoch": 1.4866881566380132,
      "grad_norm": 0.4744322896003723,
      "learning_rate": 1.2631214423938623e-05,
      "loss": 0.4444,
      "mean_token_accuracy": 0.8522366881370544,
      "num_tokens": 834299877.0,
      "step": 24905
    },
    {
      "epoch": 1.4869866284622733,
      "grad_norm": 0.48897868394851685,
      "learning_rate": 1.2622882074787071e-05,
      "loss": 0.4539,
      "mean_token_accuracy": 0.8466599106788635,
      "num_tokens": 834467557.0,
      "step": 24910
    },
    {
      "epoch": 1.4872851002865328,
      "grad_norm": 0.503420889377594,
      "learning_rate": 1.2614553349232197e-05,
      "loss": 0.4317,
      "mean_token_accuracy": 0.8563402175903321,
      "num_tokens": 834635237.0,
      "step": 24915
    },
    {
      "epoch": 1.4875835721107928,
      "grad_norm": 0.4867260158061981,
      "learning_rate": 1.2606228249302599e-05,
      "loss": 0.3895,
      "mean_token_accuracy": 0.8672491908073425,
      "num_tokens": 834802917.0,
      "step": 24920
    },
    {
      "epoch": 1.4878820439350524,
      "grad_norm": 0.49484288692474365,
      "learning_rate": 1.2597906777026023e-05,
      "loss": 0.4075,
      "mean_token_accuracy": 0.8627639412879944,
      "num_tokens": 834970597.0,
      "step": 24925
    },
    {
      "epoch": 1.4881805157593124,
      "grad_norm": 0.4858594536781311,
      "learning_rate": 1.2589588934429308e-05,
      "loss": 0.4188,
      "mean_token_accuracy": 0.8597041606903076,
      "num_tokens": 835138277.0,
      "step": 24930
    },
    {
      "epoch": 1.488478987583572,
      "grad_norm": 0.5712736248970032,
      "learning_rate": 1.2581274723538422e-05,
      "loss": 0.465,
      "mean_token_accuracy": 0.8465823769569397,
      "num_tokens": 835305957.0,
      "step": 24935
    },
    {
      "epoch": 1.488777459407832,
      "grad_norm": 0.52598637342453,
      "learning_rate": 1.2572964146378457e-05,
      "loss": 0.4273,
      "mean_token_accuracy": 0.8549862742424011,
      "num_tokens": 835473637.0,
      "step": 24940
    },
    {
      "epoch": 1.4890759312320916,
      "grad_norm": 0.47138312458992004,
      "learning_rate": 1.2564657204973578e-05,
      "loss": 0.4562,
      "mean_token_accuracy": 0.8479541897773742,
      "num_tokens": 835641317.0,
      "step": 24945
    },
    {
      "epoch": 1.4893744030563516,
      "grad_norm": 0.49905580282211304,
      "learning_rate": 1.255635390134713e-05,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8495526790618897,
      "num_tokens": 835808997.0,
      "step": 24950
    },
    {
      "epoch": 1.4896728748806112,
      "grad_norm": 0.5662775039672852,
      "learning_rate": 1.2548054237521506e-05,
      "loss": 0.461,
      "mean_token_accuracy": 0.8471012711524963,
      "num_tokens": 835976677.0,
      "step": 24955
    },
    {
      "epoch": 1.4899713467048712,
      "grad_norm": 0.5363903045654297,
      "learning_rate": 1.2539758215518255e-05,
      "loss": 0.4429,
      "mean_token_accuracy": 0.8501908779144287,
      "num_tokens": 836144357.0,
      "step": 24960
    },
    {
      "epoch": 1.4902698185291308,
      "grad_norm": 0.5376128554344177,
      "learning_rate": 1.2531465837358025e-05,
      "loss": 0.4473,
      "mean_token_accuracy": 0.8515567183494568,
      "num_tokens": 836312037.0,
      "step": 24965
    },
    {
      "epoch": 1.4905682903533906,
      "grad_norm": 0.4691319167613983,
      "learning_rate": 1.2523177105060577e-05,
      "loss": 0.4411,
      "mean_token_accuracy": 0.8522486090660095,
      "num_tokens": 836479717.0,
      "step": 24970
    },
    {
      "epoch": 1.4908667621776504,
      "grad_norm": 0.4661787450313568,
      "learning_rate": 1.2514892020644798e-05,
      "loss": 0.4675,
      "mean_token_accuracy": 0.8460515379905701,
      "num_tokens": 836647397.0,
      "step": 24975
    },
    {
      "epoch": 1.4911652340019101,
      "grad_norm": 0.5125504732131958,
      "learning_rate": 1.2506610586128648e-05,
      "loss": 0.4149,
      "mean_token_accuracy": 0.8614278912544251,
      "num_tokens": 836815077.0,
      "step": 24980
    },
    {
      "epoch": 1.49146370582617,
      "grad_norm": 0.48773860931396484,
      "learning_rate": 1.2498332803529236e-05,
      "loss": 0.4099,
      "mean_token_accuracy": 0.8640820622444153,
      "num_tokens": 836982757.0,
      "step": 24985
    },
    {
      "epoch": 1.4917621776504297,
      "grad_norm": 0.5541853904724121,
      "learning_rate": 1.249005867486277e-05,
      "loss": 0.4506,
      "mean_token_accuracy": 0.8504711866378785,
      "num_tokens": 837150437.0,
      "step": 24990
    },
    {
      "epoch": 1.4920606494746895,
      "grad_norm": 0.5276203155517578,
      "learning_rate": 1.2481788202144565e-05,
      "loss": 0.4381,
      "mean_token_accuracy": 0.85422283411026,
      "num_tokens": 837318117.0,
      "step": 24995
    },
    {
      "epoch": 1.4923591212989493,
      "grad_norm": 0.4703633785247803,
      "learning_rate": 1.2473521387389056e-05,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8546284198760986,
      "num_tokens": 837485797.0,
      "step": 25000
    },
    {
      "epoch": 1.4926575931232091,
      "grad_norm": 0.5033524036407471,
      "learning_rate": 1.2465258232609761e-05,
      "loss": 0.4217,
      "mean_token_accuracy": 0.8586842536926269,
      "num_tokens": 837653477.0,
      "step": 25005
    },
    {
      "epoch": 1.492956064947469,
      "grad_norm": 0.5081316828727722,
      "learning_rate": 1.2456998739819336e-05,
      "loss": 0.451,
      "mean_token_accuracy": 0.8490158557891846,
      "num_tokens": 837821157.0,
      "step": 25010
    },
    {
      "epoch": 1.4932545367717287,
      "grad_norm": 0.48394495248794556,
      "learning_rate": 1.2448742911029527e-05,
      "loss": 0.4494,
      "mean_token_accuracy": 0.8493916153907776,
      "num_tokens": 837988837.0,
      "step": 25015
    },
    {
      "epoch": 1.4935530085959885,
      "grad_norm": 0.49527519941329956,
      "learning_rate": 1.2440490748251197e-05,
      "loss": 0.4335,
      "mean_token_accuracy": 0.8559107661247254,
      "num_tokens": 838156517.0,
      "step": 25020
    },
    {
      "epoch": 1.4938514804202483,
      "grad_norm": 0.4975765347480774,
      "learning_rate": 1.2432242253494314e-05,
      "loss": 0.4386,
      "mean_token_accuracy": 0.8532029151916504,
      "num_tokens": 838324197.0,
      "step": 25025
    },
    {
      "epoch": 1.494149952244508,
      "grad_norm": 0.48721006512641907,
      "learning_rate": 1.2423997428767946e-05,
      "loss": 0.4118,
      "mean_token_accuracy": 0.8605153322219848,
      "num_tokens": 838491877.0,
      "step": 25030
    },
    {
      "epoch": 1.4944484240687679,
      "grad_norm": 0.535171627998352,
      "learning_rate": 1.2415756276080286e-05,
      "loss": 0.4664,
      "mean_token_accuracy": 0.8455982446670532,
      "num_tokens": 838659557.0,
      "step": 25035
    },
    {
      "epoch": 1.4947468958930277,
      "grad_norm": 0.47328364849090576,
      "learning_rate": 1.2407518797438598e-05,
      "loss": 0.4655,
      "mean_token_accuracy": 0.84479900598526,
      "num_tokens": 838827237.0,
      "step": 25040
    },
    {
      "epoch": 1.4950453677172875,
      "grad_norm": 0.4900859296321869,
      "learning_rate": 1.2399284994849285e-05,
      "loss": 0.4264,
      "mean_token_accuracy": 0.8561135649681091,
      "num_tokens": 838994917.0,
      "step": 25045
    },
    {
      "epoch": 1.4953438395415473,
      "grad_norm": 0.4779356122016907,
      "learning_rate": 1.2391054870317837e-05,
      "loss": 0.4558,
      "mean_token_accuracy": 0.8495765328407288,
      "num_tokens": 839162597.0,
      "step": 25050
    },
    {
      "epoch": 1.495642311365807,
      "grad_norm": 0.586794912815094,
      "learning_rate": 1.2382828425848854e-05,
      "loss": 0.4418,
      "mean_token_accuracy": 0.851240599155426,
      "num_tokens": 839330277.0,
      "step": 25055
    },
    {
      "epoch": 1.4959407831900668,
      "grad_norm": 0.48590776324272156,
      "learning_rate": 1.2374605663446048e-05,
      "loss": 0.4488,
      "mean_token_accuracy": 0.8508290529251099,
      "num_tokens": 839497957.0,
      "step": 25060
    },
    {
      "epoch": 1.4962392550143266,
      "grad_norm": 0.4813680052757263,
      "learning_rate": 1.2366386585112205e-05,
      "loss": 0.3965,
      "mean_token_accuracy": 0.8661517381668091,
      "num_tokens": 839665637.0,
      "step": 25065
    },
    {
      "epoch": 1.4965377268385864,
      "grad_norm": 0.5295145511627197,
      "learning_rate": 1.235817119284924e-05,
      "loss": 0.4074,
      "mean_token_accuracy": 0.8633424639701843,
      "num_tokens": 839833317.0,
      "step": 25070
    },
    {
      "epoch": 1.4968361986628462,
      "grad_norm": 0.5573604702949524,
      "learning_rate": 1.2349959488658181e-05,
      "loss": 0.4179,
      "mean_token_accuracy": 0.8604318141937256,
      "num_tokens": 840000997.0,
      "step": 25075
    },
    {
      "epoch": 1.497134670487106,
      "grad_norm": 0.4596938490867615,
      "learning_rate": 1.2341751474539115e-05,
      "loss": 0.4127,
      "mean_token_accuracy": 0.8616605043411255,
      "num_tokens": 840168677.0,
      "step": 25080
    },
    {
      "epoch": 1.4974331423113658,
      "grad_norm": 0.4987497627735138,
      "learning_rate": 1.2333547152491279e-05,
      "loss": 0.3955,
      "mean_token_accuracy": 0.8667899370193481,
      "num_tokens": 840336357.0,
      "step": 25085
    },
    {
      "epoch": 1.4977316141356256,
      "grad_norm": 0.48805469274520874,
      "learning_rate": 1.2325346524512957e-05,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8518787980079651,
      "num_tokens": 840504037.0,
      "step": 25090
    },
    {
      "epoch": 1.4980300859598854,
      "grad_norm": 0.5142251253128052,
      "learning_rate": 1.2317149592601596e-05,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8573064565658569,
      "num_tokens": 840671717.0,
      "step": 25095
    },
    {
      "epoch": 1.4983285577841452,
      "grad_norm": 0.4770118296146393,
      "learning_rate": 1.2308956358753688e-05,
      "loss": 0.4255,
      "mean_token_accuracy": 0.8563103914260864,
      "num_tokens": 840839397.0,
      "step": 25100
    },
    {
      "epoch": 1.498627029608405,
      "grad_norm": 0.5360010862350464,
      "learning_rate": 1.2300766824964855e-05,
      "loss": 0.4202,
      "mean_token_accuracy": 0.8601276397705078,
      "num_tokens": 841007077.0,
      "step": 25105
    },
    {
      "epoch": 1.4989255014326648,
      "grad_norm": 0.49961942434310913,
      "learning_rate": 1.2292580993229804e-05,
      "loss": 0.4632,
      "mean_token_accuracy": 0.845789110660553,
      "num_tokens": 841174757.0,
      "step": 25110
    },
    {
      "epoch": 1.4992239732569246,
      "grad_norm": 0.5356742143630981,
      "learning_rate": 1.2284398865542354e-05,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8493523597717285,
      "num_tokens": 841340286.0,
      "step": 25115
    },
    {
      "epoch": 1.4995224450811844,
      "grad_norm": 0.5119374990463257,
      "learning_rate": 1.2276220443895411e-05,
      "loss": 0.454,
      "mean_token_accuracy": 0.8494691729545594,
      "num_tokens": 841507966.0,
      "step": 25120
    },
    {
      "epoch": 1.4998209169054442,
      "grad_norm": 0.46571317315101624,
      "learning_rate": 1.2268045730280973e-05,
      "loss": 0.4276,
      "mean_token_accuracy": 0.8551771521568299,
      "num_tokens": 841675646.0,
      "step": 25125
    },
    {
      "epoch": 1.500119388729704,
      "grad_norm": 0.5265080332756042,
      "learning_rate": 1.2259874726690146e-05,
      "loss": 0.4639,
      "mean_token_accuracy": 0.845685076713562,
      "num_tokens": 841837651.0,
      "step": 25130
    },
    {
      "epoch": 1.5004178605539638,
      "grad_norm": 0.5621362924575806,
      "learning_rate": 1.2251707435113128e-05,
      "loss": 0.4022,
      "mean_token_accuracy": 0.8644578337669373,
      "num_tokens": 842005331.0,
      "step": 25135
    },
    {
      "epoch": 1.5007163323782235,
      "grad_norm": 0.5322445631027222,
      "learning_rate": 1.2243543857539219e-05,
      "loss": 0.455,
      "mean_token_accuracy": 0.8471847653388977,
      "num_tokens": 842173011.0,
      "step": 25140
    },
    {
      "epoch": 1.5010148042024833,
      "grad_norm": 0.5195977091789246,
      "learning_rate": 1.223538399595681e-05,
      "loss": 0.4189,
      "mean_token_accuracy": 0.86031254529953,
      "num_tokens": 842340691.0,
      "step": 25145
    },
    {
      "epoch": 1.501313276026743,
      "grad_norm": 0.4984567165374756,
      "learning_rate": 1.2227227852353367e-05,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8533878087997436,
      "num_tokens": 842508371.0,
      "step": 25150
    },
    {
      "epoch": 1.501611747851003,
      "grad_norm": 0.48683884739875793,
      "learning_rate": 1.221907542871549e-05,
      "loss": 0.4231,
      "mean_token_accuracy": 0.858057975769043,
      "num_tokens": 842676051.0,
      "step": 25155
    },
    {
      "epoch": 1.5019102196752625,
      "grad_norm": 0.5287027359008789,
      "learning_rate": 1.2210926727028856e-05,
      "loss": 0.453,
      "mean_token_accuracy": 0.8500894665718078,
      "num_tokens": 842843731.0,
      "step": 25160
    },
    {
      "epoch": 1.5022086914995225,
      "grad_norm": 0.5262054204940796,
      "learning_rate": 1.2202781749278212e-05,
      "loss": 0.4446,
      "mean_token_accuracy": 0.8515984773635864,
      "num_tokens": 843011411.0,
      "step": 25165
    },
    {
      "epoch": 1.502507163323782,
      "grad_norm": 0.4648466408252716,
      "learning_rate": 1.2194640497447428e-05,
      "loss": 0.4501,
      "mean_token_accuracy": 0.8499105215072632,
      "num_tokens": 843179091.0,
      "step": 25170
    },
    {
      "epoch": 1.502805635148042,
      "grad_norm": 0.4945337772369385,
      "learning_rate": 1.2186502973519457e-05,
      "loss": 0.4141,
      "mean_token_accuracy": 0.8615650773048401,
      "num_tokens": 843346771.0,
      "step": 25175
    },
    {
      "epoch": 1.5031041069723017,
      "grad_norm": 0.5120456218719482,
      "learning_rate": 1.217836917947634e-05,
      "loss": 0.4622,
      "mean_token_accuracy": 0.8446141123771668,
      "num_tokens": 843514451.0,
      "step": 25180
    },
    {
      "epoch": 1.5034025787965617,
      "grad_norm": 0.5191718339920044,
      "learning_rate": 1.2170239117299222e-05,
      "loss": 0.4616,
      "mean_token_accuracy": 0.8467434048652649,
      "num_tokens": 843682131.0,
      "step": 25185
    },
    {
      "epoch": 1.5037010506208213,
      "grad_norm": 0.4831700623035431,
      "learning_rate": 1.2162112788968313e-05,
      "loss": 0.4311,
      "mean_token_accuracy": 0.8565191388130188,
      "num_tokens": 843849811.0,
      "step": 25190
    },
    {
      "epoch": 1.5039995224450813,
      "grad_norm": 0.4708995521068573,
      "learning_rate": 1.2153990196462941e-05,
      "loss": 0.3956,
      "mean_token_accuracy": 0.8654121398925781,
      "num_tokens": 844017491.0,
      "step": 25195
    },
    {
      "epoch": 1.5042979942693409,
      "grad_norm": 0.4848741590976715,
      "learning_rate": 1.2145871341761512e-05,
      "loss": 0.4276,
      "mean_token_accuracy": 0.8563103914260864,
      "num_tokens": 844185171.0,
      "step": 25200
    },
    {
      "epoch": 1.5045964660936009,
      "grad_norm": 0.48994994163513184,
      "learning_rate": 1.2137756226841526e-05,
      "loss": 0.445,
      "mean_token_accuracy": 0.8521054506301879,
      "num_tokens": 844352851.0,
      "step": 25205
    },
    {
      "epoch": 1.5048949379178604,
      "grad_norm": 0.4944668412208557,
      "learning_rate": 1.212964485367956e-05,
      "loss": 0.425,
      "mean_token_accuracy": 0.8572945237159729,
      "num_tokens": 844520531.0,
      "step": 25210
    },
    {
      "epoch": 1.5051934097421205,
      "grad_norm": 0.45361489057540894,
      "learning_rate": 1.2121537224251289e-05,
      "loss": 0.3842,
      "mean_token_accuracy": 0.8697423338890076,
      "num_tokens": 844688211.0,
      "step": 25215
    },
    {
      "epoch": 1.50549188156638,
      "grad_norm": 0.49812015891075134,
      "learning_rate": 1.2113433340531496e-05,
      "loss": 0.4352,
      "mean_token_accuracy": 0.8560658454895019,
      "num_tokens": 844855891.0,
      "step": 25220
    },
    {
      "epoch": 1.50579035339064,
      "grad_norm": 0.4734099209308624,
      "learning_rate": 1.2105333204494005e-05,
      "loss": 0.3861,
      "mean_token_accuracy": 0.8691816687583923,
      "num_tokens": 845023571.0,
      "step": 25225
    },
    {
      "epoch": 1.5060888252148996,
      "grad_norm": 0.5214412808418274,
      "learning_rate": 1.2097236818111778e-05,
      "loss": 0.4605,
      "mean_token_accuracy": 0.8480973482131958,
      "num_tokens": 845191251.0,
      "step": 25230
    },
    {
      "epoch": 1.5063872970391596,
      "grad_norm": 0.5334967970848083,
      "learning_rate": 1.2089144183356805e-05,
      "loss": 0.4634,
      "mean_token_accuracy": 0.8447154879570007,
      "num_tokens": 845358931.0,
      "step": 25235
    },
    {
      "epoch": 1.5066857688634192,
      "grad_norm": 0.4756765365600586,
      "learning_rate": 1.2081055302200231e-05,
      "loss": 0.4139,
      "mean_token_accuracy": 0.8588691473007202,
      "num_tokens": 845526611.0,
      "step": 25240
    },
    {
      "epoch": 1.5069842406876792,
      "grad_norm": 0.5088101625442505,
      "learning_rate": 1.2072970176612245e-05,
      "loss": 0.4477,
      "mean_token_accuracy": 0.8507813453674317,
      "num_tokens": 845694291.0,
      "step": 25245
    },
    {
      "epoch": 1.5072827125119388,
      "grad_norm": 0.4716501832008362,
      "learning_rate": 1.2064888808562114e-05,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8575927495956421,
      "num_tokens": 845861971.0,
      "step": 25250
    },
    {
      "epoch": 1.5075811843361988,
      "grad_norm": 0.5015446543693542,
      "learning_rate": 1.2056811200018214e-05,
      "loss": 0.4239,
      "mean_token_accuracy": 0.8587737083435059,
      "num_tokens": 846029651.0,
      "step": 25255
    },
    {
      "epoch": 1.5078796561604584,
      "grad_norm": 0.49406954646110535,
      "learning_rate": 1.2048737352947995e-05,
      "loss": 0.4001,
      "mean_token_accuracy": 0.8648753523826599,
      "num_tokens": 846197331.0,
      "step": 25260
    },
    {
      "epoch": 1.5081781279847184,
      "grad_norm": 0.499738872051239,
      "learning_rate": 1.2040667269317991e-05,
      "loss": 0.4493,
      "mean_token_accuracy": 0.8507117509841919,
      "num_tokens": 846360857.0,
      "step": 25265
    },
    {
      "epoch": 1.508476599808978,
      "grad_norm": 0.5533084869384766,
      "learning_rate": 1.2032600951093829e-05,
      "loss": 0.4374,
      "mean_token_accuracy": 0.8535309433937073,
      "num_tokens": 846528537.0,
      "step": 25270
    },
    {
      "epoch": 1.508775071633238,
      "grad_norm": 0.5204278826713562,
      "learning_rate": 1.2024538400240185e-05,
      "loss": 0.4318,
      "mean_token_accuracy": 0.8561195254325866,
      "num_tokens": 846696217.0,
      "step": 25275
    },
    {
      "epoch": 1.5090735434574976,
      "grad_norm": 0.5286776423454285,
      "learning_rate": 1.2016479618720873e-05,
      "loss": 0.4405,
      "mean_token_accuracy": 0.8546880483627319,
      "num_tokens": 846863897.0,
      "step": 25280
    },
    {
      "epoch": 1.5093720152817574,
      "grad_norm": 0.5254387259483337,
      "learning_rate": 1.2008424608498733e-05,
      "loss": 0.4409,
      "mean_token_accuracy": 0.8507753729820251,
      "num_tokens": 847031577.0,
      "step": 25285
    },
    {
      "epoch": 1.5096704871060171,
      "grad_norm": 0.5032444000244141,
      "learning_rate": 1.2000373371535722e-05,
      "loss": 0.396,
      "mean_token_accuracy": 0.8657580852508545,
      "num_tokens": 847199257.0,
      "step": 25290
    },
    {
      "epoch": 1.509968958930277,
      "grad_norm": 0.477531373500824,
      "learning_rate": 1.199232590979287e-05,
      "loss": 0.4391,
      "mean_token_accuracy": 0.8542645931243896,
      "num_tokens": 847366937.0,
      "step": 25295
    },
    {
      "epoch": 1.5102674307545367,
      "grad_norm": 0.6035160422325134,
      "learning_rate": 1.198428222523028e-05,
      "loss": 0.4664,
      "mean_token_accuracy": 0.8442025542259216,
      "num_tokens": 847534617.0,
      "step": 25300
    },
    {
      "epoch": 1.5105659025787965,
      "grad_norm": 0.4623118042945862,
      "learning_rate": 1.1976242319807146e-05,
      "loss": 0.4339,
      "mean_token_accuracy": 0.8555767655372619,
      "num_tokens": 847702297.0,
      "step": 25305
    },
    {
      "epoch": 1.5108643744030563,
      "grad_norm": 0.508072018623352,
      "learning_rate": 1.1968206195481728e-05,
      "loss": 0.4473,
      "mean_token_accuracy": 0.8505487203598022,
      "num_tokens": 847869977.0,
      "step": 25310
    },
    {
      "epoch": 1.5111628462273161,
      "grad_norm": 0.5197190046310425,
      "learning_rate": 1.196017385421137e-05,
      "loss": 0.4276,
      "mean_token_accuracy": 0.8578790426254272,
      "num_tokens": 848037657.0,
      "step": 25315
    },
    {
      "epoch": 1.511461318051576,
      "grad_norm": 0.5564250349998474,
      "learning_rate": 1.1952145297952506e-05,
      "loss": 0.4551,
      "mean_token_accuracy": 0.8482047080993652,
      "num_tokens": 848205337.0,
      "step": 25320
    },
    {
      "epoch": 1.5117597898758357,
      "grad_norm": 0.5371652841567993,
      "learning_rate": 1.1944120528660633e-05,
      "loss": 0.4495,
      "mean_token_accuracy": 0.8491649866104126,
      "num_tokens": 848373017.0,
      "step": 25325
    },
    {
      "epoch": 1.5120582617000955,
      "grad_norm": 0.5175400972366333,
      "learning_rate": 1.1936099548290346e-05,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8541810870170593,
      "num_tokens": 848540697.0,
      "step": 25330
    },
    {
      "epoch": 1.5123567335243553,
      "grad_norm": 0.5128393173217773,
      "learning_rate": 1.1928082358795278e-05,
      "loss": 0.4109,
      "mean_token_accuracy": 0.8617618918418884,
      "num_tokens": 848708377.0,
      "step": 25335
    },
    {
      "epoch": 1.512655205348615,
      "grad_norm": 0.5576654076576233,
      "learning_rate": 1.1920068962128176e-05,
      "loss": 0.4522,
      "mean_token_accuracy": 0.849135160446167,
      "num_tokens": 848876057.0,
      "step": 25340
    },
    {
      "epoch": 1.5129536771728749,
      "grad_norm": 0.5023189783096313,
      "learning_rate": 1.1912059360240854e-05,
      "loss": 0.4043,
      "mean_token_accuracy": 0.8634021282196045,
      "num_tokens": 849043737.0,
      "step": 25345
    },
    {
      "epoch": 1.5132521489971347,
      "grad_norm": 0.5152266621589661,
      "learning_rate": 1.1904053555084194e-05,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8420374512672424,
      "num_tokens": 849211417.0,
      "step": 25350
    },
    {
      "epoch": 1.5135506208213945,
      "grad_norm": 0.586997389793396,
      "learning_rate": 1.1896051548608167e-05,
      "loss": 0.456,
      "mean_token_accuracy": 0.8480019092559814,
      "num_tokens": 849379097.0,
      "step": 25355
    },
    {
      "epoch": 1.5138490926456543,
      "grad_norm": 0.4987232983112335,
      "learning_rate": 1.1888053342761784e-05,
      "loss": 0.4025,
      "mean_token_accuracy": 0.8653107404708862,
      "num_tokens": 849546777.0,
      "step": 25360
    },
    {
      "epoch": 1.514147564469914,
      "grad_norm": 0.5013066530227661,
      "learning_rate": 1.188005893949319e-05,
      "loss": 0.4444,
      "mean_token_accuracy": 0.8521412372589111,
      "num_tokens": 849714457.0,
      "step": 25365
    },
    {
      "epoch": 1.5144460362941738,
      "grad_norm": 0.5315134525299072,
      "learning_rate": 1.1872068340749544e-05,
      "loss": 0.4248,
      "mean_token_accuracy": 0.857079803943634,
      "num_tokens": 849882137.0,
      "step": 25370
    },
    {
      "epoch": 1.5147445081184336,
      "grad_norm": 0.4681101143360138,
      "learning_rate": 1.1864081548477113e-05,
      "loss": 0.4298,
      "mean_token_accuracy": 0.856214964389801,
      "num_tokens": 850049817.0,
      "step": 25375
    },
    {
      "epoch": 1.5150429799426934,
      "grad_norm": 0.4866337180137634,
      "learning_rate": 1.1856098564621227e-05,
      "loss": 0.4203,
      "mean_token_accuracy": 0.858809494972229,
      "num_tokens": 850217497.0,
      "step": 25380
    },
    {
      "epoch": 1.5153414517669532,
      "grad_norm": 0.4581342041492462,
      "learning_rate": 1.1848119391126287e-05,
      "loss": 0.4043,
      "mean_token_accuracy": 0.8650721549987793,
      "num_tokens": 850385177.0,
      "step": 25385
    },
    {
      "epoch": 1.515639923591213,
      "grad_norm": 0.5076090097427368,
      "learning_rate": 1.184014402993578e-05,
      "loss": 0.392,
      "mean_token_accuracy": 0.8668555498123169,
      "num_tokens": 850552857.0,
      "step": 25390
    },
    {
      "epoch": 1.5159383954154728,
      "grad_norm": 0.5384923219680786,
      "learning_rate": 1.1832172482992234e-05,
      "loss": 0.4536,
      "mean_token_accuracy": 0.8473696827888488,
      "num_tokens": 850720537.0,
      "step": 25395
    },
    {
      "epoch": 1.5162368672397326,
      "grad_norm": 0.5698250532150269,
      "learning_rate": 1.1824204752237265e-05,
      "loss": 0.4474,
      "mean_token_accuracy": 0.8503936529159546,
      "num_tokens": 850888217.0,
      "step": 25400
    },
    {
      "epoch": 1.5165353390639924,
      "grad_norm": 0.5221206545829773,
      "learning_rate": 1.181624083961159e-05,
      "loss": 0.4284,
      "mean_token_accuracy": 0.8569962978363037,
      "num_tokens": 851055897.0,
      "step": 25405
    },
    {
      "epoch": 1.5168338108882522,
      "grad_norm": 0.5178723931312561,
      "learning_rate": 1.1808280747054933e-05,
      "loss": 0.4252,
      "mean_token_accuracy": 0.8568114042282104,
      "num_tokens": 851223577.0,
      "step": 25410
    },
    {
      "epoch": 1.517132282712512,
      "grad_norm": 0.5310178995132446,
      "learning_rate": 1.1800324476506149e-05,
      "loss": 0.449,
      "mean_token_accuracy": 0.8513897299766541,
      "num_tokens": 851391257.0,
      "step": 25415
    },
    {
      "epoch": 1.5174307545367718,
      "grad_norm": 0.5094597935676575,
      "learning_rate": 1.1792372029903107e-05,
      "loss": 0.4548,
      "mean_token_accuracy": 0.848258376121521,
      "num_tokens": 851558937.0,
      "step": 25420
    },
    {
      "epoch": 1.5177292263610314,
      "grad_norm": 0.4514775574207306,
      "learning_rate": 1.17844234091828e-05,
      "loss": 0.3802,
      "mean_token_accuracy": 0.8704043984413147,
      "num_tokens": 851726617.0,
      "step": 25425
    },
    {
      "epoch": 1.5180276981852914,
      "grad_norm": 0.4997726082801819,
      "learning_rate": 1.177647861628124e-05,
      "loss": 0.4231,
      "mean_token_accuracy": 0.8594954133033752,
      "num_tokens": 851894297.0,
      "step": 25430
    },
    {
      "epoch": 1.518326170009551,
      "grad_norm": 0.5539298057556152,
      "learning_rate": 1.1768537653133537e-05,
      "loss": 0.4642,
      "mean_token_accuracy": 0.8446081280708313,
      "num_tokens": 852061977.0,
      "step": 25435
    },
    {
      "epoch": 1.518624641833811,
      "grad_norm": 0.5309528112411499,
      "learning_rate": 1.1760600521673856e-05,
      "loss": 0.4242,
      "mean_token_accuracy": 0.8588572144508362,
      "num_tokens": 852229657.0,
      "step": 25440
    },
    {
      "epoch": 1.5189231136580705,
      "grad_norm": 0.5359094142913818,
      "learning_rate": 1.1752667223835437e-05,
      "loss": 0.4821,
      "mean_token_accuracy": 0.8404091596603394,
      "num_tokens": 852397337.0,
      "step": 25445
    },
    {
      "epoch": 1.5192215854823305,
      "grad_norm": 0.5146046876907349,
      "learning_rate": 1.1744737761550585e-05,
      "loss": 0.4429,
      "mean_token_accuracy": 0.8515865325927734,
      "num_tokens": 852565017.0,
      "step": 25450
    },
    {
      "epoch": 1.5195200573065901,
      "grad_norm": 0.4770878553390503,
      "learning_rate": 1.1736812136750649e-05,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8659608602523804,
      "num_tokens": 852732697.0,
      "step": 25455
    },
    {
      "epoch": 1.5198185291308501,
      "grad_norm": 0.47861650586128235,
      "learning_rate": 1.1728890351366077e-05,
      "loss": 0.4437,
      "mean_token_accuracy": 0.8520279049873352,
      "num_tokens": 852900377.0,
      "step": 25460
    },
    {
      "epoch": 1.5201170009551097,
      "grad_norm": 0.4809304177761078,
      "learning_rate": 1.1720972407326359e-05,
      "loss": 0.4218,
      "mean_token_accuracy": 0.8567756175994873,
      "num_tokens": 853068057.0,
      "step": 25465
    },
    {
      "epoch": 1.5204154727793697,
      "grad_norm": 0.5149447321891785,
      "learning_rate": 1.1713058306560056e-05,
      "loss": 0.43,
      "mean_token_accuracy": 0.8563581109046936,
      "num_tokens": 853235737.0,
      "step": 25470
    },
    {
      "epoch": 1.5207139446036293,
      "grad_norm": 0.48470592498779297,
      "learning_rate": 1.170514805099481e-05,
      "loss": 0.448,
      "mean_token_accuracy": 0.8501848936080932,
      "num_tokens": 853403417.0,
      "step": 25475
    },
    {
      "epoch": 1.5210124164278893,
      "grad_norm": 0.5272351503372192,
      "learning_rate": 1.1697241642557275e-05,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8430275559425354,
      "num_tokens": 853571097.0,
      "step": 25480
    },
    {
      "epoch": 1.5213108882521489,
      "grad_norm": 0.5201582908630371,
      "learning_rate": 1.1689339083173235e-05,
      "loss": 0.445,
      "mean_token_accuracy": 0.8512048244476318,
      "num_tokens": 853738777.0,
      "step": 25485
    },
    {
      "epoch": 1.521609360076409,
      "grad_norm": 0.5304452180862427,
      "learning_rate": 1.1681440374767497e-05,
      "loss": 0.4667,
      "mean_token_accuracy": 0.8431408882141114,
      "num_tokens": 853906457.0,
      "step": 25490
    },
    {
      "epoch": 1.5219078319006685,
      "grad_norm": 0.4628830850124359,
      "learning_rate": 1.1673545519263927e-05,
      "loss": 0.4317,
      "mean_token_accuracy": 0.8550936460494996,
      "num_tokens": 854074137.0,
      "step": 25495
    },
    {
      "epoch": 1.5222063037249285,
      "grad_norm": 0.5660292506217957,
      "learning_rate": 1.1665654518585468e-05,
      "loss": 0.4622,
      "mean_token_accuracy": 0.8458248853683472,
      "num_tokens": 854241817.0,
      "step": 25500
    },
    {
      "epoch": 1.522504775549188,
      "grad_norm": 0.5355969667434692,
      "learning_rate": 1.1657767374654118e-05,
      "loss": 0.4106,
      "mean_token_accuracy": 0.8627042770385742,
      "num_tokens": 854409497.0,
      "step": 25505
    },
    {
      "epoch": 1.522803247373448,
      "grad_norm": 0.5886985659599304,
      "learning_rate": 1.164988408939094e-05,
      "loss": 0.4658,
      "mean_token_accuracy": 0.8455147385597229,
      "num_tokens": 854577177.0,
      "step": 25510
    },
    {
      "epoch": 1.5231017191977076,
      "grad_norm": 0.5814493298530579,
      "learning_rate": 1.164200466471606e-05,
      "loss": 0.3958,
      "mean_token_accuracy": 0.8682237386703491,
      "num_tokens": 854739815.0,
      "step": 25515
    },
    {
      "epoch": 1.5234001910219677,
      "grad_norm": 0.4609314501285553,
      "learning_rate": 1.1634129102548641e-05,
      "loss": 0.4299,
      "mean_token_accuracy": 0.8565251111984253,
      "num_tokens": 854907495.0,
      "step": 25520
    },
    {
      "epoch": 1.5236986628462272,
      "grad_norm": 0.48527318239212036,
      "learning_rate": 1.1626257404806929e-05,
      "loss": 0.4054,
      "mean_token_accuracy": 0.8627758622169495,
      "num_tokens": 855075175.0,
      "step": 25525
    },
    {
      "epoch": 1.5239971346704873,
      "grad_norm": 0.5244532227516174,
      "learning_rate": 1.161838957340822e-05,
      "loss": 0.4487,
      "mean_token_accuracy": 0.8498926281929016,
      "num_tokens": 855242855.0,
      "step": 25530
    },
    {
      "epoch": 1.5242956064947468,
      "grad_norm": 0.5301193594932556,
      "learning_rate": 1.1610525610268882e-05,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8554813265800476,
      "num_tokens": 855410535.0,
      "step": 25535
    },
    {
      "epoch": 1.5245940783190068,
      "grad_norm": 0.5758376717567444,
      "learning_rate": 1.1602665517304308e-05,
      "loss": 0.4658,
      "mean_token_accuracy": 0.8455624461174012,
      "num_tokens": 855578215.0,
      "step": 25540
    },
    {
      "epoch": 1.5248925501432664,
      "grad_norm": 0.47314441204071045,
      "learning_rate": 1.159480929642897e-05,
      "loss": 0.4238,
      "mean_token_accuracy": 0.8574615359306336,
      "num_tokens": 855745895.0,
      "step": 25545
    },
    {
      "epoch": 1.5251910219675264,
      "grad_norm": 0.5664907693862915,
      "learning_rate": 1.1586956949556422e-05,
      "loss": 0.45,
      "mean_token_accuracy": 0.8488965630531311,
      "num_tokens": 855913575.0,
      "step": 25550
    },
    {
      "epoch": 1.525489493791786,
      "grad_norm": 0.4657769799232483,
      "learning_rate": 1.157910847859922e-05,
      "loss": 0.4174,
      "mean_token_accuracy": 0.861177396774292,
      "num_tokens": 856081255.0,
      "step": 25555
    },
    {
      "epoch": 1.5257879656160458,
      "grad_norm": 0.5686755180358887,
      "learning_rate": 1.1571263885469021e-05,
      "loss": 0.4793,
      "mean_token_accuracy": 0.8419002890586853,
      "num_tokens": 856248935.0,
      "step": 25560
    },
    {
      "epoch": 1.5260864374403056,
      "grad_norm": 0.5609865784645081,
      "learning_rate": 1.1563423172076501e-05,
      "loss": 0.46,
      "mean_token_accuracy": 0.8468090176582337,
      "num_tokens": 856416615.0,
      "step": 25565
    },
    {
      "epoch": 1.5263849092645654,
      "grad_norm": 0.5525285601615906,
      "learning_rate": 1.1555586340331431e-05,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8541989684104919,
      "num_tokens": 856584295.0,
      "step": 25570
    },
    {
      "epoch": 1.5266833810888252,
      "grad_norm": 0.4844779968261719,
      "learning_rate": 1.1547753392142613e-05,
      "loss": 0.4224,
      "mean_token_accuracy": 0.8589230298995971,
      "num_tokens": 856748145.0,
      "step": 25575
    },
    {
      "epoch": 1.526981852913085,
      "grad_norm": 0.45421233773231506,
      "learning_rate": 1.1539924329417895e-05,
      "loss": 0.4269,
      "mean_token_accuracy": 0.8569724440574646,
      "num_tokens": 856915825.0,
      "step": 25580
    },
    {
      "epoch": 1.5272803247373448,
      "grad_norm": 0.5185850262641907,
      "learning_rate": 1.1532099154064197e-05,
      "loss": 0.4113,
      "mean_token_accuracy": 0.8594715476036072,
      "num_tokens": 857083505.0,
      "step": 25585
    },
    {
      "epoch": 1.5275787965616046,
      "grad_norm": 0.463449627161026,
      "learning_rate": 1.1524277867987481e-05,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8512883305549621,
      "num_tokens": 857251185.0,
      "step": 25590
    },
    {
      "epoch": 1.5278772683858644,
      "grad_norm": 0.5653368830680847,
      "learning_rate": 1.151646047309277e-05,
      "loss": 0.4203,
      "mean_token_accuracy": 0.8600679874420166,
      "num_tokens": 857418865.0,
      "step": 25595
    },
    {
      "epoch": 1.5281757402101241,
      "grad_norm": 0.5152712464332581,
      "learning_rate": 1.1508646971284139e-05,
      "loss": 0.4209,
      "mean_token_accuracy": 0.858421802520752,
      "num_tokens": 857586545.0,
      "step": 25600
    },
    {
      "epoch": 1.528474212034384,
      "grad_norm": 0.5396140813827515,
      "learning_rate": 1.150083736446469e-05,
      "loss": 0.4749,
      "mean_token_accuracy": 0.841703450679779,
      "num_tokens": 857754225.0,
      "step": 25605
    },
    {
      "epoch": 1.5287726838586437,
      "grad_norm": 0.5097308158874512,
      "learning_rate": 1.1493031654536629e-05,
      "loss": 0.4398,
      "mean_token_accuracy": 0.853960394859314,
      "num_tokens": 857921905.0,
      "step": 25610
    },
    {
      "epoch": 1.5290711556829035,
      "grad_norm": 0.5128664970397949,
      "learning_rate": 1.1485229843401154e-05,
      "loss": 0.4348,
      "mean_token_accuracy": 0.853369927406311,
      "num_tokens": 858089585.0,
      "step": 25615
    },
    {
      "epoch": 1.5293696275071633,
      "grad_norm": 0.4968007206916809,
      "learning_rate": 1.147743193295855e-05,
      "loss": 0.4493,
      "mean_token_accuracy": 0.8519444227218628,
      "num_tokens": 858257265.0,
      "step": 25620
    },
    {
      "epoch": 1.5296680993314231,
      "grad_norm": 0.4765837788581848,
      "learning_rate": 1.146963792510814e-05,
      "loss": 0.4488,
      "mean_token_accuracy": 0.8500298142433167,
      "num_tokens": 858424945.0,
      "step": 25625
    },
    {
      "epoch": 1.529966571155683,
      "grad_norm": 0.49282023310661316,
      "learning_rate": 1.1461847821748302e-05,
      "loss": 0.4211,
      "mean_token_accuracy": 0.857706081867218,
      "num_tokens": 858592625.0,
      "step": 25630
    },
    {
      "epoch": 1.5302650429799427,
      "grad_norm": 0.5051887631416321,
      "learning_rate": 1.1454061624776463e-05,
      "loss": 0.4413,
      "mean_token_accuracy": 0.8518907308578492,
      "num_tokens": 858760305.0,
      "step": 25635
    },
    {
      "epoch": 1.5305635148042025,
      "grad_norm": 0.5638011693954468,
      "learning_rate": 1.1446279336089084e-05,
      "loss": 0.4574,
      "mean_token_accuracy": 0.8478050827980042,
      "num_tokens": 858927985.0,
      "step": 25640
    },
    {
      "epoch": 1.5308619866284623,
      "grad_norm": 0.5325712561607361,
      "learning_rate": 1.1438500957581688e-05,
      "loss": 0.4169,
      "mean_token_accuracy": 0.8607121586799622,
      "num_tokens": 859095665.0,
      "step": 25645
    },
    {
      "epoch": 1.531160458452722,
      "grad_norm": 0.5539132356643677,
      "learning_rate": 1.1430726491148846e-05,
      "loss": 0.4261,
      "mean_token_accuracy": 0.8586424946784973,
      "num_tokens": 859263345.0,
      "step": 25650
    },
    {
      "epoch": 1.5314589302769819,
      "grad_norm": 0.5448346734046936,
      "learning_rate": 1.1422955938684173e-05,
      "loss": 0.478,
      "mean_token_accuracy": 0.8418287038803101,
      "num_tokens": 859431025.0,
      "step": 25655
    },
    {
      "epoch": 1.5317574021012417,
      "grad_norm": 0.48333102464675903,
      "learning_rate": 1.1415189302080333e-05,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8548968076705933,
      "num_tokens": 859598705.0,
      "step": 25660
    },
    {
      "epoch": 1.5320558739255015,
      "grad_norm": 0.5613148808479309,
      "learning_rate": 1.1407426583229026e-05,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8573780298233032,
      "num_tokens": 859766385.0,
      "step": 25665
    },
    {
      "epoch": 1.5323543457497613,
      "grad_norm": 0.5383117198944092,
      "learning_rate": 1.1399667784021011e-05,
      "loss": 0.4193,
      "mean_token_accuracy": 0.8589884281158447,
      "num_tokens": 859934065.0,
      "step": 25670
    },
    {
      "epoch": 1.532652817574021,
      "grad_norm": 0.532345175743103,
      "learning_rate": 1.1391912906346083e-05,
      "loss": 0.481,
      "mean_token_accuracy": 0.8398604273796082,
      "num_tokens": 860101745.0,
      "step": 25675
    },
    {
      "epoch": 1.5329512893982808,
      "grad_norm": 0.6099774837493896,
      "learning_rate": 1.138416195209309e-05,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8556900739669799,
      "num_tokens": 860269425.0,
      "step": 25680
    },
    {
      "epoch": 1.5332497612225406,
      "grad_norm": 0.5399348735809326,
      "learning_rate": 1.1376414923149922e-05,
      "loss": 0.442,
      "mean_token_accuracy": 0.8528748631477356,
      "num_tokens": 860437105.0,
      "step": 25685
    },
    {
      "epoch": 1.5335482330468004,
      "grad_norm": 0.5816062688827515,
      "learning_rate": 1.1368671821403495e-05,
      "loss": 0.436,
      "mean_token_accuracy": 0.8532923817634582,
      "num_tokens": 860604785.0,
      "step": 25690
    },
    {
      "epoch": 1.5338467048710602,
      "grad_norm": 0.49152544140815735,
      "learning_rate": 1.1360932648739808e-05,
      "loss": 0.4448,
      "mean_token_accuracy": 0.8501729607582093,
      "num_tokens": 860772465.0,
      "step": 25695
    },
    {
      "epoch": 1.5341451766953198,
      "grad_norm": 0.5046047568321228,
      "learning_rate": 1.1353197407043861e-05,
      "loss": 0.4551,
      "mean_token_accuracy": 0.848496961593628,
      "num_tokens": 860940145.0,
      "step": 25700
    },
    {
      "epoch": 1.5344436485195798,
      "grad_norm": 0.5169011950492859,
      "learning_rate": 1.1345466098199722e-05,
      "loss": 0.4209,
      "mean_token_accuracy": 0.8584098815917969,
      "num_tokens": 861107825.0,
      "step": 25705
    },
    {
      "epoch": 1.5347421203438394,
      "grad_norm": 0.47887977957725525,
      "learning_rate": 1.1337738724090493e-05,
      "loss": 0.4161,
      "mean_token_accuracy": 0.8592329740524292,
      "num_tokens": 861275505.0,
      "step": 25710
    },
    {
      "epoch": 1.5350405921680994,
      "grad_norm": 0.5099368095397949,
      "learning_rate": 1.1330015286598318e-05,
      "loss": 0.4551,
      "mean_token_accuracy": 0.8475963115692139,
      "num_tokens": 861443185.0,
      "step": 25715
    },
    {
      "epoch": 1.535339063992359,
      "grad_norm": 0.5132337212562561,
      "learning_rate": 1.1322295787604394e-05,
      "loss": 0.4428,
      "mean_token_accuracy": 0.8527734637260437,
      "num_tokens": 861610865.0,
      "step": 25720
    },
    {
      "epoch": 1.535637535816619,
      "grad_norm": 0.4553278684616089,
      "learning_rate": 1.1314580228988925e-05,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8450196743011474,
      "num_tokens": 861778545.0,
      "step": 25725
    },
    {
      "epoch": 1.5359360076408786,
      "grad_norm": 0.5055566430091858,
      "learning_rate": 1.1306868612631186e-05,
      "loss": 0.4158,
      "mean_token_accuracy": 0.8602409601211548,
      "num_tokens": 861946225.0,
      "step": 25730
    },
    {
      "epoch": 1.5362344794651386,
      "grad_norm": 0.521346926689148,
      "learning_rate": 1.1299160940409503e-05,
      "loss": 0.463,
      "mean_token_accuracy": 0.845639967918396,
      "num_tokens": 862113905.0,
      "step": 25735
    },
    {
      "epoch": 1.5365329512893982,
      "grad_norm": 0.5294837951660156,
      "learning_rate": 1.1291457214201194e-05,
      "loss": 0.4248,
      "mean_token_accuracy": 0.8577120304107666,
      "num_tokens": 862281585.0,
      "step": 25740
    },
    {
      "epoch": 1.5368314231136582,
      "grad_norm": 0.5719627141952515,
      "learning_rate": 1.1283757435882668e-05,
      "loss": 0.491,
      "mean_token_accuracy": 0.8375521898269653,
      "num_tokens": 862449265.0,
      "step": 25745
    },
    {
      "epoch": 1.5371298949379177,
      "grad_norm": 0.5353347659111023,
      "learning_rate": 1.127606160732932e-05,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8414589047431946,
      "num_tokens": 862616945.0,
      "step": 25750
    },
    {
      "epoch": 1.5374283667621778,
      "grad_norm": 0.5057556629180908,
      "learning_rate": 1.126836973041564e-05,
      "loss": 0.4647,
      "mean_token_accuracy": 0.8440892338752747,
      "num_tokens": 862784625.0,
      "step": 25755
    },
    {
      "epoch": 1.5377268385864373,
      "grad_norm": 0.5216675400733948,
      "learning_rate": 1.1260681807015108e-05,
      "loss": 0.4529,
      "mean_token_accuracy": 0.8496063470840454,
      "num_tokens": 862952305.0,
      "step": 25760
    },
    {
      "epoch": 1.5380253104106973,
      "grad_norm": 0.5547958016395569,
      "learning_rate": 1.1252997839000267e-05,
      "loss": 0.4657,
      "mean_token_accuracy": 0.845276141166687,
      "num_tokens": 863119985.0,
      "step": 25765
    },
    {
      "epoch": 1.538323782234957,
      "grad_norm": 0.4714139997959137,
      "learning_rate": 1.1245317828242686e-05,
      "loss": 0.435,
      "mean_token_accuracy": 0.8532387018203735,
      "num_tokens": 863287665.0,
      "step": 25770
    },
    {
      "epoch": 1.538622254059217,
      "grad_norm": 0.5024564266204834,
      "learning_rate": 1.1237641776612974e-05,
      "loss": 0.4179,
      "mean_token_accuracy": 0.859048068523407,
      "num_tokens": 863455345.0,
      "step": 25775
    },
    {
      "epoch": 1.5389207258834765,
      "grad_norm": 0.5225319266319275,
      "learning_rate": 1.1229969685980782e-05,
      "loss": 0.4494,
      "mean_token_accuracy": 0.8506441712379456,
      "num_tokens": 863623025.0,
      "step": 25780
    },
    {
      "epoch": 1.5392191977077365,
      "grad_norm": 0.5305291414260864,
      "learning_rate": 1.1222301558214779e-05,
      "loss": 0.4546,
      "mean_token_accuracy": 0.8486043095588685,
      "num_tokens": 863790705.0,
      "step": 25785
    },
    {
      "epoch": 1.539517669531996,
      "grad_norm": 0.619536280632019,
      "learning_rate": 1.1214637395182683e-05,
      "loss": 0.4542,
      "mean_token_accuracy": 0.8495228528976441,
      "num_tokens": 863958385.0,
      "step": 25790
    },
    {
      "epoch": 1.539816141356256,
      "grad_norm": 0.4792643189430237,
      "learning_rate": 1.1206977198751243e-05,
      "loss": 0.4079,
      "mean_token_accuracy": 0.8632649302482605,
      "num_tokens": 864126065.0,
      "step": 25795
    },
    {
      "epoch": 1.5401146131805157,
      "grad_norm": 0.5104835629463196,
      "learning_rate": 1.1199320970786242e-05,
      "loss": 0.4466,
      "mean_token_accuracy": 0.8502445340156555,
      "num_tokens": 864293745.0,
      "step": 25800
    },
    {
      "epoch": 1.5404130850047757,
      "grad_norm": 0.5209883451461792,
      "learning_rate": 1.1191668713152498e-05,
      "loss": 0.4555,
      "mean_token_accuracy": 0.8478706836700439,
      "num_tokens": 864461425.0,
      "step": 25805
    },
    {
      "epoch": 1.5407115568290353,
      "grad_norm": 0.49672821164131165,
      "learning_rate": 1.118402042771385e-05,
      "loss": 0.4082,
      "mean_token_accuracy": 0.8616068124771118,
      "num_tokens": 864629105.0,
      "step": 25810
    },
    {
      "epoch": 1.5410100286532953,
      "grad_norm": 0.44503673911094666,
      "learning_rate": 1.1176376116333182e-05,
      "loss": 0.4144,
      "mean_token_accuracy": 0.8597459197044373,
      "num_tokens": 864796785.0,
      "step": 25815
    },
    {
      "epoch": 1.5413085004775549,
      "grad_norm": 0.4767151176929474,
      "learning_rate": 1.1168735780872421e-05,
      "loss": 0.4142,
      "mean_token_accuracy": 0.8602946519851684,
      "num_tokens": 864964465.0,
      "step": 25820
    },
    {
      "epoch": 1.5416069723018149,
      "grad_norm": 0.48956987261772156,
      "learning_rate": 1.1161099423192495e-05,
      "loss": 0.4408,
      "mean_token_accuracy": 0.852594530582428,
      "num_tokens": 865132145.0,
      "step": 25825
    },
    {
      "epoch": 1.5419054441260744,
      "grad_norm": 0.5427263379096985,
      "learning_rate": 1.1153467045153387e-05,
      "loss": 0.4302,
      "mean_token_accuracy": 0.8550697803497315,
      "num_tokens": 865299825.0,
      "step": 25830
    },
    {
      "epoch": 1.5422039159503342,
      "grad_norm": 0.4723684787750244,
      "learning_rate": 1.1145838648614102e-05,
      "loss": 0.3798,
      "mean_token_accuracy": 0.8718656778335572,
      "num_tokens": 865467505.0,
      "step": 25835
    },
    {
      "epoch": 1.542502387774594,
      "grad_norm": 0.4865725338459015,
      "learning_rate": 1.1138214235432683e-05,
      "loss": 0.4272,
      "mean_token_accuracy": 0.8571573376655579,
      "num_tokens": 865635185.0,
      "step": 25840
    },
    {
      "epoch": 1.5428008595988538,
      "grad_norm": 0.5468321442604065,
      "learning_rate": 1.1130593807466186e-05,
      "loss": 0.4488,
      "mean_token_accuracy": 0.849373722076416,
      "num_tokens": 865802865.0,
      "step": 25845
    },
    {
      "epoch": 1.5430993314231136,
      "grad_norm": 0.5016831755638123,
      "learning_rate": 1.1122977366570712e-05,
      "loss": 0.4036,
      "mean_token_accuracy": 0.8641238331794738,
      "num_tokens": 865970545.0,
      "step": 25850
    },
    {
      "epoch": 1.5433978032473734,
      "grad_norm": 0.5058051943778992,
      "learning_rate": 1.1115364914601382e-05,
      "loss": 0.4229,
      "mean_token_accuracy": 0.8579625368118287,
      "num_tokens": 866138225.0,
      "step": 25855
    },
    {
      "epoch": 1.5436962750716332,
      "grad_norm": 0.5117329359054565,
      "learning_rate": 1.1107756453412354e-05,
      "loss": 0.4536,
      "mean_token_accuracy": 0.8492544412612915,
      "num_tokens": 866305905.0,
      "step": 25860
    },
    {
      "epoch": 1.543994746895893,
      "grad_norm": 0.5152749419212341,
      "learning_rate": 1.1100151984856819e-05,
      "loss": 0.4568,
      "mean_token_accuracy": 0.847381591796875,
      "num_tokens": 866473585.0,
      "step": 25865
    },
    {
      "epoch": 1.5442932187201528,
      "grad_norm": 0.48024821281433105,
      "learning_rate": 1.1092551510786961e-05,
      "loss": 0.4051,
      "mean_token_accuracy": 0.8638315558433532,
      "num_tokens": 866641265.0,
      "step": 25870
    },
    {
      "epoch": 1.5445916905444126,
      "grad_norm": 0.5218802690505981,
      "learning_rate": 1.108495503305403e-05,
      "loss": 0.4201,
      "mean_token_accuracy": 0.8588930130004883,
      "num_tokens": 866808945.0,
      "step": 25875
    },
    {
      "epoch": 1.5448901623686724,
      "grad_norm": 0.498441219329834,
      "learning_rate": 1.1077362553508297e-05,
      "loss": 0.4495,
      "mean_token_accuracy": 0.8507336378097534,
      "num_tokens": 866976625.0,
      "step": 25880
    },
    {
      "epoch": 1.5451886341929322,
      "grad_norm": 0.514673113822937,
      "learning_rate": 1.1069774073999034e-05,
      "loss": 0.4682,
      "mean_token_accuracy": 0.844560420513153,
      "num_tokens": 867144305.0,
      "step": 25885
    },
    {
      "epoch": 1.545487106017192,
      "grad_norm": 0.48694169521331787,
      "learning_rate": 1.106218959637457e-05,
      "loss": 0.4335,
      "mean_token_accuracy": 0.8562507390975952,
      "num_tokens": 867311985.0,
      "step": 25890
    },
    {
      "epoch": 1.5457855778414518,
      "grad_norm": 0.533424973487854,
      "learning_rate": 1.1054609122482225e-05,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8466599106788635,
      "num_tokens": 867479665.0,
      "step": 25895
    },
    {
      "epoch": 1.5460840496657116,
      "grad_norm": 0.5263311266899109,
      "learning_rate": 1.1047032654168379e-05,
      "loss": 0.456,
      "mean_token_accuracy": 0.8474114298820495,
      "num_tokens": 867647345.0,
      "step": 25900
    },
    {
      "epoch": 1.5463825214899714,
      "grad_norm": 0.5120063424110413,
      "learning_rate": 1.1039460193278428e-05,
      "loss": 0.4173,
      "mean_token_accuracy": 0.8585888147354126,
      "num_tokens": 867815025.0,
      "step": 25905
    },
    {
      "epoch": 1.5466809933142311,
      "grad_norm": 0.5247727632522583,
      "learning_rate": 1.1031891741656762e-05,
      "loss": 0.427,
      "mean_token_accuracy": 0.8561374306678772,
      "num_tokens": 867982705.0,
      "step": 25910
    },
    {
      "epoch": 1.546979465138491,
      "grad_norm": 0.484334260225296,
      "learning_rate": 1.1024327301146833e-05,
      "loss": 0.4458,
      "mean_token_accuracy": 0.8502445459365845,
      "num_tokens": 868150385.0,
      "step": 25915
    },
    {
      "epoch": 1.5472779369627507,
      "grad_norm": 0.5441709160804749,
      "learning_rate": 1.1016766873591094e-05,
      "loss": 0.4253,
      "mean_token_accuracy": 0.8564714431762696,
      "num_tokens": 868318065.0,
      "step": 25920
    },
    {
      "epoch": 1.5475764087870105,
      "grad_norm": 0.6044000387191772,
      "learning_rate": 1.100921046083103e-05,
      "loss": 0.4515,
      "mean_token_accuracy": 0.8506501078605652,
      "num_tokens": 868485745.0,
      "step": 25925
    },
    {
      "epoch": 1.5478748806112703,
      "grad_norm": 0.4996013939380646,
      "learning_rate": 1.1001658064707148e-05,
      "loss": 0.4639,
      "mean_token_accuracy": 0.8459322452545166,
      "num_tokens": 868653425.0,
      "step": 25930
    },
    {
      "epoch": 1.5481733524355301,
      "grad_norm": 0.5255122780799866,
      "learning_rate": 1.0994109687058964e-05,
      "loss": 0.4347,
      "mean_token_accuracy": 0.8554992198944091,
      "num_tokens": 868821105.0,
      "step": 25935
    },
    {
      "epoch": 1.54847182425979,
      "grad_norm": 0.5189975500106812,
      "learning_rate": 1.0986565329725027e-05,
      "loss": 0.4452,
      "mean_token_accuracy": 0.8505308389663696,
      "num_tokens": 868988785.0,
      "step": 25940
    },
    {
      "epoch": 1.5487702960840497,
      "grad_norm": 0.47914978861808777,
      "learning_rate": 1.0979024994542908e-05,
      "loss": 0.4424,
      "mean_token_accuracy": 0.8541751265525818,
      "num_tokens": 869156465.0,
      "step": 25945
    },
    {
      "epoch": 1.5490687679083095,
      "grad_norm": 0.49758031964302063,
      "learning_rate": 1.09714886833492e-05,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8544256329536438,
      "num_tokens": 869324145.0,
      "step": 25950
    },
    {
      "epoch": 1.5493672397325693,
      "grad_norm": 0.5295895338058472,
      "learning_rate": 1.0963956397979494e-05,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8487892270088195,
      "num_tokens": 869491825.0,
      "step": 25955
    },
    {
      "epoch": 1.549665711556829,
      "grad_norm": 0.7546069025993347,
      "learning_rate": 1.0956428140268431e-05,
      "loss": 0.4829,
      "mean_token_accuracy": 0.8412846446037292,
      "num_tokens": 869651921.0,
      "step": 25960
    },
    {
      "epoch": 1.5499641833810889,
      "grad_norm": 0.4934066832065582,
      "learning_rate": 1.0948903912049663e-05,
      "loss": 0.4462,
      "mean_token_accuracy": 0.8510199189186096,
      "num_tokens": 869819601.0,
      "step": 25965
    },
    {
      "epoch": 1.5502626552053487,
      "grad_norm": 0.4560341536998749,
      "learning_rate": 1.0941383715155837e-05,
      "loss": 0.4246,
      "mean_token_accuracy": 0.8564893245697022,
      "num_tokens": 869987281.0,
      "step": 25970
    },
    {
      "epoch": 1.5505611270296082,
      "grad_norm": 0.46059396862983704,
      "learning_rate": 1.0933867551418647e-05,
      "loss": 0.3978,
      "mean_token_accuracy": 0.8638375401496887,
      "num_tokens": 870154961.0,
      "step": 25975
    },
    {
      "epoch": 1.5508595988538683,
      "grad_norm": 0.49816128611564636,
      "learning_rate": 1.0926355422668788e-05,
      "loss": 0.4384,
      "mean_token_accuracy": 0.8538828611373901,
      "num_tokens": 870322641.0,
      "step": 25980
    },
    {
      "epoch": 1.5511580706781278,
      "grad_norm": 0.47712135314941406,
      "learning_rate": 1.0918847330735981e-05,
      "loss": 0.4128,
      "mean_token_accuracy": 0.8611236929893493,
      "num_tokens": 870490321.0,
      "step": 25985
    },
    {
      "epoch": 1.5514565425023878,
      "grad_norm": 0.4922550618648529,
      "learning_rate": 1.0911343277448965e-05,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8386854410171509,
      "num_tokens": 870658001.0,
      "step": 25990
    },
    {
      "epoch": 1.5517550143266474,
      "grad_norm": 0.5132717490196228,
      "learning_rate": 1.090384326463548e-05,
      "loss": 0.4358,
      "mean_token_accuracy": 0.85394846200943,
      "num_tokens": 870825681.0,
      "step": 25995
    },
    {
      "epoch": 1.5520534861509074,
      "grad_norm": 0.5151049494743347,
      "learning_rate": 1.0896347294122297e-05,
      "loss": 0.408,
      "mean_token_accuracy": 0.8625671148300171,
      "num_tokens": 870993361.0,
      "step": 26000
    },
    {
      "epoch": 1.552351957975167,
      "grad_norm": 0.6029640436172485,
      "learning_rate": 1.08888553677352e-05,
      "loss": 0.4644,
      "mean_token_accuracy": 0.8446856737136841,
      "num_tokens": 871161041.0,
      "step": 26005
    },
    {
      "epoch": 1.552650429799427,
      "grad_norm": 0.5040461421012878,
      "learning_rate": 1.0881367487298985e-05,
      "loss": 0.4093,
      "mean_token_accuracy": 0.863008463382721,
      "num_tokens": 871328721.0,
      "step": 26010
    },
    {
      "epoch": 1.5529489016236866,
      "grad_norm": 0.4810337424278259,
      "learning_rate": 1.087388365463747e-05,
      "loss": 0.4313,
      "mean_token_accuracy": 0.8544554471969604,
      "num_tokens": 871496401.0,
      "step": 26015
    },
    {
      "epoch": 1.5532473734479466,
      "grad_norm": 0.48912104964256287,
      "learning_rate": 1.086640387157346e-05,
      "loss": 0.4274,
      "mean_token_accuracy": 0.8566324830055236,
      "num_tokens": 871664081.0,
      "step": 26020
    },
    {
      "epoch": 1.5535458452722062,
      "grad_norm": 0.5116869211196899,
      "learning_rate": 1.0858928139928822e-05,
      "loss": 0.4266,
      "mean_token_accuracy": 0.8555409789085389,
      "num_tokens": 871831761.0,
      "step": 26025
    },
    {
      "epoch": 1.5538443170964662,
      "grad_norm": 0.535221517086029,
      "learning_rate": 1.0851456461524388e-05,
      "loss": 0.4186,
      "mean_token_accuracy": 0.8599510550498962,
      "num_tokens": 871997278.0,
      "step": 26030
    },
    {
      "epoch": 1.5541427889207258,
      "grad_norm": 0.6300113201141357,
      "learning_rate": 1.0843988838180027e-05,
      "loss": 0.4623,
      "mean_token_accuracy": 0.8439712285995483,
      "num_tokens": 872157664.0,
      "step": 26035
    },
    {
      "epoch": 1.5544412607449858,
      "grad_norm": 0.5374897718429565,
      "learning_rate": 1.083652527171462e-05,
      "loss": 0.4253,
      "mean_token_accuracy": 0.8560479521751404,
      "num_tokens": 872325344.0,
      "step": 26040
    },
    {
      "epoch": 1.5547397325692454,
      "grad_norm": 0.46194443106651306,
      "learning_rate": 1.0829065763946058e-05,
      "loss": 0.4208,
      "mean_token_accuracy": 0.8587080955505371,
      "num_tokens": 872493024.0,
      "step": 26045
    },
    {
      "epoch": 1.5550382043935054,
      "grad_norm": 0.47020941972732544,
      "learning_rate": 1.0821610316691244e-05,
      "loss": 0.4302,
      "mean_token_accuracy": 0.8555469393730164,
      "num_tokens": 872660704.0,
      "step": 26050
    },
    {
      "epoch": 1.555336676217765,
      "grad_norm": 0.5209056735038757,
      "learning_rate": 1.0814158931766077e-05,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8565191507339478,
      "num_tokens": 872828384.0,
      "step": 26055
    },
    {
      "epoch": 1.555635148042025,
      "grad_norm": 0.49416953325271606,
      "learning_rate": 1.080671161098549e-05,
      "loss": 0.4032,
      "mean_token_accuracy": 0.8660324573516845,
      "num_tokens": 872996064.0,
      "step": 26060
    },
    {
      "epoch": 1.5559336198662845,
      "grad_norm": 0.5182694792747498,
      "learning_rate": 1.0799268356163413e-05,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8550604462623597,
      "num_tokens": 873163695.0,
      "step": 26065
    },
    {
      "epoch": 1.5562320916905446,
      "grad_norm": 0.5381470918655396,
      "learning_rate": 1.0791829169112788e-05,
      "loss": 0.4592,
      "mean_token_accuracy": 0.8478170037269592,
      "num_tokens": 873331375.0,
      "step": 26070
    },
    {
      "epoch": 1.5565305635148041,
      "grad_norm": 0.5272315144538879,
      "learning_rate": 1.0784394051645572e-05,
      "loss": 0.4507,
      "mean_token_accuracy": 0.8483597636222839,
      "num_tokens": 873499055.0,
      "step": 26075
    },
    {
      "epoch": 1.5568290353390641,
      "grad_norm": 0.49999555945396423,
      "learning_rate": 1.0776963005572705e-05,
      "loss": 0.419,
      "mean_token_accuracy": 0.860151493549347,
      "num_tokens": 873666735.0,
      "step": 26080
    },
    {
      "epoch": 1.5571275071633237,
      "grad_norm": 0.5012195706367493,
      "learning_rate": 1.0769536032704185e-05,
      "loss": 0.433,
      "mean_token_accuracy": 0.855725884437561,
      "num_tokens": 873834415.0,
      "step": 26085
    },
    {
      "epoch": 1.5574259789875837,
      "grad_norm": 0.44986918568611145,
      "learning_rate": 1.0762113134848969e-05,
      "loss": 0.4184,
      "mean_token_accuracy": 0.8602170944213867,
      "num_tokens": 874002095.0,
      "step": 26090
    },
    {
      "epoch": 1.5577244508118433,
      "grad_norm": 0.5704871416091919,
      "learning_rate": 1.0754694313815042e-05,
      "loss": 0.4574,
      "mean_token_accuracy": 0.8485983490943909,
      "num_tokens": 874169775.0,
      "step": 26095
    },
    {
      "epoch": 1.5580229226361033,
      "grad_norm": 0.4929519593715668,
      "learning_rate": 1.0747279571409399e-05,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8491709470748902,
      "num_tokens": 874337455.0,
      "step": 26100
    },
    {
      "epoch": 1.5583213944603629,
      "grad_norm": 0.44656768441200256,
      "learning_rate": 1.0739868909438033e-05,
      "loss": 0.4126,
      "mean_token_accuracy": 0.8603185057640076,
      "num_tokens": 874505135.0,
      "step": 26105
    },
    {
      "epoch": 1.5586198662846227,
      "grad_norm": 0.453061044216156,
      "learning_rate": 1.073246232970596e-05,
      "loss": 0.3809,
      "mean_token_accuracy": 0.8709900736808777,
      "num_tokens": 874670103.0,
      "step": 26110
    },
    {
      "epoch": 1.5589183381088825,
      "grad_norm": 0.5654840469360352,
      "learning_rate": 1.0725059834017173e-05,
      "loss": 0.4487,
      "mean_token_accuracy": 0.8499582409858704,
      "num_tokens": 874837783.0,
      "step": 26115
    },
    {
      "epoch": 1.5592168099331423,
      "grad_norm": 0.5006895661354065,
      "learning_rate": 1.0717661424174693e-05,
      "loss": 0.4271,
      "mean_token_accuracy": 0.8571871519088745,
      "num_tokens": 875005463.0,
      "step": 26120
    },
    {
      "epoch": 1.559515281757402,
      "grad_norm": 0.4911402761936188,
      "learning_rate": 1.0710267101980542e-05,
      "loss": 0.4585,
      "mean_token_accuracy": 0.8476500034332275,
      "num_tokens": 875173143.0,
      "step": 26125
    },
    {
      "epoch": 1.5598137535816619,
      "grad_norm": 0.5086870193481445,
      "learning_rate": 1.0702876869235737e-05,
      "loss": 0.4199,
      "mean_token_accuracy": 0.8591434955596924,
      "num_tokens": 875340823.0,
      "step": 26130
    },
    {
      "epoch": 1.5601122254059216,
      "grad_norm": 0.4928043484687805,
      "learning_rate": 1.069549072774032e-05,
      "loss": 0.4482,
      "mean_token_accuracy": 0.8503280401229858,
      "num_tokens": 875508503.0,
      "step": 26135
    },
    {
      "epoch": 1.5604106972301814,
      "grad_norm": 0.4967227578163147,
      "learning_rate": 1.0688108679293304e-05,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8401586532592773,
      "num_tokens": 875676183.0,
      "step": 26140
    },
    {
      "epoch": 1.5607091690544412,
      "grad_norm": 0.546454131603241,
      "learning_rate": 1.0680730725692724e-05,
      "loss": 0.4132,
      "mean_token_accuracy": 0.8598711490631104,
      "num_tokens": 875843863.0,
      "step": 26145
    },
    {
      "epoch": 1.561007640878701,
      "grad_norm": 0.4873913526535034,
      "learning_rate": 1.0673356868735637e-05,
      "loss": 0.432,
      "mean_token_accuracy": 0.8557199120521546,
      "num_tokens": 876011543.0,
      "step": 26150
    },
    {
      "epoch": 1.5613061127029608,
      "grad_norm": 0.5232744216918945,
      "learning_rate": 1.0665987110218062e-05,
      "loss": 0.4392,
      "mean_token_accuracy": 0.8530776500701904,
      "num_tokens": 876179223.0,
      "step": 26155
    },
    {
      "epoch": 1.5616045845272206,
      "grad_norm": 0.48951423168182373,
      "learning_rate": 1.0658621451935055e-05,
      "loss": 0.4136,
      "mean_token_accuracy": 0.8604199051856994,
      "num_tokens": 876346903.0,
      "step": 26160
    },
    {
      "epoch": 1.5619030563514804,
      "grad_norm": 0.4294096529483795,
      "learning_rate": 1.0651259895680635e-05,
      "loss": 0.4132,
      "mean_token_accuracy": 0.8607956528663635,
      "num_tokens": 876514583.0,
      "step": 26165
    },
    {
      "epoch": 1.5622015281757402,
      "grad_norm": 0.5300936102867126,
      "learning_rate": 1.0643902443247874e-05,
      "loss": 0.4507,
      "mean_token_accuracy": 0.8491112947463989,
      "num_tokens": 876682263.0,
      "step": 26170
    },
    {
      "epoch": 1.5625,
      "grad_norm": 0.46735334396362305,
      "learning_rate": 1.0636549096428793e-05,
      "loss": 0.4139,
      "mean_token_accuracy": 0.8602469325065613,
      "num_tokens": 876849943.0,
      "step": 26175
    },
    {
      "epoch": 1.5627984718242598,
      "grad_norm": 0.4797969162464142,
      "learning_rate": 1.0629199857014445e-05,
      "loss": 0.4595,
      "mean_token_accuracy": 0.8473978757858276,
      "num_tokens": 877011330.0,
      "step": 26180
    },
    {
      "epoch": 1.5630969436485196,
      "grad_norm": 0.5361015200614929,
      "learning_rate": 1.0621854726794869e-05,
      "loss": 0.4347,
      "mean_token_accuracy": 0.8547596216201783,
      "num_tokens": 877179010.0,
      "step": 26185
    },
    {
      "epoch": 1.5633954154727794,
      "grad_norm": 0.4962694048881531,
      "learning_rate": 1.061451370755911e-05,
      "loss": 0.4137,
      "mean_token_accuracy": 0.8613085985183716,
      "num_tokens": 877346690.0,
      "step": 26190
    },
    {
      "epoch": 1.5636938872970392,
      "grad_norm": 0.546158492565155,
      "learning_rate": 1.0607176801095215e-05,
      "loss": 0.4871,
      "mean_token_accuracy": 0.8394667744636536,
      "num_tokens": 877514370.0,
      "step": 26195
    },
    {
      "epoch": 1.563992359121299,
      "grad_norm": 0.44295719265937805,
      "learning_rate": 1.0599844009190208e-05,
      "loss": 0.4157,
      "mean_token_accuracy": 0.8599427461624145,
      "num_tokens": 877682050.0,
      "step": 26200
    },
    {
      "epoch": 1.5642908309455588,
      "grad_norm": 0.5119081139564514,
      "learning_rate": 1.0592515333630128e-05,
      "loss": 0.4415,
      "mean_token_accuracy": 0.8528867840766907,
      "num_tokens": 877849730.0,
      "step": 26205
    },
    {
      "epoch": 1.5645893027698186,
      "grad_norm": 0.5087532997131348,
      "learning_rate": 1.0585190776200026e-05,
      "loss": 0.4371,
      "mean_token_accuracy": 0.8547477126121521,
      "num_tokens": 878017410.0,
      "step": 26210
    },
    {
      "epoch": 1.5648877745940784,
      "grad_norm": 0.5864216685295105,
      "learning_rate": 1.0577870338683917e-05,
      "loss": 0.4632,
      "mean_token_accuracy": 0.8481450557708741,
      "num_tokens": 878185090.0,
      "step": 26215
    },
    {
      "epoch": 1.5651862464183381,
      "grad_norm": 0.4697224497795105,
      "learning_rate": 1.0570554022864842e-05,
      "loss": 0.3905,
      "mean_token_accuracy": 0.8672372698783875,
      "num_tokens": 878352770.0,
      "step": 26220
    },
    {
      "epoch": 1.565484718242598,
      "grad_norm": 0.5592724680900574,
      "learning_rate": 1.0563241830524803e-05,
      "loss": 0.4391,
      "mean_token_accuracy": 0.8536442875862121,
      "num_tokens": 878520450.0,
      "step": 26225
    },
    {
      "epoch": 1.5657831900668577,
      "grad_norm": 0.541803240776062,
      "learning_rate": 1.0555933763444838e-05,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8588035345077515,
      "num_tokens": 878688130.0,
      "step": 26230
    },
    {
      "epoch": 1.5660816618911175,
      "grad_norm": 0.4904037117958069,
      "learning_rate": 1.0548629823404962e-05,
      "loss": 0.4466,
      "mean_token_accuracy": 0.8503698110580444,
      "num_tokens": 878855810.0,
      "step": 26235
    },
    {
      "epoch": 1.5663801337153773,
      "grad_norm": 0.4765397012233734,
      "learning_rate": 1.0541330012184175e-05,
      "loss": 0.4132,
      "mean_token_accuracy": 0.8618573307991028,
      "num_tokens": 879023490.0,
      "step": 26240
    },
    {
      "epoch": 1.5666786055396371,
      "grad_norm": 0.5097711682319641,
      "learning_rate": 1.0534034331560482e-05,
      "loss": 0.4272,
      "mean_token_accuracy": 0.8572825908660888,
      "num_tokens": 879191170.0,
      "step": 26245
    },
    {
      "epoch": 1.5669770773638967,
      "grad_norm": 0.5480923652648926,
      "learning_rate": 1.052674278331089e-05,
      "loss": 0.4478,
      "mean_token_accuracy": 0.8509662389755249,
      "num_tokens": 879358850.0,
      "step": 26250
    },
    {
      "epoch": 1.5672755491881567,
      "grad_norm": 0.5250793099403381,
      "learning_rate": 1.0519455369211374e-05,
      "loss": 0.4558,
      "mean_token_accuracy": 0.8474293231964112,
      "num_tokens": 879526530.0,
      "step": 26255
    },
    {
      "epoch": 1.5675740210124163,
      "grad_norm": 0.4893653988838196,
      "learning_rate": 1.051217209103694e-05,
      "loss": 0.4227,
      "mean_token_accuracy": 0.8576881766319275,
      "num_tokens": 879694210.0,
      "step": 26260
    },
    {
      "epoch": 1.5678724928366763,
      "grad_norm": 0.5299640893936157,
      "learning_rate": 1.0504892950561546e-05,
      "loss": 0.4632,
      "mean_token_accuracy": 0.8466161608695983,
      "num_tokens": 879853507.0,
      "step": 26265
    },
    {
      "epoch": 1.5681709646609359,
      "grad_norm": 0.472123920917511,
      "learning_rate": 1.0497617949558165e-05,
      "loss": 0.4587,
      "mean_token_accuracy": 0.848860788345337,
      "num_tokens": 880021187.0,
      "step": 26270
    },
    {
      "epoch": 1.5684694364851959,
      "grad_norm": 0.5101821422576904,
      "learning_rate": 1.0490347089798762e-05,
      "loss": 0.4181,
      "mean_token_accuracy": 0.8592926263809204,
      "num_tokens": 880188867.0,
      "step": 26275
    },
    {
      "epoch": 1.5687679083094554,
      "grad_norm": 0.5313236713409424,
      "learning_rate": 1.0483080373054294e-05,
      "loss": 0.4323,
      "mean_token_accuracy": 0.8552606463432312,
      "num_tokens": 880356547.0,
      "step": 26280
    },
    {
      "epoch": 1.5690663801337155,
      "grad_norm": 0.5494893193244934,
      "learning_rate": 1.047581780109469e-05,
      "loss": 0.4214,
      "mean_token_accuracy": 0.8601992130279541,
      "num_tokens": 880524227.0,
      "step": 26285
    },
    {
      "epoch": 1.569364851957975,
      "grad_norm": 0.5203440189361572,
      "learning_rate": 1.0468559375688888e-05,
      "loss": 0.3706,
      "mean_token_accuracy": 0.8743468999862671,
      "num_tokens": 880691907.0,
      "step": 26290
    },
    {
      "epoch": 1.569663323782235,
      "grad_norm": 0.5092349052429199,
      "learning_rate": 1.0461305098604823e-05,
      "loss": 0.4253,
      "mean_token_accuracy": 0.8562567234039307,
      "num_tokens": 880859587.0,
      "step": 26295
    },
    {
      "epoch": 1.5699617956064946,
      "grad_norm": 0.557034969329834,
      "learning_rate": 1.0454054971609398e-05,
      "loss": 0.4403,
      "mean_token_accuracy": 0.8532804489135742,
      "num_tokens": 881027267.0,
      "step": 26300
    },
    {
      "epoch": 1.5702602674307546,
      "grad_norm": 0.4711591899394989,
      "learning_rate": 1.0446808996468512e-05,
      "loss": 0.4055,
      "mean_token_accuracy": 0.8620268106460571,
      "num_tokens": 881188675.0,
      "step": 26305
    },
    {
      "epoch": 1.5705587392550142,
      "grad_norm": 0.49493682384490967,
      "learning_rate": 1.0439567174947063e-05,
      "loss": 0.4495,
      "mean_token_accuracy": 0.850363838672638,
      "num_tokens": 881356355.0,
      "step": 26310
    },
    {
      "epoch": 1.5708572110792742,
      "grad_norm": 0.4804880917072296,
      "learning_rate": 1.0432329508808927e-05,
      "loss": 0.4157,
      "mean_token_accuracy": 0.8591256022453309,
      "num_tokens": 881524035.0,
      "step": 26315
    },
    {
      "epoch": 1.5711556829035338,
      "grad_norm": 0.5023981332778931,
      "learning_rate": 1.0425095999816978e-05,
      "loss": 0.4128,
      "mean_token_accuracy": 0.8612668514251709,
      "num_tokens": 881691715.0,
      "step": 26320
    },
    {
      "epoch": 1.5714541547277938,
      "grad_norm": 0.47381237149238586,
      "learning_rate": 1.041786664973306e-05,
      "loss": 0.4258,
      "mean_token_accuracy": 0.8583860158920288,
      "num_tokens": 881859395.0,
      "step": 26325
    },
    {
      "epoch": 1.5717526265520534,
      "grad_norm": 0.450787752866745,
      "learning_rate": 1.041064146031802e-05,
      "loss": 0.4303,
      "mean_token_accuracy": 0.8556125521659851,
      "num_tokens": 882027075.0,
      "step": 26330
    },
    {
      "epoch": 1.5720510983763134,
      "grad_norm": 0.5109887719154358,
      "learning_rate": 1.0403420433331684e-05,
      "loss": 0.4239,
      "mean_token_accuracy": 0.8576643228530884,
      "num_tokens": 882194755.0,
      "step": 26335
    },
    {
      "epoch": 1.572349570200573,
      "grad_norm": 0.4925119876861572,
      "learning_rate": 1.0396203570532867e-05,
      "loss": 0.418,
      "mean_token_accuracy": 0.8584337353706359,
      "num_tokens": 882362435.0,
      "step": 26340
    },
    {
      "epoch": 1.572648042024833,
      "grad_norm": 0.531807541847229,
      "learning_rate": 1.038899087367938e-05,
      "loss": 0.4744,
      "mean_token_accuracy": 0.8419658899307251,
      "num_tokens": 882530115.0,
      "step": 26345
    },
    {
      "epoch": 1.5729465138490926,
      "grad_norm": 0.4866146147251129,
      "learning_rate": 1.0381782344527985e-05,
      "loss": 0.3919,
      "mean_token_accuracy": 0.867732310295105,
      "num_tokens": 882697795.0,
      "step": 26350
    },
    {
      "epoch": 1.5732449856733526,
      "grad_norm": 0.5182363390922546,
      "learning_rate": 1.0374577984834479e-05,
      "loss": 0.4695,
      "mean_token_accuracy": 0.8426398634910583,
      "num_tokens": 882865475.0,
      "step": 26355
    },
    {
      "epoch": 1.5735434574976122,
      "grad_norm": 0.44843345880508423,
      "learning_rate": 1.0367377796353594e-05,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8491769075393677,
      "num_tokens": 883033155.0,
      "step": 26360
    },
    {
      "epoch": 1.5738419293218722,
      "grad_norm": 0.49718332290649414,
      "learning_rate": 1.0360181780839085e-05,
      "loss": 0.4554,
      "mean_token_accuracy": 0.8470237255096436,
      "num_tokens": 883200835.0,
      "step": 26365
    },
    {
      "epoch": 1.5741404011461317,
      "grad_norm": 0.5355473160743713,
      "learning_rate": 1.0352989940043664e-05,
      "loss": 0.4174,
      "mean_token_accuracy": 0.8623523831367492,
      "num_tokens": 883368515.0,
      "step": 26370
    },
    {
      "epoch": 1.5744388729703918,
      "grad_norm": 0.5284214615821838,
      "learning_rate": 1.034580227571904e-05,
      "loss": 0.4409,
      "mean_token_accuracy": 0.8529046893119812,
      "num_tokens": 883536195.0,
      "step": 26375
    },
    {
      "epoch": 1.5747373447946513,
      "grad_norm": 0.5034896731376648,
      "learning_rate": 1.0338618789615911e-05,
      "loss": 0.3938,
      "mean_token_accuracy": 0.8655194997787475,
      "num_tokens": 883703875.0,
      "step": 26380
    },
    {
      "epoch": 1.5750358166189111,
      "grad_norm": 0.5367690324783325,
      "learning_rate": 1.0331439483483932e-05,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8505845069885254,
      "num_tokens": 883871555.0,
      "step": 26385
    },
    {
      "epoch": 1.575334288443171,
      "grad_norm": 0.566448450088501,
      "learning_rate": 1.0324264359071761e-05,
      "loss": 0.4054,
      "mean_token_accuracy": 0.8646904468536377,
      "num_tokens": 884039235.0,
      "step": 26390
    },
    {
      "epoch": 1.5756327602674307,
      "grad_norm": 0.5155763626098633,
      "learning_rate": 1.031709341812703e-05,
      "loss": 0.4426,
      "mean_token_accuracy": 0.8529046893119812,
      "num_tokens": 884206915.0,
      "step": 26395
    },
    {
      "epoch": 1.5759312320916905,
      "grad_norm": 0.4850751459598541,
      "learning_rate": 1.030992666239636e-05,
      "loss": 0.405,
      "mean_token_accuracy": 0.8638852596282959,
      "num_tokens": 884374595.0,
      "step": 26400
    },
    {
      "epoch": 1.5762297039159503,
      "grad_norm": 0.5149714946746826,
      "learning_rate": 1.0302764093625353e-05,
      "loss": 0.4114,
      "mean_token_accuracy": 0.8620481967926026,
      "num_tokens": 884542275.0,
      "step": 26405
    },
    {
      "epoch": 1.57652817574021,
      "grad_norm": 0.48544833064079285,
      "learning_rate": 1.0295605713558565e-05,
      "loss": 0.4343,
      "mean_token_accuracy": 0.8533400893211365,
      "num_tokens": 884709955.0,
      "step": 26410
    },
    {
      "epoch": 1.5768266475644699,
      "grad_norm": 0.4907308518886566,
      "learning_rate": 1.0288451523939562e-05,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8554574728012085,
      "num_tokens": 884877635.0,
      "step": 26415
    },
    {
      "epoch": 1.5771251193887297,
      "grad_norm": 0.48904645442962646,
      "learning_rate": 1.0281301526510883e-05,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8516879439353943,
      "num_tokens": 885045315.0,
      "step": 26420
    },
    {
      "epoch": 1.5774235912129895,
      "grad_norm": 0.4829970598220825,
      "learning_rate": 1.0274155723014036e-05,
      "loss": 0.4192,
      "mean_token_accuracy": 0.8591077208518982,
      "num_tokens": 885212995.0,
      "step": 26425
    },
    {
      "epoch": 1.5777220630372493,
      "grad_norm": 0.4841747283935547,
      "learning_rate": 1.0267014115189517e-05,
      "loss": 0.4152,
      "mean_token_accuracy": 0.8611833333969117,
      "num_tokens": 885380675.0,
      "step": 26430
    },
    {
      "epoch": 1.578020534861509,
      "grad_norm": 0.5143060088157654,
      "learning_rate": 1.0259876704776792e-05,
      "loss": 0.4408,
      "mean_token_accuracy": 0.8531253695487976,
      "num_tokens": 885548355.0,
      "step": 26435
    },
    {
      "epoch": 1.5783190066857689,
      "grad_norm": 0.48538410663604736,
      "learning_rate": 1.0252743493514318e-05,
      "loss": 0.3942,
      "mean_token_accuracy": 0.8671418309211731,
      "num_tokens": 885716035.0,
      "step": 26440
    },
    {
      "epoch": 1.5786174785100286,
      "grad_norm": 0.4950398802757263,
      "learning_rate": 1.024561448313951e-05,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8537456750869751,
      "num_tokens": 885883715.0,
      "step": 26445
    },
    {
      "epoch": 1.5789159503342884,
      "grad_norm": 0.5013229846954346,
      "learning_rate": 1.0238489675388771e-05,
      "loss": 0.4297,
      "mean_token_accuracy": 0.8550280213356019,
      "num_tokens": 886051395.0,
      "step": 26450
    },
    {
      "epoch": 1.5792144221585482,
      "grad_norm": 0.4976925551891327,
      "learning_rate": 1.0231369071997486e-05,
      "loss": 0.445,
      "mean_token_accuracy": 0.8521591305732727,
      "num_tokens": 886219075.0,
      "step": 26455
    },
    {
      "epoch": 1.579512893982808,
      "grad_norm": 0.5492770075798035,
      "learning_rate": 1.0224252674700003e-05,
      "loss": 0.4322,
      "mean_token_accuracy": 0.8546105146408081,
      "num_tokens": 886386755.0,
      "step": 26460
    },
    {
      "epoch": 1.5798113658070678,
      "grad_norm": 0.4879792630672455,
      "learning_rate": 1.0217140485229661e-05,
      "loss": 0.4151,
      "mean_token_accuracy": 0.861815583705902,
      "num_tokens": 886554435.0,
      "step": 26465
    },
    {
      "epoch": 1.5801098376313276,
      "grad_norm": 0.579795777797699,
      "learning_rate": 1.0210032505318756e-05,
      "loss": 0.4468,
      "mean_token_accuracy": 0.8492305755615235,
      "num_tokens": 886722115.0,
      "step": 26470
    },
    {
      "epoch": 1.5804083094555874,
      "grad_norm": 0.48856765031814575,
      "learning_rate": 1.0202928736698561e-05,
      "loss": 0.435,
      "mean_token_accuracy": 0.8541989684104919,
      "num_tokens": 886889795.0,
      "step": 26475
    },
    {
      "epoch": 1.5807067812798472,
      "grad_norm": 0.5250730514526367,
      "learning_rate": 1.0195829181099353e-05,
      "loss": 0.4033,
      "mean_token_accuracy": 0.8631396889686584,
      "num_tokens": 887057475.0,
      "step": 26480
    },
    {
      "epoch": 1.581005253104107,
      "grad_norm": 0.5376426577568054,
      "learning_rate": 1.0188733840250343e-05,
      "loss": 0.4467,
      "mean_token_accuracy": 0.849988055229187,
      "num_tokens": 887225155.0,
      "step": 26485
    },
    {
      "epoch": 1.5813037249283668,
      "grad_norm": 0.5562255382537842,
      "learning_rate": 1.0181642715879737e-05,
      "loss": 0.4596,
      "mean_token_accuracy": 0.8456101655960083,
      "num_tokens": 887392835.0,
      "step": 26490
    },
    {
      "epoch": 1.5816021967526266,
      "grad_norm": 0.5164707899093628,
      "learning_rate": 1.0174555809714701e-05,
      "loss": 0.4326,
      "mean_token_accuracy": 0.8548789143562316,
      "num_tokens": 887560515.0,
      "step": 26495
    },
    {
      "epoch": 1.5819006685768864,
      "grad_norm": 0.5418306589126587,
      "learning_rate": 1.0167473123481395e-05,
      "loss": 0.4101,
      "mean_token_accuracy": 0.8625611424446106,
      "num_tokens": 887728195.0,
      "step": 26500
    },
    {
      "epoch": 1.5821991404011462,
      "grad_norm": 0.5183385610580444,
      "learning_rate": 1.016039465890493e-05,
      "loss": 0.4168,
      "mean_token_accuracy": 0.8602648258209229,
      "num_tokens": 887895875.0,
      "step": 26505
    },
    {
      "epoch": 1.582497612225406,
      "grad_norm": 0.5144839882850647,
      "learning_rate": 1.0153320417709395e-05,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8490874290466308,
      "num_tokens": 888063555.0,
      "step": 26510
    },
    {
      "epoch": 1.5827960840496658,
      "grad_norm": 0.49877968430519104,
      "learning_rate": 1.0146250401617856e-05,
      "loss": 0.4373,
      "mean_token_accuracy": 0.8533878087997436,
      "num_tokens": 888231235.0,
      "step": 26515
    },
    {
      "epoch": 1.5830945558739256,
      "grad_norm": 0.4924964904785156,
      "learning_rate": 1.0139184612352346e-05,
      "loss": 0.4143,
      "mean_token_accuracy": 0.860312533378601,
      "num_tokens": 888398915.0,
      "step": 26520
    },
    {
      "epoch": 1.5833930276981851,
      "grad_norm": 0.8070304989814758,
      "learning_rate": 1.0132123051633874e-05,
      "loss": 0.4514,
      "mean_token_accuracy": 0.8498687744140625,
      "num_tokens": 888566595.0,
      "step": 26525
    },
    {
      "epoch": 1.5836914995224451,
      "grad_norm": 0.4972057342529297,
      "learning_rate": 1.01250657211824e-05,
      "loss": 0.4262,
      "mean_token_accuracy": 0.8566444039344787,
      "num_tokens": 888734275.0,
      "step": 26530
    },
    {
      "epoch": 1.5839899713467047,
      "grad_norm": 0.5219476819038391,
      "learning_rate": 1.0118012622716874e-05,
      "loss": 0.4589,
      "mean_token_accuracy": 0.8484015226364136,
      "num_tokens": 888901955.0,
      "step": 26535
    },
    {
      "epoch": 1.5842884431709647,
      "grad_norm": 0.6271768808364868,
      "learning_rate": 1.0110963757955208e-05,
      "loss": 0.4679,
      "mean_token_accuracy": 0.8445663809776306,
      "num_tokens": 889069635.0,
      "step": 26540
    },
    {
      "epoch": 1.5845869149952243,
      "grad_norm": 0.48052552342414856,
      "learning_rate": 1.0103919128614284e-05,
      "loss": 0.4077,
      "mean_token_accuracy": 0.8619467973709106,
      "num_tokens": 889237315.0,
      "step": 26545
    },
    {
      "epoch": 1.5848853868194843,
      "grad_norm": 0.506382167339325,
      "learning_rate": 1.009687873640996e-05,
      "loss": 0.4318,
      "mean_token_accuracy": 0.8543302059173584,
      "num_tokens": 889404995.0,
      "step": 26550
    },
    {
      "epoch": 1.585183858643744,
      "grad_norm": 0.5315185189247131,
      "learning_rate": 1.0089842583057028e-05,
      "loss": 0.4315,
      "mean_token_accuracy": 0.8549147129058838,
      "num_tokens": 889572675.0,
      "step": 26555
    },
    {
      "epoch": 1.585482330468004,
      "grad_norm": 0.5435509085655212,
      "learning_rate": 1.0082810670269299e-05,
      "loss": 0.4426,
      "mean_token_accuracy": 0.853483247756958,
      "num_tokens": 889740355.0,
      "step": 26560
    },
    {
      "epoch": 1.5857808022922635,
      "grad_norm": 0.5118299126625061,
      "learning_rate": 1.007578299975952e-05,
      "loss": 0.4255,
      "mean_token_accuracy": 0.8571573376655579,
      "num_tokens": 889908035.0,
      "step": 26565
    },
    {
      "epoch": 1.5860792741165235,
      "grad_norm": 0.5012679100036621,
      "learning_rate": 1.00687595732394e-05,
      "loss": 0.4049,
      "mean_token_accuracy": 0.8631456613540649,
      "num_tokens": 890075715.0,
      "step": 26570
    },
    {
      "epoch": 1.586377745940783,
      "grad_norm": 0.5095944404602051,
      "learning_rate": 1.0061740392419631e-05,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8450733661651612,
      "num_tokens": 890243395.0,
      "step": 26575
    },
    {
      "epoch": 1.586676217765043,
      "grad_norm": 0.5599350333213806,
      "learning_rate": 1.0054725459009858e-05,
      "loss": 0.455,
      "mean_token_accuracy": 0.8507157325744629,
      "num_tokens": 890411075.0,
      "step": 26580
    },
    {
      "epoch": 1.5869746895893027,
      "grad_norm": 0.4542834460735321,
      "learning_rate": 1.0047714774718709e-05,
      "loss": 0.4067,
      "mean_token_accuracy": 0.8618334889411926,
      "num_tokens": 890578755.0,
      "step": 26585
    },
    {
      "epoch": 1.5872731614135627,
      "grad_norm": 0.5228456258773804,
      "learning_rate": 1.004070834125376e-05,
      "loss": 0.4544,
      "mean_token_accuracy": 0.8470476031303406,
      "num_tokens": 890746435.0,
      "step": 26590
    },
    {
      "epoch": 1.5875716332378222,
      "grad_norm": 0.5303345918655396,
      "learning_rate": 1.0033706160321551e-05,
      "loss": 0.4286,
      "mean_token_accuracy": 0.8567577362060547,
      "num_tokens": 890914115.0,
      "step": 26595
    },
    {
      "epoch": 1.5878701050620823,
      "grad_norm": 0.5150555968284607,
      "learning_rate": 1.0026708233627601e-05,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8535556197166443,
      "num_tokens": 891076157.0,
      "step": 26600
    },
    {
      "epoch": 1.5881685768863418,
      "grad_norm": 0.4887823760509491,
      "learning_rate": 1.0019714562876377e-05,
      "loss": 0.4303,
      "mean_token_accuracy": 0.8553858995437622,
      "num_tokens": 891243837.0,
      "step": 26605
    },
    {
      "epoch": 1.5884670487106018,
      "grad_norm": 0.5814841985702515,
      "learning_rate": 1.0012725149771329e-05,
      "loss": 0.4376,
      "mean_token_accuracy": 0.8544375538825989,
      "num_tokens": 891411517.0,
      "step": 26610
    },
    {
      "epoch": 1.5887655205348614,
      "grad_norm": 0.5063245296478271,
      "learning_rate": 1.000573999601484e-05,
      "loss": 0.4286,
      "mean_token_accuracy": 0.8563342452049255,
      "num_tokens": 891579197.0,
      "step": 26615
    },
    {
      "epoch": 1.5890639923591214,
      "grad_norm": 0.501069188117981,
      "learning_rate": 9.998759103308278e-06,
      "loss": 0.4112,
      "mean_token_accuracy": 0.8617678761482239,
      "num_tokens": 891746877.0,
      "step": 26620
    },
    {
      "epoch": 1.589362464183381,
      "grad_norm": 0.5059115290641785,
      "learning_rate": 9.991782473351985e-06,
      "loss": 0.4365,
      "mean_token_accuracy": 0.8556781530380249,
      "num_tokens": 891914557.0,
      "step": 26625
    },
    {
      "epoch": 1.589660936007641,
      "grad_norm": 0.5677490234375,
      "learning_rate": 9.984810107845228e-06,
      "loss": 0.4737,
      "mean_token_accuracy": 0.8436061024665833,
      "num_tokens": 892082237.0,
      "step": 26630
    },
    {
      "epoch": 1.5899594078319006,
      "grad_norm": 0.49569299817085266,
      "learning_rate": 9.977842008486263e-06,
      "loss": 0.4253,
      "mean_token_accuracy": 0.8565608978271484,
      "num_tokens": 892249917.0,
      "step": 26635
    },
    {
      "epoch": 1.5902578796561606,
      "grad_norm": 0.4933372437953949,
      "learning_rate": 9.970878176972297e-06,
      "loss": 0.4334,
      "mean_token_accuracy": 0.8551771402359009,
      "num_tokens": 892417597.0,
      "step": 26640
    },
    {
      "epoch": 1.5905563514804202,
      "grad_norm": 0.5534626245498657,
      "learning_rate": 9.963918614999502e-06,
      "loss": 0.4443,
      "mean_token_accuracy": 0.8499701857566834,
      "num_tokens": 892585277.0,
      "step": 26645
    },
    {
      "epoch": 1.5908548233046802,
      "grad_norm": 0.5127054452896118,
      "learning_rate": 9.956963324263013e-06,
      "loss": 0.4057,
      "mean_token_accuracy": 0.8625969290733337,
      "num_tokens": 892752957.0,
      "step": 26650
    },
    {
      "epoch": 1.5911532951289398,
      "grad_norm": 0.5097017288208008,
      "learning_rate": 9.95001230645691e-06,
      "loss": 0.4598,
      "mean_token_accuracy": 0.8478587508201599,
      "num_tokens": 892920637.0,
      "step": 26655
    },
    {
      "epoch": 1.5914517669531996,
      "grad_norm": 0.5153566598892212,
      "learning_rate": 9.943065563274244e-06,
      "loss": 0.4863,
      "mean_token_accuracy": 0.8390228271484375,
      "num_tokens": 893085449.0,
      "step": 26660
    },
    {
      "epoch": 1.5917502387774594,
      "grad_norm": 0.5006033182144165,
      "learning_rate": 9.936123096407024e-06,
      "loss": 0.409,
      "mean_token_accuracy": 0.8611833453178406,
      "num_tokens": 893253129.0,
      "step": 26665
    },
    {
      "epoch": 1.5920487106017192,
      "grad_norm": 0.5272590517997742,
      "learning_rate": 9.929184907546217e-06,
      "loss": 0.4513,
      "mean_token_accuracy": 0.847882616519928,
      "num_tokens": 893420809.0,
      "step": 26670
    },
    {
      "epoch": 1.592347182425979,
      "grad_norm": 0.5295668244361877,
      "learning_rate": 9.922250998381751e-06,
      "loss": 0.4247,
      "mean_token_accuracy": 0.8569724321365356,
      "num_tokens": 893588489.0,
      "step": 26675
    },
    {
      "epoch": 1.5926456542502387,
      "grad_norm": 0.5618000030517578,
      "learning_rate": 9.91532137060249e-06,
      "loss": 0.4325,
      "mean_token_accuracy": 0.8550161123275757,
      "num_tokens": 893756169.0,
      "step": 26680
    },
    {
      "epoch": 1.5929441260744985,
      "grad_norm": 0.4733685553073883,
      "learning_rate": 9.908396025896297e-06,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8540140748023987,
      "num_tokens": 893923849.0,
      "step": 26685
    },
    {
      "epoch": 1.5932425978987583,
      "grad_norm": 0.4882482588291168,
      "learning_rate": 9.901474965949953e-06,
      "loss": 0.4005,
      "mean_token_accuracy": 0.8641655802726745,
      "num_tokens": 894091529.0,
      "step": 26690
    },
    {
      "epoch": 1.5935410697230181,
      "grad_norm": 0.4893314242362976,
      "learning_rate": 9.894558192449208e-06,
      "loss": 0.4418,
      "mean_token_accuracy": 0.8521770238876343,
      "num_tokens": 894259209.0,
      "step": 26695
    },
    {
      "epoch": 1.593839541547278,
      "grad_norm": 0.5153326392173767,
      "learning_rate": 9.887645707078776e-06,
      "loss": 0.4472,
      "mean_token_accuracy": 0.8510497450828552,
      "num_tokens": 894426889.0,
      "step": 26700
    },
    {
      "epoch": 1.5941380133715377,
      "grad_norm": 0.5323935151100159,
      "learning_rate": 9.88073751152232e-06,
      "loss": 0.4482,
      "mean_token_accuracy": 0.8494453072547913,
      "num_tokens": 894594569.0,
      "step": 26705
    },
    {
      "epoch": 1.5944364851957975,
      "grad_norm": 0.5779756903648376,
      "learning_rate": 9.873833607462462e-06,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8369080305099488,
      "num_tokens": 894762249.0,
      "step": 26710
    },
    {
      "epoch": 1.5947349570200573,
      "grad_norm": 0.5010717511177063,
      "learning_rate": 9.866933996580763e-06,
      "loss": 0.4228,
      "mean_token_accuracy": 0.8586663365364074,
      "num_tokens": 894929929.0,
      "step": 26715
    },
    {
      "epoch": 1.595033428844317,
      "grad_norm": 0.4734920561313629,
      "learning_rate": 9.860038680557762e-06,
      "loss": 0.4512,
      "mean_token_accuracy": 0.8497136950492858,
      "num_tokens": 895097609.0,
      "step": 26720
    },
    {
      "epoch": 1.5953319006685769,
      "grad_norm": 0.5456564426422119,
      "learning_rate": 9.853147661072936e-06,
      "loss": 0.4087,
      "mean_token_accuracy": 0.863020396232605,
      "num_tokens": 895265289.0,
      "step": 26725
    },
    {
      "epoch": 1.5956303724928367,
      "grad_norm": 0.4554482102394104,
      "learning_rate": 9.846260939804719e-06,
      "loss": 0.3861,
      "mean_token_accuracy": 0.8693188548088073,
      "num_tokens": 895432969.0,
      "step": 26730
    },
    {
      "epoch": 1.5959288443170965,
      "grad_norm": 0.4803209900856018,
      "learning_rate": 9.839378518430506e-06,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8443278074264526,
      "num_tokens": 895600649.0,
      "step": 26735
    },
    {
      "epoch": 1.5962273161413563,
      "grad_norm": 0.4860791265964508,
      "learning_rate": 9.832500398626628e-06,
      "loss": 0.4248,
      "mean_token_accuracy": 0.8565907120704651,
      "num_tokens": 895768329.0,
      "step": 26740
    },
    {
      "epoch": 1.596525787965616,
      "grad_norm": 0.49583545327186584,
      "learning_rate": 9.825626582068384e-06,
      "loss": 0.4346,
      "mean_token_accuracy": 0.8548073410987854,
      "num_tokens": 895936009.0,
      "step": 26745
    },
    {
      "epoch": 1.5968242597898759,
      "grad_norm": 0.5204347372055054,
      "learning_rate": 9.818757070430016e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8525289297103882,
      "num_tokens": 896103689.0,
      "step": 26750
    },
    {
      "epoch": 1.5971227316141356,
      "grad_norm": 0.5683661699295044,
      "learning_rate": 9.811891865384723e-06,
      "loss": 0.4721,
      "mean_token_accuracy": 0.8433377146720886,
      "num_tokens": 896271369.0,
      "step": 26755
    },
    {
      "epoch": 1.5974212034383954,
      "grad_norm": 0.47288426756858826,
      "learning_rate": 9.805030968604651e-06,
      "loss": 0.415,
      "mean_token_accuracy": 0.8594954133033752,
      "num_tokens": 896439049.0,
      "step": 26760
    },
    {
      "epoch": 1.5977196752626552,
      "grad_norm": 0.5538458824157715,
      "learning_rate": 9.798174381760896e-06,
      "loss": 0.4291,
      "mean_token_accuracy": 0.8562626719474793,
      "num_tokens": 896606729.0,
      "step": 26765
    },
    {
      "epoch": 1.598018147086915,
      "grad_norm": 0.5352579355239868,
      "learning_rate": 9.791322106523516e-06,
      "loss": 0.4265,
      "mean_token_accuracy": 0.857079803943634,
      "num_tokens": 896774409.0,
      "step": 26770
    },
    {
      "epoch": 1.5983166189111748,
      "grad_norm": 0.5396891236305237,
      "learning_rate": 9.784474144561498e-06,
      "loss": 0.4672,
      "mean_token_accuracy": 0.845902419090271,
      "num_tokens": 896942089.0,
      "step": 26775
    },
    {
      "epoch": 1.5986150907354346,
      "grad_norm": 0.446588397026062,
      "learning_rate": 9.777630497542792e-06,
      "loss": 0.4496,
      "mean_token_accuracy": 0.8509543061256408,
      "num_tokens": 897109769.0,
      "step": 26780
    },
    {
      "epoch": 1.5989135625596944,
      "grad_norm": 0.5855042338371277,
      "learning_rate": 9.770791167134293e-06,
      "loss": 0.4575,
      "mean_token_accuracy": 0.8473398447036743,
      "num_tokens": 897277449.0,
      "step": 26785
    },
    {
      "epoch": 1.5992120343839542,
      "grad_norm": 0.48388636112213135,
      "learning_rate": 9.763956155001852e-06,
      "loss": 0.4225,
      "mean_token_accuracy": 0.85706787109375,
      "num_tokens": 897445129.0,
      "step": 26790
    },
    {
      "epoch": 1.599510506208214,
      "grad_norm": 0.4797343611717224,
      "learning_rate": 9.757125462810263e-06,
      "loss": 0.3971,
      "mean_token_accuracy": 0.8664678454399108,
      "num_tokens": 897612809.0,
      "step": 26795
    },
    {
      "epoch": 1.5998089780324736,
      "grad_norm": 0.5071563720703125,
      "learning_rate": 9.75029909222326e-06,
      "loss": 0.4709,
      "mean_token_accuracy": 0.843057382106781,
      "num_tokens": 897780489.0,
      "step": 26800
    },
    {
      "epoch": 1.6001074498567336,
      "grad_norm": 0.5073161721229553,
      "learning_rate": 9.743477044903532e-06,
      "loss": 0.4831,
      "mean_token_accuracy": 0.8407252907752991,
      "num_tokens": 897948169.0,
      "step": 26805
    },
    {
      "epoch": 1.6004059216809932,
      "grad_norm": 0.49718189239501953,
      "learning_rate": 9.73665932251272e-06,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8544196605682373,
      "num_tokens": 898115849.0,
      "step": 26810
    },
    {
      "epoch": 1.6007043935052532,
      "grad_norm": 0.4949471354484558,
      "learning_rate": 9.729845926711403e-06,
      "loss": 0.4222,
      "mean_token_accuracy": 0.8589109063148499,
      "num_tokens": 898283529.0,
      "step": 26815
    },
    {
      "epoch": 1.6010028653295127,
      "grad_norm": 0.49653738737106323,
      "learning_rate": 9.723036859159117e-06,
      "loss": 0.4221,
      "mean_token_accuracy": 0.8589282393455505,
      "num_tokens": 898447005.0,
      "step": 26820
    },
    {
      "epoch": 1.6013013371537728,
      "grad_norm": 0.5347087979316711,
      "learning_rate": 9.716232121514316e-06,
      "loss": 0.4693,
      "mean_token_accuracy": 0.8447035551071167,
      "num_tokens": 898614685.0,
      "step": 26825
    },
    {
      "epoch": 1.6015998089780323,
      "grad_norm": 0.5218964219093323,
      "learning_rate": 9.709431715434445e-06,
      "loss": 0.409,
      "mean_token_accuracy": 0.8617738366127015,
      "num_tokens": 898782365.0,
      "step": 26830
    },
    {
      "epoch": 1.6018982808022924,
      "grad_norm": 0.516520082950592,
      "learning_rate": 9.70263564257585e-06,
      "loss": 0.4309,
      "mean_token_accuracy": 0.8550817251205445,
      "num_tokens": 898950045.0,
      "step": 26835
    },
    {
      "epoch": 1.602196752626552,
      "grad_norm": 0.5449186563491821,
      "learning_rate": 9.695843904593843e-06,
      "loss": 0.4246,
      "mean_token_accuracy": 0.8581295490264893,
      "num_tokens": 899117725.0,
      "step": 26840
    },
    {
      "epoch": 1.602495224450812,
      "grad_norm": 0.49110421538352966,
      "learning_rate": 9.689056503142683e-06,
      "loss": 0.4495,
      "mean_token_accuracy": 0.8502743601799011,
      "num_tokens": 899285405.0,
      "step": 26845
    },
    {
      "epoch": 1.6027936962750715,
      "grad_norm": 0.5249834656715393,
      "learning_rate": 9.682273439875565e-06,
      "loss": 0.4201,
      "mean_token_accuracy": 0.8587916016578674,
      "num_tokens": 899453085.0,
      "step": 26850
    },
    {
      "epoch": 1.6030921680993315,
      "grad_norm": 0.5787405967712402,
      "learning_rate": 9.675494716444634e-06,
      "loss": 0.4196,
      "mean_token_accuracy": 0.85933438539505,
      "num_tokens": 899620765.0,
      "step": 26855
    },
    {
      "epoch": 1.603390639923591,
      "grad_norm": 0.5511835217475891,
      "learning_rate": 9.668720334500959e-06,
      "loss": 0.4709,
      "mean_token_accuracy": 0.8446797132492065,
      "num_tokens": 899788445.0,
      "step": 26860
    },
    {
      "epoch": 1.6036891117478511,
      "grad_norm": 0.479943186044693,
      "learning_rate": 9.661950295694575e-06,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8601156949996949,
      "num_tokens": 899956125.0,
      "step": 26865
    },
    {
      "epoch": 1.6039875835721107,
      "grad_norm": 0.5228937864303589,
      "learning_rate": 9.65518460167445e-06,
      "loss": 0.4398,
      "mean_token_accuracy": 0.8517773985862732,
      "num_tokens": 900123805.0,
      "step": 26870
    },
    {
      "epoch": 1.6042860553963707,
      "grad_norm": 0.4709329307079315,
      "learning_rate": 9.648423254088493e-06,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8529285311698913,
      "num_tokens": 900291485.0,
      "step": 26875
    },
    {
      "epoch": 1.6045845272206303,
      "grad_norm": 0.5314335823059082,
      "learning_rate": 9.64166625458356e-06,
      "loss": 0.4206,
      "mean_token_accuracy": 0.8573959231376648,
      "num_tokens": 900459165.0,
      "step": 26880
    },
    {
      "epoch": 1.6048829990448903,
      "grad_norm": 0.5271437764167786,
      "learning_rate": 9.634913604805426e-06,
      "loss": 0.3926,
      "mean_token_accuracy": 0.8686866283416748,
      "num_tokens": 900626845.0,
      "step": 26885
    },
    {
      "epoch": 1.6051814708691499,
      "grad_norm": 0.5638481378555298,
      "learning_rate": 9.62816530639884e-06,
      "loss": 0.429,
      "mean_token_accuracy": 0.8561433792114258,
      "num_tokens": 900794525.0,
      "step": 26890
    },
    {
      "epoch": 1.6054799426934099,
      "grad_norm": 0.5208526849746704,
      "learning_rate": 9.62142136100748e-06,
      "loss": 0.4191,
      "mean_token_accuracy": 0.8589824557304382,
      "num_tokens": 900962205.0,
      "step": 26895
    },
    {
      "epoch": 1.6057784145176695,
      "grad_norm": 0.4979512393474579,
      "learning_rate": 9.614681770273938e-06,
      "loss": 0.4206,
      "mean_token_accuracy": 0.8592866420745849,
      "num_tokens": 901129885.0,
      "step": 26900
    },
    {
      "epoch": 1.6060768863419295,
      "grad_norm": 0.49903905391693115,
      "learning_rate": 9.607946535839779e-06,
      "loss": 0.4445,
      "mean_token_accuracy": 0.8520637035369873,
      "num_tokens": 901297565.0,
      "step": 26905
    },
    {
      "epoch": 1.606375358166189,
      "grad_norm": 0.4726596176624298,
      "learning_rate": 9.60121565934549e-06,
      "loss": 0.417,
      "mean_token_accuracy": 0.859799587726593,
      "num_tokens": 901465245.0,
      "step": 26910
    },
    {
      "epoch": 1.606673829990449,
      "grad_norm": 0.5235370993614197,
      "learning_rate": 9.594489142430504e-06,
      "loss": 0.4423,
      "mean_token_accuracy": 0.852004063129425,
      "num_tokens": 901632925.0,
      "step": 26915
    },
    {
      "epoch": 1.6069723018147086,
      "grad_norm": 0.4798019528388977,
      "learning_rate": 9.587766986733182e-06,
      "loss": 0.4396,
      "mean_token_accuracy": 0.8529047012329102,
      "num_tokens": 901800605.0,
      "step": 26920
    },
    {
      "epoch": 1.6072707736389686,
      "grad_norm": 0.5040823221206665,
      "learning_rate": 9.581049193890832e-06,
      "loss": 0.4237,
      "mean_token_accuracy": 0.8580758810043335,
      "num_tokens": 901968285.0,
      "step": 26925
    },
    {
      "epoch": 1.6075692454632282,
      "grad_norm": 0.49416518211364746,
      "learning_rate": 9.574335765539697e-06,
      "loss": 0.4506,
      "mean_token_accuracy": 0.8500179052352905,
      "num_tokens": 902135965.0,
      "step": 26930
    },
    {
      "epoch": 1.607867717287488,
      "grad_norm": 0.5431501865386963,
      "learning_rate": 9.567626703314955e-06,
      "loss": 0.4664,
      "mean_token_accuracy": 0.8454908847808837,
      "num_tokens": 902303645.0,
      "step": 26935
    },
    {
      "epoch": 1.6081661891117478,
      "grad_norm": 0.4754214882850647,
      "learning_rate": 9.560922008850729e-06,
      "loss": 0.4107,
      "mean_token_accuracy": 0.8612907290458679,
      "num_tokens": 902471325.0,
      "step": 26940
    },
    {
      "epoch": 1.6084646609360076,
      "grad_norm": 0.4518228769302368,
      "learning_rate": 9.554221683780059e-06,
      "loss": 0.4558,
      "mean_token_accuracy": 0.8502803206443786,
      "num_tokens": 902639005.0,
      "step": 26945
    },
    {
      "epoch": 1.6087631327602674,
      "grad_norm": 0.4833535850048065,
      "learning_rate": 9.547525729734933e-06,
      "loss": 0.3867,
      "mean_token_accuracy": 0.869873559474945,
      "num_tokens": 902806685.0,
      "step": 26950
    },
    {
      "epoch": 1.6090616045845272,
      "grad_norm": 0.4745807647705078,
      "learning_rate": 9.540834148346294e-06,
      "loss": 0.436,
      "mean_token_accuracy": 0.8546074509620667,
      "num_tokens": 902972867.0,
      "step": 26955
    },
    {
      "epoch": 1.609360076408787,
      "grad_norm": 0.4893713891506195,
      "learning_rate": 9.53414694124398e-06,
      "loss": 0.4579,
      "mean_token_accuracy": 0.8480615496635437,
      "num_tokens": 903140547.0,
      "step": 26960
    },
    {
      "epoch": 1.6096585482330468,
      "grad_norm": 0.5143224596977234,
      "learning_rate": 9.527464110056795e-06,
      "loss": 0.4142,
      "mean_token_accuracy": 0.86092689037323,
      "num_tokens": 903308227.0,
      "step": 26965
    },
    {
      "epoch": 1.6099570200573066,
      "grad_norm": 0.5242037177085876,
      "learning_rate": 9.520785656412451e-06,
      "loss": 0.4504,
      "mean_token_accuracy": 0.8502564668655396,
      "num_tokens": 903475907.0,
      "step": 26970
    },
    {
      "epoch": 1.6102554918815664,
      "grad_norm": 0.49313151836395264,
      "learning_rate": 9.514111581937624e-06,
      "loss": 0.4383,
      "mean_token_accuracy": 0.8542943954467773,
      "num_tokens": 903643587.0,
      "step": 26975
    },
    {
      "epoch": 1.6105539637058262,
      "grad_norm": 0.5468960404396057,
      "learning_rate": 9.507441888257909e-06,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8423893690109253,
      "num_tokens": 903811267.0,
      "step": 26980
    },
    {
      "epoch": 1.610852435530086,
      "grad_norm": 0.4903563857078552,
      "learning_rate": 9.500776576997817e-06,
      "loss": 0.4368,
      "mean_token_accuracy": 0.8536144614219665,
      "num_tokens": 903978947.0,
      "step": 26985
    },
    {
      "epoch": 1.6111509073543457,
      "grad_norm": 0.4625544846057892,
      "learning_rate": 9.49411564978082e-06,
      "loss": 0.3955,
      "mean_token_accuracy": 0.8665871500968934,
      "num_tokens": 904146627.0,
      "step": 26990
    },
    {
      "epoch": 1.6114493791786055,
      "grad_norm": 0.5248073935508728,
      "learning_rate": 9.487459108229308e-06,
      "loss": 0.4636,
      "mean_token_accuracy": 0.8441965937614441,
      "num_tokens": 904314307.0,
      "step": 26995
    },
    {
      "epoch": 1.6117478510028653,
      "grad_norm": 0.5223250985145569,
      "learning_rate": 9.480806953964603e-06,
      "loss": 0.4274,
      "mean_token_accuracy": 0.8551348924636841,
      "num_tokens": 904475803.0,
      "step": 27000
    },
    {
      "epoch": 1.6120463228271251,
      "grad_norm": 0.5437977910041809,
      "learning_rate": 9.474159188606966e-06,
      "loss": 0.4713,
      "mean_token_accuracy": 0.8438267946243286,
      "num_tokens": 904643483.0,
      "step": 27005
    },
    {
      "epoch": 1.612344794651385,
      "grad_norm": 0.5392799377441406,
      "learning_rate": 9.467515813775565e-06,
      "loss": 0.4689,
      "mean_token_accuracy": 0.8446916460990905,
      "num_tokens": 904811163.0,
      "step": 27010
    },
    {
      "epoch": 1.6126432664756447,
      "grad_norm": 0.5598256587982178,
      "learning_rate": 9.460876831088542e-06,
      "loss": 0.4328,
      "mean_token_accuracy": 0.8536442875862121,
      "num_tokens": 904978843.0,
      "step": 27015
    },
    {
      "epoch": 1.6129417382999045,
      "grad_norm": 0.5238668322563171,
      "learning_rate": 9.454242242162925e-06,
      "loss": 0.4847,
      "mean_token_accuracy": 0.8389538288116455,
      "num_tokens": 905146523.0,
      "step": 27020
    },
    {
      "epoch": 1.6132402101241643,
      "grad_norm": 0.5520257353782654,
      "learning_rate": 9.447612048614699e-06,
      "loss": 0.4667,
      "mean_token_accuracy": 0.8442264199256897,
      "num_tokens": 905314203.0,
      "step": 27025
    },
    {
      "epoch": 1.613538681948424,
      "grad_norm": 0.5324056148529053,
      "learning_rate": 9.44098625205876e-06,
      "loss": 0.4379,
      "mean_token_accuracy": 0.8533102750778199,
      "num_tokens": 905481883.0,
      "step": 27030
    },
    {
      "epoch": 1.6138371537726839,
      "grad_norm": 0.4806233048439026,
      "learning_rate": 9.434364854108954e-06,
      "loss": 0.4011,
      "mean_token_accuracy": 0.8637361288070678,
      "num_tokens": 905649563.0,
      "step": 27035
    },
    {
      "epoch": 1.6141356255969437,
      "grad_norm": 0.48348575830459595,
      "learning_rate": 9.42774785637805e-06,
      "loss": 0.428,
      "mean_token_accuracy": 0.8563103914260864,
      "num_tokens": 905817243.0,
      "step": 27040
    },
    {
      "epoch": 1.6144340974212035,
      "grad_norm": 0.47034746408462524,
      "learning_rate": 9.421135260477723e-06,
      "loss": 0.3855,
      "mean_token_accuracy": 0.8697304010391236,
      "num_tokens": 905984923.0,
      "step": 27045
    },
    {
      "epoch": 1.6147325692454633,
      "grad_norm": 0.5088317394256592,
      "learning_rate": 9.4145270680186e-06,
      "loss": 0.4467,
      "mean_token_accuracy": 0.8514016389846801,
      "num_tokens": 906152603.0,
      "step": 27050
    },
    {
      "epoch": 1.615031041069723,
      "grad_norm": 0.576236367225647,
      "learning_rate": 9.407923280610231e-06,
      "loss": 0.4463,
      "mean_token_accuracy": 0.8515030384063721,
      "num_tokens": 906320283.0,
      "step": 27055
    },
    {
      "epoch": 1.6153295128939829,
      "grad_norm": 0.5178862810134888,
      "learning_rate": 9.401323899861088e-06,
      "loss": 0.4218,
      "mean_token_accuracy": 0.8584814429283142,
      "num_tokens": 906487963.0,
      "step": 27060
    },
    {
      "epoch": 1.6156279847182426,
      "grad_norm": 0.48196008801460266,
      "learning_rate": 9.394728927378576e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8545568346977234,
      "num_tokens": 906655643.0,
      "step": 27065
    },
    {
      "epoch": 1.6159264565425024,
      "grad_norm": 0.4702968895435333,
      "learning_rate": 9.388138364769018e-06,
      "loss": 0.4063,
      "mean_token_accuracy": 0.8634379148483277,
      "num_tokens": 906823323.0,
      "step": 27070
    },
    {
      "epoch": 1.616224928366762,
      "grad_norm": 0.5784925818443298,
      "learning_rate": 9.381552213637664e-06,
      "loss": 0.4497,
      "mean_token_accuracy": 0.8507634401321411,
      "num_tokens": 906991003.0,
      "step": 27075
    },
    {
      "epoch": 1.616523400191022,
      "grad_norm": 0.5316598415374756,
      "learning_rate": 9.374970475588697e-06,
      "loss": 0.4606,
      "mean_token_accuracy": 0.847667396068573,
      "num_tokens": 907155008.0,
      "step": 27080
    },
    {
      "epoch": 1.6168218720152816,
      "grad_norm": 0.47320249676704407,
      "learning_rate": 9.368393152225222e-06,
      "loss": 0.4331,
      "mean_token_accuracy": 0.8547477006912232,
      "num_tokens": 907322688.0,
      "step": 27085
    },
    {
      "epoch": 1.6171203438395416,
      "grad_norm": 0.45936790108680725,
      "learning_rate": 9.36182024514927e-06,
      "loss": 0.4249,
      "mean_token_accuracy": 0.8595848679542542,
      "num_tokens": 907490368.0,
      "step": 27090
    },
    {
      "epoch": 1.6174188156638012,
      "grad_norm": 0.589867353439331,
      "learning_rate": 9.35525175596178e-06,
      "loss": 0.4496,
      "mean_token_accuracy": 0.8518787980079651,
      "num_tokens": 907658048.0,
      "step": 27095
    },
    {
      "epoch": 1.6177172874880612,
      "grad_norm": 0.5166635513305664,
      "learning_rate": 9.348687686262647e-06,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8537576079368592,
      "num_tokens": 907825728.0,
      "step": 27100
    },
    {
      "epoch": 1.6180157593123208,
      "grad_norm": 0.5179538130760193,
      "learning_rate": 9.34212803765066e-06,
      "loss": 0.4534,
      "mean_token_accuracy": 0.849862813949585,
      "num_tokens": 907993408.0,
      "step": 27105
    },
    {
      "epoch": 1.6183142311365808,
      "grad_norm": 0.5585004091262817,
      "learning_rate": 9.335572811723545e-06,
      "loss": 0.4367,
      "mean_token_accuracy": 0.8537874341011047,
      "num_tokens": 908161088.0,
      "step": 27110
    },
    {
      "epoch": 1.6186127029608404,
      "grad_norm": 0.5620519518852234,
      "learning_rate": 9.329022010077947e-06,
      "loss": 0.4987,
      "mean_token_accuracy": 0.83711678981781,
      "num_tokens": 908328768.0,
      "step": 27115
    },
    {
      "epoch": 1.6189111747851004,
      "grad_norm": 0.47741344571113586,
      "learning_rate": 9.322475634309436e-06,
      "loss": 0.3985,
      "mean_token_accuracy": 0.864881956577301,
      "num_tokens": 908491032.0,
      "step": 27120
    },
    {
      "epoch": 1.61920964660936,
      "grad_norm": 0.524983823299408,
      "learning_rate": 9.31593368601251e-06,
      "loss": 0.44,
      "mean_token_accuracy": 0.8528390645980835,
      "num_tokens": 908658712.0,
      "step": 27125
    },
    {
      "epoch": 1.61950811843362,
      "grad_norm": 0.5282403230667114,
      "learning_rate": 9.309396166780567e-06,
      "loss": 0.4693,
      "mean_token_accuracy": 0.8461290717124939,
      "num_tokens": 908826392.0,
      "step": 27130
    },
    {
      "epoch": 1.6198065902578795,
      "grad_norm": 0.5588261485099792,
      "learning_rate": 9.302863078205951e-06,
      "loss": 0.451,
      "mean_token_accuracy": 0.8486937880516052,
      "num_tokens": 908994072.0,
      "step": 27135
    },
    {
      "epoch": 1.6201050620821396,
      "grad_norm": 0.4812007546424866,
      "learning_rate": 9.29633442187991e-06,
      "loss": 0.4118,
      "mean_token_accuracy": 0.8601813197135926,
      "num_tokens": 909161752.0,
      "step": 27140
    },
    {
      "epoch": 1.6204035339063991,
      "grad_norm": 0.50464928150177,
      "learning_rate": 9.289810199392622e-06,
      "loss": 0.4374,
      "mean_token_accuracy": 0.853847086429596,
      "num_tokens": 909329432.0,
      "step": 27145
    },
    {
      "epoch": 1.6207020057306591,
      "grad_norm": 0.5393331050872803,
      "learning_rate": 9.283290412333188e-06,
      "loss": 0.4707,
      "mean_token_accuracy": 0.8434748768806457,
      "num_tokens": 909497112.0,
      "step": 27150
    },
    {
      "epoch": 1.6210004775549187,
      "grad_norm": 0.4586549401283264,
      "learning_rate": 9.276775062289606e-06,
      "loss": 0.4158,
      "mean_token_accuracy": 0.861266839504242,
      "num_tokens": 909664792.0,
      "step": 27155
    },
    {
      "epoch": 1.6212989493791787,
      "grad_norm": 0.5473030209541321,
      "learning_rate": 9.27026415084883e-06,
      "loss": 0.4615,
      "mean_token_accuracy": 0.8463318586349488,
      "num_tokens": 909832472.0,
      "step": 27160
    },
    {
      "epoch": 1.6215974212034383,
      "grad_norm": 0.5064117312431335,
      "learning_rate": 9.263757679596694e-06,
      "loss": 0.4409,
      "mean_token_accuracy": 0.8542049288749695,
      "num_tokens": 910000152.0,
      "step": 27165
    },
    {
      "epoch": 1.6218958930276983,
      "grad_norm": 0.498205304145813,
      "learning_rate": 9.257255650117983e-06,
      "loss": 0.4049,
      "mean_token_accuracy": 0.8642908215522767,
      "num_tokens": 910167832.0,
      "step": 27170
    },
    {
      "epoch": 1.622194364851958,
      "grad_norm": 0.5261074900627136,
      "learning_rate": 9.250758063996376e-06,
      "loss": 0.4502,
      "mean_token_accuracy": 0.8519205570220947,
      "num_tokens": 910335512.0,
      "step": 27175
    },
    {
      "epoch": 1.622492836676218,
      "grad_norm": 0.5457990765571594,
      "learning_rate": 9.244264922814487e-06,
      "loss": 0.4497,
      "mean_token_accuracy": 0.8492544293403625,
      "num_tokens": 910503192.0,
      "step": 27180
    },
    {
      "epoch": 1.6227913085004775,
      "grad_norm": 0.5503063201904297,
      "learning_rate": 9.237776228153845e-06,
      "loss": 0.4403,
      "mean_token_accuracy": 0.8519623041152954,
      "num_tokens": 910670872.0,
      "step": 27185
    },
    {
      "epoch": 1.6230897803247375,
      "grad_norm": 0.5159018635749817,
      "learning_rate": 9.231291981594874e-06,
      "loss": 0.4118,
      "mean_token_accuracy": 0.8600083589553833,
      "num_tokens": 910838552.0,
      "step": 27190
    },
    {
      "epoch": 1.623388252148997,
      "grad_norm": 0.4987584948539734,
      "learning_rate": 9.22481218471695e-06,
      "loss": 0.4033,
      "mean_token_accuracy": 0.863873302936554,
      "num_tokens": 911006232.0,
      "step": 27195
    },
    {
      "epoch": 1.623686723973257,
      "grad_norm": 0.4894299805164337,
      "learning_rate": 9.218336839098335e-06,
      "loss": 0.4106,
      "mean_token_accuracy": 0.8630144357681274,
      "num_tokens": 911173912.0,
      "step": 27200
    },
    {
      "epoch": 1.6239851957975167,
      "grad_norm": 0.4829419255256653,
      "learning_rate": 9.211865946316223e-06,
      "loss": 0.3953,
      "mean_token_accuracy": 0.8659012198448182,
      "num_tokens": 911341592.0,
      "step": 27205
    },
    {
      "epoch": 1.6242836676217765,
      "grad_norm": 0.6027539372444153,
      "learning_rate": 9.205399507946726e-06,
      "loss": 0.4705,
      "mean_token_accuracy": 0.8444888472557068,
      "num_tokens": 911509272.0,
      "step": 27210
    },
    {
      "epoch": 1.6245821394460362,
      "grad_norm": 0.6368187069892883,
      "learning_rate": 9.198937525564856e-06,
      "loss": 0.4501,
      "mean_token_accuracy": 0.849874746799469,
      "num_tokens": 911676952.0,
      "step": 27215
    },
    {
      "epoch": 1.624880611270296,
      "grad_norm": 0.5282248258590698,
      "learning_rate": 9.192480000744544e-06,
      "loss": 0.4278,
      "mean_token_accuracy": 0.8571215748786927,
      "num_tokens": 911844632.0,
      "step": 27220
    },
    {
      "epoch": 1.6251790830945558,
      "grad_norm": 0.4703761041164398,
      "learning_rate": 9.186026935058657e-06,
      "loss": 0.4315,
      "mean_token_accuracy": 0.8558928728103637,
      "num_tokens": 912012312.0,
      "step": 27225
    },
    {
      "epoch": 1.6254775549188156,
      "grad_norm": 0.47730252146720886,
      "learning_rate": 9.179578330078939e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.859060001373291,
      "num_tokens": 912179992.0,
      "step": 27230
    },
    {
      "epoch": 1.6257760267430754,
      "grad_norm": 0.4958620071411133,
      "learning_rate": 9.173134187376078e-06,
      "loss": 0.4372,
      "mean_token_accuracy": 0.8554276466369629,
      "num_tokens": 912347672.0,
      "step": 27235
    },
    {
      "epoch": 1.6260744985673352,
      "grad_norm": 0.47102028131484985,
      "learning_rate": 9.166694508519659e-06,
      "loss": 0.4278,
      "mean_token_accuracy": 0.856316351890564,
      "num_tokens": 912515352.0,
      "step": 27240
    },
    {
      "epoch": 1.626372970391595,
      "grad_norm": 0.4946134388446808,
      "learning_rate": 9.160259295078193e-06,
      "loss": 0.4504,
      "mean_token_accuracy": 0.849749481678009,
      "num_tokens": 912683032.0,
      "step": 27245
    },
    {
      "epoch": 1.6266714422158548,
      "grad_norm": 0.5380204319953918,
      "learning_rate": 9.153828548619085e-06,
      "loss": 0.4511,
      "mean_token_accuracy": 0.8505845189094543,
      "num_tokens": 912850712.0,
      "step": 27250
    },
    {
      "epoch": 1.6269699140401146,
      "grad_norm": 0.4733681082725525,
      "learning_rate": 9.147402270708662e-06,
      "loss": 0.4287,
      "mean_token_accuracy": 0.8559346318244934,
      "num_tokens": 913018392.0,
      "step": 27255
    },
    {
      "epoch": 1.6272683858643744,
      "grad_norm": 0.4521208107471466,
      "learning_rate": 9.140980462912165e-06,
      "loss": 0.3925,
      "mean_token_accuracy": 0.867141842842102,
      "num_tokens": 913186072.0,
      "step": 27260
    },
    {
      "epoch": 1.6275668576886342,
      "grad_norm": 0.5736082196235657,
      "learning_rate": 9.134563126793749e-06,
      "loss": 0.4255,
      "mean_token_accuracy": 0.8585112690925598,
      "num_tokens": 913353752.0,
      "step": 27265
    },
    {
      "epoch": 1.627865329512894,
      "grad_norm": 0.49521011114120483,
      "learning_rate": 9.128150263916474e-06,
      "loss": 0.4516,
      "mean_token_accuracy": 0.8474114298820495,
      "num_tokens": 913521432.0,
      "step": 27270
    },
    {
      "epoch": 1.6281638013371538,
      "grad_norm": 0.5299897789955139,
      "learning_rate": 9.121741875842303e-06,
      "loss": 0.424,
      "mean_token_accuracy": 0.8581056952476501,
      "num_tokens": 913689112.0,
      "step": 27275
    },
    {
      "epoch": 1.6284622731614136,
      "grad_norm": 0.560864269733429,
      "learning_rate": 9.115337964132115e-06,
      "loss": 0.4318,
      "mean_token_accuracy": 0.8541274070739746,
      "num_tokens": 913856792.0,
      "step": 27280
    },
    {
      "epoch": 1.6287607449856734,
      "grad_norm": 0.5350287556648254,
      "learning_rate": 9.10893853034572e-06,
      "loss": 0.4502,
      "mean_token_accuracy": 0.8483597755432128,
      "num_tokens": 914024472.0,
      "step": 27285
    },
    {
      "epoch": 1.6290592168099332,
      "grad_norm": 0.5666098594665527,
      "learning_rate": 9.1025435760418e-06,
      "loss": 0.4473,
      "mean_token_accuracy": 0.8494654893875122,
      "num_tokens": 914191795.0,
      "step": 27290
    },
    {
      "epoch": 1.629357688634193,
      "grad_norm": 0.51736980676651,
      "learning_rate": 9.096153102777977e-06,
      "loss": 0.4498,
      "mean_token_accuracy": 0.850727653503418,
      "num_tokens": 914359475.0,
      "step": 27295
    },
    {
      "epoch": 1.6296561604584527,
      "grad_norm": 0.46848005056381226,
      "learning_rate": 9.089767112110747e-06,
      "loss": 0.448,
      "mean_token_accuracy": 0.8509125709533691,
      "num_tokens": 914527155.0,
      "step": 27300
    },
    {
      "epoch": 1.6299546322827125,
      "grad_norm": 0.49154558777809143,
      "learning_rate": 9.08338560559556e-06,
      "loss": 0.4371,
      "mean_token_accuracy": 0.8552069783210754,
      "num_tokens": 914694835.0,
      "step": 27305
    },
    {
      "epoch": 1.6302531041069723,
      "grad_norm": 0.5288469195365906,
      "learning_rate": 9.077008584786744e-06,
      "loss": 0.4607,
      "mean_token_accuracy": 0.8457175254821777,
      "num_tokens": 914862515.0,
      "step": 27310
    },
    {
      "epoch": 1.6305515759312321,
      "grad_norm": 0.4710169732570648,
      "learning_rate": 9.070636051237536e-06,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8464094042778015,
      "num_tokens": 915030195.0,
      "step": 27315
    },
    {
      "epoch": 1.630850047755492,
      "grad_norm": 0.5431186556816101,
      "learning_rate": 9.06426800650008e-06,
      "loss": 0.4478,
      "mean_token_accuracy": 0.851491117477417,
      "num_tokens": 915197875.0,
      "step": 27320
    },
    {
      "epoch": 1.6311485195797517,
      "grad_norm": 0.5576286315917969,
      "learning_rate": 9.057904452125438e-06,
      "loss": 0.4797,
      "mean_token_accuracy": 0.8407789587974548,
      "num_tokens": 915365555.0,
      "step": 27325
    },
    {
      "epoch": 1.6314469914040115,
      "grad_norm": 0.5282570123672485,
      "learning_rate": 9.051545389663566e-06,
      "loss": 0.4246,
      "mean_token_accuracy": 0.8584039211273193,
      "num_tokens": 915533235.0,
      "step": 27330
    },
    {
      "epoch": 1.6317454632282713,
      "grad_norm": 0.4832616448402405,
      "learning_rate": 9.045190820663342e-06,
      "loss": 0.4055,
      "mean_token_accuracy": 0.86303231716156,
      "num_tokens": 915700915.0,
      "step": 27335
    },
    {
      "epoch": 1.632043935052531,
      "grad_norm": 0.46541497111320496,
      "learning_rate": 9.038840746672523e-06,
      "loss": 0.4409,
      "mean_token_accuracy": 0.8526482224464417,
      "num_tokens": 915868595.0,
      "step": 27340
    },
    {
      "epoch": 1.6323424068767909,
      "grad_norm": 0.4830757677555084,
      "learning_rate": 9.032495169237794e-06,
      "loss": 0.4126,
      "mean_token_accuracy": 0.8616068243980408,
      "num_tokens": 916036275.0,
      "step": 27345
    },
    {
      "epoch": 1.6326408787010505,
      "grad_norm": 0.48255184292793274,
      "learning_rate": 9.026154089904737e-06,
      "loss": 0.4275,
      "mean_token_accuracy": 0.855815327167511,
      "num_tokens": 916203955.0,
      "step": 27350
    },
    {
      "epoch": 1.6329393505253105,
      "grad_norm": 0.5121918320655823,
      "learning_rate": 9.019817510217849e-06,
      "loss": 0.4352,
      "mean_token_accuracy": 0.8541751265525818,
      "num_tokens": 916371635.0,
      "step": 27355
    },
    {
      "epoch": 1.63323782234957,
      "grad_norm": 0.5057121515274048,
      "learning_rate": 9.013485431720498e-06,
      "loss": 0.4517,
      "mean_token_accuracy": 0.8485446572303772,
      "num_tokens": 916539315.0,
      "step": 27360
    },
    {
      "epoch": 1.63353629417383,
      "grad_norm": 0.5022083520889282,
      "learning_rate": 9.007157855954997e-06,
      "loss": 0.4193,
      "mean_token_accuracy": 0.8576404571533203,
      "num_tokens": 916706995.0,
      "step": 27365
    },
    {
      "epoch": 1.6338347659980896,
      "grad_norm": 0.5485715270042419,
      "learning_rate": 9.000834784462544e-06,
      "loss": 0.4334,
      "mean_token_accuracy": 0.8532386898994446,
      "num_tokens": 916874675.0,
      "step": 27370
    },
    {
      "epoch": 1.6341332378223496,
      "grad_norm": 0.5028972625732422,
      "learning_rate": 8.994516218783232e-06,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8468030571937561,
      "num_tokens": 917042355.0,
      "step": 27375
    },
    {
      "epoch": 1.6344317096466092,
      "grad_norm": 0.45500344038009644,
      "learning_rate": 8.988202160456064e-06,
      "loss": 0.4429,
      "mean_token_accuracy": 0.8534891963005066,
      "num_tokens": 917210035.0,
      "step": 27380
    },
    {
      "epoch": 1.6347301814708692,
      "grad_norm": 0.4935471713542938,
      "learning_rate": 8.981892611018949e-06,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8555827260017395,
      "num_tokens": 917377715.0,
      "step": 27385
    },
    {
      "epoch": 1.6350286532951288,
      "grad_norm": 0.4806463420391083,
      "learning_rate": 8.975587572008695e-06,
      "loss": 0.4145,
      "mean_token_accuracy": 0.8592150807380676,
      "num_tokens": 917545395.0,
      "step": 27390
    },
    {
      "epoch": 1.6353271251193888,
      "grad_norm": 0.5478640794754028,
      "learning_rate": 8.969287044961014e-06,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8494751214981079,
      "num_tokens": 917713075.0,
      "step": 27395
    },
    {
      "epoch": 1.6356255969436484,
      "grad_norm": 0.49986085295677185,
      "learning_rate": 8.962991031410507e-06,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8484313488006592,
      "num_tokens": 917880755.0,
      "step": 27400
    },
    {
      "epoch": 1.6359240687679084,
      "grad_norm": 0.5525385737419128,
      "learning_rate": 8.956699532890683e-06,
      "loss": 0.4286,
      "mean_token_accuracy": 0.856226897239685,
      "num_tokens": 918048435.0,
      "step": 27405
    },
    {
      "epoch": 1.636222540592168,
      "grad_norm": 0.4970659911632538,
      "learning_rate": 8.950412550933964e-06,
      "loss": 0.4323,
      "mean_token_accuracy": 0.8539961814880371,
      "num_tokens": 918216115.0,
      "step": 27410
    },
    {
      "epoch": 1.636521012416428,
      "grad_norm": 0.5140109658241272,
      "learning_rate": 8.944130087071651e-06,
      "loss": 0.4462,
      "mean_token_accuracy": 0.8535548210144043,
      "num_tokens": 918383795.0,
      "step": 27415
    },
    {
      "epoch": 1.6368194842406876,
      "grad_norm": 0.5214282274246216,
      "learning_rate": 8.937852142833963e-06,
      "loss": 0.4135,
      "mean_token_accuracy": 0.8604795575141907,
      "num_tokens": 918551475.0,
      "step": 27420
    },
    {
      "epoch": 1.6371179560649476,
      "grad_norm": 0.4619886875152588,
      "learning_rate": 8.931578719749997e-06,
      "loss": 0.4105,
      "mean_token_accuracy": 0.8627997159957885,
      "num_tokens": 918719155.0,
      "step": 27425
    },
    {
      "epoch": 1.6374164278892072,
      "grad_norm": 0.5684396624565125,
      "learning_rate": 8.925309819347774e-06,
      "loss": 0.4309,
      "mean_token_accuracy": 0.8536502480506897,
      "num_tokens": 918886835.0,
      "step": 27430
    },
    {
      "epoch": 1.6377148997134672,
      "grad_norm": 0.5223966836929321,
      "learning_rate": 8.919045443154189e-06,
      "loss": 0.4161,
      "mean_token_accuracy": 0.859298586845398,
      "num_tokens": 919054515.0,
      "step": 27435
    },
    {
      "epoch": 1.6380133715377267,
      "grad_norm": 0.5374763011932373,
      "learning_rate": 8.912785592695053e-06,
      "loss": 0.46,
      "mean_token_accuracy": 0.8458368182182312,
      "num_tokens": 919222195.0,
      "step": 27440
    },
    {
      "epoch": 1.6383118433619868,
      "grad_norm": 0.49368685483932495,
      "learning_rate": 8.906530269495064e-06,
      "loss": 0.4249,
      "mean_token_accuracy": 0.8565191507339478,
      "num_tokens": 919389875.0,
      "step": 27445
    },
    {
      "epoch": 1.6386103151862463,
      "grad_norm": 0.49564072489738464,
      "learning_rate": 8.900279475077827e-06,
      "loss": 0.4105,
      "mean_token_accuracy": 0.8620123982429504,
      "num_tokens": 919557555.0,
      "step": 27450
    },
    {
      "epoch": 1.6389087870105064,
      "grad_norm": 0.48721274733543396,
      "learning_rate": 8.894033210965838e-06,
      "loss": 0.4196,
      "mean_token_accuracy": 0.8584396958351135,
      "num_tokens": 919725235.0,
      "step": 27455
    },
    {
      "epoch": 1.639207258834766,
      "grad_norm": 0.6268237829208374,
      "learning_rate": 8.887791478680484e-06,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8544912338256836,
      "num_tokens": 919892915.0,
      "step": 27460
    },
    {
      "epoch": 1.639505730659026,
      "grad_norm": 0.555050790309906,
      "learning_rate": 8.881554279742056e-06,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8354646444320679,
      "num_tokens": 920060595.0,
      "step": 27465
    },
    {
      "epoch": 1.6398042024832855,
      "grad_norm": 0.4681052267551422,
      "learning_rate": 8.87532161566974e-06,
      "loss": 0.4174,
      "mean_token_accuracy": 0.8597817063331604,
      "num_tokens": 920228275.0,
      "step": 27470
    },
    {
      "epoch": 1.6401026743075455,
      "grad_norm": 0.473096638917923,
      "learning_rate": 8.869093487981619e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8590063214302063,
      "num_tokens": 920395955.0,
      "step": 27475
    },
    {
      "epoch": 1.640401146131805,
      "grad_norm": 0.506682813167572,
      "learning_rate": 8.86286989819467e-06,
      "loss": 0.4315,
      "mean_token_accuracy": 0.855839192867279,
      "num_tokens": 920563635.0,
      "step": 27480
    },
    {
      "epoch": 1.640699617956065,
      "grad_norm": 0.5237351655960083,
      "learning_rate": 8.85665084782475e-06,
      "loss": 0.4367,
      "mean_token_accuracy": 0.8536741137504578,
      "num_tokens": 920731315.0,
      "step": 27485
    },
    {
      "epoch": 1.6409980897803247,
      "grad_norm": 0.473579466342926,
      "learning_rate": 8.850436338386641e-06,
      "loss": 0.4115,
      "mean_token_accuracy": 0.8609507441520691,
      "num_tokens": 920898995.0,
      "step": 27490
    },
    {
      "epoch": 1.6412965616045845,
      "grad_norm": 0.4849102795124054,
      "learning_rate": 8.844226371393988e-06,
      "loss": 0.386,
      "mean_token_accuracy": 0.8686150550842285,
      "num_tokens": 921066675.0,
      "step": 27495
    },
    {
      "epoch": 1.6415950334288443,
      "grad_norm": 0.5197857022285461,
      "learning_rate": 8.838020948359352e-06,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8496838927268981,
      "num_tokens": 921234355.0,
      "step": 27500
    },
    {
      "epoch": 1.641893505253104,
      "grad_norm": 0.5411177277565002,
      "learning_rate": 8.831820070794169e-06,
      "loss": 0.3957,
      "mean_token_accuracy": 0.8665096044540406,
      "num_tokens": 921402035.0,
      "step": 27505
    },
    {
      "epoch": 1.6421919770773639,
      "grad_norm": 0.5621386170387268,
      "learning_rate": 8.825623740208788e-06,
      "loss": 0.4498,
      "mean_token_accuracy": 0.8501968264579773,
      "num_tokens": 921569715.0,
      "step": 27510
    },
    {
      "epoch": 1.6424904489016237,
      "grad_norm": 0.493028461933136,
      "learning_rate": 8.81943195811244e-06,
      "loss": 0.4198,
      "mean_token_accuracy": 0.858570909500122,
      "num_tokens": 921737395.0,
      "step": 27515
    },
    {
      "epoch": 1.6427889207258835,
      "grad_norm": 0.4988255500793457,
      "learning_rate": 8.813244726013233e-06,
      "loss": 0.409,
      "mean_token_accuracy": 0.861916983127594,
      "num_tokens": 921905075.0,
      "step": 27520
    },
    {
      "epoch": 1.6430873925501432,
      "grad_norm": 0.5710336565971375,
      "learning_rate": 8.807062045418191e-06,
      "loss": 0.4685,
      "mean_token_accuracy": 0.8433794617652893,
      "num_tokens": 922072755.0,
      "step": 27525
    },
    {
      "epoch": 1.643385864374403,
      "grad_norm": 0.4894435405731201,
      "learning_rate": 8.800883917833222e-06,
      "loss": 0.4221,
      "mean_token_accuracy": 0.8586842298507691,
      "num_tokens": 922240435.0,
      "step": 27530
    },
    {
      "epoch": 1.6436843361986628,
      "grad_norm": 0.5521994829177856,
      "learning_rate": 8.794710344763121e-06,
      "loss": 0.4585,
      "mean_token_accuracy": 0.8486639618873596,
      "num_tokens": 922408115.0,
      "step": 27535
    },
    {
      "epoch": 1.6439828080229226,
      "grad_norm": 0.49032309651374817,
      "learning_rate": 8.788541327711583e-06,
      "loss": 0.4351,
      "mean_token_accuracy": 0.8541930079460144,
      "num_tokens": 922575795.0,
      "step": 27540
    },
    {
      "epoch": 1.6442812798471824,
      "grad_norm": 0.545734167098999,
      "learning_rate": 8.78237686818117e-06,
      "loss": 0.4354,
      "mean_token_accuracy": 0.8546164751052856,
      "num_tokens": 922743475.0,
      "step": 27545
    },
    {
      "epoch": 1.6445797516714422,
      "grad_norm": 0.498403936624527,
      "learning_rate": 8.776216967673356e-06,
      "loss": 0.4252,
      "mean_token_accuracy": 0.858558988571167,
      "num_tokens": 922911155.0,
      "step": 27550
    },
    {
      "epoch": 1.644878223495702,
      "grad_norm": 0.4923866391181946,
      "learning_rate": 8.770061627688508e-06,
      "loss": 0.4668,
      "mean_token_accuracy": 0.8452105402946473,
      "num_tokens": 923078835.0,
      "step": 27555
    },
    {
      "epoch": 1.6451766953199618,
      "grad_norm": 0.4732860028743744,
      "learning_rate": 8.763910849725865e-06,
      "loss": 0.427,
      "mean_token_accuracy": 0.8553560853004456,
      "num_tokens": 923246515.0,
      "step": 27560
    },
    {
      "epoch": 1.6454751671442216,
      "grad_norm": 0.5003758668899536,
      "learning_rate": 8.757764635283558e-06,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8487236022949218,
      "num_tokens": 923414195.0,
      "step": 27565
    },
    {
      "epoch": 1.6457736389684814,
      "grad_norm": 0.4619987905025482,
      "learning_rate": 8.75162298585862e-06,
      "loss": 0.3876,
      "mean_token_accuracy": 0.8686269879341125,
      "num_tokens": 923581875.0,
      "step": 27570
    },
    {
      "epoch": 1.6460721107927412,
      "grad_norm": 0.5236973762512207,
      "learning_rate": 8.745485902946961e-06,
      "loss": 0.4234,
      "mean_token_accuracy": 0.8577179908752441,
      "num_tokens": 923749555.0,
      "step": 27575
    },
    {
      "epoch": 1.646370582617001,
      "grad_norm": 0.46308356523513794,
      "learning_rate": 8.739353388043375e-06,
      "loss": 0.4631,
      "mean_token_accuracy": 0.8457115650177002,
      "num_tokens": 923917235.0,
      "step": 27580
    },
    {
      "epoch": 1.6466690544412608,
      "grad_norm": 0.4999522268772125,
      "learning_rate": 8.733225442641554e-06,
      "loss": 0.3951,
      "mean_token_accuracy": 0.8672670841217041,
      "num_tokens": 924084915.0,
      "step": 27585
    },
    {
      "epoch": 1.6469675262655206,
      "grad_norm": 0.5205977559089661,
      "learning_rate": 8.727102068234073e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.8607002258300781,
      "num_tokens": 924252595.0,
      "step": 27590
    },
    {
      "epoch": 1.6472659980897804,
      "grad_norm": 0.501131534576416,
      "learning_rate": 8.720983266312387e-06,
      "loss": 0.4438,
      "mean_token_accuracy": 0.8526839971542358,
      "num_tokens": 924420275.0,
      "step": 27595
    },
    {
      "epoch": 1.6475644699140402,
      "grad_norm": 0.5764347910881042,
      "learning_rate": 8.714869038366857e-06,
      "loss": 0.4745,
      "mean_token_accuracy": 0.8423476219177246,
      "num_tokens": 924587955.0,
      "step": 27600
    },
    {
      "epoch": 1.6478629417383,
      "grad_norm": 0.4761529266834259,
      "learning_rate": 8.708759385886698e-06,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8558034062385559,
      "num_tokens": 924755635.0,
      "step": 27605
    },
    {
      "epoch": 1.6481614135625597,
      "grad_norm": 0.48771607875823975,
      "learning_rate": 8.702654310360034e-06,
      "loss": 0.4554,
      "mean_token_accuracy": 0.8486460685729981,
      "num_tokens": 924923315.0,
      "step": 27610
    },
    {
      "epoch": 1.6484598853868195,
      "grad_norm": 0.558755099773407,
      "learning_rate": 8.69655381327388e-06,
      "loss": 0.4507,
      "mean_token_accuracy": 0.8506441593170166,
      "num_tokens": 925090995.0,
      "step": 27615
    },
    {
      "epoch": 1.6487583572110793,
      "grad_norm": 0.48419713973999023,
      "learning_rate": 8.690457896114115e-06,
      "loss": 0.4397,
      "mean_token_accuracy": 0.852982223033905,
      "num_tokens": 925258675.0,
      "step": 27620
    },
    {
      "epoch": 1.649056829035339,
      "grad_norm": 0.5028923153877258,
      "learning_rate": 8.684366560365515e-06,
      "loss": 0.4479,
      "mean_token_accuracy": 0.8506203174591065,
      "num_tokens": 925426355.0,
      "step": 27625
    },
    {
      "epoch": 1.649355300859599,
      "grad_norm": 0.5347084999084473,
      "learning_rate": 8.678279807511729e-06,
      "loss": 0.3942,
      "mean_token_accuracy": 0.8680066943168641,
      "num_tokens": 925594035.0,
      "step": 27630
    },
    {
      "epoch": 1.6496537726838585,
      "grad_norm": 0.5266854763031006,
      "learning_rate": 8.67219763903531e-06,
      "loss": 0.4157,
      "mean_token_accuracy": 0.8587916135787964,
      "num_tokens": 925761715.0,
      "step": 27635
    },
    {
      "epoch": 1.6499522445081185,
      "grad_norm": 0.4946543276309967,
      "learning_rate": 8.666120056417681e-06,
      "loss": 0.4575,
      "mean_token_accuracy": 0.8478885889053345,
      "num_tokens": 925929395.0,
      "step": 27640
    },
    {
      "epoch": 1.650250716332378,
      "grad_norm": 0.46847429871559143,
      "learning_rate": 8.66004706113914e-06,
      "loss": 0.424,
      "mean_token_accuracy": 0.8566444039344787,
      "num_tokens": 926097075.0,
      "step": 27645
    },
    {
      "epoch": 1.650549188156638,
      "grad_norm": 0.506936252117157,
      "learning_rate": 8.653978654678885e-06,
      "loss": 0.4663,
      "mean_token_accuracy": 0.8446975946426392,
      "num_tokens": 926264755.0,
      "step": 27650
    },
    {
      "epoch": 1.6508476599808977,
      "grad_norm": 0.49346089363098145,
      "learning_rate": 8.647914838514985e-06,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8550280451774597,
      "num_tokens": 926432435.0,
      "step": 27655
    },
    {
      "epoch": 1.6511461318051577,
      "grad_norm": 0.4977369010448456,
      "learning_rate": 8.641855614124397e-06,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8350590467453003,
      "num_tokens": 926600115.0,
      "step": 27660
    },
    {
      "epoch": 1.6514446036294173,
      "grad_norm": 0.48802804946899414,
      "learning_rate": 8.635800982982958e-06,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8562388181686401,
      "num_tokens": 926767795.0,
      "step": 27665
    },
    {
      "epoch": 1.6517430754536773,
      "grad_norm": 0.48658645153045654,
      "learning_rate": 8.62975094656538e-06,
      "loss": 0.444,
      "mean_token_accuracy": 0.8521412372589111,
      "num_tokens": 926935475.0,
      "step": 27670
    },
    {
      "epoch": 1.6520415472779368,
      "grad_norm": 0.5168094635009766,
      "learning_rate": 8.623705506345265e-06,
      "loss": 0.3963,
      "mean_token_accuracy": 0.8662173509597778,
      "num_tokens": 927103155.0,
      "step": 27675
    },
    {
      "epoch": 1.6523400191021969,
      "grad_norm": 0.502001941204071,
      "learning_rate": 8.61766466379509e-06,
      "loss": 0.3981,
      "mean_token_accuracy": 0.8663068056106568,
      "num_tokens": 927270835.0,
      "step": 27680
    },
    {
      "epoch": 1.6526384909264564,
      "grad_norm": 0.47451192140579224,
      "learning_rate": 8.611628420386217e-06,
      "loss": 0.4107,
      "mean_token_accuracy": 0.8618871569633484,
      "num_tokens": 927438515.0,
      "step": 27685
    },
    {
      "epoch": 1.6529369627507164,
      "grad_norm": 0.4913615584373474,
      "learning_rate": 8.60559677758887e-06,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8519324779510498,
      "num_tokens": 927606195.0,
      "step": 27690
    },
    {
      "epoch": 1.653235434574976,
      "grad_norm": 0.5603580474853516,
      "learning_rate": 8.599569736872187e-06,
      "loss": 0.4343,
      "mean_token_accuracy": 0.8542526602745056,
      "num_tokens": 927773875.0,
      "step": 27695
    },
    {
      "epoch": 1.653533906399236,
      "grad_norm": 0.47387024760246277,
      "learning_rate": 8.593547299704164e-06,
      "loss": 0.405,
      "mean_token_accuracy": 0.8633007287979126,
      "num_tokens": 927941555.0,
      "step": 27700
    },
    {
      "epoch": 1.6538323782234956,
      "grad_norm": 0.45370379090309143,
      "learning_rate": 8.587529467551658e-06,
      "loss": 0.4143,
      "mean_token_accuracy": 0.859948706626892,
      "num_tokens": 928109235.0,
      "step": 27705
    },
    {
      "epoch": 1.6541308500477556,
      "grad_norm": 0.5351717472076416,
      "learning_rate": 8.581516241880439e-06,
      "loss": 0.4458,
      "mean_token_accuracy": 0.8508529186248779,
      "num_tokens": 928276915.0,
      "step": 27710
    },
    {
      "epoch": 1.6544293218720152,
      "grad_norm": 0.49247390031814575,
      "learning_rate": 8.57550762415513e-06,
      "loss": 0.4148,
      "mean_token_accuracy": 0.8607300400733948,
      "num_tokens": 928444595.0,
      "step": 27715
    },
    {
      "epoch": 1.6547277936962752,
      "grad_norm": 0.5350109338760376,
      "learning_rate": 8.56950361583925e-06,
      "loss": 0.4478,
      "mean_token_accuracy": 0.8499940276145935,
      "num_tokens": 928612275.0,
      "step": 27720
    },
    {
      "epoch": 1.6550262655205348,
      "grad_norm": 0.5484545230865479,
      "learning_rate": 8.563504218395182e-06,
      "loss": 0.4396,
      "mean_token_accuracy": 0.8516998529434204,
      "num_tokens": 928779955.0,
      "step": 27725
    },
    {
      "epoch": 1.6553247373447948,
      "grad_norm": 0.5197243690490723,
      "learning_rate": 8.557509433284188e-06,
      "loss": 0.4114,
      "mean_token_accuracy": 0.8616783857345581,
      "num_tokens": 928947635.0,
      "step": 27730
    },
    {
      "epoch": 1.6556232091690544,
      "grad_norm": 0.5030907988548279,
      "learning_rate": 8.551519261966408e-06,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8522963166236878,
      "num_tokens": 929115315.0,
      "step": 27735
    },
    {
      "epoch": 1.6559216809933144,
      "grad_norm": 0.5298928618431091,
      "learning_rate": 8.545533705900863e-06,
      "loss": 0.4654,
      "mean_token_accuracy": 0.8443397402763366,
      "num_tokens": 929282995.0,
      "step": 27740
    },
    {
      "epoch": 1.656220152817574,
      "grad_norm": 0.5053024291992188,
      "learning_rate": 8.53955276654544e-06,
      "loss": 0.445,
      "mean_token_accuracy": 0.8511749982833863,
      "num_tokens": 929450675.0,
      "step": 27745
    },
    {
      "epoch": 1.656518624641834,
      "grad_norm": 0.4837656319141388,
      "learning_rate": 8.533576445356917e-06,
      "loss": 0.4117,
      "mean_token_accuracy": 0.8607598781585694,
      "num_tokens": 929618355.0,
      "step": 27750
    },
    {
      "epoch": 1.6568170964660935,
      "grad_norm": 0.5115906000137329,
      "learning_rate": 8.527604743790919e-06,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8451926589012146,
      "num_tokens": 929786035.0,
      "step": 27755
    },
    {
      "epoch": 1.6571155682903533,
      "grad_norm": 0.6006373763084412,
      "learning_rate": 8.52163766330199e-06,
      "loss": 0.4402,
      "mean_token_accuracy": 0.8521770358085632,
      "num_tokens": 929953715.0,
      "step": 27760
    },
    {
      "epoch": 1.6574140401146131,
      "grad_norm": 0.4715985357761383,
      "learning_rate": 8.515675205343501e-06,
      "loss": 0.4159,
      "mean_token_accuracy": 0.8614815711975098,
      "num_tokens": 930121395.0,
      "step": 27765
    },
    {
      "epoch": 1.657712511938873,
      "grad_norm": 0.5042600035667419,
      "learning_rate": 8.509717371367732e-06,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8507097721099853,
      "num_tokens": 930289075.0,
      "step": 27770
    },
    {
      "epoch": 1.6580109837631327,
      "grad_norm": 0.4789464771747589,
      "learning_rate": 8.503764162825805e-06,
      "loss": 0.4566,
      "mean_token_accuracy": 0.847518789768219,
      "num_tokens": 930456755.0,
      "step": 27775
    },
    {
      "epoch": 1.6583094555873925,
      "grad_norm": 0.4957612454891205,
      "learning_rate": 8.497815581167755e-06,
      "loss": 0.4177,
      "mean_token_accuracy": 0.8605153203010559,
      "num_tokens": 930624435.0,
      "step": 27780
    },
    {
      "epoch": 1.6586079274116523,
      "grad_norm": 0.5574946999549866,
      "learning_rate": 8.491871627842465e-06,
      "loss": 0.4383,
      "mean_token_accuracy": 0.8546403408050537,
      "num_tokens": 930792115.0,
      "step": 27785
    },
    {
      "epoch": 1.658906399235912,
      "grad_norm": 0.49140918254852295,
      "learning_rate": 8.485932304297683e-06,
      "loss": 0.4454,
      "mean_token_accuracy": 0.8529762506484986,
      "num_tokens": 930959795.0,
      "step": 27790
    },
    {
      "epoch": 1.659204871060172,
      "grad_norm": 0.5171493887901306,
      "learning_rate": 8.479997611980048e-06,
      "loss": 0.4541,
      "mean_token_accuracy": 0.8486520290374756,
      "num_tokens": 931127475.0,
      "step": 27795
    },
    {
      "epoch": 1.6595033428844317,
      "grad_norm": 0.49158546328544617,
      "learning_rate": 8.474067552335062e-06,
      "loss": 0.3906,
      "mean_token_accuracy": 0.8663068175315857,
      "num_tokens": 931295155.0,
      "step": 27800
    },
    {
      "epoch": 1.6598018147086915,
      "grad_norm": 0.48990312218666077,
      "learning_rate": 8.468142126807105e-06,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8553020596504212,
      "num_tokens": 931449666.0,
      "step": 27805
    },
    {
      "epoch": 1.6601002865329513,
      "grad_norm": 0.5473856329917908,
      "learning_rate": 8.462221336839427e-06,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8515865325927734,
      "num_tokens": 931617346.0,
      "step": 27810
    },
    {
      "epoch": 1.660398758357211,
      "grad_norm": 0.4636194407939911,
      "learning_rate": 8.456305183874128e-06,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8540737271308899,
      "num_tokens": 931785026.0,
      "step": 27815
    },
    {
      "epoch": 1.6606972301814709,
      "grad_norm": 0.5502971410751343,
      "learning_rate": 8.450393669352215e-06,
      "loss": 0.4251,
      "mean_token_accuracy": 0.8557437658309937,
      "num_tokens": 931952706.0,
      "step": 27820
    },
    {
      "epoch": 1.6609957020057307,
      "grad_norm": 0.4992155432701111,
      "learning_rate": 8.444486794713538e-06,
      "loss": 0.4104,
      "mean_token_accuracy": 0.8617140412330627,
      "num_tokens": 932119688.0,
      "step": 27825
    },
    {
      "epoch": 1.6612941738299905,
      "grad_norm": 0.5216702818870544,
      "learning_rate": 8.438584561396826e-06,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8459441781044006,
      "num_tokens": 932287368.0,
      "step": 27830
    },
    {
      "epoch": 1.6615926456542502,
      "grad_norm": 0.5434436798095703,
      "learning_rate": 8.432686970839674e-06,
      "loss": 0.4404,
      "mean_token_accuracy": 0.8518191695213317,
      "num_tokens": 932455048.0,
      "step": 27835
    },
    {
      "epoch": 1.66189111747851,
      "grad_norm": 0.47585082054138184,
      "learning_rate": 8.426794024478555e-06,
      "loss": 0.4079,
      "mean_token_accuracy": 0.8623523831367492,
      "num_tokens": 932622728.0,
      "step": 27840
    },
    {
      "epoch": 1.6621895893027698,
      "grad_norm": 0.456572026014328,
      "learning_rate": 8.420905723748809e-06,
      "loss": 0.421,
      "mean_token_accuracy": 0.8603841185569763,
      "num_tokens": 932790408.0,
      "step": 27845
    },
    {
      "epoch": 1.6624880611270296,
      "grad_norm": 0.48674291372299194,
      "learning_rate": 8.415022070084626e-06,
      "loss": 0.4261,
      "mean_token_accuracy": 0.8582906007766724,
      "num_tokens": 932958088.0,
      "step": 27850
    },
    {
      "epoch": 1.6627865329512894,
      "grad_norm": 0.4801638722419739,
      "learning_rate": 8.409143064919088e-06,
      "loss": 0.4466,
      "mean_token_accuracy": 0.8507037878036499,
      "num_tokens": 933125768.0,
      "step": 27855
    },
    {
      "epoch": 1.6630850047755492,
      "grad_norm": 0.48785632848739624,
      "learning_rate": 8.403268709684135e-06,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8466360449790955,
      "num_tokens": 933293448.0,
      "step": 27860
    },
    {
      "epoch": 1.663383476599809,
      "grad_norm": 0.4722157418727875,
      "learning_rate": 8.397399005810569e-06,
      "loss": 0.4201,
      "mean_token_accuracy": 0.8584218144416809,
      "num_tokens": 933461128.0,
      "step": 27865
    },
    {
      "epoch": 1.6636819484240688,
      "grad_norm": 0.4849564731121063,
      "learning_rate": 8.391533954728075e-06,
      "loss": 0.4372,
      "mean_token_accuracy": 0.85422283411026,
      "num_tokens": 933628808.0,
      "step": 27870
    },
    {
      "epoch": 1.6639804202483286,
      "grad_norm": 0.4901125133037567,
      "learning_rate": 8.385673557865184e-06,
      "loss": 0.4394,
      "mean_token_accuracy": 0.8514575719833374,
      "num_tokens": 933790061.0,
      "step": 27875
    },
    {
      "epoch": 1.6642788920725884,
      "grad_norm": 0.4593963623046875,
      "learning_rate": 8.379817816649308e-06,
      "loss": 0.4223,
      "mean_token_accuracy": 0.8571036577224731,
      "num_tokens": 933957741.0,
      "step": 27880
    },
    {
      "epoch": 1.6645773638968482,
      "grad_norm": 0.48224207758903503,
      "learning_rate": 8.373966732506722e-06,
      "loss": 0.467,
      "mean_token_accuracy": 0.8445186734199523,
      "num_tokens": 934125421.0,
      "step": 27885
    },
    {
      "epoch": 1.664875835721108,
      "grad_norm": 0.4888492524623871,
      "learning_rate": 8.368120306862563e-06,
      "loss": 0.4487,
      "mean_token_accuracy": 0.8504950404167175,
      "num_tokens": 934293101.0,
      "step": 27890
    },
    {
      "epoch": 1.6651743075453678,
      "grad_norm": 0.5116983652114868,
      "learning_rate": 8.362278541140845e-06,
      "loss": 0.3966,
      "mean_token_accuracy": 0.8655374050140381,
      "num_tokens": 934460781.0,
      "step": 27895
    },
    {
      "epoch": 1.6654727793696273,
      "grad_norm": 0.5528643727302551,
      "learning_rate": 8.356441436764415e-06,
      "loss": 0.4638,
      "mean_token_accuracy": 0.8443516492843628,
      "num_tokens": 934628461.0,
      "step": 27900
    },
    {
      "epoch": 1.6657712511938874,
      "grad_norm": 0.5356369614601135,
      "learning_rate": 8.350608995155035e-06,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8570857644081116,
      "num_tokens": 934796141.0,
      "step": 27905
    },
    {
      "epoch": 1.666069723018147,
      "grad_norm": 0.472017377614975,
      "learning_rate": 8.344781217733285e-06,
      "loss": 0.4183,
      "mean_token_accuracy": 0.8590242266654968,
      "num_tokens": 934963821.0,
      "step": 27910
    },
    {
      "epoch": 1.666368194842407,
      "grad_norm": 0.5102182626724243,
      "learning_rate": 8.338958105918639e-06,
      "loss": 0.4345,
      "mean_token_accuracy": 0.8555946588516236,
      "num_tokens": 935131501.0,
      "step": 27915
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.4919207692146301,
      "learning_rate": 8.333139661129415e-06,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8502326130867004,
      "num_tokens": 935299181.0,
      "step": 27920
    },
    {
      "epoch": 1.6669651384909265,
      "grad_norm": 0.6054666042327881,
      "learning_rate": 8.327325884782804e-06,
      "loss": 0.4499,
      "mean_token_accuracy": 0.8494810819625854,
      "num_tokens": 935466861.0,
      "step": 27925
    },
    {
      "epoch": 1.667263610315186,
      "grad_norm": 0.5438222885131836,
      "learning_rate": 8.321516778294872e-06,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8481868028640747,
      "num_tokens": 935634541.0,
      "step": 27930
    },
    {
      "epoch": 1.6675620821394461,
      "grad_norm": 0.5048179626464844,
      "learning_rate": 8.315712343080512e-06,
      "loss": 0.4254,
      "mean_token_accuracy": 0.85670405626297,
      "num_tokens": 935802221.0,
      "step": 27935
    },
    {
      "epoch": 1.6678605539637057,
      "grad_norm": 0.540630578994751,
      "learning_rate": 8.309912580553512e-06,
      "loss": 0.4135,
      "mean_token_accuracy": 0.8610700249671936,
      "num_tokens": 935969901.0,
      "step": 27940
    },
    {
      "epoch": 1.6681590257879657,
      "grad_norm": 0.44495153427124023,
      "learning_rate": 8.304117492126519e-06,
      "loss": 0.4127,
      "mean_token_accuracy": 0.8618275046348571,
      "num_tokens": 936137581.0,
      "step": 27945
    },
    {
      "epoch": 1.6684574976122253,
      "grad_norm": 0.5471038222312927,
      "learning_rate": 8.298327079211022e-06,
      "loss": 0.4465,
      "mean_token_accuracy": 0.8510318517684936,
      "num_tokens": 936305261.0,
      "step": 27950
    },
    {
      "epoch": 1.6687559694364853,
      "grad_norm": 0.562255859375,
      "learning_rate": 8.29254134321739e-06,
      "loss": 0.4418,
      "mean_token_accuracy": 0.85333411693573,
      "num_tokens": 936472941.0,
      "step": 27955
    },
    {
      "epoch": 1.6690544412607449,
      "grad_norm": 0.5070235133171082,
      "learning_rate": 8.286760285554836e-06,
      "loss": 0.4117,
      "mean_token_accuracy": 0.8602230787277222,
      "num_tokens": 936640621.0,
      "step": 27960
    },
    {
      "epoch": 1.6693529130850049,
      "grad_norm": 0.47316086292266846,
      "learning_rate": 8.280983907631452e-06,
      "loss": 0.4227,
      "mean_token_accuracy": 0.8590838670730591,
      "num_tokens": 936808301.0,
      "step": 27965
    },
    {
      "epoch": 1.6696513849092645,
      "grad_norm": 0.5139144062995911,
      "learning_rate": 8.275212210854188e-06,
      "loss": 0.4366,
      "mean_token_accuracy": 0.8539424896240234,
      "num_tokens": 936975981.0,
      "step": 27970
    },
    {
      "epoch": 1.6699498567335245,
      "grad_norm": 0.524645984172821,
      "learning_rate": 8.269445196628831e-06,
      "loss": 0.4456,
      "mean_token_accuracy": 0.8517237186431885,
      "num_tokens": 937143661.0,
      "step": 27975
    },
    {
      "epoch": 1.670248328557784,
      "grad_norm": 0.5532966256141663,
      "learning_rate": 8.263682866360051e-06,
      "loss": 0.424,
      "mean_token_accuracy": 0.8589824676513672,
      "num_tokens": 937311341.0,
      "step": 27980
    },
    {
      "epoch": 1.670546800382044,
      "grad_norm": 0.4807977080345154,
      "learning_rate": 8.257925221451373e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.8570380449295044,
      "num_tokens": 937479021.0,
      "step": 27985
    },
    {
      "epoch": 1.6708452722063036,
      "grad_norm": 0.547822117805481,
      "learning_rate": 8.252172263305183e-06,
      "loss": 0.4259,
      "mean_token_accuracy": 0.8576702833175659,
      "num_tokens": 937646701.0,
      "step": 27990
    },
    {
      "epoch": 1.6711437440305636,
      "grad_norm": 0.5031469464302063,
      "learning_rate": 8.246423993322706e-06,
      "loss": 0.4524,
      "mean_token_accuracy": 0.8491291999816895,
      "num_tokens": 937814381.0,
      "step": 27995
    },
    {
      "epoch": 1.6714422158548232,
      "grad_norm": 0.5416647791862488,
      "learning_rate": 8.240680412904041e-06,
      "loss": 0.4221,
      "mean_token_accuracy": 0.858183228969574,
      "num_tokens": 937982061.0,
      "step": 28000
    },
    {
      "epoch": 1.6717406876790832,
      "grad_norm": 0.483253538608551,
      "learning_rate": 8.23494152344815e-06,
      "loss": 0.4238,
      "mean_token_accuracy": 0.8573720693588257,
      "num_tokens": 938149741.0,
      "step": 28005
    },
    {
      "epoch": 1.6720391595033428,
      "grad_norm": 0.5158764123916626,
      "learning_rate": 8.229207326352845e-06,
      "loss": 0.4256,
      "mean_token_accuracy": 0.8569068312644958,
      "num_tokens": 938317421.0,
      "step": 28010
    },
    {
      "epoch": 1.6723376313276028,
      "grad_norm": 0.5189338326454163,
      "learning_rate": 8.223477823014797e-06,
      "loss": 0.4145,
      "mean_token_accuracy": 0.8611952662467957,
      "num_tokens": 938485101.0,
      "step": 28015
    },
    {
      "epoch": 1.6726361031518624,
      "grad_norm": 0.4930899143218994,
      "learning_rate": 8.217753014829527e-06,
      "loss": 0.4107,
      "mean_token_accuracy": 0.8626685023307801,
      "num_tokens": 938652781.0,
      "step": 28020
    },
    {
      "epoch": 1.6729345749761224,
      "grad_norm": 0.4693681001663208,
      "learning_rate": 8.21203290319141e-06,
      "loss": 0.4444,
      "mean_token_accuracy": 0.8507097721099853,
      "num_tokens": 938820461.0,
      "step": 28025
    },
    {
      "epoch": 1.673233046800382,
      "grad_norm": 0.5533576607704163,
      "learning_rate": 8.20631748949371e-06,
      "loss": 0.442,
      "mean_token_accuracy": 0.8511093974113464,
      "num_tokens": 938988141.0,
      "step": 28030
    },
    {
      "epoch": 1.6735315186246418,
      "grad_norm": 0.4884834885597229,
      "learning_rate": 8.200606775128497e-06,
      "loss": 0.4369,
      "mean_token_accuracy": 0.8536621689796448,
      "num_tokens": 939155821.0,
      "step": 28035
    },
    {
      "epoch": 1.6738299904489016,
      "grad_norm": 0.46458420157432556,
      "learning_rate": 8.194900761486731e-06,
      "loss": 0.4325,
      "mean_token_accuracy": 0.8541333794593811,
      "num_tokens": 939323501.0,
      "step": 28040
    },
    {
      "epoch": 1.6741284622731614,
      "grad_norm": 0.5253254175186157,
      "learning_rate": 8.189199449958215e-06,
      "loss": 0.4428,
      "mean_token_accuracy": 0.8522843837738037,
      "num_tokens": 939491181.0,
      "step": 28045
    },
    {
      "epoch": 1.6744269340974212,
      "grad_norm": 0.5520123243331909,
      "learning_rate": 8.183502841931611e-06,
      "loss": 0.4209,
      "mean_token_accuracy": 0.857819390296936,
      "num_tokens": 939658861.0,
      "step": 28050
    },
    {
      "epoch": 1.674725405921681,
      "grad_norm": 0.45845434069633484,
      "learning_rate": 8.177810938794436e-06,
      "loss": 0.409,
      "mean_token_accuracy": 0.8619587302207947,
      "num_tokens": 939826541.0,
      "step": 28055
    },
    {
      "epoch": 1.6750238777459407,
      "grad_norm": 0.49971920251846313,
      "learning_rate": 8.172123741933047e-06,
      "loss": 0.399,
      "mean_token_accuracy": 0.8653644323348999,
      "num_tokens": 939994221.0,
      "step": 28060
    },
    {
      "epoch": 1.6753223495702005,
      "grad_norm": 0.5177424550056458,
      "learning_rate": 8.166441252732675e-06,
      "loss": 0.4149,
      "mean_token_accuracy": 0.859686267375946,
      "num_tokens": 940161901.0,
      "step": 28065
    },
    {
      "epoch": 1.6756208213944603,
      "grad_norm": 0.49712035059928894,
      "learning_rate": 8.16076347257739e-06,
      "loss": 0.4011,
      "mean_token_accuracy": 0.8641894340515137,
      "num_tokens": 940329581.0,
      "step": 28070
    },
    {
      "epoch": 1.6759192932187201,
      "grad_norm": 0.5571178197860718,
      "learning_rate": 8.155090402850124e-06,
      "loss": 0.4376,
      "mean_token_accuracy": 0.8536383032798767,
      "num_tokens": 940497261.0,
      "step": 28075
    },
    {
      "epoch": 1.67621776504298,
      "grad_norm": 0.5189345479011536,
      "learning_rate": 8.149422044932664e-06,
      "loss": 0.4335,
      "mean_token_accuracy": 0.854461395740509,
      "num_tokens": 940664941.0,
      "step": 28080
    },
    {
      "epoch": 1.6765162368672397,
      "grad_norm": 0.44967299699783325,
      "learning_rate": 8.143758400205622e-06,
      "loss": 0.4055,
      "mean_token_accuracy": 0.8618453979492188,
      "num_tokens": 940832621.0,
      "step": 28085
    },
    {
      "epoch": 1.6768147086914995,
      "grad_norm": 0.5450178384780884,
      "learning_rate": 8.138099470048508e-06,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8467613220214844,
      "num_tokens": 941000301.0,
      "step": 28090
    },
    {
      "epoch": 1.6771131805157593,
      "grad_norm": 0.5511917471885681,
      "learning_rate": 8.132445255839644e-06,
      "loss": 0.4214,
      "mean_token_accuracy": 0.8596087336540222,
      "num_tokens": 941167981.0,
      "step": 28095
    },
    {
      "epoch": 1.677411652340019,
      "grad_norm": 0.4799546003341675,
      "learning_rate": 8.126795758956225e-06,
      "loss": 0.4539,
      "mean_token_accuracy": 0.8484134674072266,
      "num_tokens": 941335661.0,
      "step": 28100
    },
    {
      "epoch": 1.677710124164279,
      "grad_norm": 0.4885207414627075,
      "learning_rate": 8.121150980774279e-06,
      "loss": 0.4232,
      "mean_token_accuracy": 0.8574794292449951,
      "num_tokens": 941503341.0,
      "step": 28105
    },
    {
      "epoch": 1.6780085959885387,
      "grad_norm": 0.4217756688594818,
      "learning_rate": 8.115510922668706e-06,
      "loss": 0.3685,
      "mean_token_accuracy": 0.8754443645477294,
      "num_tokens": 941671021.0,
      "step": 28110
    },
    {
      "epoch": 1.6783070678127985,
      "grad_norm": 0.5287328958511353,
      "learning_rate": 8.109875586013253e-06,
      "loss": 0.4501,
      "mean_token_accuracy": 0.8497495055198669,
      "num_tokens": 941838701.0,
      "step": 28115
    },
    {
      "epoch": 1.6786055396370583,
      "grad_norm": 0.522629976272583,
      "learning_rate": 8.104244972180496e-06,
      "loss": 0.4474,
      "mean_token_accuracy": 0.8518430233001709,
      "num_tokens": 942006381.0,
      "step": 28120
    },
    {
      "epoch": 1.678904011461318,
      "grad_norm": 0.6132490038871765,
      "learning_rate": 8.098619082541881e-06,
      "loss": 0.4715,
      "mean_token_accuracy": 0.8445246338844299,
      "num_tokens": 942174061.0,
      "step": 28125
    },
    {
      "epoch": 1.6792024832855779,
      "grad_norm": 0.4500133693218231,
      "learning_rate": 8.092997918467697e-06,
      "loss": 0.4139,
      "mean_token_accuracy": 0.8597578406333923,
      "num_tokens": 942341741.0,
      "step": 28130
    },
    {
      "epoch": 1.6795009551098377,
      "grad_norm": 0.41929033398628235,
      "learning_rate": 8.08738148132708e-06,
      "loss": 0.4064,
      "mean_token_accuracy": 0.8644816875457764,
      "num_tokens": 942509421.0,
      "step": 28135
    },
    {
      "epoch": 1.6797994269340975,
      "grad_norm": 0.5036787986755371,
      "learning_rate": 8.081769772488025e-06,
      "loss": 0.4311,
      "mean_token_accuracy": 0.8543898344039917,
      "num_tokens": 942677101.0,
      "step": 28140
    },
    {
      "epoch": 1.6800978987583572,
      "grad_norm": 0.4599270820617676,
      "learning_rate": 8.076162793317357e-06,
      "loss": 0.4101,
      "mean_token_accuracy": 0.8629428505897522,
      "num_tokens": 942844781.0,
      "step": 28145
    },
    {
      "epoch": 1.680396370582617,
      "grad_norm": 0.48597437143325806,
      "learning_rate": 8.070560545180763e-06,
      "loss": 0.4391,
      "mean_token_accuracy": 0.8520935416221619,
      "num_tokens": 943012461.0,
      "step": 28150
    },
    {
      "epoch": 1.6806948424068768,
      "grad_norm": 0.54108726978302,
      "learning_rate": 8.064963029442776e-06,
      "loss": 0.4854,
      "mean_token_accuracy": 0.842067277431488,
      "num_tokens": 943180141.0,
      "step": 28155
    },
    {
      "epoch": 1.6809933142311366,
      "grad_norm": 0.4786975681781769,
      "learning_rate": 8.059370247466773e-06,
      "loss": 0.4354,
      "mean_token_accuracy": 0.8550280213356019,
      "num_tokens": 943347821.0,
      "step": 28160
    },
    {
      "epoch": 1.6812917860553964,
      "grad_norm": 0.4980897903442383,
      "learning_rate": 8.053782200614978e-06,
      "loss": 0.4173,
      "mean_token_accuracy": 0.8611356258392334,
      "num_tokens": 943515501.0,
      "step": 28165
    },
    {
      "epoch": 1.6815902578796562,
      "grad_norm": 0.47392842173576355,
      "learning_rate": 8.048198890248463e-06,
      "loss": 0.4102,
      "mean_token_accuracy": 0.860640573501587,
      "num_tokens": 943683181.0,
      "step": 28170
    },
    {
      "epoch": 1.6818887297039158,
      "grad_norm": 0.49055197834968567,
      "learning_rate": 8.042620317727152e-06,
      "loss": 0.4206,
      "mean_token_accuracy": 0.8582130551338196,
      "num_tokens": 943850861.0,
      "step": 28175
    },
    {
      "epoch": 1.6821872015281758,
      "grad_norm": 0.5143933296203613,
      "learning_rate": 8.037046484409801e-06,
      "loss": 0.425,
      "mean_token_accuracy": 0.8564356565475464,
      "num_tokens": 944018541.0,
      "step": 28180
    },
    {
      "epoch": 1.6824856733524354,
      "grad_norm": 0.5638523697853088,
      "learning_rate": 8.031477391654025e-06,
      "loss": 0.4715,
      "mean_token_accuracy": 0.8440176606178283,
      "num_tokens": 944186221.0,
      "step": 28185
    },
    {
      "epoch": 1.6827841451766954,
      "grad_norm": 0.6349928379058838,
      "learning_rate": 8.025913040816278e-06,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8538411140441895,
      "num_tokens": 944353901.0,
      "step": 28190
    },
    {
      "epoch": 1.683082617000955,
      "grad_norm": 0.4978715181350708,
      "learning_rate": 8.020353433251859e-06,
      "loss": 0.3895,
      "mean_token_accuracy": 0.8682273626327515,
      "num_tokens": 944521581.0,
      "step": 28195
    },
    {
      "epoch": 1.683381088825215,
      "grad_norm": 0.5366897583007812,
      "learning_rate": 8.014798570314918e-06,
      "loss": 0.4763,
      "mean_token_accuracy": 0.8423953294754029,
      "num_tokens": 944689261.0,
      "step": 28200
    },
    {
      "epoch": 1.6836795606494745,
      "grad_norm": 0.5215397477149963,
      "learning_rate": 8.009248453358438e-06,
      "loss": 0.4289,
      "mean_token_accuracy": 0.855851125717163,
      "num_tokens": 944856941.0,
      "step": 28205
    },
    {
      "epoch": 1.6839780324737346,
      "grad_norm": 0.5060926675796509,
      "learning_rate": 8.003703083734254e-06,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8561254739761353,
      "num_tokens": 945024621.0,
      "step": 28210
    },
    {
      "epoch": 1.6842765042979941,
      "grad_norm": 0.6309110522270203,
      "learning_rate": 7.998162462793046e-06,
      "loss": 0.4059,
      "mean_token_accuracy": 0.862018370628357,
      "num_tokens": 945192301.0,
      "step": 28215
    },
    {
      "epoch": 1.6845749761222542,
      "grad_norm": 0.46858254075050354,
      "learning_rate": 7.992626591884331e-06,
      "loss": 0.4039,
      "mean_token_accuracy": 0.8643146634101868,
      "num_tokens": 945359981.0,
      "step": 28220
    },
    {
      "epoch": 1.6848734479465137,
      "grad_norm": 0.5501070022583008,
      "learning_rate": 7.98709547235648e-06,
      "loss": 0.4729,
      "mean_token_accuracy": 0.8422581434249878,
      "num_tokens": 945527661.0,
      "step": 28225
    },
    {
      "epoch": 1.6851719197707737,
      "grad_norm": 0.4850921034812927,
      "learning_rate": 7.981569105556685e-06,
      "loss": 0.3941,
      "mean_token_accuracy": 0.8667780041694642,
      "num_tokens": 945695341.0,
      "step": 28230
    },
    {
      "epoch": 1.6854703915950333,
      "grad_norm": 0.4399625360965729,
      "learning_rate": 7.976047492831012e-06,
      "loss": 0.4256,
      "mean_token_accuracy": 0.8575629234313965,
      "num_tokens": 945863021.0,
      "step": 28235
    },
    {
      "epoch": 1.6857688634192933,
      "grad_norm": 0.5460418462753296,
      "learning_rate": 7.97053063552434e-06,
      "loss": 0.4751,
      "mean_token_accuracy": 0.843546462059021,
      "num_tokens": 946030701.0,
      "step": 28240
    },
    {
      "epoch": 1.686067335243553,
      "grad_norm": 0.48071375489234924,
      "learning_rate": 7.9650185349804e-06,
      "loss": 0.4543,
      "mean_token_accuracy": 0.8498270273208618,
      "num_tokens": 946198381.0,
      "step": 28245
    },
    {
      "epoch": 1.686365807067813,
      "grad_norm": 0.503944993019104,
      "learning_rate": 7.959511192541772e-06,
      "loss": 0.4395,
      "mean_token_accuracy": 0.8524394512176514,
      "num_tokens": 946366061.0,
      "step": 28250
    },
    {
      "epoch": 1.6866642788920725,
      "grad_norm": 0.4849456548690796,
      "learning_rate": 7.954008609549872e-06,
      "loss": 0.4476,
      "mean_token_accuracy": 0.8522247314453125,
      "num_tokens": 946533741.0,
      "step": 28255
    },
    {
      "epoch": 1.6869627507163325,
      "grad_norm": 0.5044650435447693,
      "learning_rate": 7.948510787344956e-06,
      "loss": 0.428,
      "mean_token_accuracy": 0.8562388062477112,
      "num_tokens": 946701421.0,
      "step": 28260
    },
    {
      "epoch": 1.687261222540592,
      "grad_norm": 0.47206827998161316,
      "learning_rate": 7.94301772726611e-06,
      "loss": 0.4515,
      "mean_token_accuracy": 0.8497972130775452,
      "num_tokens": 946869101.0,
      "step": 28265
    },
    {
      "epoch": 1.687559694364852,
      "grad_norm": 0.4965664744377136,
      "learning_rate": 7.937529430651282e-06,
      "loss": 0.4011,
      "mean_token_accuracy": 0.8655254721641541,
      "num_tokens": 947036781.0,
      "step": 28270
    },
    {
      "epoch": 1.6878581661891117,
      "grad_norm": 0.4931280016899109,
      "learning_rate": 7.932045898837238e-06,
      "loss": 0.4639,
      "mean_token_accuracy": 0.8454550862312317,
      "num_tokens": 947204461.0,
      "step": 28275
    },
    {
      "epoch": 1.6881566380133717,
      "grad_norm": 0.49421343207359314,
      "learning_rate": 7.9265671331596e-06,
      "loss": 0.4502,
      "mean_token_accuracy": 0.8504592537879944,
      "num_tokens": 947372141.0,
      "step": 28280
    },
    {
      "epoch": 1.6884551098376313,
      "grad_norm": 0.5166566371917725,
      "learning_rate": 7.921093134952827e-06,
      "loss": 0.409,
      "mean_token_accuracy": 0.8620002031326294,
      "num_tokens": 947532664.0,
      "step": 28285
    },
    {
      "epoch": 1.6887535816618913,
      "grad_norm": 0.48326945304870605,
      "learning_rate": 7.915623905550194e-06,
      "loss": 0.4539,
      "mean_token_accuracy": 0.8495586276054382,
      "num_tokens": 947700344.0,
      "step": 28290
    },
    {
      "epoch": 1.6890520534861508,
      "grad_norm": 0.5510462522506714,
      "learning_rate": 7.910159446283852e-06,
      "loss": 0.4258,
      "mean_token_accuracy": 0.8554395794868469,
      "num_tokens": 947868024.0,
      "step": 28295
    },
    {
      "epoch": 1.6893505253104109,
      "grad_norm": 0.4928211271762848,
      "learning_rate": 7.904699758484768e-06,
      "loss": 0.4386,
      "mean_token_accuracy": 0.8537516236305237,
      "num_tokens": 948035704.0,
      "step": 28300
    },
    {
      "epoch": 1.6896489971346704,
      "grad_norm": 0.4887271225452423,
      "learning_rate": 7.899244843482742e-06,
      "loss": 0.4257,
      "mean_token_accuracy": 0.8570857524871827,
      "num_tokens": 948203384.0,
      "step": 28305
    },
    {
      "epoch": 1.6899474689589304,
      "grad_norm": 0.5313078165054321,
      "learning_rate": 7.89379470260642e-06,
      "loss": 0.4325,
      "mean_token_accuracy": 0.8548013806343079,
      "num_tokens": 948371064.0,
      "step": 28310
    },
    {
      "epoch": 1.69024594078319,
      "grad_norm": 0.5151989459991455,
      "learning_rate": 7.888349337183289e-06,
      "loss": 0.4271,
      "mean_token_accuracy": 0.8573422312736512,
      "num_tokens": 948538744.0,
      "step": 28315
    },
    {
      "epoch": 1.6905444126074498,
      "grad_norm": 0.5010346174240112,
      "learning_rate": 7.882908748539672e-06,
      "loss": 0.4466,
      "mean_token_accuracy": 0.8504592776298523,
      "num_tokens": 948706424.0,
      "step": 28320
    },
    {
      "epoch": 1.6908428844317096,
      "grad_norm": 0.4591065049171448,
      "learning_rate": 7.877472938000713e-06,
      "loss": 0.4088,
      "mean_token_accuracy": 0.8633484244346619,
      "num_tokens": 948874104.0,
      "step": 28325
    },
    {
      "epoch": 1.6911413562559694,
      "grad_norm": 0.5350884795188904,
      "learning_rate": 7.872041906890412e-06,
      "loss": 0.4453,
      "mean_token_accuracy": 0.8519742369651795,
      "num_tokens": 949041784.0,
      "step": 28330
    },
    {
      "epoch": 1.6914398280802292,
      "grad_norm": 0.5192052721977234,
      "learning_rate": 7.866615656531593e-06,
      "loss": 0.4101,
      "mean_token_accuracy": 0.8620004773139953,
      "num_tokens": 949209464.0,
      "step": 28335
    },
    {
      "epoch": 1.691738299904489,
      "grad_norm": 0.4882083833217621,
      "learning_rate": 7.861194188245922e-06,
      "loss": 0.3833,
      "mean_token_accuracy": 0.8693725347518921,
      "num_tokens": 949377144.0,
      "step": 28340
    },
    {
      "epoch": 1.6920367717287488,
      "grad_norm": 0.5025365352630615,
      "learning_rate": 7.855777503353904e-06,
      "loss": 0.4753,
      "mean_token_accuracy": 0.8442443013191223,
      "num_tokens": 949544824.0,
      "step": 28345
    },
    {
      "epoch": 1.6923352435530086,
      "grad_norm": 0.5009467005729675,
      "learning_rate": 7.850365603174858e-06,
      "loss": 0.4136,
      "mean_token_accuracy": 0.8606942653656006,
      "num_tokens": 949712504.0,
      "step": 28350
    },
    {
      "epoch": 1.6926337153772684,
      "grad_norm": 0.520879328250885,
      "learning_rate": 7.84495848902696e-06,
      "loss": 0.4373,
      "mean_token_accuracy": 0.8529941439628601,
      "num_tokens": 949880184.0,
      "step": 28355
    },
    {
      "epoch": 1.6929321872015282,
      "grad_norm": 0.4755707383155823,
      "learning_rate": 7.839556162227218e-06,
      "loss": 0.4562,
      "mean_token_accuracy": 0.8494751214981079,
      "num_tokens": 950047864.0,
      "step": 28360
    },
    {
      "epoch": 1.693230659025788,
      "grad_norm": 0.5176428556442261,
      "learning_rate": 7.834158624091459e-06,
      "loss": 0.4074,
      "mean_token_accuracy": 0.8618990778923035,
      "num_tokens": 950215544.0,
      "step": 28365
    },
    {
      "epoch": 1.6935291308500477,
      "grad_norm": 0.4631257951259613,
      "learning_rate": 7.828765875934362e-06,
      "loss": 0.403,
      "mean_token_accuracy": 0.8632470369338989,
      "num_tokens": 950383224.0,
      "step": 28370
    },
    {
      "epoch": 1.6938276026743075,
      "grad_norm": 0.5002923011779785,
      "learning_rate": 7.823377919069416e-06,
      "loss": 0.3853,
      "mean_token_accuracy": 0.8705057859420776,
      "num_tokens": 950550904.0,
      "step": 28375
    },
    {
      "epoch": 1.6941260744985673,
      "grad_norm": 0.5075061321258545,
      "learning_rate": 7.81799475480897e-06,
      "loss": 0.4351,
      "mean_token_accuracy": 0.8541930079460144,
      "num_tokens": 950718584.0,
      "step": 28380
    },
    {
      "epoch": 1.6944245463228271,
      "grad_norm": 0.4791085422039032,
      "learning_rate": 7.812616384464197e-06,
      "loss": 0.44,
      "mean_token_accuracy": 0.8523440361022949,
      "num_tokens": 950886264.0,
      "step": 28385
    },
    {
      "epoch": 1.694723018147087,
      "grad_norm": 0.5743136405944824,
      "learning_rate": 7.807242809345084e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.8563044190406799,
      "num_tokens": 951053944.0,
      "step": 28390
    },
    {
      "epoch": 1.6950214899713467,
      "grad_norm": 0.5197674632072449,
      "learning_rate": 7.801874030760472e-06,
      "loss": 0.4496,
      "mean_token_accuracy": 0.8507097840309144,
      "num_tokens": 951221624.0,
      "step": 28395
    },
    {
      "epoch": 1.6953199617956065,
      "grad_norm": 0.4603707790374756,
      "learning_rate": 7.796510050018025e-06,
      "loss": 0.447,
      "mean_token_accuracy": 0.8499463200569153,
      "num_tokens": 951389304.0,
      "step": 28400
    },
    {
      "epoch": 1.6956184336198663,
      "grad_norm": 0.47777071595191956,
      "learning_rate": 7.791150868424243e-06,
      "loss": 0.4135,
      "mean_token_accuracy": 0.8609805583953858,
      "num_tokens": 951556984.0,
      "step": 28405
    },
    {
      "epoch": 1.695916905444126,
      "grad_norm": 0.43646612763404846,
      "learning_rate": 7.785796487284453e-06,
      "loss": 0.4107,
      "mean_token_accuracy": 0.862155556678772,
      "num_tokens": 951724664.0,
      "step": 28410
    },
    {
      "epoch": 1.696215377268386,
      "grad_norm": 0.5291107296943665,
      "learning_rate": 7.780446907902802e-06,
      "loss": 0.432,
      "mean_token_accuracy": 0.85520099401474,
      "num_tokens": 951892344.0,
      "step": 28415
    },
    {
      "epoch": 1.6965138490926457,
      "grad_norm": 0.53330397605896,
      "learning_rate": 7.775102131582296e-06,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8504354119300842,
      "num_tokens": 952060024.0,
      "step": 28420
    },
    {
      "epoch": 1.6968123209169055,
      "grad_norm": 0.5128780603408813,
      "learning_rate": 7.769762159624742e-06,
      "loss": 0.4181,
      "mean_token_accuracy": 0.8613086104393005,
      "num_tokens": 952227704.0,
      "step": 28425
    },
    {
      "epoch": 1.6971107927411653,
      "grad_norm": 0.5765726566314697,
      "learning_rate": 7.764426993330798e-06,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8495944023132325,
      "num_tokens": 952395384.0,
      "step": 28430
    },
    {
      "epoch": 1.697409264565425,
      "grad_norm": 0.5101821422576904,
      "learning_rate": 7.759096633999927e-06,
      "loss": 0.409,
      "mean_token_accuracy": 0.8609089851379395,
      "num_tokens": 952563064.0,
      "step": 28435
    },
    {
      "epoch": 1.6977077363896849,
      "grad_norm": 0.6067535877227783,
      "learning_rate": 7.753771082930452e-06,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8458964586257934,
      "num_tokens": 952730744.0,
      "step": 28440
    },
    {
      "epoch": 1.6980062082139447,
      "grad_norm": 0.4924018979072571,
      "learning_rate": 7.748450341419505e-06,
      "loss": 0.4066,
      "mean_token_accuracy": 0.8639150738716126,
      "num_tokens": 952898424.0,
      "step": 28445
    },
    {
      "epoch": 1.6983046800382042,
      "grad_norm": 0.5162074565887451,
      "learning_rate": 7.743134410763048e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.859435772895813,
      "num_tokens": 953066104.0,
      "step": 28450
    },
    {
      "epoch": 1.6986031518624642,
      "grad_norm": 0.5504987835884094,
      "learning_rate": 7.737823292255873e-06,
      "loss": 0.44,
      "mean_token_accuracy": 0.8522247433662414,
      "num_tokens": 953233784.0,
      "step": 28455
    },
    {
      "epoch": 1.6989016236867238,
      "grad_norm": 0.48993080854415894,
      "learning_rate": 7.732516987191602e-06,
      "loss": 0.4297,
      "mean_token_accuracy": 0.8549683928489685,
      "num_tokens": 953401464.0,
      "step": 28460
    },
    {
      "epoch": 1.6992000955109838,
      "grad_norm": 0.5028998255729675,
      "learning_rate": 7.727215496862687e-06,
      "loss": 0.4329,
      "mean_token_accuracy": 0.8567756175994873,
      "num_tokens": 953569144.0,
      "step": 28465
    },
    {
      "epoch": 1.6994985673352434,
      "grad_norm": 0.5005002617835999,
      "learning_rate": 7.721918822560404e-06,
      "loss": 0.4003,
      "mean_token_accuracy": 0.8648267984390259,
      "num_tokens": 953732227.0,
      "step": 28470
    },
    {
      "epoch": 1.6997970391595034,
      "grad_norm": 0.49301546812057495,
      "learning_rate": 7.71662696557485e-06,
      "loss": 0.4218,
      "mean_token_accuracy": 0.8585291624069213,
      "num_tokens": 953899907.0,
      "step": 28475
    },
    {
      "epoch": 1.700095510983763,
      "grad_norm": 0.5593149065971375,
      "learning_rate": 7.711339927194958e-06,
      "loss": 0.4376,
      "mean_token_accuracy": 0.853083610534668,
      "num_tokens": 954067587.0,
      "step": 28480
    },
    {
      "epoch": 1.700393982808023,
      "grad_norm": 0.5250505805015564,
      "learning_rate": 7.706057708708486e-06,
      "loss": 0.4142,
      "mean_token_accuracy": 0.8604497194290162,
      "num_tokens": 954235267.0,
      "step": 28485
    },
    {
      "epoch": 1.7006924546322826,
      "grad_norm": 0.5128530859947205,
      "learning_rate": 7.700780311402008e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.8612728118896484,
      "num_tokens": 954402947.0,
      "step": 28490
    },
    {
      "epoch": 1.7009909264565426,
      "grad_norm": 0.5120176672935486,
      "learning_rate": 7.695507736560945e-06,
      "loss": 0.4246,
      "mean_token_accuracy": 0.8568531632423401,
      "num_tokens": 954570627.0,
      "step": 28495
    },
    {
      "epoch": 1.7012893982808022,
      "grad_norm": 0.4883759617805481,
      "learning_rate": 7.690239985469511e-06,
      "loss": 0.4116,
      "mean_token_accuracy": 0.8602588534355163,
      "num_tokens": 954738307.0,
      "step": 28500
    },
    {
      "epoch": 1.7015878701050622,
      "grad_norm": 0.5529108643531799,
      "learning_rate": 7.684977059410783e-06,
      "loss": 0.4784,
      "mean_token_accuracy": 0.8417511582374573,
      "num_tokens": 954905987.0,
      "step": 28505
    },
    {
      "epoch": 1.7018863419293218,
      "grad_norm": 0.5632097125053406,
      "learning_rate": 7.679718959666635e-06,
      "loss": 0.4082,
      "mean_token_accuracy": 0.8618931174278259,
      "num_tokens": 955073667.0,
      "step": 28510
    },
    {
      "epoch": 1.7021848137535818,
      "grad_norm": 0.7264323234558105,
      "learning_rate": 7.674465687517773e-06,
      "loss": 0.4837,
      "mean_token_accuracy": 0.8404151201248169,
      "num_tokens": 955241347.0,
      "step": 28515
    },
    {
      "epoch": 1.7024832855778413,
      "grad_norm": 0.5806310772895813,
      "learning_rate": 7.669217244243728e-06,
      "loss": 0.4629,
      "mean_token_accuracy": 0.8464153647422791,
      "num_tokens": 955409027.0,
      "step": 28520
    },
    {
      "epoch": 1.7027817574021014,
      "grad_norm": 0.5168312191963196,
      "learning_rate": 7.66397363112286e-06,
      "loss": 0.4383,
      "mean_token_accuracy": 0.8536860466003418,
      "num_tokens": 955576707.0,
      "step": 28525
    },
    {
      "epoch": 1.703080229226361,
      "grad_norm": 0.5000343322753906,
      "learning_rate": 7.65873484943235e-06,
      "loss": 0.4408,
      "mean_token_accuracy": 0.8526422500610351,
      "num_tokens": 955744387.0,
      "step": 28530
    },
    {
      "epoch": 1.703378701050621,
      "grad_norm": 0.5200855731964111,
      "learning_rate": 7.653500900448189e-06,
      "loss": 0.4255,
      "mean_token_accuracy": 0.8565907120704651,
      "num_tokens": 955912067.0,
      "step": 28535
    },
    {
      "epoch": 1.7036771728748805,
      "grad_norm": 0.5474066734313965,
      "learning_rate": 7.648271785445209e-06,
      "loss": 0.4445,
      "mean_token_accuracy": 0.8520517587661743,
      "num_tokens": 956079747.0,
      "step": 28540
    },
    {
      "epoch": 1.7039756446991405,
      "grad_norm": 0.5423592329025269,
      "learning_rate": 7.643047505697055e-06,
      "loss": 0.4958,
      "mean_token_accuracy": 0.833472490310669,
      "num_tokens": 956247427.0,
      "step": 28545
    },
    {
      "epoch": 1.7042741165234,
      "grad_norm": 0.4779508113861084,
      "learning_rate": 7.6378280624762e-06,
      "loss": 0.4067,
      "mean_token_accuracy": 0.8627579569816589,
      "num_tokens": 956415107.0,
      "step": 28550
    },
    {
      "epoch": 1.7045725883476601,
      "grad_norm": 0.5067100524902344,
      "learning_rate": 7.63261345705394e-06,
      "loss": 0.4401,
      "mean_token_accuracy": 0.850626266002655,
      "num_tokens": 956582787.0,
      "step": 28555
    },
    {
      "epoch": 1.7048710601719197,
      "grad_norm": 0.5187618136405945,
      "learning_rate": 7.6274036907003705e-06,
      "loss": 0.4082,
      "mean_token_accuracy": 0.8630621552467346,
      "num_tokens": 956750467.0,
      "step": 28560
    },
    {
      "epoch": 1.7051695319961797,
      "grad_norm": 0.5184881687164307,
      "learning_rate": 7.622198764684449e-06,
      "loss": 0.4116,
      "mean_token_accuracy": 0.8616008639335633,
      "num_tokens": 956918147.0,
      "step": 28565
    },
    {
      "epoch": 1.7054680038204393,
      "grad_norm": 0.450073778629303,
      "learning_rate": 7.616998680273916e-06,
      "loss": 0.402,
      "mean_token_accuracy": 0.8652153253555298,
      "num_tokens": 957085827.0,
      "step": 28570
    },
    {
      "epoch": 1.7057664756446993,
      "grad_norm": 0.48244667053222656,
      "learning_rate": 7.611803438735351e-06,
      "loss": 0.399,
      "mean_token_accuracy": 0.865000593662262,
      "num_tokens": 957253507.0,
      "step": 28575
    },
    {
      "epoch": 1.7060649474689589,
      "grad_norm": 0.46460282802581787,
      "learning_rate": 7.606613041334154e-06,
      "loss": 0.4208,
      "mean_token_accuracy": 0.8585828542709351,
      "num_tokens": 957421187.0,
      "step": 28580
    },
    {
      "epoch": 1.7063634192932189,
      "grad_norm": 0.5835782885551453,
      "learning_rate": 7.601427489334541e-06,
      "loss": 0.4493,
      "mean_token_accuracy": 0.8491709470748902,
      "num_tokens": 957588867.0,
      "step": 28585
    },
    {
      "epoch": 1.7066618911174785,
      "grad_norm": 0.48928967118263245,
      "learning_rate": 7.5962467839995524e-06,
      "loss": 0.4152,
      "mean_token_accuracy": 0.8602827072143555,
      "num_tokens": 957756547.0,
      "step": 28590
    },
    {
      "epoch": 1.7069603629417383,
      "grad_norm": 0.5073524713516235,
      "learning_rate": 7.591070926591036e-06,
      "loss": 0.4228,
      "mean_token_accuracy": 0.8580221891403198,
      "num_tokens": 957924227.0,
      "step": 28595
    },
    {
      "epoch": 1.707258834765998,
      "grad_norm": 0.504292368888855,
      "learning_rate": 7.585899918369673e-06,
      "loss": 0.439,
      "mean_token_accuracy": 0.8524633169174194,
      "num_tokens": 958091907.0,
      "step": 28600
    },
    {
      "epoch": 1.7075573065902578,
      "grad_norm": 0.5035752058029175,
      "learning_rate": 7.580733760594956e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.8580639362335205,
      "num_tokens": 958259587.0,
      "step": 28605
    },
    {
      "epoch": 1.7078557784145176,
      "grad_norm": 0.5165441036224365,
      "learning_rate": 7.575572454525198e-06,
      "loss": 0.4323,
      "mean_token_accuracy": 0.854485273361206,
      "num_tokens": 958427267.0,
      "step": 28610
    },
    {
      "epoch": 1.7081542502387774,
      "grad_norm": 0.5164929032325745,
      "learning_rate": 7.570416001417537e-06,
      "loss": 0.4346,
      "mean_token_accuracy": 0.852117371559143,
      "num_tokens": 958594947.0,
      "step": 28615
    },
    {
      "epoch": 1.7084527220630372,
      "grad_norm": 0.5457221269607544,
      "learning_rate": 7.565264402527912e-06,
      "loss": 0.458,
      "mean_token_accuracy": 0.8475068569183349,
      "num_tokens": 958762627.0,
      "step": 28620
    },
    {
      "epoch": 1.708751193887297,
      "grad_norm": 0.6608583331108093,
      "learning_rate": 7.560117659111092e-06,
      "loss": 0.4455,
      "mean_token_accuracy": 0.851389718055725,
      "num_tokens": 958930307.0,
      "step": 28625
    },
    {
      "epoch": 1.7090496657115568,
      "grad_norm": 0.4773944318294525,
      "learning_rate": 7.5549757724206724e-06,
      "loss": 0.4187,
      "mean_token_accuracy": 0.8601634263992309,
      "num_tokens": 959097987.0,
      "step": 28630
    },
    {
      "epoch": 1.7093481375358166,
      "grad_norm": 0.5381609797477722,
      "learning_rate": 7.5498387437090425e-06,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8539424896240234,
      "num_tokens": 959265667.0,
      "step": 28635
    },
    {
      "epoch": 1.7096466093600764,
      "grad_norm": 0.5077849626541138,
      "learning_rate": 7.544706574227423e-06,
      "loss": 0.4035,
      "mean_token_accuracy": 0.8643146634101868,
      "num_tokens": 959433347.0,
      "step": 28640
    },
    {
      "epoch": 1.7099450811843362,
      "grad_norm": 0.5975263118743896,
      "learning_rate": 7.5395792652258514e-06,
      "loss": 0.4528,
      "mean_token_accuracy": 0.8489264011383056,
      "num_tokens": 959601027.0,
      "step": 28645
    },
    {
      "epoch": 1.710243553008596,
      "grad_norm": 0.5666967630386353,
      "learning_rate": 7.53445681795318e-06,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8485506534576416,
      "num_tokens": 959768707.0,
      "step": 28650
    },
    {
      "epoch": 1.7105420248328558,
      "grad_norm": 0.4827417731285095,
      "learning_rate": 7.5293392336570676e-06,
      "loss": 0.4247,
      "mean_token_accuracy": 0.8574674963951111,
      "num_tokens": 959936387.0,
      "step": 28655
    },
    {
      "epoch": 1.7108404966571156,
      "grad_norm": 0.48321419954299927,
      "learning_rate": 7.524226513584001e-06,
      "loss": 0.4498,
      "mean_token_accuracy": 0.8502242684364318,
      "num_tokens": 960101809.0,
      "step": 28660
    },
    {
      "epoch": 1.7111389684813754,
      "grad_norm": 0.5237590670585632,
      "learning_rate": 7.51911865897928e-06,
      "loss": 0.4706,
      "mean_token_accuracy": 0.845162832736969,
      "num_tokens": 960269489.0,
      "step": 28665
    },
    {
      "epoch": 1.7114374403056352,
      "grad_norm": 0.5307829976081848,
      "learning_rate": 7.514015671087012e-06,
      "loss": 0.4312,
      "mean_token_accuracy": 0.8544971942901611,
      "num_tokens": 960437169.0,
      "step": 28670
    },
    {
      "epoch": 1.711735912129895,
      "grad_norm": 0.5036895871162415,
      "learning_rate": 7.508917551150131e-06,
      "loss": 0.414,
      "mean_token_accuracy": 0.860914945602417,
      "num_tokens": 960604849.0,
      "step": 28675
    },
    {
      "epoch": 1.7120343839541547,
      "grad_norm": 0.5041924118995667,
      "learning_rate": 7.503824300410371e-06,
      "loss": 0.4439,
      "mean_token_accuracy": 0.8513181567192077,
      "num_tokens": 960772529.0,
      "step": 28680
    },
    {
      "epoch": 1.7123328557784145,
      "grad_norm": 0.5621993541717529,
      "learning_rate": 7.498735920108284e-06,
      "loss": 0.4432,
      "mean_token_accuracy": 0.8518370389938354,
      "num_tokens": 960940209.0,
      "step": 28685
    },
    {
      "epoch": 1.7126313276026743,
      "grad_norm": 0.5141382813453674,
      "learning_rate": 7.493652411483251e-06,
      "loss": 0.4087,
      "mean_token_accuracy": 0.8628951430320739,
      "num_tokens": 961107889.0,
      "step": 28690
    },
    {
      "epoch": 1.7129297994269341,
      "grad_norm": 0.5095019936561584,
      "learning_rate": 7.488573775773447e-06,
      "loss": 0.4097,
      "mean_token_accuracy": 0.8610103845596313,
      "num_tokens": 961275569.0,
      "step": 28695
    },
    {
      "epoch": 1.713228271251194,
      "grad_norm": 0.5133291482925415,
      "learning_rate": 7.483500014215872e-06,
      "loss": 0.4543,
      "mean_token_accuracy": 0.8493200540542603,
      "num_tokens": 961443249.0,
      "step": 28700
    },
    {
      "epoch": 1.7135267430754537,
      "grad_norm": 0.48671406507492065,
      "learning_rate": 7.4784311280463205e-06,
      "loss": 0.444,
      "mean_token_accuracy": 0.8509841322898865,
      "num_tokens": 961610929.0,
      "step": 28705
    },
    {
      "epoch": 1.7138252148997135,
      "grad_norm": 0.4508340060710907,
      "learning_rate": 7.47336711849943e-06,
      "loss": 0.4076,
      "mean_token_accuracy": 0.8617439985275268,
      "num_tokens": 961778609.0,
      "step": 28710
    },
    {
      "epoch": 1.7141236867239733,
      "grad_norm": 0.4830595850944519,
      "learning_rate": 7.468307986808631e-06,
      "loss": 0.4253,
      "mean_token_accuracy": 0.8575331091880798,
      "num_tokens": 961946289.0,
      "step": 28715
    },
    {
      "epoch": 1.714422158548233,
      "grad_norm": 0.4971616268157959,
      "learning_rate": 7.4632537342061605e-06,
      "loss": 0.4057,
      "mean_token_accuracy": 0.8620004773139953,
      "num_tokens": 962113969.0,
      "step": 28720
    },
    {
      "epoch": 1.7147206303724927,
      "grad_norm": 0.5192204117774963,
      "learning_rate": 7.458204361923078e-06,
      "loss": 0.4302,
      "mean_token_accuracy": 0.854570722579956,
      "num_tokens": 962281259.0,
      "step": 28725
    },
    {
      "epoch": 1.7150191021967527,
      "grad_norm": 0.5822002291679382,
      "learning_rate": 7.4531598711892524e-06,
      "loss": 0.4206,
      "mean_token_accuracy": 0.8583144426345826,
      "num_tokens": 962448939.0,
      "step": 28730
    },
    {
      "epoch": 1.7153175740210123,
      "grad_norm": 0.5741744041442871,
      "learning_rate": 7.448120263233369e-06,
      "loss": 0.4437,
      "mean_token_accuracy": 0.852266502380371,
      "num_tokens": 962616619.0,
      "step": 28735
    },
    {
      "epoch": 1.7156160458452723,
      "grad_norm": 0.5148447155952454,
      "learning_rate": 7.443085539282902e-06,
      "loss": 0.436,
      "mean_token_accuracy": 0.8529404759407043,
      "num_tokens": 962784299.0,
      "step": 28740
    },
    {
      "epoch": 1.7159145176695318,
      "grad_norm": 0.5515783429145813,
      "learning_rate": 7.438055700564164e-06,
      "loss": 0.46,
      "mean_token_accuracy": 0.8464213252067566,
      "num_tokens": 962951979.0,
      "step": 28745
    },
    {
      "epoch": 1.7162129894937919,
      "grad_norm": 0.5184022784233093,
      "learning_rate": 7.4330307483022594e-06,
      "loss": 0.3965,
      "mean_token_accuracy": 0.8646069407463074,
      "num_tokens": 963119659.0,
      "step": 28750
    },
    {
      "epoch": 1.7165114613180514,
      "grad_norm": 0.5345397591590881,
      "learning_rate": 7.428010683721109e-06,
      "loss": 0.4516,
      "mean_token_accuracy": 0.8486043095588685,
      "num_tokens": 963287339.0,
      "step": 28755
    },
    {
      "epoch": 1.7168099331423115,
      "grad_norm": 0.5140678882598877,
      "learning_rate": 7.422995508043451e-06,
      "loss": 0.3867,
      "mean_token_accuracy": 0.8687820553779602,
      "num_tokens": 963455019.0,
      "step": 28760
    },
    {
      "epoch": 1.717108404966571,
      "grad_norm": 0.5007964968681335,
      "learning_rate": 7.417985222490803e-06,
      "loss": 0.4303,
      "mean_token_accuracy": 0.8557855248451233,
      "num_tokens": 963622699.0,
      "step": 28765
    },
    {
      "epoch": 1.717406876790831,
      "grad_norm": 0.514384388923645,
      "learning_rate": 7.412979828283531e-06,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8426458239555359,
      "num_tokens": 963790379.0,
      "step": 28770
    },
    {
      "epoch": 1.7177053486150906,
      "grad_norm": 0.5809042453765869,
      "learning_rate": 7.407979326640795e-06,
      "loss": 0.4444,
      "mean_token_accuracy": 0.853095555305481,
      "num_tokens": 963958059.0,
      "step": 28775
    },
    {
      "epoch": 1.7180038204393506,
      "grad_norm": 0.46283888816833496,
      "learning_rate": 7.40298371878054e-06,
      "loss": 0.4431,
      "mean_token_accuracy": 0.8503161191940307,
      "num_tokens": 964125739.0,
      "step": 28780
    },
    {
      "epoch": 1.7183022922636102,
      "grad_norm": 0.5773378610610962,
      "learning_rate": 7.397993005919551e-06,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8577418327331543,
      "num_tokens": 964293419.0,
      "step": 28785
    },
    {
      "epoch": 1.7186007640878702,
      "grad_norm": 0.5771684050559998,
      "learning_rate": 7.393007189273404e-06,
      "loss": 0.4561,
      "mean_token_accuracy": 0.8474770426750183,
      "num_tokens": 964461099.0,
      "step": 28790
    },
    {
      "epoch": 1.7188992359121298,
      "grad_norm": 0.527180016040802,
      "learning_rate": 7.38802627005649e-06,
      "loss": 0.4289,
      "mean_token_accuracy": 0.856316351890564,
      "num_tokens": 964628779.0,
      "step": 28795
    },
    {
      "epoch": 1.7191977077363898,
      "grad_norm": 0.5195962190628052,
      "learning_rate": 7.3830502494820066e-06,
      "loss": 0.4409,
      "mean_token_accuracy": 0.8521770238876343,
      "num_tokens": 964796459.0,
      "step": 28800
    },
    {
      "epoch": 1.7194961795606494,
      "grad_norm": 0.5164055228233337,
      "learning_rate": 7.3780791287619496e-06,
      "loss": 0.4091,
      "mean_token_accuracy": 0.8626327157020569,
      "num_tokens": 964964139.0,
      "step": 28805
    },
    {
      "epoch": 1.7197946513849094,
      "grad_norm": 0.5179728865623474,
      "learning_rate": 7.373112909107126e-06,
      "loss": 0.4101,
      "mean_token_accuracy": 0.8619348645210266,
      "num_tokens": 965131819.0,
      "step": 28810
    },
    {
      "epoch": 1.720093123209169,
      "grad_norm": 0.5365899801254272,
      "learning_rate": 7.368151591727153e-06,
      "loss": 0.419,
      "mean_token_accuracy": 0.857455563545227,
      "num_tokens": 965299499.0,
      "step": 28815
    },
    {
      "epoch": 1.720391595033429,
      "grad_norm": 0.5447962284088135,
      "learning_rate": 7.363195177830451e-06,
      "loss": 0.431,
      "mean_token_accuracy": 0.8541273832321167,
      "num_tokens": 965467179.0,
      "step": 28820
    },
    {
      "epoch": 1.7206900668576885,
      "grad_norm": 0.48601073026657104,
      "learning_rate": 7.358243668624254e-06,
      "loss": 0.4281,
      "mean_token_accuracy": 0.8566503643989563,
      "num_tokens": 965634859.0,
      "step": 28825
    },
    {
      "epoch": 1.7209885386819486,
      "grad_norm": 0.5295308828353882,
      "learning_rate": 7.353297065314572e-06,
      "loss": 0.4307,
      "mean_token_accuracy": 0.8569008708000183,
      "num_tokens": 965802539.0,
      "step": 28830
    },
    {
      "epoch": 1.7212870105062081,
      "grad_norm": 0.5698486566543579,
      "learning_rate": 7.348355369106268e-06,
      "loss": 0.4398,
      "mean_token_accuracy": 0.8530418872833252,
      "num_tokens": 965970219.0,
      "step": 28835
    },
    {
      "epoch": 1.7215854823304682,
      "grad_norm": 0.4656085968017578,
      "learning_rate": 7.343418581202964e-06,
      "loss": 0.3996,
      "mean_token_accuracy": 0.8655970335006714,
      "num_tokens": 966137899.0,
      "step": 28840
    },
    {
      "epoch": 1.7218839541547277,
      "grad_norm": 0.5593969225883484,
      "learning_rate": 7.3384867028071135e-06,
      "loss": 0.4967,
      "mean_token_accuracy": 0.8361266851425171,
      "num_tokens": 966305579.0,
      "step": 28845
    },
    {
      "epoch": 1.7221824259789877,
      "grad_norm": 0.4897625148296356,
      "learning_rate": 7.333559735119966e-06,
      "loss": 0.4224,
      "mean_token_accuracy": 0.8582011222839355,
      "num_tokens": 966473259.0,
      "step": 28850
    },
    {
      "epoch": 1.7224808978032473,
      "grad_norm": 0.5123838186264038,
      "learning_rate": 7.328637679341572e-06,
      "loss": 0.3894,
      "mean_token_accuracy": 0.8688894152641297,
      "num_tokens": 966640939.0,
      "step": 28855
    },
    {
      "epoch": 1.7227793696275073,
      "grad_norm": 0.5210784077644348,
      "learning_rate": 7.3237205366708e-06,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8524513959884643,
      "num_tokens": 966808619.0,
      "step": 28860
    },
    {
      "epoch": 1.723077841451767,
      "grad_norm": 0.4530964195728302,
      "learning_rate": 7.318808308305295e-06,
      "loss": 0.4,
      "mean_token_accuracy": 0.8650363922119141,
      "num_tokens": 966976299.0,
      "step": 28865
    },
    {
      "epoch": 1.7233763132760267,
      "grad_norm": 0.5902482271194458,
      "learning_rate": 7.31390099544153e-06,
      "loss": 0.4581,
      "mean_token_accuracy": 0.8467791914939881,
      "num_tokens": 967143979.0,
      "step": 28870
    },
    {
      "epoch": 1.7236747851002865,
      "grad_norm": 0.5239464640617371,
      "learning_rate": 7.308998599274772e-06,
      "loss": 0.4526,
      "mean_token_accuracy": 0.8492007613182068,
      "num_tokens": 967311659.0,
      "step": 28875
    },
    {
      "epoch": 1.7239732569245463,
      "grad_norm": 0.4668653905391693,
      "learning_rate": 7.304101120999086e-06,
      "loss": 0.396,
      "mean_token_accuracy": 0.86689133644104,
      "num_tokens": 967479339.0,
      "step": 28880
    },
    {
      "epoch": 1.724271728748806,
      "grad_norm": 0.4923475682735443,
      "learning_rate": 7.299208561807351e-06,
      "loss": 0.4622,
      "mean_token_accuracy": 0.8459918856620788,
      "num_tokens": 967647019.0,
      "step": 28885
    },
    {
      "epoch": 1.7245702005730659,
      "grad_norm": 0.5477373003959656,
      "learning_rate": 7.294320922891225e-06,
      "loss": 0.4121,
      "mean_token_accuracy": 0.861916983127594,
      "num_tokens": 967814699.0,
      "step": 28890
    },
    {
      "epoch": 1.7248686723973257,
      "grad_norm": 0.4551163613796234,
      "learning_rate": 7.289438205441201e-06,
      "loss": 0.4339,
      "mean_token_accuracy": 0.8527881026268005,
      "num_tokens": 967978525.0,
      "step": 28895
    },
    {
      "epoch": 1.7251671442215855,
      "grad_norm": 0.5083862543106079,
      "learning_rate": 7.284560410646544e-06,
      "loss": 0.4332,
      "mean_token_accuracy": 0.8555469393730164,
      "num_tokens": 968146205.0,
      "step": 28900
    },
    {
      "epoch": 1.7254656160458453,
      "grad_norm": 0.46898481249809265,
      "learning_rate": 7.279687539695332e-06,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8623463988304139,
      "num_tokens": 968313885.0,
      "step": 28905
    },
    {
      "epoch": 1.725764087870105,
      "grad_norm": 0.49112948775291443,
      "learning_rate": 7.274819593774443e-06,
      "loss": 0.4341,
      "mean_token_accuracy": 0.8541274070739746,
      "num_tokens": 968481565.0,
      "step": 28910
    },
    {
      "epoch": 1.7260625596943648,
      "grad_norm": 0.5248199701309204,
      "learning_rate": 7.269956574069557e-06,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8411189317703247,
      "num_tokens": 968649245.0,
      "step": 28915
    },
    {
      "epoch": 1.7263610315186246,
      "grad_norm": 0.49649155139923096,
      "learning_rate": 7.265098481765156e-06,
      "loss": 0.4202,
      "mean_token_accuracy": 0.8583621621131897,
      "num_tokens": 968816925.0,
      "step": 28920
    },
    {
      "epoch": 1.7266595033428844,
      "grad_norm": 0.510306715965271,
      "learning_rate": 7.260245318044506e-06,
      "loss": 0.3827,
      "mean_token_accuracy": 0.8696409344673157,
      "num_tokens": 968984605.0,
      "step": 28925
    },
    {
      "epoch": 1.7269579751671442,
      "grad_norm": 0.4668762683868408,
      "learning_rate": 7.255397084089695e-06,
      "loss": 0.4285,
      "mean_token_accuracy": 0.8577836155891418,
      "num_tokens": 969152285.0,
      "step": 28930
    },
    {
      "epoch": 1.727256446991404,
      "grad_norm": 0.4967258870601654,
      "learning_rate": 7.250553781081597e-06,
      "loss": 0.4556,
      "mean_token_accuracy": 0.8465525388717652,
      "num_tokens": 969319965.0,
      "step": 28935
    },
    {
      "epoch": 1.7275549188156638,
      "grad_norm": 0.4535907804965973,
      "learning_rate": 7.245715410199885e-06,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8593045592308044,
      "num_tokens": 969487645.0,
      "step": 28940
    },
    {
      "epoch": 1.7278533906399236,
      "grad_norm": 0.5136157274246216,
      "learning_rate": 7.2408819726230425e-06,
      "loss": 0.3981,
      "mean_token_accuracy": 0.8662173509597778,
      "num_tokens": 969655325.0,
      "step": 28945
    },
    {
      "epoch": 1.7281518624641834,
      "grad_norm": 0.5177064538002014,
      "learning_rate": 7.2360534695283314e-06,
      "loss": 0.4551,
      "mean_token_accuracy": 0.8487236142158509,
      "num_tokens": 969823005.0,
      "step": 28950
    },
    {
      "epoch": 1.7284503342884432,
      "grad_norm": 0.4764386713504791,
      "learning_rate": 7.231229902091827e-06,
      "loss": 0.4167,
      "mean_token_accuracy": 0.8607300400733948,
      "num_tokens": 969990685.0,
      "step": 28955
    },
    {
      "epoch": 1.728748806112703,
      "grad_norm": 0.5229875445365906,
      "learning_rate": 7.226411271488398e-06,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8314565300941468,
      "num_tokens": 970158365.0,
      "step": 28960
    },
    {
      "epoch": 1.7290472779369628,
      "grad_norm": 0.5183318257331848,
      "learning_rate": 7.221597578891713e-06,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8390671610832214,
      "num_tokens": 970326045.0,
      "step": 28965
    },
    {
      "epoch": 1.7293457497612226,
      "grad_norm": 0.48091086745262146,
      "learning_rate": 7.216788825474233e-06,
      "loss": 0.5109,
      "mean_token_accuracy": 0.8319217443466187,
      "num_tokens": 970493725.0,
      "step": 28970
    },
    {
      "epoch": 1.7296442215854824,
      "grad_norm": 0.5449036359786987,
      "learning_rate": 7.211985012407219e-06,
      "loss": 0.4715,
      "mean_token_accuracy": 0.8427252888679504,
      "num_tokens": 970654365.0,
      "step": 28975
    },
    {
      "epoch": 1.7299426934097422,
      "grad_norm": 0.5691753625869751,
      "learning_rate": 7.207186140860733e-06,
      "loss": 0.4263,
      "mean_token_accuracy": 0.857568895816803,
      "num_tokens": 970822045.0,
      "step": 28980
    },
    {
      "epoch": 1.730241165234002,
      "grad_norm": 0.5598579049110413,
      "learning_rate": 7.202392212003623e-06,
      "loss": 0.4246,
      "mean_token_accuracy": 0.8589526534080505,
      "num_tokens": 970989725.0,
      "step": 28985
    },
    {
      "epoch": 1.7305396370582617,
      "grad_norm": 0.5520233511924744,
      "learning_rate": 7.197603227003538e-06,
      "loss": 0.4255,
      "mean_token_accuracy": 0.8578790307044983,
      "num_tokens": 971157405.0,
      "step": 28990
    },
    {
      "epoch": 1.7308381088825215,
      "grad_norm": 0.4802533686161041,
      "learning_rate": 7.192819187026928e-06,
      "loss": 0.4604,
      "mean_token_accuracy": 0.8450733661651612,
      "num_tokens": 971325085.0,
      "step": 28995
    },
    {
      "epoch": 1.7311365807067811,
      "grad_norm": 0.5109880566596985,
      "learning_rate": 7.188040093239032e-06,
      "loss": 0.4186,
      "mean_token_accuracy": 0.8593582153320313,
      "num_tokens": 971492765.0,
      "step": 29000
    },
    {
      "epoch": 1.7314350525310411,
      "grad_norm": 0.4747738540172577,
      "learning_rate": 7.183265946803892e-06,
      "loss": 0.4051,
      "mean_token_accuracy": 0.8643803000450134,
      "num_tokens": 971660445.0,
      "step": 29005
    },
    {
      "epoch": 1.7317335243553007,
      "grad_norm": 0.5008866190910339,
      "learning_rate": 7.178496748884329e-06,
      "loss": 0.4108,
      "mean_token_accuracy": 0.8623046636581421,
      "num_tokens": 971828125.0,
      "step": 29010
    },
    {
      "epoch": 1.7320319961795607,
      "grad_norm": 0.5050458312034607,
      "learning_rate": 7.173732500641972e-06,
      "loss": 0.4519,
      "mean_token_accuracy": 0.8488190412521363,
      "num_tokens": 971995805.0,
      "step": 29015
    },
    {
      "epoch": 1.7323304680038203,
      "grad_norm": 0.5497936606407166,
      "learning_rate": 7.168973203237253e-06,
      "loss": 0.4582,
      "mean_token_accuracy": 0.8469521641731262,
      "num_tokens": 972163485.0,
      "step": 29020
    },
    {
      "epoch": 1.7326289398280803,
      "grad_norm": 0.5496740341186523,
      "learning_rate": 7.164218857829373e-06,
      "loss": 0.4554,
      "mean_token_accuracy": 0.8479840159416199,
      "num_tokens": 972331165.0,
      "step": 29025
    },
    {
      "epoch": 1.7329274116523399,
      "grad_norm": 0.5461985468864441,
      "learning_rate": 7.15946946557635e-06,
      "loss": 0.4564,
      "mean_token_accuracy": 0.8468149900436401,
      "num_tokens": 972498845.0,
      "step": 29030
    },
    {
      "epoch": 1.7332258834766,
      "grad_norm": 0.5176361203193665,
      "learning_rate": 7.1547250276349746e-06,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8657998323440552,
      "num_tokens": 972666525.0,
      "step": 29035
    },
    {
      "epoch": 1.7335243553008595,
      "grad_norm": 0.5617021322250366,
      "learning_rate": 7.149985545160852e-06,
      "loss": 0.4602,
      "mean_token_accuracy": 0.8467255115509034,
      "num_tokens": 972834205.0,
      "step": 29040
    },
    {
      "epoch": 1.7338228271251195,
      "grad_norm": 0.5165064334869385,
      "learning_rate": 7.1452510193083745e-06,
      "loss": 0.459,
      "mean_token_accuracy": 0.8488667726516723,
      "num_tokens": 973001885.0,
      "step": 29045
    },
    {
      "epoch": 1.734121298949379,
      "grad_norm": 0.5426809191703796,
      "learning_rate": 7.1405214512307085e-06,
      "loss": 0.4268,
      "mean_token_accuracy": 0.8566265106201172,
      "num_tokens": 973169565.0,
      "step": 29050
    },
    {
      "epoch": 1.734419770773639,
      "grad_norm": 0.5158417820930481,
      "learning_rate": 7.135796842079841e-06,
      "loss": 0.4816,
      "mean_token_accuracy": 0.8405284523963928,
      "num_tokens": 973337245.0,
      "step": 29055
    },
    {
      "epoch": 1.7347182425978986,
      "grad_norm": 0.5401809811592102,
      "learning_rate": 7.13107719300653e-06,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8447095274925231,
      "num_tokens": 973504925.0,
      "step": 29060
    },
    {
      "epoch": 1.7350167144221587,
      "grad_norm": 0.5143449902534485,
      "learning_rate": 7.126362505160339e-06,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8531253695487976,
      "num_tokens": 973672605.0,
      "step": 29065
    },
    {
      "epoch": 1.7353151862464182,
      "grad_norm": 0.47009438276290894,
      "learning_rate": 7.1216527796896126e-06,
      "loss": 0.4591,
      "mean_token_accuracy": 0.8464630842208862,
      "num_tokens": 973840285.0,
      "step": 29070
    },
    {
      "epoch": 1.7356136580706782,
      "grad_norm": 0.5764166116714478,
      "learning_rate": 7.11694801774149e-06,
      "loss": 0.4459,
      "mean_token_accuracy": 0.8515089988708496,
      "num_tokens": 974007965.0,
      "step": 29075
    },
    {
      "epoch": 1.7359121298949378,
      "grad_norm": 0.5124257802963257,
      "learning_rate": 7.112248220461908e-06,
      "loss": 0.453,
      "mean_token_accuracy": 0.8497912406921386,
      "num_tokens": 974175645.0,
      "step": 29080
    },
    {
      "epoch": 1.7362106017191978,
      "grad_norm": 0.4700307548046112,
      "learning_rate": 7.107553388995586e-06,
      "loss": 0.4199,
      "mean_token_accuracy": 0.8593761086463928,
      "num_tokens": 974343325.0,
      "step": 29085
    },
    {
      "epoch": 1.7365090735434574,
      "grad_norm": 0.45867449045181274,
      "learning_rate": 7.102863524486042e-06,
      "loss": 0.4006,
      "mean_token_accuracy": 0.8648157000541687,
      "num_tokens": 974511005.0,
      "step": 29090
    },
    {
      "epoch": 1.7368075453677174,
      "grad_norm": 0.5127758979797363,
      "learning_rate": 7.098178628075568e-06,
      "loss": 0.4498,
      "mean_token_accuracy": 0.8496123075485229,
      "num_tokens": 974678685.0,
      "step": 29095
    },
    {
      "epoch": 1.737106017191977,
      "grad_norm": 0.5111568570137024,
      "learning_rate": 7.0934987009052655e-06,
      "loss": 0.4442,
      "mean_token_accuracy": 0.8514195442199707,
      "num_tokens": 974846365.0,
      "step": 29100
    },
    {
      "epoch": 1.737404489016237,
      "grad_norm": 0.5325424671173096,
      "learning_rate": 7.0888237441150164e-06,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8515328764915466,
      "num_tokens": 975014045.0,
      "step": 29105
    },
    {
      "epoch": 1.7377029608404966,
      "grad_norm": 0.520170271396637,
      "learning_rate": 7.084153758843492e-06,
      "loss": 0.4009,
      "mean_token_accuracy": 0.8634319543838501,
      "num_tokens": 975181725.0,
      "step": 29110
    },
    {
      "epoch": 1.7380014326647566,
      "grad_norm": 0.5719833970069885,
      "learning_rate": 7.079488746228152e-06,
      "loss": 0.4187,
      "mean_token_accuracy": 0.8595729470252991,
      "num_tokens": 975349405.0,
      "step": 29115
    },
    {
      "epoch": 1.7382999044890162,
      "grad_norm": 0.48435789346694946,
      "learning_rate": 7.074828707405247e-06,
      "loss": 0.4097,
      "mean_token_accuracy": 0.8607479453086853,
      "num_tokens": 975517085.0,
      "step": 29120
    },
    {
      "epoch": 1.7385983763132762,
      "grad_norm": 0.45030462741851807,
      "learning_rate": 7.0701736435098155e-06,
      "loss": 0.3969,
      "mean_token_accuracy": 0.866628885269165,
      "num_tokens": 975684765.0,
      "step": 29125
    },
    {
      "epoch": 1.7388968481375358,
      "grad_norm": 0.47363996505737305,
      "learning_rate": 7.065523555675688e-06,
      "loss": 0.4299,
      "mean_token_accuracy": 0.8558690190315247,
      "num_tokens": 975852445.0,
      "step": 29130
    },
    {
      "epoch": 1.7391953199617958,
      "grad_norm": 0.5388246774673462,
      "learning_rate": 7.060878445035475e-06,
      "loss": 0.4663,
      "mean_token_accuracy": 0.8441667675971984,
      "num_tokens": 976020125.0,
      "step": 29135
    },
    {
      "epoch": 1.7394937917860553,
      "grad_norm": 0.5351729393005371,
      "learning_rate": 7.0562383127205755e-06,
      "loss": 0.4287,
      "mean_token_accuracy": 0.8562984466552734,
      "num_tokens": 976187805.0,
      "step": 29140
    },
    {
      "epoch": 1.7397922636103151,
      "grad_norm": 0.5127094984054565,
      "learning_rate": 7.051603159861185e-06,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8518012642860413,
      "num_tokens": 976355485.0,
      "step": 29145
    },
    {
      "epoch": 1.740090735434575,
      "grad_norm": 0.47143083810806274,
      "learning_rate": 7.046972987586278e-06,
      "loss": 0.4619,
      "mean_token_accuracy": 0.8455684065818787,
      "num_tokens": 976523165.0,
      "step": 29150
    },
    {
      "epoch": 1.7403892072588347,
      "grad_norm": 0.5048629641532898,
      "learning_rate": 7.042347797023627e-06,
      "loss": 0.4374,
      "mean_token_accuracy": 0.8541870474815368,
      "num_tokens": 976690845.0,
      "step": 29155
    },
    {
      "epoch": 1.7406876790830945,
      "grad_norm": 0.5515156388282776,
      "learning_rate": 7.037727589299763e-06,
      "loss": 0.4351,
      "mean_token_accuracy": 0.8531372904777527,
      "num_tokens": 976858525.0,
      "step": 29160
    },
    {
      "epoch": 1.7409861509073543,
      "grad_norm": 0.5150281190872192,
      "learning_rate": 7.033112365540046e-06,
      "loss": 0.4113,
      "mean_token_accuracy": 0.8639532208442688,
      "num_tokens": 977020654.0,
      "step": 29165
    },
    {
      "epoch": 1.741284622731614,
      "grad_norm": 0.4680708348751068,
      "learning_rate": 7.0285021268685815e-06,
      "loss": 0.3796,
      "mean_token_accuracy": 0.8723666787147522,
      "num_tokens": 977188334.0,
      "step": 29170
    },
    {
      "epoch": 1.741583094555874,
      "grad_norm": 0.5442808866500854,
      "learning_rate": 7.0238968744082895e-06,
      "loss": 0.4461,
      "mean_token_accuracy": 0.8513360381126404,
      "num_tokens": 977356014.0,
      "step": 29175
    },
    {
      "epoch": 1.7418815663801337,
      "grad_norm": 0.5612311959266663,
      "learning_rate": 7.019296609280852e-06,
      "loss": 0.469,
      "mean_token_accuracy": 0.8423177838325501,
      "num_tokens": 977523694.0,
      "step": 29180
    },
    {
      "epoch": 1.7421800382043935,
      "grad_norm": 0.5339342355728149,
      "learning_rate": 7.01470133260676e-06,
      "loss": 0.4202,
      "mean_token_accuracy": 0.8588035225868225,
      "num_tokens": 977691374.0,
      "step": 29185
    },
    {
      "epoch": 1.7424785100286533,
      "grad_norm": 0.47362253069877625,
      "learning_rate": 7.010111045505276e-06,
      "loss": 0.4004,
      "mean_token_accuracy": 0.8640641808509827,
      "num_tokens": 977859054.0,
      "step": 29190
    },
    {
      "epoch": 1.742776981852913,
      "grad_norm": 0.49893632531166077,
      "learning_rate": 7.0055257490944435e-06,
      "loss": 0.4327,
      "mean_token_accuracy": 0.8548312067985535,
      "num_tokens": 978026734.0,
      "step": 29195
    },
    {
      "epoch": 1.7430754536771729,
      "grad_norm": 0.498502254486084,
      "learning_rate": 7.0009454444911e-06,
      "loss": 0.4054,
      "mean_token_accuracy": 0.8631158351898194,
      "num_tokens": 978194414.0,
      "step": 29200
    },
    {
      "epoch": 1.7433739255014327,
      "grad_norm": 0.47744449973106384,
      "learning_rate": 6.996370132810863e-06,
      "loss": 0.4117,
      "mean_token_accuracy": 0.8615889191627503,
      "num_tokens": 978362094.0,
      "step": 29205
    },
    {
      "epoch": 1.7436723973256925,
      "grad_norm": 0.5168974995613098,
      "learning_rate": 6.991799815168132e-06,
      "loss": 0.4116,
      "mean_token_accuracy": 0.861678397655487,
      "num_tokens": 978529774.0,
      "step": 29210
    },
    {
      "epoch": 1.7439708691499523,
      "grad_norm": 0.5573127269744873,
      "learning_rate": 6.987234492676101e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.8568233370780944,
      "num_tokens": 978697454.0,
      "step": 29215
    },
    {
      "epoch": 1.744269340974212,
      "grad_norm": 0.5329184532165527,
      "learning_rate": 6.982674166446721e-06,
      "loss": 0.4745,
      "mean_token_accuracy": 0.8439878225326538,
      "num_tokens": 978865134.0,
      "step": 29220
    },
    {
      "epoch": 1.7445678127984718,
      "grad_norm": 0.5054529309272766,
      "learning_rate": 6.978118837590762e-06,
      "loss": 0.4428,
      "mean_token_accuracy": 0.852266502380371,
      "num_tokens": 979032814.0,
      "step": 29225
    },
    {
      "epoch": 1.7448662846227316,
      "grad_norm": 0.5723335146903992,
      "learning_rate": 6.973568507217748e-06,
      "loss": 0.414,
      "mean_token_accuracy": 0.8620839715003967,
      "num_tokens": 979200494.0,
      "step": 29230
    },
    {
      "epoch": 1.7451647564469914,
      "grad_norm": 0.5810419917106628,
      "learning_rate": 6.969023176435998e-06,
      "loss": 0.4487,
      "mean_token_accuracy": 0.8496063590049744,
      "num_tokens": 979368174.0,
      "step": 29235
    },
    {
      "epoch": 1.7454632282712512,
      "grad_norm": 0.48232221603393555,
      "learning_rate": 6.964482846352613e-06,
      "loss": 0.4197,
      "mean_token_accuracy": 0.8599069476127624,
      "num_tokens": 979535854.0,
      "step": 29240
    },
    {
      "epoch": 1.745761700095511,
      "grad_norm": 0.5374194383621216,
      "learning_rate": 6.9599475180734715e-06,
      "loss": 0.4154,
      "mean_token_accuracy": 0.8601992130279541,
      "num_tokens": 979703534.0,
      "step": 29245
    },
    {
      "epoch": 1.7460601719197708,
      "grad_norm": 0.5285108089447021,
      "learning_rate": 6.955417192703242e-06,
      "loss": 0.4761,
      "mean_token_accuracy": 0.8415483593940735,
      "num_tokens": 979871214.0,
      "step": 29250
    },
    {
      "epoch": 1.7463586437440306,
      "grad_norm": 0.4871658980846405,
      "learning_rate": 6.950891871345362e-06,
      "loss": 0.4199,
      "mean_token_accuracy": 0.8586902022361755,
      "num_tokens": 980038894.0,
      "step": 29255
    },
    {
      "epoch": 1.7466571155682904,
      "grad_norm": 0.4737515449523926,
      "learning_rate": 6.946371555102058e-06,
      "loss": 0.4323,
      "mean_token_accuracy": 0.8547477006912232,
      "num_tokens": 980206574.0,
      "step": 29260
    },
    {
      "epoch": 1.7469555873925502,
      "grad_norm": 0.5292275547981262,
      "learning_rate": 6.941856245074341e-06,
      "loss": 0.4226,
      "mean_token_accuracy": 0.8580042958259583,
      "num_tokens": 980374254.0,
      "step": 29265
    },
    {
      "epoch": 1.74725405921681,
      "grad_norm": 0.4743780791759491,
      "learning_rate": 6.937345942361992e-06,
      "loss": 0.4303,
      "mean_token_accuracy": 0.8553918600082397,
      "num_tokens": 980541934.0,
      "step": 29270
    },
    {
      "epoch": 1.7475525310410696,
      "grad_norm": 0.5489332675933838,
      "learning_rate": 6.932840648063588e-06,
      "loss": 0.4259,
      "mean_token_accuracy": 0.8564177393913269,
      "num_tokens": 980709614.0,
      "step": 29275
    },
    {
      "epoch": 1.7478510028653296,
      "grad_norm": 0.5025535225868225,
      "learning_rate": 6.928340363276464e-06,
      "loss": 0.4545,
      "mean_token_accuracy": 0.847667908668518,
      "num_tokens": 980877294.0,
      "step": 29280
    },
    {
      "epoch": 1.7481494746895891,
      "grad_norm": 0.5074628591537476,
      "learning_rate": 6.9238450890967575e-06,
      "loss": 0.4301,
      "mean_token_accuracy": 0.8545985817909241,
      "num_tokens": 981044974.0,
      "step": 29285
    },
    {
      "epoch": 1.7484479465138492,
      "grad_norm": 0.5280393362045288,
      "learning_rate": 6.919354826619369e-06,
      "loss": 0.4493,
      "mean_token_accuracy": 0.8487355351448059,
      "num_tokens": 981212654.0,
      "step": 29290
    },
    {
      "epoch": 1.7487464183381087,
      "grad_norm": 0.4847422242164612,
      "learning_rate": 6.914869576937988e-06,
      "loss": 0.4602,
      "mean_token_accuracy": 0.8480973482131958,
      "num_tokens": 981380334.0,
      "step": 29295
    },
    {
      "epoch": 1.7490448901623687,
      "grad_norm": 0.48247066140174866,
      "learning_rate": 6.910389341145085e-06,
      "loss": 0.4243,
      "mean_token_accuracy": 0.8567696452140808,
      "num_tokens": 981548014.0,
      "step": 29300
    },
    {
      "epoch": 1.7493433619866283,
      "grad_norm": 0.5005117654800415,
      "learning_rate": 6.905914120331891e-06,
      "loss": 0.4526,
      "mean_token_accuracy": 0.8484253883361816,
      "num_tokens": 981715694.0,
      "step": 29305
    },
    {
      "epoch": 1.7496418338108883,
      "grad_norm": 0.5716114640235901,
      "learning_rate": 6.901443915588443e-06,
      "loss": 0.4315,
      "mean_token_accuracy": 0.8556662440299988,
      "num_tokens": 981883374.0,
      "step": 29310
    },
    {
      "epoch": 1.749940305635148,
      "grad_norm": 0.6131590008735657,
      "learning_rate": 6.8969787280035306e-06,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8557735919952393,
      "num_tokens": 982051054.0,
      "step": 29315
    },
    {
      "epoch": 1.750238777459408,
      "grad_norm": 0.5128840804100037,
      "learning_rate": 6.892518558664739e-06,
      "loss": 0.4395,
      "mean_token_accuracy": 0.8533997297286987,
      "num_tokens": 982218734.0,
      "step": 29320
    },
    {
      "epoch": 1.7505372492836675,
      "grad_norm": 0.49456241726875305,
      "learning_rate": 6.888063408658421e-06,
      "loss": 0.3973,
      "mean_token_accuracy": 0.8649588465690613,
      "num_tokens": 982386414.0,
      "step": 29325
    },
    {
      "epoch": 1.7508357211079275,
      "grad_norm": 0.49524593353271484,
      "learning_rate": 6.883613279069712e-06,
      "loss": 0.398,
      "mean_token_accuracy": 0.8655433654785156,
      "num_tokens": 982554094.0,
      "step": 29330
    },
    {
      "epoch": 1.751134192932187,
      "grad_norm": 0.4886665940284729,
      "learning_rate": 6.879168170982531e-06,
      "loss": 0.4201,
      "mean_token_accuracy": 0.8586782693862915,
      "num_tokens": 982721774.0,
      "step": 29335
    },
    {
      "epoch": 1.751432664756447,
      "grad_norm": 0.49355655908584595,
      "learning_rate": 6.874728085479551e-06,
      "loss": 0.4182,
      "mean_token_accuracy": 0.8605272531509399,
      "num_tokens": 982889454.0,
      "step": 29340
    },
    {
      "epoch": 1.7517311365807067,
      "grad_norm": 0.507889449596405,
      "learning_rate": 6.870293023642244e-06,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8461708307266236,
      "num_tokens": 983057134.0,
      "step": 29345
    },
    {
      "epoch": 1.7520296084049667,
      "grad_norm": 0.4929352104663849,
      "learning_rate": 6.8658629865508585e-06,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8423476099967957,
      "num_tokens": 983224814.0,
      "step": 29350
    },
    {
      "epoch": 1.7523280802292263,
      "grad_norm": 0.5128406882286072,
      "learning_rate": 6.861437975284399e-06,
      "loss": 0.4205,
      "mean_token_accuracy": 0.8580937623977661,
      "num_tokens": 983392494.0,
      "step": 29355
    },
    {
      "epoch": 1.7526265520534863,
      "grad_norm": 0.5166106224060059,
      "learning_rate": 6.857017990920672e-06,
      "loss": 0.4722,
      "mean_token_accuracy": 0.8444411396980286,
      "num_tokens": 983560174.0,
      "step": 29360
    },
    {
      "epoch": 1.7529250238777458,
      "grad_norm": 0.5328486561775208,
      "learning_rate": 6.852603034536231e-06,
      "loss": 0.4444,
      "mean_token_accuracy": 0.8511332392692565,
      "num_tokens": 983727854.0,
      "step": 29365
    },
    {
      "epoch": 1.7532234957020059,
      "grad_norm": 0.48859071731567383,
      "learning_rate": 6.84819310720643e-06,
      "loss": 0.4608,
      "mean_token_accuracy": 0.8448586344718934,
      "num_tokens": 983895534.0,
      "step": 29370
    },
    {
      "epoch": 1.7535219675262654,
      "grad_norm": 0.5125442743301392,
      "learning_rate": 6.8437882100053946e-06,
      "loss": 0.4601,
      "mean_token_accuracy": 0.8469879508018494,
      "num_tokens": 984063214.0,
      "step": 29375
    },
    {
      "epoch": 1.7538204393505255,
      "grad_norm": 0.5006015300750732,
      "learning_rate": 6.839388344006004e-06,
      "loss": 0.4298,
      "mean_token_accuracy": 0.8558272600173951,
      "num_tokens": 984230894.0,
      "step": 29380
    },
    {
      "epoch": 1.754118911174785,
      "grad_norm": 0.5291780233383179,
      "learning_rate": 6.83499351027993e-06,
      "loss": 0.4062,
      "mean_token_accuracy": 0.862919008731842,
      "num_tokens": 984398574.0,
      "step": 29385
    },
    {
      "epoch": 1.754417382999045,
      "grad_norm": 0.5332061052322388,
      "learning_rate": 6.830603709897625e-06,
      "loss": 0.4418,
      "mean_token_accuracy": 0.8513121843338013,
      "num_tokens": 984566254.0,
      "step": 29390
    },
    {
      "epoch": 1.7547158548233046,
      "grad_norm": 0.5164974927902222,
      "learning_rate": 6.8262189439283004e-06,
      "loss": 0.4455,
      "mean_token_accuracy": 0.8509960651397706,
      "num_tokens": 984733934.0,
      "step": 29395
    },
    {
      "epoch": 1.7550143266475646,
      "grad_norm": 0.47510188817977905,
      "learning_rate": 6.82183921343994e-06,
      "loss": 0.4042,
      "mean_token_accuracy": 0.8634796619415284,
      "num_tokens": 984901614.0,
      "step": 29400
    },
    {
      "epoch": 1.7553127984718242,
      "grad_norm": 0.5421191453933716,
      "learning_rate": 6.81746451949932e-06,
      "loss": 0.4726,
      "mean_token_accuracy": 0.842317771911621,
      "num_tokens": 985069294.0,
      "step": 29405
    },
    {
      "epoch": 1.7556112702960842,
      "grad_norm": 0.48657265305519104,
      "learning_rate": 6.813094863171968e-06,
      "loss": 0.3967,
      "mean_token_accuracy": 0.8663724184036254,
      "num_tokens": 985236974.0,
      "step": 29410
    },
    {
      "epoch": 1.7559097421203438,
      "grad_norm": 0.4703061282634735,
      "learning_rate": 6.808730245522199e-06,
      "loss": 0.4188,
      "mean_token_accuracy": 0.8590003728866578,
      "num_tokens": 985404654.0,
      "step": 29415
    },
    {
      "epoch": 1.7562082139446036,
      "grad_norm": 0.5072619915008545,
      "learning_rate": 6.804370667613103e-06,
      "loss": 0.4632,
      "mean_token_accuracy": 0.8465823769569397,
      "num_tokens": 985572334.0,
      "step": 29420
    },
    {
      "epoch": 1.7565066857688634,
      "grad_norm": 0.5081788897514343,
      "learning_rate": 6.800016130506521e-06,
      "loss": 0.4306,
      "mean_token_accuracy": 0.8566324710845947,
      "num_tokens": 985740014.0,
      "step": 29425
    },
    {
      "epoch": 1.7568051575931232,
      "grad_norm": 0.4417765140533447,
      "learning_rate": 6.795666635263083e-06,
      "loss": 0.4129,
      "mean_token_accuracy": 0.8616784095764161,
      "num_tokens": 985907694.0,
      "step": 29430
    },
    {
      "epoch": 1.757103629417383,
      "grad_norm": 0.5713768601417542,
      "learning_rate": 6.7913221829422055e-06,
      "loss": 0.4495,
      "mean_token_accuracy": 0.8488071084022522,
      "num_tokens": 986075374.0,
      "step": 29435
    },
    {
      "epoch": 1.7574021012416428,
      "grad_norm": 0.5122180581092834,
      "learning_rate": 6.786982774602044e-06,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8512942910194397,
      "num_tokens": 986243054.0,
      "step": 29440
    },
    {
      "epoch": 1.7577005730659025,
      "grad_norm": 0.4664669632911682,
      "learning_rate": 6.782648411299544e-06,
      "loss": 0.4132,
      "mean_token_accuracy": 0.8611415982246399,
      "num_tokens": 986410734.0,
      "step": 29445
    },
    {
      "epoch": 1.7579990448901623,
      "grad_norm": 0.5467097163200378,
      "learning_rate": 6.778319094090427e-06,
      "loss": 0.4528,
      "mean_token_accuracy": 0.8484194159507752,
      "num_tokens": 986578414.0,
      "step": 29450
    },
    {
      "epoch": 1.7582975167144221,
      "grad_norm": 0.5622749924659729,
      "learning_rate": 6.77399482402917e-06,
      "loss": 0.4422,
      "mean_token_accuracy": 0.8527436494827271,
      "num_tokens": 986746094.0,
      "step": 29455
    },
    {
      "epoch": 1.758595988538682,
      "grad_norm": 0.5136642456054688,
      "learning_rate": 6.7696756021690356e-06,
      "loss": 0.421,
      "mean_token_accuracy": 0.8591971874237061,
      "num_tokens": 986913774.0,
      "step": 29460
    },
    {
      "epoch": 1.7588944603629417,
      "grad_norm": 0.5292744636535645,
      "learning_rate": 6.765361429562042e-06,
      "loss": 0.4342,
      "mean_token_accuracy": 0.8541631937026978,
      "num_tokens": 987081454.0,
      "step": 29465
    },
    {
      "epoch": 1.7591929321872015,
      "grad_norm": 0.5027296543121338,
      "learning_rate": 6.761052307258992e-06,
      "loss": 0.4121,
      "mean_token_accuracy": 0.8607479453086853,
      "num_tokens": 987249134.0,
      "step": 29470
    },
    {
      "epoch": 1.7594914040114613,
      "grad_norm": 0.5739250779151917,
      "learning_rate": 6.756748236309446e-06,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8539365410804749,
      "num_tokens": 987416814.0,
      "step": 29475
    },
    {
      "epoch": 1.759789875835721,
      "grad_norm": 0.48803815245628357,
      "learning_rate": 6.752449217761744e-06,
      "loss": 0.4259,
      "mean_token_accuracy": 0.8570976853370667,
      "num_tokens": 987584494.0,
      "step": 29480
    },
    {
      "epoch": 1.760088347659981,
      "grad_norm": 0.4690641760826111,
      "learning_rate": 6.748155252662996e-06,
      "loss": 0.4255,
      "mean_token_accuracy": 0.8577955484390258,
      "num_tokens": 987752174.0,
      "step": 29485
    },
    {
      "epoch": 1.7603868194842407,
      "grad_norm": 0.4950862228870392,
      "learning_rate": 6.743866342059066e-06,
      "loss": 0.4608,
      "mean_token_accuracy": 0.845288074016571,
      "num_tokens": 987919854.0,
      "step": 29490
    },
    {
      "epoch": 1.7606852913085005,
      "grad_norm": 0.5685482025146484,
      "learning_rate": 6.739582486994608e-06,
      "loss": 0.4432,
      "mean_token_accuracy": 0.8523320913314819,
      "num_tokens": 988087534.0,
      "step": 29495
    },
    {
      "epoch": 1.7609837631327603,
      "grad_norm": 0.48449113965034485,
      "learning_rate": 6.735303688513025e-06,
      "loss": 0.4256,
      "mean_token_accuracy": 0.859173321723938,
      "num_tokens": 988255214.0,
      "step": 29500
    },
    {
      "epoch": 1.76128223495702,
      "grad_norm": 0.4949661195278168,
      "learning_rate": 6.731029947656505e-06,
      "loss": 0.4143,
      "mean_token_accuracy": 0.8599248647689819,
      "num_tokens": 988422894.0,
      "step": 29505
    },
    {
      "epoch": 1.7615807067812799,
      "grad_norm": 0.46910420060157776,
      "learning_rate": 6.726761265465985e-06,
      "loss": 0.4161,
      "mean_token_accuracy": 0.8598831176757813,
      "num_tokens": 988590574.0,
      "step": 29510
    },
    {
      "epoch": 1.7618791786055397,
      "grad_norm": 0.548302173614502,
      "learning_rate": 6.722497642981192e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8521591305732727,
      "num_tokens": 988758254.0,
      "step": 29515
    },
    {
      "epoch": 1.7621776504297995,
      "grad_norm": 0.5400492548942566,
      "learning_rate": 6.7182390812406125e-06,
      "loss": 0.4209,
      "mean_token_accuracy": 0.8587379217147827,
      "num_tokens": 988925934.0,
      "step": 29520
    },
    {
      "epoch": 1.7624761222540593,
      "grad_norm": 0.531758189201355,
      "learning_rate": 6.713985581281487e-06,
      "loss": 0.4372,
      "mean_token_accuracy": 0.853996193408966,
      "num_tokens": 989093614.0,
      "step": 29525
    },
    {
      "epoch": 1.762774594078319,
      "grad_norm": 0.48997655510902405,
      "learning_rate": 6.70973714413984e-06,
      "loss": 0.4511,
      "mean_token_accuracy": 0.8480317234992981,
      "num_tokens": 989261294.0,
      "step": 29530
    },
    {
      "epoch": 1.7630730659025788,
      "grad_norm": 0.4956251084804535,
      "learning_rate": 6.705493770850455e-06,
      "loss": 0.4228,
      "mean_token_accuracy": 0.8595908403396606,
      "num_tokens": 989428974.0,
      "step": 29535
    },
    {
      "epoch": 1.7633715377268386,
      "grad_norm": 0.5295732617378235,
      "learning_rate": 6.701255462446881e-06,
      "loss": 0.4159,
      "mean_token_accuracy": 0.860926878452301,
      "num_tokens": 989596654.0,
      "step": 29540
    },
    {
      "epoch": 1.7636700095510984,
      "grad_norm": 0.5092737078666687,
      "learning_rate": 6.6970222199614474e-06,
      "loss": 0.4296,
      "mean_token_accuracy": 0.8560718059539795,
      "num_tokens": 989764334.0,
      "step": 29545
    },
    {
      "epoch": 1.7639684813753582,
      "grad_norm": 0.5308535695075989,
      "learning_rate": 6.692794044425228e-06,
      "loss": 0.4366,
      "mean_token_accuracy": 0.8535667419433594,
      "num_tokens": 989932014.0,
      "step": 29550
    },
    {
      "epoch": 1.764266953199618,
      "grad_norm": 0.48232805728912354,
      "learning_rate": 6.688570936868076e-06,
      "loss": 0.4279,
      "mean_token_accuracy": 0.8573243379592895,
      "num_tokens": 990099694.0,
      "step": 29555
    },
    {
      "epoch": 1.7645654250238776,
      "grad_norm": 0.5460061430931091,
      "learning_rate": 6.684352898318606e-06,
      "loss": 0.4795,
      "mean_token_accuracy": 0.839448893070221,
      "num_tokens": 990267374.0,
      "step": 29560
    },
    {
      "epoch": 1.7648638968481376,
      "grad_norm": 0.5096883177757263,
      "learning_rate": 6.680139929804202e-06,
      "loss": 0.4549,
      "mean_token_accuracy": 0.8480734825134277,
      "num_tokens": 990435054.0,
      "step": 29565
    },
    {
      "epoch": 1.7651623686723972,
      "grad_norm": 0.5248125195503235,
      "learning_rate": 6.675932032351008e-06,
      "loss": 0.4159,
      "mean_token_accuracy": 0.8602469205856323,
      "num_tokens": 990602734.0,
      "step": 29570
    },
    {
      "epoch": 1.7654608404966572,
      "grad_norm": 0.4921119511127472,
      "learning_rate": 6.671729206983934e-06,
      "loss": 0.4502,
      "mean_token_accuracy": 0.8517058253288269,
      "num_tokens": 990770414.0,
      "step": 29575
    },
    {
      "epoch": 1.7657593123209168,
      "grad_norm": 0.5222781896591187,
      "learning_rate": 6.667531454726664e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8724323034286499,
      "num_tokens": 990938094.0,
      "step": 29580
    },
    {
      "epoch": 1.7660577841451768,
      "grad_norm": 0.5048891305923462,
      "learning_rate": 6.663338776601626e-06,
      "loss": 0.4473,
      "mean_token_accuracy": 0.850751531124115,
      "num_tokens": 991105774.0,
      "step": 29585
    },
    {
      "epoch": 1.7663562559694364,
      "grad_norm": 0.49026796221733093,
      "learning_rate": 6.659151173630029e-06,
      "loss": 0.4047,
      "mean_token_accuracy": 0.864398181438446,
      "num_tokens": 991273454.0,
      "step": 29590
    },
    {
      "epoch": 1.7666547277936964,
      "grad_norm": 0.555770218372345,
      "learning_rate": 6.6549686468318454e-06,
      "loss": 0.4178,
      "mean_token_accuracy": 0.8583025217056275,
      "num_tokens": 991441134.0,
      "step": 29595
    },
    {
      "epoch": 1.766953199617956,
      "grad_norm": 0.4972131848335266,
      "learning_rate": 6.6507911972258e-06,
      "loss": 0.4104,
      "mean_token_accuracy": 0.8608612656593323,
      "num_tokens": 991608814.0,
      "step": 29600
    },
    {
      "epoch": 1.767251671442216,
      "grad_norm": 0.48660045862197876,
      "learning_rate": 6.646618825829394e-06,
      "loss": 0.4345,
      "mean_token_accuracy": 0.8554919004440308,
      "num_tokens": 991771670.0,
      "step": 29605
    },
    {
      "epoch": 1.7675501432664755,
      "grad_norm": 0.5368556380271912,
      "learning_rate": 6.642451533658879e-06,
      "loss": 0.4054,
      "mean_token_accuracy": 0.8636287689208985,
      "num_tokens": 991939350.0,
      "step": 29610
    },
    {
      "epoch": 1.7678486150907355,
      "grad_norm": 0.46552684903144836,
      "learning_rate": 6.638289321729283e-06,
      "loss": 0.4149,
      "mean_token_accuracy": 0.8622509837150574,
      "num_tokens": 992107030.0,
      "step": 29615
    },
    {
      "epoch": 1.7681470869149951,
      "grad_norm": 0.5402354598045349,
      "learning_rate": 6.634132191054379e-06,
      "loss": 0.4233,
      "mean_token_accuracy": 0.8583919763565063,
      "num_tokens": 992274710.0,
      "step": 29620
    },
    {
      "epoch": 1.7684455587392551,
      "grad_norm": 0.5115271210670471,
      "learning_rate": 6.6299801426467234e-06,
      "loss": 0.4332,
      "mean_token_accuracy": 0.8553441524505615,
      "num_tokens": 992442390.0,
      "step": 29625
    },
    {
      "epoch": 1.7687440305635147,
      "grad_norm": 0.49103888869285583,
      "learning_rate": 6.625833177517627e-06,
      "loss": 0.3992,
      "mean_token_accuracy": 0.8636526226997375,
      "num_tokens": 992610070.0,
      "step": 29630
    },
    {
      "epoch": 1.7690425023877747,
      "grad_norm": 0.4704672694206238,
      "learning_rate": 6.62169129667714e-06,
      "loss": 0.4018,
      "mean_token_accuracy": 0.8645353674888611,
      "num_tokens": 992777750.0,
      "step": 29635
    },
    {
      "epoch": 1.7693409742120343,
      "grad_norm": 0.571395754814148,
      "learning_rate": 6.617554501134118e-06,
      "loss": 0.4396,
      "mean_token_accuracy": 0.8529345035552979,
      "num_tokens": 992945430.0,
      "step": 29640
    },
    {
      "epoch": 1.7696394460362943,
      "grad_norm": 0.5068790316581726,
      "learning_rate": 6.61342279189614e-06,
      "loss": 0.412,
      "mean_token_accuracy": 0.8606405854225159,
      "num_tokens": 993113110.0,
      "step": 29645
    },
    {
      "epoch": 1.7699379178605539,
      "grad_norm": 0.5302010774612427,
      "learning_rate": 6.609296169969559e-06,
      "loss": 0.4267,
      "mean_token_accuracy": 0.8559704184532165,
      "num_tokens": 993280790.0,
      "step": 29650
    },
    {
      "epoch": 1.770236389684814,
      "grad_norm": 0.4594547748565674,
      "learning_rate": 6.6051746363594975e-06,
      "loss": 0.4412,
      "mean_token_accuracy": 0.8518251180648804,
      "num_tokens": 993448470.0,
      "step": 29655
    },
    {
      "epoch": 1.7705348615090735,
      "grad_norm": 0.5209713578224182,
      "learning_rate": 6.601058192069829e-06,
      "loss": 0.472,
      "mean_token_accuracy": 0.8433198094367981,
      "num_tokens": 993616150.0,
      "step": 29660
    },
    {
      "epoch": 1.7708333333333335,
      "grad_norm": 0.4633740782737732,
      "learning_rate": 6.596946838103189e-06,
      "loss": 0.379,
      "mean_token_accuracy": 0.8718776106834412,
      "num_tokens": 993783830.0,
      "step": 29665
    },
    {
      "epoch": 1.771131805157593,
      "grad_norm": 0.5139845013618469,
      "learning_rate": 6.592840575460968e-06,
      "loss": 0.4091,
      "mean_token_accuracy": 0.8618036508560181,
      "num_tokens": 993951510.0,
      "step": 29670
    },
    {
      "epoch": 1.771430276981853,
      "grad_norm": 0.4802762269973755,
      "learning_rate": 6.588739405143328e-06,
      "loss": 0.4141,
      "mean_token_accuracy": 0.8604795455932617,
      "num_tokens": 994119190.0,
      "step": 29675
    },
    {
      "epoch": 1.7717287488061126,
      "grad_norm": 0.5031047463417053,
      "learning_rate": 6.584643328149185e-06,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8474412441253663,
      "num_tokens": 994286870.0,
      "step": 29680
    },
    {
      "epoch": 1.7720272206303727,
      "grad_norm": 0.5848226547241211,
      "learning_rate": 6.5805523454762125e-06,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8492186546325684,
      "num_tokens": 994454550.0,
      "step": 29685
    },
    {
      "epoch": 1.7723256924546322,
      "grad_norm": 0.5669364929199219,
      "learning_rate": 6.576466458120848e-06,
      "loss": 0.4503,
      "mean_token_accuracy": 0.8499164938926697,
      "num_tokens": 994622230.0,
      "step": 29690
    },
    {
      "epoch": 1.772624164278892,
      "grad_norm": 0.5281694531440735,
      "learning_rate": 6.572385667078273e-06,
      "loss": 0.442,
      "mean_token_accuracy": 0.8521710515022278,
      "num_tokens": 994789910.0,
      "step": 29695
    },
    {
      "epoch": 1.7729226361031518,
      "grad_norm": 0.5648983716964722,
      "learning_rate": 6.568309973342457e-06,
      "loss": 0.4608,
      "mean_token_accuracy": 0.8459382057189941,
      "num_tokens": 994957590.0,
      "step": 29700
    },
    {
      "epoch": 1.7732211079274116,
      "grad_norm": 0.5047057867050171,
      "learning_rate": 6.564239377906098e-06,
      "loss": 0.4217,
      "mean_token_accuracy": 0.8572646856307984,
      "num_tokens": 995125270.0,
      "step": 29705
    },
    {
      "epoch": 1.7735195797516714,
      "grad_norm": 0.48366889357566833,
      "learning_rate": 6.5601738817606685e-06,
      "loss": 0.421,
      "mean_token_accuracy": 0.8602588653564454,
      "num_tokens": 995292950.0,
      "step": 29710
    },
    {
      "epoch": 1.7738180515759312,
      "grad_norm": 0.4254077970981598,
      "learning_rate": 6.556113485896395e-06,
      "loss": 0.3889,
      "mean_token_accuracy": 0.8687104821205139,
      "num_tokens": 995460630.0,
      "step": 29715
    },
    {
      "epoch": 1.774116523400191,
      "grad_norm": 0.48525968194007874,
      "learning_rate": 6.55205819130226e-06,
      "loss": 0.4345,
      "mean_token_accuracy": 0.8535667538642884,
      "num_tokens": 995628310.0,
      "step": 29720
    },
    {
      "epoch": 1.7744149952244508,
      "grad_norm": 0.5002026557922363,
      "learning_rate": 6.548007998966012e-06,
      "loss": 0.4077,
      "mean_token_accuracy": 0.8615650534629822,
      "num_tokens": 995795990.0,
      "step": 29725
    },
    {
      "epoch": 1.7747134670487106,
      "grad_norm": 0.4627630114555359,
      "learning_rate": 6.5439629098741435e-06,
      "loss": 0.4213,
      "mean_token_accuracy": 0.8598294258117676,
      "num_tokens": 995963670.0,
      "step": 29730
    },
    {
      "epoch": 1.7750119388729704,
      "grad_norm": 0.495576947927475,
      "learning_rate": 6.539922925011911e-06,
      "loss": 0.4279,
      "mean_token_accuracy": 0.856340229511261,
      "num_tokens": 996131350.0,
      "step": 29735
    },
    {
      "epoch": 1.7753104106972302,
      "grad_norm": 0.5531829595565796,
      "learning_rate": 6.535888045363328e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.8580281496047973,
      "num_tokens": 996299030.0,
      "step": 29740
    },
    {
      "epoch": 1.77560888252149,
      "grad_norm": 0.4722798466682434,
      "learning_rate": 6.531858271911165e-06,
      "loss": 0.4172,
      "mean_token_accuracy": 0.8583681344985962,
      "num_tokens": 996466710.0,
      "step": 29745
    },
    {
      "epoch": 1.7759073543457498,
      "grad_norm": 0.410942405462265,
      "learning_rate": 6.527833605636953e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8734045147895813,
      "num_tokens": 996634390.0,
      "step": 29750
    },
    {
      "epoch": 1.7762058261700095,
      "grad_norm": 0.46298593282699585,
      "learning_rate": 6.5238140475209625e-06,
      "loss": 0.4172,
      "mean_token_accuracy": 0.8595013737678527,
      "num_tokens": 996802070.0,
      "step": 29755
    },
    {
      "epoch": 1.7765042979942693,
      "grad_norm": 0.5453138947486877,
      "learning_rate": 6.519799598542235e-06,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8545150876045227,
      "num_tokens": 996969750.0,
      "step": 29760
    },
    {
      "epoch": 1.7768027698185291,
      "grad_norm": 0.5334903597831726,
      "learning_rate": 6.515790259678572e-06,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8477156281471252,
      "num_tokens": 997137430.0,
      "step": 29765
    },
    {
      "epoch": 1.777101241642789,
      "grad_norm": 0.5010488629341125,
      "learning_rate": 6.511786031906511e-06,
      "loss": 0.3825,
      "mean_token_accuracy": 0.8697840809822083,
      "num_tokens": 997305110.0,
      "step": 29770
    },
    {
      "epoch": 1.7773997134670487,
      "grad_norm": 0.5318519473075867,
      "learning_rate": 6.507786916201359e-06,
      "loss": 0.3842,
      "mean_token_accuracy": 0.8707205057144165,
      "num_tokens": 997472790.0,
      "step": 29775
    },
    {
      "epoch": 1.7776981852913085,
      "grad_norm": 0.5182591676712036,
      "learning_rate": 6.503792913537177e-06,
      "loss": 0.4347,
      "mean_token_accuracy": 0.8557556986808776,
      "num_tokens": 997640470.0,
      "step": 29780
    },
    {
      "epoch": 1.7779966571155683,
      "grad_norm": 0.49705302715301514,
      "learning_rate": 6.499804024886776e-06,
      "loss": 0.4526,
      "mean_token_accuracy": 0.8486997485160828,
      "num_tokens": 997808150.0,
      "step": 29785
    },
    {
      "epoch": 1.778295128939828,
      "grad_norm": 0.5338529944419861,
      "learning_rate": 6.495820251221728e-06,
      "loss": 0.4119,
      "mean_token_accuracy": 0.8605806112289429,
      "num_tokens": 997969428.0,
      "step": 29790
    },
    {
      "epoch": 1.778593600764088,
      "grad_norm": 0.5711192488670349,
      "learning_rate": 6.491841593512346e-06,
      "loss": 0.4334,
      "mean_token_accuracy": 0.8553143262863159,
      "num_tokens": 998137108.0,
      "step": 29795
    },
    {
      "epoch": 1.7788920725883477,
      "grad_norm": 0.5651690363883972,
      "learning_rate": 6.4878680527277105e-06,
      "loss": 0.4488,
      "mean_token_accuracy": 0.8490635871887207,
      "num_tokens": 998304788.0,
      "step": 29800
    },
    {
      "epoch": 1.7791905444126075,
      "grad_norm": 0.515377938747406,
      "learning_rate": 6.483899629835651e-06,
      "loss": 0.4077,
      "mean_token_accuracy": 0.86315758228302,
      "num_tokens": 998472468.0,
      "step": 29805
    },
    {
      "epoch": 1.7794890162368673,
      "grad_norm": 0.5820438265800476,
      "learning_rate": 6.479936325802752e-06,
      "loss": 0.4448,
      "mean_token_accuracy": 0.8510855317115784,
      "num_tokens": 998640148.0,
      "step": 29810
    },
    {
      "epoch": 1.779787488061127,
      "grad_norm": 0.5081514716148376,
      "learning_rate": 6.475978141594345e-06,
      "loss": 0.4274,
      "mean_token_accuracy": 0.8555827260017395,
      "num_tokens": 998807828.0,
      "step": 29815
    },
    {
      "epoch": 1.7800859598853869,
      "grad_norm": 0.5771610736846924,
      "learning_rate": 6.472025078174514e-06,
      "loss": 0.4841,
      "mean_token_accuracy": 0.8405165195465087,
      "num_tokens": 998975508.0,
      "step": 29820
    },
    {
      "epoch": 1.7803844317096467,
      "grad_norm": 0.5426960587501526,
      "learning_rate": 6.4680771365061175e-06,
      "loss": 0.4183,
      "mean_token_accuracy": 0.8575092315673828,
      "num_tokens": 999143188.0,
      "step": 29825
    },
    {
      "epoch": 1.7806829035339065,
      "grad_norm": 0.4830979108810425,
      "learning_rate": 6.4641343175507336e-06,
      "loss": 0.4017,
      "mean_token_accuracy": 0.8635810613632202,
      "num_tokens": 999310868.0,
      "step": 29830
    },
    {
      "epoch": 1.780981375358166,
      "grad_norm": 0.5529038310050964,
      "learning_rate": 6.460196622268717e-06,
      "loss": 0.4335,
      "mean_token_accuracy": 0.854461407661438,
      "num_tokens": 999478548.0,
      "step": 29835
    },
    {
      "epoch": 1.781279847182426,
      "grad_norm": 0.5055702924728394,
      "learning_rate": 6.4562640516191595e-06,
      "loss": 0.4004,
      "mean_token_accuracy": 0.8650065660476685,
      "num_tokens": 999646228.0,
      "step": 29840
    },
    {
      "epoch": 1.7815783190066856,
      "grad_norm": 0.530823290348053,
      "learning_rate": 6.452336606559915e-06,
      "loss": 0.4214,
      "mean_token_accuracy": 0.8566145777702332,
      "num_tokens": 999813908.0,
      "step": 29845
    },
    {
      "epoch": 1.7818767908309456,
      "grad_norm": 0.5236109495162964,
      "learning_rate": 6.448414288047592e-06,
      "loss": 0.4201,
      "mean_token_accuracy": 0.8586305618286133,
      "num_tokens": 999981588.0,
      "step": 29850
    },
    {
      "epoch": 1.7821752626552052,
      "grad_norm": 0.4977428913116455,
      "learning_rate": 6.444497097037532e-06,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8490516543388367,
      "num_tokens": 1000149268.0,
      "step": 29855
    },
    {
      "epoch": 1.7824737344794652,
      "grad_norm": 0.5108789205551147,
      "learning_rate": 6.4405850344838425e-06,
      "loss": 0.4315,
      "mean_token_accuracy": 0.8559883236885071,
      "num_tokens": 1000316948.0,
      "step": 29860
    },
    {
      "epoch": 1.7827722063037248,
      "grad_norm": 0.4991239905357361,
      "learning_rate": 6.436678101339383e-06,
      "loss": 0.3895,
      "mean_token_accuracy": 0.8672253370285035,
      "num_tokens": 1000484628.0,
      "step": 29865
    },
    {
      "epoch": 1.7830706781279848,
      "grad_norm": 0.5139074325561523,
      "learning_rate": 6.432776298555755e-06,
      "loss": 0.4312,
      "mean_token_accuracy": 0.8549803376197815,
      "num_tokens": 1000652308.0,
      "step": 29870
    },
    {
      "epoch": 1.7833691499522444,
      "grad_norm": 0.5655705332756042,
      "learning_rate": 6.428879627083317e-06,
      "loss": 0.4463,
      "mean_token_accuracy": 0.8506799459457397,
      "num_tokens": 1000819988.0,
      "step": 29875
    },
    {
      "epoch": 1.7836676217765044,
      "grad_norm": 0.507956862449646,
      "learning_rate": 6.424988087871175e-06,
      "loss": 0.4131,
      "mean_token_accuracy": 0.8603363990783691,
      "num_tokens": 1000987668.0,
      "step": 29880
    },
    {
      "epoch": 1.783966093600764,
      "grad_norm": 0.5413893461227417,
      "learning_rate": 6.421101681867181e-06,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8563103914260864,
      "num_tokens": 1001155348.0,
      "step": 29885
    },
    {
      "epoch": 1.784264565425024,
      "grad_norm": 0.49874091148376465,
      "learning_rate": 6.417220410017946e-06,
      "loss": 0.4314,
      "mean_token_accuracy": 0.855851125717163,
      "num_tokens": 1001323028.0,
      "step": 29890
    },
    {
      "epoch": 1.7845630372492836,
      "grad_norm": 0.6097025871276855,
      "learning_rate": 6.413344273268823e-06,
      "loss": 0.4512,
      "mean_token_accuracy": 0.8496063470840454,
      "num_tokens": 1001490708.0,
      "step": 29895
    },
    {
      "epoch": 1.7848615090735436,
      "grad_norm": 0.5162869691848755,
      "learning_rate": 6.409473272563918e-06,
      "loss": 0.4323,
      "mean_token_accuracy": 0.8546641945838929,
      "num_tokens": 1001658388.0,
      "step": 29900
    },
    {
      "epoch": 1.7851599808978031,
      "grad_norm": 0.5996779799461365,
      "learning_rate": 6.405607408846083e-06,
      "loss": 0.4293,
      "mean_token_accuracy": 0.856942617893219,
      "num_tokens": 1001826068.0,
      "step": 29905
    },
    {
      "epoch": 1.7854584527220632,
      "grad_norm": 0.5370012521743774,
      "learning_rate": 6.401746683056927e-06,
      "loss": 0.4577,
      "mean_token_accuracy": 0.8476500153541565,
      "num_tokens": 1001993748.0,
      "step": 29910
    },
    {
      "epoch": 1.7857569245463227,
      "grad_norm": 0.5540712475776672,
      "learning_rate": 6.397891096136789e-06,
      "loss": 0.4736,
      "mean_token_accuracy": 0.8423714637756348,
      "num_tokens": 1002161428.0,
      "step": 29915
    },
    {
      "epoch": 1.7860553963705827,
      "grad_norm": 0.4775422513484955,
      "learning_rate": 6.394040649024777e-06,
      "loss": 0.4254,
      "mean_token_accuracy": 0.8566086292266846,
      "num_tokens": 1002329108.0,
      "step": 29920
    },
    {
      "epoch": 1.7863538681948423,
      "grad_norm": 0.45298928022384644,
      "learning_rate": 6.390195342658732e-06,
      "loss": 0.4006,
      "mean_token_accuracy": 0.8650840878486633,
      "num_tokens": 1002496788.0,
      "step": 29925
    },
    {
      "epoch": 1.7866523400191023,
      "grad_norm": 0.457699716091156,
      "learning_rate": 6.386355177975259e-06,
      "loss": 0.4396,
      "mean_token_accuracy": 0.8517356634140014,
      "num_tokens": 1002664468.0,
      "step": 29930
    },
    {
      "epoch": 1.786950811843362,
      "grad_norm": 0.49803197383880615,
      "learning_rate": 6.382520155909696e-06,
      "loss": 0.4217,
      "mean_token_accuracy": 0.8583979368209839,
      "num_tokens": 1002832148.0,
      "step": 29935
    },
    {
      "epoch": 1.787249283667622,
      "grad_norm": 0.5499160885810852,
      "learning_rate": 6.378690277396133e-06,
      "loss": 0.4345,
      "mean_token_accuracy": 0.8553560853004456,
      "num_tokens": 1002999828.0,
      "step": 29940
    },
    {
      "epoch": 1.7875477554918815,
      "grad_norm": 0.5389583706855774,
      "learning_rate": 6.374865543367406e-06,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8422223448753356,
      "num_tokens": 1003167508.0,
      "step": 29945
    },
    {
      "epoch": 1.7878462273161415,
      "grad_norm": 0.5455766916275024,
      "learning_rate": 6.371045954755098e-06,
      "loss": 0.4444,
      "mean_token_accuracy": 0.8525766491889953,
      "num_tokens": 1003335188.0,
      "step": 29950
    },
    {
      "epoch": 1.788144699140401,
      "grad_norm": 0.4734271764755249,
      "learning_rate": 6.3672315124895475e-06,
      "loss": 0.4064,
      "mean_token_accuracy": 0.8629249691963196,
      "num_tokens": 1003502868.0,
      "step": 29955
    },
    {
      "epoch": 1.788443170964661,
      "grad_norm": 0.5003581643104553,
      "learning_rate": 6.36342221749983e-06,
      "loss": 0.4504,
      "mean_token_accuracy": 0.8496063590049744,
      "num_tokens": 1003670548.0,
      "step": 29960
    },
    {
      "epoch": 1.7887416427889207,
      "grad_norm": 0.53315269947052,
      "learning_rate": 6.359618070713763e-06,
      "loss": 0.4222,
      "mean_token_accuracy": 0.8570141911506652,
      "num_tokens": 1003838228.0,
      "step": 29965
    },
    {
      "epoch": 1.7890401146131805,
      "grad_norm": 0.5205037593841553,
      "learning_rate": 6.3558190730579254e-06,
      "loss": 0.4318,
      "mean_token_accuracy": 0.8561970710754394,
      "num_tokens": 1004005908.0,
      "step": 29970
    },
    {
      "epoch": 1.7893385864374403,
      "grad_norm": 0.49853748083114624,
      "learning_rate": 6.352025225457628e-06,
      "loss": 0.4156,
      "mean_token_accuracy": 0.859310507774353,
      "num_tokens": 1004173588.0,
      "step": 29975
    },
    {
      "epoch": 1.7896370582617,
      "grad_norm": 0.4901517629623413,
      "learning_rate": 6.348236528836933e-06,
      "loss": 0.426,
      "mean_token_accuracy": 0.8556977391242981,
      "num_tokens": 1004336302.0,
      "step": 29980
    },
    {
      "epoch": 1.7899355300859598,
      "grad_norm": 0.4826200604438782,
      "learning_rate": 6.3444529841186495e-06,
      "loss": 0.431,
      "mean_token_accuracy": 0.8551771402359009,
      "num_tokens": 1004503982.0,
      "step": 29985
    },
    {
      "epoch": 1.7902340019102196,
      "grad_norm": 0.4971453845500946,
      "learning_rate": 6.340674592224327e-06,
      "loss": 0.4506,
      "mean_token_accuracy": 0.8485446691513061,
      "num_tokens": 1004671662.0,
      "step": 29990
    },
    {
      "epoch": 1.7905324737344794,
      "grad_norm": 0.4662124216556549,
      "learning_rate": 6.3369013540742694e-06,
      "loss": 0.3942,
      "mean_token_accuracy": 0.868233323097229,
      "num_tokens": 1004839342.0,
      "step": 29995
    },
    {
      "epoch": 1.7908309455587392,
      "grad_norm": 0.5221835970878601,
      "learning_rate": 6.333133270587506e-06,
      "loss": 0.4243,
      "mean_token_accuracy": 0.8579923629760742,
      "num_tokens": 1005007022.0,
      "step": 30000
    },
    {
      "epoch": 1.791129417382999,
      "grad_norm": 0.5196115970611572,
      "learning_rate": 6.329370342681833e-06,
      "loss": 0.4223,
      "mean_token_accuracy": 0.8587856411933898,
      "num_tokens": 1005174702.0,
      "step": 30005
    },
    {
      "epoch": 1.7914278892072588,
      "grad_norm": 0.4869367480278015,
      "learning_rate": 6.3256125712737755e-06,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8585052967071534,
      "num_tokens": 1005342382.0,
      "step": 30010
    },
    {
      "epoch": 1.7917263610315186,
      "grad_norm": 0.4613809287548065,
      "learning_rate": 6.3218599572786144e-06,
      "loss": 0.4149,
      "mean_token_accuracy": 0.8595669746398926,
      "num_tokens": 1005510062.0,
      "step": 30015
    },
    {
      "epoch": 1.7920248328557784,
      "grad_norm": 0.54756098985672,
      "learning_rate": 6.318112501610367e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8521114230155945,
      "num_tokens": 1005677742.0,
      "step": 30020
    },
    {
      "epoch": 1.7923233046800382,
      "grad_norm": 0.5355969071388245,
      "learning_rate": 6.314370205181783e-06,
      "loss": 0.4551,
      "mean_token_accuracy": 0.8488011479377746,
      "num_tokens": 1005845422.0,
      "step": 30025
    },
    {
      "epoch": 1.792621776504298,
      "grad_norm": 0.45079246163368225,
      "learning_rate": 6.310633068904386e-06,
      "loss": 0.4153,
      "mean_token_accuracy": 0.8608195066452027,
      "num_tokens": 1006013102.0,
      "step": 30030
    },
    {
      "epoch": 1.7929202483285578,
      "grad_norm": 0.5127429366111755,
      "learning_rate": 6.3069010936884135e-06,
      "loss": 0.4152,
      "mean_token_accuracy": 0.8595431089401245,
      "num_tokens": 1006180782.0,
      "step": 30035
    },
    {
      "epoch": 1.7932187201528176,
      "grad_norm": 0.5833861231803894,
      "learning_rate": 6.303174280442858e-06,
      "loss": 0.452,
      "mean_token_accuracy": 0.8508409857749939,
      "num_tokens": 1006348462.0,
      "step": 30040
    },
    {
      "epoch": 1.7935171919770774,
      "grad_norm": 0.502337634563446,
      "learning_rate": 6.299452630075456e-06,
      "loss": 0.4288,
      "mean_token_accuracy": 0.8551592588424682,
      "num_tokens": 1006516142.0,
      "step": 30045
    },
    {
      "epoch": 1.7938156638013372,
      "grad_norm": 0.5145968794822693,
      "learning_rate": 6.295736143492683e-06,
      "loss": 0.4381,
      "mean_token_accuracy": 0.8540617823600769,
      "num_tokens": 1006683822.0,
      "step": 30050
    },
    {
      "epoch": 1.794114135625597,
      "grad_norm": 0.48628824949264526,
      "learning_rate": 6.292024821599762e-06,
      "loss": 0.4722,
      "mean_token_accuracy": 0.8416378259658813,
      "num_tokens": 1006851502.0,
      "step": 30055
    },
    {
      "epoch": 1.7944126074498568,
      "grad_norm": 0.5729851126670837,
      "learning_rate": 6.288318665300645e-06,
      "loss": 0.4725,
      "mean_token_accuracy": 0.843922221660614,
      "num_tokens": 1007019182.0,
      "step": 30060
    },
    {
      "epoch": 1.7947110792741165,
      "grad_norm": 0.46259039640426636,
      "learning_rate": 6.284617675498041e-06,
      "loss": 0.3711,
      "mean_token_accuracy": 0.8740904331207275,
      "num_tokens": 1007186862.0,
      "step": 30065
    },
    {
      "epoch": 1.7950095510983763,
      "grad_norm": 0.4978337585926056,
      "learning_rate": 6.280921853093394e-06,
      "loss": 0.3941,
      "mean_token_accuracy": 0.8675533771514893,
      "num_tokens": 1007354542.0,
      "step": 30070
    },
    {
      "epoch": 1.7953080229226361,
      "grad_norm": 0.5172361135482788,
      "learning_rate": 6.277231198986885e-06,
      "loss": 0.4487,
      "mean_token_accuracy": 0.8510974526405335,
      "num_tokens": 1007522222.0,
      "step": 30075
    },
    {
      "epoch": 1.795606494746896,
      "grad_norm": 0.4804052412509918,
      "learning_rate": 6.273545714077454e-06,
      "loss": 0.424,
      "mean_token_accuracy": 0.8576524019241333,
      "num_tokens": 1007689902.0,
      "step": 30080
    },
    {
      "epoch": 1.7959049665711557,
      "grad_norm": 0.4806613624095917,
      "learning_rate": 6.269865399262753e-06,
      "loss": 0.3956,
      "mean_token_accuracy": 0.8665453910827636,
      "num_tokens": 1007857582.0,
      "step": 30085
    },
    {
      "epoch": 1.7962034383954155,
      "grad_norm": 0.5243979692459106,
      "learning_rate": 6.266190255439194e-06,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8571633100509644,
      "num_tokens": 1008025262.0,
      "step": 30090
    },
    {
      "epoch": 1.7965019102196753,
      "grad_norm": 0.5323292016983032,
      "learning_rate": 6.262520283501938e-06,
      "loss": 0.4229,
      "mean_token_accuracy": 0.8592687606811523,
      "num_tokens": 1008192942.0,
      "step": 30095
    },
    {
      "epoch": 1.796800382043935,
      "grad_norm": 0.5139323472976685,
      "learning_rate": 6.258855484344862e-06,
      "loss": 0.4318,
      "mean_token_accuracy": 0.8544256210327148,
      "num_tokens": 1008360622.0,
      "step": 30100
    },
    {
      "epoch": 1.797098853868195,
      "grad_norm": 0.4848935008049011,
      "learning_rate": 6.255195858860602e-06,
      "loss": 0.397,
      "mean_token_accuracy": 0.865752112865448,
      "num_tokens": 1008528302.0,
      "step": 30105
    },
    {
      "epoch": 1.7973973256924545,
      "grad_norm": 0.5255609750747681,
      "learning_rate": 6.251541407940517e-06,
      "loss": 0.4215,
      "mean_token_accuracy": 0.8580639481544494,
      "num_tokens": 1008695982.0,
      "step": 30110
    },
    {
      "epoch": 1.7976957975167145,
      "grad_norm": 0.49564024806022644,
      "learning_rate": 6.247892132474729e-06,
      "loss": 0.4461,
      "mean_token_accuracy": 0.8502326130867004,
      "num_tokens": 1008863662.0,
      "step": 30115
    },
    {
      "epoch": 1.797994269340974,
      "grad_norm": 0.5161488652229309,
      "learning_rate": 6.244248033352084e-06,
      "loss": 0.4301,
      "mean_token_accuracy": 0.8570976972579956,
      "num_tokens": 1009031342.0,
      "step": 30120
    },
    {
      "epoch": 1.798292741165234,
      "grad_norm": 0.4784206748008728,
      "learning_rate": 6.240609111460163e-06,
      "loss": 0.4134,
      "mean_token_accuracy": 0.858821427822113,
      "num_tokens": 1009199022.0,
      "step": 30125
    },
    {
      "epoch": 1.7985912129894936,
      "grad_norm": 0.5764876008033752,
      "learning_rate": 6.236975367685295e-06,
      "loss": 0.3997,
      "mean_token_accuracy": 0.8661755919456482,
      "num_tokens": 1009366702.0,
      "step": 30130
    },
    {
      "epoch": 1.7988896848137537,
      "grad_norm": 0.5112322568893433,
      "learning_rate": 6.23334680291255e-06,
      "loss": 0.4148,
      "mean_token_accuracy": 0.8599308013916016,
      "num_tokens": 1009534382.0,
      "step": 30135
    },
    {
      "epoch": 1.7991881566380132,
      "grad_norm": 0.47579818964004517,
      "learning_rate": 6.229723418025729e-06,
      "loss": 0.3801,
      "mean_token_accuracy": 0.8712871193885803,
      "num_tokens": 1009702062.0,
      "step": 30140
    },
    {
      "epoch": 1.7994866284622733,
      "grad_norm": 0.4868124723434448,
      "learning_rate": 6.226105213907373e-06,
      "loss": 0.4179,
      "mean_token_accuracy": 0.8603721737861634,
      "num_tokens": 1009869742.0,
      "step": 30145
    },
    {
      "epoch": 1.7997851002865328,
      "grad_norm": 0.5073914527893066,
      "learning_rate": 6.222492191438758e-06,
      "loss": 0.4627,
      "mean_token_accuracy": 0.8444172739982605,
      "num_tokens": 1010037422.0,
      "step": 30150
    },
    {
      "epoch": 1.8000835721107928,
      "grad_norm": 0.5191833972930908,
      "learning_rate": 6.218884351499913e-06,
      "loss": 0.4261,
      "mean_token_accuracy": 0.8585053086280823,
      "num_tokens": 1010205102.0,
      "step": 30155
    },
    {
      "epoch": 1.8003820439350524,
      "grad_norm": 0.4738912880420685,
      "learning_rate": 6.215281694969587e-06,
      "loss": 0.3829,
      "mean_token_accuracy": 0.8716271042823791,
      "num_tokens": 1010372782.0,
      "step": 30160
    },
    {
      "epoch": 1.8006805157593124,
      "grad_norm": 0.5197004079818726,
      "learning_rate": 6.211684222725278e-06,
      "loss": 0.401,
      "mean_token_accuracy": 0.8648276329040527,
      "num_tokens": 1010540462.0,
      "step": 30165
    },
    {
      "epoch": 1.800978987583572,
      "grad_norm": 0.5202313661575317,
      "learning_rate": 6.20809193564321e-06,
      "loss": 0.4036,
      "mean_token_accuracy": 0.8631754636764526,
      "num_tokens": 1010708142.0,
      "step": 30170
    },
    {
      "epoch": 1.801277459407832,
      "grad_norm": 0.5416523814201355,
      "learning_rate": 6.204504834598354e-06,
      "loss": 0.462,
      "mean_token_accuracy": 0.8463437795639038,
      "num_tokens": 1010875822.0,
      "step": 30175
    },
    {
      "epoch": 1.8015759312320916,
      "grad_norm": 0.4549430012702942,
      "learning_rate": 6.200922920464421e-06,
      "loss": 0.4142,
      "mean_token_accuracy": 0.8605928778648376,
      "num_tokens": 1011043502.0,
      "step": 30180
    },
    {
      "epoch": 1.8018744030563516,
      "grad_norm": 0.5418521165847778,
      "learning_rate": 6.197346194113842e-06,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8541631817817688,
      "num_tokens": 1011211182.0,
      "step": 30185
    },
    {
      "epoch": 1.8021728748806112,
      "grad_norm": 0.5701766610145569,
      "learning_rate": 6.1937746564178e-06,
      "loss": 0.4103,
      "mean_token_accuracy": 0.8625790357589722,
      "num_tokens": 1011378862.0,
      "step": 30190
    },
    {
      "epoch": 1.8024713467048712,
      "grad_norm": 0.52357017993927,
      "learning_rate": 6.190208308246208e-06,
      "loss": 0.4213,
      "mean_token_accuracy": 0.8581474423408508,
      "num_tokens": 1011546542.0,
      "step": 30195
    },
    {
      "epoch": 1.8027698185291308,
      "grad_norm": 0.5457358360290527,
      "learning_rate": 6.186647150467716e-06,
      "loss": 0.4454,
      "mean_token_accuracy": 0.8509125471115112,
      "num_tokens": 1011714222.0,
      "step": 30200
    },
    {
      "epoch": 1.8030682903533908,
      "grad_norm": 0.539545476436615,
      "learning_rate": 6.183091183949715e-06,
      "loss": 0.4601,
      "mean_token_accuracy": 0.8465406179428101,
      "num_tokens": 1011881902.0,
      "step": 30205
    },
    {
      "epoch": 1.8033667621776504,
      "grad_norm": 0.49626877903938293,
      "learning_rate": 6.179540409558316e-06,
      "loss": 0.4137,
      "mean_token_accuracy": 0.8608970522880555,
      "num_tokens": 1012049582.0,
      "step": 30210
    },
    {
      "epoch": 1.8036652340019104,
      "grad_norm": 0.49910977482795715,
      "learning_rate": 6.175994828158381e-06,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8575629353523254,
      "num_tokens": 1012217262.0,
      "step": 30215
    },
    {
      "epoch": 1.80396370582617,
      "grad_norm": 0.5258999466896057,
      "learning_rate": 6.172454440613502e-06,
      "loss": 0.4367,
      "mean_token_accuracy": 0.8520577430725098,
      "num_tokens": 1012384942.0,
      "step": 30220
    },
    {
      "epoch": 1.80426217765043,
      "grad_norm": 0.4913251996040344,
      "learning_rate": 6.168919247786005e-06,
      "loss": 0.4112,
      "mean_token_accuracy": 0.8620839834213256,
      "num_tokens": 1012552622.0,
      "step": 30225
    },
    {
      "epoch": 1.8045606494746895,
      "grad_norm": 0.4836638569831848,
      "learning_rate": 6.165389250536957e-06,
      "loss": 0.4311,
      "mean_token_accuracy": 0.8542824864387513,
      "num_tokens": 1012720302.0,
      "step": 30230
    },
    {
      "epoch": 1.8048591212989495,
      "grad_norm": 0.5360139012336731,
      "learning_rate": 6.161864449726139e-06,
      "loss": 0.4506,
      "mean_token_accuracy": 0.8507157325744629,
      "num_tokens": 1012887982.0,
      "step": 30235
    },
    {
      "epoch": 1.8051575931232091,
      "grad_norm": 0.5072572231292725,
      "learning_rate": 6.158344846212099e-06,
      "loss": 0.4152,
      "mean_token_accuracy": 0.8602051615715027,
      "num_tokens": 1013055662.0,
      "step": 30240
    },
    {
      "epoch": 1.805456064947469,
      "grad_norm": 0.5142276287078857,
      "learning_rate": 6.154830440852092e-06,
      "loss": 0.3919,
      "mean_token_accuracy": 0.8670165777206421,
      "num_tokens": 1013223342.0,
      "step": 30245
    },
    {
      "epoch": 1.8057545367717287,
      "grad_norm": 0.5045875906944275,
      "learning_rate": 6.151321234502116e-06,
      "loss": 0.418,
      "mean_token_accuracy": 0.8589705348014831,
      "num_tokens": 1013391022.0,
      "step": 30250
    },
    {
      "epoch": 1.8060530085959885,
      "grad_norm": 0.5202556848526001,
      "learning_rate": 6.147817228016904e-06,
      "loss": 0.4196,
      "mean_token_accuracy": 0.8584038972854614,
      "num_tokens": 1013558702.0,
      "step": 30255
    },
    {
      "epoch": 1.8063514804202483,
      "grad_norm": 0.500163197517395,
      "learning_rate": 6.144318422249924e-06,
      "loss": 0.4447,
      "mean_token_accuracy": 0.852105462551117,
      "num_tokens": 1013726382.0,
      "step": 30260
    },
    {
      "epoch": 1.806649952244508,
      "grad_norm": 0.5260593891143799,
      "learning_rate": 6.140824818053378e-06,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8506381869316101,
      "num_tokens": 1013894062.0,
      "step": 30265
    },
    {
      "epoch": 1.8069484240687679,
      "grad_norm": 0.4753541350364685,
      "learning_rate": 6.137336416278186e-06,
      "loss": 0.4088,
      "mean_token_accuracy": 0.8640641808509827,
      "num_tokens": 1014061742.0,
      "step": 30270
    },
    {
      "epoch": 1.8072468958930277,
      "grad_norm": 0.47642311453819275,
      "learning_rate": 6.133853217774019e-06,
      "loss": 0.3893,
      "mean_token_accuracy": 0.8676488041877747,
      "num_tokens": 1014229422.0,
      "step": 30275
    },
    {
      "epoch": 1.8075453677172875,
      "grad_norm": 0.523440420627594,
      "learning_rate": 6.130375223389276e-06,
      "loss": 0.3957,
      "mean_token_accuracy": 0.86489919424057,
      "num_tokens": 1014397102.0,
      "step": 30280
    },
    {
      "epoch": 1.8078438395415473,
      "grad_norm": 0.5024499893188477,
      "learning_rate": 6.126902433971086e-06,
      "loss": 0.4415,
      "mean_token_accuracy": 0.8528868079185485,
      "num_tokens": 1014564782.0,
      "step": 30285
    },
    {
      "epoch": 1.808142311365807,
      "grad_norm": 0.4526674747467041,
      "learning_rate": 6.123434850365314e-06,
      "loss": 0.3982,
      "mean_token_accuracy": 0.8654956459999085,
      "num_tokens": 1014732462.0,
      "step": 30290
    },
    {
      "epoch": 1.8084407831900668,
      "grad_norm": 0.510155200958252,
      "learning_rate": 6.119972473416544e-06,
      "loss": 0.4131,
      "mean_token_accuracy": 0.8604735612869263,
      "num_tokens": 1014900142.0,
      "step": 30295
    },
    {
      "epoch": 1.8087392550143266,
      "grad_norm": 0.47911566495895386,
      "learning_rate": 6.116515303968114e-06,
      "loss": 0.4057,
      "mean_token_accuracy": 0.8639985799789429,
      "num_tokens": 1015067822.0,
      "step": 30300
    },
    {
      "epoch": 1.8090377268385864,
      "grad_norm": 0.5345131158828735,
      "learning_rate": 6.1130633428620705e-06,
      "loss": 0.4371,
      "mean_token_accuracy": 0.8538768887519836,
      "num_tokens": 1015235502.0,
      "step": 30305
    },
    {
      "epoch": 1.8093361986628462,
      "grad_norm": 0.5456638336181641,
      "learning_rate": 6.109616590939211e-06,
      "loss": 0.4267,
      "mean_token_accuracy": 0.8576643228530884,
      "num_tokens": 1015403182.0,
      "step": 30310
    },
    {
      "epoch": 1.809634670487106,
      "grad_norm": 0.5286844968795776,
      "learning_rate": 6.1061750490390504e-06,
      "loss": 0.4282,
      "mean_token_accuracy": 0.8555171132087708,
      "num_tokens": 1015570862.0,
      "step": 30315
    },
    {
      "epoch": 1.8099331423113658,
      "grad_norm": 0.5617997050285339,
      "learning_rate": 6.102738717999842e-06,
      "loss": 0.4752,
      "mean_token_accuracy": 0.8416020512580872,
      "num_tokens": 1015738542.0,
      "step": 30320
    },
    {
      "epoch": 1.8102316141356256,
      "grad_norm": 0.4980163872241974,
      "learning_rate": 6.099307598658568e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8586007237434388,
      "num_tokens": 1015906222.0,
      "step": 30325
    },
    {
      "epoch": 1.8105300859598854,
      "grad_norm": 0.4981996715068817,
      "learning_rate": 6.095881691850939e-06,
      "loss": 0.427,
      "mean_token_accuracy": 0.8558928847312928,
      "num_tokens": 1016073902.0,
      "step": 30330
    },
    {
      "epoch": 1.8108285577841452,
      "grad_norm": 0.5117992758750916,
      "learning_rate": 6.092460998411397e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.8573660969734191,
      "num_tokens": 1016241582.0,
      "step": 30335
    },
    {
      "epoch": 1.811127029608405,
      "grad_norm": 0.5172623991966248,
      "learning_rate": 6.089045519173118e-06,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8579685091972351,
      "num_tokens": 1016409262.0,
      "step": 30340
    },
    {
      "epoch": 1.8114255014326648,
      "grad_norm": 0.5884139537811279,
      "learning_rate": 6.085635254968003e-06,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8454789519309998,
      "num_tokens": 1016576942.0,
      "step": 30345
    },
    {
      "epoch": 1.8117239732569246,
      "grad_norm": 0.532335638999939,
      "learning_rate": 6.082230206626689e-06,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8511392116546631,
      "num_tokens": 1016744622.0,
      "step": 30350
    },
    {
      "epoch": 1.8120224450811844,
      "grad_norm": 0.48214107751846313,
      "learning_rate": 6.078830374978531e-06,
      "loss": 0.4089,
      "mean_token_accuracy": 0.8626148223876953,
      "num_tokens": 1016912302.0,
      "step": 30355
    },
    {
      "epoch": 1.8123209169054442,
      "grad_norm": 0.5337048768997192,
      "learning_rate": 6.075435760851626e-06,
      "loss": 0.4478,
      "mean_token_accuracy": 0.8519205570220947,
      "num_tokens": 1017079982.0,
      "step": 30360
    },
    {
      "epoch": 1.812619388729704,
      "grad_norm": 0.5389066338539124,
      "learning_rate": 6.072046365072793e-06,
      "loss": 0.4507,
      "mean_token_accuracy": 0.8505367994308471,
      "num_tokens": 1017247662.0,
      "step": 30365
    },
    {
      "epoch": 1.8129178605539638,
      "grad_norm": 0.515815258026123,
      "learning_rate": 6.068662188467582e-06,
      "loss": 0.4133,
      "mean_token_accuracy": 0.8625551581382751,
      "num_tokens": 1017415342.0,
      "step": 30370
    },
    {
      "epoch": 1.8132163323782235,
      "grad_norm": 0.4780847728252411,
      "learning_rate": 6.065283231860274e-06,
      "loss": 0.4383,
      "mean_token_accuracy": 0.8542824745178222,
      "num_tokens": 1017583022.0,
      "step": 30375
    },
    {
      "epoch": 1.8135148042024833,
      "grad_norm": 0.5041197538375854,
      "learning_rate": 6.061909496073872e-06,
      "loss": 0.4329,
      "mean_token_accuracy": 0.8557795524597168,
      "num_tokens": 1017750702.0,
      "step": 30380
    },
    {
      "epoch": 1.813813276026743,
      "grad_norm": 0.5082278847694397,
      "learning_rate": 6.05854098193012e-06,
      "loss": 0.4529,
      "mean_token_accuracy": 0.849021828174591,
      "num_tokens": 1017918382.0,
      "step": 30385
    },
    {
      "epoch": 1.814111747851003,
      "grad_norm": 0.5243868827819824,
      "learning_rate": 6.055177690249469e-06,
      "loss": 0.4374,
      "mean_token_accuracy": 0.8529166221618653,
      "num_tokens": 1018086062.0,
      "step": 30390
    },
    {
      "epoch": 1.8144102196752625,
      "grad_norm": 0.5364031195640564,
      "learning_rate": 6.051819621851121e-06,
      "loss": 0.3963,
      "mean_token_accuracy": 0.8664738178253174,
      "num_tokens": 1018253742.0,
      "step": 30395
    },
    {
      "epoch": 1.8147086914995225,
      "grad_norm": 0.46478471159935,
      "learning_rate": 6.0484667775529895e-06,
      "loss": 0.404,
      "mean_token_accuracy": 0.8640164613723755,
      "num_tokens": 1018421422.0,
      "step": 30400
    },
    {
      "epoch": 1.815007163323782,
      "grad_norm": 0.48922744393348694,
      "learning_rate": 6.045119158171727e-06,
      "loss": 0.3969,
      "mean_token_accuracy": 0.866628897190094,
      "num_tokens": 1018589102.0,
      "step": 30405
    },
    {
      "epoch": 1.815305635148042,
      "grad_norm": 0.583345353603363,
      "learning_rate": 6.041776764522707e-06,
      "loss": 0.457,
      "mean_token_accuracy": 0.8483120560646057,
      "num_tokens": 1018756782.0,
      "step": 30410
    },
    {
      "epoch": 1.8156041069723017,
      "grad_norm": 0.5307941436767578,
      "learning_rate": 6.038439597420024e-06,
      "loss": 0.4051,
      "mean_token_accuracy": 0.8624537825584412,
      "num_tokens": 1018924462.0,
      "step": 30415
    },
    {
      "epoch": 1.8159025787965617,
      "grad_norm": 0.5566131472587585,
      "learning_rate": 6.035107657676513e-06,
      "loss": 0.4351,
      "mean_token_accuracy": 0.8534355163574219,
      "num_tokens": 1019092142.0,
      "step": 30420
    },
    {
      "epoch": 1.8162010506208213,
      "grad_norm": 0.4648083448410034,
      "learning_rate": 6.0317809461037324e-06,
      "loss": 0.392,
      "mean_token_accuracy": 0.8670404314994812,
      "num_tokens": 1019259822.0,
      "step": 30425
    },
    {
      "epoch": 1.8164995224450813,
      "grad_norm": 0.4756131172180176,
      "learning_rate": 6.028459463511958e-06,
      "loss": 0.4008,
      "mean_token_accuracy": 0.8640701293945312,
      "num_tokens": 1019427502.0,
      "step": 30430
    },
    {
      "epoch": 1.8167979942693409,
      "grad_norm": 0.5145769715309143,
      "learning_rate": 6.0251432107102015e-06,
      "loss": 0.4107,
      "mean_token_accuracy": 0.8606525182723999,
      "num_tokens": 1019595182.0,
      "step": 30435
    },
    {
      "epoch": 1.8170964660936009,
      "grad_norm": 0.5484784841537476,
      "learning_rate": 6.021832188506193e-06,
      "loss": 0.4058,
      "mean_token_accuracy": 0.8620064496994019,
      "num_tokens": 1019762862.0,
      "step": 30440
    },
    {
      "epoch": 1.8173949379178604,
      "grad_norm": 0.5580423474311829,
      "learning_rate": 6.018526397706399e-06,
      "loss": 0.4609,
      "mean_token_accuracy": 0.8464332580566406,
      "num_tokens": 1019930542.0,
      "step": 30445
    },
    {
      "epoch": 1.8176934097421205,
      "grad_norm": 0.6324069499969482,
      "learning_rate": 6.015225839116006e-06,
      "loss": 0.4179,
      "mean_token_accuracy": 0.8596206545829773,
      "num_tokens": 1020098222.0,
      "step": 30450
    },
    {
      "epoch": 1.81799188156638,
      "grad_norm": 0.45643118023872375,
      "learning_rate": 6.0119305135389205e-06,
      "loss": 0.413,
      "mean_token_accuracy": 0.8611237049102783,
      "num_tokens": 1020265902.0,
      "step": 30455
    },
    {
      "epoch": 1.81829035339064,
      "grad_norm": 0.5177029967308044,
      "learning_rate": 6.008640421777781e-06,
      "loss": 0.403,
      "mean_token_accuracy": 0.8635035157203674,
      "num_tokens": 1020433582.0,
      "step": 30460
    },
    {
      "epoch": 1.8185888252148996,
      "grad_norm": 0.5298128128051758,
      "learning_rate": 6.005355564633954e-06,
      "loss": 0.4297,
      "mean_token_accuracy": 0.8553620457649231,
      "num_tokens": 1020601262.0,
      "step": 30465
    },
    {
      "epoch": 1.8188872970391596,
      "grad_norm": 0.5516754984855652,
      "learning_rate": 6.002075942907528e-06,
      "loss": 0.4449,
      "mean_token_accuracy": 0.8510676264762879,
      "num_tokens": 1020768942.0,
      "step": 30470
    },
    {
      "epoch": 1.8191857688634192,
      "grad_norm": 0.5264891386032104,
      "learning_rate": 5.998801557397308e-06,
      "loss": 0.4343,
      "mean_token_accuracy": 0.8535369157791137,
      "num_tokens": 1020936622.0,
      "step": 30475
    },
    {
      "epoch": 1.8194842406876792,
      "grad_norm": 0.5354034900665283,
      "learning_rate": 5.995532408900836e-06,
      "loss": 0.444,
      "mean_token_accuracy": 0.8525706768035889,
      "num_tokens": 1021104302.0,
      "step": 30480
    },
    {
      "epoch": 1.8197827125119388,
      "grad_norm": 0.5818821787834167,
      "learning_rate": 5.992268498214374e-06,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8516939043998718,
      "num_tokens": 1021271982.0,
      "step": 30485
    },
    {
      "epoch": 1.8200811843361988,
      "grad_norm": 0.5167210698127747,
      "learning_rate": 5.989009826132905e-06,
      "loss": 0.4558,
      "mean_token_accuracy": 0.8479064702987671,
      "num_tokens": 1021439662.0,
      "step": 30490
    },
    {
      "epoch": 1.8203796561604584,
      "grad_norm": 0.50235915184021,
      "learning_rate": 5.985756393450143e-06,
      "loss": 0.4145,
      "mean_token_accuracy": 0.8593641877174377,
      "num_tokens": 1021607342.0,
      "step": 30495
    },
    {
      "epoch": 1.8206781279847184,
      "grad_norm": 0.4829601049423218,
      "learning_rate": 5.982508200958512e-06,
      "loss": 0.4185,
      "mean_token_accuracy": 0.858057975769043,
      "num_tokens": 1021775022.0,
      "step": 30500
    },
    {
      "epoch": 1.820976599808978,
      "grad_norm": 0.5765755772590637,
      "learning_rate": 5.979265249449182e-06,
      "loss": 0.4443,
      "mean_token_accuracy": 0.85235595703125,
      "num_tokens": 1021942702.0,
      "step": 30505
    },
    {
      "epoch": 1.821275071633238,
      "grad_norm": 0.6675260663032532,
      "learning_rate": 5.976027539712028e-06,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8453059792518616,
      "num_tokens": 1022110382.0,
      "step": 30510
    },
    {
      "epoch": 1.8215735434574976,
      "grad_norm": 0.6030436754226685,
      "learning_rate": 5.972795072535652e-06,
      "loss": 0.4436,
      "mean_token_accuracy": 0.8529643177986145,
      "num_tokens": 1022278062.0,
      "step": 30515
    },
    {
      "epoch": 1.8218720152817574,
      "grad_norm": 0.5055321455001831,
      "learning_rate": 5.969567848707382e-06,
      "loss": 0.407,
      "mean_token_accuracy": 0.8628235697746277,
      "num_tokens": 1022445742.0,
      "step": 30520
    },
    {
      "epoch": 1.8221704871060171,
      "grad_norm": 0.5094475150108337,
      "learning_rate": 5.9663458690132715e-06,
      "loss": 0.4299,
      "mean_token_accuracy": 0.8551950216293335,
      "num_tokens": 1022613422.0,
      "step": 30525
    },
    {
      "epoch": 1.822468958930277,
      "grad_norm": 0.5414401292800903,
      "learning_rate": 5.963129134238087e-06,
      "loss": 0.4438,
      "mean_token_accuracy": 0.8517654657363891,
      "num_tokens": 1022781102.0,
      "step": 30530
    },
    {
      "epoch": 1.8227674307545367,
      "grad_norm": 0.5134124755859375,
      "learning_rate": 5.9599176451653305e-06,
      "loss": 0.4275,
      "mean_token_accuracy": 0.8573004961013794,
      "num_tokens": 1022948782.0,
      "step": 30535
    },
    {
      "epoch": 1.8230659025787965,
      "grad_norm": 0.5222249627113342,
      "learning_rate": 5.956711402577213e-06,
      "loss": 0.4075,
      "mean_token_accuracy": 0.8628056764602661,
      "num_tokens": 1023116462.0,
      "step": 30540
    },
    {
      "epoch": 1.8233643744030563,
      "grad_norm": 0.5483086705207825,
      "learning_rate": 5.953510407254677e-06,
      "loss": 0.4206,
      "mean_token_accuracy": 0.8589884281158447,
      "num_tokens": 1023284142.0,
      "step": 30545
    },
    {
      "epoch": 1.8236628462273161,
      "grad_norm": 0.5105926394462585,
      "learning_rate": 5.950314659977386e-06,
      "loss": 0.4368,
      "mean_token_accuracy": 0.855338191986084,
      "num_tokens": 1023451822.0,
      "step": 30550
    },
    {
      "epoch": 1.823961318051576,
      "grad_norm": 0.48213180899620056,
      "learning_rate": 5.947124161523719e-06,
      "loss": 0.4238,
      "mean_token_accuracy": 0.8586245775222778,
      "num_tokens": 1023619502.0,
      "step": 30555
    },
    {
      "epoch": 1.8242597898758357,
      "grad_norm": 0.48848965764045715,
      "learning_rate": 5.943938912670785e-06,
      "loss": 0.4308,
      "mean_token_accuracy": 0.8550578713417053,
      "num_tokens": 1023787182.0,
      "step": 30560
    },
    {
      "epoch": 1.8245582617000955,
      "grad_norm": 0.4824397563934326,
      "learning_rate": 5.940758914194404e-06,
      "loss": 0.431,
      "mean_token_accuracy": 0.8564058184623718,
      "num_tokens": 1023954862.0,
      "step": 30565
    },
    {
      "epoch": 1.8248567335243553,
      "grad_norm": 0.49501392245292664,
      "learning_rate": 5.937584166869134e-06,
      "loss": 0.4153,
      "mean_token_accuracy": 0.8611058115959167,
      "num_tokens": 1024122542.0,
      "step": 30570
    },
    {
      "epoch": 1.825155205348615,
      "grad_norm": 0.4652409255504608,
      "learning_rate": 5.934414671468232e-06,
      "loss": 0.4041,
      "mean_token_accuracy": 0.8645293951034546,
      "num_tokens": 1024290222.0,
      "step": 30575
    },
    {
      "epoch": 1.8254536771728749,
      "grad_norm": 0.48045143485069275,
      "learning_rate": 5.931250428763696e-06,
      "loss": 0.378,
      "mean_token_accuracy": 0.8716092228889465,
      "num_tokens": 1024457902.0,
      "step": 30580
    },
    {
      "epoch": 1.8257521489971347,
      "grad_norm": 0.5033023357391357,
      "learning_rate": 5.928091439526226e-06,
      "loss": 0.3789,
      "mean_token_accuracy": 0.8707622766494751,
      "num_tokens": 1024625582.0,
      "step": 30585
    },
    {
      "epoch": 1.8260506208213945,
      "grad_norm": 0.5048587322235107,
      "learning_rate": 5.924937704525259e-06,
      "loss": 0.416,
      "mean_token_accuracy": 0.8606584668159485,
      "num_tokens": 1024793262.0,
      "step": 30590
    },
    {
      "epoch": 1.8263490926456543,
      "grad_norm": 0.5610945820808411,
      "learning_rate": 5.921789224528949e-06,
      "loss": 0.4605,
      "mean_token_accuracy": 0.8468030452728271,
      "num_tokens": 1024960942.0,
      "step": 30595
    },
    {
      "epoch": 1.826647564469914,
      "grad_norm": 0.5296248197555542,
      "learning_rate": 5.918646000304156e-06,
      "loss": 0.4155,
      "mean_token_accuracy": 0.8622748494148255,
      "num_tokens": 1025128622.0,
      "step": 30600
    },
    {
      "epoch": 1.8269460362941738,
      "grad_norm": 0.5809566974639893,
      "learning_rate": 5.915508032616478e-06,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8490635871887207,
      "num_tokens": 1025296302.0,
      "step": 30605
    },
    {
      "epoch": 1.8272445081184336,
      "grad_norm": 0.5218061804771423,
      "learning_rate": 5.912375322230222e-06,
      "loss": 0.4565,
      "mean_token_accuracy": 0.848884642124176,
      "num_tokens": 1025463982.0,
      "step": 30610
    },
    {
      "epoch": 1.8275429799426934,
      "grad_norm": 0.49793973565101624,
      "learning_rate": 5.909247869908417e-06,
      "loss": 0.4328,
      "mean_token_accuracy": 0.854825246334076,
      "num_tokens": 1025631662.0,
      "step": 30615
    },
    {
      "epoch": 1.8278414517669532,
      "grad_norm": 0.4951666593551636,
      "learning_rate": 5.906125676412817e-06,
      "loss": 0.407,
      "mean_token_accuracy": 0.8635810613632202,
      "num_tokens": 1025799342.0,
      "step": 30620
    },
    {
      "epoch": 1.828139923591213,
      "grad_norm": 0.5134680271148682,
      "learning_rate": 5.903008742503879e-06,
      "loss": 0.4169,
      "mean_token_accuracy": 0.8579685091972351,
      "num_tokens": 1025967022.0,
      "step": 30625
    },
    {
      "epoch": 1.8284383954154728,
      "grad_norm": 0.5246886610984802,
      "learning_rate": 5.899897068940805e-06,
      "loss": 0.4037,
      "mean_token_accuracy": 0.8634021162986756,
      "num_tokens": 1026134702.0,
      "step": 30630
    },
    {
      "epoch": 1.8287368672397326,
      "grad_norm": 0.5466708540916443,
      "learning_rate": 5.8967906564814866e-06,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8419599175453186,
      "num_tokens": 1026302382.0,
      "step": 30635
    },
    {
      "epoch": 1.8290353390639924,
      "grad_norm": 0.4633078873157501,
      "learning_rate": 5.8936895058825545e-06,
      "loss": 0.4183,
      "mean_token_accuracy": 0.8600679874420166,
      "num_tokens": 1026470062.0,
      "step": 30640
    },
    {
      "epoch": 1.8293338108882522,
      "grad_norm": 0.4778256118297577,
      "learning_rate": 5.89059361789935e-06,
      "loss": 0.4117,
      "mean_token_accuracy": 0.8613741993904114,
      "num_tokens": 1026637742.0,
      "step": 30645
    },
    {
      "epoch": 1.829632282712512,
      "grad_norm": 0.5174081921577454,
      "learning_rate": 5.887502993285934e-06,
      "loss": 0.4258,
      "mean_token_accuracy": 0.8562209129333496,
      "num_tokens": 1026805422.0,
      "step": 30650
    },
    {
      "epoch": 1.8299307545367718,
      "grad_norm": 0.46859246492385864,
      "learning_rate": 5.884417632795086e-06,
      "loss": 0.4118,
      "mean_token_accuracy": 0.8600739598274231,
      "num_tokens": 1026973102.0,
      "step": 30655
    },
    {
      "epoch": 1.8302292263610314,
      "grad_norm": 0.45793935656547546,
      "learning_rate": 5.8813375371783e-06,
      "loss": 0.4128,
      "mean_token_accuracy": 0.861165452003479,
      "num_tokens": 1027140782.0,
      "step": 30660
    },
    {
      "epoch": 1.8305276981852914,
      "grad_norm": 0.5306112766265869,
      "learning_rate": 5.878262707185792e-06,
      "loss": 0.4176,
      "mean_token_accuracy": 0.8616724252700806,
      "num_tokens": 1027308462.0,
      "step": 30665
    },
    {
      "epoch": 1.830826170009551,
      "grad_norm": 0.5088845491409302,
      "learning_rate": 5.875193143566493e-06,
      "loss": 0.4141,
      "mean_token_accuracy": 0.8621555447578431,
      "num_tokens": 1027476142.0,
      "step": 30670
    },
    {
      "epoch": 1.831124641833811,
      "grad_norm": 0.4992761015892029,
      "learning_rate": 5.8721288470680525e-06,
      "loss": 0.4296,
      "mean_token_accuracy": 0.8566980838775635,
      "num_tokens": 1027643822.0,
      "step": 30675
    },
    {
      "epoch": 1.8314231136580705,
      "grad_norm": 0.5080541372299194,
      "learning_rate": 5.869069818436838e-06,
      "loss": 0.3732,
      "mean_token_accuracy": 0.8748896598815918,
      "num_tokens": 1027811502.0,
      "step": 30680
    },
    {
      "epoch": 1.8317215854823305,
      "grad_norm": 0.5191560983657837,
      "learning_rate": 5.866016058417927e-06,
      "loss": 0.4143,
      "mean_token_accuracy": 0.8610700130462646,
      "num_tokens": 1027979182.0,
      "step": 30685
    },
    {
      "epoch": 1.8320200573065901,
      "grad_norm": 0.495977520942688,
      "learning_rate": 5.862967567755124e-06,
      "loss": 0.417,
      "mean_token_accuracy": 0.8614875316619873,
      "num_tokens": 1028146862.0,
      "step": 30690
    },
    {
      "epoch": 1.8323185291308501,
      "grad_norm": 0.48497673869132996,
      "learning_rate": 5.859924347190942e-06,
      "loss": 0.389,
      "mean_token_accuracy": 0.8695345759391785,
      "num_tokens": 1028314302.0,
      "step": 30695
    },
    {
      "epoch": 1.8326170009551097,
      "grad_norm": 0.4743047058582306,
      "learning_rate": 5.856886397466617e-06,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8495467066764831,
      "num_tokens": 1028481982.0,
      "step": 30700
    },
    {
      "epoch": 1.8329154727793697,
      "grad_norm": 0.5589334964752197,
      "learning_rate": 5.853853719322099e-06,
      "loss": 0.4554,
      "mean_token_accuracy": 0.8488071084022522,
      "num_tokens": 1028649662.0,
      "step": 30705
    },
    {
      "epoch": 1.8332139446036293,
      "grad_norm": 0.5527417063713074,
      "learning_rate": 5.850826313496043e-06,
      "loss": 0.4206,
      "mean_token_accuracy": 0.8598830938339234,
      "num_tokens": 1028817342.0,
      "step": 30710
    },
    {
      "epoch": 1.8335124164278893,
      "grad_norm": 0.5253627300262451,
      "learning_rate": 5.8478041807258425e-06,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8509185314178467,
      "num_tokens": 1028985022.0,
      "step": 30715
    },
    {
      "epoch": 1.8338108882521489,
      "grad_norm": 0.5398898124694824,
      "learning_rate": 5.844787321747584e-06,
      "loss": 0.4285,
      "mean_token_accuracy": 0.855952525138855,
      "num_tokens": 1029152702.0,
      "step": 30720
    },
    {
      "epoch": 1.834109360076409,
      "grad_norm": 0.575414776802063,
      "learning_rate": 5.841775737296084e-06,
      "loss": 0.4511,
      "mean_token_accuracy": 0.8481927633285522,
      "num_tokens": 1029320382.0,
      "step": 30725
    },
    {
      "epoch": 1.8344078319006685,
      "grad_norm": 0.5164406895637512,
      "learning_rate": 5.838769428104867e-06,
      "loss": 0.4312,
      "mean_token_accuracy": 0.8546045660972595,
      "num_tokens": 1029488062.0,
      "step": 30730
    },
    {
      "epoch": 1.8347063037249285,
      "grad_norm": 0.5350559949874878,
      "learning_rate": 5.835768394906176e-06,
      "loss": 0.4204,
      "mean_token_accuracy": 0.8575271368026733,
      "num_tokens": 1029655742.0,
      "step": 30735
    },
    {
      "epoch": 1.835004775549188,
      "grad_norm": 0.46778568625450134,
      "learning_rate": 5.832772638430973e-06,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8587438821792602,
      "num_tokens": 1029823422.0,
      "step": 30740
    },
    {
      "epoch": 1.835303247373448,
      "grad_norm": 0.52099609375,
      "learning_rate": 5.829782159408919e-06,
      "loss": 0.4327,
      "mean_token_accuracy": 0.8545697569847107,
      "num_tokens": 1029990264.0,
      "step": 30745
    },
    {
      "epoch": 1.8356017191977076,
      "grad_norm": 0.49678269028663635,
      "learning_rate": 5.826796958568407e-06,
      "loss": 0.4147,
      "mean_token_accuracy": 0.8594536542892456,
      "num_tokens": 1030157944.0,
      "step": 30750
    },
    {
      "epoch": 1.8359001910219677,
      "grad_norm": 0.5183316469192505,
      "learning_rate": 5.823817036636541e-06,
      "loss": 0.4461,
      "mean_token_accuracy": 0.8510139584541321,
      "num_tokens": 1030325624.0,
      "step": 30755
    },
    {
      "epoch": 1.8361986628462272,
      "grad_norm": 0.48756346106529236,
      "learning_rate": 5.820842394339129e-06,
      "loss": 0.4095,
      "mean_token_accuracy": 0.8618014693260193,
      "num_tokens": 1030491801.0,
      "step": 30760
    },
    {
      "epoch": 1.8364971346704873,
      "grad_norm": 0.5023455619812012,
      "learning_rate": 5.817873032400707e-06,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8708338379859925,
      "num_tokens": 1030659481.0,
      "step": 30765
    },
    {
      "epoch": 1.8367956064947468,
      "grad_norm": 0.5409082174301147,
      "learning_rate": 5.814908951544509e-06,
      "loss": 0.464,
      "mean_token_accuracy": 0.8456936597824096,
      "num_tokens": 1030827161.0,
      "step": 30770
    },
    {
      "epoch": 1.8370940783190068,
      "grad_norm": 0.48703309893608093,
      "learning_rate": 5.811950152492503e-06,
      "loss": 0.422,
      "mean_token_accuracy": 0.8569008827209472,
      "num_tokens": 1030994841.0,
      "step": 30775
    },
    {
      "epoch": 1.8373925501432664,
      "grad_norm": 0.5385554432868958,
      "learning_rate": 5.8089966359653494e-06,
      "loss": 0.4417,
      "mean_token_accuracy": 0.8538112759590148,
      "num_tokens": 1031162521.0,
      "step": 30780
    },
    {
      "epoch": 1.8376910219675264,
      "grad_norm": 0.5271511673927307,
      "learning_rate": 5.806048402682435e-06,
      "loss": 0.45,
      "mean_token_accuracy": 0.8497017741203308,
      "num_tokens": 1031330201.0,
      "step": 30785
    },
    {
      "epoch": 1.837989493791786,
      "grad_norm": 0.50701504945755,
      "learning_rate": 5.803105453361856e-06,
      "loss": 0.3984,
      "mean_token_accuracy": 0.866115951538086,
      "num_tokens": 1031497881.0,
      "step": 30790
    },
    {
      "epoch": 1.8382879656160458,
      "grad_norm": 0.4824557602405548,
      "learning_rate": 5.800167788720423e-06,
      "loss": 0.3946,
      "mean_token_accuracy": 0.8666587233543396,
      "num_tokens": 1031665561.0,
      "step": 30795
    },
    {
      "epoch": 1.8385864374403056,
      "grad_norm": 0.5363814234733582,
      "learning_rate": 5.797235409473661e-06,
      "loss": 0.4524,
      "mean_token_accuracy": 0.8493081331253052,
      "num_tokens": 1031833241.0,
      "step": 30800
    },
    {
      "epoch": 1.8388849092645654,
      "grad_norm": 0.5110020637512207,
      "learning_rate": 5.794308316335798e-06,
      "loss": 0.4058,
      "mean_token_accuracy": 0.8607777714729309,
      "num_tokens": 1032000921.0,
      "step": 30805
    },
    {
      "epoch": 1.8391833810888252,
      "grad_norm": 0.5159292221069336,
      "learning_rate": 5.791386510019786e-06,
      "loss": 0.4365,
      "mean_token_accuracy": 0.851741623878479,
      "num_tokens": 1032168601.0,
      "step": 30810
    },
    {
      "epoch": 1.839481852913085,
      "grad_norm": 0.5154623985290527,
      "learning_rate": 5.788469991237288e-06,
      "loss": 0.4223,
      "mean_token_accuracy": 0.8582667231559753,
      "num_tokens": 1032336281.0,
      "step": 30815
    },
    {
      "epoch": 1.8397803247373448,
      "grad_norm": 0.5138291716575623,
      "learning_rate": 5.78555876069867e-06,
      "loss": 0.4061,
      "mean_token_accuracy": 0.8626983165740967,
      "num_tokens": 1032503961.0,
      "step": 30820
    },
    {
      "epoch": 1.8400787965616046,
      "grad_norm": 0.5058844685554504,
      "learning_rate": 5.78265281911302e-06,
      "loss": 0.4252,
      "mean_token_accuracy": 0.856214952468872,
      "num_tokens": 1032671641.0,
      "step": 30825
    },
    {
      "epoch": 1.8403772683858644,
      "grad_norm": 0.5139539837837219,
      "learning_rate": 5.779752167188129e-06,
      "loss": 0.391,
      "mean_token_accuracy": 0.867881429195404,
      "num_tokens": 1032839321.0,
      "step": 30830
    },
    {
      "epoch": 1.8406757402101241,
      "grad_norm": 0.5310842394828796,
      "learning_rate": 5.776856805630508e-06,
      "loss": 0.457,
      "mean_token_accuracy": 0.8469521760940552,
      "num_tokens": 1033007001.0,
      "step": 30835
    },
    {
      "epoch": 1.840974212034384,
      "grad_norm": 0.5083591341972351,
      "learning_rate": 5.773966735145379e-06,
      "loss": 0.4439,
      "mean_token_accuracy": 0.8496719598770142,
      "num_tokens": 1033174681.0,
      "step": 30840
    },
    {
      "epoch": 1.8412726838586437,
      "grad_norm": 0.5695140957832336,
      "learning_rate": 5.771081956436667e-06,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8546045541763305,
      "num_tokens": 1033342361.0,
      "step": 30845
    },
    {
      "epoch": 1.8415711556829035,
      "grad_norm": 0.557837963104248,
      "learning_rate": 5.768202470207014e-06,
      "loss": 0.4544,
      "mean_token_accuracy": 0.8481629490852356,
      "num_tokens": 1033510041.0,
      "step": 30850
    },
    {
      "epoch": 1.8418696275071633,
      "grad_norm": 0.5029990077018738,
      "learning_rate": 5.7653282771577705e-06,
      "loss": 0.4244,
      "mean_token_accuracy": 0.8581653237342834,
      "num_tokens": 1033677721.0,
      "step": 30855
    },
    {
      "epoch": 1.8421680993314231,
      "grad_norm": 0.5400208830833435,
      "learning_rate": 5.762459377989002e-06,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8476798295974731,
      "num_tokens": 1033845401.0,
      "step": 30860
    },
    {
      "epoch": 1.842466571155683,
      "grad_norm": 0.44514694809913635,
      "learning_rate": 5.759595773399483e-06,
      "loss": 0.4171,
      "mean_token_accuracy": 0.8604079723358155,
      "num_tokens": 1034013081.0,
      "step": 30865
    },
    {
      "epoch": 1.8427650429799427,
      "grad_norm": 0.49066242575645447,
      "learning_rate": 5.756737464086695e-06,
      "loss": 0.4167,
      "mean_token_accuracy": 0.8607479333877563,
      "num_tokens": 1034180761.0,
      "step": 30870
    },
    {
      "epoch": 1.8430635148042025,
      "grad_norm": 0.47519221901893616,
      "learning_rate": 5.753884450746831e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.8672492027282714,
      "num_tokens": 1034348441.0,
      "step": 30875
    },
    {
      "epoch": 1.8433619866284623,
      "grad_norm": 0.4528384804725647,
      "learning_rate": 5.7510367340747965e-06,
      "loss": 0.3736,
      "mean_token_accuracy": 0.8736132621765137,
      "num_tokens": 1034516121.0,
      "step": 30880
    },
    {
      "epoch": 1.843660458452722,
      "grad_norm": 0.512131929397583,
      "learning_rate": 5.7481943147642065e-06,
      "loss": 0.4181,
      "mean_token_accuracy": 0.8582190155982972,
      "num_tokens": 1034683801.0,
      "step": 30885
    },
    {
      "epoch": 1.8439589302769819,
      "grad_norm": 0.451454222202301,
      "learning_rate": 5.745357193507382e-06,
      "loss": 0.4176,
      "mean_token_accuracy": 0.8610879063606263,
      "num_tokens": 1034851481.0,
      "step": 30890
    },
    {
      "epoch": 1.8442574021012417,
      "grad_norm": 0.507235050201416,
      "learning_rate": 5.742525370995356e-06,
      "loss": 0.3913,
      "mean_token_accuracy": 0.867517602443695,
      "num_tokens": 1035019161.0,
      "step": 30895
    },
    {
      "epoch": 1.8445558739255015,
      "grad_norm": 0.5010097026824951,
      "learning_rate": 5.739698847917875e-06,
      "loss": 0.4036,
      "mean_token_accuracy": 0.8648056149482727,
      "num_tokens": 1035182536.0,
      "step": 30900
    },
    {
      "epoch": 1.8448543457497613,
      "grad_norm": 0.49267691373825073,
      "learning_rate": 5.736877624963391e-06,
      "loss": 0.4026,
      "mean_token_accuracy": 0.8634002923965454,
      "num_tokens": 1035345405.0,
      "step": 30905
    },
    {
      "epoch": 1.845152817574021,
      "grad_norm": 0.5309245586395264,
      "learning_rate": 5.734061702819062e-06,
      "loss": 0.4389,
      "mean_token_accuracy": 0.8526780486106873,
      "num_tokens": 1035513085.0,
      "step": 30910
    },
    {
      "epoch": 1.8454512893982808,
      "grad_norm": 0.5029569864273071,
      "learning_rate": 5.731251082170753e-06,
      "loss": 0.4231,
      "mean_token_accuracy": 0.8572885513305664,
      "num_tokens": 1035680765.0,
      "step": 30915
    },
    {
      "epoch": 1.8457497612225406,
      "grad_norm": 0.4928448498249054,
      "learning_rate": 5.728445763703051e-06,
      "loss": 0.4143,
      "mean_token_accuracy": 0.8612489461898803,
      "num_tokens": 1035848445.0,
      "step": 30920
    },
    {
      "epoch": 1.8460482330468004,
      "grad_norm": 0.49496352672576904,
      "learning_rate": 5.725645748099244e-06,
      "loss": 0.4179,
      "mean_token_accuracy": 0.8594178676605224,
      "num_tokens": 1036016125.0,
      "step": 30925
    },
    {
      "epoch": 1.8463467048710602,
      "grad_norm": 0.5211257934570312,
      "learning_rate": 5.72285103604132e-06,
      "loss": 0.4233,
      "mean_token_accuracy": 0.8571275115013123,
      "num_tokens": 1036183805.0,
      "step": 30930
    },
    {
      "epoch": 1.8466451766953198,
      "grad_norm": 0.5315496325492859,
      "learning_rate": 5.720061628209986e-06,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8540737390518188,
      "num_tokens": 1036351485.0,
      "step": 30935
    },
    {
      "epoch": 1.8469436485195798,
      "grad_norm": 0.4949491024017334,
      "learning_rate": 5.717277525284655e-06,
      "loss": 0.4245,
      "mean_token_accuracy": 0.8582309484481812,
      "num_tokens": 1036519165.0,
      "step": 30940
    },
    {
      "epoch": 1.8472421203438394,
      "grad_norm": 0.5251035094261169,
      "learning_rate": 5.714498727943445e-06,
      "loss": 0.4002,
      "mean_token_accuracy": 0.8639925837516784,
      "num_tokens": 1036686845.0,
      "step": 30945
    },
    {
      "epoch": 1.8475405921680994,
      "grad_norm": 0.5131856203079224,
      "learning_rate": 5.711725236863185e-06,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8512644648551941,
      "num_tokens": 1036854525.0,
      "step": 30950
    },
    {
      "epoch": 1.847839063992359,
      "grad_norm": 0.5783919095993042,
      "learning_rate": 5.7089570527194075e-06,
      "loss": 0.4403,
      "mean_token_accuracy": 0.8528092622756958,
      "num_tokens": 1037022205.0,
      "step": 30955
    },
    {
      "epoch": 1.848137535816619,
      "grad_norm": 0.5095967650413513,
      "learning_rate": 5.70619417618636e-06,
      "loss": 0.4258,
      "mean_token_accuracy": 0.8565430045127869,
      "num_tokens": 1037189885.0,
      "step": 30960
    },
    {
      "epoch": 1.8484360076408786,
      "grad_norm": 0.45795413851737976,
      "learning_rate": 5.703436607936984e-06,
      "loss": 0.3981,
      "mean_token_accuracy": 0.8656566858291626,
      "num_tokens": 1037357565.0,
      "step": 30965
    },
    {
      "epoch": 1.8487344794651386,
      "grad_norm": 0.47025933861732483,
      "learning_rate": 5.700684348642941e-06,
      "loss": 0.3886,
      "mean_token_accuracy": 0.8674400568008422,
      "num_tokens": 1037525245.0,
      "step": 30970
    },
    {
      "epoch": 1.8490329512893982,
      "grad_norm": 0.4744865894317627,
      "learning_rate": 5.697937398974596e-06,
      "loss": 0.4174,
      "mean_token_accuracy": 0.8595848679542542,
      "num_tokens": 1037692925.0,
      "step": 30975
    },
    {
      "epoch": 1.8493314231136582,
      "grad_norm": 0.47892388701438904,
      "learning_rate": 5.695195759601016e-06,
      "loss": 0.3958,
      "mean_token_accuracy": 0.8661994576454163,
      "num_tokens": 1037860605.0,
      "step": 30980
    },
    {
      "epoch": 1.8496298949379177,
      "grad_norm": 0.4974650740623474,
      "learning_rate": 5.6924594311899844e-06,
      "loss": 0.4008,
      "mean_token_accuracy": 0.865263032913208,
      "num_tokens": 1038028285.0,
      "step": 30985
    },
    {
      "epoch": 1.8499283667621778,
      "grad_norm": 0.4725826680660248,
      "learning_rate": 5.689728414407974e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8601992130279541,
      "num_tokens": 1038195965.0,
      "step": 30990
    },
    {
      "epoch": 1.8502268385864373,
      "grad_norm": 0.5021836161613464,
      "learning_rate": 5.6870027099201805e-06,
      "loss": 0.4397,
      "mean_token_accuracy": 0.8525169849395752,
      "num_tokens": 1038363645.0,
      "step": 30995
    },
    {
      "epoch": 1.8505253104106973,
      "grad_norm": 0.5390616059303284,
      "learning_rate": 5.6842823183905e-06,
      "loss": 0.405,
      "mean_token_accuracy": 0.8640283823013306,
      "num_tokens": 1038531325.0,
      "step": 31000
    },
    {
      "epoch": 1.850823782234957,
      "grad_norm": 0.5269567370414734,
      "learning_rate": 5.681567240481531e-06,
      "loss": 0.4363,
      "mean_token_accuracy": 0.8528390884399414,
      "num_tokens": 1038699005.0,
      "step": 31005
    },
    {
      "epoch": 1.851122254059217,
      "grad_norm": 0.4809347987174988,
      "learning_rate": 5.678857476854587e-06,
      "loss": 0.4282,
      "mean_token_accuracy": 0.8572050571441651,
      "num_tokens": 1038866685.0,
      "step": 31010
    },
    {
      "epoch": 1.8514207258834765,
      "grad_norm": 0.5040122866630554,
      "learning_rate": 5.676153028169674e-06,
      "loss": 0.4321,
      "mean_token_accuracy": 0.854360020160675,
      "num_tokens": 1039034365.0,
      "step": 31015
    },
    {
      "epoch": 1.8517191977077365,
      "grad_norm": 0.5039914846420288,
      "learning_rate": 5.67345389508551e-06,
      "loss": 0.4338,
      "mean_token_accuracy": 0.8536263942718506,
      "num_tokens": 1039202045.0,
      "step": 31020
    },
    {
      "epoch": 1.852017669531996,
      "grad_norm": 0.49863916635513306,
      "learning_rate": 5.670760078259525e-06,
      "loss": 0.4026,
      "mean_token_accuracy": 0.865263032913208,
      "num_tokens": 1039369725.0,
      "step": 31025
    },
    {
      "epoch": 1.852316141356256,
      "grad_norm": 0.493740975856781,
      "learning_rate": 5.668071578347844e-06,
      "loss": 0.4279,
      "mean_token_accuracy": 0.857055950164795,
      "num_tokens": 1039537405.0,
      "step": 31030
    },
    {
      "epoch": 1.8526146131805157,
      "grad_norm": 0.5113048553466797,
      "learning_rate": 5.665388396005303e-06,
      "loss": 0.3987,
      "mean_token_accuracy": 0.8637361288070678,
      "num_tokens": 1039705085.0,
      "step": 31035
    },
    {
      "epoch": 1.8529130850047757,
      "grad_norm": 0.5170607566833496,
      "learning_rate": 5.662710531885434e-06,
      "loss": 0.4453,
      "mean_token_accuracy": 0.8512048244476318,
      "num_tokens": 1039872765.0,
      "step": 31040
    },
    {
      "epoch": 1.8532115568290353,
      "grad_norm": 0.5332815647125244,
      "learning_rate": 5.660037986640491e-06,
      "loss": 0.438,
      "mean_token_accuracy": 0.8533162355422974,
      "num_tokens": 1040040445.0,
      "step": 31045
    },
    {
      "epoch": 1.8535100286532953,
      "grad_norm": 0.5392531752586365,
      "learning_rate": 5.657370760921413e-06,
      "loss": 0.4272,
      "mean_token_accuracy": 0.8567696452140808,
      "num_tokens": 1040208125.0,
      "step": 31050
    },
    {
      "epoch": 1.8538085004775549,
      "grad_norm": 0.4770124554634094,
      "learning_rate": 5.6547088553778545e-06,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8585530281066894,
      "num_tokens": 1040375805.0,
      "step": 31055
    },
    {
      "epoch": 1.8541069723018149,
      "grad_norm": 0.5287703275680542,
      "learning_rate": 5.652052270658172e-06,
      "loss": 0.4705,
      "mean_token_accuracy": 0.8429619550704956,
      "num_tokens": 1040543485.0,
      "step": 31060
    },
    {
      "epoch": 1.8544054441260744,
      "grad_norm": 0.49176666140556335,
      "learning_rate": 5.6494010074094265e-06,
      "loss": 0.4423,
      "mean_token_accuracy": 0.8530597686767578,
      "num_tokens": 1040711165.0,
      "step": 31065
    },
    {
      "epoch": 1.8547039159503342,
      "grad_norm": 0.5171224474906921,
      "learning_rate": 5.646755066277384e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8530299425125122,
      "num_tokens": 1040878845.0,
      "step": 31070
    },
    {
      "epoch": 1.855002387774594,
      "grad_norm": 0.4887007772922516,
      "learning_rate": 5.644114447906508e-06,
      "loss": 0.4481,
      "mean_token_accuracy": 0.8498687744140625,
      "num_tokens": 1041046525.0,
      "step": 31075
    },
    {
      "epoch": 1.8553008595988538,
      "grad_norm": 0.49851658940315247,
      "learning_rate": 5.6414791529399684e-06,
      "loss": 0.4355,
      "mean_token_accuracy": 0.8541274070739746,
      "num_tokens": 1041214205.0,
      "step": 31080
    },
    {
      "epoch": 1.8555993314231136,
      "grad_norm": 0.4828285872936249,
      "learning_rate": 5.638849182019645e-06,
      "loss": 0.427,
      "mean_token_accuracy": 0.8561732053756714,
      "num_tokens": 1041381885.0,
      "step": 31085
    },
    {
      "epoch": 1.8558978032473734,
      "grad_norm": 0.5649977922439575,
      "learning_rate": 5.636224535786117e-06,
      "loss": 0.4407,
      "mean_token_accuracy": 0.8528686165809631,
      "num_tokens": 1041543202.0,
      "step": 31090
    },
    {
      "epoch": 1.8561962750716332,
      "grad_norm": 0.5959926247596741,
      "learning_rate": 5.6336052148786605e-06,
      "loss": 0.4319,
      "mean_token_accuracy": 0.8543063282966614,
      "num_tokens": 1041710882.0,
      "step": 31095
    },
    {
      "epoch": 1.856494746895893,
      "grad_norm": 0.4736967086791992,
      "learning_rate": 5.6309912199352595e-06,
      "loss": 0.4475,
      "mean_token_accuracy": 0.8506322264671325,
      "num_tokens": 1041878562.0,
      "step": 31100
    },
    {
      "epoch": 1.8567932187201528,
      "grad_norm": 0.5529529452323914,
      "learning_rate": 5.628382551592605e-06,
      "loss": 0.4591,
      "mean_token_accuracy": 0.8482762813568115,
      "num_tokens": 1042046242.0,
      "step": 31105
    },
    {
      "epoch": 1.8570916905444126,
      "grad_norm": 0.4477444291114807,
      "learning_rate": 5.625779210486084e-06,
      "loss": 0.4064,
      "mean_token_accuracy": 0.8632768869400025,
      "num_tokens": 1042213922.0,
      "step": 31110
    },
    {
      "epoch": 1.8573901623686724,
      "grad_norm": 0.512795090675354,
      "learning_rate": 5.623181197249785e-06,
      "loss": 0.4731,
      "mean_token_accuracy": 0.8422879695892334,
      "num_tokens": 1042381602.0,
      "step": 31115
    },
    {
      "epoch": 1.8576886341929322,
      "grad_norm": 0.5364097952842712,
      "learning_rate": 5.620588512516507e-06,
      "loss": 0.4087,
      "mean_token_accuracy": 0.8622569441795349,
      "num_tokens": 1042549282.0,
      "step": 31120
    },
    {
      "epoch": 1.857987106017192,
      "grad_norm": 0.5044190287590027,
      "learning_rate": 5.618001156917744e-06,
      "loss": 0.4242,
      "mean_token_accuracy": 0.8563938975334168,
      "num_tokens": 1042716962.0,
      "step": 31125
    },
    {
      "epoch": 1.8582855778414518,
      "grad_norm": 0.4930665194988251,
      "learning_rate": 5.6154191310836975e-06,
      "loss": 0.4285,
      "mean_token_accuracy": 0.8569485902786255,
      "num_tokens": 1042884642.0,
      "step": 31130
    },
    {
      "epoch": 1.8585840496657116,
      "grad_norm": 0.43919238448143005,
      "learning_rate": 5.612842435643263e-06,
      "loss": 0.4016,
      "mean_token_accuracy": 0.8655254602432251,
      "num_tokens": 1043052322.0,
      "step": 31135
    },
    {
      "epoch": 1.8588825214899714,
      "grad_norm": 0.5250810980796814,
      "learning_rate": 5.610271071224042e-06,
      "loss": 0.4134,
      "mean_token_accuracy": 0.8605869054794312,
      "num_tokens": 1043220002.0,
      "step": 31140
    },
    {
      "epoch": 1.8591809933142311,
      "grad_norm": 0.4844275712966919,
      "learning_rate": 5.6077050384523405e-06,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8537755131721496,
      "num_tokens": 1043387682.0,
      "step": 31145
    },
    {
      "epoch": 1.859479465138491,
      "grad_norm": 0.573715090751648,
      "learning_rate": 5.605144337953162e-06,
      "loss": 0.4765,
      "mean_token_accuracy": 0.8424907445907592,
      "num_tokens": 1043555362.0,
      "step": 31150
    },
    {
      "epoch": 1.8597779369627507,
      "grad_norm": 0.47320133447647095,
      "learning_rate": 5.602588970350216e-06,
      "loss": 0.4385,
      "mean_token_accuracy": 0.8532029151916504,
      "num_tokens": 1043723042.0,
      "step": 31155
    },
    {
      "epoch": 1.8600764087870105,
      "grad_norm": 0.6306964755058289,
      "learning_rate": 5.6000389362659035e-06,
      "loss": 0.4136,
      "mean_token_accuracy": 0.86192946434021,
      "num_tokens": 1043883614.0,
      "step": 31160
    },
    {
      "epoch": 1.8603748806112703,
      "grad_norm": 0.477536678314209,
      "learning_rate": 5.597494236321334e-06,
      "loss": 0.4103,
      "mean_token_accuracy": 0.8599069476127624,
      "num_tokens": 1044051294.0,
      "step": 31165
    },
    {
      "epoch": 1.8606733524355301,
      "grad_norm": 0.5094892382621765,
      "learning_rate": 5.594954871136321e-06,
      "loss": 0.4436,
      "mean_token_accuracy": 0.8513777852058411,
      "num_tokens": 1044218974.0,
      "step": 31170
    },
    {
      "epoch": 1.86097182425979,
      "grad_norm": 0.4471518397331238,
      "learning_rate": 5.592420841329364e-06,
      "loss": 0.4208,
      "mean_token_accuracy": 0.858320415019989,
      "num_tokens": 1044386654.0,
      "step": 31175
    },
    {
      "epoch": 1.8612702960840497,
      "grad_norm": 0.5190898776054382,
      "learning_rate": 5.589892147517684e-06,
      "loss": 0.4323,
      "mean_token_accuracy": 0.8539067149162293,
      "num_tokens": 1044554334.0,
      "step": 31180
    },
    {
      "epoch": 1.8615687679083095,
      "grad_norm": 0.5199024677276611,
      "learning_rate": 5.587368790317184e-06,
      "loss": 0.4238,
      "mean_token_accuracy": 0.8567040443420411,
      "num_tokens": 1044722014.0,
      "step": 31185
    },
    {
      "epoch": 1.8618672397325693,
      "grad_norm": 0.5011622905731201,
      "learning_rate": 5.584850770342475e-06,
      "loss": 0.4493,
      "mean_token_accuracy": 0.8490337610244751,
      "num_tokens": 1044889694.0,
      "step": 31190
    },
    {
      "epoch": 1.862165711556829,
      "grad_norm": 0.5112608075141907,
      "learning_rate": 5.582338088206869e-06,
      "loss": 0.4384,
      "mean_token_accuracy": 0.8545508623123169,
      "num_tokens": 1045057374.0,
      "step": 31195
    },
    {
      "epoch": 1.8624641833810889,
      "grad_norm": 0.5377089381217957,
      "learning_rate": 5.579830744522375e-06,
      "loss": 0.4339,
      "mean_token_accuracy": 0.8547954201698303,
      "num_tokens": 1045225054.0,
      "step": 31200
    },
    {
      "epoch": 1.8627626552053487,
      "grad_norm": 0.5588016510009766,
      "learning_rate": 5.577328739899703e-06,
      "loss": 0.4548,
      "mean_token_accuracy": 0.8471967101097106,
      "num_tokens": 1045392734.0,
      "step": 31205
    },
    {
      "epoch": 1.8630611270296082,
      "grad_norm": 0.6160690784454346,
      "learning_rate": 5.574832074948264e-06,
      "loss": 0.4379,
      "mean_token_accuracy": 0.8541095018386841,
      "num_tokens": 1045560414.0,
      "step": 31210
    },
    {
      "epoch": 1.8633595988538683,
      "grad_norm": 0.5562750101089478,
      "learning_rate": 5.572340750276166e-06,
      "loss": 0.4241,
      "mean_token_accuracy": 0.8574674844741821,
      "num_tokens": 1045728094.0,
      "step": 31215
    },
    {
      "epoch": 1.8636580706781278,
      "grad_norm": 0.5025776624679565,
      "learning_rate": 5.569854766490215e-06,
      "loss": 0.4739,
      "mean_token_accuracy": 0.8429500102996826,
      "num_tokens": 1045895774.0,
      "step": 31220
    },
    {
      "epoch": 1.8639565425023878,
      "grad_norm": 0.5171743035316467,
      "learning_rate": 5.567374124195916e-06,
      "loss": 0.4428,
      "mean_token_accuracy": 0.8515209436416626,
      "num_tokens": 1046063454.0,
      "step": 31225
    },
    {
      "epoch": 1.8642550143266474,
      "grad_norm": 0.49668410420417786,
      "learning_rate": 5.564898823997484e-06,
      "loss": 0.4135,
      "mean_token_accuracy": 0.8606405854225159,
      "num_tokens": 1046231134.0,
      "step": 31230
    },
    {
      "epoch": 1.8645534861509074,
      "grad_norm": 0.5170344114303589,
      "learning_rate": 5.562428866497817e-06,
      "loss": 0.409,
      "mean_token_accuracy": 0.8612370371818543,
      "num_tokens": 1046398814.0,
      "step": 31235
    },
    {
      "epoch": 1.864851957975167,
      "grad_norm": 0.4819069802761078,
      "learning_rate": 5.559964252298523e-06,
      "loss": 0.3992,
      "mean_token_accuracy": 0.8660205125808715,
      "num_tokens": 1046566494.0,
      "step": 31240
    },
    {
      "epoch": 1.865150429799427,
      "grad_norm": 0.5160818099975586,
      "learning_rate": 5.557504981999899e-06,
      "loss": 0.4216,
      "mean_token_accuracy": 0.857932710647583,
      "num_tokens": 1046734174.0,
      "step": 31245
    },
    {
      "epoch": 1.8654489016236866,
      "grad_norm": 0.46007150411605835,
      "learning_rate": 5.555051056200949e-06,
      "loss": 0.3957,
      "mean_token_accuracy": 0.8672014832496643,
      "num_tokens": 1046901854.0,
      "step": 31250
    },
    {
      "epoch": 1.8657473734479466,
      "grad_norm": 0.4852820336818695,
      "learning_rate": 5.552602475499375e-06,
      "loss": 0.412,
      "mean_token_accuracy": 0.8617917180061341,
      "num_tokens": 1047069534.0,
      "step": 31255
    },
    {
      "epoch": 1.8660458452722062,
      "grad_norm": 0.5058529376983643,
      "learning_rate": 5.5501592404915664e-06,
      "loss": 0.3849,
      "mean_token_accuracy": 0.8685196161270141,
      "num_tokens": 1047237214.0,
      "step": 31260
    },
    {
      "epoch": 1.8663443170964662,
      "grad_norm": 0.526887834072113,
      "learning_rate": 5.547721351772624e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8532088875770569,
      "num_tokens": 1047404894.0,
      "step": 31265
    },
    {
      "epoch": 1.8666427889207258,
      "grad_norm": 0.5444082021713257,
      "learning_rate": 5.54528880993634e-06,
      "loss": 0.4298,
      "mean_token_accuracy": 0.854324221611023,
      "num_tokens": 1047572574.0,
      "step": 31270
    },
    {
      "epoch": 1.8669412607449858,
      "grad_norm": 0.4989696741104126,
      "learning_rate": 5.542861615575201e-06,
      "loss": 0.4205,
      "mean_token_accuracy": 0.8583621621131897,
      "num_tokens": 1047740254.0,
      "step": 31275
    },
    {
      "epoch": 1.8672397325692454,
      "grad_norm": 0.5281450748443604,
      "learning_rate": 5.540439769280403e-06,
      "loss": 0.4368,
      "mean_token_accuracy": 0.8532088756561279,
      "num_tokens": 1047907934.0,
      "step": 31280
    },
    {
      "epoch": 1.8675382043935054,
      "grad_norm": 0.4540364444255829,
      "learning_rate": 5.538023271641821e-06,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8563342452049255,
      "num_tokens": 1048075614.0,
      "step": 31285
    },
    {
      "epoch": 1.867836676217765,
      "grad_norm": 0.4731082022190094,
      "learning_rate": 5.535612123248045e-06,
      "loss": 0.4109,
      "mean_token_accuracy": 0.8609893798828125,
      "num_tokens": 1048238899.0,
      "step": 31290
    },
    {
      "epoch": 1.868135148042025,
      "grad_norm": 0.6066111922264099,
      "learning_rate": 5.533206324686354e-06,
      "loss": 0.4409,
      "mean_token_accuracy": 0.853483247756958,
      "num_tokens": 1048406579.0,
      "step": 31295
    },
    {
      "epoch": 1.8684336198662845,
      "grad_norm": 0.5273709893226624,
      "learning_rate": 5.5308058765427205e-06,
      "loss": 0.4191,
      "mean_token_accuracy": 0.8587259888648987,
      "num_tokens": 1048574259.0,
      "step": 31300
    },
    {
      "epoch": 1.8687320916905446,
      "grad_norm": 0.49483057856559753,
      "learning_rate": 5.528410779401821e-06,
      "loss": 0.4204,
      "mean_token_accuracy": 0.8597041726112366,
      "num_tokens": 1048741939.0,
      "step": 31305
    },
    {
      "epoch": 1.8690305635148041,
      "grad_norm": 0.5066244602203369,
      "learning_rate": 5.526021033847026e-06,
      "loss": 0.4282,
      "mean_token_accuracy": 0.8569843769073486,
      "num_tokens": 1048909619.0,
      "step": 31310
    },
    {
      "epoch": 1.8693290353390641,
      "grad_norm": 0.48427721858024597,
      "learning_rate": 5.523636640460405e-06,
      "loss": 0.4212,
      "mean_token_accuracy": 0.8600918531417847,
      "num_tokens": 1049077299.0,
      "step": 31315
    },
    {
      "epoch": 1.8696275071633237,
      "grad_norm": 0.46243947744369507,
      "learning_rate": 5.521257599822711e-06,
      "loss": 0.4414,
      "mean_token_accuracy": 0.8525647044181823,
      "num_tokens": 1049244979.0,
      "step": 31320
    },
    {
      "epoch": 1.8699259789875837,
      "grad_norm": 0.5038584470748901,
      "learning_rate": 5.518883912513413e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8539246201515198,
      "num_tokens": 1049412659.0,
      "step": 31325
    },
    {
      "epoch": 1.8702244508118433,
      "grad_norm": 0.48492860794067383,
      "learning_rate": 5.5165155791106635e-06,
      "loss": 0.4196,
      "mean_token_accuracy": 0.8589764952659606,
      "num_tokens": 1049580339.0,
      "step": 31330
    },
    {
      "epoch": 1.8705229226361033,
      "grad_norm": 0.5504201054573059,
      "learning_rate": 5.514152600191311e-06,
      "loss": 0.4257,
      "mean_token_accuracy": 0.8546463131904602,
      "num_tokens": 1049748019.0,
      "step": 31335
    },
    {
      "epoch": 1.8708213944603629,
      "grad_norm": 0.5138734579086304,
      "learning_rate": 5.511794976330908e-06,
      "loss": 0.44,
      "mean_token_accuracy": 0.8532685160636901,
      "num_tokens": 1049915699.0,
      "step": 31340
    },
    {
      "epoch": 1.8711198662846227,
      "grad_norm": 0.5428465604782104,
      "learning_rate": 5.509442708103692e-06,
      "loss": 0.4199,
      "mean_token_accuracy": 0.859071922302246,
      "num_tokens": 1050083379.0,
      "step": 31345
    },
    {
      "epoch": 1.8714183381088825,
      "grad_norm": 0.554255485534668,
      "learning_rate": 5.507095796082603e-06,
      "loss": 0.4291,
      "mean_token_accuracy": 0.8551055669784546,
      "num_tokens": 1050251059.0,
      "step": 31350
    },
    {
      "epoch": 1.8717168099331423,
      "grad_norm": 0.4803463816642761,
      "learning_rate": 5.504754240839277e-06,
      "loss": 0.3983,
      "mean_token_accuracy": 0.8641178727149963,
      "num_tokens": 1050418739.0,
      "step": 31355
    },
    {
      "epoch": 1.872015281757402,
      "grad_norm": 0.5302906632423401,
      "learning_rate": 5.502418042944041e-06,
      "loss": 0.4397,
      "mean_token_accuracy": 0.8524573564529419,
      "num_tokens": 1050586419.0,
      "step": 31360
    },
    {
      "epoch": 1.8723137535816619,
      "grad_norm": 0.5589760541915894,
      "learning_rate": 5.500087202965919e-06,
      "loss": 0.4184,
      "mean_token_accuracy": 0.8608970522880555,
      "num_tokens": 1050754099.0,
      "step": 31365
    },
    {
      "epoch": 1.8726122254059216,
      "grad_norm": 0.5340816378593445,
      "learning_rate": 5.497761721472629e-06,
      "loss": 0.4462,
      "mean_token_accuracy": 0.8521531581878662,
      "num_tokens": 1050921779.0,
      "step": 31370
    },
    {
      "epoch": 1.8729106972301814,
      "grad_norm": 0.556317150592804,
      "learning_rate": 5.495441599030591e-06,
      "loss": 0.4465,
      "mean_token_accuracy": 0.8506620526313782,
      "num_tokens": 1051089459.0,
      "step": 31375
    },
    {
      "epoch": 1.8732091690544412,
      "grad_norm": 0.5510767102241516,
      "learning_rate": 5.493126836204907e-06,
      "loss": 0.4374,
      "mean_token_accuracy": 0.8553322315216064,
      "num_tokens": 1051257139.0,
      "step": 31380
    },
    {
      "epoch": 1.873507640878701,
      "grad_norm": 0.4932164251804352,
      "learning_rate": 5.490817433559382e-06,
      "loss": 0.4202,
      "mean_token_accuracy": 0.8583919882774353,
      "num_tokens": 1051424819.0,
      "step": 31385
    },
    {
      "epoch": 1.8738061127029608,
      "grad_norm": 0.4719121754169464,
      "learning_rate": 5.488513391656512e-06,
      "loss": 0.4132,
      "mean_token_accuracy": 0.8617320656776428,
      "num_tokens": 1051592499.0,
      "step": 31390
    },
    {
      "epoch": 1.8741045845272206,
      "grad_norm": 0.46552151441574097,
      "learning_rate": 5.486214711057493e-06,
      "loss": 0.364,
      "mean_token_accuracy": 0.8763569116592407,
      "num_tokens": 1051760179.0,
      "step": 31395
    },
    {
      "epoch": 1.8744030563514804,
      "grad_norm": 0.514227569103241,
      "learning_rate": 5.483921392322212e-06,
      "loss": 0.4231,
      "mean_token_accuracy": 0.8592568278312683,
      "num_tokens": 1051927859.0,
      "step": 31400
    },
    {
      "epoch": 1.8747015281757402,
      "grad_norm": 0.5355833768844604,
      "learning_rate": 5.481633436009242e-06,
      "loss": 0.4254,
      "mean_token_accuracy": 0.8574853897094726,
      "num_tokens": 1052095539.0,
      "step": 31405
    },
    {
      "epoch": 1.875,
      "grad_norm": 0.5531699061393738,
      "learning_rate": 5.479350842675863e-06,
      "loss": 0.4084,
      "mean_token_accuracy": 0.8619646906852723,
      "num_tokens": 1052263219.0,
      "step": 31410
    },
    {
      "epoch": 1.8752984718242598,
      "grad_norm": 0.5218827128410339,
      "learning_rate": 5.477073612878039e-06,
      "loss": 0.47,
      "mean_token_accuracy": 0.8432780504226685,
      "num_tokens": 1052430899.0,
      "step": 31415
    },
    {
      "epoch": 1.8755969436485196,
      "grad_norm": 0.5016190409660339,
      "learning_rate": 5.474801747170434e-06,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8582011222839355,
      "num_tokens": 1052598579.0,
      "step": 31420
    },
    {
      "epoch": 1.8758954154727794,
      "grad_norm": 0.5406301617622375,
      "learning_rate": 5.472535246106404e-06,
      "loss": 0.445,
      "mean_token_accuracy": 0.851604449748993,
      "num_tokens": 1052766259.0,
      "step": 31425
    },
    {
      "epoch": 1.8761938872970392,
      "grad_norm": 0.4535290002822876,
      "learning_rate": 5.470274110237993e-06,
      "loss": 0.3857,
      "mean_token_accuracy": 0.8698556661605835,
      "num_tokens": 1052933939.0,
      "step": 31430
    },
    {
      "epoch": 1.876492359121299,
      "grad_norm": 0.5444434285163879,
      "learning_rate": 5.468018340115945e-06,
      "loss": 0.4305,
      "mean_token_accuracy": 0.855839216709137,
      "num_tokens": 1053101619.0,
      "step": 31435
    },
    {
      "epoch": 1.8767908309455588,
      "grad_norm": 0.4442603588104248,
      "learning_rate": 5.465767936289695e-06,
      "loss": 0.4033,
      "mean_token_accuracy": 0.863169503211975,
      "num_tokens": 1053269299.0,
      "step": 31440
    },
    {
      "epoch": 1.8770893027698186,
      "grad_norm": 0.5314321517944336,
      "learning_rate": 5.4635228993073675e-06,
      "loss": 0.4498,
      "mean_token_accuracy": 0.8506203055381775,
      "num_tokens": 1053436979.0,
      "step": 31445
    },
    {
      "epoch": 1.8773877745940784,
      "grad_norm": 0.5004502534866333,
      "learning_rate": 5.461283229715787e-06,
      "loss": 0.4576,
      "mean_token_accuracy": 0.8473816037178039,
      "num_tokens": 1053604659.0,
      "step": 31450
    },
    {
      "epoch": 1.8776862464183381,
      "grad_norm": 0.48730525374412537,
      "learning_rate": 5.4590489280604606e-06,
      "loss": 0.4277,
      "mean_token_accuracy": 0.8568591237068176,
      "num_tokens": 1053772339.0,
      "step": 31455
    },
    {
      "epoch": 1.877984718242598,
      "grad_norm": 0.5040608644485474,
      "learning_rate": 5.456819994885603e-06,
      "loss": 0.3913,
      "mean_token_accuracy": 0.8682571887969971,
      "num_tokens": 1053940019.0,
      "step": 31460
    },
    {
      "epoch": 1.8782831900668577,
      "grad_norm": 0.45529037714004517,
      "learning_rate": 5.454596430734101e-06,
      "loss": 0.4069,
      "mean_token_accuracy": 0.8635810613632202,
      "num_tokens": 1054107699.0,
      "step": 31465
    },
    {
      "epoch": 1.8785816618911175,
      "grad_norm": 0.5126983523368835,
      "learning_rate": 5.452378236147551e-06,
      "loss": 0.4399,
      "mean_token_accuracy": 0.8538411140441895,
      "num_tokens": 1054275379.0,
      "step": 31470
    },
    {
      "epoch": 1.8788801337153773,
      "grad_norm": 0.4560665786266327,
      "learning_rate": 5.450165411666235e-06,
      "loss": 0.45,
      "mean_token_accuracy": 0.8493021726608276,
      "num_tokens": 1054443059.0,
      "step": 31475
    },
    {
      "epoch": 1.8791786055396371,
      "grad_norm": 0.460692822933197,
      "learning_rate": 5.447957957829126e-06,
      "loss": 0.4081,
      "mean_token_accuracy": 0.8611177444458008,
      "num_tokens": 1054610739.0,
      "step": 31480
    },
    {
      "epoch": 1.8794770773638967,
      "grad_norm": 0.5564628839492798,
      "learning_rate": 5.4457558751738905e-06,
      "loss": 0.4584,
      "mean_token_accuracy": 0.8453656315803528,
      "num_tokens": 1054778419.0,
      "step": 31485
    },
    {
      "epoch": 1.8797755491881567,
      "grad_norm": 0.5538517832756042,
      "learning_rate": 5.443559164236885e-06,
      "loss": 0.4225,
      "mean_token_accuracy": 0.8589824557304382,
      "num_tokens": 1054946099.0,
      "step": 31490
    },
    {
      "epoch": 1.8800740210124163,
      "grad_norm": 0.5153970122337341,
      "learning_rate": 5.4413678255531606e-06,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8511034250259399,
      "num_tokens": 1055113779.0,
      "step": 31495
    },
    {
      "epoch": 1.8803724928366763,
      "grad_norm": 0.5142152905464172,
      "learning_rate": 5.43918185965646e-06,
      "loss": 0.4147,
      "mean_token_accuracy": 0.8613205313682556,
      "num_tokens": 1055281459.0,
      "step": 31500
    },
    {
      "epoch": 1.8806709646609359,
      "grad_norm": 0.4856979250907898,
      "learning_rate": 5.437001267079212e-06,
      "loss": 0.4205,
      "mean_token_accuracy": 0.8587975621223449,
      "num_tokens": 1055449139.0,
      "step": 31505
    },
    {
      "epoch": 1.8809694364851959,
      "grad_norm": 0.5209598541259766,
      "learning_rate": 5.434826048352544e-06,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8500059723854065,
      "num_tokens": 1055616819.0,
      "step": 31510
    },
    {
      "epoch": 1.8812679083094554,
      "grad_norm": 0.5124226808547974,
      "learning_rate": 5.4326562040062634e-06,
      "loss": 0.3999,
      "mean_token_accuracy": 0.8635631561279297,
      "num_tokens": 1055784499.0,
      "step": 31515
    },
    {
      "epoch": 1.8815663801337155,
      "grad_norm": 0.519205629825592,
      "learning_rate": 5.430491734568883e-06,
      "loss": 0.3998,
      "mean_token_accuracy": 0.8645413398742676,
      "num_tokens": 1055952179.0,
      "step": 31520
    },
    {
      "epoch": 1.881864851957975,
      "grad_norm": 0.5213624238967896,
      "learning_rate": 5.428332640567598e-06,
      "loss": 0.4081,
      "mean_token_accuracy": 0.8631814360618592,
      "num_tokens": 1056119859.0,
      "step": 31525
    },
    {
      "epoch": 1.882163323782235,
      "grad_norm": 0.48710131645202637,
      "learning_rate": 5.426178922528295e-06,
      "loss": 0.4177,
      "mean_token_accuracy": 0.8594834804534912,
      "num_tokens": 1056287539.0,
      "step": 31530
    },
    {
      "epoch": 1.8824617956064946,
      "grad_norm": 0.46749168634414673,
      "learning_rate": 5.4240305809755495e-06,
      "loss": 0.4369,
      "mean_token_accuracy": 0.8547298192977906,
      "num_tokens": 1056455219.0,
      "step": 31535
    },
    {
      "epoch": 1.8827602674307546,
      "grad_norm": 0.5485461950302124,
      "learning_rate": 5.421887616432633e-06,
      "loss": 0.4318,
      "mean_token_accuracy": 0.8540021419525147,
      "num_tokens": 1056622899.0,
      "step": 31540
    },
    {
      "epoch": 1.8830587392550142,
      "grad_norm": 0.4804892838001251,
      "learning_rate": 5.419750029421506e-06,
      "loss": 0.426,
      "mean_token_accuracy": 0.8559584975242615,
      "num_tokens": 1056790579.0,
      "step": 31545
    },
    {
      "epoch": 1.8833572110792742,
      "grad_norm": 0.5282833576202393,
      "learning_rate": 5.417617820462809e-06,
      "loss": 0.4234,
      "mean_token_accuracy": 0.8564952969551086,
      "num_tokens": 1056958259.0,
      "step": 31550
    },
    {
      "epoch": 1.8836556829035338,
      "grad_norm": 0.4980647563934326,
      "learning_rate": 5.4154909900758855e-06,
      "loss": 0.4247,
      "mean_token_accuracy": 0.8575629234313965,
      "num_tokens": 1057125939.0,
      "step": 31555
    },
    {
      "epoch": 1.8839541547277938,
      "grad_norm": 0.48746436834335327,
      "learning_rate": 5.413369538778768e-06,
      "loss": 0.4367,
      "mean_token_accuracy": 0.8544136881828308,
      "num_tokens": 1057293619.0,
      "step": 31560
    },
    {
      "epoch": 1.8842526265520534,
      "grad_norm": 0.5036244988441467,
      "learning_rate": 5.411253467088171e-06,
      "loss": 0.4309,
      "mean_token_accuracy": 0.8553441524505615,
      "num_tokens": 1057461299.0,
      "step": 31565
    },
    {
      "epoch": 1.8845510983763134,
      "grad_norm": 0.6265799403190613,
      "learning_rate": 5.409142775519507e-06,
      "loss": 0.4513,
      "mean_token_accuracy": 0.848646080493927,
      "num_tokens": 1057628979.0,
      "step": 31570
    },
    {
      "epoch": 1.884849570200573,
      "grad_norm": 0.4896012842655182,
      "learning_rate": 5.407037464586865e-06,
      "loss": 0.3926,
      "mean_token_accuracy": 0.8680305361747742,
      "num_tokens": 1057796659.0,
      "step": 31575
    },
    {
      "epoch": 1.885148042024833,
      "grad_norm": 0.5121934413909912,
      "learning_rate": 5.404937534803038e-06,
      "loss": 0.4555,
      "mean_token_accuracy": 0.8478885769844056,
      "num_tokens": 1057964339.0,
      "step": 31580
    },
    {
      "epoch": 1.8854465138490926,
      "grad_norm": 0.46774473786354065,
      "learning_rate": 5.402842986679505e-06,
      "loss": 0.4186,
      "mean_token_accuracy": 0.8576344966888427,
      "num_tokens": 1058132019.0,
      "step": 31585
    },
    {
      "epoch": 1.8857449856733526,
      "grad_norm": 0.5030744671821594,
      "learning_rate": 5.400753820726429e-06,
      "loss": 0.3922,
      "mean_token_accuracy": 0.8677024960517883,
      "num_tokens": 1058299699.0,
      "step": 31590
    },
    {
      "epoch": 1.8860434574976122,
      "grad_norm": 0.4890047609806061,
      "learning_rate": 5.398670037452664e-06,
      "loss": 0.3895,
      "mean_token_accuracy": 0.8691935896873474,
      "num_tokens": 1058467379.0,
      "step": 31595
    },
    {
      "epoch": 1.8863419293218722,
      "grad_norm": 0.5055318474769592,
      "learning_rate": 5.396591637365756e-06,
      "loss": 0.4318,
      "mean_token_accuracy": 0.8578642725944519,
      "num_tokens": 1058628606.0,
      "step": 31600
    },
    {
      "epoch": 1.8866404011461317,
      "grad_norm": 0.49970972537994385,
      "learning_rate": 5.394518620971936e-06,
      "loss": 0.3923,
      "mean_token_accuracy": 0.8688059210777282,
      "num_tokens": 1058796286.0,
      "step": 31605
    },
    {
      "epoch": 1.8869388729703918,
      "grad_norm": 0.6755126714706421,
      "learning_rate": 5.39245098877613e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8520517587661743,
      "num_tokens": 1058963966.0,
      "step": 31610
    },
    {
      "epoch": 1.8872373447946513,
      "grad_norm": 0.5788027048110962,
      "learning_rate": 5.390388741281942e-06,
      "loss": 0.4076,
      "mean_token_accuracy": 0.8621138095855713,
      "num_tokens": 1059131646.0,
      "step": 31615
    },
    {
      "epoch": 1.8875358166189111,
      "grad_norm": 0.46830010414123535,
      "learning_rate": 5.3883318789916735e-06,
      "loss": 0.3929,
      "mean_token_accuracy": 0.8661636590957642,
      "num_tokens": 1059299326.0,
      "step": 31620
    },
    {
      "epoch": 1.887834288443171,
      "grad_norm": 0.49278074502944946,
      "learning_rate": 5.386280402406308e-06,
      "loss": 0.4455,
      "mean_token_accuracy": 0.8518072247505188,
      "num_tokens": 1059467006.0,
      "step": 31625
    },
    {
      "epoch": 1.8881327602674307,
      "grad_norm": 0.4522901475429535,
      "learning_rate": 5.384234312025525e-06,
      "loss": 0.4092,
      "mean_token_accuracy": 0.8618513584136963,
      "num_tokens": 1059634686.0,
      "step": 31630
    },
    {
      "epoch": 1.8884312320916905,
      "grad_norm": 0.5867912173271179,
      "learning_rate": 5.382193608347684e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8537039279937744,
      "num_tokens": 1059802366.0,
      "step": 31635
    },
    {
      "epoch": 1.8887297039159503,
      "grad_norm": 0.5196858644485474,
      "learning_rate": 5.380158291869836e-06,
      "loss": 0.4183,
      "mean_token_accuracy": 0.8600083470344544,
      "num_tokens": 1059970046.0,
      "step": 31640
    },
    {
      "epoch": 1.88902817574021,
      "grad_norm": 0.5334433317184448,
      "learning_rate": 5.378128363087728e-06,
      "loss": 0.4235,
      "mean_token_accuracy": 0.8578790426254272,
      "num_tokens": 1060137726.0,
      "step": 31645
    },
    {
      "epoch": 1.8893266475644699,
      "grad_norm": 0.48552897572517395,
      "learning_rate": 5.376103822495774e-06,
      "loss": 0.4468,
      "mean_token_accuracy": 0.851503050327301,
      "num_tokens": 1060305406.0,
      "step": 31650
    },
    {
      "epoch": 1.8896251193887297,
      "grad_norm": 0.5461321473121643,
      "learning_rate": 5.374084670587097e-06,
      "loss": 0.39,
      "mean_token_accuracy": 0.8679887771606445,
      "num_tokens": 1060473086.0,
      "step": 31655
    },
    {
      "epoch": 1.8899235912129895,
      "grad_norm": 0.523984968662262,
      "learning_rate": 5.372070907853496e-06,
      "loss": 0.4169,
      "mean_token_accuracy": 0.8611892938613892,
      "num_tokens": 1060640766.0,
      "step": 31660
    },
    {
      "epoch": 1.8902220630372493,
      "grad_norm": 0.4599439799785614,
      "learning_rate": 5.370062534785459e-06,
      "loss": 0.3978,
      "mean_token_accuracy": 0.8666467905044556,
      "num_tokens": 1060808446.0,
      "step": 31665
    },
    {
      "epoch": 1.890520534861509,
      "grad_norm": 0.46077027916908264,
      "learning_rate": 5.368059551872164e-06,
      "loss": 0.4009,
      "mean_token_accuracy": 0.8674877762794495,
      "num_tokens": 1060976126.0,
      "step": 31670
    },
    {
      "epoch": 1.8908190066857689,
      "grad_norm": 0.6345248818397522,
      "learning_rate": 5.366061959601475e-06,
      "loss": 0.4813,
      "mean_token_accuracy": 0.8408385992050171,
      "num_tokens": 1061143806.0,
      "step": 31675
    },
    {
      "epoch": 1.8911174785100286,
      "grad_norm": 0.5159919261932373,
      "learning_rate": 5.36406975845994e-06,
      "loss": 0.4338,
      "mean_token_accuracy": 0.8544196605682373,
      "num_tokens": 1061311486.0,
      "step": 31680
    },
    {
      "epoch": 1.8914159503342884,
      "grad_norm": 0.5581896305084229,
      "learning_rate": 5.3620829489327995e-06,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8435345411300659,
      "num_tokens": 1061479166.0,
      "step": 31685
    },
    {
      "epoch": 1.8917144221585482,
      "grad_norm": 0.5385413765907288,
      "learning_rate": 5.360101531503973e-06,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8542287945747375,
      "num_tokens": 1061646846.0,
      "step": 31690
    },
    {
      "epoch": 1.892012893982808,
      "grad_norm": 0.5238235592842102,
      "learning_rate": 5.358125506656078e-06,
      "loss": 0.4502,
      "mean_token_accuracy": 0.8497596025466919,
      "num_tokens": 1061809829.0,
      "step": 31695
    },
    {
      "epoch": 1.8923113658070678,
      "grad_norm": 0.4854484796524048,
      "learning_rate": 5.356154874870405e-06,
      "loss": 0.4288,
      "mean_token_accuracy": 0.8591136693954468,
      "num_tokens": 1061977509.0,
      "step": 31700
    },
    {
      "epoch": 1.8926098376313276,
      "grad_norm": 0.48504146933555603,
      "learning_rate": 5.354189636626943e-06,
      "loss": 0.4302,
      "mean_token_accuracy": 0.8553322196006775,
      "num_tokens": 1062145189.0,
      "step": 31705
    },
    {
      "epoch": 1.8929083094555874,
      "grad_norm": 0.47814470529556274,
      "learning_rate": 5.352229792404358e-06,
      "loss": 0.4002,
      "mean_token_accuracy": 0.8656745910644531,
      "num_tokens": 1062312869.0,
      "step": 31710
    },
    {
      "epoch": 1.8932067812798472,
      "grad_norm": 0.449599951505661,
      "learning_rate": 5.350275342680009e-06,
      "loss": 0.42,
      "mean_token_accuracy": 0.8595669865608215,
      "num_tokens": 1062480549.0,
      "step": 31715
    },
    {
      "epoch": 1.893505253104107,
      "grad_norm": 0.54246586561203,
      "learning_rate": 5.3483262879299366e-06,
      "loss": 0.4323,
      "mean_token_accuracy": 0.8552845001220704,
      "num_tokens": 1062648229.0,
      "step": 31720
    },
    {
      "epoch": 1.8938037249283668,
      "grad_norm": 0.5128061175346375,
      "learning_rate": 5.3463826286288706e-06,
      "loss": 0.4036,
      "mean_token_accuracy": 0.863897168636322,
      "num_tokens": 1062815909.0,
      "step": 31725
    },
    {
      "epoch": 1.8941021967526266,
      "grad_norm": 0.4644211232662201,
      "learning_rate": 5.3444443652502245e-06,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8552725672721863,
      "num_tokens": 1062983589.0,
      "step": 31730
    },
    {
      "epoch": 1.8944006685768864,
      "grad_norm": 0.5072385668754578,
      "learning_rate": 5.342511498266098e-06,
      "loss": 0.4259,
      "mean_token_accuracy": 0.8572289228439331,
      "num_tokens": 1063151269.0,
      "step": 31735
    },
    {
      "epoch": 1.8946991404011462,
      "grad_norm": 0.47007593512535095,
      "learning_rate": 5.340584028147276e-06,
      "loss": 0.4058,
      "mean_token_accuracy": 0.8622569441795349,
      "num_tokens": 1063318949.0,
      "step": 31740
    },
    {
      "epoch": 1.894997612225406,
      "grad_norm": 0.5132487416267395,
      "learning_rate": 5.338661955363228e-06,
      "loss": 0.4432,
      "mean_token_accuracy": 0.8513360261917114,
      "num_tokens": 1063486629.0,
      "step": 31745
    },
    {
      "epoch": 1.8952960840496658,
      "grad_norm": 0.564643383026123,
      "learning_rate": 5.336745280382114e-06,
      "loss": 0.4611,
      "mean_token_accuracy": 0.8468448162078858,
      "num_tokens": 1063654309.0,
      "step": 31750
    },
    {
      "epoch": 1.8955945558739256,
      "grad_norm": 0.5347476005554199,
      "learning_rate": 5.334834003670775e-06,
      "loss": 0.4208,
      "mean_token_accuracy": 0.8596743464469909,
      "num_tokens": 1063821989.0,
      "step": 31755
    },
    {
      "epoch": 1.8958930276981851,
      "grad_norm": 0.5305520296096802,
      "learning_rate": 5.332928125694736e-06,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8477991104125977,
      "num_tokens": 1063989669.0,
      "step": 31760
    },
    {
      "epoch": 1.8961914995224451,
      "grad_norm": 0.47084951400756836,
      "learning_rate": 5.331027646918208e-06,
      "loss": 0.3921,
      "mean_token_accuracy": 0.8669390320777893,
      "num_tokens": 1064157349.0,
      "step": 31765
    },
    {
      "epoch": 1.8964899713467047,
      "grad_norm": 0.5179524421691895,
      "learning_rate": 5.329132567804089e-06,
      "loss": 0.4047,
      "mean_token_accuracy": 0.8634856224060059,
      "num_tokens": 1064325029.0,
      "step": 31770
    },
    {
      "epoch": 1.8967884431709647,
      "grad_norm": 0.5258401036262512,
      "learning_rate": 5.327242888813961e-06,
      "loss": 0.4379,
      "mean_token_accuracy": 0.8546284198760986,
      "num_tokens": 1064492709.0,
      "step": 31775
    },
    {
      "epoch": 1.8970869149952243,
      "grad_norm": 0.49087873101234436,
      "learning_rate": 5.325358610408088e-06,
      "loss": 0.4253,
      "mean_token_accuracy": 0.858195161819458,
      "num_tokens": 1064660389.0,
      "step": 31780
    },
    {
      "epoch": 1.8973853868194843,
      "grad_norm": 0.4951891303062439,
      "learning_rate": 5.323479733045425e-06,
      "loss": 0.4595,
      "mean_token_accuracy": 0.8468448162078858,
      "num_tokens": 1064828069.0,
      "step": 31785
    },
    {
      "epoch": 1.897683858643744,
      "grad_norm": 0.6353661417961121,
      "learning_rate": 5.3216062571836056e-06,
      "loss": 0.4642,
      "mean_token_accuracy": 0.8454431533813477,
      "num_tokens": 1064995749.0,
      "step": 31790
    },
    {
      "epoch": 1.897982330468004,
      "grad_norm": 0.4828473925590515,
      "learning_rate": 5.319738183278945e-06,
      "loss": 0.4168,
      "mean_token_accuracy": 0.8611058115959167,
      "num_tokens": 1065163429.0,
      "step": 31795
    },
    {
      "epoch": 1.8982808022922635,
      "grad_norm": 0.4772716164588928,
      "learning_rate": 5.3178755117864525e-06,
      "loss": 0.4037,
      "mean_token_accuracy": 0.8639210224151611,
      "num_tokens": 1065331109.0,
      "step": 31800
    },
    {
      "epoch": 1.8985792741165235,
      "grad_norm": 0.4815249741077423,
      "learning_rate": 5.316018243159813e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.858547055721283,
      "num_tokens": 1065498789.0,
      "step": 31805
    },
    {
      "epoch": 1.898877745940783,
      "grad_norm": 0.6053141951560974,
      "learning_rate": 5.3141663778514004e-06,
      "loss": 0.4133,
      "mean_token_accuracy": 0.8614815831184387,
      "num_tokens": 1065666469.0,
      "step": 31810
    },
    {
      "epoch": 1.899176217765043,
      "grad_norm": 0.5123937129974365,
      "learning_rate": 5.31231991631227e-06,
      "loss": 0.4425,
      "mean_token_accuracy": 0.8516640901565552,
      "num_tokens": 1065834149.0,
      "step": 31815
    },
    {
      "epoch": 1.8994746895893027,
      "grad_norm": 0.5880845785140991,
      "learning_rate": 5.310478858992162e-06,
      "loss": 0.4058,
      "mean_token_accuracy": 0.8639866471290588,
      "num_tokens": 1066001829.0,
      "step": 31820
    },
    {
      "epoch": 1.8997731614135627,
      "grad_norm": 0.5158835053443909,
      "learning_rate": 5.308643206339497e-06,
      "loss": 0.4392,
      "mean_token_accuracy": 0.8542884349822998,
      "num_tokens": 1066169509.0,
      "step": 31825
    },
    {
      "epoch": 1.9000716332378222,
      "grad_norm": 0.5261027216911316,
      "learning_rate": 5.306812958801389e-06,
      "loss": 0.4494,
      "mean_token_accuracy": 0.8503936648368835,
      "num_tokens": 1066337189.0,
      "step": 31830
    },
    {
      "epoch": 1.9003701050620823,
      "grad_norm": 0.5400513410568237,
      "learning_rate": 5.304988116823622e-06,
      "loss": 0.408,
      "mean_token_accuracy": 0.8631993293762207,
      "num_tokens": 1066504869.0,
      "step": 31835
    },
    {
      "epoch": 1.9006685768863418,
      "grad_norm": 0.4870010018348694,
      "learning_rate": 5.3031686808506745e-06,
      "loss": 0.4145,
      "mean_token_accuracy": 0.8592210531234741,
      "num_tokens": 1066672549.0,
      "step": 31840
    },
    {
      "epoch": 1.9009670487106018,
      "grad_norm": 0.4953862428665161,
      "learning_rate": 5.3013546513256966e-06,
      "loss": 0.4341,
      "mean_token_accuracy": 0.85431067943573,
      "num_tokens": 1066834087.0,
      "step": 31845
    },
    {
      "epoch": 1.9012655205348614,
      "grad_norm": 0.4556969702243805,
      "learning_rate": 5.299546028690539e-06,
      "loss": 0.3759,
      "mean_token_accuracy": 0.8721579313278198,
      "num_tokens": 1067001767.0,
      "step": 31850
    },
    {
      "epoch": 1.9015639923591214,
      "grad_norm": 0.4724929630756378,
      "learning_rate": 5.2977428133857145e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.859799599647522,
      "num_tokens": 1067169447.0,
      "step": 31855
    },
    {
      "epoch": 1.901862464183381,
      "grad_norm": 0.6419077515602112,
      "learning_rate": 5.295945005850437e-06,
      "loss": 0.423,
      "mean_token_accuracy": 0.8584874153137207,
      "num_tokens": 1067337127.0,
      "step": 31860
    },
    {
      "epoch": 1.902160936007641,
      "grad_norm": 0.4826868772506714,
      "learning_rate": 5.294152606522592e-06,
      "loss": 0.4126,
      "mean_token_accuracy": 0.8613205313682556,
      "num_tokens": 1067504807.0,
      "step": 31865
    },
    {
      "epoch": 1.9024594078319006,
      "grad_norm": 0.5089836716651917,
      "learning_rate": 5.292365615838752e-06,
      "loss": 0.4231,
      "mean_token_accuracy": 0.8571513652801513,
      "num_tokens": 1067672487.0,
      "step": 31870
    },
    {
      "epoch": 1.9027578796561606,
      "grad_norm": 0.5602514743804932,
      "learning_rate": 5.290584034234177e-06,
      "loss": 0.4452,
      "mean_token_accuracy": 0.850852906703949,
      "num_tokens": 1067840167.0,
      "step": 31875
    },
    {
      "epoch": 1.9030563514804202,
      "grad_norm": 0.5133985877037048,
      "learning_rate": 5.288807862142793e-06,
      "loss": 0.442,
      "mean_token_accuracy": 0.8537635684013367,
      "num_tokens": 1068007847.0,
      "step": 31880
    },
    {
      "epoch": 1.9033548233046802,
      "grad_norm": 0.5091225504875183,
      "learning_rate": 5.287037099997229e-06,
      "loss": 0.4267,
      "mean_token_accuracy": 0.8579506158828736,
      "num_tokens": 1068175527.0,
      "step": 31885
    },
    {
      "epoch": 1.9036532951289398,
      "grad_norm": 0.4815044105052948,
      "learning_rate": 5.285271748228784e-06,
      "loss": 0.4212,
      "mean_token_accuracy": 0.8574734568595886,
      "num_tokens": 1068343207.0,
      "step": 31890
    },
    {
      "epoch": 1.9039517669531996,
      "grad_norm": 0.5389668345451355,
      "learning_rate": 5.28351180726744e-06,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8600381731986999,
      "num_tokens": 1068510887.0,
      "step": 31895
    },
    {
      "epoch": 1.9042502387774594,
      "grad_norm": 0.5164958834648132,
      "learning_rate": 5.281757277541867e-06,
      "loss": 0.4123,
      "mean_token_accuracy": 0.8624955177307129,
      "num_tokens": 1068678567.0,
      "step": 31900
    },
    {
      "epoch": 1.9045487106017192,
      "grad_norm": 0.5868775248527527,
      "learning_rate": 5.280008159479409e-06,
      "loss": 0.4498,
      "mean_token_accuracy": 0.8485267758369446,
      "num_tokens": 1068846247.0,
      "step": 31905
    },
    {
      "epoch": 1.904847182425979,
      "grad_norm": 0.546893298625946,
      "learning_rate": 5.278264453506097e-06,
      "loss": 0.4205,
      "mean_token_accuracy": 0.859435760974884,
      "num_tokens": 1069013927.0,
      "step": 31910
    },
    {
      "epoch": 1.9051456542502387,
      "grad_norm": 0.5343213677406311,
      "learning_rate": 5.276526160046645e-06,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8547059535980225,
      "num_tokens": 1069181607.0,
      "step": 31915
    },
    {
      "epoch": 1.9054441260744985,
      "grad_norm": 0.5360780954360962,
      "learning_rate": 5.274793279524446e-06,
      "loss": 0.4455,
      "mean_token_accuracy": 0.8515746235847473,
      "num_tokens": 1069349287.0,
      "step": 31920
    },
    {
      "epoch": 1.9057425978987583,
      "grad_norm": 0.5584276914596558,
      "learning_rate": 5.273065812361573e-06,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8572766423225403,
      "num_tokens": 1069516967.0,
      "step": 31925
    },
    {
      "epoch": 1.9060410697230181,
      "grad_norm": 0.4997557997703552,
      "learning_rate": 5.271343758978782e-06,
      "loss": 0.4302,
      "mean_token_accuracy": 0.855350112915039,
      "num_tokens": 1069684647.0,
      "step": 31930
    },
    {
      "epoch": 1.906339541547278,
      "grad_norm": 0.4914553761482239,
      "learning_rate": 5.269627119795515e-06,
      "loss": 0.4389,
      "mean_token_accuracy": 0.8529166221618653,
      "num_tokens": 1069852327.0,
      "step": 31935
    },
    {
      "epoch": 1.9066380133715377,
      "grad_norm": 0.5483037829399109,
      "learning_rate": 5.2679158952298896e-06,
      "loss": 0.4457,
      "mean_token_accuracy": 0.851753544807434,
      "num_tokens": 1070020007.0,
      "step": 31940
    },
    {
      "epoch": 1.9069364851957975,
      "grad_norm": 0.5162957310676575,
      "learning_rate": 5.266210085698702e-06,
      "loss": 0.4217,
      "mean_token_accuracy": 0.8590838551521301,
      "num_tokens": 1070187687.0,
      "step": 31945
    },
    {
      "epoch": 1.9072349570200573,
      "grad_norm": 0.4837557375431061,
      "learning_rate": 5.264509691617439e-06,
      "loss": 0.4312,
      "mean_token_accuracy": 0.8554097533226013,
      "num_tokens": 1070355367.0,
      "step": 31950
    },
    {
      "epoch": 1.907533428844317,
      "grad_norm": 0.4790009558200836,
      "learning_rate": 5.262814713400257e-06,
      "loss": 0.4101,
      "mean_token_accuracy": 0.8619229316711425,
      "num_tokens": 1070523047.0,
      "step": 31955
    },
    {
      "epoch": 1.9078319006685769,
      "grad_norm": 0.4639718532562256,
      "learning_rate": 5.261125151460007e-06,
      "loss": 0.4425,
      "mean_token_accuracy": 0.8534056901931762,
      "num_tokens": 1070690727.0,
      "step": 31960
    },
    {
      "epoch": 1.9081303724928367,
      "grad_norm": 0.510834813117981,
      "learning_rate": 5.259441006208208e-06,
      "loss": 0.433,
      "mean_token_accuracy": 0.85520099401474,
      "num_tokens": 1070858407.0,
      "step": 31965
    },
    {
      "epoch": 1.9084288443170965,
      "grad_norm": 0.5318284630775452,
      "learning_rate": 5.257762278055063e-06,
      "loss": 0.4317,
      "mean_token_accuracy": 0.8547596454620361,
      "num_tokens": 1071026087.0,
      "step": 31970
    },
    {
      "epoch": 1.9087273161413563,
      "grad_norm": 0.47995248436927795,
      "learning_rate": 5.256088967409464e-06,
      "loss": 0.4152,
      "mean_token_accuracy": 0.8603781461715698,
      "num_tokens": 1071193767.0,
      "step": 31975
    },
    {
      "epoch": 1.909025787965616,
      "grad_norm": 0.47716400027275085,
      "learning_rate": 5.254421074678971e-06,
      "loss": 0.4011,
      "mean_token_accuracy": 0.8657461524009704,
      "num_tokens": 1071361447.0,
      "step": 31980
    },
    {
      "epoch": 1.9093242597898759,
      "grad_norm": 0.5465705394744873,
      "learning_rate": 5.252758600269831e-06,
      "loss": 0.4401,
      "mean_token_accuracy": 0.8544375658035278,
      "num_tokens": 1071529127.0,
      "step": 31985
    },
    {
      "epoch": 1.9096227316141356,
      "grad_norm": 0.5060298442840576,
      "learning_rate": 5.25110154458697e-06,
      "loss": 0.4327,
      "mean_token_accuracy": 0.8546045541763305,
      "num_tokens": 1071696807.0,
      "step": 31990
    },
    {
      "epoch": 1.9099212034383954,
      "grad_norm": 0.5056952834129333,
      "learning_rate": 5.249449908033996e-06,
      "loss": 0.4499,
      "mean_token_accuracy": 0.8500894665718078,
      "num_tokens": 1071864487.0,
      "step": 31995
    },
    {
      "epoch": 1.9102196752626552,
      "grad_norm": 0.5316416025161743,
      "learning_rate": 5.2478036910131975e-06,
      "loss": 0.4394,
      "mean_token_accuracy": 0.8522068500518799,
      "num_tokens": 1072032167.0,
      "step": 32000
    },
    {
      "epoch": 1.910518147086915,
      "grad_norm": 0.47567927837371826,
      "learning_rate": 5.246162893925537e-06,
      "loss": 0.4376,
      "mean_token_accuracy": 0.8528152227401733,
      "num_tokens": 1072199847.0,
      "step": 32005
    },
    {
      "epoch": 1.9108166189111748,
      "grad_norm": 0.4773764908313751,
      "learning_rate": 5.244527517170661e-06,
      "loss": 0.3972,
      "mean_token_accuracy": 0.8659668326377868,
      "num_tokens": 1072367527.0,
      "step": 32010
    },
    {
      "epoch": 1.9111150907354346,
      "grad_norm": 0.5136639475822449,
      "learning_rate": 5.2428975611468986e-06,
      "loss": 0.4329,
      "mean_token_accuracy": 0.8539246082305908,
      "num_tokens": 1072535207.0,
      "step": 32015
    },
    {
      "epoch": 1.9114135625596944,
      "grad_norm": 0.5270606875419617,
      "learning_rate": 5.241273026251253e-06,
      "loss": 0.4688,
      "mean_token_accuracy": 0.8458964586257934,
      "num_tokens": 1072702887.0,
      "step": 32020
    },
    {
      "epoch": 1.9117120343839542,
      "grad_norm": 0.48470228910446167,
      "learning_rate": 5.239653912879412e-06,
      "loss": 0.3988,
      "mean_token_accuracy": 0.8664738059043884,
      "num_tokens": 1072870567.0,
      "step": 32025
    },
    {
      "epoch": 1.912010506208214,
      "grad_norm": 0.5084724426269531,
      "learning_rate": 5.238040221425739e-06,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8526601433753968,
      "num_tokens": 1073038247.0,
      "step": 32030
    },
    {
      "epoch": 1.9123089780324736,
      "grad_norm": 0.4969509541988373,
      "learning_rate": 5.23643195228328e-06,
      "loss": 0.4188,
      "mean_token_accuracy": 0.8598532795906066,
      "num_tokens": 1073205927.0,
      "step": 32035
    },
    {
      "epoch": 1.9126074498567336,
      "grad_norm": 0.46751248836517334,
      "learning_rate": 5.234829105843755e-06,
      "loss": 0.4131,
      "mean_token_accuracy": 0.8601634263992309,
      "num_tokens": 1073373607.0,
      "step": 32040
    },
    {
      "epoch": 1.9129059216809932,
      "grad_norm": 0.5296719670295715,
      "learning_rate": 5.233231682497572e-06,
      "loss": 0.4247,
      "mean_token_accuracy": 0.8574018955230713,
      "num_tokens": 1073541287.0,
      "step": 32045
    },
    {
      "epoch": 1.9132043935052532,
      "grad_norm": 0.5356879234313965,
      "learning_rate": 5.231639682633807e-06,
      "loss": 0.4232,
      "mean_token_accuracy": 0.8574496030807495,
      "num_tokens": 1073708967.0,
      "step": 32050
    },
    {
      "epoch": 1.9135028653295127,
      "grad_norm": 0.570656418800354,
      "learning_rate": 5.2300531066402265e-06,
      "loss": 0.4588,
      "mean_token_accuracy": 0.845890486240387,
      "num_tokens": 1073876647.0,
      "step": 32055
    },
    {
      "epoch": 1.9138013371537728,
      "grad_norm": 0.5327288508415222,
      "learning_rate": 5.228471954903269e-06,
      "loss": 0.4463,
      "mean_token_accuracy": 0.8516640782356262,
      "num_tokens": 1074044327.0,
      "step": 32060
    },
    {
      "epoch": 1.9140998089780323,
      "grad_norm": 0.4700440466403961,
      "learning_rate": 5.22689622780805e-06,
      "loss": 0.4564,
      "mean_token_accuracy": 0.8459441781044006,
      "num_tokens": 1074212007.0,
      "step": 32065
    },
    {
      "epoch": 1.9143982808022924,
      "grad_norm": 0.4614141583442688,
      "learning_rate": 5.225325925738369e-06,
      "loss": 0.4181,
      "mean_token_accuracy": 0.8588274002075196,
      "num_tokens": 1074379687.0,
      "step": 32070
    },
    {
      "epoch": 1.914696752626552,
      "grad_norm": 0.48530611395835876,
      "learning_rate": 5.223761049076703e-06,
      "loss": 0.4126,
      "mean_token_accuracy": 0.8612131834030151,
      "num_tokens": 1074547367.0,
      "step": 32075
    },
    {
      "epoch": 1.914995224450812,
      "grad_norm": 0.5136609077453613,
      "learning_rate": 5.222201598204203e-06,
      "loss": 0.4595,
      "mean_token_accuracy": 0.846767270565033,
      "num_tokens": 1074715047.0,
      "step": 32080
    },
    {
      "epoch": 1.9152936962750715,
      "grad_norm": 0.5924073457717896,
      "learning_rate": 5.220647573500711e-06,
      "loss": 0.4637,
      "mean_token_accuracy": 0.8435643553733826,
      "num_tokens": 1074882727.0,
      "step": 32085
    },
    {
      "epoch": 1.9155921680993315,
      "grad_norm": 0.4726608991622925,
      "learning_rate": 5.219098975344729e-06,
      "loss": 0.4426,
      "mean_token_accuracy": 0.8532506346702575,
      "num_tokens": 1075050407.0,
      "step": 32090
    },
    {
      "epoch": 1.915890639923591,
      "grad_norm": 0.500882625579834,
      "learning_rate": 5.217555804113447e-06,
      "loss": 0.425,
      "mean_token_accuracy": 0.858570909500122,
      "num_tokens": 1075218087.0,
      "step": 32095
    },
    {
      "epoch": 1.9161891117478511,
      "grad_norm": 0.47746843099594116,
      "learning_rate": 5.2160180601827385e-06,
      "loss": 0.4118,
      "mean_token_accuracy": 0.8609924793243409,
      "num_tokens": 1075385767.0,
      "step": 32100
    },
    {
      "epoch": 1.9164875835721107,
      "grad_norm": 0.5892366766929626,
      "learning_rate": 5.214485743927147e-06,
      "loss": 0.4262,
      "mean_token_accuracy": 0.8567994832992554,
      "num_tokens": 1075553447.0,
      "step": 32105
    },
    {
      "epoch": 1.9167860553963707,
      "grad_norm": 0.4851658046245575,
      "learning_rate": 5.212958855719894e-06,
      "loss": 0.399,
      "mean_token_accuracy": 0.866139817237854,
      "num_tokens": 1075721127.0,
      "step": 32110
    },
    {
      "epoch": 1.9170845272206303,
      "grad_norm": 0.507503092288971,
      "learning_rate": 5.211437395932885e-06,
      "loss": 0.4205,
      "mean_token_accuracy": 0.8584158420562744,
      "num_tokens": 1075888807.0,
      "step": 32115
    },
    {
      "epoch": 1.9173829990448903,
      "grad_norm": 0.5060835480690002,
      "learning_rate": 5.209921364936699e-06,
      "loss": 0.4262,
      "mean_token_accuracy": 0.8586245894432067,
      "num_tokens": 1076056487.0,
      "step": 32120
    },
    {
      "epoch": 1.9176814708691499,
      "grad_norm": 0.4905909597873688,
      "learning_rate": 5.20841076310059e-06,
      "loss": 0.4297,
      "mean_token_accuracy": 0.8556363940238952,
      "num_tokens": 1076224167.0,
      "step": 32125
    },
    {
      "epoch": 1.9179799426934099,
      "grad_norm": 0.4943788945674896,
      "learning_rate": 5.206905590792496e-06,
      "loss": 0.3928,
      "mean_token_accuracy": 0.8674698948860169,
      "num_tokens": 1076391847.0,
      "step": 32130
    },
    {
      "epoch": 1.9182784145176695,
      "grad_norm": 0.5155355334281921,
      "learning_rate": 5.205405848379028e-06,
      "loss": 0.4182,
      "mean_token_accuracy": 0.8585709214210511,
      "num_tokens": 1076559527.0,
      "step": 32135
    },
    {
      "epoch": 1.9185768863419295,
      "grad_norm": 0.49372220039367676,
      "learning_rate": 5.203911536225477e-06,
      "loss": 0.3931,
      "mean_token_accuracy": 0.8676786422729492,
      "num_tokens": 1076727207.0,
      "step": 32140
    },
    {
      "epoch": 1.918875358166189,
      "grad_norm": 0.47030818462371826,
      "learning_rate": 5.202422654695812e-06,
      "loss": 0.3998,
      "mean_token_accuracy": 0.863998556137085,
      "num_tokens": 1076894887.0,
      "step": 32145
    },
    {
      "epoch": 1.919173829990449,
      "grad_norm": 0.474126398563385,
      "learning_rate": 5.200939204152674e-06,
      "loss": 0.3981,
      "mean_token_accuracy": 0.8657998442649841,
      "num_tokens": 1077062567.0,
      "step": 32150
    },
    {
      "epoch": 1.9194723018147086,
      "grad_norm": 0.5253416895866394,
      "learning_rate": 5.199461184957383e-06,
      "loss": 0.4365,
      "mean_token_accuracy": 0.8545150876045227,
      "num_tokens": 1077230247.0,
      "step": 32155
    },
    {
      "epoch": 1.9197707736389686,
      "grad_norm": 0.4943583309650421,
      "learning_rate": 5.197988597469946e-06,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8502147197723389,
      "num_tokens": 1077397927.0,
      "step": 32160
    },
    {
      "epoch": 1.9200692454632282,
      "grad_norm": 0.49374493956565857,
      "learning_rate": 5.196521442049033e-06,
      "loss": 0.4134,
      "mean_token_accuracy": 0.8596027612686157,
      "num_tokens": 1077565607.0,
      "step": 32165
    },
    {
      "epoch": 1.920367717287488,
      "grad_norm": 0.5343686938285828,
      "learning_rate": 5.195059719052e-06,
      "loss": 0.413,
      "mean_token_accuracy": 0.859811520576477,
      "num_tokens": 1077733287.0,
      "step": 32170
    },
    {
      "epoch": 1.9206661891117478,
      "grad_norm": 0.506992518901825,
      "learning_rate": 5.1936034288348686e-06,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8493021607398987,
      "num_tokens": 1077900967.0,
      "step": 32175
    },
    {
      "epoch": 1.9209646609360076,
      "grad_norm": 0.455807089805603,
      "learning_rate": 5.192152571752357e-06,
      "loss": 0.4073,
      "mean_token_accuracy": 0.8615710377693176,
      "num_tokens": 1078068647.0,
      "step": 32180
    },
    {
      "epoch": 1.9212631327602674,
      "grad_norm": 0.5381808280944824,
      "learning_rate": 5.190707148157839e-06,
      "loss": 0.4379,
      "mean_token_accuracy": 0.8528032898902893,
      "num_tokens": 1078236327.0,
      "step": 32185
    },
    {
      "epoch": 1.9215616045845272,
      "grad_norm": 0.4487593472003937,
      "learning_rate": 5.189267158403377e-06,
      "loss": 0.3877,
      "mean_token_accuracy": 0.8674579501152039,
      "num_tokens": 1078404007.0,
      "step": 32190
    },
    {
      "epoch": 1.921860076408787,
      "grad_norm": 0.49918490648269653,
      "learning_rate": 5.187832602839708e-06,
      "loss": 0.3994,
      "mean_token_accuracy": 0.8661048412322998,
      "num_tokens": 1078568782.0,
      "step": 32195
    },
    {
      "epoch": 1.9221585482330468,
      "grad_norm": 0.4842585027217865,
      "learning_rate": 5.186403481816245e-06,
      "loss": 0.4226,
      "mean_token_accuracy": 0.8582070827484131,
      "num_tokens": 1078736462.0,
      "step": 32200
    },
    {
      "epoch": 1.9224570200573066,
      "grad_norm": 0.5461395978927612,
      "learning_rate": 5.1849797956810755e-06,
      "loss": 0.4263,
      "mean_token_accuracy": 0.8561254858970642,
      "num_tokens": 1078904142.0,
      "step": 32205
    },
    {
      "epoch": 1.9227554918815664,
      "grad_norm": 0.46837109327316284,
      "learning_rate": 5.183561544780965e-06,
      "loss": 0.4019,
      "mean_token_accuracy": 0.8645532608032227,
      "num_tokens": 1079071822.0,
      "step": 32210
    },
    {
      "epoch": 1.9230539637058262,
      "grad_norm": 0.4826766550540924,
      "learning_rate": 5.182148729461353e-06,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8577597498893738,
      "num_tokens": 1079239502.0,
      "step": 32215
    },
    {
      "epoch": 1.923352435530086,
      "grad_norm": 0.5636767148971558,
      "learning_rate": 5.180741350066359e-06,
      "loss": 0.4307,
      "mean_token_accuracy": 0.8561314582824707,
      "num_tokens": 1079407182.0,
      "step": 32220
    },
    {
      "epoch": 1.9236509073543457,
      "grad_norm": 0.551056981086731,
      "learning_rate": 5.179339406938776e-06,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8547715663909912,
      "num_tokens": 1079574862.0,
      "step": 32225
    },
    {
      "epoch": 1.9239493791786055,
      "grad_norm": 0.4722616672515869,
      "learning_rate": 5.1779429004200735e-06,
      "loss": 0.4177,
      "mean_token_accuracy": 0.8593761205673218,
      "num_tokens": 1079742542.0,
      "step": 32230
    },
    {
      "epoch": 1.9242478510028653,
      "grad_norm": 0.5256747603416443,
      "learning_rate": 5.1765518308503945e-06,
      "loss": 0.442,
      "mean_token_accuracy": 0.851228678226471,
      "num_tokens": 1079910222.0,
      "step": 32235
    },
    {
      "epoch": 1.9245463228271251,
      "grad_norm": 0.47784456610679626,
      "learning_rate": 5.17516619856856e-06,
      "loss": 0.4031,
      "mean_token_accuracy": 0.8624074697494507,
      "num_tokens": 1080074740.0,
      "step": 32240
    },
    {
      "epoch": 1.924844794651385,
      "grad_norm": 0.5293857455253601,
      "learning_rate": 5.1737860039120695e-06,
      "loss": 0.4571,
      "mean_token_accuracy": 0.8470476031303406,
      "num_tokens": 1080242420.0,
      "step": 32245
    },
    {
      "epoch": 1.9251432664756447,
      "grad_norm": 0.5418432950973511,
      "learning_rate": 5.172411247217088e-06,
      "loss": 0.4235,
      "mean_token_accuracy": 0.8572467923164367,
      "num_tokens": 1080410100.0,
      "step": 32250
    },
    {
      "epoch": 1.9254417382999045,
      "grad_norm": 0.593043863773346,
      "learning_rate": 5.171041928818472e-06,
      "loss": 0.4409,
      "mean_token_accuracy": 0.8518787980079651,
      "num_tokens": 1080577780.0,
      "step": 32255
    },
    {
      "epoch": 1.9257402101241643,
      "grad_norm": 0.5288111567497253,
      "learning_rate": 5.169678049049736e-06,
      "loss": 0.4429,
      "mean_token_accuracy": 0.8513718366622924,
      "num_tokens": 1080745460.0,
      "step": 32260
    },
    {
      "epoch": 1.926038681948424,
      "grad_norm": 0.5180296897888184,
      "learning_rate": 5.168319608243083e-06,
      "loss": 0.3802,
      "mean_token_accuracy": 0.8718835830688476,
      "num_tokens": 1080913140.0,
      "step": 32265
    },
    {
      "epoch": 1.9263371537726839,
      "grad_norm": 0.49369266629219055,
      "learning_rate": 5.166966606729387e-06,
      "loss": 0.4178,
      "mean_token_accuracy": 0.8581056714057922,
      "num_tokens": 1081080820.0,
      "step": 32270
    },
    {
      "epoch": 1.9266356255969437,
      "grad_norm": 0.5329632759094238,
      "learning_rate": 5.165619044838193e-06,
      "loss": 0.4047,
      "mean_token_accuracy": 0.8635154604911804,
      "num_tokens": 1081248500.0,
      "step": 32275
    },
    {
      "epoch": 1.9269340974212035,
      "grad_norm": 0.4576363265514374,
      "learning_rate": 5.164276922897726e-06,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8699689865112304,
      "num_tokens": 1081416180.0,
      "step": 32280
    },
    {
      "epoch": 1.9272325692454633,
      "grad_norm": 0.5308344960212708,
      "learning_rate": 5.162940241234883e-06,
      "loss": 0.3867,
      "mean_token_accuracy": 0.8699570536613465,
      "num_tokens": 1081583860.0,
      "step": 32285
    },
    {
      "epoch": 1.927531041069723,
      "grad_norm": 0.47972264885902405,
      "learning_rate": 5.161609000175242e-06,
      "loss": 0.4052,
      "mean_token_accuracy": 0.862656569480896,
      "num_tokens": 1081751540.0,
      "step": 32290
    },
    {
      "epoch": 1.9278295128939829,
      "grad_norm": 0.5196604132652283,
      "learning_rate": 5.160283200043046e-06,
      "loss": 0.4096,
      "mean_token_accuracy": 0.8632172226905823,
      "num_tokens": 1081919220.0,
      "step": 32295
    },
    {
      "epoch": 1.9281279847182426,
      "grad_norm": 0.517647385597229,
      "learning_rate": 5.158962841161222e-06,
      "loss": 0.4347,
      "mean_token_accuracy": 0.8538410902023316,
      "num_tokens": 1082086900.0,
      "step": 32300
    },
    {
      "epoch": 1.9284264565425024,
      "grad_norm": 0.49499738216400146,
      "learning_rate": 5.157647923851366e-06,
      "loss": 0.4092,
      "mean_token_accuracy": 0.8620660901069641,
      "num_tokens": 1082254580.0,
      "step": 32305
    },
    {
      "epoch": 1.928724928366762,
      "grad_norm": 0.48807811737060547,
      "learning_rate": 5.156338448433749e-06,
      "loss": 0.4236,
      "mean_token_accuracy": 0.8564773917198181,
      "num_tokens": 1082422260.0,
      "step": 32310
    },
    {
      "epoch": 1.929023400191022,
      "grad_norm": 0.4935206174850464,
      "learning_rate": 5.155034415227325e-06,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8550996065139771,
      "num_tokens": 1082589940.0,
      "step": 32315
    },
    {
      "epoch": 1.9293218720152816,
      "grad_norm": 0.5093145370483398,
      "learning_rate": 5.153735824549705e-06,
      "loss": 0.4024,
      "mean_token_accuracy": 0.8654777407646179,
      "num_tokens": 1082757620.0,
      "step": 32320
    },
    {
      "epoch": 1.9296203438395416,
      "grad_norm": 0.5231690406799316,
      "learning_rate": 5.15244267671719e-06,
      "loss": 0.4831,
      "mean_token_accuracy": 0.8399558663368225,
      "num_tokens": 1082925300.0,
      "step": 32325
    },
    {
      "epoch": 1.9299188156638012,
      "grad_norm": 0.6406030058860779,
      "learning_rate": 5.151154972044753e-06,
      "loss": 0.4368,
      "mean_token_accuracy": 0.8529205799102784,
      "num_tokens": 1083084912.0,
      "step": 32330
    },
    {
      "epoch": 1.9302172874880612,
      "grad_norm": 0.5277084112167358,
      "learning_rate": 5.149872710846032e-06,
      "loss": 0.3843,
      "mean_token_accuracy": 0.8710843324661255,
      "num_tokens": 1083252592.0,
      "step": 32335
    },
    {
      "epoch": 1.9305157593123208,
      "grad_norm": 0.5303910970687866,
      "learning_rate": 5.148595893433348e-06,
      "loss": 0.3972,
      "mean_token_accuracy": 0.8653465390205384,
      "num_tokens": 1083420272.0,
      "step": 32340
    },
    {
      "epoch": 1.9308142311365808,
      "grad_norm": 0.5570942759513855,
      "learning_rate": 5.147324520117695e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8592687487602234,
      "num_tokens": 1083587952.0,
      "step": 32345
    },
    {
      "epoch": 1.9311127029608404,
      "grad_norm": 0.4771854877471924,
      "learning_rate": 5.146058591208738e-06,
      "loss": 0.4309,
      "mean_token_accuracy": 0.8555171132087708,
      "num_tokens": 1083755632.0,
      "step": 32350
    },
    {
      "epoch": 1.9314111747851004,
      "grad_norm": 0.5093052387237549,
      "learning_rate": 5.144798107014818e-06,
      "loss": 0.4331,
      "mean_token_accuracy": 0.8540677547454834,
      "num_tokens": 1083923312.0,
      "step": 32355
    },
    {
      "epoch": 1.93170964660936,
      "grad_norm": 0.5296303629875183,
      "learning_rate": 5.143543067842946e-06,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8556423664093018,
      "num_tokens": 1084090992.0,
      "step": 32360
    },
    {
      "epoch": 1.93200811843362,
      "grad_norm": 0.5807662606239319,
      "learning_rate": 5.142293473998816e-06,
      "loss": 0.4529,
      "mean_token_accuracy": 0.8488369345664978,
      "num_tokens": 1084258672.0,
      "step": 32365
    },
    {
      "epoch": 1.9323065902578795,
      "grad_norm": 0.5080680847167969,
      "learning_rate": 5.141049325786785e-06,
      "loss": 0.4131,
      "mean_token_accuracy": 0.8619766116142273,
      "num_tokens": 1084426352.0,
      "step": 32370
    },
    {
      "epoch": 1.9326050620821396,
      "grad_norm": 0.5295805335044861,
      "learning_rate": 5.139810623509891e-06,
      "loss": 0.4124,
      "mean_token_accuracy": 0.8604735612869263,
      "num_tokens": 1084594032.0,
      "step": 32375
    },
    {
      "epoch": 1.9329035339063991,
      "grad_norm": 0.4883458614349365,
      "learning_rate": 5.138577367469842e-06,
      "loss": 0.4146,
      "mean_token_accuracy": 0.8606703996658325,
      "num_tokens": 1084761712.0,
      "step": 32380
    },
    {
      "epoch": 1.9332020057306591,
      "grad_norm": 0.5473781824111938,
      "learning_rate": 5.137349557967016e-06,
      "loss": 0.4292,
      "mean_token_accuracy": 0.8561433792114258,
      "num_tokens": 1084929392.0,
      "step": 32385
    },
    {
      "epoch": 1.9335004775549187,
      "grad_norm": 0.50545734167099,
      "learning_rate": 5.136127195300477e-06,
      "loss": 0.399,
      "mean_token_accuracy": 0.8654419779777527,
      "num_tokens": 1085097072.0,
      "step": 32390
    },
    {
      "epoch": 1.9337989493791787,
      "grad_norm": 0.5239991545677185,
      "learning_rate": 5.134910279767947e-06,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8564117908477783,
      "num_tokens": 1085264752.0,
      "step": 32395
    },
    {
      "epoch": 1.9340974212034383,
      "grad_norm": 0.5179317593574524,
      "learning_rate": 5.133698811665831e-06,
      "loss": 0.4254,
      "mean_token_accuracy": 0.8578372955322265,
      "num_tokens": 1085432432.0,
      "step": 32400
    },
    {
      "epoch": 1.9343958930276983,
      "grad_norm": 0.47603145241737366,
      "learning_rate": 5.1324927912892035e-06,
      "loss": 0.4208,
      "mean_token_accuracy": 0.8594954133033752,
      "num_tokens": 1085600112.0,
      "step": 32405
    },
    {
      "epoch": 1.934694364851958,
      "grad_norm": 0.4627433717250824,
      "learning_rate": 5.1312922189318115e-06,
      "loss": 0.4162,
      "mean_token_accuracy": 0.8601634263992309,
      "num_tokens": 1085767792.0,
      "step": 32410
    },
    {
      "epoch": 1.934992836676218,
      "grad_norm": 0.5203522443771362,
      "learning_rate": 5.130097094886084e-06,
      "loss": 0.4203,
      "mean_token_accuracy": 0.8578671097755433,
      "num_tokens": 1085935472.0,
      "step": 32415
    },
    {
      "epoch": 1.9352913085004775,
      "grad_norm": 0.5076498985290527,
      "learning_rate": 5.128907419443105e-06,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8469998836517334,
      "num_tokens": 1086103152.0,
      "step": 32420
    },
    {
      "epoch": 1.9355897803247375,
      "grad_norm": 0.5263394713401794,
      "learning_rate": 5.127723192892649e-06,
      "loss": 0.421,
      "mean_token_accuracy": 0.8593641757965088,
      "num_tokens": 1086270832.0,
      "step": 32425
    },
    {
      "epoch": 1.935888252148997,
      "grad_norm": 0.4940546751022339,
      "learning_rate": 5.1265444155231545e-06,
      "loss": 0.4358,
      "mean_token_accuracy": 0.853984260559082,
      "num_tokens": 1086438512.0,
      "step": 32430
    },
    {
      "epoch": 1.936186723973257,
      "grad_norm": 0.5298051238059998,
      "learning_rate": 5.125371087621734e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.8595968008041381,
      "num_tokens": 1086606192.0,
      "step": 32435
    },
    {
      "epoch": 1.9364851957975167,
      "grad_norm": 0.5094213485717773,
      "learning_rate": 5.124203209474177e-06,
      "loss": 0.4104,
      "mean_token_accuracy": 0.8609089851379395,
      "num_tokens": 1086773872.0,
      "step": 32440
    },
    {
      "epoch": 1.9367836676217765,
      "grad_norm": 0.5221966505050659,
      "learning_rate": 5.123040781364933e-06,
      "loss": 0.3862,
      "mean_token_accuracy": 0.8694321751594544,
      "num_tokens": 1086941552.0,
      "step": 32445
    },
    {
      "epoch": 1.9370821394460362,
      "grad_norm": 0.6008448004722595,
      "learning_rate": 5.121883803577139e-06,
      "loss": 0.393,
      "mean_token_accuracy": 0.8662648558616638,
      "num_tokens": 1087102373.0,
      "step": 32450
    },
    {
      "epoch": 1.937380611270296,
      "grad_norm": 0.5770822167396545,
      "learning_rate": 5.120732276392598e-06,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8495765328407288,
      "num_tokens": 1087270053.0,
      "step": 32455
    },
    {
      "epoch": 1.9376790830945558,
      "grad_norm": 0.49772778153419495,
      "learning_rate": 5.119586200091786e-06,
      "loss": 0.4313,
      "mean_token_accuracy": 0.8565728187561035,
      "num_tokens": 1087437733.0,
      "step": 32460
    },
    {
      "epoch": 1.9379775549188156,
      "grad_norm": 0.4981571435928345,
      "learning_rate": 5.1184455749538475e-06,
      "loss": 0.4155,
      "mean_token_accuracy": 0.8609507322311402,
      "num_tokens": 1087605413.0,
      "step": 32465
    },
    {
      "epoch": 1.9382760267430754,
      "grad_norm": 0.49466049671173096,
      "learning_rate": 5.117310401256608e-06,
      "loss": 0.3992,
      "mean_token_accuracy": 0.865877366065979,
      "num_tokens": 1087773093.0,
      "step": 32470
    },
    {
      "epoch": 1.9385744985673352,
      "grad_norm": 0.5097780823707581,
      "learning_rate": 5.116180679276559e-06,
      "loss": 0.4184,
      "mean_token_accuracy": 0.8581295371055603,
      "num_tokens": 1087940773.0,
      "step": 32475
    },
    {
      "epoch": 1.938872970391595,
      "grad_norm": 0.5160642266273499,
      "learning_rate": 5.1150564092888595e-06,
      "loss": 0.4102,
      "mean_token_accuracy": 0.8621078252792358,
      "num_tokens": 1088108453.0,
      "step": 32480
    },
    {
      "epoch": 1.9391714422158548,
      "grad_norm": 0.4997353255748749,
      "learning_rate": 5.113937591567351e-06,
      "loss": 0.4079,
      "mean_token_accuracy": 0.8624478101730346,
      "num_tokens": 1088276133.0,
      "step": 32485
    },
    {
      "epoch": 1.9394699140401146,
      "grad_norm": 0.5170464515686035,
      "learning_rate": 5.112824226384544e-06,
      "loss": 0.3919,
      "mean_token_accuracy": 0.8661696314811707,
      "num_tokens": 1088443813.0,
      "step": 32490
    },
    {
      "epoch": 1.9397683858643744,
      "grad_norm": 0.4755837619304657,
      "learning_rate": 5.111716314011615e-06,
      "loss": 0.4015,
      "mean_token_accuracy": 0.8652153134346008,
      "num_tokens": 1088611493.0,
      "step": 32495
    },
    {
      "epoch": 1.9400668576886342,
      "grad_norm": 0.5550814867019653,
      "learning_rate": 5.110613854718418e-06,
      "loss": 0.4629,
      "mean_token_accuracy": 0.845914363861084,
      "num_tokens": 1088779173.0,
      "step": 32500
    },
    {
      "epoch": 1.940365329512894,
      "grad_norm": 0.5384535789489746,
      "learning_rate": 5.109516848773475e-06,
      "loss": 0.4052,
      "mean_token_accuracy": 0.864147686958313,
      "num_tokens": 1088946853.0,
      "step": 32505
    },
    {
      "epoch": 1.9406638013371538,
      "grad_norm": 0.509691596031189,
      "learning_rate": 5.108425296443988e-06,
      "loss": 0.4459,
      "mean_token_accuracy": 0.8503161191940307,
      "num_tokens": 1089114533.0,
      "step": 32510
    },
    {
      "epoch": 1.9409622731614136,
      "grad_norm": 0.513257622718811,
      "learning_rate": 5.107339197995821e-06,
      "loss": 0.4403,
      "mean_token_accuracy": 0.8517058372497559,
      "num_tokens": 1089282213.0,
      "step": 32515
    },
    {
      "epoch": 1.9412607449856734,
      "grad_norm": 0.5266803503036499,
      "learning_rate": 5.10625855369351e-06,
      "loss": 0.4619,
      "mean_token_accuracy": 0.8452702045440674,
      "num_tokens": 1089449893.0,
      "step": 32520
    },
    {
      "epoch": 1.9415592168099332,
      "grad_norm": 0.4870196282863617,
      "learning_rate": 5.105183363800273e-06,
      "loss": 0.376,
      "mean_token_accuracy": 0.872969114780426,
      "num_tokens": 1089617573.0,
      "step": 32525
    },
    {
      "epoch": 1.941857688634193,
      "grad_norm": 0.5021434426307678,
      "learning_rate": 5.104113628577987e-06,
      "loss": 0.4144,
      "mean_token_accuracy": 0.8604377985000611,
      "num_tokens": 1089785253.0,
      "step": 32530
    },
    {
      "epoch": 1.9421561604584527,
      "grad_norm": 0.4860079288482666,
      "learning_rate": 5.10304934828721e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8605749845504761,
      "num_tokens": 1089952933.0,
      "step": 32535
    },
    {
      "epoch": 1.9424546322827125,
      "grad_norm": 0.5217337012290955,
      "learning_rate": 5.10199052318716e-06,
      "loss": 0.4469,
      "mean_token_accuracy": 0.851616358757019,
      "num_tokens": 1090120613.0,
      "step": 32540
    },
    {
      "epoch": 1.9427531041069723,
      "grad_norm": 0.5213215947151184,
      "learning_rate": 5.1009371535357426e-06,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8478766441345215,
      "num_tokens": 1090288293.0,
      "step": 32545
    },
    {
      "epoch": 1.9430515759312321,
      "grad_norm": 0.4800375699996948,
      "learning_rate": 5.099889239589518e-06,
      "loss": 0.4028,
      "mean_token_accuracy": 0.8645174741744995,
      "num_tokens": 1090455973.0,
      "step": 32550
    },
    {
      "epoch": 1.943350047755492,
      "grad_norm": 0.5107091665267944,
      "learning_rate": 5.0988467816037265e-06,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8474352836608887,
      "num_tokens": 1090623653.0,
      "step": 32555
    },
    {
      "epoch": 1.9436485195797517,
      "grad_norm": 0.5225998163223267,
      "learning_rate": 5.097809779832283e-06,
      "loss": 0.4317,
      "mean_token_accuracy": 0.856715977191925,
      "num_tokens": 1090791333.0,
      "step": 32560
    },
    {
      "epoch": 1.9439469914040115,
      "grad_norm": 0.5601538419723511,
      "learning_rate": 5.096778234527763e-06,
      "loss": 0.4564,
      "mean_token_accuracy": 0.8473517775535584,
      "num_tokens": 1090959013.0,
      "step": 32565
    },
    {
      "epoch": 1.9442454632282713,
      "grad_norm": 0.5448104739189148,
      "learning_rate": 5.095752145941418e-06,
      "loss": 0.4373,
      "mean_token_accuracy": 0.8525348901748657,
      "num_tokens": 1091126693.0,
      "step": 32570
    },
    {
      "epoch": 1.944543935052531,
      "grad_norm": 0.5191338658332825,
      "learning_rate": 5.094731514323175e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8580042958259583,
      "num_tokens": 1091294373.0,
      "step": 32575
    },
    {
      "epoch": 1.9448424068767909,
      "grad_norm": 0.5559952855110168,
      "learning_rate": 5.093716339921625e-06,
      "loss": 0.4302,
      "mean_token_accuracy": 0.8550101399421692,
      "num_tokens": 1091462053.0,
      "step": 32580
    },
    {
      "epoch": 1.9451408787010505,
      "grad_norm": 0.5127229690551758,
      "learning_rate": 5.092706622984035e-06,
      "loss": 0.4206,
      "mean_token_accuracy": 0.8593582391738892,
      "num_tokens": 1091629733.0,
      "step": 32585
    },
    {
      "epoch": 1.9454393505253105,
      "grad_norm": 0.5035390257835388,
      "learning_rate": 5.091702363756335e-06,
      "loss": 0.4309,
      "mean_token_accuracy": 0.8548431396484375,
      "num_tokens": 1091797413.0,
      "step": 32590
    },
    {
      "epoch": 1.94573782234957,
      "grad_norm": 0.4593779742717743,
      "learning_rate": 5.090703562483136e-06,
      "loss": 0.4675,
      "mean_token_accuracy": 0.8449481129646301,
      "num_tokens": 1091965093.0,
      "step": 32595
    },
    {
      "epoch": 1.94603629417383,
      "grad_norm": 0.5089210867881775,
      "learning_rate": 5.089710219407714e-06,
      "loss": 0.4299,
      "mean_token_accuracy": 0.8553322196006775,
      "num_tokens": 1092132773.0,
      "step": 32600
    },
    {
      "epoch": 1.9463347659980896,
      "grad_norm": 0.4984305500984192,
      "learning_rate": 5.088722334772011e-06,
      "loss": 0.4529,
      "mean_token_accuracy": 0.8480973362922668,
      "num_tokens": 1092300453.0,
      "step": 32605
    },
    {
      "epoch": 1.9466332378223496,
      "grad_norm": 0.4983685314655304,
      "learning_rate": 5.08773990881665e-06,
      "loss": 0.4046,
      "mean_token_accuracy": 0.8634021162986756,
      "num_tokens": 1092468133.0,
      "step": 32610
    },
    {
      "epoch": 1.9469317096466092,
      "grad_norm": 0.5268809795379639,
      "learning_rate": 5.086762941780916e-06,
      "loss": 0.4,
      "mean_token_accuracy": 0.8654538989067078,
      "num_tokens": 1092635813.0,
      "step": 32615
    },
    {
      "epoch": 1.9472301814708692,
      "grad_norm": 0.5293128490447998,
      "learning_rate": 5.085791433902771e-06,
      "loss": 0.422,
      "mean_token_accuracy": 0.8590182542800904,
      "num_tokens": 1092803493.0,
      "step": 32620
    },
    {
      "epoch": 1.9475286532951288,
      "grad_norm": 0.5257720947265625,
      "learning_rate": 5.084825385418841e-06,
      "loss": 0.4396,
      "mean_token_accuracy": 0.8520100235939025,
      "num_tokens": 1092971173.0,
      "step": 32625
    },
    {
      "epoch": 1.9478271251193888,
      "grad_norm": 0.5668218731880188,
      "learning_rate": 5.083864796564422e-06,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8553143382072449,
      "num_tokens": 1093138853.0,
      "step": 32630
    },
    {
      "epoch": 1.9481255969436484,
      "grad_norm": 0.5278427004814148,
      "learning_rate": 5.082909667573488e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8592031598091125,
      "num_tokens": 1093306533.0,
      "step": 32635
    },
    {
      "epoch": 1.9484240687679084,
      "grad_norm": 0.47221750020980835,
      "learning_rate": 5.081959998678678e-06,
      "loss": 0.4394,
      "mean_token_accuracy": 0.85449720621109,
      "num_tokens": 1093474213.0,
      "step": 32640
    },
    {
      "epoch": 1.948722540592168,
      "grad_norm": 0.518292248249054,
      "learning_rate": 5.081015790111298e-06,
      "loss": 0.4305,
      "mean_token_accuracy": 0.855463445186615,
      "num_tokens": 1093641893.0,
      "step": 32645
    },
    {
      "epoch": 1.949021012416428,
      "grad_norm": 0.4561622440814972,
      "learning_rate": 5.080077042101328e-06,
      "loss": 0.4478,
      "mean_token_accuracy": 0.8493558287620544,
      "num_tokens": 1093809573.0,
      "step": 32650
    },
    {
      "epoch": 1.9493194842406876,
      "grad_norm": 0.48174574971199036,
      "learning_rate": 5.079143754877423e-06,
      "loss": 0.4469,
      "mean_token_accuracy": 0.8514851450920105,
      "num_tokens": 1093977253.0,
      "step": 32655
    },
    {
      "epoch": 1.9496179560649476,
      "grad_norm": 0.4775007963180542,
      "learning_rate": 5.078215928666896e-06,
      "loss": 0.4152,
      "mean_token_accuracy": 0.8596803069114685,
      "num_tokens": 1094144933.0,
      "step": 32660
    },
    {
      "epoch": 1.9499164278892072,
      "grad_norm": 0.5278202891349792,
      "learning_rate": 5.077293563695738e-06,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8539902091026306,
      "num_tokens": 1094312613.0,
      "step": 32665
    },
    {
      "epoch": 1.9502148997134672,
      "grad_norm": 0.5718221068382263,
      "learning_rate": 5.076376660188608e-06,
      "loss": 0.4381,
      "mean_token_accuracy": 0.8539961695671081,
      "num_tokens": 1094480293.0,
      "step": 32670
    },
    {
      "epoch": 1.9505133715377267,
      "grad_norm": 0.4924035370349884,
      "learning_rate": 5.075465218368835e-06,
      "loss": 0.3835,
      "mean_token_accuracy": 0.8699451208114624,
      "num_tokens": 1094647973.0,
      "step": 32675
    },
    {
      "epoch": 1.9508118433619868,
      "grad_norm": 0.5240120887756348,
      "learning_rate": 5.074559238458418e-06,
      "loss": 0.4284,
      "mean_token_accuracy": 0.8556304335594177,
      "num_tokens": 1094815653.0,
      "step": 32680
    },
    {
      "epoch": 1.9511103151862463,
      "grad_norm": 0.45406651496887207,
      "learning_rate": 5.073658720678026e-06,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8576762676239014,
      "num_tokens": 1094983333.0,
      "step": 32685
    },
    {
      "epoch": 1.9514087870105064,
      "grad_norm": 0.47419631481170654,
      "learning_rate": 5.072763665246994e-06,
      "loss": 0.3863,
      "mean_token_accuracy": 0.8681617498397827,
      "num_tokens": 1095151013.0,
      "step": 32690
    },
    {
      "epoch": 1.951707258834766,
      "grad_norm": 0.4790462851524353,
      "learning_rate": 5.071874072383329e-06,
      "loss": 0.3921,
      "mean_token_accuracy": 0.8667780041694642,
      "num_tokens": 1095318693.0,
      "step": 32695
    },
    {
      "epoch": 1.952005730659026,
      "grad_norm": 0.5091140270233154,
      "learning_rate": 5.07098994230371e-06,
      "loss": 0.4055,
      "mean_token_accuracy": 0.8613980770111084,
      "num_tokens": 1095486373.0,
      "step": 32700
    },
    {
      "epoch": 1.9523042024832855,
      "grad_norm": 0.5264834761619568,
      "learning_rate": 5.070111275223485e-06,
      "loss": 0.4116,
      "mean_token_accuracy": 0.8620064377784729,
      "num_tokens": 1095654053.0,
      "step": 32705
    },
    {
      "epoch": 1.9526026743075455,
      "grad_norm": 0.5464398264884949,
      "learning_rate": 5.069238071356666e-06,
      "loss": 0.3965,
      "mean_token_accuracy": 0.8665632843971253,
      "num_tokens": 1095821733.0,
      "step": 32710
    },
    {
      "epoch": 1.952901146131805,
      "grad_norm": 0.5134994387626648,
      "learning_rate": 5.068370330915939e-06,
      "loss": 0.445,
      "mean_token_accuracy": 0.8497137069702149,
      "num_tokens": 1095989413.0,
      "step": 32715
    },
    {
      "epoch": 1.953199617956065,
      "grad_norm": 0.4360191822052002,
      "learning_rate": 5.0675080541126604e-06,
      "loss": 0.3706,
      "mean_token_accuracy": 0.8747405409812927,
      "num_tokens": 1096157093.0,
      "step": 32720
    },
    {
      "epoch": 1.9534980897803247,
      "grad_norm": 0.5409044623374939,
      "learning_rate": 5.066651241156847e-06,
      "loss": 0.4359,
      "mean_token_accuracy": 0.8537755012512207,
      "num_tokens": 1096324773.0,
      "step": 32725
    },
    {
      "epoch": 1.9537965616045845,
      "grad_norm": 0.47892406582832336,
      "learning_rate": 5.0657998922572e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8529106497764587,
      "num_tokens": 1096492453.0,
      "step": 32730
    },
    {
      "epoch": 1.9540950334288443,
      "grad_norm": 0.49769166111946106,
      "learning_rate": 5.064954007621076e-06,
      "loss": 0.3899,
      "mean_token_accuracy": 0.8674698829650879,
      "num_tokens": 1096660133.0,
      "step": 32735
    },
    {
      "epoch": 1.954393505253104,
      "grad_norm": 0.5040245652198792,
      "learning_rate": 5.064113587454505e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8539365410804749,
      "num_tokens": 1096827813.0,
      "step": 32740
    },
    {
      "epoch": 1.9546919770773639,
      "grad_norm": 0.4996117651462555,
      "learning_rate": 5.063278631962192e-06,
      "loss": 0.3836,
      "mean_token_accuracy": 0.8699451327323914,
      "num_tokens": 1096995493.0,
      "step": 32745
    },
    {
      "epoch": 1.9549904489016237,
      "grad_norm": 0.4961124062538147,
      "learning_rate": 5.062449141347501e-06,
      "loss": 0.4251,
      "mean_token_accuracy": 0.8567100167274475,
      "num_tokens": 1097163173.0,
      "step": 32750
    },
    {
      "epoch": 1.9552889207258835,
      "grad_norm": 0.4479682743549347,
      "learning_rate": 5.061625115812472e-06,
      "loss": 0.4003,
      "mean_token_accuracy": 0.8652451276779175,
      "num_tokens": 1097330853.0,
      "step": 32755
    },
    {
      "epoch": 1.9555873925501432,
      "grad_norm": 0.548488199710846,
      "learning_rate": 5.060806555557811e-06,
      "loss": 0.4109,
      "mean_token_accuracy": 0.8621853828430176,
      "num_tokens": 1097498533.0,
      "step": 32760
    },
    {
      "epoch": 1.955885864374403,
      "grad_norm": 0.5609689354896545,
      "learning_rate": 5.059993460782892e-06,
      "loss": 0.4446,
      "mean_token_accuracy": 0.8500775337219239,
      "num_tokens": 1097666213.0,
      "step": 32765
    },
    {
      "epoch": 1.9561843361986628,
      "grad_norm": 0.47095489501953125,
      "learning_rate": 5.059185831685762e-06,
      "loss": 0.4309,
      "mean_token_accuracy": 0.8558749794960022,
      "num_tokens": 1097833893.0,
      "step": 32770
    },
    {
      "epoch": 1.9564828080229226,
      "grad_norm": 0.5144312977790833,
      "learning_rate": 5.058383668463131e-06,
      "loss": 0.4152,
      "mean_token_accuracy": 0.8596982002258301,
      "num_tokens": 1098001573.0,
      "step": 32775
    },
    {
      "epoch": 1.9567812798471824,
      "grad_norm": 0.4622931182384491,
      "learning_rate": 5.057586971310384e-06,
      "loss": 0.368,
      "mean_token_accuracy": 0.8749791264533997,
      "num_tokens": 1098169253.0,
      "step": 32780
    },
    {
      "epoch": 1.9570797516714422,
      "grad_norm": 0.5186492204666138,
      "learning_rate": 5.05679574042157e-06,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8546463131904602,
      "num_tokens": 1098336933.0,
      "step": 32785
    },
    {
      "epoch": 1.957378223495702,
      "grad_norm": 0.47548261284828186,
      "learning_rate": 5.056009975989405e-06,
      "loss": 0.4,
      "mean_token_accuracy": 0.863748061656952,
      "num_tokens": 1098504613.0,
      "step": 32790
    },
    {
      "epoch": 1.9576766953199618,
      "grad_norm": 0.5142004489898682,
      "learning_rate": 5.055229678205281e-06,
      "loss": 0.4526,
      "mean_token_accuracy": 0.8496302127838135,
      "num_tokens": 1098672293.0,
      "step": 32795
    },
    {
      "epoch": 1.9579751671442216,
      "grad_norm": 0.4857657253742218,
      "learning_rate": 5.054454847259252e-06,
      "loss": 0.4216,
      "mean_token_accuracy": 0.8572408437728882,
      "num_tokens": 1098839973.0,
      "step": 32800
    },
    {
      "epoch": 1.9582736389684814,
      "grad_norm": 0.5145917534828186,
      "learning_rate": 5.053685483340041e-06,
      "loss": 0.4209,
      "mean_token_accuracy": 0.8593462944030762,
      "num_tokens": 1099007653.0,
      "step": 32805
    },
    {
      "epoch": 1.9585721107927412,
      "grad_norm": 0.4593253433704376,
      "learning_rate": 5.052921586635044e-06,
      "loss": 0.3843,
      "mean_token_accuracy": 0.8702910661697387,
      "num_tokens": 1099175333.0,
      "step": 32810
    },
    {
      "epoch": 1.958870582617001,
      "grad_norm": 0.543697714805603,
      "learning_rate": 5.052163157330317e-06,
      "loss": 0.4267,
      "mean_token_accuracy": 0.8570320844650269,
      "num_tokens": 1099343013.0,
      "step": 32815
    },
    {
      "epoch": 1.9591690544412608,
      "grad_norm": 0.46874719858169556,
      "learning_rate": 5.051410195610596e-06,
      "loss": 0.3923,
      "mean_token_accuracy": 0.8681200027465821,
      "num_tokens": 1099510693.0,
      "step": 32820
    },
    {
      "epoch": 1.9594675262655206,
      "grad_norm": 0.5890623331069946,
      "learning_rate": 5.0506627016592744e-06,
      "loss": 0.4119,
      "mean_token_accuracy": 0.8600143194198608,
      "num_tokens": 1099678373.0,
      "step": 32825
    },
    {
      "epoch": 1.9597659980897804,
      "grad_norm": 0.5843247175216675,
      "learning_rate": 5.049920675658418e-06,
      "loss": 0.4121,
      "mean_token_accuracy": 0.8618871450424195,
      "num_tokens": 1099846053.0,
      "step": 32830
    },
    {
      "epoch": 1.9600644699140402,
      "grad_norm": 0.4814905524253845,
      "learning_rate": 5.049184117788762e-06,
      "loss": 0.4166,
      "mean_token_accuracy": 0.860914945602417,
      "num_tokens": 1100013733.0,
      "step": 32835
    },
    {
      "epoch": 1.9603629417383,
      "grad_norm": 0.5177224278450012,
      "learning_rate": 5.0484530282297106e-06,
      "loss": 0.4146,
      "mean_token_accuracy": 0.8597220659255982,
      "num_tokens": 1100181413.0,
      "step": 32840
    },
    {
      "epoch": 1.9606614135625597,
      "grad_norm": 0.5069063901901245,
      "learning_rate": 5.04772740715933e-06,
      "loss": 0.4041,
      "mean_token_accuracy": 0.8631635546684265,
      "num_tokens": 1100349093.0,
      "step": 32845
    },
    {
      "epoch": 1.9609598853868195,
      "grad_norm": 0.5217315554618835,
      "learning_rate": 5.047007254754361e-06,
      "loss": 0.4234,
      "mean_token_accuracy": 0.8576046824455261,
      "num_tokens": 1100516773.0,
      "step": 32850
    },
    {
      "epoch": 1.9612583572110793,
      "grad_norm": 0.46630823612213135,
      "learning_rate": 5.046292571190211e-06,
      "loss": 0.3977,
      "mean_token_accuracy": 0.8659966588020325,
      "num_tokens": 1100684453.0,
      "step": 32855
    },
    {
      "epoch": 1.961556829035339,
      "grad_norm": 0.5049127340316772,
      "learning_rate": 5.045583356640954e-06,
      "loss": 0.4402,
      "mean_token_accuracy": 0.8517120480537415,
      "num_tokens": 1100850815.0,
      "step": 32860
    },
    {
      "epoch": 1.961855300859599,
      "grad_norm": 0.5101924538612366,
      "learning_rate": 5.044879611279331e-06,
      "loss": 0.4311,
      "mean_token_accuracy": 0.8554395914077759,
      "num_tokens": 1101018495.0,
      "step": 32865
    },
    {
      "epoch": 1.9621537726838585,
      "grad_norm": 0.5113304257392883,
      "learning_rate": 5.044181335276751e-06,
      "loss": 0.4564,
      "mean_token_accuracy": 0.8484909772872925,
      "num_tokens": 1101186175.0,
      "step": 32870
    },
    {
      "epoch": 1.9624522445081185,
      "grad_norm": 0.5390164852142334,
      "learning_rate": 5.043488528803296e-06,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8523141980171204,
      "num_tokens": 1101353855.0,
      "step": 32875
    },
    {
      "epoch": 1.962750716332378,
      "grad_norm": 0.5232124328613281,
      "learning_rate": 5.042801192027708e-06,
      "loss": 0.3973,
      "mean_token_accuracy": 0.8659191131591797,
      "num_tokens": 1101521535.0,
      "step": 32880
    },
    {
      "epoch": 1.963049188156638,
      "grad_norm": 0.5050330758094788,
      "learning_rate": 5.0421193251174005e-06,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8562626719474793,
      "num_tokens": 1101689215.0,
      "step": 32885
    },
    {
      "epoch": 1.9633476599808977,
      "grad_norm": 0.5037065744400024,
      "learning_rate": 5.041442928238459e-06,
      "loss": 0.4104,
      "mean_token_accuracy": 0.8612787842750549,
      "num_tokens": 1101856895.0,
      "step": 32890
    },
    {
      "epoch": 1.9636461318051577,
      "grad_norm": 0.5262560844421387,
      "learning_rate": 5.0407720015556295e-06,
      "loss": 0.4331,
      "mean_token_accuracy": 0.8548431277275086,
      "num_tokens": 1102024575.0,
      "step": 32895
    },
    {
      "epoch": 1.9639446036294173,
      "grad_norm": 0.5172950625419617,
      "learning_rate": 5.040106545232327e-06,
      "loss": 0.4096,
      "mean_token_accuracy": 0.8620899438858032,
      "num_tokens": 1102192255.0,
      "step": 32900
    },
    {
      "epoch": 1.9642430754536773,
      "grad_norm": 0.6064838767051697,
      "learning_rate": 5.03944655943064e-06,
      "loss": 0.4078,
      "mean_token_accuracy": 0.8617559313774109,
      "num_tokens": 1102359935.0,
      "step": 32905
    },
    {
      "epoch": 1.9645415472779368,
      "grad_norm": 0.4952913522720337,
      "learning_rate": 5.038792044311317e-06,
      "loss": 0.4189,
      "mean_token_accuracy": 0.8586424946784973,
      "num_tokens": 1102527615.0,
      "step": 32910
    },
    {
      "epoch": 1.9648400191021969,
      "grad_norm": 0.4856053292751312,
      "learning_rate": 5.0381430000337765e-06,
      "loss": 0.3827,
      "mean_token_accuracy": 0.8710127711296082,
      "num_tokens": 1102695295.0,
      "step": 32915
    },
    {
      "epoch": 1.9651384909264564,
      "grad_norm": 0.5069352984428406,
      "learning_rate": 5.037499426756109e-06,
      "loss": 0.4577,
      "mean_token_accuracy": 0.8478229761123657,
      "num_tokens": 1102862975.0,
      "step": 32920
    },
    {
      "epoch": 1.9654369627507164,
      "grad_norm": 0.5154653787612915,
      "learning_rate": 5.0368613246350645e-06,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8479184031486511,
      "num_tokens": 1103030655.0,
      "step": 32925
    },
    {
      "epoch": 1.965735434574976,
      "grad_norm": 0.5819794535636902,
      "learning_rate": 5.036228693826068e-06,
      "loss": 0.421,
      "mean_token_accuracy": 0.8591673612594605,
      "num_tokens": 1103198335.0,
      "step": 32930
    },
    {
      "epoch": 1.966033906399236,
      "grad_norm": 0.5436822175979614,
      "learning_rate": 5.035601534483207e-06,
      "loss": 0.407,
      "mean_token_accuracy": 0.8614935040473938,
      "num_tokens": 1103366015.0,
      "step": 32935
    },
    {
      "epoch": 1.9663323782234956,
      "grad_norm": 0.48451313376426697,
      "learning_rate": 5.034979846759236e-06,
      "loss": 0.4602,
      "mean_token_accuracy": 0.8455326199531555,
      "num_tokens": 1103533695.0,
      "step": 32940
    },
    {
      "epoch": 1.9666308500477556,
      "grad_norm": 0.5387483239173889,
      "learning_rate": 5.03436363080558e-06,
      "loss": 0.4106,
      "mean_token_accuracy": 0.8616485714912414,
      "num_tokens": 1103701375.0,
      "step": 32945
    },
    {
      "epoch": 1.9669293218720152,
      "grad_norm": 0.4694807529449463,
      "learning_rate": 5.03375288677233e-06,
      "loss": 0.4009,
      "mean_token_accuracy": 0.8634259700775146,
      "num_tokens": 1103869055.0,
      "step": 32950
    },
    {
      "epoch": 1.9672277936962752,
      "grad_norm": 0.49442437291145325,
      "learning_rate": 5.033147614808247e-06,
      "loss": 0.4256,
      "mean_token_accuracy": 0.8560479640960693,
      "num_tokens": 1104036735.0,
      "step": 32955
    },
    {
      "epoch": 1.9675262655205348,
      "grad_norm": 0.4747960865497589,
      "learning_rate": 5.032547815060747e-06,
      "loss": 0.419,
      "mean_token_accuracy": 0.858958613872528,
      "num_tokens": 1104204415.0,
      "step": 32960
    },
    {
      "epoch": 1.9678247373447948,
      "grad_norm": 0.48176291584968567,
      "learning_rate": 5.031953487675933e-06,
      "loss": 0.4481,
      "mean_token_accuracy": 0.8515388369560242,
      "num_tokens": 1104372095.0,
      "step": 32965
    },
    {
      "epoch": 1.9681232091690544,
      "grad_norm": 0.4866582155227661,
      "learning_rate": 5.031364632798559e-06,
      "loss": 0.4034,
      "mean_token_accuracy": 0.8639150619506836,
      "num_tokens": 1104539775.0,
      "step": 32970
    },
    {
      "epoch": 1.9684216809933144,
      "grad_norm": 0.48427248001098633,
      "learning_rate": 5.030781250572053e-06,
      "loss": 0.3808,
      "mean_token_accuracy": 0.8713348507881165,
      "num_tokens": 1104707455.0,
      "step": 32975
    },
    {
      "epoch": 1.968720152817574,
      "grad_norm": 0.5667198300361633,
      "learning_rate": 5.030203341138504e-06,
      "loss": 0.4192,
      "mean_token_accuracy": 0.8588691353797913,
      "num_tokens": 1104875135.0,
      "step": 32980
    },
    {
      "epoch": 1.969018624641834,
      "grad_norm": 0.5106105208396912,
      "learning_rate": 5.029630904638678e-06,
      "loss": 0.4168,
      "mean_token_accuracy": 0.8606465458869934,
      "num_tokens": 1105042815.0,
      "step": 32985
    },
    {
      "epoch": 1.9693170964660935,
      "grad_norm": 0.4783756136894226,
      "learning_rate": 5.029063941212001e-06,
      "loss": 0.4382,
      "mean_token_accuracy": 0.8531193971633911,
      "num_tokens": 1105210495.0,
      "step": 32990
    },
    {
      "epoch": 1.9696155682903533,
      "grad_norm": 0.5211960673332214,
      "learning_rate": 5.028502450996568e-06,
      "loss": 0.3857,
      "mean_token_accuracy": 0.8693964123725891,
      "num_tokens": 1105378175.0,
      "step": 32995
    },
    {
      "epoch": 1.9699140401146131,
      "grad_norm": 0.5239377617835999,
      "learning_rate": 5.027946434129137e-06,
      "loss": 0.4134,
      "mean_token_accuracy": 0.8608791470527649,
      "num_tokens": 1105545855.0,
      "step": 33000
    },
    {
      "epoch": 1.970212511938873,
      "grad_norm": 0.4839612543582916,
      "learning_rate": 5.027395890745141e-06,
      "loss": 0.4018,
      "mean_token_accuracy": 0.8651616454124451,
      "num_tokens": 1105713535.0,
      "step": 33005
    },
    {
      "epoch": 1.9705109837631327,
      "grad_norm": 0.5277972221374512,
      "learning_rate": 5.026850820978671e-06,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8563783884048461,
      "num_tokens": 1105876527.0,
      "step": 33010
    },
    {
      "epoch": 1.9708094555873925,
      "grad_norm": 0.4896654486656189,
      "learning_rate": 5.0263112249624925e-06,
      "loss": 0.4208,
      "mean_token_accuracy": 0.8595192670822144,
      "num_tokens": 1106044207.0,
      "step": 33015
    },
    {
      "epoch": 1.9711079274116523,
      "grad_norm": 0.5849815607070923,
      "learning_rate": 5.025777102828029e-06,
      "loss": 0.4675,
      "mean_token_accuracy": 0.8446141004562377,
      "num_tokens": 1106211887.0,
      "step": 33020
    },
    {
      "epoch": 1.971406399235912,
      "grad_norm": 0.5098403692245483,
      "learning_rate": 5.0252484547053815e-06,
      "loss": 0.4147,
      "mean_token_accuracy": 0.8612155199050904,
      "num_tokens": 1106373449.0,
      "step": 33025
    },
    {
      "epoch": 1.971704871060172,
      "grad_norm": 0.5544062852859497,
      "learning_rate": 5.0247252807233074e-06,
      "loss": 0.431,
      "mean_token_accuracy": 0.8559704184532165,
      "num_tokens": 1106541129.0,
      "step": 33030
    },
    {
      "epoch": 1.9720033428844317,
      "grad_norm": 0.4739290177822113,
      "learning_rate": 5.0242075810092364e-06,
      "loss": 0.3962,
      "mean_token_accuracy": 0.8666467785835266,
      "num_tokens": 1106708809.0,
      "step": 33035
    },
    {
      "epoch": 1.9723018147086915,
      "grad_norm": 0.48968541622161865,
      "learning_rate": 5.023695355689267e-06,
      "loss": 0.4206,
      "mean_token_accuracy": 0.858057975769043,
      "num_tokens": 1106876489.0,
      "step": 33040
    },
    {
      "epoch": 1.9726002865329513,
      "grad_norm": 0.546718955039978,
      "learning_rate": 5.023188604888156e-06,
      "loss": 0.479,
      "mean_token_accuracy": 0.8403674006462097,
      "num_tokens": 1107044169.0,
      "step": 33045
    },
    {
      "epoch": 1.972898758357211,
      "grad_norm": 0.4721791744232178,
      "learning_rate": 5.022687328729336e-06,
      "loss": 0.419,
      "mean_token_accuracy": 0.859185254573822,
      "num_tokens": 1107211849.0,
      "step": 33050
    },
    {
      "epoch": 1.9731972301814709,
      "grad_norm": 0.5139625072479248,
      "learning_rate": 5.022191527334902e-06,
      "loss": 0.4434,
      "mean_token_accuracy": 0.8497316002845764,
      "num_tokens": 1107379529.0,
      "step": 33055
    },
    {
      "epoch": 1.9734957020057307,
      "grad_norm": 0.5600574612617493,
      "learning_rate": 5.021701200825614e-06,
      "loss": 0.4695,
      "mean_token_accuracy": 0.843546462059021,
      "num_tokens": 1107547209.0,
      "step": 33060
    },
    {
      "epoch": 1.9737941738299905,
      "grad_norm": 0.5157943964004517,
      "learning_rate": 5.0212163493208995e-06,
      "loss": 0.4167,
      "mean_token_accuracy": 0.8595789074897766,
      "num_tokens": 1107714889.0,
      "step": 33065
    },
    {
      "epoch": 1.9740926456542502,
      "grad_norm": 0.5059912204742432,
      "learning_rate": 5.020736972938854e-06,
      "loss": 0.4472,
      "mean_token_accuracy": 0.848872709274292,
      "num_tokens": 1107882569.0,
      "step": 33070
    },
    {
      "epoch": 1.97439111747851,
      "grad_norm": 0.5495331287384033,
      "learning_rate": 5.0202630717962385e-06,
      "loss": 0.4162,
      "mean_token_accuracy": 0.8586782932281494,
      "num_tokens": 1108050249.0,
      "step": 33075
    },
    {
      "epoch": 1.9746895893027698,
      "grad_norm": 0.5167054533958435,
      "learning_rate": 5.01979464600848e-06,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8533699154853821,
      "num_tokens": 1108217929.0,
      "step": 33080
    },
    {
      "epoch": 1.9749880611270296,
      "grad_norm": 0.4671654999256134,
      "learning_rate": 5.019331695689673e-06,
      "loss": 0.4025,
      "mean_token_accuracy": 0.8629547834396363,
      "num_tokens": 1108385609.0,
      "step": 33085
    },
    {
      "epoch": 1.9752865329512894,
      "grad_norm": 0.5264710783958435,
      "learning_rate": 5.018874220952578e-06,
      "loss": 0.4018,
      "mean_token_accuracy": 0.8641238331794738,
      "num_tokens": 1108553289.0,
      "step": 33090
    },
    {
      "epoch": 1.9755850047755492,
      "grad_norm": 0.4915428161621094,
      "learning_rate": 5.01842222190862e-06,
      "loss": 0.4233,
      "mean_token_accuracy": 0.8572289109230041,
      "num_tokens": 1108720969.0,
      "step": 33095
    },
    {
      "epoch": 1.975883476599809,
      "grad_norm": 0.580143392086029,
      "learning_rate": 5.017975698667892e-06,
      "loss": 0.4386,
      "mean_token_accuracy": 0.8535727143287659,
      "num_tokens": 1108888649.0,
      "step": 33100
    },
    {
      "epoch": 1.9761819484240688,
      "grad_norm": 0.5151272416114807,
      "learning_rate": 5.0175346513391506e-06,
      "loss": 0.423,
      "mean_token_accuracy": 0.8572527766227722,
      "num_tokens": 1109056329.0,
      "step": 33105
    },
    {
      "epoch": 1.9764804202483286,
      "grad_norm": 0.5739167332649231,
      "learning_rate": 5.017099080029827e-06,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8560061931610108,
      "num_tokens": 1109224009.0,
      "step": 33110
    },
    {
      "epoch": 1.9767788920725884,
      "grad_norm": 0.5494800209999084,
      "learning_rate": 5.016668984846008e-06,
      "loss": 0.3928,
      "mean_token_accuracy": 0.8678158164024353,
      "num_tokens": 1109391689.0,
      "step": 33115
    },
    {
      "epoch": 1.9770773638968482,
      "grad_norm": 0.4673916697502136,
      "learning_rate": 5.016244365892452e-06,
      "loss": 0.4051,
      "mean_token_accuracy": 0.8637182235717773,
      "num_tokens": 1109559369.0,
      "step": 33120
    },
    {
      "epoch": 1.977375835721108,
      "grad_norm": 0.44602450728416443,
      "learning_rate": 5.0158252232725845e-06,
      "loss": 0.4042,
      "mean_token_accuracy": 0.8626327037811279,
      "num_tokens": 1109727049.0,
      "step": 33125
    },
    {
      "epoch": 1.9776743075453678,
      "grad_norm": 0.5123845338821411,
      "learning_rate": 5.015411557088493e-06,
      "loss": 0.4368,
      "mean_token_accuracy": 0.8515925168991089,
      "num_tokens": 1109894729.0,
      "step": 33130
    },
    {
      "epoch": 1.9779727793696273,
      "grad_norm": 0.5120893716812134,
      "learning_rate": 5.015003367440934e-06,
      "loss": 0.4247,
      "mean_token_accuracy": 0.8575330972671509,
      "num_tokens": 1110062409.0,
      "step": 33135
    },
    {
      "epoch": 1.9782712511938874,
      "grad_norm": 0.47194841504096985,
      "learning_rate": 5.014600654429334e-06,
      "loss": 0.3922,
      "mean_token_accuracy": 0.8669569253921509,
      "num_tokens": 1110230089.0,
      "step": 33140
    },
    {
      "epoch": 1.978569723018147,
      "grad_norm": 0.5535193085670471,
      "learning_rate": 5.014203418151775e-06,
      "loss": 0.4346,
      "mean_token_accuracy": 0.8549027800559997,
      "num_tokens": 1110397769.0,
      "step": 33145
    },
    {
      "epoch": 1.978868194842407,
      "grad_norm": 0.5228692293167114,
      "learning_rate": 5.013811658705015e-06,
      "loss": 0.4278,
      "mean_token_accuracy": 0.8584695219993591,
      "num_tokens": 1110565449.0,
      "step": 33150
    },
    {
      "epoch": 1.9791666666666665,
      "grad_norm": 0.553502082824707,
      "learning_rate": 5.01342537618447e-06,
      "loss": 0.4309,
      "mean_token_accuracy": 0.8556781649589539,
      "num_tokens": 1110733129.0,
      "step": 33155
    },
    {
      "epoch": 1.9794651384909265,
      "grad_norm": 0.5194050669670105,
      "learning_rate": 5.013044570684232e-06,
      "loss": 0.4332,
      "mean_token_accuracy": 0.8560121774673461,
      "num_tokens": 1110900809.0,
      "step": 33160
    },
    {
      "epoch": 1.979763610315186,
      "grad_norm": 0.5111545324325562,
      "learning_rate": 5.01266924229705e-06,
      "loss": 0.4112,
      "mean_token_accuracy": 0.860199224948883,
      "num_tokens": 1111068489.0,
      "step": 33165
    },
    {
      "epoch": 1.9800620821394461,
      "grad_norm": 0.4741707146167755,
      "learning_rate": 5.012299391114342e-06,
      "loss": 0.4181,
      "mean_token_accuracy": 0.8597339868545533,
      "num_tokens": 1111236169.0,
      "step": 33170
    },
    {
      "epoch": 1.9803605539637057,
      "grad_norm": 0.5976281762123108,
      "learning_rate": 5.011935017226192e-06,
      "loss": 0.4043,
      "mean_token_accuracy": 0.8634558081626892,
      "num_tokens": 1111403849.0,
      "step": 33175
    },
    {
      "epoch": 1.9806590257879657,
      "grad_norm": 0.5199026465415955,
      "learning_rate": 5.011576120721352e-06,
      "loss": 0.4067,
      "mean_token_accuracy": 0.8623583555221558,
      "num_tokens": 1111571529.0,
      "step": 33180
    },
    {
      "epoch": 1.9809574976122253,
      "grad_norm": 0.4533449113368988,
      "learning_rate": 5.0112227016872364e-06,
      "loss": 0.4223,
      "mean_token_accuracy": 0.8586782932281494,
      "num_tokens": 1111739209.0,
      "step": 33185
    },
    {
      "epoch": 1.9812559694364853,
      "grad_norm": 0.5162702798843384,
      "learning_rate": 5.010874760209927e-06,
      "loss": 0.4301,
      "mean_token_accuracy": 0.8562746047973633,
      "num_tokens": 1111906889.0,
      "step": 33190
    },
    {
      "epoch": 1.9815544412607449,
      "grad_norm": 0.5097695589065552,
      "learning_rate": 5.0105322963741725e-06,
      "loss": 0.4564,
      "mean_token_accuracy": 0.8489800810813903,
      "num_tokens": 1112074569.0,
      "step": 33195
    },
    {
      "epoch": 1.9818529130850049,
      "grad_norm": 0.49297916889190674,
      "learning_rate": 5.010195310263383e-06,
      "loss": 0.3998,
      "mean_token_accuracy": 0.8655552983283996,
      "num_tokens": 1112242249.0,
      "step": 33200
    },
    {
      "epoch": 1.9821513849092645,
      "grad_norm": 0.5222697854042053,
      "learning_rate": 5.0098638019596404e-06,
      "loss": 0.4177,
      "mean_token_accuracy": 0.8594477057456971,
      "num_tokens": 1112409929.0,
      "step": 33205
    },
    {
      "epoch": 1.9824498567335245,
      "grad_norm": 0.5047992467880249,
      "learning_rate": 5.00953777154369e-06,
      "loss": 0.418,
      "mean_token_accuracy": 0.8590838551521301,
      "num_tokens": 1112577609.0,
      "step": 33210
    },
    {
      "epoch": 1.982748328557784,
      "grad_norm": 0.5256698131561279,
      "learning_rate": 5.009217219094941e-06,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8532148361206054,
      "num_tokens": 1112745289.0,
      "step": 33215
    },
    {
      "epoch": 1.983046800382044,
      "grad_norm": 0.5669993162155151,
      "learning_rate": 5.008902144691471e-06,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8474114298820495,
      "num_tokens": 1112912969.0,
      "step": 33220
    },
    {
      "epoch": 1.9833452722063036,
      "grad_norm": 0.49203798174858093,
      "learning_rate": 5.008592548410022e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.852105462551117,
      "num_tokens": 1113080649.0,
      "step": 33225
    },
    {
      "epoch": 1.9836437440305636,
      "grad_norm": 0.511444628238678,
      "learning_rate": 5.008288430326001e-06,
      "loss": 0.4418,
      "mean_token_accuracy": 0.8524275422096252,
      "num_tokens": 1113248329.0,
      "step": 33230
    },
    {
      "epoch": 1.9839422158548232,
      "grad_norm": 0.4787483513355255,
      "learning_rate": 5.007989790513481e-06,
      "loss": 0.4039,
      "mean_token_accuracy": 0.8646009922027588,
      "num_tokens": 1113416009.0,
      "step": 33235
    },
    {
      "epoch": 1.9842406876790832,
      "grad_norm": 0.49464672803878784,
      "learning_rate": 5.007696629045205e-06,
      "loss": 0.4124,
      "mean_token_accuracy": 0.8605809450149536,
      "num_tokens": 1113583689.0,
      "step": 33240
    },
    {
      "epoch": 1.9845391595033428,
      "grad_norm": 0.4983403980731964,
      "learning_rate": 5.0074089459925755e-06,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8521472096443177,
      "num_tokens": 1113751369.0,
      "step": 33245
    },
    {
      "epoch": 1.9848376313276028,
      "grad_norm": 0.553785502910614,
      "learning_rate": 5.0071267414256605e-06,
      "loss": 0.4299,
      "mean_token_accuracy": 0.8551592469215393,
      "num_tokens": 1113919049.0,
      "step": 33250
    },
    {
      "epoch": 1.9851361031518624,
      "grad_norm": 0.503053605556488,
      "learning_rate": 5.006850015413201e-06,
      "loss": 0.4165,
      "mean_token_accuracy": 0.8602290272712707,
      "num_tokens": 1114086729.0,
      "step": 33255
    },
    {
      "epoch": 1.9854345749761224,
      "grad_norm": 0.6152521371841431,
      "learning_rate": 5.006578768022595e-06,
      "loss": 0.4511,
      "mean_token_accuracy": 0.850715720653534,
      "num_tokens": 1114254409.0,
      "step": 33260
    },
    {
      "epoch": 1.985733046800382,
      "grad_norm": 0.47155895829200745,
      "learning_rate": 5.00631299931991e-06,
      "loss": 0.4256,
      "mean_token_accuracy": 0.8568531632423401,
      "num_tokens": 1114422089.0,
      "step": 33265
    },
    {
      "epoch": 1.9860315186246418,
      "grad_norm": 0.45770949125289917,
      "learning_rate": 5.006052709369881e-06,
      "loss": 0.4136,
      "mean_token_accuracy": 0.8623821973800659,
      "num_tokens": 1114589769.0,
      "step": 33270
    },
    {
      "epoch": 1.9863299904489016,
      "grad_norm": 0.5494801998138428,
      "learning_rate": 5.005797898235906e-06,
      "loss": 0.4718,
      "mean_token_accuracy": 0.8433913826942444,
      "num_tokens": 1114757449.0,
      "step": 33275
    },
    {
      "epoch": 1.9866284622731614,
      "grad_norm": 0.5019585490226746,
      "learning_rate": 5.005548565980046e-06,
      "loss": 0.4449,
      "mean_token_accuracy": 0.850501012802124,
      "num_tokens": 1114925129.0,
      "step": 33280
    },
    {
      "epoch": 1.9869269340974212,
      "grad_norm": 0.49741071462631226,
      "learning_rate": 5.005304712663035e-06,
      "loss": 0.4449,
      "mean_token_accuracy": 0.8513837575912475,
      "num_tokens": 1115092809.0,
      "step": 33285
    },
    {
      "epoch": 1.987225405921681,
      "grad_norm": 0.4714791476726532,
      "learning_rate": 5.005066338344263e-06,
      "loss": 0.4112,
      "mean_token_accuracy": 0.8611952781677246,
      "num_tokens": 1115260489.0,
      "step": 33290
    },
    {
      "epoch": 1.9875238777459407,
      "grad_norm": 0.50583416223526,
      "learning_rate": 5.004833443081796e-06,
      "loss": 0.4276,
      "mean_token_accuracy": 0.8571155905723572,
      "num_tokens": 1115428169.0,
      "step": 33295
    },
    {
      "epoch": 1.9878223495702005,
      "grad_norm": 0.5397236943244934,
      "learning_rate": 5.004606026932356e-06,
      "loss": 0.4563,
      "mean_token_accuracy": 0.8470356583595275,
      "num_tokens": 1115595849.0,
      "step": 33300
    },
    {
      "epoch": 1.9881208213944603,
      "grad_norm": 0.5348944067955017,
      "learning_rate": 5.004384089951338e-06,
      "loss": 0.4351,
      "mean_token_accuracy": 0.8554216980934143,
      "num_tokens": 1115763529.0,
      "step": 33305
    },
    {
      "epoch": 1.9884192932187201,
      "grad_norm": 0.4901181757450104,
      "learning_rate": 5.004167632192794e-06,
      "loss": 0.4269,
      "mean_token_accuracy": 0.856817364692688,
      "num_tokens": 1115931209.0,
      "step": 33310
    },
    {
      "epoch": 1.98871776504298,
      "grad_norm": 0.48394614458084106,
      "learning_rate": 5.003956653709448e-06,
      "loss": 0.4253,
      "mean_token_accuracy": 0.8571633100509644,
      "num_tokens": 1116098889.0,
      "step": 33315
    },
    {
      "epoch": 1.9890162368672397,
      "grad_norm": 0.5433171391487122,
      "learning_rate": 5.003751154552688e-06,
      "loss": 0.4372,
      "mean_token_accuracy": 0.8538411021232605,
      "num_tokens": 1116266569.0,
      "step": 33320
    },
    {
      "epoch": 1.9893147086914995,
      "grad_norm": 0.5413157939910889,
      "learning_rate": 5.003551134772569e-06,
      "loss": 0.4102,
      "mean_token_accuracy": 0.8618573307991028,
      "num_tokens": 1116434249.0,
      "step": 33325
    },
    {
      "epoch": 1.9896131805157593,
      "grad_norm": 0.5299882888793945,
      "learning_rate": 5.003356594417809e-06,
      "loss": 0.4193,
      "mean_token_accuracy": 0.8580997228622437,
      "num_tokens": 1116601929.0,
      "step": 33330
    },
    {
      "epoch": 1.989911652340019,
      "grad_norm": 0.4970306158065796,
      "learning_rate": 5.0031675335357895e-06,
      "loss": 0.4112,
      "mean_token_accuracy": 0.8617618918418884,
      "num_tokens": 1116769609.0,
      "step": 33335
    },
    {
      "epoch": 1.990210124164279,
      "grad_norm": 0.47278285026550293,
      "learning_rate": 5.0029839521725605e-06,
      "loss": 0.426,
      "mean_token_accuracy": 0.858421802520752,
      "num_tokens": 1116937289.0,
      "step": 33340
    },
    {
      "epoch": 1.9905085959885387,
      "grad_norm": 0.4842239320278168,
      "learning_rate": 5.002805850372837e-06,
      "loss": 0.4031,
      "mean_token_accuracy": 0.8642132878303528,
      "num_tokens": 1117104969.0,
      "step": 33345
    },
    {
      "epoch": 1.9908070678127985,
      "grad_norm": 0.5508589744567871,
      "learning_rate": 5.002633228180002e-06,
      "loss": 0.4413,
      "mean_token_accuracy": 0.8531373023986817,
      "num_tokens": 1117272649.0,
      "step": 33350
    },
    {
      "epoch": 1.9911055396370583,
      "grad_norm": 0.49138855934143066,
      "learning_rate": 5.002466085636097e-06,
      "loss": 0.4588,
      "mean_token_accuracy": 0.8479780554771423,
      "num_tokens": 1117440329.0,
      "step": 33355
    },
    {
      "epoch": 1.991404011461318,
      "grad_norm": 0.5040826797485352,
      "learning_rate": 5.002304422781831e-06,
      "loss": 0.441,
      "mean_token_accuracy": 0.8517058372497559,
      "num_tokens": 1117608009.0,
      "step": 33360
    },
    {
      "epoch": 1.9917024832855779,
      "grad_norm": 0.4997211694717407,
      "learning_rate": 5.002148239656586e-06,
      "loss": 0.4086,
      "mean_token_accuracy": 0.8632410883903503,
      "num_tokens": 1117775689.0,
      "step": 33365
    },
    {
      "epoch": 1.9920009551098377,
      "grad_norm": 0.6780403852462769,
      "learning_rate": 5.001997536298399e-06,
      "loss": 0.4413,
      "mean_token_accuracy": 0.8516461968421936,
      "num_tokens": 1117943369.0,
      "step": 33370
    },
    {
      "epoch": 1.9922994269340975,
      "grad_norm": 0.5391092300415039,
      "learning_rate": 5.001852312743976e-06,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8525408625602722,
      "num_tokens": 1118111049.0,
      "step": 33375
    },
    {
      "epoch": 1.9925978987583572,
      "grad_norm": 0.4946630001068115,
      "learning_rate": 5.001712569028694e-06,
      "loss": 0.4145,
      "mean_token_accuracy": 0.8602588534355163,
      "num_tokens": 1118278729.0,
      "step": 33380
    },
    {
      "epoch": 1.992896370582617,
      "grad_norm": 0.5029065608978271,
      "learning_rate": 5.001578305186583e-06,
      "loss": 0.439,
      "mean_token_accuracy": 0.8532923817634582,
      "num_tokens": 1118446409.0,
      "step": 33385
    },
    {
      "epoch": 1.9931948424068768,
      "grad_norm": 0.5339170694351196,
      "learning_rate": 5.00144952125035e-06,
      "loss": 0.4476,
      "mean_token_accuracy": 0.849886679649353,
      "num_tokens": 1118614089.0,
      "step": 33390
    },
    {
      "epoch": 1.9934933142311366,
      "grad_norm": 0.5761162638664246,
      "learning_rate": 5.001326217251364e-06,
      "loss": 0.4348,
      "mean_token_accuracy": 0.8539246082305908,
      "num_tokens": 1118781769.0,
      "step": 33395
    },
    {
      "epoch": 1.9937917860553964,
      "grad_norm": 0.4588549733161926,
      "learning_rate": 5.001208393219653e-06,
      "loss": 0.4167,
      "mean_token_accuracy": 0.8589824676513672,
      "num_tokens": 1118949449.0,
      "step": 33400
    },
    {
      "epoch": 1.9940902578796562,
      "grad_norm": 0.5122301578521729,
      "learning_rate": 5.001096049183919e-06,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8490909337997437,
      "num_tokens": 1119115055.0,
      "step": 33405
    },
    {
      "epoch": 1.9943887297039158,
      "grad_norm": 0.5342676043510437,
      "learning_rate": 5.000989185171524e-06,
      "loss": 0.4385,
      "mean_token_accuracy": 0.8538709282875061,
      "num_tokens": 1119282735.0,
      "step": 33410
    },
    {
      "epoch": 1.9946872015281758,
      "grad_norm": 0.4799768328666687,
      "learning_rate": 5.000887801208497e-06,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8442085146903991,
      "num_tokens": 1119450415.0,
      "step": 33415
    },
    {
      "epoch": 1.9949856733524354,
      "grad_norm": 0.5027874708175659,
      "learning_rate": 5.000791897319534e-06,
      "loss": 0.406,
      "mean_token_accuracy": 0.8625551700592041,
      "num_tokens": 1119618095.0,
      "step": 33420
    },
    {
      "epoch": 1.9952841451766954,
      "grad_norm": 0.5278739929199219,
      "learning_rate": 5.0007014735279895e-06,
      "loss": 0.4203,
      "mean_token_accuracy": 0.8578969359397888,
      "num_tokens": 1119785775.0,
      "step": 33425
    },
    {
      "epoch": 1.995582617000955,
      "grad_norm": 0.5141813158988953,
      "learning_rate": 5.000616529855894e-06,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8492305874824524,
      "num_tokens": 1119953455.0,
      "step": 33430
    },
    {
      "epoch": 1.995881088825215,
      "grad_norm": 0.6113985776901245,
      "learning_rate": 5.000537066323929e-06,
      "loss": 0.4255,
      "mean_token_accuracy": 0.8578372836112976,
      "num_tokens": 1120121135.0,
      "step": 33435
    },
    {
      "epoch": 1.9961795606494745,
      "grad_norm": 0.5158660411834717,
      "learning_rate": 5.000463082951458e-06,
      "loss": 0.4125,
      "mean_token_accuracy": 0.8619825959205627,
      "num_tokens": 1120288815.0,
      "step": 33440
    },
    {
      "epoch": 1.9964780324737346,
      "grad_norm": 0.49674826860427856,
      "learning_rate": 5.000394579756494e-06,
      "loss": 0.4249,
      "mean_token_accuracy": 0.8567219376564026,
      "num_tokens": 1120456495.0,
      "step": 33445
    },
    {
      "epoch": 1.9967765042979941,
      "grad_norm": 0.6172782778739929,
      "learning_rate": 5.000331556755729e-06,
      "loss": 0.4158,
      "mean_token_accuracy": 0.8608806848526,
      "num_tokens": 1120618264.0,
      "step": 33450
    },
    {
      "epoch": 1.9970749761222542,
      "grad_norm": 0.5504373908042908,
      "learning_rate": 5.000274013964507e-06,
      "loss": 0.4341,
      "mean_token_accuracy": 0.8539544343948364,
      "num_tokens": 1120785944.0,
      "step": 33455
    },
    {
      "epoch": 1.9973734479465137,
      "grad_norm": 0.4685095548629761,
      "learning_rate": 5.000221951396845e-06,
      "loss": 0.4103,
      "mean_token_accuracy": 0.8622092366218567,
      "num_tokens": 1120953624.0,
      "step": 33460
    },
    {
      "epoch": 1.9976719197707737,
      "grad_norm": 0.4722028076648712,
      "learning_rate": 5.000175369065427e-06,
      "loss": 0.42,
      "mean_token_accuracy": 0.8578611493110657,
      "num_tokens": 1121121304.0,
      "step": 33465
    },
    {
      "epoch": 1.9979703915950333,
      "grad_norm": 0.4826979339122772,
      "learning_rate": 5.000134266981596e-06,
      "loss": 0.3938,
      "mean_token_accuracy": 0.8672253370285035,
      "num_tokens": 1121288984.0,
      "step": 33470
    },
    {
      "epoch": 1.9982688634192933,
      "grad_norm": 0.4782162308692932,
      "learning_rate": 5.000098645155365e-06,
      "loss": 0.4515,
      "mean_token_accuracy": 0.8497673869132996,
      "num_tokens": 1121456664.0,
      "step": 33475
    },
    {
      "epoch": 1.998567335243553,
      "grad_norm": 0.48977309465408325,
      "learning_rate": 5.0000685035954095e-06,
      "loss": 0.4259,
      "mean_token_accuracy": 0.8563461780548096,
      "num_tokens": 1121624344.0,
      "step": 33480
    },
    {
      "epoch": 1.998865807067813,
      "grad_norm": 0.44192495942115784,
      "learning_rate": 5.000043842309071e-06,
      "loss": 0.4209,
      "mean_token_accuracy": 0.8587379336357117,
      "num_tokens": 1121792024.0,
      "step": 33485
    },
    {
      "epoch": 1.9991642788920725,
      "grad_norm": 0.4938274621963501,
      "learning_rate": 5.000024661302356e-06,
      "loss": 0.4278,
      "mean_token_accuracy": 0.8574078559875489,
      "num_tokens": 1121959704.0,
      "step": 33490
    },
    {
      "epoch": 1.9994627507163325,
      "grad_norm": 0.4794931411743164,
      "learning_rate": 5.000010960579939e-06,
      "loss": 0.4223,
      "mean_token_accuracy": 0.8573720693588257,
      "num_tokens": 1122127384.0,
      "step": 33495
    },
    {
      "epoch": 1.999761222540592,
      "grad_norm": 0.4582333564758301,
      "learning_rate": 5.0000027401451516e-06,
      "loss": 0.4157,
      "mean_token_accuracy": 0.8597220659255982,
      "num_tokens": 1122295064.0,
      "step": 33500
    },
    {
      "epoch": 2.0,
      "step": 33504,
      "total_flos": 8.833714643953254e+18,
      "train_loss": 0.0,
      "train_runtime": 2.2868,
      "train_samples_per_second": 58602.159,
      "train_steps_per_second": 14651.196
    }
  ],
  "logging_steps": 5,
  "max_steps": 33504,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 2000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.833714643953254e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}