{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 4815,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00020768431983385254,
      "grad_norm": 17.47925567626953,
      "learning_rate": 8.000000000000001e-07,
      "loss": 2.1801,
      "step": 1
    },
    {
      "epoch": 0.0004153686396677051,
      "grad_norm": 21.894914627075195,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 2.249,
      "step": 2
    },
    {
      "epoch": 0.0006230529595015577,
      "grad_norm": 25.090269088745117,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 2.3123,
      "step": 3
    },
    {
      "epoch": 0.0008307372793354102,
      "grad_norm": 18.762964248657227,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 2.1709,
      "step": 4
    },
    {
      "epoch": 0.0010384215991692627,
      "grad_norm": 18.856515884399414,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.1823,
      "step": 5
    },
    {
      "epoch": 0.0012461059190031153,
      "grad_norm": 12.56984806060791,
      "learning_rate": 4.800000000000001e-06,
      "loss": 2.0798,
      "step": 6
    },
    {
      "epoch": 0.0014537902388369677,
      "grad_norm": 8.624156951904297,
      "learning_rate": 5.600000000000001e-06,
      "loss": 1.9833,
      "step": 7
    },
    {
      "epoch": 0.0016614745586708203,
      "grad_norm": 7.026400089263916,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 2.0748,
      "step": 8
    },
    {
      "epoch": 0.001869158878504673,
      "grad_norm": 4.460483074188232,
      "learning_rate": 7.2000000000000005e-06,
      "loss": 1.9404,
      "step": 9
    },
    {
      "epoch": 0.0020768431983385254,
      "grad_norm": 3.6536319255828857,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.0135,
      "step": 10
    },
    {
      "epoch": 0.002284527518172378,
      "grad_norm": 2.5768752098083496,
      "learning_rate": 8.8e-06,
      "loss": 1.8598,
      "step": 11
    },
    {
      "epoch": 0.0024922118380062306,
      "grad_norm": 2.892141103744507,
      "learning_rate": 9.600000000000001e-06,
      "loss": 1.8686,
      "step": 12
    },
    {
      "epoch": 0.0026998961578400833,
      "grad_norm": 2.858818531036377,
      "learning_rate": 1.04e-05,
      "loss": 1.8634,
      "step": 13
    },
    {
      "epoch": 0.0029075804776739354,
      "grad_norm": 2.683091163635254,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 2.0208,
      "step": 14
    },
    {
      "epoch": 0.003115264797507788,
      "grad_norm": 2.7056639194488525,
      "learning_rate": 1.2e-05,
      "loss": 1.9497,
      "step": 15
    },
    {
      "epoch": 0.0033229491173416407,
      "grad_norm": 1.5998775959014893,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 1.8816,
      "step": 16
    },
    {
      "epoch": 0.0035306334371754933,
      "grad_norm": 1.1934099197387695,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 1.8246,
      "step": 17
    },
    {
      "epoch": 0.003738317757009346,
      "grad_norm": 1.1701968908309937,
      "learning_rate": 1.4400000000000001e-05,
      "loss": 1.8578,
      "step": 18
    },
    {
      "epoch": 0.003946002076843198,
      "grad_norm": 1.1254523992538452,
      "learning_rate": 1.5200000000000002e-05,
      "loss": 1.8861,
      "step": 19
    },
    {
      "epoch": 0.004153686396677051,
      "grad_norm": 0.8316693305969238,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 1.7194,
      "step": 20
    },
    {
      "epoch": 0.004361370716510903,
      "grad_norm": 0.8501921892166138,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 1.8474,
      "step": 21
    },
    {
      "epoch": 0.004569055036344756,
      "grad_norm": 0.9284434914588928,
      "learning_rate": 1.76e-05,
      "loss": 1.9258,
      "step": 22
    },
    {
      "epoch": 0.004776739356178609,
      "grad_norm": 0.9898029565811157,
      "learning_rate": 1.8400000000000003e-05,
      "loss": 1.8441,
      "step": 23
    },
    {
      "epoch": 0.004984423676012461,
      "grad_norm": 0.8852891325950623,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 1.8521,
      "step": 24
    },
    {
      "epoch": 0.005192107995846314,
      "grad_norm": 0.9528728723526001,
      "learning_rate": 2e-05,
      "loss": 1.8485,
      "step": 25
    },
    {
      "epoch": 0.0053997923156801665,
      "grad_norm": 0.928016185760498,
      "learning_rate": 2.08e-05,
      "loss": 1.8883,
      "step": 26
    },
    {
      "epoch": 0.005607476635514018,
      "grad_norm": 0.9373347163200378,
      "learning_rate": 2.1600000000000003e-05,
      "loss": 1.891,
      "step": 27
    },
    {
      "epoch": 0.005815160955347871,
      "grad_norm": 0.9534879326820374,
      "learning_rate": 2.2400000000000002e-05,
      "loss": 1.8144,
      "step": 28
    },
    {
      "epoch": 0.0060228452751817235,
      "grad_norm": 0.7779879570007324,
      "learning_rate": 2.32e-05,
      "loss": 1.8153,
      "step": 29
    },
    {
      "epoch": 0.006230529595015576,
      "grad_norm": 0.8465295433998108,
      "learning_rate": 2.4e-05,
      "loss": 1.794,
      "step": 30
    },
    {
      "epoch": 0.006438213914849429,
      "grad_norm": 0.797340452671051,
      "learning_rate": 2.4800000000000003e-05,
      "loss": 1.8324,
      "step": 31
    },
    {
      "epoch": 0.006645898234683281,
      "grad_norm": 0.7843337655067444,
      "learning_rate": 2.5600000000000002e-05,
      "loss": 1.729,
      "step": 32
    },
    {
      "epoch": 0.006853582554517134,
      "grad_norm": 0.7067154049873352,
      "learning_rate": 2.6400000000000005e-05,
      "loss": 1.7337,
      "step": 33
    },
    {
      "epoch": 0.007061266874350987,
      "grad_norm": 0.7495686411857605,
      "learning_rate": 2.7200000000000004e-05,
      "loss": 1.8074,
      "step": 34
    },
    {
      "epoch": 0.007268951194184839,
      "grad_norm": 0.7126153707504272,
      "learning_rate": 2.8e-05,
      "loss": 1.7377,
      "step": 35
    },
    {
      "epoch": 0.007476635514018692,
      "grad_norm": 0.7612162828445435,
      "learning_rate": 2.8800000000000002e-05,
      "loss": 1.775,
      "step": 36
    },
    {
      "epoch": 0.0076843198338525445,
      "grad_norm": 0.7538896799087524,
      "learning_rate": 2.96e-05,
      "loss": 1.7478,
      "step": 37
    },
    {
      "epoch": 0.007892004153686396,
      "grad_norm": 0.8245300650596619,
      "learning_rate": 3.0400000000000004e-05,
      "loss": 1.769,
      "step": 38
    },
    {
      "epoch": 0.00809968847352025,
      "grad_norm": 0.7681639790534973,
      "learning_rate": 3.1200000000000006e-05,
      "loss": 1.7789,
      "step": 39
    },
    {
      "epoch": 0.008307372793354102,
      "grad_norm": 0.7439791560173035,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 1.8245,
      "step": 40
    },
    {
      "epoch": 0.008515057113187955,
      "grad_norm": 0.7001054286956787,
      "learning_rate": 3.28e-05,
      "loss": 1.8516,
      "step": 41
    },
    {
      "epoch": 0.008722741433021807,
      "grad_norm": 0.7122063040733337,
      "learning_rate": 3.3600000000000004e-05,
      "loss": 1.7698,
      "step": 42
    },
    {
      "epoch": 0.008930425752855659,
      "grad_norm": 0.6748693585395813,
      "learning_rate": 3.44e-05,
      "loss": 1.7525,
      "step": 43
    },
    {
      "epoch": 0.009138110072689512,
      "grad_norm": 0.7190465927124023,
      "learning_rate": 3.52e-05,
      "loss": 1.7725,
      "step": 44
    },
    {
      "epoch": 0.009345794392523364,
      "grad_norm": 0.6779818534851074,
      "learning_rate": 3.6e-05,
      "loss": 1.7658,
      "step": 45
    },
    {
      "epoch": 0.009553478712357217,
      "grad_norm": 0.6692732572555542,
      "learning_rate": 3.680000000000001e-05,
      "loss": 1.698,
      "step": 46
    },
    {
      "epoch": 0.009761163032191069,
      "grad_norm": 0.6866817474365234,
      "learning_rate": 3.76e-05,
      "loss": 1.8105,
      "step": 47
    },
    {
      "epoch": 0.009968847352024923,
      "grad_norm": 0.6470504999160767,
      "learning_rate": 3.8400000000000005e-05,
      "loss": 1.6455,
      "step": 48
    },
    {
      "epoch": 0.010176531671858774,
      "grad_norm": 0.681182324886322,
      "learning_rate": 3.9200000000000004e-05,
      "loss": 1.6234,
      "step": 49
    },
    {
      "epoch": 0.010384215991692628,
      "grad_norm": 0.6945406198501587,
      "learning_rate": 4e-05,
      "loss": 1.8115,
      "step": 50
    },
    {
      "epoch": 0.01059190031152648,
      "grad_norm": 0.635296642780304,
      "learning_rate": 4.08e-05,
      "loss": 1.752,
      "step": 51
    },
    {
      "epoch": 0.010799584631360333,
      "grad_norm": 0.6995029449462891,
      "learning_rate": 4.16e-05,
      "loss": 1.7869,
      "step": 52
    },
    {
      "epoch": 0.011007268951194185,
      "grad_norm": 0.6692876815795898,
      "learning_rate": 4.240000000000001e-05,
      "loss": 1.7784,
      "step": 53
    },
    {
      "epoch": 0.011214953271028037,
      "grad_norm": 0.6663792133331299,
      "learning_rate": 4.3200000000000007e-05,
      "loss": 1.7742,
      "step": 54
    },
    {
      "epoch": 0.01142263759086189,
      "grad_norm": 0.6469627022743225,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 1.7418,
      "step": 55
    },
    {
      "epoch": 0.011630321910695742,
      "grad_norm": 0.6327781677246094,
      "learning_rate": 4.4800000000000005e-05,
      "loss": 1.7479,
      "step": 56
    },
    {
      "epoch": 0.011838006230529595,
      "grad_norm": 0.6531626582145691,
      "learning_rate": 4.56e-05,
      "loss": 1.7121,
      "step": 57
    },
    {
      "epoch": 0.012045690550363447,
      "grad_norm": 0.6819879412651062,
      "learning_rate": 4.64e-05,
      "loss": 1.8296,
      "step": 58
    },
    {
      "epoch": 0.0122533748701973,
      "grad_norm": 0.6116477847099304,
      "learning_rate": 4.72e-05,
      "loss": 1.6922,
      "step": 59
    },
    {
      "epoch": 0.012461059190031152,
      "grad_norm": 0.6569270491600037,
      "learning_rate": 4.8e-05,
      "loss": 1.7661,
      "step": 60
    },
    {
      "epoch": 0.012668743509865006,
      "grad_norm": 0.6796997785568237,
      "learning_rate": 4.88e-05,
      "loss": 1.7959,
      "step": 61
    },
    {
      "epoch": 0.012876427829698858,
      "grad_norm": 0.7139971256256104,
      "learning_rate": 4.9600000000000006e-05,
      "loss": 1.7764,
      "step": 62
    },
    {
      "epoch": 0.013084112149532711,
      "grad_norm": 0.7374599575996399,
      "learning_rate": 5.0400000000000005e-05,
      "loss": 1.7271,
      "step": 63
    },
    {
      "epoch": 0.013291796469366563,
      "grad_norm": 0.6610848307609558,
      "learning_rate": 5.1200000000000004e-05,
      "loss": 1.8526,
      "step": 64
    },
    {
      "epoch": 0.013499480789200415,
      "grad_norm": 0.742107093334198,
      "learning_rate": 5.2000000000000004e-05,
      "loss": 1.754,
      "step": 65
    },
    {
      "epoch": 0.013707165109034268,
      "grad_norm": 0.6820899844169617,
      "learning_rate": 5.280000000000001e-05,
      "loss": 1.7355,
      "step": 66
    },
    {
      "epoch": 0.01391484942886812,
      "grad_norm": 0.649463951587677,
      "learning_rate": 5.360000000000001e-05,
      "loss": 1.6453,
      "step": 67
    },
    {
      "epoch": 0.014122533748701973,
      "grad_norm": 0.7077696323394775,
      "learning_rate": 5.440000000000001e-05,
      "loss": 1.7962,
      "step": 68
    },
    {
      "epoch": 0.014330218068535825,
      "grad_norm": 0.6177412867546082,
      "learning_rate": 5.52e-05,
      "loss": 1.6742,
      "step": 69
    },
    {
      "epoch": 0.014537902388369679,
      "grad_norm": 0.7839769721031189,
      "learning_rate": 5.6e-05,
      "loss": 1.8364,
      "step": 70
    },
    {
      "epoch": 0.01474558670820353,
      "grad_norm": 0.6374268531799316,
      "learning_rate": 5.6800000000000005e-05,
      "loss": 1.7423,
      "step": 71
    },
    {
      "epoch": 0.014953271028037384,
      "grad_norm": 0.6933043599128723,
      "learning_rate": 5.7600000000000004e-05,
      "loss": 1.7669,
      "step": 72
    },
    {
      "epoch": 0.015160955347871236,
      "grad_norm": 0.641764760017395,
      "learning_rate": 5.84e-05,
      "loss": 1.7328,
      "step": 73
    },
    {
      "epoch": 0.015368639667705089,
      "grad_norm": 0.6187850832939148,
      "learning_rate": 5.92e-05,
      "loss": 1.6375,
      "step": 74
    },
    {
      "epoch": 0.01557632398753894,
      "grad_norm": 0.6676532030105591,
      "learning_rate": 6.000000000000001e-05,
      "loss": 1.7784,
      "step": 75
    },
    {
      "epoch": 0.015784008307372793,
      "grad_norm": 0.6391186118125916,
      "learning_rate": 6.080000000000001e-05,
      "loss": 1.6554,
      "step": 76
    },
    {
      "epoch": 0.015991692627206646,
      "grad_norm": 0.6301543116569519,
      "learning_rate": 6.16e-05,
      "loss": 1.9057,
      "step": 77
    },
    {
      "epoch": 0.0161993769470405,
      "grad_norm": 0.6136651039123535,
      "learning_rate": 6.240000000000001e-05,
      "loss": 1.6371,
      "step": 78
    },
    {
      "epoch": 0.01640706126687435,
      "grad_norm": 0.6152656674385071,
      "learning_rate": 6.32e-05,
      "loss": 1.7339,
      "step": 79
    },
    {
      "epoch": 0.016614745586708203,
      "grad_norm": 0.6085280179977417,
      "learning_rate": 6.400000000000001e-05,
      "loss": 1.7106,
      "step": 80
    },
    {
      "epoch": 0.016822429906542057,
      "grad_norm": 0.6468527913093567,
      "learning_rate": 6.48e-05,
      "loss": 1.7993,
      "step": 81
    },
    {
      "epoch": 0.01703011422637591,
      "grad_norm": 0.6175693273544312,
      "learning_rate": 6.56e-05,
      "loss": 1.7434,
      "step": 82
    },
    {
      "epoch": 0.01723779854620976,
      "grad_norm": 0.5871699452400208,
      "learning_rate": 6.64e-05,
      "loss": 1.6838,
      "step": 83
    },
    {
      "epoch": 0.017445482866043614,
      "grad_norm": 0.6958571672439575,
      "learning_rate": 6.720000000000001e-05,
      "loss": 1.7389,
      "step": 84
    },
    {
      "epoch": 0.017653167185877467,
      "grad_norm": 0.5976506471633911,
      "learning_rate": 6.8e-05,
      "loss": 1.7073,
      "step": 85
    },
    {
      "epoch": 0.017860851505711317,
      "grad_norm": 0.6843888163566589,
      "learning_rate": 6.88e-05,
      "loss": 1.7339,
      "step": 86
    },
    {
      "epoch": 0.01806853582554517,
      "grad_norm": 0.5901737213134766,
      "learning_rate": 6.960000000000001e-05,
      "loss": 1.7577,
      "step": 87
    },
    {
      "epoch": 0.018276220145379024,
      "grad_norm": 0.6460241675376892,
      "learning_rate": 7.04e-05,
      "loss": 1.7482,
      "step": 88
    },
    {
      "epoch": 0.018483904465212878,
      "grad_norm": 0.6465390920639038,
      "learning_rate": 7.120000000000001e-05,
      "loss": 1.8085,
      "step": 89
    },
    {
      "epoch": 0.018691588785046728,
      "grad_norm": 0.6187453866004944,
      "learning_rate": 7.2e-05,
      "loss": 1.7507,
      "step": 90
    },
    {
      "epoch": 0.01889927310488058,
      "grad_norm": 0.6101999282836914,
      "learning_rate": 7.280000000000001e-05,
      "loss": 1.7488,
      "step": 91
    },
    {
      "epoch": 0.019106957424714435,
      "grad_norm": 0.5893586874008179,
      "learning_rate": 7.360000000000001e-05,
      "loss": 1.6846,
      "step": 92
    },
    {
      "epoch": 0.019314641744548288,
      "grad_norm": 0.6104580163955688,
      "learning_rate": 7.44e-05,
      "loss": 1.6573,
      "step": 93
    },
    {
      "epoch": 0.019522326064382138,
      "grad_norm": 0.6451858878135681,
      "learning_rate": 7.52e-05,
      "loss": 1.8052,
      "step": 94
    },
    {
      "epoch": 0.01973001038421599,
      "grad_norm": 0.5865317583084106,
      "learning_rate": 7.6e-05,
      "loss": 1.6534,
      "step": 95
    },
    {
      "epoch": 0.019937694704049845,
      "grad_norm": 0.6278175711631775,
      "learning_rate": 7.680000000000001e-05,
      "loss": 1.6898,
      "step": 96
    },
    {
      "epoch": 0.020145379023883695,
      "grad_norm": 0.5943008661270142,
      "learning_rate": 7.76e-05,
      "loss": 1.7049,
      "step": 97
    },
    {
      "epoch": 0.02035306334371755,
      "grad_norm": 0.6195578575134277,
      "learning_rate": 7.840000000000001e-05,
      "loss": 1.7527,
      "step": 98
    },
    {
      "epoch": 0.020560747663551402,
      "grad_norm": 0.6871982216835022,
      "learning_rate": 7.92e-05,
      "loss": 1.708,
      "step": 99
    },
    {
      "epoch": 0.020768431983385256,
      "grad_norm": 0.6266918182373047,
      "learning_rate": 8e-05,
      "loss": 1.781,
      "step": 100
    },
    {
      "epoch": 0.020976116303219106,
      "grad_norm": 0.7398074269294739,
      "learning_rate": 8e-05,
      "loss": 1.7166,
      "step": 101
    },
    {
      "epoch": 0.02118380062305296,
      "grad_norm": 0.6563434600830078,
      "learning_rate": 8e-05,
      "loss": 1.8624,
      "step": 102
    },
    {
      "epoch": 0.021391484942886813,
      "grad_norm": 0.7264277338981628,
      "learning_rate": 8e-05,
      "loss": 1.7253,
      "step": 103
    },
    {
      "epoch": 0.021599169262720666,
      "grad_norm": 0.6289424896240234,
      "learning_rate": 8e-05,
      "loss": 1.7794,
      "step": 104
    },
    {
      "epoch": 0.021806853582554516,
      "grad_norm": 0.6097893714904785,
      "learning_rate": 8e-05,
      "loss": 1.7288,
      "step": 105
    },
    {
      "epoch": 0.02201453790238837,
      "grad_norm": 0.5839325785636902,
      "learning_rate": 8e-05,
      "loss": 1.6222,
      "step": 106
    },
    {
      "epoch": 0.022222222222222223,
      "grad_norm": 0.5372262001037598,
      "learning_rate": 8e-05,
      "loss": 1.6274,
      "step": 107
    },
    {
      "epoch": 0.022429906542056073,
      "grad_norm": 0.5895355343818665,
      "learning_rate": 8e-05,
      "loss": 1.7382,
      "step": 108
    },
    {
      "epoch": 0.022637590861889927,
      "grad_norm": 0.6392791867256165,
      "learning_rate": 8e-05,
      "loss": 1.7517,
      "step": 109
    },
    {
      "epoch": 0.02284527518172378,
      "grad_norm": 0.6434172987937927,
      "learning_rate": 8e-05,
      "loss": 1.7384,
      "step": 110
    },
    {
      "epoch": 0.023052959501557634,
      "grad_norm": 0.5970573425292969,
      "learning_rate": 8e-05,
      "loss": 1.7531,
      "step": 111
    },
    {
      "epoch": 0.023260643821391484,
      "grad_norm": 0.5773294568061829,
      "learning_rate": 8e-05,
      "loss": 1.6521,
      "step": 112
    },
    {
      "epoch": 0.023468328141225337,
      "grad_norm": 0.6489471197128296,
      "learning_rate": 8e-05,
      "loss": 1.7838,
      "step": 113
    },
    {
      "epoch": 0.02367601246105919,
      "grad_norm": 0.6172572374343872,
      "learning_rate": 8e-05,
      "loss": 1.7552,
      "step": 114
    },
    {
      "epoch": 0.023883696780893044,
      "grad_norm": 0.5731709003448486,
      "learning_rate": 8e-05,
      "loss": 1.6842,
      "step": 115
    },
    {
      "epoch": 0.024091381100726894,
      "grad_norm": 0.5704280138015747,
      "learning_rate": 8e-05,
      "loss": 1.7002,
      "step": 116
    },
    {
      "epoch": 0.024299065420560748,
      "grad_norm": 0.5774348378181458,
      "learning_rate": 8e-05,
      "loss": 1.7487,
      "step": 117
    },
    {
      "epoch": 0.0245067497403946,
      "grad_norm": 0.5606811046600342,
      "learning_rate": 8e-05,
      "loss": 1.6928,
      "step": 118
    },
    {
      "epoch": 0.02471443406022845,
      "grad_norm": 0.5781872868537903,
      "learning_rate": 8e-05,
      "loss": 1.7106,
      "step": 119
    },
    {
      "epoch": 0.024922118380062305,
      "grad_norm": 0.5436711311340332,
      "learning_rate": 8e-05,
      "loss": 1.6909,
      "step": 120
    },
    {
      "epoch": 0.025129802699896158,
      "grad_norm": 0.5731891393661499,
      "learning_rate": 8e-05,
      "loss": 1.6783,
      "step": 121
    },
    {
      "epoch": 0.02533748701973001,
      "grad_norm": 0.5782831907272339,
      "learning_rate": 8e-05,
      "loss": 1.715,
      "step": 122
    },
    {
      "epoch": 0.02554517133956386,
      "grad_norm": 0.6293315291404724,
      "learning_rate": 8e-05,
      "loss": 1.7052,
      "step": 123
    },
    {
      "epoch": 0.025752855659397715,
      "grad_norm": 0.5766347646713257,
      "learning_rate": 8e-05,
      "loss": 1.7466,
      "step": 124
    },
    {
      "epoch": 0.02596053997923157,
      "grad_norm": 0.6577507257461548,
      "learning_rate": 8e-05,
      "loss": 1.7412,
      "step": 125
    },
    {
      "epoch": 0.026168224299065422,
      "grad_norm": 0.5595094561576843,
      "learning_rate": 8e-05,
      "loss": 1.7464,
      "step": 126
    },
    {
      "epoch": 0.026375908618899272,
      "grad_norm": 0.5962421894073486,
      "learning_rate": 8e-05,
      "loss": 1.6721,
      "step": 127
    },
    {
      "epoch": 0.026583592938733126,
      "grad_norm": 0.5711417198181152,
      "learning_rate": 8e-05,
      "loss": 1.7608,
      "step": 128
    },
    {
      "epoch": 0.02679127725856698,
      "grad_norm": 0.6157316565513611,
      "learning_rate": 8e-05,
      "loss": 1.7212,
      "step": 129
    },
    {
      "epoch": 0.02699896157840083,
      "grad_norm": 0.5418056845664978,
      "learning_rate": 8e-05,
      "loss": 1.6652,
      "step": 130
    },
    {
      "epoch": 0.027206645898234683,
      "grad_norm": 0.5753270983695984,
      "learning_rate": 8e-05,
      "loss": 1.7269,
      "step": 131
    },
    {
      "epoch": 0.027414330218068536,
      "grad_norm": 0.5521407127380371,
      "learning_rate": 8e-05,
      "loss": 1.7053,
      "step": 132
    },
    {
      "epoch": 0.02762201453790239,
      "grad_norm": 0.5795088410377502,
      "learning_rate": 8e-05,
      "loss": 1.8084,
      "step": 133
    },
    {
      "epoch": 0.02782969885773624,
      "grad_norm": 0.5171025991439819,
      "learning_rate": 8e-05,
      "loss": 1.6639,
      "step": 134
    },
    {
      "epoch": 0.028037383177570093,
      "grad_norm": 0.5902402997016907,
      "learning_rate": 8e-05,
      "loss": 1.6855,
      "step": 135
    },
    {
      "epoch": 0.028245067497403947,
      "grad_norm": 0.535606861114502,
      "learning_rate": 8e-05,
      "loss": 1.7269,
      "step": 136
    },
    {
      "epoch": 0.0284527518172378,
      "grad_norm": 0.61818528175354,
      "learning_rate": 8e-05,
      "loss": 1.7777,
      "step": 137
    },
    {
      "epoch": 0.02866043613707165,
      "grad_norm": 0.5806450843811035,
      "learning_rate": 8e-05,
      "loss": 1.738,
      "step": 138
    },
    {
      "epoch": 0.028868120456905504,
      "grad_norm": 0.5649462342262268,
      "learning_rate": 8e-05,
      "loss": 1.6814,
      "step": 139
    },
    {
      "epoch": 0.029075804776739357,
      "grad_norm": 0.5286867022514343,
      "learning_rate": 8e-05,
      "loss": 1.6813,
      "step": 140
    },
    {
      "epoch": 0.029283489096573207,
      "grad_norm": 0.5503925085067749,
      "learning_rate": 8e-05,
      "loss": 1.6656,
      "step": 141
    },
    {
      "epoch": 0.02949117341640706,
      "grad_norm": 0.5422264933586121,
      "learning_rate": 8e-05,
      "loss": 1.6465,
      "step": 142
    },
    {
      "epoch": 0.029698857736240914,
      "grad_norm": 0.5629477500915527,
      "learning_rate": 8e-05,
      "loss": 1.7515,
      "step": 143
    },
    {
      "epoch": 0.029906542056074768,
      "grad_norm": 0.5476765036582947,
      "learning_rate": 8e-05,
      "loss": 1.7234,
      "step": 144
    },
    {
      "epoch": 0.030114226375908618,
      "grad_norm": 0.5577101111412048,
      "learning_rate": 8e-05,
      "loss": 1.6463,
      "step": 145
    },
    {
      "epoch": 0.03032191069574247,
      "grad_norm": 0.566056489944458,
      "learning_rate": 8e-05,
      "loss": 1.7592,
      "step": 146
    },
    {
      "epoch": 0.030529595015576325,
      "grad_norm": 0.5325253009796143,
      "learning_rate": 8e-05,
      "loss": 1.7202,
      "step": 147
    },
    {
      "epoch": 0.030737279335410178,
      "grad_norm": 0.569216787815094,
      "learning_rate": 8e-05,
      "loss": 1.7454,
      "step": 148
    },
    {
      "epoch": 0.030944963655244028,
      "grad_norm": 0.5541402101516724,
      "learning_rate": 8e-05,
      "loss": 1.5909,
      "step": 149
    },
    {
      "epoch": 0.03115264797507788,
      "grad_norm": 0.5762909650802612,
      "learning_rate": 8e-05,
      "loss": 1.7057,
      "step": 150
    },
    {
      "epoch": 0.03136033229491173,
      "grad_norm": 0.5259131193161011,
      "learning_rate": 8e-05,
      "loss": 1.624,
      "step": 151
    },
    {
      "epoch": 0.031568016614745585,
      "grad_norm": 0.5526155233383179,
      "learning_rate": 8e-05,
      "loss": 1.6645,
      "step": 152
    },
    {
      "epoch": 0.03177570093457944,
      "grad_norm": 0.5420238375663757,
      "learning_rate": 8e-05,
      "loss": 1.6899,
      "step": 153
    },
    {
      "epoch": 0.03198338525441329,
      "grad_norm": 0.5291162133216858,
      "learning_rate": 8e-05,
      "loss": 1.7062,
      "step": 154
    },
    {
      "epoch": 0.032191069574247146,
      "grad_norm": 0.5605945587158203,
      "learning_rate": 8e-05,
      "loss": 1.7297,
      "step": 155
    },
    {
      "epoch": 0.032398753894081,
      "grad_norm": 0.5243666768074036,
      "learning_rate": 8e-05,
      "loss": 1.6641,
      "step": 156
    },
    {
      "epoch": 0.03260643821391485,
      "grad_norm": 0.5971183180809021,
      "learning_rate": 8e-05,
      "loss": 1.6927,
      "step": 157
    },
    {
      "epoch": 0.0328141225337487,
      "grad_norm": 0.5431556105613708,
      "learning_rate": 8e-05,
      "loss": 1.6636,
      "step": 158
    },
    {
      "epoch": 0.03302180685358255,
      "grad_norm": 0.5140684247016907,
      "learning_rate": 8e-05,
      "loss": 1.6315,
      "step": 159
    },
    {
      "epoch": 0.033229491173416406,
      "grad_norm": 0.518284022808075,
      "learning_rate": 8e-05,
      "loss": 1.6613,
      "step": 160
    },
    {
      "epoch": 0.03343717549325026,
      "grad_norm": 0.5419729948043823,
      "learning_rate": 8e-05,
      "loss": 1.7693,
      "step": 161
    },
    {
      "epoch": 0.03364485981308411,
      "grad_norm": 0.5171647667884827,
      "learning_rate": 8e-05,
      "loss": 1.6841,
      "step": 162
    },
    {
      "epoch": 0.03385254413291797,
      "grad_norm": 0.5692399740219116,
      "learning_rate": 8e-05,
      "loss": 1.7286,
      "step": 163
    },
    {
      "epoch": 0.03406022845275182,
      "grad_norm": 0.5294033288955688,
      "learning_rate": 8e-05,
      "loss": 1.7233,
      "step": 164
    },
    {
      "epoch": 0.03426791277258567,
      "grad_norm": 0.5355773568153381,
      "learning_rate": 8e-05,
      "loss": 1.748,
      "step": 165
    },
    {
      "epoch": 0.03447559709241952,
      "grad_norm": 0.5389161705970764,
      "learning_rate": 8e-05,
      "loss": 1.7231,
      "step": 166
    },
    {
      "epoch": 0.034683281412253374,
      "grad_norm": 0.5242868065834045,
      "learning_rate": 8e-05,
      "loss": 1.759,
      "step": 167
    },
    {
      "epoch": 0.03489096573208723,
      "grad_norm": 0.5554731488227844,
      "learning_rate": 8e-05,
      "loss": 1.7271,
      "step": 168
    },
    {
      "epoch": 0.03509865005192108,
      "grad_norm": 0.603803277015686,
      "learning_rate": 8e-05,
      "loss": 1.7537,
      "step": 169
    },
    {
      "epoch": 0.035306334371754934,
      "grad_norm": 0.5199013948440552,
      "learning_rate": 8e-05,
      "loss": 1.6379,
      "step": 170
    },
    {
      "epoch": 0.03551401869158879,
      "grad_norm": 0.5927119851112366,
      "learning_rate": 8e-05,
      "loss": 1.6706,
      "step": 171
    },
    {
      "epoch": 0.035721703011422634,
      "grad_norm": 0.5246160626411438,
      "learning_rate": 8e-05,
      "loss": 1.6838,
      "step": 172
    },
    {
      "epoch": 0.03592938733125649,
      "grad_norm": 0.568761944770813,
      "learning_rate": 8e-05,
      "loss": 1.6608,
      "step": 173
    },
    {
      "epoch": 0.03613707165109034,
      "grad_norm": 0.5335773825645447,
      "learning_rate": 8e-05,
      "loss": 1.6861,
      "step": 174
    },
    {
      "epoch": 0.036344755970924195,
      "grad_norm": 0.5661566257476807,
      "learning_rate": 8e-05,
      "loss": 1.7069,
      "step": 175
    },
    {
      "epoch": 0.03655244029075805,
      "grad_norm": 0.550425112247467,
      "learning_rate": 8e-05,
      "loss": 1.7824,
      "step": 176
    },
    {
      "epoch": 0.0367601246105919,
      "grad_norm": 0.5691404938697815,
      "learning_rate": 8e-05,
      "loss": 1.7234,
      "step": 177
    },
    {
      "epoch": 0.036967808930425755,
      "grad_norm": 0.5029447078704834,
      "learning_rate": 8e-05,
      "loss": 1.592,
      "step": 178
    },
    {
      "epoch": 0.03717549325025961,
      "grad_norm": 0.5434035062789917,
      "learning_rate": 8e-05,
      "loss": 1.76,
      "step": 179
    },
    {
      "epoch": 0.037383177570093455,
      "grad_norm": 0.5171444416046143,
      "learning_rate": 8e-05,
      "loss": 1.6538,
      "step": 180
    },
    {
      "epoch": 0.03759086188992731,
      "grad_norm": 0.5830038785934448,
      "learning_rate": 8e-05,
      "loss": 1.8091,
      "step": 181
    },
    {
      "epoch": 0.03779854620976116,
      "grad_norm": 0.5217128992080688,
      "learning_rate": 8e-05,
      "loss": 1.7283,
      "step": 182
    },
    {
      "epoch": 0.038006230529595016,
      "grad_norm": 0.6341305375099182,
      "learning_rate": 8e-05,
      "loss": 1.7947,
      "step": 183
    },
    {
      "epoch": 0.03821391484942887,
      "grad_norm": 0.5161294341087341,
      "learning_rate": 8e-05,
      "loss": 1.6634,
      "step": 184
    },
    {
      "epoch": 0.03842159916926272,
      "grad_norm": 0.6006560325622559,
      "learning_rate": 8e-05,
      "loss": 1.6709,
      "step": 185
    },
    {
      "epoch": 0.038629283489096576,
      "grad_norm": 0.5234457850456238,
      "learning_rate": 8e-05,
      "loss": 1.695,
      "step": 186
    },
    {
      "epoch": 0.03883696780893042,
      "grad_norm": 0.5401941537857056,
      "learning_rate": 8e-05,
      "loss": 1.6625,
      "step": 187
    },
    {
      "epoch": 0.039044652128764276,
      "grad_norm": 0.5211032032966614,
      "learning_rate": 8e-05,
      "loss": 1.6295,
      "step": 188
    },
    {
      "epoch": 0.03925233644859813,
      "grad_norm": 0.5172737240791321,
      "learning_rate": 8e-05,
      "loss": 1.5994,
      "step": 189
    },
    {
      "epoch": 0.03946002076843198,
      "grad_norm": 0.5083449482917786,
      "learning_rate": 8e-05,
      "loss": 1.6803,
      "step": 190
    },
    {
      "epoch": 0.03966770508826584,
      "grad_norm": 0.5440491437911987,
      "learning_rate": 8e-05,
      "loss": 1.7009,
      "step": 191
    },
    {
      "epoch": 0.03987538940809969,
      "grad_norm": 0.6146543622016907,
      "learning_rate": 8e-05,
      "loss": 1.7213,
      "step": 192
    },
    {
      "epoch": 0.040083073727933544,
      "grad_norm": 0.5229283571243286,
      "learning_rate": 8e-05,
      "loss": 1.6891,
      "step": 193
    },
    {
      "epoch": 0.04029075804776739,
      "grad_norm": 0.6130139827728271,
      "learning_rate": 8e-05,
      "loss": 1.7485,
      "step": 194
    },
    {
      "epoch": 0.040498442367601244,
      "grad_norm": 0.5257254242897034,
      "learning_rate": 8e-05,
      "loss": 1.6668,
      "step": 195
    },
    {
      "epoch": 0.0407061266874351,
      "grad_norm": 0.5981298089027405,
      "learning_rate": 8e-05,
      "loss": 1.7177,
      "step": 196
    },
    {
      "epoch": 0.04091381100726895,
      "grad_norm": 0.48841404914855957,
      "learning_rate": 8e-05,
      "loss": 1.6452,
      "step": 197
    },
    {
      "epoch": 0.041121495327102804,
      "grad_norm": 0.6227656006813049,
      "learning_rate": 8e-05,
      "loss": 1.6978,
      "step": 198
    },
    {
      "epoch": 0.04132917964693666,
      "grad_norm": 0.5224297046661377,
      "learning_rate": 8e-05,
      "loss": 1.7488,
      "step": 199
    },
    {
      "epoch": 0.04153686396677051,
      "grad_norm": 0.6549935936927795,
      "learning_rate": 8e-05,
      "loss": 1.759,
      "step": 200
    },
    {
      "epoch": 0.041744548286604365,
      "grad_norm": 0.5142560601234436,
      "learning_rate": 8e-05,
      "loss": 1.6678,
      "step": 201
    },
    {
      "epoch": 0.04195223260643821,
      "grad_norm": 0.53005051612854,
      "learning_rate": 8e-05,
      "loss": 1.6766,
      "step": 202
    },
    {
      "epoch": 0.042159916926272065,
      "grad_norm": 0.5193365812301636,
      "learning_rate": 8e-05,
      "loss": 1.6739,
      "step": 203
    },
    {
      "epoch": 0.04236760124610592,
      "grad_norm": 0.5400128364562988,
      "learning_rate": 8e-05,
      "loss": 1.7213,
      "step": 204
    },
    {
      "epoch": 0.04257528556593977,
      "grad_norm": 0.5410746932029724,
      "learning_rate": 8e-05,
      "loss": 1.721,
      "step": 205
    },
    {
      "epoch": 0.042782969885773625,
      "grad_norm": 0.4970126450061798,
      "learning_rate": 8e-05,
      "loss": 1.6326,
      "step": 206
    },
    {
      "epoch": 0.04299065420560748,
      "grad_norm": 0.5962429642677307,
      "learning_rate": 8e-05,
      "loss": 1.7096,
      "step": 207
    },
    {
      "epoch": 0.04319833852544133,
      "grad_norm": 0.5132887363433838,
      "learning_rate": 8e-05,
      "loss": 1.7047,
      "step": 208
    },
    {
      "epoch": 0.04340602284527518,
      "grad_norm": 0.5727942585945129,
      "learning_rate": 8e-05,
      "loss": 1.6485,
      "step": 209
    },
    {
      "epoch": 0.04361370716510903,
      "grad_norm": 0.516907811164856,
      "learning_rate": 8e-05,
      "loss": 1.693,
      "step": 210
    },
    {
      "epoch": 0.043821391484942886,
      "grad_norm": 0.5103829503059387,
      "learning_rate": 8e-05,
      "loss": 1.6644,
      "step": 211
    },
    {
      "epoch": 0.04402907580477674,
      "grad_norm": 0.5416673421859741,
      "learning_rate": 8e-05,
      "loss": 1.7322,
      "step": 212
    },
    {
      "epoch": 0.04423676012461059,
      "grad_norm": 0.4991842806339264,
      "learning_rate": 8e-05,
      "loss": 1.6948,
      "step": 213
    },
    {
      "epoch": 0.044444444444444446,
      "grad_norm": 0.5187154412269592,
      "learning_rate": 8e-05,
      "loss": 1.5997,
      "step": 214
    },
    {
      "epoch": 0.0446521287642783,
      "grad_norm": 0.5057274699211121,
      "learning_rate": 8e-05,
      "loss": 1.7124,
      "step": 215
    },
    {
      "epoch": 0.044859813084112146,
      "grad_norm": 0.522167980670929,
      "learning_rate": 8e-05,
      "loss": 1.646,
      "step": 216
    },
    {
      "epoch": 0.045067497403946,
      "grad_norm": 0.5210246443748474,
      "learning_rate": 8e-05,
      "loss": 1.7006,
      "step": 217
    },
    {
      "epoch": 0.04527518172377985,
      "grad_norm": 0.5107594728469849,
      "learning_rate": 8e-05,
      "loss": 1.6886,
      "step": 218
    },
    {
      "epoch": 0.04548286604361371,
      "grad_norm": 0.546195387840271,
      "learning_rate": 8e-05,
      "loss": 1.6922,
      "step": 219
    },
    {
      "epoch": 0.04569055036344756,
      "grad_norm": 0.5224754214286804,
      "learning_rate": 8e-05,
      "loss": 1.7163,
      "step": 220
    },
    {
      "epoch": 0.045898234683281414,
      "grad_norm": 0.538343071937561,
      "learning_rate": 8e-05,
      "loss": 1.6758,
      "step": 221
    },
    {
      "epoch": 0.04610591900311527,
      "grad_norm": 0.5105881094932556,
      "learning_rate": 8e-05,
      "loss": 1.7446,
      "step": 222
    },
    {
      "epoch": 0.04631360332294912,
      "grad_norm": 0.5250230431556702,
      "learning_rate": 8e-05,
      "loss": 1.7269,
      "step": 223
    },
    {
      "epoch": 0.04652128764278297,
      "grad_norm": 0.5123667120933533,
      "learning_rate": 8e-05,
      "loss": 1.6593,
      "step": 224
    },
    {
      "epoch": 0.04672897196261682,
      "grad_norm": 0.5394571423530579,
      "learning_rate": 8e-05,
      "loss": 1.6677,
      "step": 225
    },
    {
      "epoch": 0.046936656282450674,
      "grad_norm": 0.5013480186462402,
      "learning_rate": 8e-05,
      "loss": 1.6936,
      "step": 226
    },
    {
      "epoch": 0.04714434060228453,
      "grad_norm": 0.501785933971405,
      "learning_rate": 8e-05,
      "loss": 1.5588,
      "step": 227
    },
    {
      "epoch": 0.04735202492211838,
      "grad_norm": 0.5234807729721069,
      "learning_rate": 8e-05,
      "loss": 1.7372,
      "step": 228
    },
    {
      "epoch": 0.047559709241952235,
      "grad_norm": 0.5221506357192993,
      "learning_rate": 8e-05,
      "loss": 1.6345,
      "step": 229
    },
    {
      "epoch": 0.04776739356178609,
      "grad_norm": 0.5281847715377808,
      "learning_rate": 8e-05,
      "loss": 1.6074,
      "step": 230
    },
    {
      "epoch": 0.047975077881619935,
      "grad_norm": 0.5333276987075806,
      "learning_rate": 8e-05,
      "loss": 1.6483,
      "step": 231
    },
    {
      "epoch": 0.04818276220145379,
      "grad_norm": 0.49470800161361694,
      "learning_rate": 8e-05,
      "loss": 1.6387,
      "step": 232
    },
    {
      "epoch": 0.04839044652128764,
      "grad_norm": 0.5513780117034912,
      "learning_rate": 8e-05,
      "loss": 1.6956,
      "step": 233
    },
    {
      "epoch": 0.048598130841121495,
      "grad_norm": 0.48948371410369873,
      "learning_rate": 8e-05,
      "loss": 1.6548,
      "step": 234
    },
    {
      "epoch": 0.04880581516095535,
      "grad_norm": 0.5753876566886902,
      "learning_rate": 8e-05,
      "loss": 1.7143,
      "step": 235
    },
    {
      "epoch": 0.0490134994807892,
      "grad_norm": 0.5774050951004028,
      "learning_rate": 8e-05,
      "loss": 1.6995,
      "step": 236
    },
    {
      "epoch": 0.049221183800623056,
      "grad_norm": 0.5836580395698547,
      "learning_rate": 8e-05,
      "loss": 1.75,
      "step": 237
    },
    {
      "epoch": 0.0494288681204569,
      "grad_norm": 0.6421523094177246,
      "learning_rate": 8e-05,
      "loss": 1.7579,
      "step": 238
    },
    {
      "epoch": 0.049636552440290756,
      "grad_norm": 0.5229347944259644,
      "learning_rate": 8e-05,
      "loss": 1.6735,
      "step": 239
    },
    {
      "epoch": 0.04984423676012461,
      "grad_norm": 0.6658037304878235,
      "learning_rate": 8e-05,
      "loss": 1.7808,
      "step": 240
    },
    {
      "epoch": 0.05005192107995846,
      "grad_norm": 0.5098143219947815,
      "learning_rate": 8e-05,
      "loss": 1.6961,
      "step": 241
    },
    {
      "epoch": 0.050259605399792316,
      "grad_norm": 0.5492331385612488,
      "learning_rate": 8e-05,
      "loss": 1.6589,
      "step": 242
    },
    {
      "epoch": 0.05046728971962617,
      "grad_norm": 0.5122767686843872,
      "learning_rate": 8e-05,
      "loss": 1.688,
      "step": 243
    },
    {
      "epoch": 0.05067497403946002,
      "grad_norm": 0.5612162351608276,
      "learning_rate": 8e-05,
      "loss": 1.7156,
      "step": 244
    },
    {
      "epoch": 0.05088265835929388,
      "grad_norm": 0.5456447601318359,
      "learning_rate": 8e-05,
      "loss": 1.668,
      "step": 245
    },
    {
      "epoch": 0.05109034267912772,
      "grad_norm": 0.5987516641616821,
      "learning_rate": 8e-05,
      "loss": 1.7434,
      "step": 246
    },
    {
      "epoch": 0.05129802699896158,
      "grad_norm": 0.5334711074829102,
      "learning_rate": 8e-05,
      "loss": 1.6684,
      "step": 247
    },
    {
      "epoch": 0.05150571131879543,
      "grad_norm": 0.5441988706588745,
      "learning_rate": 8e-05,
      "loss": 1.667,
      "step": 248
    },
    {
      "epoch": 0.051713395638629284,
      "grad_norm": 0.5460416078567505,
      "learning_rate": 8e-05,
      "loss": 1.6615,
      "step": 249
    },
    {
      "epoch": 0.05192107995846314,
      "grad_norm": 0.5425359010696411,
      "learning_rate": 8e-05,
      "loss": 1.6637,
      "step": 250
    },
    {
      "epoch": 0.05212876427829699,
      "grad_norm": 0.5430195927619934,
      "learning_rate": 8e-05,
      "loss": 1.6521,
      "step": 251
    },
    {
      "epoch": 0.052336448598130844,
      "grad_norm": 0.4934813678264618,
      "learning_rate": 8e-05,
      "loss": 1.7507,
      "step": 252
    },
    {
      "epoch": 0.05254413291796469,
      "grad_norm": 0.5606212019920349,
      "learning_rate": 8e-05,
      "loss": 1.754,
      "step": 253
    },
    {
      "epoch": 0.052751817237798544,
      "grad_norm": 0.5214331746101379,
      "learning_rate": 8e-05,
      "loss": 1.676,
      "step": 254
    },
    {
      "epoch": 0.0529595015576324,
      "grad_norm": 0.60882568359375,
      "learning_rate": 8e-05,
      "loss": 1.7517,
      "step": 255
    },
    {
      "epoch": 0.05316718587746625,
      "grad_norm": 0.5168995261192322,
      "learning_rate": 8e-05,
      "loss": 1.6832,
      "step": 256
    },
    {
      "epoch": 0.053374870197300105,
      "grad_norm": 0.5288822650909424,
      "learning_rate": 8e-05,
      "loss": 1.6817,
      "step": 257
    },
    {
      "epoch": 0.05358255451713396,
      "grad_norm": 0.563677966594696,
      "learning_rate": 8e-05,
      "loss": 1.7,
      "step": 258
    },
    {
      "epoch": 0.05379023883696781,
      "grad_norm": 0.5071162581443787,
      "learning_rate": 8e-05,
      "loss": 1.72,
      "step": 259
    },
    {
      "epoch": 0.05399792315680166,
      "grad_norm": 0.5226413607597351,
      "learning_rate": 8e-05,
      "loss": 1.6503,
      "step": 260
    },
    {
      "epoch": 0.05420560747663551,
      "grad_norm": 0.5036090016365051,
      "learning_rate": 8e-05,
      "loss": 1.671,
      "step": 261
    },
    {
      "epoch": 0.054413291796469365,
      "grad_norm": 0.5291279554367065,
      "learning_rate": 8e-05,
      "loss": 1.6547,
      "step": 262
    },
    {
      "epoch": 0.05462097611630322,
      "grad_norm": 0.5152414441108704,
      "learning_rate": 8e-05,
      "loss": 1.7235,
      "step": 263
    },
    {
      "epoch": 0.05482866043613707,
      "grad_norm": 0.5221415758132935,
      "learning_rate": 8e-05,
      "loss": 1.7344,
      "step": 264
    },
    {
      "epoch": 0.055036344755970926,
      "grad_norm": 0.5111775398254395,
      "learning_rate": 8e-05,
      "loss": 1.6625,
      "step": 265
    },
    {
      "epoch": 0.05524402907580478,
      "grad_norm": 0.49026650190353394,
      "learning_rate": 8e-05,
      "loss": 1.626,
      "step": 266
    },
    {
      "epoch": 0.05545171339563863,
      "grad_norm": 0.5113672614097595,
      "learning_rate": 8e-05,
      "loss": 1.7152,
      "step": 267
    },
    {
      "epoch": 0.05565939771547248,
      "grad_norm": 0.5143501162528992,
      "learning_rate": 8e-05,
      "loss": 1.6065,
      "step": 268
    },
    {
      "epoch": 0.05586708203530633,
      "grad_norm": 0.548305094242096,
      "learning_rate": 8e-05,
      "loss": 1.7169,
      "step": 269
    },
    {
      "epoch": 0.056074766355140186,
      "grad_norm": 0.5203267931938171,
      "learning_rate": 8e-05,
      "loss": 1.7894,
      "step": 270
    },
    {
      "epoch": 0.05628245067497404,
      "grad_norm": 0.5775648951530457,
      "learning_rate": 8e-05,
      "loss": 1.6821,
      "step": 271
    },
    {
      "epoch": 0.05649013499480789,
      "grad_norm": 0.48842960596084595,
      "learning_rate": 8e-05,
      "loss": 1.6095,
      "step": 272
    },
    {
      "epoch": 0.05669781931464175,
      "grad_norm": 0.5582982301712036,
      "learning_rate": 8e-05,
      "loss": 1.7071,
      "step": 273
    },
    {
      "epoch": 0.0569055036344756,
      "grad_norm": 0.4931994378566742,
      "learning_rate": 8e-05,
      "loss": 1.6754,
      "step": 274
    },
    {
      "epoch": 0.05711318795430945,
      "grad_norm": 0.5403690934181213,
      "learning_rate": 8e-05,
      "loss": 1.6986,
      "step": 275
    },
    {
      "epoch": 0.0573208722741433,
      "grad_norm": 0.5050585269927979,
      "learning_rate": 8e-05,
      "loss": 1.7073,
      "step": 276
    },
    {
      "epoch": 0.057528556593977154,
      "grad_norm": 0.5011170506477356,
      "learning_rate": 8e-05,
      "loss": 1.6561,
      "step": 277
    },
    {
      "epoch": 0.05773624091381101,
      "grad_norm": 0.5033249258995056,
      "learning_rate": 8e-05,
      "loss": 1.7003,
      "step": 278
    },
    {
      "epoch": 0.05794392523364486,
      "grad_norm": 0.5251483917236328,
      "learning_rate": 8e-05,
      "loss": 1.6675,
      "step": 279
    },
    {
      "epoch": 0.058151609553478714,
      "grad_norm": 0.5164880752563477,
      "learning_rate": 8e-05,
      "loss": 1.7132,
      "step": 280
    },
    {
      "epoch": 0.05835929387331257,
      "grad_norm": 0.4889441728591919,
      "learning_rate": 8e-05,
      "loss": 1.7148,
      "step": 281
    },
    {
      "epoch": 0.058566978193146414,
      "grad_norm": 0.49667251110076904,
      "learning_rate": 8e-05,
      "loss": 1.7136,
      "step": 282
    },
    {
      "epoch": 0.05877466251298027,
      "grad_norm": 0.5121577382087708,
      "learning_rate": 8e-05,
      "loss": 1.764,
      "step": 283
    },
    {
      "epoch": 0.05898234683281412,
      "grad_norm": 0.48562902212142944,
      "learning_rate": 8e-05,
      "loss": 1.7124,
      "step": 284
    },
    {
      "epoch": 0.059190031152647975,
      "grad_norm": 0.4913182556629181,
      "learning_rate": 8e-05,
      "loss": 1.6498,
      "step": 285
    },
    {
      "epoch": 0.05939771547248183,
      "grad_norm": 0.5064699053764343,
      "learning_rate": 8e-05,
      "loss": 1.598,
      "step": 286
    },
    {
      "epoch": 0.05960539979231568,
      "grad_norm": 0.553032398223877,
      "learning_rate": 8e-05,
      "loss": 1.705,
      "step": 287
    },
    {
      "epoch": 0.059813084112149535,
      "grad_norm": 0.5344775319099426,
      "learning_rate": 8e-05,
      "loss": 1.7229,
      "step": 288
    },
    {
      "epoch": 0.06002076843198339,
      "grad_norm": 0.5126497149467468,
      "learning_rate": 8e-05,
      "loss": 1.7099,
      "step": 289
    },
    {
      "epoch": 0.060228452751817235,
      "grad_norm": 0.4790019690990448,
      "learning_rate": 8e-05,
      "loss": 1.6468,
      "step": 290
    },
    {
      "epoch": 0.06043613707165109,
      "grad_norm": 0.5178213715553284,
      "learning_rate": 8e-05,
      "loss": 1.729,
      "step": 291
    },
    {
      "epoch": 0.06064382139148494,
      "grad_norm": 0.4989386796951294,
      "learning_rate": 8e-05,
      "loss": 1.7297,
      "step": 292
    },
    {
      "epoch": 0.060851505711318796,
      "grad_norm": 0.49952176213264465,
      "learning_rate": 8e-05,
      "loss": 1.7531,
      "step": 293
    },
    {
      "epoch": 0.06105919003115265,
      "grad_norm": 0.5082775354385376,
      "learning_rate": 8e-05,
      "loss": 1.7071,
      "step": 294
    },
    {
      "epoch": 0.0612668743509865,
      "grad_norm": 0.4975671172142029,
      "learning_rate": 8e-05,
      "loss": 1.6984,
      "step": 295
    },
    {
      "epoch": 0.061474558670820356,
      "grad_norm": 0.5003031492233276,
      "learning_rate": 8e-05,
      "loss": 1.657,
      "step": 296
    },
    {
      "epoch": 0.0616822429906542,
      "grad_norm": 0.524743378162384,
      "learning_rate": 8e-05,
      "loss": 1.6763,
      "step": 297
    },
    {
      "epoch": 0.061889927310488056,
      "grad_norm": 0.5136753916740417,
      "learning_rate": 8e-05,
      "loss": 1.6642,
      "step": 298
    },
    {
      "epoch": 0.06209761163032191,
      "grad_norm": 0.55498868227005,
      "learning_rate": 8e-05,
      "loss": 1.7189,
      "step": 299
    },
    {
      "epoch": 0.06230529595015576,
      "grad_norm": 0.468551367521286,
      "learning_rate": 8e-05,
      "loss": 1.6853,
      "step": 300
    },
    {
      "epoch": 0.06251298026998961,
      "grad_norm": 0.5169484615325928,
      "learning_rate": 8e-05,
      "loss": 1.6934,
      "step": 301
    },
    {
      "epoch": 0.06272066458982346,
      "grad_norm": 0.4962407052516937,
      "learning_rate": 8e-05,
      "loss": 1.67,
      "step": 302
    },
    {
      "epoch": 0.06292834890965732,
      "grad_norm": 0.5252827405929565,
      "learning_rate": 8e-05,
      "loss": 1.7051,
      "step": 303
    },
    {
      "epoch": 0.06313603322949117,
      "grad_norm": 0.49051743745803833,
      "learning_rate": 8e-05,
      "loss": 1.6435,
      "step": 304
    },
    {
      "epoch": 0.06334371754932502,
      "grad_norm": 0.49855688214302063,
      "learning_rate": 8e-05,
      "loss": 1.6876,
      "step": 305
    },
    {
      "epoch": 0.06355140186915888,
      "grad_norm": 0.5097089409828186,
      "learning_rate": 8e-05,
      "loss": 1.7275,
      "step": 306
    },
    {
      "epoch": 0.06375908618899273,
      "grad_norm": 0.4844972491264343,
      "learning_rate": 8e-05,
      "loss": 1.596,
      "step": 307
    },
    {
      "epoch": 0.06396677050882658,
      "grad_norm": 0.4773125946521759,
      "learning_rate": 8e-05,
      "loss": 1.6302,
      "step": 308
    },
    {
      "epoch": 0.06417445482866044,
      "grad_norm": 0.4860425293445587,
      "learning_rate": 8e-05,
      "loss": 1.6042,
      "step": 309
    },
    {
      "epoch": 0.06438213914849429,
      "grad_norm": 0.49147120118141174,
      "learning_rate": 8e-05,
      "loss": 1.6496,
      "step": 310
    },
    {
      "epoch": 0.06458982346832814,
      "grad_norm": 0.4904959797859192,
      "learning_rate": 8e-05,
      "loss": 1.6267,
      "step": 311
    },
    {
      "epoch": 0.064797507788162,
      "grad_norm": 0.5017173886299133,
      "learning_rate": 8e-05,
      "loss": 1.6827,
      "step": 312
    },
    {
      "epoch": 0.06500519210799585,
      "grad_norm": 0.5119773149490356,
      "learning_rate": 8e-05,
      "loss": 1.7423,
      "step": 313
    },
    {
      "epoch": 0.0652128764278297,
      "grad_norm": 0.5187619924545288,
      "learning_rate": 8e-05,
      "loss": 1.6919,
      "step": 314
    },
    {
      "epoch": 0.06542056074766354,
      "grad_norm": 0.499828577041626,
      "learning_rate": 8e-05,
      "loss": 1.6802,
      "step": 315
    },
    {
      "epoch": 0.0656282450674974,
      "grad_norm": 0.5605705976486206,
      "learning_rate": 8e-05,
      "loss": 1.7527,
      "step": 316
    },
    {
      "epoch": 0.06583592938733125,
      "grad_norm": 0.477637380361557,
      "learning_rate": 8e-05,
      "loss": 1.6605,
      "step": 317
    },
    {
      "epoch": 0.0660436137071651,
      "grad_norm": 0.5356731414794922,
      "learning_rate": 8e-05,
      "loss": 1.6564,
      "step": 318
    },
    {
      "epoch": 0.06625129802699896,
      "grad_norm": 0.4828336834907532,
      "learning_rate": 8e-05,
      "loss": 1.5889,
      "step": 319
    },
    {
      "epoch": 0.06645898234683281,
      "grad_norm": 0.5663031339645386,
      "learning_rate": 8e-05,
      "loss": 1.7158,
      "step": 320
    },
    {
      "epoch": 0.06666666666666667,
      "grad_norm": 0.5123764872550964,
      "learning_rate": 8e-05,
      "loss": 1.6212,
      "step": 321
    },
    {
      "epoch": 0.06687435098650052,
      "grad_norm": 0.5193995833396912,
      "learning_rate": 8e-05,
      "loss": 1.6672,
      "step": 322
    },
    {
      "epoch": 0.06708203530633437,
      "grad_norm": 0.612371563911438,
      "learning_rate": 8e-05,
      "loss": 1.7297,
      "step": 323
    },
    {
      "epoch": 0.06728971962616823,
      "grad_norm": 0.5306211709976196,
      "learning_rate": 8e-05,
      "loss": 1.5693,
      "step": 324
    },
    {
      "epoch": 0.06749740394600208,
      "grad_norm": 0.5609772801399231,
      "learning_rate": 8e-05,
      "loss": 1.6011,
      "step": 325
    },
    {
      "epoch": 0.06770508826583593,
      "grad_norm": 0.5140113234519958,
      "learning_rate": 8e-05,
      "loss": 1.6519,
      "step": 326
    },
    {
      "epoch": 0.06791277258566979,
      "grad_norm": 0.5529564023017883,
      "learning_rate": 8e-05,
      "loss": 1.6217,
      "step": 327
    },
    {
      "epoch": 0.06812045690550364,
      "grad_norm": 0.48979297280311584,
      "learning_rate": 8e-05,
      "loss": 1.6044,
      "step": 328
    },
    {
      "epoch": 0.0683281412253375,
      "grad_norm": 0.5272225141525269,
      "learning_rate": 8e-05,
      "loss": 1.7409,
      "step": 329
    },
    {
      "epoch": 0.06853582554517133,
      "grad_norm": 0.5034517645835876,
      "learning_rate": 8e-05,
      "loss": 1.6181,
      "step": 330
    },
    {
      "epoch": 0.06874350986500519,
      "grad_norm": 0.4756111800670624,
      "learning_rate": 8e-05,
      "loss": 1.6358,
      "step": 331
    },
    {
      "epoch": 0.06895119418483904,
      "grad_norm": 0.5277131199836731,
      "learning_rate": 8e-05,
      "loss": 1.7289,
      "step": 332
    },
    {
      "epoch": 0.0691588785046729,
      "grad_norm": 0.4864356815814972,
      "learning_rate": 8e-05,
      "loss": 1.6573,
      "step": 333
    },
    {
      "epoch": 0.06936656282450675,
      "grad_norm": 0.49576079845428467,
      "learning_rate": 8e-05,
      "loss": 1.7073,
      "step": 334
    },
    {
      "epoch": 0.0695742471443406,
      "grad_norm": 0.4868834614753723,
      "learning_rate": 8e-05,
      "loss": 1.6662,
      "step": 335
    },
    {
      "epoch": 0.06978193146417445,
      "grad_norm": 0.5043144822120667,
      "learning_rate": 8e-05,
      "loss": 1.6672,
      "step": 336
    },
    {
      "epoch": 0.06998961578400831,
      "grad_norm": 0.49576571583747864,
      "learning_rate": 8e-05,
      "loss": 1.7137,
      "step": 337
    },
    {
      "epoch": 0.07019730010384216,
      "grad_norm": 0.4854160249233246,
      "learning_rate": 8e-05,
      "loss": 1.6409,
      "step": 338
    },
    {
      "epoch": 0.07040498442367601,
      "grad_norm": 0.48695647716522217,
      "learning_rate": 8e-05,
      "loss": 1.6114,
      "step": 339
    },
    {
      "epoch": 0.07061266874350987,
      "grad_norm": 0.4941887855529785,
      "learning_rate": 8e-05,
      "loss": 1.6843,
      "step": 340
    },
    {
      "epoch": 0.07082035306334372,
      "grad_norm": 0.49328118562698364,
      "learning_rate": 8e-05,
      "loss": 1.6837,
      "step": 341
    },
    {
      "epoch": 0.07102803738317758,
      "grad_norm": 0.4951527714729309,
      "learning_rate": 8e-05,
      "loss": 1.6745,
      "step": 342
    },
    {
      "epoch": 0.07123572170301143,
      "grad_norm": 0.49303266406059265,
      "learning_rate": 8e-05,
      "loss": 1.6726,
      "step": 343
    },
    {
      "epoch": 0.07144340602284527,
      "grad_norm": 0.48809322714805603,
      "learning_rate": 8e-05,
      "loss": 1.6716,
      "step": 344
    },
    {
      "epoch": 0.07165109034267912,
      "grad_norm": 0.5064594745635986,
      "learning_rate": 8e-05,
      "loss": 1.7361,
      "step": 345
    },
    {
      "epoch": 0.07185877466251298,
      "grad_norm": 0.5080369710922241,
      "learning_rate": 8e-05,
      "loss": 1.6929,
      "step": 346
    },
    {
      "epoch": 0.07206645898234683,
      "grad_norm": 0.4928899109363556,
      "learning_rate": 8e-05,
      "loss": 1.7132,
      "step": 347
    },
    {
      "epoch": 0.07227414330218068,
      "grad_norm": 0.47712448239326477,
      "learning_rate": 8e-05,
      "loss": 1.6334,
      "step": 348
    },
    {
      "epoch": 0.07248182762201454,
      "grad_norm": 0.513994038105011,
      "learning_rate": 8e-05,
      "loss": 1.699,
      "step": 349
    },
    {
      "epoch": 0.07268951194184839,
      "grad_norm": 0.5016348958015442,
      "learning_rate": 8e-05,
      "loss": 1.634,
      "step": 350
    },
    {
      "epoch": 0.07289719626168224,
      "grad_norm": 0.5038484334945679,
      "learning_rate": 8e-05,
      "loss": 1.6643,
      "step": 351
    },
    {
      "epoch": 0.0731048805815161,
      "grad_norm": 0.4928259551525116,
      "learning_rate": 8e-05,
      "loss": 1.6646,
      "step": 352
    },
    {
      "epoch": 0.07331256490134995,
      "grad_norm": 0.5103161334991455,
      "learning_rate": 8e-05,
      "loss": 1.6494,
      "step": 353
    },
    {
      "epoch": 0.0735202492211838,
      "grad_norm": 0.47103309631347656,
      "learning_rate": 8e-05,
      "loss": 1.6299,
      "step": 354
    },
    {
      "epoch": 0.07372793354101766,
      "grad_norm": 0.5037011504173279,
      "learning_rate": 8e-05,
      "loss": 1.6283,
      "step": 355
    },
    {
      "epoch": 0.07393561786085151,
      "grad_norm": 0.5026373267173767,
      "learning_rate": 8e-05,
      "loss": 1.7157,
      "step": 356
    },
    {
      "epoch": 0.07414330218068536,
      "grad_norm": 0.5053650140762329,
      "learning_rate": 8e-05,
      "loss": 1.6587,
      "step": 357
    },
    {
      "epoch": 0.07435098650051922,
      "grad_norm": 0.496429443359375,
      "learning_rate": 8e-05,
      "loss": 1.7041,
      "step": 358
    },
    {
      "epoch": 0.07455867082035306,
      "grad_norm": 0.47459688782691956,
      "learning_rate": 8e-05,
      "loss": 1.6136,
      "step": 359
    },
    {
      "epoch": 0.07476635514018691,
      "grad_norm": 0.5140553116798401,
      "learning_rate": 8e-05,
      "loss": 1.6625,
      "step": 360
    },
    {
      "epoch": 0.07497403946002076,
      "grad_norm": 0.515044093132019,
      "learning_rate": 8e-05,
      "loss": 1.711,
      "step": 361
    },
    {
      "epoch": 0.07518172377985462,
      "grad_norm": 0.5119950771331787,
      "learning_rate": 8e-05,
      "loss": 1.6852,
      "step": 362
    },
    {
      "epoch": 0.07538940809968847,
      "grad_norm": 0.5062432289123535,
      "learning_rate": 8e-05,
      "loss": 1.6892,
      "step": 363
    },
    {
      "epoch": 0.07559709241952232,
      "grad_norm": 0.5517569780349731,
      "learning_rate": 8e-05,
      "loss": 1.6883,
      "step": 364
    },
    {
      "epoch": 0.07580477673935618,
      "grad_norm": 0.48322999477386475,
      "learning_rate": 8e-05,
      "loss": 1.6383,
      "step": 365
    },
    {
      "epoch": 0.07601246105919003,
      "grad_norm": 0.47941911220550537,
      "learning_rate": 8e-05,
      "loss": 1.6032,
      "step": 366
    },
    {
      "epoch": 0.07622014537902388,
      "grad_norm": 0.4892639219760895,
      "learning_rate": 8e-05,
      "loss": 1.6862,
      "step": 367
    },
    {
      "epoch": 0.07642782969885774,
      "grad_norm": 0.48457542061805725,
      "learning_rate": 8e-05,
      "loss": 1.6253,
      "step": 368
    },
    {
      "epoch": 0.07663551401869159,
      "grad_norm": 0.48371532559394836,
      "learning_rate": 8e-05,
      "loss": 1.7072,
      "step": 369
    },
    {
      "epoch": 0.07684319833852545,
      "grad_norm": 0.5162315964698792,
      "learning_rate": 8e-05,
      "loss": 1.6952,
      "step": 370
    },
    {
      "epoch": 0.0770508826583593,
      "grad_norm": 0.494991660118103,
      "learning_rate": 8e-05,
      "loss": 1.7292,
      "step": 371
    },
    {
      "epoch": 0.07725856697819315,
      "grad_norm": 0.510187029838562,
      "learning_rate": 8e-05,
      "loss": 1.6835,
      "step": 372
    },
    {
      "epoch": 0.077466251298027,
      "grad_norm": 0.47694623470306396,
      "learning_rate": 8e-05,
      "loss": 1.6809,
      "step": 373
    },
    {
      "epoch": 0.07767393561786085,
      "grad_norm": 0.4950973093509674,
      "learning_rate": 8e-05,
      "loss": 1.6691,
      "step": 374
    },
    {
      "epoch": 0.0778816199376947,
      "grad_norm": 0.4885307848453522,
      "learning_rate": 8e-05,
      "loss": 1.6682,
      "step": 375
    },
    {
      "epoch": 0.07808930425752855,
      "grad_norm": 0.5185424089431763,
      "learning_rate": 8e-05,
      "loss": 1.698,
      "step": 376
    },
    {
      "epoch": 0.0782969885773624,
      "grad_norm": 0.48990127444267273,
      "learning_rate": 8e-05,
      "loss": 1.6877,
      "step": 377
    },
    {
      "epoch": 0.07850467289719626,
      "grad_norm": 0.5352708101272583,
      "learning_rate": 8e-05,
      "loss": 1.6141,
      "step": 378
    },
    {
      "epoch": 0.07871235721703011,
      "grad_norm": 0.48971933126449585,
      "learning_rate": 8e-05,
      "loss": 1.6644,
      "step": 379
    },
    {
      "epoch": 0.07892004153686397,
      "grad_norm": 0.5039674639701843,
      "learning_rate": 8e-05,
      "loss": 1.6772,
      "step": 380
    },
    {
      "epoch": 0.07912772585669782,
      "grad_norm": 0.5058568120002747,
      "learning_rate": 8e-05,
      "loss": 1.6109,
      "step": 381
    },
    {
      "epoch": 0.07933541017653167,
      "grad_norm": 0.5172670483589172,
      "learning_rate": 8e-05,
      "loss": 1.6229,
      "step": 382
    },
    {
      "epoch": 0.07954309449636553,
      "grad_norm": 0.4776594340801239,
      "learning_rate": 8e-05,
      "loss": 1.6517,
      "step": 383
    },
    {
      "epoch": 0.07975077881619938,
      "grad_norm": 0.5346738696098328,
      "learning_rate": 8e-05,
      "loss": 1.6985,
      "step": 384
    },
    {
      "epoch": 0.07995846313603323,
      "grad_norm": 0.5013498067855835,
      "learning_rate": 8e-05,
      "loss": 1.6297,
      "step": 385
    },
    {
      "epoch": 0.08016614745586709,
      "grad_norm": 0.6052170991897583,
      "learning_rate": 8e-05,
      "loss": 1.6397,
      "step": 386
    },
    {
      "epoch": 0.08037383177570094,
      "grad_norm": 0.4749467074871063,
      "learning_rate": 8e-05,
      "loss": 1.6263,
      "step": 387
    },
    {
      "epoch": 0.08058151609553478,
      "grad_norm": 0.4913139045238495,
      "learning_rate": 8e-05,
      "loss": 1.6091,
      "step": 388
    },
    {
      "epoch": 0.08078920041536863,
      "grad_norm": 0.5050566792488098,
      "learning_rate": 8e-05,
      "loss": 1.6801,
      "step": 389
    },
    {
      "epoch": 0.08099688473520249,
      "grad_norm": 0.4727949798107147,
      "learning_rate": 8e-05,
      "loss": 1.6141,
      "step": 390
    },
    {
      "epoch": 0.08120456905503634,
      "grad_norm": 0.4729914367198944,
      "learning_rate": 8e-05,
      "loss": 1.5569,
      "step": 391
    },
    {
      "epoch": 0.0814122533748702,
      "grad_norm": 0.4925561249256134,
      "learning_rate": 8e-05,
      "loss": 1.7113,
      "step": 392
    },
    {
      "epoch": 0.08161993769470405,
      "grad_norm": 0.49586695432662964,
      "learning_rate": 8e-05,
      "loss": 1.6699,
      "step": 393
    },
    {
      "epoch": 0.0818276220145379,
      "grad_norm": 0.5150527954101562,
      "learning_rate": 8e-05,
      "loss": 1.6564,
      "step": 394
    },
    {
      "epoch": 0.08203530633437175,
      "grad_norm": 0.5019596219062805,
      "learning_rate": 8e-05,
      "loss": 1.6275,
      "step": 395
    },
    {
      "epoch": 0.08224299065420561,
      "grad_norm": 0.4876788258552551,
      "learning_rate": 8e-05,
      "loss": 1.5836,
      "step": 396
    },
    {
      "epoch": 0.08245067497403946,
      "grad_norm": 0.49811601638793945,
      "learning_rate": 8e-05,
      "loss": 1.6608,
      "step": 397
    },
    {
      "epoch": 0.08265835929387332,
      "grad_norm": 0.5187938809394836,
      "learning_rate": 8e-05,
      "loss": 1.6999,
      "step": 398
    },
    {
      "epoch": 0.08286604361370717,
      "grad_norm": 0.4808737635612488,
      "learning_rate": 8e-05,
      "loss": 1.6004,
      "step": 399
    },
    {
      "epoch": 0.08307372793354102,
      "grad_norm": 0.515282392501831,
      "learning_rate": 8e-05,
      "loss": 1.7455,
      "step": 400
    },
    {
      "epoch": 0.08328141225337488,
      "grad_norm": 0.5063252449035645,
      "learning_rate": 8e-05,
      "loss": 1.6955,
      "step": 401
    },
    {
      "epoch": 0.08348909657320873,
      "grad_norm": 0.5026871562004089,
      "learning_rate": 8e-05,
      "loss": 1.7067,
      "step": 402
    },
    {
      "epoch": 0.08369678089304257,
      "grad_norm": 0.4903314709663391,
      "learning_rate": 8e-05,
      "loss": 1.6333,
      "step": 403
    },
    {
      "epoch": 0.08390446521287642,
      "grad_norm": 0.4954453110694885,
      "learning_rate": 8e-05,
      "loss": 1.5555,
      "step": 404
    },
    {
      "epoch": 0.08411214953271028,
      "grad_norm": 0.5026134252548218,
      "learning_rate": 8e-05,
      "loss": 1.6969,
      "step": 405
    },
    {
      "epoch": 0.08431983385254413,
      "grad_norm": 0.49587976932525635,
      "learning_rate": 8e-05,
      "loss": 1.651,
      "step": 406
    },
    {
      "epoch": 0.08452751817237798,
      "grad_norm": 0.5106424689292908,
      "learning_rate": 8e-05,
      "loss": 1.7021,
      "step": 407
    },
    {
      "epoch": 0.08473520249221184,
      "grad_norm": 0.5195350646972656,
      "learning_rate": 8e-05,
      "loss": 1.7539,
      "step": 408
    },
    {
      "epoch": 0.08494288681204569,
      "grad_norm": 0.501416027545929,
      "learning_rate": 8e-05,
      "loss": 1.725,
      "step": 409
    },
    {
      "epoch": 0.08515057113187954,
      "grad_norm": 0.48874419927597046,
      "learning_rate": 8e-05,
      "loss": 1.7037,
      "step": 410
    },
    {
      "epoch": 0.0853582554517134,
      "grad_norm": 0.4698718190193176,
      "learning_rate": 8e-05,
      "loss": 1.6509,
      "step": 411
    },
    {
      "epoch": 0.08556593977154725,
      "grad_norm": 0.4783203899860382,
      "learning_rate": 8e-05,
      "loss": 1.6171,
      "step": 412
    },
    {
      "epoch": 0.0857736240913811,
      "grad_norm": 0.481812983751297,
      "learning_rate": 8e-05,
      "loss": 1.6676,
      "step": 413
    },
    {
      "epoch": 0.08598130841121496,
      "grad_norm": 0.481001615524292,
      "learning_rate": 8e-05,
      "loss": 1.7201,
      "step": 414
    },
    {
      "epoch": 0.08618899273104881,
      "grad_norm": 0.48555779457092285,
      "learning_rate": 8e-05,
      "loss": 1.6573,
      "step": 415
    },
    {
      "epoch": 0.08639667705088266,
      "grad_norm": 0.49405309557914734,
      "learning_rate": 8e-05,
      "loss": 1.6688,
      "step": 416
    },
    {
      "epoch": 0.08660436137071652,
      "grad_norm": 0.5051506757736206,
      "learning_rate": 8e-05,
      "loss": 1.7261,
      "step": 417
    },
    {
      "epoch": 0.08681204569055036,
      "grad_norm": 0.48117661476135254,
      "learning_rate": 8e-05,
      "loss": 1.6696,
      "step": 418
    },
    {
      "epoch": 0.08701973001038421,
      "grad_norm": 0.4655582904815674,
      "learning_rate": 8e-05,
      "loss": 1.5525,
      "step": 419
    },
    {
      "epoch": 0.08722741433021806,
      "grad_norm": 0.4815155863761902,
      "learning_rate": 8e-05,
      "loss": 1.6482,
      "step": 420
    },
    {
      "epoch": 0.08743509865005192,
      "grad_norm": 0.4939725697040558,
      "learning_rate": 8e-05,
      "loss": 1.6752,
      "step": 421
    },
    {
      "epoch": 0.08764278296988577,
      "grad_norm": 0.4941619634628296,
      "learning_rate": 8e-05,
      "loss": 1.658,
      "step": 422
    },
    {
      "epoch": 0.08785046728971962,
      "grad_norm": 0.5332744121551514,
      "learning_rate": 8e-05,
      "loss": 1.6118,
      "step": 423
    },
    {
      "epoch": 0.08805815160955348,
      "grad_norm": 0.5089534521102905,
      "learning_rate": 8e-05,
      "loss": 1.6835,
      "step": 424
    },
    {
      "epoch": 0.08826583592938733,
      "grad_norm": 0.5659358501434326,
      "learning_rate": 8e-05,
      "loss": 1.646,
      "step": 425
    },
    {
      "epoch": 0.08847352024922119,
      "grad_norm": 0.49219921231269836,
      "learning_rate": 8e-05,
      "loss": 1.6452,
      "step": 426
    },
    {
      "epoch": 0.08868120456905504,
      "grad_norm": 0.5574254989624023,
      "learning_rate": 8e-05,
      "loss": 1.6968,
      "step": 427
    },
    {
      "epoch": 0.08888888888888889,
      "grad_norm": 0.4780307412147522,
      "learning_rate": 8e-05,
      "loss": 1.6332,
      "step": 428
    },
    {
      "epoch": 0.08909657320872275,
      "grad_norm": 0.4689270555973053,
      "learning_rate": 8e-05,
      "loss": 1.6342,
      "step": 429
    },
    {
      "epoch": 0.0893042575285566,
      "grad_norm": 0.5344724655151367,
      "learning_rate": 8e-05,
      "loss": 1.6307,
      "step": 430
    },
    {
      "epoch": 0.08951194184839045,
      "grad_norm": 0.4806777536869049,
      "learning_rate": 8e-05,
      "loss": 1.6404,
      "step": 431
    },
    {
      "epoch": 0.08971962616822429,
      "grad_norm": 0.5253303050994873,
      "learning_rate": 8e-05,
      "loss": 1.6228,
      "step": 432
    },
    {
      "epoch": 0.08992731048805815,
      "grad_norm": 0.49155253171920776,
      "learning_rate": 8e-05,
      "loss": 1.6423,
      "step": 433
    },
    {
      "epoch": 0.090134994807892,
      "grad_norm": 0.5021010637283325,
      "learning_rate": 8e-05,
      "loss": 1.6646,
      "step": 434
    },
    {
      "epoch": 0.09034267912772585,
      "grad_norm": 0.5126837491989136,
      "learning_rate": 8e-05,
      "loss": 1.6719,
      "step": 435
    },
    {
      "epoch": 0.0905503634475597,
      "grad_norm": 0.48987722396850586,
      "learning_rate": 8e-05,
      "loss": 1.6355,
      "step": 436
    },
    {
      "epoch": 0.09075804776739356,
      "grad_norm": 0.46434321999549866,
      "learning_rate": 8e-05,
      "loss": 1.6236,
      "step": 437
    },
    {
      "epoch": 0.09096573208722741,
      "grad_norm": 0.4966569244861603,
      "learning_rate": 8e-05,
      "loss": 1.7529,
      "step": 438
    },
    {
      "epoch": 0.09117341640706127,
      "grad_norm": 0.5068960189819336,
      "learning_rate": 8e-05,
      "loss": 1.718,
      "step": 439
    },
    {
      "epoch": 0.09138110072689512,
      "grad_norm": 0.46872174739837646,
      "learning_rate": 8e-05,
      "loss": 1.5913,
      "step": 440
    },
    {
      "epoch": 0.09158878504672897,
      "grad_norm": 0.5093649625778198,
      "learning_rate": 8e-05,
      "loss": 1.7232,
      "step": 441
    },
    {
      "epoch": 0.09179646936656283,
      "grad_norm": 0.48676973581314087,
      "learning_rate": 8e-05,
      "loss": 1.6004,
      "step": 442
    },
    {
      "epoch": 0.09200415368639668,
      "grad_norm": 0.5160446763038635,
      "learning_rate": 8e-05,
      "loss": 1.7038,
      "step": 443
    },
    {
      "epoch": 0.09221183800623053,
      "grad_norm": 0.4958758056163788,
      "learning_rate": 8e-05,
      "loss": 1.6985,
      "step": 444
    },
    {
      "epoch": 0.09241952232606439,
      "grad_norm": 0.5180954337120056,
      "learning_rate": 8e-05,
      "loss": 1.636,
      "step": 445
    },
    {
      "epoch": 0.09262720664589824,
      "grad_norm": 0.5121625661849976,
      "learning_rate": 8e-05,
      "loss": 1.6144,
      "step": 446
    },
    {
      "epoch": 0.09283489096573208,
      "grad_norm": 0.49673452973365784,
      "learning_rate": 8e-05,
      "loss": 1.5977,
      "step": 447
    },
    {
      "epoch": 0.09304257528556593,
      "grad_norm": 0.4997375011444092,
      "learning_rate": 8e-05,
      "loss": 1.686,
      "step": 448
    },
    {
      "epoch": 0.09325025960539979,
      "grad_norm": 0.49643149971961975,
      "learning_rate": 8e-05,
      "loss": 1.6721,
      "step": 449
    },
    {
      "epoch": 0.09345794392523364,
      "grad_norm": 0.4948190152645111,
      "learning_rate": 8e-05,
      "loss": 1.657,
      "step": 450
    },
    {
      "epoch": 0.0936656282450675,
      "grad_norm": 0.4922389090061188,
      "learning_rate": 8e-05,
      "loss": 1.65,
      "step": 451
    },
    {
      "epoch": 0.09387331256490135,
      "grad_norm": 0.5069667100906372,
      "learning_rate": 8e-05,
      "loss": 1.7523,
      "step": 452
    },
    {
      "epoch": 0.0940809968847352,
      "grad_norm": 0.4961357116699219,
      "learning_rate": 8e-05,
      "loss": 1.752,
      "step": 453
    },
    {
      "epoch": 0.09428868120456906,
      "grad_norm": 0.5226089358329773,
      "learning_rate": 8e-05,
      "loss": 1.6604,
      "step": 454
    },
    {
      "epoch": 0.09449636552440291,
      "grad_norm": 0.47510433197021484,
      "learning_rate": 8e-05,
      "loss": 1.7006,
      "step": 455
    },
    {
      "epoch": 0.09470404984423676,
      "grad_norm": 0.5109766125679016,
      "learning_rate": 8e-05,
      "loss": 1.7016,
      "step": 456
    },
    {
      "epoch": 0.09491173416407062,
      "grad_norm": 0.4836071729660034,
      "learning_rate": 8e-05,
      "loss": 1.6091,
      "step": 457
    },
    {
      "epoch": 0.09511941848390447,
      "grad_norm": 0.523279070854187,
      "learning_rate": 8e-05,
      "loss": 1.6313,
      "step": 458
    },
    {
      "epoch": 0.09532710280373832,
      "grad_norm": 0.5089949369430542,
      "learning_rate": 8e-05,
      "loss": 1.6241,
      "step": 459
    },
    {
      "epoch": 0.09553478712357218,
      "grad_norm": 0.509610652923584,
      "learning_rate": 8e-05,
      "loss": 1.6721,
      "step": 460
    },
    {
      "epoch": 0.09574247144340603,
      "grad_norm": 0.4768628776073456,
      "learning_rate": 8e-05,
      "loss": 1.5789,
      "step": 461
    },
    {
      "epoch": 0.09595015576323987,
      "grad_norm": 0.5115273594856262,
      "learning_rate": 8e-05,
      "loss": 1.7081,
      "step": 462
    },
    {
      "epoch": 0.09615784008307372,
      "grad_norm": 0.4946316182613373,
      "learning_rate": 8e-05,
      "loss": 1.6407,
      "step": 463
    },
    {
      "epoch": 0.09636552440290758,
      "grad_norm": 0.5241323113441467,
      "learning_rate": 8e-05,
      "loss": 1.7573,
      "step": 464
    },
    {
      "epoch": 0.09657320872274143,
      "grad_norm": 0.47573548555374146,
      "learning_rate": 8e-05,
      "loss": 1.5695,
      "step": 465
    },
    {
      "epoch": 0.09678089304257528,
      "grad_norm": 0.49498170614242554,
      "learning_rate": 8e-05,
      "loss": 1.6743,
      "step": 466
    },
    {
      "epoch": 0.09698857736240914,
      "grad_norm": 0.48741909861564636,
      "learning_rate": 8e-05,
      "loss": 1.645,
      "step": 467
    },
    {
      "epoch": 0.09719626168224299,
      "grad_norm": 0.498971551656723,
      "learning_rate": 8e-05,
      "loss": 1.6154,
      "step": 468
    },
    {
      "epoch": 0.09740394600207684,
      "grad_norm": 0.5093749165534973,
      "learning_rate": 8e-05,
      "loss": 1.7089,
      "step": 469
    },
    {
      "epoch": 0.0976116303219107,
      "grad_norm": 0.5314822793006897,
      "learning_rate": 8e-05,
      "loss": 1.7294,
      "step": 470
    },
    {
      "epoch": 0.09781931464174455,
      "grad_norm": 0.5050390958786011,
      "learning_rate": 8e-05,
      "loss": 1.6589,
      "step": 471
    },
    {
      "epoch": 0.0980269989615784,
      "grad_norm": 0.4838104546070099,
      "learning_rate": 8e-05,
      "loss": 1.6221,
      "step": 472
    },
    {
      "epoch": 0.09823468328141226,
      "grad_norm": 0.5459195375442505,
      "learning_rate": 8e-05,
      "loss": 1.6315,
      "step": 473
    },
    {
      "epoch": 0.09844236760124611,
      "grad_norm": 0.511231541633606,
      "learning_rate": 8e-05,
      "loss": 1.7314,
      "step": 474
    },
    {
      "epoch": 0.09865005192107996,
      "grad_norm": 0.5203128457069397,
      "learning_rate": 8e-05,
      "loss": 1.6929,
      "step": 475
    },
    {
      "epoch": 0.0988577362409138,
      "grad_norm": 0.509968101978302,
      "learning_rate": 8e-05,
      "loss": 1.5961,
      "step": 476
    },
    {
      "epoch": 0.09906542056074766,
      "grad_norm": 0.45741814374923706,
      "learning_rate": 8e-05,
      "loss": 1.5686,
      "step": 477
    },
    {
      "epoch": 0.09927310488058151,
      "grad_norm": 0.49833589792251587,
      "learning_rate": 8e-05,
      "loss": 1.6054,
      "step": 478
    },
    {
      "epoch": 0.09948078920041536,
      "grad_norm": 0.4866011142730713,
      "learning_rate": 8e-05,
      "loss": 1.5743,
      "step": 479
    },
    {
      "epoch": 0.09968847352024922,
      "grad_norm": 0.48203375935554504,
      "learning_rate": 8e-05,
      "loss": 1.6288,
      "step": 480
    },
    {
      "epoch": 0.09989615784008307,
      "grad_norm": 0.48031631112098694,
      "learning_rate": 8e-05,
      "loss": 1.6257,
      "step": 481
    },
    {
      "epoch": 0.10010384215991693,
      "grad_norm": 0.4775748550891876,
      "learning_rate": 8e-05,
      "loss": 1.582,
      "step": 482
    },
    {
      "epoch": 0.10031152647975078,
      "grad_norm": 0.5151557326316833,
      "learning_rate": 8e-05,
      "loss": 1.6394,
      "step": 483
    },
    {
      "epoch": 0.10051921079958463,
      "grad_norm": 0.47847479581832886,
      "learning_rate": 8e-05,
      "loss": 1.6036,
      "step": 484
    },
    {
      "epoch": 0.10072689511941849,
      "grad_norm": 0.5062543153762817,
      "learning_rate": 8e-05,
      "loss": 1.6405,
      "step": 485
    },
    {
      "epoch": 0.10093457943925234,
      "grad_norm": 0.46788135170936584,
      "learning_rate": 8e-05,
      "loss": 1.6106,
      "step": 486
    },
    {
      "epoch": 0.10114226375908619,
      "grad_norm": 0.47770899534225464,
      "learning_rate": 8e-05,
      "loss": 1.6511,
      "step": 487
    },
    {
      "epoch": 0.10134994807892005,
      "grad_norm": 0.5121338963508606,
      "learning_rate": 8e-05,
      "loss": 1.6954,
      "step": 488
    },
    {
      "epoch": 0.1015576323987539,
      "grad_norm": 0.5268594026565552,
      "learning_rate": 8e-05,
      "loss": 1.7471,
      "step": 489
    },
    {
      "epoch": 0.10176531671858775,
      "grad_norm": 0.4916820526123047,
      "learning_rate": 8e-05,
      "loss": 1.6428,
      "step": 490
    },
    {
      "epoch": 0.10197300103842159,
      "grad_norm": 0.4892474412918091,
      "learning_rate": 8e-05,
      "loss": 1.6466,
      "step": 491
    },
    {
      "epoch": 0.10218068535825545,
      "grad_norm": 0.48951107263565063,
      "learning_rate": 8e-05,
      "loss": 1.7112,
      "step": 492
    },
    {
      "epoch": 0.1023883696780893,
      "grad_norm": 0.4870230555534363,
      "learning_rate": 8e-05,
      "loss": 1.7277,
      "step": 493
    },
    {
      "epoch": 0.10259605399792315,
      "grad_norm": 0.5075536370277405,
      "learning_rate": 8e-05,
      "loss": 1.6364,
      "step": 494
    },
    {
      "epoch": 0.102803738317757,
      "grad_norm": 0.5025159120559692,
      "learning_rate": 8e-05,
      "loss": 1.6866,
      "step": 495
    },
    {
      "epoch": 0.10301142263759086,
      "grad_norm": 0.49492985010147095,
      "learning_rate": 8e-05,
      "loss": 1.6144,
      "step": 496
    },
    {
      "epoch": 0.10321910695742471,
      "grad_norm": 0.48841235041618347,
      "learning_rate": 8e-05,
      "loss": 1.6522,
      "step": 497
    },
    {
      "epoch": 0.10342679127725857,
      "grad_norm": 0.47402575612068176,
      "learning_rate": 8e-05,
      "loss": 1.5525,
      "step": 498
    },
    {
      "epoch": 0.10363447559709242,
      "grad_norm": 0.4916023910045624,
      "learning_rate": 8e-05,
      "loss": 1.6744,
      "step": 499
    },
    {
      "epoch": 0.10384215991692627,
      "grad_norm": 0.4725121259689331,
      "learning_rate": 8e-05,
      "loss": 1.5623,
      "step": 500
    },
    {
      "epoch": 0.10404984423676013,
      "grad_norm": 0.4987272322177887,
      "learning_rate": 8e-05,
      "loss": 1.6302,
      "step": 501
    },
    {
      "epoch": 0.10425752855659398,
      "grad_norm": 0.48811638355255127,
      "learning_rate": 8e-05,
      "loss": 1.6124,
      "step": 502
    },
    {
      "epoch": 0.10446521287642783,
      "grad_norm": 0.5044394731521606,
      "learning_rate": 8e-05,
      "loss": 1.7075,
      "step": 503
    },
    {
      "epoch": 0.10467289719626169,
      "grad_norm": 0.4859680235385895,
      "learning_rate": 8e-05,
      "loss": 1.6975,
      "step": 504
    },
    {
      "epoch": 0.10488058151609553,
      "grad_norm": 0.528005063533783,
      "learning_rate": 8e-05,
      "loss": 1.6129,
      "step": 505
    },
    {
      "epoch": 0.10508826583592938,
      "grad_norm": 0.5075289011001587,
      "learning_rate": 8e-05,
      "loss": 1.7663,
      "step": 506
    },
    {
      "epoch": 0.10529595015576323,
      "grad_norm": 0.5202328562736511,
      "learning_rate": 8e-05,
      "loss": 1.685,
      "step": 507
    },
    {
      "epoch": 0.10550363447559709,
      "grad_norm": 0.47399210929870605,
      "learning_rate": 8e-05,
      "loss": 1.612,
      "step": 508
    },
    {
      "epoch": 0.10571131879543094,
      "grad_norm": 0.48944804072380066,
      "learning_rate": 8e-05,
      "loss": 1.5956,
      "step": 509
    },
    {
      "epoch": 0.1059190031152648,
      "grad_norm": 0.5143656730651855,
      "learning_rate": 8e-05,
      "loss": 1.6964,
      "step": 510
    },
    {
      "epoch": 0.10612668743509865,
      "grad_norm": 0.4810107350349426,
      "learning_rate": 8e-05,
      "loss": 1.6548,
      "step": 511
    },
    {
      "epoch": 0.1063343717549325,
      "grad_norm": 0.48855581879615784,
      "learning_rate": 8e-05,
      "loss": 1.637,
      "step": 512
    },
    {
      "epoch": 0.10654205607476636,
      "grad_norm": 0.477912038564682,
      "learning_rate": 8e-05,
      "loss": 1.6137,
      "step": 513
    },
    {
      "epoch": 0.10674974039460021,
      "grad_norm": 0.49473705887794495,
      "learning_rate": 8e-05,
      "loss": 1.6657,
      "step": 514
    },
    {
      "epoch": 0.10695742471443406,
      "grad_norm": 0.4874957501888275,
      "learning_rate": 8e-05,
      "loss": 1.5984,
      "step": 515
    },
    {
      "epoch": 0.10716510903426792,
      "grad_norm": 0.47925207018852234,
      "learning_rate": 8e-05,
      "loss": 1.6142,
      "step": 516
    },
    {
      "epoch": 0.10737279335410177,
      "grad_norm": 0.49199312925338745,
      "learning_rate": 8e-05,
      "loss": 1.6706,
      "step": 517
    },
    {
      "epoch": 0.10758047767393562,
      "grad_norm": 0.4763036072254181,
      "learning_rate": 8e-05,
      "loss": 1.6835,
      "step": 518
    },
    {
      "epoch": 0.10778816199376948,
      "grad_norm": 0.5105186700820923,
      "learning_rate": 8e-05,
      "loss": 1.6513,
      "step": 519
    },
    {
      "epoch": 0.10799584631360332,
      "grad_norm": 0.48001575469970703,
      "learning_rate": 8e-05,
      "loss": 1.6549,
      "step": 520
    },
    {
      "epoch": 0.10820353063343717,
      "grad_norm": 0.5005422830581665,
      "learning_rate": 8e-05,
      "loss": 1.6499,
      "step": 521
    },
    {
      "epoch": 0.10841121495327102,
      "grad_norm": 0.4944564402103424,
      "learning_rate": 8e-05,
      "loss": 1.6491,
      "step": 522
    },
    {
      "epoch": 0.10861889927310488,
      "grad_norm": 0.47744297981262207,
      "learning_rate": 8e-05,
      "loss": 1.5596,
      "step": 523
    },
    {
      "epoch": 0.10882658359293873,
      "grad_norm": 0.4701041877269745,
      "learning_rate": 8e-05,
      "loss": 1.6377,
      "step": 524
    },
    {
      "epoch": 0.10903426791277258,
      "grad_norm": 0.5091982483863831,
      "learning_rate": 8e-05,
      "loss": 1.7002,
      "step": 525
    },
    {
      "epoch": 0.10924195223260644,
      "grad_norm": 0.4894920587539673,
      "learning_rate": 8e-05,
      "loss": 1.6587,
      "step": 526
    },
    {
      "epoch": 0.10944963655244029,
      "grad_norm": 0.48994290828704834,
      "learning_rate": 8e-05,
      "loss": 1.6763,
      "step": 527
    },
    {
      "epoch": 0.10965732087227414,
      "grad_norm": 0.49817031621932983,
      "learning_rate": 8e-05,
      "loss": 1.6896,
      "step": 528
    },
    {
      "epoch": 0.109865005192108,
      "grad_norm": 0.4927424490451813,
      "learning_rate": 8e-05,
      "loss": 1.634,
      "step": 529
    },
    {
      "epoch": 0.11007268951194185,
      "grad_norm": 0.46971508860588074,
      "learning_rate": 8e-05,
      "loss": 1.6402,
      "step": 530
    },
    {
      "epoch": 0.1102803738317757,
      "grad_norm": 0.5013573169708252,
      "learning_rate": 8e-05,
      "loss": 1.6279,
      "step": 531
    },
    {
      "epoch": 0.11048805815160956,
      "grad_norm": 0.494057297706604,
      "learning_rate": 8e-05,
      "loss": 1.6284,
      "step": 532
    },
    {
      "epoch": 0.11069574247144341,
      "grad_norm": 0.49397385120391846,
      "learning_rate": 8e-05,
      "loss": 1.6885,
      "step": 533
    },
    {
      "epoch": 0.11090342679127727,
      "grad_norm": 0.5148130059242249,
      "learning_rate": 8e-05,
      "loss": 1.6592,
      "step": 534
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 0.4772680401802063,
      "learning_rate": 8e-05,
      "loss": 1.6081,
      "step": 535
    },
    {
      "epoch": 0.11131879543094496,
      "grad_norm": 0.5171798467636108,
      "learning_rate": 8e-05,
      "loss": 1.7278,
      "step": 536
    },
    {
      "epoch": 0.11152647975077881,
      "grad_norm": 0.47712597250938416,
      "learning_rate": 8e-05,
      "loss": 1.6092,
      "step": 537
    },
    {
      "epoch": 0.11173416407061267,
      "grad_norm": 0.47499656677246094,
      "learning_rate": 8e-05,
      "loss": 1.598,
      "step": 538
    },
    {
      "epoch": 0.11194184839044652,
      "grad_norm": 0.4703918397426605,
      "learning_rate": 8e-05,
      "loss": 1.6449,
      "step": 539
    },
    {
      "epoch": 0.11214953271028037,
      "grad_norm": 0.5092900991439819,
      "learning_rate": 8e-05,
      "loss": 1.6808,
      "step": 540
    },
    {
      "epoch": 0.11235721703011423,
      "grad_norm": 0.48537686467170715,
      "learning_rate": 8e-05,
      "loss": 1.6026,
      "step": 541
    },
    {
      "epoch": 0.11256490134994808,
      "grad_norm": 0.49423104524612427,
      "learning_rate": 8e-05,
      "loss": 1.7332,
      "step": 542
    },
    {
      "epoch": 0.11277258566978193,
      "grad_norm": 0.48707374930381775,
      "learning_rate": 8e-05,
      "loss": 1.6686,
      "step": 543
    },
    {
      "epoch": 0.11298026998961579,
      "grad_norm": 0.4906916618347168,
      "learning_rate": 8e-05,
      "loss": 1.6518,
      "step": 544
    },
    {
      "epoch": 0.11318795430944964,
      "grad_norm": 0.4747893512248993,
      "learning_rate": 8e-05,
      "loss": 1.6452,
      "step": 545
    },
    {
      "epoch": 0.1133956386292835,
      "grad_norm": 0.4864533245563507,
      "learning_rate": 8e-05,
      "loss": 1.6477,
      "step": 546
    },
    {
      "epoch": 0.11360332294911735,
      "grad_norm": 0.469113826751709,
      "learning_rate": 8e-05,
      "loss": 1.5731,
      "step": 547
    },
    {
      "epoch": 0.1138110072689512,
      "grad_norm": 0.48744997382164,
      "learning_rate": 8e-05,
      "loss": 1.6906,
      "step": 548
    },
    {
      "epoch": 0.11401869158878504,
      "grad_norm": 0.4924449920654297,
      "learning_rate": 8e-05,
      "loss": 1.5568,
      "step": 549
    },
    {
      "epoch": 0.1142263759086189,
      "grad_norm": 0.46387964487075806,
      "learning_rate": 8e-05,
      "loss": 1.6574,
      "step": 550
    },
    {
      "epoch": 0.11443406022845275,
      "grad_norm": 0.49000611901283264,
      "learning_rate": 8e-05,
      "loss": 1.6969,
      "step": 551
    },
    {
      "epoch": 0.1146417445482866,
      "grad_norm": 0.5043030977249146,
      "learning_rate": 8e-05,
      "loss": 1.6773,
      "step": 552
    },
    {
      "epoch": 0.11484942886812045,
      "grad_norm": 0.5181741118431091,
      "learning_rate": 8e-05,
      "loss": 1.6705,
      "step": 553
    },
    {
      "epoch": 0.11505711318795431,
      "grad_norm": 0.517162561416626,
      "learning_rate": 8e-05,
      "loss": 1.7016,
      "step": 554
    },
    {
      "epoch": 0.11526479750778816,
      "grad_norm": 0.5007038116455078,
      "learning_rate": 8e-05,
      "loss": 1.6708,
      "step": 555
    },
    {
      "epoch": 0.11547248182762201,
      "grad_norm": 0.49101993441581726,
      "learning_rate": 8e-05,
      "loss": 1.6829,
      "step": 556
    },
    {
      "epoch": 0.11568016614745587,
      "grad_norm": 0.47777554392814636,
      "learning_rate": 8e-05,
      "loss": 1.6972,
      "step": 557
    },
    {
      "epoch": 0.11588785046728972,
      "grad_norm": 0.5146732330322266,
      "learning_rate": 8e-05,
      "loss": 1.6615,
      "step": 558
    },
    {
      "epoch": 0.11609553478712357,
      "grad_norm": 0.4889799654483795,
      "learning_rate": 8e-05,
      "loss": 1.6389,
      "step": 559
    },
    {
      "epoch": 0.11630321910695743,
      "grad_norm": 0.48216649889945984,
      "learning_rate": 8e-05,
      "loss": 1.7034,
      "step": 560
    },
    {
      "epoch": 0.11651090342679128,
      "grad_norm": 0.466871052980423,
      "learning_rate": 8e-05,
      "loss": 1.5976,
      "step": 561
    },
    {
      "epoch": 0.11671858774662514,
      "grad_norm": 0.4961405396461487,
      "learning_rate": 8e-05,
      "loss": 1.6173,
      "step": 562
    },
    {
      "epoch": 0.11692627206645899,
      "grad_norm": 0.5158309936523438,
      "learning_rate": 8e-05,
      "loss": 1.7243,
      "step": 563
    },
    {
      "epoch": 0.11713395638629283,
      "grad_norm": 0.5011987686157227,
      "learning_rate": 8e-05,
      "loss": 1.6861,
      "step": 564
    },
    {
      "epoch": 0.11734164070612668,
      "grad_norm": 0.535325825214386,
      "learning_rate": 8e-05,
      "loss": 1.799,
      "step": 565
    },
    {
      "epoch": 0.11754932502596054,
      "grad_norm": 0.4968489110469818,
      "learning_rate": 8e-05,
      "loss": 1.7096,
      "step": 566
    },
    {
      "epoch": 0.11775700934579439,
      "grad_norm": 0.4858616292476654,
      "learning_rate": 8e-05,
      "loss": 1.6617,
      "step": 567
    },
    {
      "epoch": 0.11796469366562824,
      "grad_norm": 0.49563148617744446,
      "learning_rate": 8e-05,
      "loss": 1.6681,
      "step": 568
    },
    {
      "epoch": 0.1181723779854621,
      "grad_norm": 0.49028831720352173,
      "learning_rate": 8e-05,
      "loss": 1.6425,
      "step": 569
    },
    {
      "epoch": 0.11838006230529595,
      "grad_norm": 0.4926828145980835,
      "learning_rate": 8e-05,
      "loss": 1.6134,
      "step": 570
    },
    {
      "epoch": 0.1185877466251298,
      "grad_norm": 0.4846639037132263,
      "learning_rate": 8e-05,
      "loss": 1.6194,
      "step": 571
    },
    {
      "epoch": 0.11879543094496366,
      "grad_norm": 0.4756839871406555,
      "learning_rate": 8e-05,
      "loss": 1.6454,
      "step": 572
    },
    {
      "epoch": 0.11900311526479751,
      "grad_norm": 0.4722893536090851,
      "learning_rate": 8e-05,
      "loss": 1.6123,
      "step": 573
    },
    {
      "epoch": 0.11921079958463136,
      "grad_norm": 0.4920216202735901,
      "learning_rate": 8e-05,
      "loss": 1.6916,
      "step": 574
    },
    {
      "epoch": 0.11941848390446522,
      "grad_norm": 0.48332494497299194,
      "learning_rate": 8e-05,
      "loss": 1.6414,
      "step": 575
    },
    {
      "epoch": 0.11962616822429907,
      "grad_norm": 0.5009651780128479,
      "learning_rate": 8e-05,
      "loss": 1.6119,
      "step": 576
    },
    {
      "epoch": 0.11983385254413292,
      "grad_norm": 0.4851033389568329,
      "learning_rate": 8e-05,
      "loss": 1.6018,
      "step": 577
    },
    {
      "epoch": 0.12004153686396678,
      "grad_norm": 0.4941191077232361,
      "learning_rate": 8e-05,
      "loss": 1.6982,
      "step": 578
    },
    {
      "epoch": 0.12024922118380062,
      "grad_norm": 0.4699910879135132,
      "learning_rate": 8e-05,
      "loss": 1.6004,
      "step": 579
    },
    {
      "epoch": 0.12045690550363447,
      "grad_norm": 0.47666141390800476,
      "learning_rate": 8e-05,
      "loss": 1.6523,
      "step": 580
    },
    {
      "epoch": 0.12066458982346832,
      "grad_norm": 0.5250875949859619,
      "learning_rate": 8e-05,
      "loss": 1.7131,
      "step": 581
    },
    {
      "epoch": 0.12087227414330218,
      "grad_norm": 0.47818005084991455,
      "learning_rate": 8e-05,
      "loss": 1.6228,
      "step": 582
    },
    {
      "epoch": 0.12107995846313603,
      "grad_norm": 0.46727368235588074,
      "learning_rate": 8e-05,
      "loss": 1.5709,
      "step": 583
    },
    {
      "epoch": 0.12128764278296988,
      "grad_norm": 0.4783129394054413,
      "learning_rate": 8e-05,
      "loss": 1.6809,
      "step": 584
    },
    {
      "epoch": 0.12149532710280374,
      "grad_norm": 0.48866942524909973,
      "learning_rate": 8e-05,
      "loss": 1.6795,
      "step": 585
    },
    {
      "epoch": 0.12170301142263759,
      "grad_norm": 0.4863775968551636,
      "learning_rate": 8e-05,
      "loss": 1.6499,
      "step": 586
    },
    {
      "epoch": 0.12191069574247144,
      "grad_norm": 0.48488086462020874,
      "learning_rate": 8e-05,
      "loss": 1.6765,
      "step": 587
    },
    {
      "epoch": 0.1221183800623053,
      "grad_norm": 0.5019093751907349,
      "learning_rate": 8e-05,
      "loss": 1.6547,
      "step": 588
    },
    {
      "epoch": 0.12232606438213915,
      "grad_norm": 0.484720379114151,
      "learning_rate": 8e-05,
      "loss": 1.6438,
      "step": 589
    },
    {
      "epoch": 0.122533748701973,
      "grad_norm": 0.48393747210502625,
      "learning_rate": 8e-05,
      "loss": 1.6595,
      "step": 590
    },
    {
      "epoch": 0.12274143302180686,
      "grad_norm": 0.5197237133979797,
      "learning_rate": 8e-05,
      "loss": 1.6346,
      "step": 591
    },
    {
      "epoch": 0.12294911734164071,
      "grad_norm": 0.4648591876029968,
      "learning_rate": 8e-05,
      "loss": 1.6076,
      "step": 592
    },
    {
      "epoch": 0.12315680166147455,
      "grad_norm": 0.4920279085636139,
      "learning_rate": 8e-05,
      "loss": 1.6787,
      "step": 593
    },
    {
      "epoch": 0.1233644859813084,
      "grad_norm": 0.48412156105041504,
      "learning_rate": 8e-05,
      "loss": 1.6896,
      "step": 594
    },
    {
      "epoch": 0.12357217030114226,
      "grad_norm": 0.4906349778175354,
      "learning_rate": 8e-05,
      "loss": 1.6746,
      "step": 595
    },
    {
      "epoch": 0.12377985462097611,
      "grad_norm": 0.49094000458717346,
      "learning_rate": 8e-05,
      "loss": 1.6618,
      "step": 596
    },
    {
      "epoch": 0.12398753894080997,
      "grad_norm": 0.48553135991096497,
      "learning_rate": 8e-05,
      "loss": 1.7193,
      "step": 597
    },
    {
      "epoch": 0.12419522326064382,
      "grad_norm": 0.48377707600593567,
      "learning_rate": 8e-05,
      "loss": 1.6537,
      "step": 598
    },
    {
      "epoch": 0.12440290758047767,
      "grad_norm": 0.47426944971084595,
      "learning_rate": 8e-05,
      "loss": 1.6248,
      "step": 599
    },
    {
      "epoch": 0.12461059190031153,
      "grad_norm": 0.4826415777206421,
      "learning_rate": 8e-05,
      "loss": 1.6596,
      "step": 600
    },
    {
      "epoch": 0.12481827622014538,
      "grad_norm": 0.49226781725883484,
      "learning_rate": 8e-05,
      "loss": 1.7157,
      "step": 601
    },
    {
      "epoch": 0.12502596053997922,
      "grad_norm": 0.46622025966644287,
      "learning_rate": 8e-05,
      "loss": 1.61,
      "step": 602
    },
    {
      "epoch": 0.1252336448598131,
      "grad_norm": 0.4895409345626831,
      "learning_rate": 8e-05,
      "loss": 1.6629,
      "step": 603
    },
    {
      "epoch": 0.12544132917964693,
      "grad_norm": 0.4662224352359772,
      "learning_rate": 8e-05,
      "loss": 1.6188,
      "step": 604
    },
    {
      "epoch": 0.1256490134994808,
      "grad_norm": 0.4780591130256653,
      "learning_rate": 8e-05,
      "loss": 1.5856,
      "step": 605
    },
    {
      "epoch": 0.12585669781931463,
      "grad_norm": 0.49765658378601074,
      "learning_rate": 8e-05,
      "loss": 1.613,
      "step": 606
    },
    {
      "epoch": 0.1260643821391485,
      "grad_norm": 0.4997657239437103,
      "learning_rate": 8e-05,
      "loss": 1.6758,
      "step": 607
    },
    {
      "epoch": 0.12627206645898234,
      "grad_norm": 0.48155564069747925,
      "learning_rate": 8e-05,
      "loss": 1.6179,
      "step": 608
    },
    {
      "epoch": 0.1264797507788162,
      "grad_norm": 0.4809686541557312,
      "learning_rate": 8e-05,
      "loss": 1.5715,
      "step": 609
    },
    {
      "epoch": 0.12668743509865005,
      "grad_norm": 0.5093618035316467,
      "learning_rate": 8e-05,
      "loss": 1.7242,
      "step": 610
    },
    {
      "epoch": 0.12689511941848391,
      "grad_norm": 0.4809294641017914,
      "learning_rate": 8e-05,
      "loss": 1.6881,
      "step": 611
    },
    {
      "epoch": 0.12710280373831775,
      "grad_norm": 0.4776981472969055,
      "learning_rate": 8e-05,
      "loss": 1.6183,
      "step": 612
    },
    {
      "epoch": 0.12731048805815162,
      "grad_norm": 0.4951392412185669,
      "learning_rate": 8e-05,
      "loss": 1.7086,
      "step": 613
    },
    {
      "epoch": 0.12751817237798546,
      "grad_norm": 0.48191651701927185,
      "learning_rate": 8e-05,
      "loss": 1.6385,
      "step": 614
    },
    {
      "epoch": 0.1277258566978193,
      "grad_norm": 0.4886009693145752,
      "learning_rate": 8e-05,
      "loss": 1.618,
      "step": 615
    },
    {
      "epoch": 0.12793354101765317,
      "grad_norm": 0.49545979499816895,
      "learning_rate": 8e-05,
      "loss": 1.6361,
      "step": 616
    },
    {
      "epoch": 0.128141225337487,
      "grad_norm": 0.49346309900283813,
      "learning_rate": 8e-05,
      "loss": 1.693,
      "step": 617
    },
    {
      "epoch": 0.12834890965732088,
      "grad_norm": 0.5284665822982788,
      "learning_rate": 8e-05,
      "loss": 1.6347,
      "step": 618
    },
    {
      "epoch": 0.12855659397715472,
      "grad_norm": 0.5081282258033752,
      "learning_rate": 8e-05,
      "loss": 1.6581,
      "step": 619
    },
    {
      "epoch": 0.12876427829698858,
      "grad_norm": 0.5118349194526672,
      "learning_rate": 8e-05,
      "loss": 1.7038,
      "step": 620
    },
    {
      "epoch": 0.12897196261682242,
      "grad_norm": 0.49056780338287354,
      "learning_rate": 8e-05,
      "loss": 1.7102,
      "step": 621
    },
    {
      "epoch": 0.1291796469366563,
      "grad_norm": 0.4869656562805176,
      "learning_rate": 8e-05,
      "loss": 1.619,
      "step": 622
    },
    {
      "epoch": 0.12938733125649013,
      "grad_norm": 0.5157124996185303,
      "learning_rate": 8e-05,
      "loss": 1.6398,
      "step": 623
    },
    {
      "epoch": 0.129595015576324,
      "grad_norm": 0.48705726861953735,
      "learning_rate": 8e-05,
      "loss": 1.6655,
      "step": 624
    },
    {
      "epoch": 0.12980269989615784,
      "grad_norm": 0.46967780590057373,
      "learning_rate": 8e-05,
      "loss": 1.6131,
      "step": 625
    },
    {
      "epoch": 0.1300103842159917,
      "grad_norm": 0.47689181566238403,
      "learning_rate": 8e-05,
      "loss": 1.6086,
      "step": 626
    },
    {
      "epoch": 0.13021806853582554,
      "grad_norm": 0.48610159754753113,
      "learning_rate": 8e-05,
      "loss": 1.6463,
      "step": 627
    },
    {
      "epoch": 0.1304257528556594,
      "grad_norm": 0.4769737124443054,
      "learning_rate": 8e-05,
      "loss": 1.57,
      "step": 628
    },
    {
      "epoch": 0.13063343717549325,
      "grad_norm": 0.47366201877593994,
      "learning_rate": 8e-05,
      "loss": 1.6771,
      "step": 629
    },
    {
      "epoch": 0.1308411214953271,
      "grad_norm": 0.5110802054405212,
      "learning_rate": 8e-05,
      "loss": 1.6377,
      "step": 630
    },
    {
      "epoch": 0.13104880581516096,
      "grad_norm": 0.5230108499526978,
      "learning_rate": 8e-05,
      "loss": 1.6979,
      "step": 631
    },
    {
      "epoch": 0.1312564901349948,
      "grad_norm": 0.4913478493690491,
      "learning_rate": 8e-05,
      "loss": 1.6687,
      "step": 632
    },
    {
      "epoch": 0.13146417445482866,
      "grad_norm": 0.5109069347381592,
      "learning_rate": 8e-05,
      "loss": 1.6655,
      "step": 633
    },
    {
      "epoch": 0.1316718587746625,
      "grad_norm": 0.5116170644760132,
      "learning_rate": 8e-05,
      "loss": 1.7231,
      "step": 634
    },
    {
      "epoch": 0.13187954309449637,
      "grad_norm": 0.5041365623474121,
      "learning_rate": 8e-05,
      "loss": 1.6218,
      "step": 635
    },
    {
      "epoch": 0.1320872274143302,
      "grad_norm": 0.4839491546154022,
      "learning_rate": 8e-05,
      "loss": 1.6518,
      "step": 636
    },
    {
      "epoch": 0.13229491173416408,
      "grad_norm": 0.4792550802230835,
      "learning_rate": 8e-05,
      "loss": 1.6506,
      "step": 637
    },
    {
      "epoch": 0.13250259605399792,
      "grad_norm": 0.48295944929122925,
      "learning_rate": 8e-05,
      "loss": 1.6901,
      "step": 638
    },
    {
      "epoch": 0.13271028037383178,
      "grad_norm": 0.4973006844520569,
      "learning_rate": 8e-05,
      "loss": 1.6432,
      "step": 639
    },
    {
      "epoch": 0.13291796469366562,
      "grad_norm": 0.4931349456310272,
      "learning_rate": 8e-05,
      "loss": 1.6903,
      "step": 640
    },
    {
      "epoch": 0.1331256490134995,
      "grad_norm": 0.5359208583831787,
      "learning_rate": 8e-05,
      "loss": 1.6107,
      "step": 641
    },
    {
      "epoch": 0.13333333333333333,
      "grad_norm": 0.5108036994934082,
      "learning_rate": 8e-05,
      "loss": 1.743,
      "step": 642
    },
    {
      "epoch": 0.1335410176531672,
      "grad_norm": 0.53212970495224,
      "learning_rate": 8e-05,
      "loss": 1.6529,
      "step": 643
    },
    {
      "epoch": 0.13374870197300104,
      "grad_norm": 0.4618084132671356,
      "learning_rate": 8e-05,
      "loss": 1.6473,
      "step": 644
    },
    {
      "epoch": 0.13395638629283488,
      "grad_norm": 0.4782806932926178,
      "learning_rate": 8e-05,
      "loss": 1.6267,
      "step": 645
    },
    {
      "epoch": 0.13416407061266875,
      "grad_norm": 0.5155082941055298,
      "learning_rate": 8e-05,
      "loss": 1.6114,
      "step": 646
    },
    {
      "epoch": 0.13437175493250259,
      "grad_norm": 0.5027504563331604,
      "learning_rate": 8e-05,
      "loss": 1.6559,
      "step": 647
    },
    {
      "epoch": 0.13457943925233645,
      "grad_norm": 0.5325614213943481,
      "learning_rate": 8e-05,
      "loss": 1.5957,
      "step": 648
    },
    {
      "epoch": 0.1347871235721703,
      "grad_norm": 0.4770219922065735,
      "learning_rate": 8e-05,
      "loss": 1.5635,
      "step": 649
    },
    {
      "epoch": 0.13499480789200416,
      "grad_norm": 0.4837128818035126,
      "learning_rate": 8e-05,
      "loss": 1.5886,
      "step": 650
    },
    {
      "epoch": 0.135202492211838,
      "grad_norm": 0.5071947574615479,
      "learning_rate": 8e-05,
      "loss": 1.7164,
      "step": 651
    },
    {
      "epoch": 0.13541017653167187,
      "grad_norm": 0.48948243260383606,
      "learning_rate": 8e-05,
      "loss": 1.6695,
      "step": 652
    },
    {
      "epoch": 0.1356178608515057,
      "grad_norm": 0.49877387285232544,
      "learning_rate": 8e-05,
      "loss": 1.6044,
      "step": 653
    },
    {
      "epoch": 0.13582554517133957,
      "grad_norm": 0.48378750681877136,
      "learning_rate": 8e-05,
      "loss": 1.6118,
      "step": 654
    },
    {
      "epoch": 0.1360332294911734,
      "grad_norm": 0.49697884917259216,
      "learning_rate": 8e-05,
      "loss": 1.6245,
      "step": 655
    },
    {
      "epoch": 0.13624091381100728,
      "grad_norm": 0.5250710248947144,
      "learning_rate": 8e-05,
      "loss": 1.5699,
      "step": 656
    },
    {
      "epoch": 0.13644859813084112,
      "grad_norm": 0.49164098501205444,
      "learning_rate": 8e-05,
      "loss": 1.7009,
      "step": 657
    },
    {
      "epoch": 0.136656282450675,
      "grad_norm": 0.51181560754776,
      "learning_rate": 8e-05,
      "loss": 1.6749,
      "step": 658
    },
    {
      "epoch": 0.13686396677050883,
      "grad_norm": 0.4895937442779541,
      "learning_rate": 8e-05,
      "loss": 1.6471,
      "step": 659
    },
    {
      "epoch": 0.13707165109034267,
      "grad_norm": 0.500437319278717,
      "learning_rate": 8e-05,
      "loss": 1.6039,
      "step": 660
    },
    {
      "epoch": 0.13727933541017653,
      "grad_norm": 0.49318140745162964,
      "learning_rate": 8e-05,
      "loss": 1.6222,
      "step": 661
    },
    {
      "epoch": 0.13748701973001037,
      "grad_norm": 0.47766220569610596,
      "learning_rate": 8e-05,
      "loss": 1.6307,
      "step": 662
    },
    {
      "epoch": 0.13769470404984424,
      "grad_norm": 0.526156485080719,
      "learning_rate": 8e-05,
      "loss": 1.6572,
      "step": 663
    },
    {
      "epoch": 0.13790238836967808,
      "grad_norm": 0.48735764622688293,
      "learning_rate": 8e-05,
      "loss": 1.6393,
      "step": 664
    },
    {
      "epoch": 0.13811007268951195,
      "grad_norm": 0.49642348289489746,
      "learning_rate": 8e-05,
      "loss": 1.6371,
      "step": 665
    },
    {
      "epoch": 0.1383177570093458,
      "grad_norm": 0.5099259614944458,
      "learning_rate": 8e-05,
      "loss": 1.6536,
      "step": 666
    },
    {
      "epoch": 0.13852544132917965,
      "grad_norm": 0.5075072646141052,
      "learning_rate": 8e-05,
      "loss": 1.6389,
      "step": 667
    },
    {
      "epoch": 0.1387331256490135,
      "grad_norm": 0.49564462900161743,
      "learning_rate": 8e-05,
      "loss": 1.6303,
      "step": 668
    },
    {
      "epoch": 0.13894080996884736,
      "grad_norm": 0.468453049659729,
      "learning_rate": 8e-05,
      "loss": 1.6019,
      "step": 669
    },
    {
      "epoch": 0.1391484942886812,
      "grad_norm": 0.4859846234321594,
      "learning_rate": 8e-05,
      "loss": 1.6411,
      "step": 670
    },
    {
      "epoch": 0.13935617860851507,
      "grad_norm": 0.468565970659256,
      "learning_rate": 8e-05,
      "loss": 1.5784,
      "step": 671
    },
    {
      "epoch": 0.1395638629283489,
      "grad_norm": 0.4832054674625397,
      "learning_rate": 8e-05,
      "loss": 1.6492,
      "step": 672
    },
    {
      "epoch": 0.13977154724818278,
      "grad_norm": 0.49340352416038513,
      "learning_rate": 8e-05,
      "loss": 1.6563,
      "step": 673
    },
    {
      "epoch": 0.13997923156801662,
      "grad_norm": 0.5209794044494629,
      "learning_rate": 8e-05,
      "loss": 1.7233,
      "step": 674
    },
    {
      "epoch": 0.14018691588785046,
      "grad_norm": 0.48782119154930115,
      "learning_rate": 8e-05,
      "loss": 1.6472,
      "step": 675
    },
    {
      "epoch": 0.14039460020768432,
      "grad_norm": 0.5135282874107361,
      "learning_rate": 8e-05,
      "loss": 1.6647,
      "step": 676
    },
    {
      "epoch": 0.14060228452751816,
      "grad_norm": 0.5062872171401978,
      "learning_rate": 8e-05,
      "loss": 1.6414,
      "step": 677
    },
    {
      "epoch": 0.14080996884735203,
      "grad_norm": 0.4905490577220917,
      "learning_rate": 8e-05,
      "loss": 1.5628,
      "step": 678
    },
    {
      "epoch": 0.14101765316718587,
      "grad_norm": 0.4919457733631134,
      "learning_rate": 8e-05,
      "loss": 1.6491,
      "step": 679
    },
    {
      "epoch": 0.14122533748701974,
      "grad_norm": 0.5233182907104492,
      "learning_rate": 8e-05,
      "loss": 1.6324,
      "step": 680
    },
    {
      "epoch": 0.14143302180685358,
      "grad_norm": 0.4853593111038208,
      "learning_rate": 8e-05,
      "loss": 1.6134,
      "step": 681
    },
    {
      "epoch": 0.14164070612668744,
      "grad_norm": 0.48515936732292175,
      "learning_rate": 8e-05,
      "loss": 1.5967,
      "step": 682
    },
    {
      "epoch": 0.14184839044652128,
      "grad_norm": 0.5354658961296082,
      "learning_rate": 8e-05,
      "loss": 1.5981,
      "step": 683
    },
    {
      "epoch": 0.14205607476635515,
      "grad_norm": 0.48015713691711426,
      "learning_rate": 8e-05,
      "loss": 1.6279,
      "step": 684
    },
    {
      "epoch": 0.142263759086189,
      "grad_norm": 0.5790321230888367,
      "learning_rate": 8e-05,
      "loss": 1.7114,
      "step": 685
    },
    {
      "epoch": 0.14247144340602286,
      "grad_norm": 0.46386808156967163,
      "learning_rate": 8e-05,
      "loss": 1.6201,
      "step": 686
    },
    {
      "epoch": 0.1426791277258567,
      "grad_norm": 0.5704835653305054,
      "learning_rate": 8e-05,
      "loss": 1.6524,
      "step": 687
    },
    {
      "epoch": 0.14288681204569054,
      "grad_norm": 0.4924614727497101,
      "learning_rate": 8e-05,
      "loss": 1.7339,
      "step": 688
    },
    {
      "epoch": 0.1430944963655244,
      "grad_norm": 0.5158697962760925,
      "learning_rate": 8e-05,
      "loss": 1.6294,
      "step": 689
    },
    {
      "epoch": 0.14330218068535824,
      "grad_norm": 0.5031910538673401,
      "learning_rate": 8e-05,
      "loss": 1.5753,
      "step": 690
    },
    {
      "epoch": 0.1435098650051921,
      "grad_norm": 0.4993326961994171,
      "learning_rate": 8e-05,
      "loss": 1.7069,
      "step": 691
    },
    {
      "epoch": 0.14371754932502595,
      "grad_norm": 0.5596215724945068,
      "learning_rate": 8e-05,
      "loss": 1.6887,
      "step": 692
    },
    {
      "epoch": 0.14392523364485982,
      "grad_norm": 0.464817613363266,
      "learning_rate": 8e-05,
      "loss": 1.5974,
      "step": 693
    },
    {
      "epoch": 0.14413291796469366,
      "grad_norm": 0.5313472151756287,
      "learning_rate": 8e-05,
      "loss": 1.6584,
      "step": 694
    },
    {
      "epoch": 0.14434060228452752,
      "grad_norm": 0.470864862203598,
      "learning_rate": 8e-05,
      "loss": 1.5668,
      "step": 695
    },
    {
      "epoch": 0.14454828660436136,
      "grad_norm": 0.4789741039276123,
      "learning_rate": 8e-05,
      "loss": 1.6432,
      "step": 696
    },
    {
      "epoch": 0.14475597092419523,
      "grad_norm": 0.4912903606891632,
      "learning_rate": 8e-05,
      "loss": 1.5999,
      "step": 697
    },
    {
      "epoch": 0.14496365524402907,
      "grad_norm": 0.49190813302993774,
      "learning_rate": 8e-05,
      "loss": 1.6045,
      "step": 698
    },
    {
      "epoch": 0.14517133956386294,
      "grad_norm": 0.48753032088279724,
      "learning_rate": 8e-05,
      "loss": 1.6548,
      "step": 699
    },
    {
      "epoch": 0.14537902388369678,
      "grad_norm": 0.47778916358947754,
      "learning_rate": 8e-05,
      "loss": 1.5859,
      "step": 700
    },
    {
      "epoch": 0.14558670820353065,
      "grad_norm": 0.48212990164756775,
      "learning_rate": 8e-05,
      "loss": 1.6022,
      "step": 701
    },
    {
      "epoch": 0.14579439252336449,
      "grad_norm": 0.5078028440475464,
      "learning_rate": 8e-05,
      "loss": 1.7241,
      "step": 702
    },
    {
      "epoch": 0.14600207684319833,
      "grad_norm": 0.49921220541000366,
      "learning_rate": 8e-05,
      "loss": 1.7156,
      "step": 703
    },
    {
      "epoch": 0.1462097611630322,
      "grad_norm": 0.5023846626281738,
      "learning_rate": 8e-05,
      "loss": 1.71,
      "step": 704
    },
    {
      "epoch": 0.14641744548286603,
      "grad_norm": 0.4875635802745819,
      "learning_rate": 8e-05,
      "loss": 1.6838,
      "step": 705
    },
    {
      "epoch": 0.1466251298026999,
      "grad_norm": 0.477684885263443,
      "learning_rate": 8e-05,
      "loss": 1.6201,
      "step": 706
    },
    {
      "epoch": 0.14683281412253374,
      "grad_norm": 0.4834393858909607,
      "learning_rate": 8e-05,
      "loss": 1.6228,
      "step": 707
    },
    {
      "epoch": 0.1470404984423676,
      "grad_norm": 0.4797709286212921,
      "learning_rate": 8e-05,
      "loss": 1.589,
      "step": 708
    },
    {
      "epoch": 0.14724818276220145,
      "grad_norm": 0.4814303517341614,
      "learning_rate": 8e-05,
      "loss": 1.6916,
      "step": 709
    },
    {
      "epoch": 0.1474558670820353,
      "grad_norm": 0.48855525255203247,
      "learning_rate": 8e-05,
      "loss": 1.6431,
      "step": 710
    },
    {
      "epoch": 0.14766355140186915,
      "grad_norm": 0.4643092453479767,
      "learning_rate": 8e-05,
      "loss": 1.6707,
      "step": 711
    },
    {
      "epoch": 0.14787123572170302,
      "grad_norm": 0.4828472435474396,
      "learning_rate": 8e-05,
      "loss": 1.6574,
      "step": 712
    },
    {
      "epoch": 0.14807892004153686,
      "grad_norm": 0.5036082863807678,
      "learning_rate": 8e-05,
      "loss": 1.7352,
      "step": 713
    },
    {
      "epoch": 0.14828660436137073,
      "grad_norm": 0.49272510409355164,
      "learning_rate": 8e-05,
      "loss": 1.676,
      "step": 714
    },
    {
      "epoch": 0.14849428868120457,
      "grad_norm": 0.4979461431503296,
      "learning_rate": 8e-05,
      "loss": 1.6097,
      "step": 715
    },
    {
      "epoch": 0.14870197300103843,
      "grad_norm": 0.4923213720321655,
      "learning_rate": 8e-05,
      "loss": 1.5899,
      "step": 716
    },
    {
      "epoch": 0.14890965732087227,
      "grad_norm": 0.48915570974349976,
      "learning_rate": 8e-05,
      "loss": 1.5901,
      "step": 717
    },
    {
      "epoch": 0.1491173416407061,
      "grad_norm": 0.4918684661388397,
      "learning_rate": 8e-05,
      "loss": 1.6548,
      "step": 718
    },
    {
      "epoch": 0.14932502596053998,
      "grad_norm": 0.47110480070114136,
      "learning_rate": 8e-05,
      "loss": 1.6423,
      "step": 719
    },
    {
      "epoch": 0.14953271028037382,
      "grad_norm": 0.5042193531990051,
      "learning_rate": 8e-05,
      "loss": 1.6419,
      "step": 720
    },
    {
      "epoch": 0.1497403946002077,
      "grad_norm": 0.4912029504776001,
      "learning_rate": 8e-05,
      "loss": 1.6902,
      "step": 721
    },
    {
      "epoch": 0.14994807892004153,
      "grad_norm": 0.4772588312625885,
      "learning_rate": 8e-05,
      "loss": 1.572,
      "step": 722
    },
    {
      "epoch": 0.1501557632398754,
      "grad_norm": 0.4800530970096588,
      "learning_rate": 8e-05,
      "loss": 1.6327,
      "step": 723
    },
    {
      "epoch": 0.15036344755970923,
      "grad_norm": 0.5064852237701416,
      "learning_rate": 8e-05,
      "loss": 1.6544,
      "step": 724
    },
    {
      "epoch": 0.1505711318795431,
      "grad_norm": 0.4814448654651642,
      "learning_rate": 8e-05,
      "loss": 1.6141,
      "step": 725
    },
    {
      "epoch": 0.15077881619937694,
      "grad_norm": 0.497315376996994,
      "learning_rate": 8e-05,
      "loss": 1.7214,
      "step": 726
    },
    {
      "epoch": 0.1509865005192108,
      "grad_norm": 0.4816543459892273,
      "learning_rate": 8e-05,
      "loss": 1.6629,
      "step": 727
    },
    {
      "epoch": 0.15119418483904465,
      "grad_norm": 0.4759445786476135,
      "learning_rate": 8e-05,
      "loss": 1.5397,
      "step": 728
    },
    {
      "epoch": 0.15140186915887852,
      "grad_norm": 0.4926449954509735,
      "learning_rate": 8e-05,
      "loss": 1.6486,
      "step": 729
    },
    {
      "epoch": 0.15160955347871236,
      "grad_norm": 0.4702441990375519,
      "learning_rate": 8e-05,
      "loss": 1.5585,
      "step": 730
    },
    {
      "epoch": 0.15181723779854622,
      "grad_norm": 0.48316627740859985,
      "learning_rate": 8e-05,
      "loss": 1.6167,
      "step": 731
    },
    {
      "epoch": 0.15202492211838006,
      "grad_norm": 0.5125194191932678,
      "learning_rate": 8e-05,
      "loss": 1.6876,
      "step": 732
    },
    {
      "epoch": 0.1522326064382139,
      "grad_norm": 0.4879029393196106,
      "learning_rate": 8e-05,
      "loss": 1.6086,
      "step": 733
    },
    {
      "epoch": 0.15244029075804777,
      "grad_norm": 0.48626405000686646,
      "learning_rate": 8e-05,
      "loss": 1.5933,
      "step": 734
    },
    {
      "epoch": 0.1526479750778816,
      "grad_norm": 0.504183828830719,
      "learning_rate": 8e-05,
      "loss": 1.6104,
      "step": 735
    },
    {
      "epoch": 0.15285565939771548,
      "grad_norm": 0.4829785227775574,
      "learning_rate": 8e-05,
      "loss": 1.6874,
      "step": 736
    },
    {
      "epoch": 0.15306334371754932,
      "grad_norm": 0.5468804240226746,
      "learning_rate": 8e-05,
      "loss": 1.7613,
      "step": 737
    },
    {
      "epoch": 0.15327102803738318,
      "grad_norm": 0.5711941719055176,
      "learning_rate": 8e-05,
      "loss": 1.6815,
      "step": 738
    },
    {
      "epoch": 0.15347871235721702,
      "grad_norm": 0.4895186722278595,
      "learning_rate": 8e-05,
      "loss": 1.6076,
      "step": 739
    },
    {
      "epoch": 0.1536863966770509,
      "grad_norm": 0.5142952799797058,
      "learning_rate": 8e-05,
      "loss": 1.7081,
      "step": 740
    },
    {
      "epoch": 0.15389408099688473,
      "grad_norm": 0.498079776763916,
      "learning_rate": 8e-05,
      "loss": 1.6074,
      "step": 741
    },
    {
      "epoch": 0.1541017653167186,
      "grad_norm": 0.47368738055229187,
      "learning_rate": 8e-05,
      "loss": 1.6465,
      "step": 742
    },
    {
      "epoch": 0.15430944963655244,
      "grad_norm": 0.5251184105873108,
      "learning_rate": 8e-05,
      "loss": 1.7324,
      "step": 743
    },
    {
      "epoch": 0.1545171339563863,
      "grad_norm": 0.49982884526252747,
      "learning_rate": 8e-05,
      "loss": 1.7063,
      "step": 744
    },
    {
      "epoch": 0.15472481827622014,
      "grad_norm": 0.4818509519100189,
      "learning_rate": 8e-05,
      "loss": 1.6719,
      "step": 745
    },
    {
      "epoch": 0.154932502596054,
      "grad_norm": 0.48312175273895264,
      "learning_rate": 8e-05,
      "loss": 1.6511,
      "step": 746
    },
    {
      "epoch": 0.15514018691588785,
      "grad_norm": 0.4745786190032959,
      "learning_rate": 8e-05,
      "loss": 1.6419,
      "step": 747
    },
    {
      "epoch": 0.1553478712357217,
      "grad_norm": 0.5050838589668274,
      "learning_rate": 8e-05,
      "loss": 1.6486,
      "step": 748
    },
    {
      "epoch": 0.15555555555555556,
      "grad_norm": 0.4608010947704315,
      "learning_rate": 8e-05,
      "loss": 1.5949,
      "step": 749
    },
    {
      "epoch": 0.1557632398753894,
      "grad_norm": 0.499937504529953,
      "learning_rate": 8e-05,
      "loss": 1.6977,
      "step": 750
    },
    {
      "epoch": 0.15597092419522327,
      "grad_norm": 0.48160862922668457,
      "learning_rate": 8e-05,
      "loss": 1.6612,
      "step": 751
    },
    {
      "epoch": 0.1561786085150571,
      "grad_norm": 0.47431129217147827,
      "learning_rate": 8e-05,
      "loss": 1.5773,
      "step": 752
    },
    {
      "epoch": 0.15638629283489097,
      "grad_norm": 0.48737314343452454,
      "learning_rate": 8e-05,
      "loss": 1.5839,
      "step": 753
    },
    {
      "epoch": 0.1565939771547248,
      "grad_norm": 0.49073323607444763,
      "learning_rate": 8e-05,
      "loss": 1.5981,
      "step": 754
    },
    {
      "epoch": 0.15680166147455868,
      "grad_norm": 0.4805930554866791,
      "learning_rate": 8e-05,
      "loss": 1.5883,
      "step": 755
    },
    {
      "epoch": 0.15700934579439252,
      "grad_norm": 0.4898761212825775,
      "learning_rate": 8e-05,
      "loss": 1.6197,
      "step": 756
    },
    {
      "epoch": 0.15721703011422639,
      "grad_norm": 0.48299679160118103,
      "learning_rate": 8e-05,
      "loss": 1.6251,
      "step": 757
    },
    {
      "epoch": 0.15742471443406023,
      "grad_norm": 0.483012318611145,
      "learning_rate": 8e-05,
      "loss": 1.5927,
      "step": 758
    },
    {
      "epoch": 0.1576323987538941,
      "grad_norm": 0.49116986989974976,
      "learning_rate": 8e-05,
      "loss": 1.6458,
      "step": 759
    },
    {
      "epoch": 0.15784008307372793,
      "grad_norm": 0.4684026837348938,
      "learning_rate": 8e-05,
      "loss": 1.585,
      "step": 760
    },
    {
      "epoch": 0.1580477673935618,
      "grad_norm": 0.4952329099178314,
      "learning_rate": 8e-05,
      "loss": 1.6516,
      "step": 761
    },
    {
      "epoch": 0.15825545171339564,
      "grad_norm": 0.4795088768005371,
      "learning_rate": 8e-05,
      "loss": 1.6788,
      "step": 762
    },
    {
      "epoch": 0.15846313603322948,
      "grad_norm": 0.481624960899353,
      "learning_rate": 8e-05,
      "loss": 1.577,
      "step": 763
    },
    {
      "epoch": 0.15867082035306335,
      "grad_norm": 0.48516589403152466,
      "learning_rate": 8e-05,
      "loss": 1.6067,
      "step": 764
    },
    {
      "epoch": 0.1588785046728972,
      "grad_norm": 0.5102306604385376,
      "learning_rate": 8e-05,
      "loss": 1.7154,
      "step": 765
    },
    {
      "epoch": 0.15908618899273105,
      "grad_norm": 0.48237496614456177,
      "learning_rate": 8e-05,
      "loss": 1.6296,
      "step": 766
    },
    {
      "epoch": 0.1592938733125649,
      "grad_norm": 0.5044882297515869,
      "learning_rate": 8e-05,
      "loss": 1.6751,
      "step": 767
    },
    {
      "epoch": 0.15950155763239876,
      "grad_norm": 0.47297602891921997,
      "learning_rate": 8e-05,
      "loss": 1.6108,
      "step": 768
    },
    {
      "epoch": 0.1597092419522326,
      "grad_norm": 0.5018360018730164,
      "learning_rate": 8e-05,
      "loss": 1.6738,
      "step": 769
    },
    {
      "epoch": 0.15991692627206647,
      "grad_norm": 0.49589648842811584,
      "learning_rate": 8e-05,
      "loss": 1.6752,
      "step": 770
    },
    {
      "epoch": 0.1601246105919003,
      "grad_norm": 0.4882931411266327,
      "learning_rate": 8e-05,
      "loss": 1.666,
      "step": 771
    },
    {
      "epoch": 0.16033229491173417,
      "grad_norm": 0.4880520701408386,
      "learning_rate": 8e-05,
      "loss": 1.5941,
      "step": 772
    },
    {
      "epoch": 0.16053997923156801,
      "grad_norm": 0.4973679780960083,
      "learning_rate": 8e-05,
      "loss": 1.6388,
      "step": 773
    },
    {
      "epoch": 0.16074766355140188,
      "grad_norm": 0.494547039270401,
      "learning_rate": 8e-05,
      "loss": 1.6878,
      "step": 774
    },
    {
      "epoch": 0.16095534787123572,
      "grad_norm": 0.4797399938106537,
      "learning_rate": 8e-05,
      "loss": 1.6462,
      "step": 775
    },
    {
      "epoch": 0.16116303219106956,
      "grad_norm": 0.4973697364330292,
      "learning_rate": 8e-05,
      "loss": 1.6277,
      "step": 776
    },
    {
      "epoch": 0.16137071651090343,
      "grad_norm": 0.4668048322200775,
      "learning_rate": 8e-05,
      "loss": 1.5988,
      "step": 777
    },
    {
      "epoch": 0.16157840083073727,
      "grad_norm": 0.49718210101127625,
      "learning_rate": 8e-05,
      "loss": 1.6764,
      "step": 778
    },
    {
      "epoch": 0.16178608515057114,
      "grad_norm": 0.48876386880874634,
      "learning_rate": 8e-05,
      "loss": 1.6308,
      "step": 779
    },
    {
      "epoch": 0.16199376947040497,
      "grad_norm": 0.4705182909965515,
      "learning_rate": 8e-05,
      "loss": 1.6355,
      "step": 780
    },
    {
      "epoch": 0.16220145379023884,
      "grad_norm": 0.48368439078330994,
      "learning_rate": 8e-05,
      "loss": 1.6519,
      "step": 781
    },
    {
      "epoch": 0.16240913811007268,
      "grad_norm": 0.479199081659317,
      "learning_rate": 8e-05,
      "loss": 1.6046,
      "step": 782
    },
    {
      "epoch": 0.16261682242990655,
      "grad_norm": 0.4942733943462372,
      "learning_rate": 8e-05,
      "loss": 1.6996,
      "step": 783
    },
    {
      "epoch": 0.1628245067497404,
      "grad_norm": 0.5083856582641602,
      "learning_rate": 8e-05,
      "loss": 1.7383,
      "step": 784
    },
    {
      "epoch": 0.16303219106957426,
      "grad_norm": 0.46334561705589294,
      "learning_rate": 8e-05,
      "loss": 1.6044,
      "step": 785
    },
    {
      "epoch": 0.1632398753894081,
      "grad_norm": 0.5042205452919006,
      "learning_rate": 8e-05,
      "loss": 1.6773,
      "step": 786
    },
    {
      "epoch": 0.16344755970924196,
      "grad_norm": 0.4855109453201294,
      "learning_rate": 8e-05,
      "loss": 1.6655,
      "step": 787
    },
    {
      "epoch": 0.1636552440290758,
      "grad_norm": 0.4775260090827942,
      "learning_rate": 8e-05,
      "loss": 1.5763,
      "step": 788
    },
    {
      "epoch": 0.16386292834890967,
      "grad_norm": 0.49258020520210266,
      "learning_rate": 8e-05,
      "loss": 1.6628,
      "step": 789
    },
    {
      "epoch": 0.1640706126687435,
      "grad_norm": 0.4889313280582428,
      "learning_rate": 8e-05,
      "loss": 1.6056,
      "step": 790
    },
    {
      "epoch": 0.16427829698857735,
      "grad_norm": 0.49056270718574524,
      "learning_rate": 8e-05,
      "loss": 1.6066,
      "step": 791
    },
    {
      "epoch": 0.16448598130841122,
      "grad_norm": 0.47176501154899597,
      "learning_rate": 8e-05,
      "loss": 1.6171,
      "step": 792
    },
    {
      "epoch": 0.16469366562824506,
      "grad_norm": 0.5094092488288879,
      "learning_rate": 8e-05,
      "loss": 1.6178,
      "step": 793
    },
    {
      "epoch": 0.16490134994807892,
      "grad_norm": 0.4841829240322113,
      "learning_rate": 8e-05,
      "loss": 1.5962,
      "step": 794
    },
    {
      "epoch": 0.16510903426791276,
      "grad_norm": 0.4807954728603363,
      "learning_rate": 8e-05,
      "loss": 1.5233,
      "step": 795
    },
    {
      "epoch": 0.16531671858774663,
      "grad_norm": 0.4819432497024536,
      "learning_rate": 8e-05,
      "loss": 1.6909,
      "step": 796
    },
    {
      "epoch": 0.16552440290758047,
      "grad_norm": 0.513128936290741,
      "learning_rate": 8e-05,
      "loss": 1.5513,
      "step": 797
    },
    {
      "epoch": 0.16573208722741434,
      "grad_norm": 0.4752349555492401,
      "learning_rate": 8e-05,
      "loss": 1.5488,
      "step": 798
    },
    {
      "epoch": 0.16593977154724818,
      "grad_norm": 0.48235711455345154,
      "learning_rate": 8e-05,
      "loss": 1.6281,
      "step": 799
    },
    {
      "epoch": 0.16614745586708204,
      "grad_norm": 0.49711811542510986,
      "learning_rate": 8e-05,
      "loss": 1.6796,
      "step": 800
    },
    {
      "epoch": 0.16635514018691588,
      "grad_norm": 0.5144046545028687,
      "learning_rate": 8e-05,
      "loss": 1.6711,
      "step": 801
    },
    {
      "epoch": 0.16656282450674975,
      "grad_norm": 0.506710410118103,
      "learning_rate": 8e-05,
      "loss": 1.6507,
      "step": 802
    },
    {
      "epoch": 0.1667705088265836,
      "grad_norm": 0.5138517618179321,
      "learning_rate": 8e-05,
      "loss": 1.6207,
      "step": 803
    },
    {
      "epoch": 0.16697819314641746,
      "grad_norm": 0.480153352022171,
      "learning_rate": 8e-05,
      "loss": 1.6877,
      "step": 804
    },
    {
      "epoch": 0.1671858774662513,
      "grad_norm": 0.5096755623817444,
      "learning_rate": 8e-05,
      "loss": 1.7058,
      "step": 805
    },
    {
      "epoch": 0.16739356178608514,
      "grad_norm": 0.49383774399757385,
      "learning_rate": 8e-05,
      "loss": 1.6983,
      "step": 806
    },
    {
      "epoch": 0.167601246105919,
      "grad_norm": 0.5019107460975647,
      "learning_rate": 8e-05,
      "loss": 1.6543,
      "step": 807
    },
    {
      "epoch": 0.16780893042575284,
      "grad_norm": 0.48990175127983093,
      "learning_rate": 8e-05,
      "loss": 1.6206,
      "step": 808
    },
    {
      "epoch": 0.1680166147455867,
      "grad_norm": 0.4764305055141449,
      "learning_rate": 8e-05,
      "loss": 1.6453,
      "step": 809
    },
    {
      "epoch": 0.16822429906542055,
      "grad_norm": 0.5044610500335693,
      "learning_rate": 8e-05,
      "loss": 1.6739,
      "step": 810
    },
    {
      "epoch": 0.16843198338525442,
      "grad_norm": 0.48134395480155945,
      "learning_rate": 8e-05,
      "loss": 1.6153,
      "step": 811
    },
    {
      "epoch": 0.16863966770508826,
      "grad_norm": 0.4967174232006073,
      "learning_rate": 8e-05,
      "loss": 1.677,
      "step": 812
    },
    {
      "epoch": 0.16884735202492213,
      "grad_norm": 0.47364965081214905,
      "learning_rate": 8e-05,
      "loss": 1.555,
      "step": 813
    },
    {
      "epoch": 0.16905503634475597,
      "grad_norm": 0.49788594245910645,
      "learning_rate": 8e-05,
      "loss": 1.6635,
      "step": 814
    },
    {
      "epoch": 0.16926272066458983,
      "grad_norm": 0.49295979738235474,
      "learning_rate": 8e-05,
      "loss": 1.6627,
      "step": 815
    },
    {
      "epoch": 0.16947040498442367,
      "grad_norm": 0.5006287693977356,
      "learning_rate": 8e-05,
      "loss": 1.6183,
      "step": 816
    },
    {
      "epoch": 0.16967808930425754,
      "grad_norm": 0.4745385944843292,
      "learning_rate": 8e-05,
      "loss": 1.585,
      "step": 817
    },
    {
      "epoch": 0.16988577362409138,
      "grad_norm": 0.5007401704788208,
      "learning_rate": 8e-05,
      "loss": 1.6358,
      "step": 818
    },
    {
      "epoch": 0.17009345794392525,
      "grad_norm": 0.4691254198551178,
      "learning_rate": 8e-05,
      "loss": 1.605,
      "step": 819
    },
    {
      "epoch": 0.1703011422637591,
      "grad_norm": 0.48528677225112915,
      "learning_rate": 8e-05,
      "loss": 1.6889,
      "step": 820
    },
    {
      "epoch": 0.17050882658359293,
      "grad_norm": 0.48160579800605774,
      "learning_rate": 8e-05,
      "loss": 1.6284,
      "step": 821
    },
    {
      "epoch": 0.1707165109034268,
      "grad_norm": 0.4866841435432434,
      "learning_rate": 8e-05,
      "loss": 1.635,
      "step": 822
    },
    {
      "epoch": 0.17092419522326063,
      "grad_norm": 0.4998115003108978,
      "learning_rate": 8e-05,
      "loss": 1.7034,
      "step": 823
    },
    {
      "epoch": 0.1711318795430945,
      "grad_norm": 0.49009475111961365,
      "learning_rate": 8e-05,
      "loss": 1.6505,
      "step": 824
    },
    {
      "epoch": 0.17133956386292834,
      "grad_norm": 0.47515684366226196,
      "learning_rate": 8e-05,
      "loss": 1.6409,
      "step": 825
    },
    {
      "epoch": 0.1715472481827622,
      "grad_norm": 0.5075555443763733,
      "learning_rate": 8e-05,
      "loss": 1.6623,
      "step": 826
    },
    {
      "epoch": 0.17175493250259605,
      "grad_norm": 0.4834977984428406,
      "learning_rate": 8e-05,
      "loss": 1.6343,
      "step": 827
    },
    {
      "epoch": 0.17196261682242991,
      "grad_norm": 0.4810510575771332,
      "learning_rate": 8e-05,
      "loss": 1.5518,
      "step": 828
    },
    {
      "epoch": 0.17217030114226375,
      "grad_norm": 0.5000099539756775,
      "learning_rate": 8e-05,
      "loss": 1.6423,
      "step": 829
    },
    {
      "epoch": 0.17237798546209762,
      "grad_norm": 0.4854555130004883,
      "learning_rate": 8e-05,
      "loss": 1.6345,
      "step": 830
    },
    {
      "epoch": 0.17258566978193146,
      "grad_norm": 0.47438621520996094,
      "learning_rate": 8e-05,
      "loss": 1.5127,
      "step": 831
    },
    {
      "epoch": 0.17279335410176533,
      "grad_norm": 0.4803098440170288,
      "learning_rate": 8e-05,
      "loss": 1.5919,
      "step": 832
    },
    {
      "epoch": 0.17300103842159917,
      "grad_norm": 0.48187175393104553,
      "learning_rate": 8e-05,
      "loss": 1.6423,
      "step": 833
    },
    {
      "epoch": 0.17320872274143304,
      "grad_norm": 0.4821400046348572,
      "learning_rate": 8e-05,
      "loss": 1.6798,
      "step": 834
    },
    {
      "epoch": 0.17341640706126688,
      "grad_norm": 0.4715557098388672,
      "learning_rate": 8e-05,
      "loss": 1.6162,
      "step": 835
    },
    {
      "epoch": 0.17362409138110071,
      "grad_norm": 0.48765188455581665,
      "learning_rate": 8e-05,
      "loss": 1.6846,
      "step": 836
    },
    {
      "epoch": 0.17383177570093458,
      "grad_norm": 0.5081665515899658,
      "learning_rate": 8e-05,
      "loss": 1.6857,
      "step": 837
    },
    {
      "epoch": 0.17403946002076842,
      "grad_norm": 0.49220722913742065,
      "learning_rate": 8e-05,
      "loss": 1.6495,
      "step": 838
    },
    {
      "epoch": 0.1742471443406023,
      "grad_norm": 0.461845338344574,
      "learning_rate": 8e-05,
      "loss": 1.6098,
      "step": 839
    },
    {
      "epoch": 0.17445482866043613,
      "grad_norm": 0.4715192914009094,
      "learning_rate": 8e-05,
      "loss": 1.5672,
      "step": 840
    },
    {
      "epoch": 0.17466251298027,
      "grad_norm": 0.4969162940979004,
      "learning_rate": 8e-05,
      "loss": 1.6727,
      "step": 841
    },
    {
      "epoch": 0.17487019730010384,
      "grad_norm": 0.4748702645301819,
      "learning_rate": 8e-05,
      "loss": 1.6493,
      "step": 842
    },
    {
      "epoch": 0.1750778816199377,
      "grad_norm": 0.47799915075302124,
      "learning_rate": 8e-05,
      "loss": 1.5978,
      "step": 843
    },
    {
      "epoch": 0.17528556593977154,
      "grad_norm": 0.4815102815628052,
      "learning_rate": 8e-05,
      "loss": 1.587,
      "step": 844
    },
    {
      "epoch": 0.1754932502596054,
      "grad_norm": 0.4794101417064667,
      "learning_rate": 8e-05,
      "loss": 1.5926,
      "step": 845
    },
    {
      "epoch": 0.17570093457943925,
      "grad_norm": 0.480980783700943,
      "learning_rate": 8e-05,
      "loss": 1.6106,
      "step": 846
    },
    {
      "epoch": 0.17590861889927312,
      "grad_norm": 0.483927458524704,
      "learning_rate": 8e-05,
      "loss": 1.6169,
      "step": 847
    },
    {
      "epoch": 0.17611630321910696,
      "grad_norm": 0.4977688193321228,
      "learning_rate": 8e-05,
      "loss": 1.6233,
      "step": 848
    },
    {
      "epoch": 0.1763239875389408,
      "grad_norm": 0.4976138770580292,
      "learning_rate": 8e-05,
      "loss": 1.7223,
      "step": 849
    },
    {
      "epoch": 0.17653167185877466,
      "grad_norm": 0.4990401864051819,
      "learning_rate": 8e-05,
      "loss": 1.6844,
      "step": 850
    },
    {
      "epoch": 0.1767393561786085,
      "grad_norm": 0.5020548105239868,
      "learning_rate": 8e-05,
      "loss": 1.6672,
      "step": 851
    },
    {
      "epoch": 0.17694704049844237,
      "grad_norm": 0.5174390077590942,
      "learning_rate": 8e-05,
      "loss": 1.634,
      "step": 852
    },
    {
      "epoch": 0.1771547248182762,
      "grad_norm": 0.4973217248916626,
      "learning_rate": 8e-05,
      "loss": 1.6551,
      "step": 853
    },
    {
      "epoch": 0.17736240913811008,
      "grad_norm": 0.504899263381958,
      "learning_rate": 8e-05,
      "loss": 1.6016,
      "step": 854
    },
    {
      "epoch": 0.17757009345794392,
      "grad_norm": 0.5085044503211975,
      "learning_rate": 8e-05,
      "loss": 1.6352,
      "step": 855
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 0.4816834330558777,
      "learning_rate": 8e-05,
      "loss": 1.5832,
      "step": 856
    },
    {
      "epoch": 0.17798546209761162,
      "grad_norm": 0.5155865550041199,
      "learning_rate": 8e-05,
      "loss": 1.6151,
      "step": 857
    },
    {
      "epoch": 0.1781931464174455,
      "grad_norm": 0.48705244064331055,
      "learning_rate": 8e-05,
      "loss": 1.648,
      "step": 858
    },
    {
      "epoch": 0.17840083073727933,
      "grad_norm": 0.5086380243301392,
      "learning_rate": 8e-05,
      "loss": 1.6601,
      "step": 859
    },
    {
      "epoch": 0.1786085150571132,
      "grad_norm": 0.46647846698760986,
      "learning_rate": 8e-05,
      "loss": 1.6007,
      "step": 860
    },
    {
      "epoch": 0.17881619937694704,
      "grad_norm": 0.4949185252189636,
      "learning_rate": 8e-05,
      "loss": 1.6181,
      "step": 861
    },
    {
      "epoch": 0.1790238836967809,
      "grad_norm": 0.4831303656101227,
      "learning_rate": 8e-05,
      "loss": 1.6037,
      "step": 862
    },
    {
      "epoch": 0.17923156801661475,
      "grad_norm": 0.5041294693946838,
      "learning_rate": 8e-05,
      "loss": 1.6582,
      "step": 863
    },
    {
      "epoch": 0.17943925233644858,
      "grad_norm": 0.5045524835586548,
      "learning_rate": 8e-05,
      "loss": 1.6141,
      "step": 864
    },
    {
      "epoch": 0.17964693665628245,
      "grad_norm": 0.49055367708206177,
      "learning_rate": 8e-05,
      "loss": 1.6371,
      "step": 865
    },
    {
      "epoch": 0.1798546209761163,
      "grad_norm": 0.5056166648864746,
      "learning_rate": 8e-05,
      "loss": 1.6949,
      "step": 866
    },
    {
      "epoch": 0.18006230529595016,
      "grad_norm": 0.4675775170326233,
      "learning_rate": 8e-05,
      "loss": 1.5697,
      "step": 867
    },
    {
      "epoch": 0.180269989615784,
      "grad_norm": 0.490346223115921,
      "learning_rate": 8e-05,
      "loss": 1.6092,
      "step": 868
    },
    {
      "epoch": 0.18047767393561787,
      "grad_norm": 0.4684089720249176,
      "learning_rate": 8e-05,
      "loss": 1.5648,
      "step": 869
    },
    {
      "epoch": 0.1806853582554517,
      "grad_norm": 0.46101874113082886,
      "learning_rate": 8e-05,
      "loss": 1.566,
      "step": 870
    },
    {
      "epoch": 0.18089304257528557,
      "grad_norm": 0.4861421287059784,
      "learning_rate": 8e-05,
      "loss": 1.6511,
      "step": 871
    },
    {
      "epoch": 0.1811007268951194,
      "grad_norm": 0.48315370082855225,
      "learning_rate": 8e-05,
      "loss": 1.6527,
      "step": 872
    },
    {
      "epoch": 0.18130841121495328,
      "grad_norm": 0.48477229475975037,
      "learning_rate": 8e-05,
      "loss": 1.6477,
      "step": 873
    },
    {
      "epoch": 0.18151609553478712,
      "grad_norm": 0.4834623336791992,
      "learning_rate": 8e-05,
      "loss": 1.5799,
      "step": 874
    },
    {
      "epoch": 0.181723779854621,
      "grad_norm": 0.4903808534145355,
      "learning_rate": 8e-05,
      "loss": 1.6191,
      "step": 875
    },
    {
      "epoch": 0.18193146417445483,
      "grad_norm": 0.4885776937007904,
      "learning_rate": 8e-05,
      "loss": 1.6815,
      "step": 876
    },
    {
      "epoch": 0.1821391484942887,
      "grad_norm": 0.48751139640808105,
      "learning_rate": 8e-05,
      "loss": 1.6905,
      "step": 877
    },
    {
      "epoch": 0.18234683281412253,
      "grad_norm": 0.48983898758888245,
      "learning_rate": 8e-05,
      "loss": 1.5902,
      "step": 878
    },
    {
      "epoch": 0.18255451713395637,
      "grad_norm": 0.47432875633239746,
      "learning_rate": 8e-05,
      "loss": 1.6571,
      "step": 879
    },
    {
      "epoch": 0.18276220145379024,
      "grad_norm": 0.4667503535747528,
      "learning_rate": 8e-05,
      "loss": 1.6198,
      "step": 880
    },
    {
      "epoch": 0.18296988577362408,
      "grad_norm": 0.5021142363548279,
      "learning_rate": 8e-05,
      "loss": 1.7175,
      "step": 881
    },
    {
      "epoch": 0.18317757009345795,
      "grad_norm": 0.46607160568237305,
      "learning_rate": 8e-05,
      "loss": 1.5967,
      "step": 882
    },
    {
      "epoch": 0.1833852544132918,
      "grad_norm": 0.4718491733074188,
      "learning_rate": 8e-05,
      "loss": 1.5974,
      "step": 883
    },
    {
      "epoch": 0.18359293873312565,
      "grad_norm": 0.49445825815200806,
      "learning_rate": 8e-05,
      "loss": 1.6271,
      "step": 884
    },
    {
      "epoch": 0.1838006230529595,
      "grad_norm": 0.46848443150520325,
      "learning_rate": 8e-05,
      "loss": 1.6061,
      "step": 885
    },
    {
      "epoch": 0.18400830737279336,
      "grad_norm": 0.46459993720054626,
      "learning_rate": 8e-05,
      "loss": 1.5613,
      "step": 886
    },
    {
      "epoch": 0.1842159916926272,
      "grad_norm": 0.47281578183174133,
      "learning_rate": 8e-05,
      "loss": 1.704,
      "step": 887
    },
    {
      "epoch": 0.18442367601246107,
      "grad_norm": 0.456195592880249,
      "learning_rate": 8e-05,
      "loss": 1.559,
      "step": 888
    },
    {
      "epoch": 0.1846313603322949,
      "grad_norm": 0.5002410411834717,
      "learning_rate": 8e-05,
      "loss": 1.6179,
      "step": 889
    },
    {
      "epoch": 0.18483904465212878,
      "grad_norm": 0.4889829158782959,
      "learning_rate": 8e-05,
      "loss": 1.6081,
      "step": 890
    },
    {
      "epoch": 0.18504672897196262,
      "grad_norm": 0.5022076964378357,
      "learning_rate": 8e-05,
      "loss": 1.629,
      "step": 891
    },
    {
      "epoch": 0.18525441329179648,
      "grad_norm": 0.496134489774704,
      "learning_rate": 8e-05,
      "loss": 1.611,
      "step": 892
    },
    {
      "epoch": 0.18546209761163032,
      "grad_norm": 0.4800807237625122,
      "learning_rate": 8e-05,
      "loss": 1.68,
      "step": 893
    },
    {
      "epoch": 0.18566978193146416,
      "grad_norm": 0.4833254814147949,
      "learning_rate": 8e-05,
      "loss": 1.6493,
      "step": 894
    },
    {
      "epoch": 0.18587746625129803,
      "grad_norm": 0.47736915946006775,
      "learning_rate": 8e-05,
      "loss": 1.5986,
      "step": 895
    },
    {
      "epoch": 0.18608515057113187,
      "grad_norm": 0.5091724395751953,
      "learning_rate": 8e-05,
      "loss": 1.6303,
      "step": 896
    },
    {
      "epoch": 0.18629283489096574,
      "grad_norm": 0.46926015615463257,
      "learning_rate": 8e-05,
      "loss": 1.5778,
      "step": 897
    },
    {
      "epoch": 0.18650051921079958,
      "grad_norm": 0.48769834637641907,
      "learning_rate": 8e-05,
      "loss": 1.6396,
      "step": 898
    },
    {
      "epoch": 0.18670820353063344,
      "grad_norm": 0.47755125164985657,
      "learning_rate": 8e-05,
      "loss": 1.6374,
      "step": 899
    },
    {
      "epoch": 0.18691588785046728,
      "grad_norm": 0.4844597280025482,
      "learning_rate": 8e-05,
      "loss": 1.6072,
      "step": 900
    },
    {
      "epoch": 0.18712357217030115,
      "grad_norm": 0.5078486800193787,
      "learning_rate": 8e-05,
      "loss": 1.6898,
      "step": 901
    },
    {
      "epoch": 0.187331256490135,
      "grad_norm": 0.4972101151943207,
      "learning_rate": 8e-05,
      "loss": 1.7502,
      "step": 902
    },
    {
      "epoch": 0.18753894080996886,
      "grad_norm": 0.4829871952533722,
      "learning_rate": 8e-05,
      "loss": 1.615,
      "step": 903
    },
    {
      "epoch": 0.1877466251298027,
      "grad_norm": 0.5250853300094604,
      "learning_rate": 8e-05,
      "loss": 1.6795,
      "step": 904
    },
    {
      "epoch": 0.18795430944963656,
      "grad_norm": 0.4843866229057312,
      "learning_rate": 8e-05,
      "loss": 1.6664,
      "step": 905
    },
    {
      "epoch": 0.1881619937694704,
      "grad_norm": 0.4833875894546509,
      "learning_rate": 8e-05,
      "loss": 1.5387,
      "step": 906
    },
    {
      "epoch": 0.18836967808930427,
      "grad_norm": 0.5005854368209839,
      "learning_rate": 8e-05,
      "loss": 1.5622,
      "step": 907
    },
    {
      "epoch": 0.1885773624091381,
      "grad_norm": 0.47214603424072266,
      "learning_rate": 8e-05,
      "loss": 1.6011,
      "step": 908
    },
    {
      "epoch": 0.18878504672897195,
      "grad_norm": 0.5020946264266968,
      "learning_rate": 8e-05,
      "loss": 1.6588,
      "step": 909
    },
    {
      "epoch": 0.18899273104880582,
      "grad_norm": 0.5007855892181396,
      "learning_rate": 8e-05,
      "loss": 1.6688,
      "step": 910
    },
    {
      "epoch": 0.18920041536863966,
      "grad_norm": 0.5172010064125061,
      "learning_rate": 8e-05,
      "loss": 1.6751,
      "step": 911
    },
    {
      "epoch": 0.18940809968847352,
      "grad_norm": 0.4866928160190582,
      "learning_rate": 8e-05,
      "loss": 1.56,
      "step": 912
    },
    {
      "epoch": 0.18961578400830736,
      "grad_norm": 0.47611889243125916,
      "learning_rate": 8e-05,
      "loss": 1.5304,
      "step": 913
    },
    {
      "epoch": 0.18982346832814123,
      "grad_norm": 0.4758463203907013,
      "learning_rate": 8e-05,
      "loss": 1.6201,
      "step": 914
    },
    {
      "epoch": 0.19003115264797507,
      "grad_norm": 0.4987952709197998,
      "learning_rate": 8e-05,
      "loss": 1.6785,
      "step": 915
    },
    {
      "epoch": 0.19023883696780894,
      "grad_norm": 0.5098257064819336,
      "learning_rate": 8e-05,
      "loss": 1.6758,
      "step": 916
    },
    {
      "epoch": 0.19044652128764278,
      "grad_norm": 0.4878323972225189,
      "learning_rate": 8e-05,
      "loss": 1.5964,
      "step": 917
    },
    {
      "epoch": 0.19065420560747665,
      "grad_norm": 0.49030864238739014,
      "learning_rate": 8e-05,
      "loss": 1.6166,
      "step": 918
    },
    {
      "epoch": 0.19086188992731049,
      "grad_norm": 0.4641314446926117,
      "learning_rate": 8e-05,
      "loss": 1.6265,
      "step": 919
    },
    {
      "epoch": 0.19106957424714435,
      "grad_norm": 0.49231404066085815,
      "learning_rate": 8e-05,
      "loss": 1.6691,
      "step": 920
    },
    {
      "epoch": 0.1912772585669782,
      "grad_norm": 0.5089201331138611,
      "learning_rate": 8e-05,
      "loss": 1.6911,
      "step": 921
    },
    {
      "epoch": 0.19148494288681206,
      "grad_norm": 0.4495040774345398,
      "learning_rate": 8e-05,
      "loss": 1.5579,
      "step": 922
    },
    {
      "epoch": 0.1916926272066459,
      "grad_norm": 0.47652044892311096,
      "learning_rate": 8e-05,
      "loss": 1.5685,
      "step": 923
    },
    {
      "epoch": 0.19190031152647974,
      "grad_norm": 0.4827858805656433,
      "learning_rate": 8e-05,
      "loss": 1.6168,
      "step": 924
    },
    {
      "epoch": 0.1921079958463136,
      "grad_norm": 0.47965332865715027,
      "learning_rate": 8e-05,
      "loss": 1.5983,
      "step": 925
    },
    {
      "epoch": 0.19231568016614745,
      "grad_norm": 0.4766390919685364,
      "learning_rate": 8e-05,
      "loss": 1.599,
      "step": 926
    },
    {
      "epoch": 0.1925233644859813,
      "grad_norm": 0.48412805795669556,
      "learning_rate": 8e-05,
      "loss": 1.6749,
      "step": 927
    },
    {
      "epoch": 0.19273104880581515,
      "grad_norm": 0.47555628418922424,
      "learning_rate": 8e-05,
      "loss": 1.5841,
      "step": 928
    },
    {
      "epoch": 0.19293873312564902,
      "grad_norm": 0.48913460969924927,
      "learning_rate": 8e-05,
      "loss": 1.6512,
      "step": 929
    },
    {
      "epoch": 0.19314641744548286,
      "grad_norm": 0.47544533014297485,
      "learning_rate": 8e-05,
      "loss": 1.6341,
      "step": 930
    },
    {
      "epoch": 0.19335410176531673,
      "grad_norm": 0.4919319450855255,
      "learning_rate": 8e-05,
      "loss": 1.6381,
      "step": 931
    },
    {
      "epoch": 0.19356178608515057,
      "grad_norm": 0.49583807587623596,
      "learning_rate": 8e-05,
      "loss": 1.6419,
      "step": 932
    },
    {
      "epoch": 0.19376947040498443,
      "grad_norm": 0.5070441961288452,
      "learning_rate": 8e-05,
      "loss": 1.6657,
      "step": 933
    },
    {
      "epoch": 0.19397715472481827,
      "grad_norm": 0.5061896443367004,
      "learning_rate": 8e-05,
      "loss": 1.6081,
      "step": 934
    },
    {
      "epoch": 0.19418483904465214,
      "grad_norm": 0.5091527104377747,
      "learning_rate": 8e-05,
      "loss": 1.6912,
      "step": 935
    },
    {
      "epoch": 0.19439252336448598,
      "grad_norm": 0.5117039084434509,
      "learning_rate": 8e-05,
      "loss": 1.632,
      "step": 936
    },
    {
      "epoch": 0.19460020768431982,
      "grad_norm": 0.4791244864463806,
      "learning_rate": 8e-05,
      "loss": 1.6062,
      "step": 937
    },
    {
      "epoch": 0.1948078920041537,
      "grad_norm": 0.5442696213722229,
      "learning_rate": 8e-05,
      "loss": 1.6536,
      "step": 938
    },
    {
      "epoch": 0.19501557632398753,
      "grad_norm": 0.48307886719703674,
      "learning_rate": 8e-05,
      "loss": 1.5979,
      "step": 939
    },
    {
      "epoch": 0.1952232606438214,
      "grad_norm": 0.5369094014167786,
      "learning_rate": 8e-05,
      "loss": 1.5968,
      "step": 940
    },
    {
      "epoch": 0.19543094496365523,
      "grad_norm": 0.4949534237384796,
      "learning_rate": 8e-05,
      "loss": 1.6489,
      "step": 941
    },
    {
      "epoch": 0.1956386292834891,
      "grad_norm": 0.5377048850059509,
      "learning_rate": 8e-05,
      "loss": 1.5288,
      "step": 942
    },
    {
      "epoch": 0.19584631360332294,
      "grad_norm": 0.49918225407600403,
      "learning_rate": 8e-05,
      "loss": 1.5978,
      "step": 943
    },
    {
      "epoch": 0.1960539979231568,
      "grad_norm": 0.5020906329154968,
      "learning_rate": 8e-05,
      "loss": 1.6248,
      "step": 944
    },
    {
      "epoch": 0.19626168224299065,
      "grad_norm": 0.5110490918159485,
      "learning_rate": 8e-05,
      "loss": 1.5623,
      "step": 945
    },
    {
      "epoch": 0.19646936656282452,
      "grad_norm": 0.4651744067668915,
      "learning_rate": 8e-05,
      "loss": 1.554,
      "step": 946
    },
    {
      "epoch": 0.19667705088265836,
      "grad_norm": 0.5259853005409241,
      "learning_rate": 8e-05,
      "loss": 1.619,
      "step": 947
    },
    {
      "epoch": 0.19688473520249222,
      "grad_norm": 0.5092069506645203,
      "learning_rate": 8e-05,
      "loss": 1.6075,
      "step": 948
    },
    {
      "epoch": 0.19709241952232606,
      "grad_norm": 0.47999849915504456,
      "learning_rate": 8e-05,
      "loss": 1.4621,
      "step": 949
    },
    {
      "epoch": 0.19730010384215993,
      "grad_norm": 0.5035858750343323,
      "learning_rate": 8e-05,
      "loss": 1.5704,
      "step": 950
    },
    {
      "epoch": 0.19750778816199377,
      "grad_norm": 0.5029184222221375,
      "learning_rate": 8e-05,
      "loss": 1.691,
      "step": 951
    },
    {
      "epoch": 0.1977154724818276,
      "grad_norm": 0.5592398047447205,
      "learning_rate": 8e-05,
      "loss": 1.6728,
      "step": 952
    },
    {
      "epoch": 0.19792315680166148,
      "grad_norm": 0.5204004645347595,
      "learning_rate": 8e-05,
      "loss": 1.5816,
      "step": 953
    },
    {
      "epoch": 0.19813084112149532,
      "grad_norm": 0.5047582983970642,
      "learning_rate": 8e-05,
      "loss": 1.6473,
      "step": 954
    },
    {
      "epoch": 0.19833852544132918,
      "grad_norm": 0.5084308385848999,
      "learning_rate": 8e-05,
      "loss": 1.6297,
      "step": 955
    },
    {
      "epoch": 0.19854620976116302,
      "grad_norm": 0.4851769506931305,
      "learning_rate": 8e-05,
      "loss": 1.676,
      "step": 956
    },
    {
      "epoch": 0.1987538940809969,
      "grad_norm": 0.48817214369773865,
      "learning_rate": 8e-05,
      "loss": 1.6213,
      "step": 957
    },
    {
      "epoch": 0.19896157840083073,
      "grad_norm": 0.5137952566146851,
      "learning_rate": 8e-05,
      "loss": 1.6807,
      "step": 958
    },
    {
      "epoch": 0.1991692627206646,
      "grad_norm": 0.494804322719574,
      "learning_rate": 8e-05,
      "loss": 1.633,
      "step": 959
    },
    {
      "epoch": 0.19937694704049844,
      "grad_norm": 0.4694535434246063,
      "learning_rate": 8e-05,
      "loss": 1.5782,
      "step": 960
    },
    {
      "epoch": 0.1995846313603323,
      "grad_norm": 0.4649278223514557,
      "learning_rate": 8e-05,
      "loss": 1.5167,
      "step": 961
    },
    {
      "epoch": 0.19979231568016614,
      "grad_norm": 0.49113568663597107,
      "learning_rate": 8e-05,
      "loss": 1.63,
      "step": 962
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.4700087308883667,
      "learning_rate": 8e-05,
      "loss": 1.5999,
      "step": 963
    },
    {
      "epoch": 0.20020768431983385,
      "grad_norm": 0.4888930320739746,
      "learning_rate": 8e-05,
      "loss": 1.6966,
      "step": 964
    },
    {
      "epoch": 0.20041536863966772,
      "grad_norm": 0.4866234362125397,
      "learning_rate": 8e-05,
      "loss": 1.6195,
      "step": 965
    },
    {
      "epoch": 0.20062305295950156,
      "grad_norm": 0.5145866870880127,
      "learning_rate": 8e-05,
      "loss": 1.6721,
      "step": 966
    },
    {
      "epoch": 0.2008307372793354,
      "grad_norm": 0.47302988171577454,
      "learning_rate": 8e-05,
      "loss": 1.6644,
      "step": 967
    },
    {
      "epoch": 0.20103842159916926,
      "grad_norm": 0.48115986585617065,
      "learning_rate": 8e-05,
      "loss": 1.6557,
      "step": 968
    },
    {
      "epoch": 0.2012461059190031,
      "grad_norm": 0.4676402807235718,
      "learning_rate": 8e-05,
      "loss": 1.5842,
      "step": 969
    },
    {
      "epoch": 0.20145379023883697,
      "grad_norm": 0.4860267639160156,
      "learning_rate": 8e-05,
      "loss": 1.6519,
      "step": 970
    },
    {
      "epoch": 0.2016614745586708,
      "grad_norm": 0.4892720580101013,
      "learning_rate": 8e-05,
      "loss": 1.5545,
      "step": 971
    },
    {
      "epoch": 0.20186915887850468,
      "grad_norm": 0.46676620841026306,
      "learning_rate": 8e-05,
      "loss": 1.553,
      "step": 972
    },
    {
      "epoch": 0.20207684319833852,
      "grad_norm": 0.47388455271720886,
      "learning_rate": 8e-05,
      "loss": 1.5923,
      "step": 973
    },
    {
      "epoch": 0.20228452751817239,
      "grad_norm": 0.4880634546279907,
      "learning_rate": 8e-05,
      "loss": 1.5832,
      "step": 974
    },
    {
      "epoch": 0.20249221183800623,
      "grad_norm": 0.4873626232147217,
      "learning_rate": 8e-05,
      "loss": 1.6224,
      "step": 975
    },
    {
      "epoch": 0.2026998961578401,
      "grad_norm": 0.5022002458572388,
      "learning_rate": 8e-05,
      "loss": 1.6834,
      "step": 976
    },
    {
      "epoch": 0.20290758047767393,
      "grad_norm": 0.47066470980644226,
      "learning_rate": 8e-05,
      "loss": 1.5887,
      "step": 977
    },
    {
      "epoch": 0.2031152647975078,
      "grad_norm": 0.4675566554069519,
      "learning_rate": 8e-05,
      "loss": 1.6026,
      "step": 978
    },
    {
      "epoch": 0.20332294911734164,
      "grad_norm": 0.5087757110595703,
      "learning_rate": 8e-05,
      "loss": 1.6192,
      "step": 979
    },
    {
      "epoch": 0.2035306334371755,
      "grad_norm": 0.49187105894088745,
      "learning_rate": 8e-05,
      "loss": 1.6126,
      "step": 980
    },
    {
      "epoch": 0.20373831775700935,
      "grad_norm": 0.4877180755138397,
      "learning_rate": 8e-05,
      "loss": 1.6781,
      "step": 981
    },
    {
      "epoch": 0.20394600207684319,
      "grad_norm": 0.47267571091651917,
      "learning_rate": 8e-05,
      "loss": 1.646,
      "step": 982
    },
    {
      "epoch": 0.20415368639667705,
      "grad_norm": 0.4767378270626068,
      "learning_rate": 8e-05,
      "loss": 1.5953,
      "step": 983
    },
    {
      "epoch": 0.2043613707165109,
      "grad_norm": 0.48661693930625916,
      "learning_rate": 8e-05,
      "loss": 1.5469,
      "step": 984
    },
    {
      "epoch": 0.20456905503634476,
      "grad_norm": 0.4928230047225952,
      "learning_rate": 8e-05,
      "loss": 1.7054,
      "step": 985
    },
    {
      "epoch": 0.2047767393561786,
      "grad_norm": 0.4975184202194214,
      "learning_rate": 8e-05,
      "loss": 1.594,
      "step": 986
    },
    {
      "epoch": 0.20498442367601247,
      "grad_norm": 0.4876691997051239,
      "learning_rate": 8e-05,
      "loss": 1.6356,
      "step": 987
    },
    {
      "epoch": 0.2051921079958463,
      "grad_norm": 0.4785061776638031,
      "learning_rate": 8e-05,
      "loss": 1.6646,
      "step": 988
    },
    {
      "epoch": 0.20539979231568017,
      "grad_norm": 0.49012821912765503,
      "learning_rate": 8e-05,
      "loss": 1.689,
      "step": 989
    },
    {
      "epoch": 0.205607476635514,
      "grad_norm": 0.45855405926704407,
      "learning_rate": 8e-05,
      "loss": 1.6318,
      "step": 990
    },
    {
      "epoch": 0.20581516095534788,
      "grad_norm": 0.4825817048549652,
      "learning_rate": 8e-05,
      "loss": 1.6189,
      "step": 991
    },
    {
      "epoch": 0.20602284527518172,
      "grad_norm": 0.4723818004131317,
      "learning_rate": 8e-05,
      "loss": 1.5096,
      "step": 992
    },
    {
      "epoch": 0.2062305295950156,
      "grad_norm": 0.46535414457321167,
      "learning_rate": 8e-05,
      "loss": 1.6112,
      "step": 993
    },
    {
      "epoch": 0.20643821391484943,
      "grad_norm": 0.4890800714492798,
      "learning_rate": 8e-05,
      "loss": 1.6472,
      "step": 994
    },
    {
      "epoch": 0.2066458982346833,
      "grad_norm": 0.48340532183647156,
      "learning_rate": 8e-05,
      "loss": 1.6347,
      "step": 995
    },
    {
      "epoch": 0.20685358255451713,
      "grad_norm": 0.490472674369812,
      "learning_rate": 8e-05,
      "loss": 1.6745,
      "step": 996
    },
    {
      "epoch": 0.20706126687435097,
      "grad_norm": 0.5247419476509094,
      "learning_rate": 8e-05,
      "loss": 1.7016,
      "step": 997
    },
    {
      "epoch": 0.20726895119418484,
      "grad_norm": 0.4798457622528076,
      "learning_rate": 8e-05,
      "loss": 1.6581,
      "step": 998
    },
    {
      "epoch": 0.20747663551401868,
      "grad_norm": 0.5022425651550293,
      "learning_rate": 8e-05,
      "loss": 1.6491,
      "step": 999
    },
    {
      "epoch": 0.20768431983385255,
      "grad_norm": 0.5252692699432373,
      "learning_rate": 8e-05,
      "loss": 1.624,
      "step": 1000
    },
    {
      "epoch": 0.2078920041536864,
      "grad_norm": 0.4893631041049957,
      "learning_rate": 8e-05,
      "loss": 1.6016,
      "step": 1001
    },
    {
      "epoch": 0.20809968847352026,
      "grad_norm": 0.49601900577545166,
      "learning_rate": 8e-05,
      "loss": 1.6679,
      "step": 1002
    },
    {
      "epoch": 0.2083073727933541,
      "grad_norm": 0.484885036945343,
      "learning_rate": 8e-05,
      "loss": 1.5687,
      "step": 1003
    },
    {
      "epoch": 0.20851505711318796,
      "grad_norm": 0.4795399606227875,
      "learning_rate": 8e-05,
      "loss": 1.5629,
      "step": 1004
    },
    {
      "epoch": 0.2087227414330218,
      "grad_norm": 0.5045424699783325,
      "learning_rate": 8e-05,
      "loss": 1.6474,
      "step": 1005
    },
    {
      "epoch": 0.20893042575285567,
      "grad_norm": 0.49389609694480896,
      "learning_rate": 8e-05,
      "loss": 1.6446,
      "step": 1006
    },
    {
      "epoch": 0.2091381100726895,
      "grad_norm": 0.5181095600128174,
      "learning_rate": 8e-05,
      "loss": 1.6578,
      "step": 1007
    },
    {
      "epoch": 0.20934579439252338,
      "grad_norm": 0.4928913414478302,
      "learning_rate": 8e-05,
      "loss": 1.6385,
      "step": 1008
    },
    {
      "epoch": 0.20955347871235722,
      "grad_norm": 0.5067183971405029,
      "learning_rate": 8e-05,
      "loss": 1.5462,
      "step": 1009
    },
    {
      "epoch": 0.20976116303219106,
      "grad_norm": 0.46599841117858887,
      "learning_rate": 8e-05,
      "loss": 1.6113,
      "step": 1010
    },
    {
      "epoch": 0.20996884735202492,
      "grad_norm": 0.5027761459350586,
      "learning_rate": 8e-05,
      "loss": 1.6009,
      "step": 1011
    },
    {
      "epoch": 0.21017653167185876,
      "grad_norm": 0.5024246573448181,
      "learning_rate": 8e-05,
      "loss": 1.5819,
      "step": 1012
    },
    {
      "epoch": 0.21038421599169263,
      "grad_norm": 0.4743015468120575,
      "learning_rate": 8e-05,
      "loss": 1.6115,
      "step": 1013
    },
    {
      "epoch": 0.21059190031152647,
      "grad_norm": 0.5321120619773865,
      "learning_rate": 8e-05,
      "loss": 1.6409,
      "step": 1014
    },
    {
      "epoch": 0.21079958463136034,
      "grad_norm": 0.4797392189502716,
      "learning_rate": 8e-05,
      "loss": 1.5711,
      "step": 1015
    },
    {
      "epoch": 0.21100726895119418,
      "grad_norm": 0.5282681584358215,
      "learning_rate": 8e-05,
      "loss": 1.6369,
      "step": 1016
    },
    {
      "epoch": 0.21121495327102804,
      "grad_norm": 0.4802643954753876,
      "learning_rate": 8e-05,
      "loss": 1.6363,
      "step": 1017
    },
    {
      "epoch": 0.21142263759086188,
      "grad_norm": 0.479918897151947,
      "learning_rate": 8e-05,
      "loss": 1.6828,
      "step": 1018
    },
    {
      "epoch": 0.21163032191069575,
      "grad_norm": 0.485012412071228,
      "learning_rate": 8e-05,
      "loss": 1.6451,
      "step": 1019
    },
    {
      "epoch": 0.2118380062305296,
      "grad_norm": 0.47825416922569275,
      "learning_rate": 8e-05,
      "loss": 1.6492,
      "step": 1020
    },
    {
      "epoch": 0.21204569055036346,
      "grad_norm": 0.48356735706329346,
      "learning_rate": 8e-05,
      "loss": 1.5729,
      "step": 1021
    },
    {
      "epoch": 0.2122533748701973,
      "grad_norm": 0.4711926579475403,
      "learning_rate": 8e-05,
      "loss": 1.5919,
      "step": 1022
    },
    {
      "epoch": 0.21246105919003117,
      "grad_norm": 0.4887542128562927,
      "learning_rate": 8e-05,
      "loss": 1.6684,
      "step": 1023
    },
    {
      "epoch": 0.212668743509865,
      "grad_norm": 0.47914424538612366,
      "learning_rate": 8e-05,
      "loss": 1.5889,
      "step": 1024
    },
    {
      "epoch": 0.21287642782969884,
      "grad_norm": 0.49408769607543945,
      "learning_rate": 8e-05,
      "loss": 1.663,
      "step": 1025
    },
    {
      "epoch": 0.2130841121495327,
      "grad_norm": 0.49186617136001587,
      "learning_rate": 8e-05,
      "loss": 1.657,
      "step": 1026
    },
    {
      "epoch": 0.21329179646936655,
      "grad_norm": 0.5070930123329163,
      "learning_rate": 8e-05,
      "loss": 1.6457,
      "step": 1027
    },
    {
      "epoch": 0.21349948078920042,
      "grad_norm": 0.4801136553287506,
      "learning_rate": 8e-05,
      "loss": 1.6029,
      "step": 1028
    },
    {
      "epoch": 0.21370716510903426,
      "grad_norm": 0.4819825291633606,
      "learning_rate": 8e-05,
      "loss": 1.6109,
      "step": 1029
    },
    {
      "epoch": 0.21391484942886813,
      "grad_norm": 0.5037396550178528,
      "learning_rate": 8e-05,
      "loss": 1.6871,
      "step": 1030
    },
    {
      "epoch": 0.21412253374870197,
      "grad_norm": 0.5130371451377869,
      "learning_rate": 8e-05,
      "loss": 1.7099,
      "step": 1031
    },
    {
      "epoch": 0.21433021806853583,
      "grad_norm": 0.5002063512802124,
      "learning_rate": 8e-05,
      "loss": 1.6737,
      "step": 1032
    },
    {
      "epoch": 0.21453790238836967,
      "grad_norm": 0.4774034917354584,
      "learning_rate": 8e-05,
      "loss": 1.6461,
      "step": 1033
    },
    {
      "epoch": 0.21474558670820354,
      "grad_norm": 0.48661088943481445,
      "learning_rate": 8e-05,
      "loss": 1.5844,
      "step": 1034
    },
    {
      "epoch": 0.21495327102803738,
      "grad_norm": 0.4912685453891754,
      "learning_rate": 8e-05,
      "loss": 1.6028,
      "step": 1035
    },
    {
      "epoch": 0.21516095534787125,
      "grad_norm": 0.5158385634422302,
      "learning_rate": 8e-05,
      "loss": 1.6254,
      "step": 1036
    },
    {
      "epoch": 0.2153686396677051,
      "grad_norm": 0.4866856038570404,
      "learning_rate": 8e-05,
      "loss": 1.654,
      "step": 1037
    },
    {
      "epoch": 0.21557632398753895,
      "grad_norm": 0.5178074836730957,
      "learning_rate": 8e-05,
      "loss": 1.6212,
      "step": 1038
    },
    {
      "epoch": 0.2157840083073728,
      "grad_norm": 0.48483380675315857,
      "learning_rate": 8e-05,
      "loss": 1.6116,
      "step": 1039
    },
    {
      "epoch": 0.21599169262720663,
      "grad_norm": 0.47141504287719727,
      "learning_rate": 8e-05,
      "loss": 1.6042,
      "step": 1040
    },
    {
      "epoch": 0.2161993769470405,
      "grad_norm": 0.5006888508796692,
      "learning_rate": 8e-05,
      "loss": 1.6242,
      "step": 1041
    },
    {
      "epoch": 0.21640706126687434,
      "grad_norm": 0.47855380177497864,
      "learning_rate": 8e-05,
      "loss": 1.5661,
      "step": 1042
    },
    {
      "epoch": 0.2166147455867082,
      "grad_norm": 0.5185720324516296,
      "learning_rate": 8e-05,
      "loss": 1.6747,
      "step": 1043
    },
    {
      "epoch": 0.21682242990654205,
      "grad_norm": 0.4679183065891266,
      "learning_rate": 8e-05,
      "loss": 1.534,
      "step": 1044
    },
    {
      "epoch": 0.21703011422637591,
      "grad_norm": 0.5030350685119629,
      "learning_rate": 8e-05,
      "loss": 1.624,
      "step": 1045
    },
    {
      "epoch": 0.21723779854620975,
      "grad_norm": 0.48393726348876953,
      "learning_rate": 8e-05,
      "loss": 1.6531,
      "step": 1046
    },
    {
      "epoch": 0.21744548286604362,
      "grad_norm": 0.4825136065483093,
      "learning_rate": 8e-05,
      "loss": 1.6242,
      "step": 1047
    },
    {
      "epoch": 0.21765316718587746,
      "grad_norm": 0.4966621696949005,
      "learning_rate": 8e-05,
      "loss": 1.6368,
      "step": 1048
    },
    {
      "epoch": 0.21786085150571133,
      "grad_norm": 0.48175907135009766,
      "learning_rate": 8e-05,
      "loss": 1.6121,
      "step": 1049
    },
    {
      "epoch": 0.21806853582554517,
      "grad_norm": 0.4931303858757019,
      "learning_rate": 8e-05,
      "loss": 1.6452,
      "step": 1050
    },
    {
      "epoch": 0.21827622014537904,
      "grad_norm": 0.4742048680782318,
      "learning_rate": 8e-05,
      "loss": 1.6037,
      "step": 1051
    },
    {
      "epoch": 0.21848390446521287,
      "grad_norm": 0.4972934126853943,
      "learning_rate": 8e-05,
      "loss": 1.6522,
      "step": 1052
    },
    {
      "epoch": 0.21869158878504674,
      "grad_norm": 0.49021288752555847,
      "learning_rate": 8e-05,
      "loss": 1.6386,
      "step": 1053
    },
    {
      "epoch": 0.21889927310488058,
      "grad_norm": 0.4739423096179962,
      "learning_rate": 8e-05,
      "loss": 1.6267,
      "step": 1054
    },
    {
      "epoch": 0.21910695742471442,
      "grad_norm": 0.49233463406562805,
      "learning_rate": 8e-05,
      "loss": 1.6757,
      "step": 1055
    },
    {
      "epoch": 0.2193146417445483,
      "grad_norm": 0.4631211757659912,
      "learning_rate": 8e-05,
      "loss": 1.5198,
      "step": 1056
    },
    {
      "epoch": 0.21952232606438213,
      "grad_norm": 0.5057953596115112,
      "learning_rate": 8e-05,
      "loss": 1.6831,
      "step": 1057
    },
    {
      "epoch": 0.219730010384216,
      "grad_norm": 0.46444934606552124,
      "learning_rate": 8e-05,
      "loss": 1.6107,
      "step": 1058
    },
    {
      "epoch": 0.21993769470404984,
      "grad_norm": 0.48006969690322876,
      "learning_rate": 8e-05,
      "loss": 1.6356,
      "step": 1059
    },
    {
      "epoch": 0.2201453790238837,
      "grad_norm": 0.47697484493255615,
      "learning_rate": 8e-05,
      "loss": 1.605,
      "step": 1060
    },
    {
      "epoch": 0.22035306334371754,
      "grad_norm": 0.45909324288368225,
      "learning_rate": 8e-05,
      "loss": 1.6047,
      "step": 1061
    },
    {
      "epoch": 0.2205607476635514,
      "grad_norm": 0.45758333802223206,
      "learning_rate": 8e-05,
      "loss": 1.5597,
      "step": 1062
    },
    {
      "epoch": 0.22076843198338525,
      "grad_norm": 0.49077147245407104,
      "learning_rate": 8e-05,
      "loss": 1.6428,
      "step": 1063
    },
    {
      "epoch": 0.22097611630321912,
      "grad_norm": 0.49720385670661926,
      "learning_rate": 8e-05,
      "loss": 1.6476,
      "step": 1064
    },
    {
      "epoch": 0.22118380062305296,
      "grad_norm": 0.5079086422920227,
      "learning_rate": 8e-05,
      "loss": 1.6564,
      "step": 1065
    },
    {
      "epoch": 0.22139148494288682,
      "grad_norm": 0.47034189105033875,
      "learning_rate": 8e-05,
      "loss": 1.5323,
      "step": 1066
    },
    {
      "epoch": 0.22159916926272066,
      "grad_norm": 0.5000166893005371,
      "learning_rate": 8e-05,
      "loss": 1.6268,
      "step": 1067
    },
    {
      "epoch": 0.22180685358255453,
      "grad_norm": 0.47734546661376953,
      "learning_rate": 8e-05,
      "loss": 1.6071,
      "step": 1068
    },
    {
      "epoch": 0.22201453790238837,
      "grad_norm": 0.49549058079719543,
      "learning_rate": 8e-05,
      "loss": 1.6152,
      "step": 1069
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 0.4837387502193451,
      "learning_rate": 8e-05,
      "loss": 1.6049,
      "step": 1070
    },
    {
      "epoch": 0.22242990654205608,
      "grad_norm": 0.46780574321746826,
      "learning_rate": 8e-05,
      "loss": 1.5809,
      "step": 1071
    },
    {
      "epoch": 0.22263759086188992,
      "grad_norm": 0.4783675968647003,
      "learning_rate": 8e-05,
      "loss": 1.5655,
      "step": 1072
    },
    {
      "epoch": 0.22284527518172378,
      "grad_norm": 0.4945738613605499,
      "learning_rate": 8e-05,
      "loss": 1.6605,
      "step": 1073
    },
    {
      "epoch": 0.22305295950155762,
      "grad_norm": 0.5201888084411621,
      "learning_rate": 8e-05,
      "loss": 1.6003,
      "step": 1074
    },
    {
      "epoch": 0.2232606438213915,
      "grad_norm": 0.5067545175552368,
      "learning_rate": 8e-05,
      "loss": 1.6575,
      "step": 1075
    },
    {
      "epoch": 0.22346832814122533,
      "grad_norm": 0.4898606538772583,
      "learning_rate": 8e-05,
      "loss": 1.6554,
      "step": 1076
    },
    {
      "epoch": 0.2236760124610592,
      "grad_norm": 0.5530943274497986,
      "learning_rate": 8e-05,
      "loss": 1.6715,
      "step": 1077
    },
    {
      "epoch": 0.22388369678089304,
      "grad_norm": 0.4874219596385956,
      "learning_rate": 8e-05,
      "loss": 1.6182,
      "step": 1078
    },
    {
      "epoch": 0.2240913811007269,
      "grad_norm": 0.5321915745735168,
      "learning_rate": 8e-05,
      "loss": 1.6416,
      "step": 1079
    },
    {
      "epoch": 0.22429906542056074,
      "grad_norm": 0.46676793694496155,
      "learning_rate": 8e-05,
      "loss": 1.6459,
      "step": 1080
    },
    {
      "epoch": 0.2245067497403946,
      "grad_norm": 0.4740728735923767,
      "learning_rate": 8e-05,
      "loss": 1.5671,
      "step": 1081
    },
    {
      "epoch": 0.22471443406022845,
      "grad_norm": 0.48530635237693787,
      "learning_rate": 8e-05,
      "loss": 1.5294,
      "step": 1082
    },
    {
      "epoch": 0.22492211838006232,
      "grad_norm": 0.4611906111240387,
      "learning_rate": 8e-05,
      "loss": 1.6403,
      "step": 1083
    },
    {
      "epoch": 0.22512980269989616,
      "grad_norm": 0.4903087913990021,
      "learning_rate": 8e-05,
      "loss": 1.6264,
      "step": 1084
    },
    {
      "epoch": 0.22533748701973,
      "grad_norm": 0.486613392829895,
      "learning_rate": 8e-05,
      "loss": 1.6315,
      "step": 1085
    },
    {
      "epoch": 0.22554517133956387,
      "grad_norm": 0.5356956124305725,
      "learning_rate": 8e-05,
      "loss": 1.6754,
      "step": 1086
    },
    {
      "epoch": 0.2257528556593977,
      "grad_norm": 0.5136194825172424,
      "learning_rate": 8e-05,
      "loss": 1.6052,
      "step": 1087
    },
    {
      "epoch": 0.22596053997923157,
      "grad_norm": 0.49597111344337463,
      "learning_rate": 8e-05,
      "loss": 1.6403,
      "step": 1088
    },
    {
      "epoch": 0.2261682242990654,
      "grad_norm": 0.48345381021499634,
      "learning_rate": 8e-05,
      "loss": 1.6072,
      "step": 1089
    },
    {
      "epoch": 0.22637590861889928,
      "grad_norm": 0.505710780620575,
      "learning_rate": 8e-05,
      "loss": 1.5855,
      "step": 1090
    },
    {
      "epoch": 0.22658359293873312,
      "grad_norm": 0.5041759014129639,
      "learning_rate": 8e-05,
      "loss": 1.6558,
      "step": 1091
    },
    {
      "epoch": 0.226791277258567,
      "grad_norm": 0.5124122500419617,
      "learning_rate": 8e-05,
      "loss": 1.6171,
      "step": 1092
    },
    {
      "epoch": 0.22699896157840083,
      "grad_norm": 0.5028002262115479,
      "learning_rate": 8e-05,
      "loss": 1.667,
      "step": 1093
    },
    {
      "epoch": 0.2272066458982347,
      "grad_norm": 0.5096011161804199,
      "learning_rate": 8e-05,
      "loss": 1.6043,
      "step": 1094
    },
    {
      "epoch": 0.22741433021806853,
      "grad_norm": 0.4749337136745453,
      "learning_rate": 8e-05,
      "loss": 1.6256,
      "step": 1095
    },
    {
      "epoch": 0.2276220145379024,
      "grad_norm": 0.4887832701206207,
      "learning_rate": 8e-05,
      "loss": 1.6197,
      "step": 1096
    },
    {
      "epoch": 0.22782969885773624,
      "grad_norm": 0.4867877662181854,
      "learning_rate": 8e-05,
      "loss": 1.6318,
      "step": 1097
    },
    {
      "epoch": 0.22803738317757008,
      "grad_norm": 0.45605990290641785,
      "learning_rate": 8e-05,
      "loss": 1.5584,
      "step": 1098
    },
    {
      "epoch": 0.22824506749740395,
      "grad_norm": 0.48189279437065125,
      "learning_rate": 8e-05,
      "loss": 1.618,
      "step": 1099
    },
    {
      "epoch": 0.2284527518172378,
      "grad_norm": 0.510169506072998,
      "learning_rate": 8e-05,
      "loss": 1.6289,
      "step": 1100
    },
    {
      "epoch": 0.22866043613707165,
      "grad_norm": 0.47950589656829834,
      "learning_rate": 8e-05,
      "loss": 1.6601,
      "step": 1101
    },
    {
      "epoch": 0.2288681204569055,
      "grad_norm": 0.5073772072792053,
      "learning_rate": 8e-05,
      "loss": 1.6404,
      "step": 1102
    },
    {
      "epoch": 0.22907580477673936,
      "grad_norm": 0.47775015234947205,
      "learning_rate": 8e-05,
      "loss": 1.6235,
      "step": 1103
    },
    {
      "epoch": 0.2292834890965732,
      "grad_norm": 0.4975552558898926,
      "learning_rate": 8e-05,
      "loss": 1.6022,
      "step": 1104
    },
    {
      "epoch": 0.22949117341640707,
      "grad_norm": 0.49373266100883484,
      "learning_rate": 8e-05,
      "loss": 1.631,
      "step": 1105
    },
    {
      "epoch": 0.2296988577362409,
      "grad_norm": 0.5255943536758423,
      "learning_rate": 8e-05,
      "loss": 1.6067,
      "step": 1106
    },
    {
      "epoch": 0.22990654205607478,
      "grad_norm": 0.4758288264274597,
      "learning_rate": 8e-05,
      "loss": 1.5626,
      "step": 1107
    },
    {
      "epoch": 0.23011422637590861,
      "grad_norm": 0.4850333631038666,
      "learning_rate": 8e-05,
      "loss": 1.5805,
      "step": 1108
    },
    {
      "epoch": 0.23032191069574248,
      "grad_norm": 0.4803512990474701,
      "learning_rate": 8e-05,
      "loss": 1.5439,
      "step": 1109
    },
    {
      "epoch": 0.23052959501557632,
      "grad_norm": 0.4789465665817261,
      "learning_rate": 8e-05,
      "loss": 1.6022,
      "step": 1110
    },
    {
      "epoch": 0.2307372793354102,
      "grad_norm": 0.5042217969894409,
      "learning_rate": 8e-05,
      "loss": 1.6692,
      "step": 1111
    },
    {
      "epoch": 0.23094496365524403,
      "grad_norm": 0.497745543718338,
      "learning_rate": 8e-05,
      "loss": 1.6743,
      "step": 1112
    },
    {
      "epoch": 0.23115264797507787,
      "grad_norm": 0.45986318588256836,
      "learning_rate": 8e-05,
      "loss": 1.5806,
      "step": 1113
    },
    {
      "epoch": 0.23136033229491174,
      "grad_norm": 0.4838648736476898,
      "learning_rate": 8e-05,
      "loss": 1.653,
      "step": 1114
    },
    {
      "epoch": 0.23156801661474558,
      "grad_norm": 0.4738873243331909,
      "learning_rate": 8e-05,
      "loss": 1.5853,
      "step": 1115
    },
    {
      "epoch": 0.23177570093457944,
      "grad_norm": 0.48498794436454773,
      "learning_rate": 8e-05,
      "loss": 1.673,
      "step": 1116
    },
    {
      "epoch": 0.23198338525441328,
      "grad_norm": 0.4791082739830017,
      "learning_rate": 8e-05,
      "loss": 1.5446,
      "step": 1117
    },
    {
      "epoch": 0.23219106957424715,
      "grad_norm": 0.4974493086338043,
      "learning_rate": 8e-05,
      "loss": 1.5747,
      "step": 1118
    },
    {
      "epoch": 0.232398753894081,
      "grad_norm": 0.5050761103630066,
      "learning_rate": 8e-05,
      "loss": 1.5965,
      "step": 1119
    },
    {
      "epoch": 0.23260643821391486,
      "grad_norm": 0.47682154178619385,
      "learning_rate": 8e-05,
      "loss": 1.581,
      "step": 1120
    },
    {
      "epoch": 0.2328141225337487,
      "grad_norm": 0.5180225372314453,
      "learning_rate": 8e-05,
      "loss": 1.6094,
      "step": 1121
    },
    {
      "epoch": 0.23302180685358256,
      "grad_norm": 0.5043385028839111,
      "learning_rate": 8e-05,
      "loss": 1.6338,
      "step": 1122
    },
    {
      "epoch": 0.2332294911734164,
      "grad_norm": 0.5274357199668884,
      "learning_rate": 8e-05,
      "loss": 1.6203,
      "step": 1123
    },
    {
      "epoch": 0.23343717549325027,
      "grad_norm": 0.49157971143722534,
      "learning_rate": 8e-05,
      "loss": 1.5854,
      "step": 1124
    },
    {
      "epoch": 0.2336448598130841,
      "grad_norm": 0.5090160369873047,
      "learning_rate": 8e-05,
      "loss": 1.6155,
      "step": 1125
    },
    {
      "epoch": 0.23385254413291798,
      "grad_norm": 0.4960342049598694,
      "learning_rate": 8e-05,
      "loss": 1.6501,
      "step": 1126
    },
    {
      "epoch": 0.23406022845275182,
      "grad_norm": 0.49579086899757385,
      "learning_rate": 8e-05,
      "loss": 1.6299,
      "step": 1127
    },
    {
      "epoch": 0.23426791277258566,
      "grad_norm": 0.5103940963745117,
      "learning_rate": 8e-05,
      "loss": 1.5807,
      "step": 1128
    },
    {
      "epoch": 0.23447559709241952,
      "grad_norm": 0.47810351848602295,
      "learning_rate": 8e-05,
      "loss": 1.6593,
      "step": 1129
    },
    {
      "epoch": 0.23468328141225336,
      "grad_norm": 0.48690786957740784,
      "learning_rate": 8e-05,
      "loss": 1.5911,
      "step": 1130
    },
    {
      "epoch": 0.23489096573208723,
      "grad_norm": 0.5075724720954895,
      "learning_rate": 8e-05,
      "loss": 1.6673,
      "step": 1131
    },
    {
      "epoch": 0.23509865005192107,
      "grad_norm": 0.5023790001869202,
      "learning_rate": 8e-05,
      "loss": 1.6244,
      "step": 1132
    },
    {
      "epoch": 0.23530633437175494,
      "grad_norm": 0.49479010701179504,
      "learning_rate": 8e-05,
      "loss": 1.5765,
      "step": 1133
    },
    {
      "epoch": 0.23551401869158878,
      "grad_norm": 0.47907403111457825,
      "learning_rate": 8e-05,
      "loss": 1.6108,
      "step": 1134
    },
    {
      "epoch": 0.23572170301142265,
      "grad_norm": 0.5096808075904846,
      "learning_rate": 8e-05,
      "loss": 1.5931,
      "step": 1135
    },
    {
      "epoch": 0.23592938733125648,
      "grad_norm": 0.4935376048088074,
      "learning_rate": 8e-05,
      "loss": 1.6679,
      "step": 1136
    },
    {
      "epoch": 0.23613707165109035,
      "grad_norm": 0.4955846667289734,
      "learning_rate": 8e-05,
      "loss": 1.6786,
      "step": 1137
    },
    {
      "epoch": 0.2363447559709242,
      "grad_norm": 0.5047092437744141,
      "learning_rate": 8e-05,
      "loss": 1.5719,
      "step": 1138
    },
    {
      "epoch": 0.23655244029075806,
      "grad_norm": 0.4767501950263977,
      "learning_rate": 8e-05,
      "loss": 1.5391,
      "step": 1139
    },
    {
      "epoch": 0.2367601246105919,
      "grad_norm": 0.49148985743522644,
      "learning_rate": 8e-05,
      "loss": 1.6281,
      "step": 1140
    },
    {
      "epoch": 0.23696780893042577,
      "grad_norm": 0.4789850413799286,
      "learning_rate": 8e-05,
      "loss": 1.541,
      "step": 1141
    },
    {
      "epoch": 0.2371754932502596,
      "grad_norm": 0.47179505228996277,
      "learning_rate": 8e-05,
      "loss": 1.56,
      "step": 1142
    },
    {
      "epoch": 0.23738317757009345,
      "grad_norm": 0.4852232038974762,
      "learning_rate": 8e-05,
      "loss": 1.6166,
      "step": 1143
    },
    {
      "epoch": 0.2375908618899273,
      "grad_norm": 0.4789161682128906,
      "learning_rate": 8e-05,
      "loss": 1.6051,
      "step": 1144
    },
    {
      "epoch": 0.23779854620976115,
      "grad_norm": 0.5053293108940125,
      "learning_rate": 8e-05,
      "loss": 1.6531,
      "step": 1145
    },
    {
      "epoch": 0.23800623052959502,
      "grad_norm": 0.48226839303970337,
      "learning_rate": 8e-05,
      "loss": 1.6138,
      "step": 1146
    },
    {
      "epoch": 0.23821391484942886,
      "grad_norm": 0.4766741693019867,
      "learning_rate": 8e-05,
      "loss": 1.5757,
      "step": 1147
    },
    {
      "epoch": 0.23842159916926273,
      "grad_norm": 0.4817132353782654,
      "learning_rate": 8e-05,
      "loss": 1.5919,
      "step": 1148
    },
    {
      "epoch": 0.23862928348909657,
      "grad_norm": 0.48039570450782776,
      "learning_rate": 8e-05,
      "loss": 1.5237,
      "step": 1149
    },
    {
      "epoch": 0.23883696780893043,
      "grad_norm": 0.539954662322998,
      "learning_rate": 8e-05,
      "loss": 1.7231,
      "step": 1150
    },
    {
      "epoch": 0.23904465212876427,
      "grad_norm": 0.49873772263526917,
      "learning_rate": 8e-05,
      "loss": 1.6099,
      "step": 1151
    },
    {
      "epoch": 0.23925233644859814,
      "grad_norm": 0.5055965185165405,
      "learning_rate": 8e-05,
      "loss": 1.6688,
      "step": 1152
    },
    {
      "epoch": 0.23946002076843198,
      "grad_norm": 0.4793665409088135,
      "learning_rate": 8e-05,
      "loss": 1.6186,
      "step": 1153
    },
    {
      "epoch": 0.23966770508826585,
      "grad_norm": 0.49260565638542175,
      "learning_rate": 8e-05,
      "loss": 1.5712,
      "step": 1154
    },
    {
      "epoch": 0.2398753894080997,
      "grad_norm": 0.48806118965148926,
      "learning_rate": 8e-05,
      "loss": 1.6346,
      "step": 1155
    },
    {
      "epoch": 0.24008307372793355,
      "grad_norm": 0.4877453148365021,
      "learning_rate": 8e-05,
      "loss": 1.5825,
      "step": 1156
    },
    {
      "epoch": 0.2402907580477674,
      "grad_norm": 0.4928903877735138,
      "learning_rate": 8e-05,
      "loss": 1.6033,
      "step": 1157
    },
    {
      "epoch": 0.24049844236760123,
      "grad_norm": 0.4725443124771118,
      "learning_rate": 8e-05,
      "loss": 1.6135,
      "step": 1158
    },
    {
      "epoch": 0.2407061266874351,
      "grad_norm": 0.47031083703041077,
      "learning_rate": 8e-05,
      "loss": 1.6,
      "step": 1159
    },
    {
      "epoch": 0.24091381100726894,
      "grad_norm": 0.46935370564460754,
      "learning_rate": 8e-05,
      "loss": 1.574,
      "step": 1160
    },
    {
      "epoch": 0.2411214953271028,
      "grad_norm": 0.4930771291255951,
      "learning_rate": 8e-05,
      "loss": 1.5769,
      "step": 1161
    },
    {
      "epoch": 0.24132917964693665,
      "grad_norm": 0.4878680408000946,
      "learning_rate": 8e-05,
      "loss": 1.6057,
      "step": 1162
    },
    {
      "epoch": 0.24153686396677052,
      "grad_norm": 0.46639183163642883,
      "learning_rate": 8e-05,
      "loss": 1.5862,
      "step": 1163
    },
    {
      "epoch": 0.24174454828660435,
      "grad_norm": 0.48250576853752136,
      "learning_rate": 8e-05,
      "loss": 1.5948,
      "step": 1164
    },
    {
      "epoch": 0.24195223260643822,
      "grad_norm": 0.4780838191509247,
      "learning_rate": 8e-05,
      "loss": 1.5429,
      "step": 1165
    },
    {
      "epoch": 0.24215991692627206,
      "grad_norm": 0.4783189296722412,
      "learning_rate": 8e-05,
      "loss": 1.6205,
      "step": 1166
    },
    {
      "epoch": 0.24236760124610593,
      "grad_norm": 0.4721772372722626,
      "learning_rate": 8e-05,
      "loss": 1.5667,
      "step": 1167
    },
    {
      "epoch": 0.24257528556593977,
      "grad_norm": 0.49488070607185364,
      "learning_rate": 8e-05,
      "loss": 1.6791,
      "step": 1168
    },
    {
      "epoch": 0.24278296988577364,
      "grad_norm": 0.49183887243270874,
      "learning_rate": 8e-05,
      "loss": 1.6051,
      "step": 1169
    },
    {
      "epoch": 0.24299065420560748,
      "grad_norm": 0.4702599346637726,
      "learning_rate": 8e-05,
      "loss": 1.5553,
      "step": 1170
    },
    {
      "epoch": 0.24319833852544132,
      "grad_norm": 0.48021695017814636,
      "learning_rate": 8e-05,
      "loss": 1.5804,
      "step": 1171
    },
    {
      "epoch": 0.24340602284527518,
      "grad_norm": 0.5404965281486511,
      "learning_rate": 8e-05,
      "loss": 1.6724,
      "step": 1172
    },
    {
      "epoch": 0.24361370716510902,
      "grad_norm": 0.5167496204376221,
      "learning_rate": 8e-05,
      "loss": 1.5631,
      "step": 1173
    },
    {
      "epoch": 0.2438213914849429,
      "grad_norm": 0.4974764883518219,
      "learning_rate": 8e-05,
      "loss": 1.5615,
      "step": 1174
    },
    {
      "epoch": 0.24402907580477673,
      "grad_norm": 0.4905969202518463,
      "learning_rate": 8e-05,
      "loss": 1.6023,
      "step": 1175
    },
    {
      "epoch": 0.2442367601246106,
      "grad_norm": 0.5358709692955017,
      "learning_rate": 8e-05,
      "loss": 1.5555,
      "step": 1176
    },
    {
      "epoch": 0.24444444444444444,
      "grad_norm": 0.49605727195739746,
      "learning_rate": 8e-05,
      "loss": 1.6356,
      "step": 1177
    },
    {
      "epoch": 0.2446521287642783,
      "grad_norm": 0.4891799986362457,
      "learning_rate": 8e-05,
      "loss": 1.6877,
      "step": 1178
    },
    {
      "epoch": 0.24485981308411214,
      "grad_norm": 0.4895115792751312,
      "learning_rate": 8e-05,
      "loss": 1.5594,
      "step": 1179
    },
    {
      "epoch": 0.245067497403946,
      "grad_norm": 0.49354806542396545,
      "learning_rate": 8e-05,
      "loss": 1.6815,
      "step": 1180
    },
    {
      "epoch": 0.24527518172377985,
      "grad_norm": 0.4897477924823761,
      "learning_rate": 8e-05,
      "loss": 1.6378,
      "step": 1181
    },
    {
      "epoch": 0.24548286604361372,
      "grad_norm": 0.4873749613761902,
      "learning_rate": 8e-05,
      "loss": 1.5757,
      "step": 1182
    },
    {
      "epoch": 0.24569055036344756,
      "grad_norm": 0.464875191450119,
      "learning_rate": 8e-05,
      "loss": 1.5587,
      "step": 1183
    },
    {
      "epoch": 0.24589823468328142,
      "grad_norm": 0.4927029311656952,
      "learning_rate": 8e-05,
      "loss": 1.6541,
      "step": 1184
    },
    {
      "epoch": 0.24610591900311526,
      "grad_norm": 0.4638020694255829,
      "learning_rate": 8e-05,
      "loss": 1.5619,
      "step": 1185
    },
    {
      "epoch": 0.2463136033229491,
      "grad_norm": 0.4894510805606842,
      "learning_rate": 8e-05,
      "loss": 1.5729,
      "step": 1186
    },
    {
      "epoch": 0.24652128764278297,
      "grad_norm": 0.4695272743701935,
      "learning_rate": 8e-05,
      "loss": 1.6059,
      "step": 1187
    },
    {
      "epoch": 0.2467289719626168,
      "grad_norm": 0.47015345096588135,
      "learning_rate": 8e-05,
      "loss": 1.6093,
      "step": 1188
    },
    {
      "epoch": 0.24693665628245068,
      "grad_norm": 0.4933643043041229,
      "learning_rate": 8e-05,
      "loss": 1.5702,
      "step": 1189
    },
    {
      "epoch": 0.24714434060228452,
      "grad_norm": 0.4731082022190094,
      "learning_rate": 8e-05,
      "loss": 1.5177,
      "step": 1190
    },
    {
      "epoch": 0.24735202492211839,
      "grad_norm": 0.48732051253318787,
      "learning_rate": 8e-05,
      "loss": 1.6784,
      "step": 1191
    },
    {
      "epoch": 0.24755970924195222,
      "grad_norm": 0.5021098256111145,
      "learning_rate": 8e-05,
      "loss": 1.6564,
      "step": 1192
    },
    {
      "epoch": 0.2477673935617861,
      "grad_norm": 0.48108044266700745,
      "learning_rate": 8e-05,
      "loss": 1.6051,
      "step": 1193
    },
    {
      "epoch": 0.24797507788161993,
      "grad_norm": 0.47929883003234863,
      "learning_rate": 8e-05,
      "loss": 1.6104,
      "step": 1194
    },
    {
      "epoch": 0.2481827622014538,
      "grad_norm": 0.502677321434021,
      "learning_rate": 8e-05,
      "loss": 1.6059,
      "step": 1195
    },
    {
      "epoch": 0.24839044652128764,
      "grad_norm": 0.49433907866477966,
      "learning_rate": 8e-05,
      "loss": 1.6257,
      "step": 1196
    },
    {
      "epoch": 0.2485981308411215,
      "grad_norm": 0.4825282394886017,
      "learning_rate": 8e-05,
      "loss": 1.6134,
      "step": 1197
    },
    {
      "epoch": 0.24880581516095535,
      "grad_norm": 0.4730031192302704,
      "learning_rate": 8e-05,
      "loss": 1.581,
      "step": 1198
    },
    {
      "epoch": 0.2490134994807892,
      "grad_norm": 0.4965001940727234,
      "learning_rate": 8e-05,
      "loss": 1.6166,
      "step": 1199
    },
    {
      "epoch": 0.24922118380062305,
      "grad_norm": 0.49881839752197266,
      "learning_rate": 8e-05,
      "loss": 1.6278,
      "step": 1200
    },
    {
      "epoch": 0.2494288681204569,
      "grad_norm": 0.484718382358551,
      "learning_rate": 8e-05,
      "loss": 1.6499,
      "step": 1201
    },
    {
      "epoch": 0.24963655244029076,
      "grad_norm": 0.4792778491973877,
      "learning_rate": 8e-05,
      "loss": 1.6444,
      "step": 1202
    },
    {
      "epoch": 0.2498442367601246,
      "grad_norm": 0.4959498643875122,
      "learning_rate": 8e-05,
      "loss": 1.6745,
      "step": 1203
    },
    {
      "epoch": 0.25005192107995844,
      "grad_norm": 0.48365911841392517,
      "learning_rate": 8e-05,
      "loss": 1.5993,
      "step": 1204
    },
    {
      "epoch": 0.2502596053997923,
      "grad_norm": 0.5093377828598022,
      "learning_rate": 8e-05,
      "loss": 1.628,
      "step": 1205
    },
    {
      "epoch": 0.2504672897196262,
      "grad_norm": 0.486716091632843,
      "learning_rate": 8e-05,
      "loss": 1.6436,
      "step": 1206
    },
    {
      "epoch": 0.25067497403946004,
      "grad_norm": 0.4940541982650757,
      "learning_rate": 8e-05,
      "loss": 1.5606,
      "step": 1207
    },
    {
      "epoch": 0.25088265835929385,
      "grad_norm": 0.4985447824001312,
      "learning_rate": 8e-05,
      "loss": 1.7143,
      "step": 1208
    },
    {
      "epoch": 0.2510903426791277,
      "grad_norm": 0.49395254254341125,
      "learning_rate": 8e-05,
      "loss": 1.5876,
      "step": 1209
    },
    {
      "epoch": 0.2512980269989616,
      "grad_norm": 0.4942306876182556,
      "learning_rate": 8e-05,
      "loss": 1.6473,
      "step": 1210
    },
    {
      "epoch": 0.25150571131879546,
      "grad_norm": 0.48080599308013916,
      "learning_rate": 8e-05,
      "loss": 1.6188,
      "step": 1211
    },
    {
      "epoch": 0.25171339563862927,
      "grad_norm": 0.4913437068462372,
      "learning_rate": 8e-05,
      "loss": 1.6626,
      "step": 1212
    },
    {
      "epoch": 0.25192107995846313,
      "grad_norm": 0.4762611985206604,
      "learning_rate": 8e-05,
      "loss": 1.5715,
      "step": 1213
    },
    {
      "epoch": 0.252128764278297,
      "grad_norm": 0.49467694759368896,
      "learning_rate": 8e-05,
      "loss": 1.6428,
      "step": 1214
    },
    {
      "epoch": 0.2523364485981308,
      "grad_norm": 0.49399566650390625,
      "learning_rate": 8e-05,
      "loss": 1.5658,
      "step": 1215
    },
    {
      "epoch": 0.2525441329179647,
      "grad_norm": 0.47850465774536133,
      "learning_rate": 8e-05,
      "loss": 1.6057,
      "step": 1216
    },
    {
      "epoch": 0.25275181723779855,
      "grad_norm": 0.5025778412818909,
      "learning_rate": 8e-05,
      "loss": 1.6746,
      "step": 1217
    },
    {
      "epoch": 0.2529595015576324,
      "grad_norm": 0.4746226966381073,
      "learning_rate": 8e-05,
      "loss": 1.5973,
      "step": 1218
    },
    {
      "epoch": 0.2531671858774662,
      "grad_norm": 0.490772008895874,
      "learning_rate": 8e-05,
      "loss": 1.6176,
      "step": 1219
    },
    {
      "epoch": 0.2533748701973001,
      "grad_norm": 0.477453351020813,
      "learning_rate": 8e-05,
      "loss": 1.6027,
      "step": 1220
    },
    {
      "epoch": 0.25358255451713396,
      "grad_norm": 0.48569169640541077,
      "learning_rate": 8e-05,
      "loss": 1.6144,
      "step": 1221
    },
    {
      "epoch": 0.25379023883696783,
      "grad_norm": 0.48016172647476196,
      "learning_rate": 8e-05,
      "loss": 1.612,
      "step": 1222
    },
    {
      "epoch": 0.25399792315680164,
      "grad_norm": 0.458385169506073,
      "learning_rate": 8e-05,
      "loss": 1.4938,
      "step": 1223
    },
    {
      "epoch": 0.2542056074766355,
      "grad_norm": 0.4843856692314148,
      "learning_rate": 8e-05,
      "loss": 1.5878,
      "step": 1224
    },
    {
      "epoch": 0.2544132917964694,
      "grad_norm": 0.5237738490104675,
      "learning_rate": 8e-05,
      "loss": 1.6753,
      "step": 1225
    },
    {
      "epoch": 0.25462097611630324,
      "grad_norm": 0.5184034109115601,
      "learning_rate": 8e-05,
      "loss": 1.5755,
      "step": 1226
    },
    {
      "epoch": 0.25482866043613706,
      "grad_norm": 0.5014573931694031,
      "learning_rate": 8e-05,
      "loss": 1.6439,
      "step": 1227
    },
    {
      "epoch": 0.2550363447559709,
      "grad_norm": 0.4746113419532776,
      "learning_rate": 8e-05,
      "loss": 1.6612,
      "step": 1228
    },
    {
      "epoch": 0.2552440290758048,
      "grad_norm": 0.48897260427474976,
      "learning_rate": 8e-05,
      "loss": 1.5836,
      "step": 1229
    },
    {
      "epoch": 0.2554517133956386,
      "grad_norm": 0.47817346453666687,
      "learning_rate": 8e-05,
      "loss": 1.5826,
      "step": 1230
    },
    {
      "epoch": 0.25565939771547247,
      "grad_norm": 0.5092668533325195,
      "learning_rate": 8e-05,
      "loss": 1.6462,
      "step": 1231
    },
    {
      "epoch": 0.25586708203530634,
      "grad_norm": 0.4775659739971161,
      "learning_rate": 8e-05,
      "loss": 1.5857,
      "step": 1232
    },
    {
      "epoch": 0.2560747663551402,
      "grad_norm": 0.5099481344223022,
      "learning_rate": 8e-05,
      "loss": 1.592,
      "step": 1233
    },
    {
      "epoch": 0.256282450674974,
      "grad_norm": 0.4921092987060547,
      "learning_rate": 8e-05,
      "loss": 1.5091,
      "step": 1234
    },
    {
      "epoch": 0.2564901349948079,
      "grad_norm": 0.5136047601699829,
      "learning_rate": 8e-05,
      "loss": 1.6556,
      "step": 1235
    },
    {
      "epoch": 0.25669781931464175,
      "grad_norm": 0.4959682822227478,
      "learning_rate": 8e-05,
      "loss": 1.6347,
      "step": 1236
    },
    {
      "epoch": 0.2569055036344756,
      "grad_norm": 0.5049384236335754,
      "learning_rate": 8e-05,
      "loss": 1.6096,
      "step": 1237
    },
    {
      "epoch": 0.25711318795430943,
      "grad_norm": 0.5177755951881409,
      "learning_rate": 8e-05,
      "loss": 1.5408,
      "step": 1238
    },
    {
      "epoch": 0.2573208722741433,
      "grad_norm": 0.4715554416179657,
      "learning_rate": 8e-05,
      "loss": 1.6184,
      "step": 1239
    },
    {
      "epoch": 0.25752855659397716,
      "grad_norm": 0.5331630110740662,
      "learning_rate": 8e-05,
      "loss": 1.6457,
      "step": 1240
    },
    {
      "epoch": 0.25773624091381103,
      "grad_norm": 0.48646849393844604,
      "learning_rate": 8e-05,
      "loss": 1.6077,
      "step": 1241
    },
    {
      "epoch": 0.25794392523364484,
      "grad_norm": 0.48699328303337097,
      "learning_rate": 8e-05,
      "loss": 1.5695,
      "step": 1242
    },
    {
      "epoch": 0.2581516095534787,
      "grad_norm": 0.4788655936717987,
      "learning_rate": 8e-05,
      "loss": 1.6047,
      "step": 1243
    },
    {
      "epoch": 0.2583592938733126,
      "grad_norm": 0.49768000841140747,
      "learning_rate": 8e-05,
      "loss": 1.6321,
      "step": 1244
    },
    {
      "epoch": 0.2585669781931464,
      "grad_norm": 0.4728749096393585,
      "learning_rate": 8e-05,
      "loss": 1.4864,
      "step": 1245
    },
    {
      "epoch": 0.25877466251298026,
      "grad_norm": 0.4823184311389923,
      "learning_rate": 8e-05,
      "loss": 1.5463,
      "step": 1246
    },
    {
      "epoch": 0.2589823468328141,
      "grad_norm": 0.49181875586509705,
      "learning_rate": 8e-05,
      "loss": 1.6372,
      "step": 1247
    },
    {
      "epoch": 0.259190031152648,
      "grad_norm": 0.48405319452285767,
      "learning_rate": 8e-05,
      "loss": 1.598,
      "step": 1248
    },
    {
      "epoch": 0.2593977154724818,
      "grad_norm": 0.5048245787620544,
      "learning_rate": 8e-05,
      "loss": 1.5521,
      "step": 1249
    },
    {
      "epoch": 0.25960539979231567,
      "grad_norm": 0.4842333495616913,
      "learning_rate": 8e-05,
      "loss": 1.6642,
      "step": 1250
    },
    {
      "epoch": 0.25981308411214954,
      "grad_norm": 0.48424315452575684,
      "learning_rate": 8e-05,
      "loss": 1.6036,
      "step": 1251
    },
    {
      "epoch": 0.2600207684319834,
      "grad_norm": 0.46493804454803467,
      "learning_rate": 8e-05,
      "loss": 1.6679,
      "step": 1252
    },
    {
      "epoch": 0.2602284527518172,
      "grad_norm": 0.5107417106628418,
      "learning_rate": 8e-05,
      "loss": 1.6116,
      "step": 1253
    },
    {
      "epoch": 0.2604361370716511,
      "grad_norm": 0.48230957984924316,
      "learning_rate": 8e-05,
      "loss": 1.5848,
      "step": 1254
    },
    {
      "epoch": 0.26064382139148495,
      "grad_norm": 0.4716993570327759,
      "learning_rate": 8e-05,
      "loss": 1.5599,
      "step": 1255
    },
    {
      "epoch": 0.2608515057113188,
      "grad_norm": 0.4995378255844116,
      "learning_rate": 8e-05,
      "loss": 1.5162,
      "step": 1256
    },
    {
      "epoch": 0.26105919003115263,
      "grad_norm": 0.49914610385894775,
      "learning_rate": 8e-05,
      "loss": 1.5889,
      "step": 1257
    },
    {
      "epoch": 0.2612668743509865,
      "grad_norm": 0.5149269104003906,
      "learning_rate": 8e-05,
      "loss": 1.5693,
      "step": 1258
    },
    {
      "epoch": 0.26147455867082037,
      "grad_norm": 0.4664236903190613,
      "learning_rate": 8e-05,
      "loss": 1.5911,
      "step": 1259
    },
    {
      "epoch": 0.2616822429906542,
      "grad_norm": 0.4691005349159241,
      "learning_rate": 8e-05,
      "loss": 1.5378,
      "step": 1260
    },
    {
      "epoch": 0.26188992731048805,
      "grad_norm": 0.5125289559364319,
      "learning_rate": 8e-05,
      "loss": 1.6207,
      "step": 1261
    },
    {
      "epoch": 0.2620976116303219,
      "grad_norm": 0.5016699433326721,
      "learning_rate": 8e-05,
      "loss": 1.7096,
      "step": 1262
    },
    {
      "epoch": 0.2623052959501558,
      "grad_norm": 0.4931148290634155,
      "learning_rate": 8e-05,
      "loss": 1.535,
      "step": 1263
    },
    {
      "epoch": 0.2625129802699896,
      "grad_norm": 0.4818016588687897,
      "learning_rate": 8e-05,
      "loss": 1.6163,
      "step": 1264
    },
    {
      "epoch": 0.26272066458982346,
      "grad_norm": 0.4864917993545532,
      "learning_rate": 8e-05,
      "loss": 1.5272,
      "step": 1265
    },
    {
      "epoch": 0.26292834890965733,
      "grad_norm": 0.4974689483642578,
      "learning_rate": 8e-05,
      "loss": 1.6396,
      "step": 1266
    },
    {
      "epoch": 0.2631360332294912,
      "grad_norm": 0.48342207074165344,
      "learning_rate": 8e-05,
      "loss": 1.5803,
      "step": 1267
    },
    {
      "epoch": 0.263343717549325,
      "grad_norm": 0.5663976073265076,
      "learning_rate": 8e-05,
      "loss": 1.6336,
      "step": 1268
    },
    {
      "epoch": 0.2635514018691589,
      "grad_norm": 0.5170464515686035,
      "learning_rate": 8e-05,
      "loss": 1.6435,
      "step": 1269
    },
    {
      "epoch": 0.26375908618899274,
      "grad_norm": 0.49931710958480835,
      "learning_rate": 8e-05,
      "loss": 1.697,
      "step": 1270
    },
    {
      "epoch": 0.2639667705088266,
      "grad_norm": 0.4786977767944336,
      "learning_rate": 8e-05,
      "loss": 1.5923,
      "step": 1271
    },
    {
      "epoch": 0.2641744548286604,
      "grad_norm": 0.490163117647171,
      "learning_rate": 8e-05,
      "loss": 1.5944,
      "step": 1272
    },
    {
      "epoch": 0.2643821391484943,
      "grad_norm": 0.469289630651474,
      "learning_rate": 8e-05,
      "loss": 1.4951,
      "step": 1273
    },
    {
      "epoch": 0.26458982346832816,
      "grad_norm": 0.49278539419174194,
      "learning_rate": 8e-05,
      "loss": 1.5941,
      "step": 1274
    },
    {
      "epoch": 0.26479750778816197,
      "grad_norm": 0.4842006266117096,
      "learning_rate": 8e-05,
      "loss": 1.5923,
      "step": 1275
    },
    {
      "epoch": 0.26500519210799584,
      "grad_norm": 0.5030384659767151,
      "learning_rate": 8e-05,
      "loss": 1.6361,
      "step": 1276
    },
    {
      "epoch": 0.2652128764278297,
      "grad_norm": 0.4949890971183777,
      "learning_rate": 8e-05,
      "loss": 1.6389,
      "step": 1277
    },
    {
      "epoch": 0.26542056074766357,
      "grad_norm": 0.4904261827468872,
      "learning_rate": 8e-05,
      "loss": 1.5555,
      "step": 1278
    },
    {
      "epoch": 0.2656282450674974,
      "grad_norm": 0.48278945684432983,
      "learning_rate": 8e-05,
      "loss": 1.64,
      "step": 1279
    },
    {
      "epoch": 0.26583592938733125,
      "grad_norm": 0.48540300130844116,
      "learning_rate": 8e-05,
      "loss": 1.6054,
      "step": 1280
    },
    {
      "epoch": 0.2660436137071651,
      "grad_norm": 0.5015822649002075,
      "learning_rate": 8e-05,
      "loss": 1.5868,
      "step": 1281
    },
    {
      "epoch": 0.266251298026999,
      "grad_norm": 0.49217545986175537,
      "learning_rate": 8e-05,
      "loss": 1.6145,
      "step": 1282
    },
    {
      "epoch": 0.2664589823468328,
      "grad_norm": 0.5260308384895325,
      "learning_rate": 8e-05,
      "loss": 1.6333,
      "step": 1283
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 0.5329016447067261,
      "learning_rate": 8e-05,
      "loss": 1.7274,
      "step": 1284
    },
    {
      "epoch": 0.26687435098650053,
      "grad_norm": 0.5138272047042847,
      "learning_rate": 8e-05,
      "loss": 1.6601,
      "step": 1285
    },
    {
      "epoch": 0.2670820353063344,
      "grad_norm": 0.4883299171924591,
      "learning_rate": 8e-05,
      "loss": 1.5988,
      "step": 1286
    },
    {
      "epoch": 0.2672897196261682,
      "grad_norm": 0.5150203108787537,
      "learning_rate": 8e-05,
      "loss": 1.6732,
      "step": 1287
    },
    {
      "epoch": 0.2674974039460021,
      "grad_norm": 0.49123409390449524,
      "learning_rate": 8e-05,
      "loss": 1.5287,
      "step": 1288
    },
    {
      "epoch": 0.26770508826583594,
      "grad_norm": 0.4945281744003296,
      "learning_rate": 8e-05,
      "loss": 1.6321,
      "step": 1289
    },
    {
      "epoch": 0.26791277258566976,
      "grad_norm": 0.5029784440994263,
      "learning_rate": 8e-05,
      "loss": 1.6214,
      "step": 1290
    },
    {
      "epoch": 0.2681204569055036,
      "grad_norm": 0.46249303221702576,
      "learning_rate": 8e-05,
      "loss": 1.4891,
      "step": 1291
    },
    {
      "epoch": 0.2683281412253375,
      "grad_norm": 0.5195319652557373,
      "learning_rate": 8e-05,
      "loss": 1.6059,
      "step": 1292
    },
    {
      "epoch": 0.26853582554517136,
      "grad_norm": 0.5203794836997986,
      "learning_rate": 8e-05,
      "loss": 1.6673,
      "step": 1293
    },
    {
      "epoch": 0.26874350986500517,
      "grad_norm": 0.4765598773956299,
      "learning_rate": 8e-05,
      "loss": 1.6292,
      "step": 1294
    },
    {
      "epoch": 0.26895119418483904,
      "grad_norm": 0.4790004789829254,
      "learning_rate": 8e-05,
      "loss": 1.5275,
      "step": 1295
    },
    {
      "epoch": 0.2691588785046729,
      "grad_norm": 0.486456036567688,
      "learning_rate": 8e-05,
      "loss": 1.5916,
      "step": 1296
    },
    {
      "epoch": 0.2693665628245068,
      "grad_norm": 0.4780852496623993,
      "learning_rate": 8e-05,
      "loss": 1.6143,
      "step": 1297
    },
    {
      "epoch": 0.2695742471443406,
      "grad_norm": 0.5063662528991699,
      "learning_rate": 8e-05,
      "loss": 1.598,
      "step": 1298
    },
    {
      "epoch": 0.26978193146417445,
      "grad_norm": 0.4856896996498108,
      "learning_rate": 8e-05,
      "loss": 1.6217,
      "step": 1299
    },
    {
      "epoch": 0.2699896157840083,
      "grad_norm": 0.4855417013168335,
      "learning_rate": 8e-05,
      "loss": 1.5746,
      "step": 1300
    },
    {
      "epoch": 0.2701973001038422,
      "grad_norm": 0.4749685227870941,
      "learning_rate": 8e-05,
      "loss": 1.529,
      "step": 1301
    },
    {
      "epoch": 0.270404984423676,
      "grad_norm": 0.4671476185321808,
      "learning_rate": 8e-05,
      "loss": 1.5717,
      "step": 1302
    },
    {
      "epoch": 0.27061266874350987,
      "grad_norm": 0.49062880873680115,
      "learning_rate": 8e-05,
      "loss": 1.5799,
      "step": 1303
    },
    {
      "epoch": 0.27082035306334373,
      "grad_norm": 0.4858238995075226,
      "learning_rate": 8e-05,
      "loss": 1.6249,
      "step": 1304
    },
    {
      "epoch": 0.27102803738317754,
      "grad_norm": 0.4788179397583008,
      "learning_rate": 8e-05,
      "loss": 1.5742,
      "step": 1305
    },
    {
      "epoch": 0.2712357217030114,
      "grad_norm": 0.5083610415458679,
      "learning_rate": 8e-05,
      "loss": 1.5634,
      "step": 1306
    },
    {
      "epoch": 0.2714434060228453,
      "grad_norm": 0.4782449007034302,
      "learning_rate": 8e-05,
      "loss": 1.593,
      "step": 1307
    },
    {
      "epoch": 0.27165109034267915,
      "grad_norm": 0.5111173391342163,
      "learning_rate": 8e-05,
      "loss": 1.5827,
      "step": 1308
    },
    {
      "epoch": 0.27185877466251296,
      "grad_norm": 0.4896087646484375,
      "learning_rate": 8e-05,
      "loss": 1.5713,
      "step": 1309
    },
    {
      "epoch": 0.2720664589823468,
      "grad_norm": 0.50101238489151,
      "learning_rate": 8e-05,
      "loss": 1.5708,
      "step": 1310
    },
    {
      "epoch": 0.2722741433021807,
      "grad_norm": 0.4787428677082062,
      "learning_rate": 8e-05,
      "loss": 1.53,
      "step": 1311
    },
    {
      "epoch": 0.27248182762201456,
      "grad_norm": 0.4917222857475281,
      "learning_rate": 8e-05,
      "loss": 1.6704,
      "step": 1312
    },
    {
      "epoch": 0.2726895119418484,
      "grad_norm": 0.4935707449913025,
      "learning_rate": 8e-05,
      "loss": 1.5838,
      "step": 1313
    },
    {
      "epoch": 0.27289719626168224,
      "grad_norm": 0.48440101742744446,
      "learning_rate": 8e-05,
      "loss": 1.6517,
      "step": 1314
    },
    {
      "epoch": 0.2731048805815161,
      "grad_norm": 0.47382527589797974,
      "learning_rate": 8e-05,
      "loss": 1.5884,
      "step": 1315
    },
    {
      "epoch": 0.27331256490135,
      "grad_norm": 0.5190213322639465,
      "learning_rate": 8e-05,
      "loss": 1.6797,
      "step": 1316
    },
    {
      "epoch": 0.2735202492211838,
      "grad_norm": 0.49640318751335144,
      "learning_rate": 8e-05,
      "loss": 1.6594,
      "step": 1317
    },
    {
      "epoch": 0.27372793354101765,
      "grad_norm": 0.5151790380477905,
      "learning_rate": 8e-05,
      "loss": 1.5557,
      "step": 1318
    },
    {
      "epoch": 0.2739356178608515,
      "grad_norm": 0.47358784079551697,
      "learning_rate": 8e-05,
      "loss": 1.5639,
      "step": 1319
    },
    {
      "epoch": 0.27414330218068533,
      "grad_norm": 0.5438617467880249,
      "learning_rate": 8e-05,
      "loss": 1.6362,
      "step": 1320
    },
    {
      "epoch": 0.2743509865005192,
      "grad_norm": 0.4804224967956543,
      "learning_rate": 8e-05,
      "loss": 1.5946,
      "step": 1321
    },
    {
      "epoch": 0.27455867082035307,
      "grad_norm": 0.5258082151412964,
      "learning_rate": 8e-05,
      "loss": 1.5561,
      "step": 1322
    },
    {
      "epoch": 0.27476635514018694,
      "grad_norm": 0.48843201994895935,
      "learning_rate": 8e-05,
      "loss": 1.5366,
      "step": 1323
    },
    {
      "epoch": 0.27497403946002075,
      "grad_norm": 0.47955572605133057,
      "learning_rate": 8e-05,
      "loss": 1.6056,
      "step": 1324
    },
    {
      "epoch": 0.2751817237798546,
      "grad_norm": 0.4964466392993927,
      "learning_rate": 8e-05,
      "loss": 1.6292,
      "step": 1325
    },
    {
      "epoch": 0.2753894080996885,
      "grad_norm": 0.4776870608329773,
      "learning_rate": 8e-05,
      "loss": 1.6147,
      "step": 1326
    },
    {
      "epoch": 0.27559709241952235,
      "grad_norm": 0.49412986636161804,
      "learning_rate": 8e-05,
      "loss": 1.596,
      "step": 1327
    },
    {
      "epoch": 0.27580477673935616,
      "grad_norm": 0.49872222542762756,
      "learning_rate": 8e-05,
      "loss": 1.622,
      "step": 1328
    },
    {
      "epoch": 0.27601246105919003,
      "grad_norm": 0.48729175329208374,
      "learning_rate": 8e-05,
      "loss": 1.6105,
      "step": 1329
    },
    {
      "epoch": 0.2762201453790239,
      "grad_norm": 0.4850686490535736,
      "learning_rate": 8e-05,
      "loss": 1.5978,
      "step": 1330
    },
    {
      "epoch": 0.27642782969885776,
      "grad_norm": 0.49125537276268005,
      "learning_rate": 8e-05,
      "loss": 1.5584,
      "step": 1331
    },
    {
      "epoch": 0.2766355140186916,
      "grad_norm": 0.5107133984565735,
      "learning_rate": 8e-05,
      "loss": 1.6258,
      "step": 1332
    },
    {
      "epoch": 0.27684319833852544,
      "grad_norm": 0.4897415339946747,
      "learning_rate": 8e-05,
      "loss": 1.6198,
      "step": 1333
    },
    {
      "epoch": 0.2770508826583593,
      "grad_norm": 0.4734158217906952,
      "learning_rate": 8e-05,
      "loss": 1.5732,
      "step": 1334
    },
    {
      "epoch": 0.2772585669781931,
      "grad_norm": 0.5000379681587219,
      "learning_rate": 8e-05,
      "loss": 1.6496,
      "step": 1335
    },
    {
      "epoch": 0.277466251298027,
      "grad_norm": 0.5036786198616028,
      "learning_rate": 8e-05,
      "loss": 1.6218,
      "step": 1336
    },
    {
      "epoch": 0.27767393561786086,
      "grad_norm": 0.4840192198753357,
      "learning_rate": 8e-05,
      "loss": 1.5044,
      "step": 1337
    },
    {
      "epoch": 0.2778816199376947,
      "grad_norm": 0.4741503894329071,
      "learning_rate": 8e-05,
      "loss": 1.5801,
      "step": 1338
    },
    {
      "epoch": 0.27808930425752854,
      "grad_norm": 0.488294392824173,
      "learning_rate": 8e-05,
      "loss": 1.597,
      "step": 1339
    },
    {
      "epoch": 0.2782969885773624,
      "grad_norm": 0.5086274147033691,
      "learning_rate": 8e-05,
      "loss": 1.6069,
      "step": 1340
    },
    {
      "epoch": 0.27850467289719627,
      "grad_norm": 0.510127604007721,
      "learning_rate": 8e-05,
      "loss": 1.596,
      "step": 1341
    },
    {
      "epoch": 0.27871235721703014,
      "grad_norm": 0.46870115399360657,
      "learning_rate": 8e-05,
      "loss": 1.5364,
      "step": 1342
    },
    {
      "epoch": 0.27892004153686395,
      "grad_norm": 0.45464587211608887,
      "learning_rate": 8e-05,
      "loss": 1.5018,
      "step": 1343
    },
    {
      "epoch": 0.2791277258566978,
      "grad_norm": 0.48707786202430725,
      "learning_rate": 8e-05,
      "loss": 1.5411,
      "step": 1344
    },
    {
      "epoch": 0.2793354101765317,
      "grad_norm": 0.4926038086414337,
      "learning_rate": 8e-05,
      "loss": 1.6213,
      "step": 1345
    },
    {
      "epoch": 0.27954309449636555,
      "grad_norm": 0.48424461483955383,
      "learning_rate": 8e-05,
      "loss": 1.6463,
      "step": 1346
    },
    {
      "epoch": 0.27975077881619936,
      "grad_norm": 0.48776450753211975,
      "learning_rate": 8e-05,
      "loss": 1.6108,
      "step": 1347
    },
    {
      "epoch": 0.27995846313603323,
      "grad_norm": 0.4794577360153198,
      "learning_rate": 8e-05,
      "loss": 1.5981,
      "step": 1348
    },
    {
      "epoch": 0.2801661474558671,
      "grad_norm": 0.4882426857948303,
      "learning_rate": 8e-05,
      "loss": 1.6774,
      "step": 1349
    },
    {
      "epoch": 0.2803738317757009,
      "grad_norm": 0.5013483166694641,
      "learning_rate": 8e-05,
      "loss": 1.6659,
      "step": 1350
    },
    {
      "epoch": 0.2805815160955348,
      "grad_norm": 0.4758848547935486,
      "learning_rate": 8e-05,
      "loss": 1.5826,
      "step": 1351
    },
    {
      "epoch": 0.28078920041536864,
      "grad_norm": 0.4662100374698639,
      "learning_rate": 8e-05,
      "loss": 1.5151,
      "step": 1352
    },
    {
      "epoch": 0.2809968847352025,
      "grad_norm": 0.47722524404525757,
      "learning_rate": 8e-05,
      "loss": 1.5534,
      "step": 1353
    },
    {
      "epoch": 0.2812045690550363,
      "grad_norm": 0.5073485374450684,
      "learning_rate": 8e-05,
      "loss": 1.6202,
      "step": 1354
    },
    {
      "epoch": 0.2814122533748702,
      "grad_norm": 0.49153366684913635,
      "learning_rate": 8e-05,
      "loss": 1.5921,
      "step": 1355
    },
    {
      "epoch": 0.28161993769470406,
      "grad_norm": 0.4862796664237976,
      "learning_rate": 8e-05,
      "loss": 1.5981,
      "step": 1356
    },
    {
      "epoch": 0.2818276220145379,
      "grad_norm": 0.49322280287742615,
      "learning_rate": 8e-05,
      "loss": 1.6894,
      "step": 1357
    },
    {
      "epoch": 0.28203530633437174,
      "grad_norm": 0.48775988817214966,
      "learning_rate": 8e-05,
      "loss": 1.6123,
      "step": 1358
    },
    {
      "epoch": 0.2822429906542056,
      "grad_norm": 0.47947344183921814,
      "learning_rate": 8e-05,
      "loss": 1.5709,
      "step": 1359
    },
    {
      "epoch": 0.2824506749740395,
      "grad_norm": 0.5054206848144531,
      "learning_rate": 8e-05,
      "loss": 1.6541,
      "step": 1360
    },
    {
      "epoch": 0.28265835929387334,
      "grad_norm": 0.4824366569519043,
      "learning_rate": 8e-05,
      "loss": 1.5929,
      "step": 1361
    },
    {
      "epoch": 0.28286604361370715,
      "grad_norm": 0.49978917837142944,
      "learning_rate": 8e-05,
      "loss": 1.6151,
      "step": 1362
    },
    {
      "epoch": 0.283073727933541,
      "grad_norm": 0.4934566915035248,
      "learning_rate": 8e-05,
      "loss": 1.6641,
      "step": 1363
    },
    {
      "epoch": 0.2832814122533749,
      "grad_norm": 0.4824499785900116,
      "learning_rate": 8e-05,
      "loss": 1.6189,
      "step": 1364
    },
    {
      "epoch": 0.2834890965732087,
      "grad_norm": 0.47832030057907104,
      "learning_rate": 8e-05,
      "loss": 1.5083,
      "step": 1365
    },
    {
      "epoch": 0.28369678089304257,
      "grad_norm": 0.4860023260116577,
      "learning_rate": 8e-05,
      "loss": 1.565,
      "step": 1366
    },
    {
      "epoch": 0.28390446521287643,
      "grad_norm": 0.48774632811546326,
      "learning_rate": 8e-05,
      "loss": 1.638,
      "step": 1367
    },
    {
      "epoch": 0.2841121495327103,
      "grad_norm": 0.47733739018440247,
      "learning_rate": 8e-05,
      "loss": 1.6151,
      "step": 1368
    },
    {
      "epoch": 0.2843198338525441,
      "grad_norm": 0.48727935552597046,
      "learning_rate": 8e-05,
      "loss": 1.605,
      "step": 1369
    },
    {
      "epoch": 0.284527518172378,
      "grad_norm": 0.4770117998123169,
      "learning_rate": 8e-05,
      "loss": 1.6759,
      "step": 1370
    },
    {
      "epoch": 0.28473520249221185,
      "grad_norm": 0.49704787135124207,
      "learning_rate": 8e-05,
      "loss": 1.6116,
      "step": 1371
    },
    {
      "epoch": 0.2849428868120457,
      "grad_norm": 0.48451000452041626,
      "learning_rate": 8e-05,
      "loss": 1.5776,
      "step": 1372
    },
    {
      "epoch": 0.2851505711318795,
      "grad_norm": 0.49385419487953186,
      "learning_rate": 8e-05,
      "loss": 1.6142,
      "step": 1373
    },
    {
      "epoch": 0.2853582554517134,
      "grad_norm": 0.49187561869621277,
      "learning_rate": 8e-05,
      "loss": 1.6076,
      "step": 1374
    },
    {
      "epoch": 0.28556593977154726,
      "grad_norm": 0.4695932865142822,
      "learning_rate": 8e-05,
      "loss": 1.5801,
      "step": 1375
    },
    {
      "epoch": 0.2857736240913811,
      "grad_norm": 0.49953845143318176,
      "learning_rate": 8e-05,
      "loss": 1.617,
      "step": 1376
    },
    {
      "epoch": 0.28598130841121494,
      "grad_norm": 0.49016183614730835,
      "learning_rate": 8e-05,
      "loss": 1.6751,
      "step": 1377
    },
    {
      "epoch": 0.2861889927310488,
      "grad_norm": 0.5039032101631165,
      "learning_rate": 8e-05,
      "loss": 1.5667,
      "step": 1378
    },
    {
      "epoch": 0.2863966770508827,
      "grad_norm": 0.5153660178184509,
      "learning_rate": 8e-05,
      "loss": 1.5351,
      "step": 1379
    },
    {
      "epoch": 0.2866043613707165,
      "grad_norm": 0.4923986792564392,
      "learning_rate": 8e-05,
      "loss": 1.6022,
      "step": 1380
    },
    {
      "epoch": 0.28681204569055035,
      "grad_norm": 0.4953305125236511,
      "learning_rate": 8e-05,
      "loss": 1.6172,
      "step": 1381
    },
    {
      "epoch": 0.2870197300103842,
      "grad_norm": 0.48334407806396484,
      "learning_rate": 8e-05,
      "loss": 1.5416,
      "step": 1382
    },
    {
      "epoch": 0.2872274143302181,
      "grad_norm": 0.4850659966468811,
      "learning_rate": 8e-05,
      "loss": 1.6704,
      "step": 1383
    },
    {
      "epoch": 0.2874350986500519,
      "grad_norm": 0.5282303690910339,
      "learning_rate": 8e-05,
      "loss": 1.6134,
      "step": 1384
    },
    {
      "epoch": 0.28764278296988577,
      "grad_norm": 0.4880530834197998,
      "learning_rate": 8e-05,
      "loss": 1.6108,
      "step": 1385
    },
    {
      "epoch": 0.28785046728971964,
      "grad_norm": 0.5376881957054138,
      "learning_rate": 8e-05,
      "loss": 1.6852,
      "step": 1386
    },
    {
      "epoch": 0.2880581516095535,
      "grad_norm": 0.5560894012451172,
      "learning_rate": 8e-05,
      "loss": 1.677,
      "step": 1387
    },
    {
      "epoch": 0.2882658359293873,
      "grad_norm": 0.5074867606163025,
      "learning_rate": 8e-05,
      "loss": 1.5932,
      "step": 1388
    },
    {
      "epoch": 0.2884735202492212,
      "grad_norm": 0.4841192066669464,
      "learning_rate": 8e-05,
      "loss": 1.5258,
      "step": 1389
    },
    {
      "epoch": 0.28868120456905505,
      "grad_norm": 0.5013852715492249,
      "learning_rate": 8e-05,
      "loss": 1.5784,
      "step": 1390
    },
    {
      "epoch": 0.28888888888888886,
      "grad_norm": 0.494810551404953,
      "learning_rate": 8e-05,
      "loss": 1.5967,
      "step": 1391
    },
    {
      "epoch": 0.28909657320872273,
      "grad_norm": 0.49166232347488403,
      "learning_rate": 8e-05,
      "loss": 1.5865,
      "step": 1392
    },
    {
      "epoch": 0.2893042575285566,
      "grad_norm": 0.5110693573951721,
      "learning_rate": 8e-05,
      "loss": 1.5969,
      "step": 1393
    },
    {
      "epoch": 0.28951194184839046,
      "grad_norm": 0.542017936706543,
      "learning_rate": 8e-05,
      "loss": 1.6099,
      "step": 1394
    },
    {
      "epoch": 0.2897196261682243,
      "grad_norm": 0.5205735564231873,
      "learning_rate": 8e-05,
      "loss": 1.6194,
      "step": 1395
    },
    {
      "epoch": 0.28992731048805814,
      "grad_norm": 0.542843222618103,
      "learning_rate": 8e-05,
      "loss": 1.5778,
      "step": 1396
    },
    {
      "epoch": 0.290134994807892,
      "grad_norm": 0.5020806789398193,
      "learning_rate": 8e-05,
      "loss": 1.6645,
      "step": 1397
    },
    {
      "epoch": 0.2903426791277259,
      "grad_norm": 0.48713013529777527,
      "learning_rate": 8e-05,
      "loss": 1.5229,
      "step": 1398
    },
    {
      "epoch": 0.2905503634475597,
      "grad_norm": 0.49321314692497253,
      "learning_rate": 8e-05,
      "loss": 1.6144,
      "step": 1399
    },
    {
      "epoch": 0.29075804776739356,
      "grad_norm": 0.4920049011707306,
      "learning_rate": 8e-05,
      "loss": 1.6081,
      "step": 1400
    },
    {
      "epoch": 0.2909657320872274,
      "grad_norm": 0.4814404249191284,
      "learning_rate": 8e-05,
      "loss": 1.5424,
      "step": 1401
    },
    {
      "epoch": 0.2911734164070613,
      "grad_norm": 0.48672202229499817,
      "learning_rate": 8e-05,
      "loss": 1.5617,
      "step": 1402
    },
    {
      "epoch": 0.2913811007268951,
      "grad_norm": 0.4993645250797272,
      "learning_rate": 8e-05,
      "loss": 1.5552,
      "step": 1403
    },
    {
      "epoch": 0.29158878504672897,
      "grad_norm": 0.493221640586853,
      "learning_rate": 8e-05,
      "loss": 1.6083,
      "step": 1404
    },
    {
      "epoch": 0.29179646936656284,
      "grad_norm": 0.5242977738380432,
      "learning_rate": 8e-05,
      "loss": 1.6794,
      "step": 1405
    },
    {
      "epoch": 0.29200415368639665,
      "grad_norm": 0.5010468363761902,
      "learning_rate": 8e-05,
      "loss": 1.6118,
      "step": 1406
    },
    {
      "epoch": 0.2922118380062305,
      "grad_norm": 0.4755447506904602,
      "learning_rate": 8e-05,
      "loss": 1.5602,
      "step": 1407
    },
    {
      "epoch": 0.2924195223260644,
      "grad_norm": 0.48869359493255615,
      "learning_rate": 8e-05,
      "loss": 1.6312,
      "step": 1408
    },
    {
      "epoch": 0.29262720664589825,
      "grad_norm": 0.505214512348175,
      "learning_rate": 8e-05,
      "loss": 1.7103,
      "step": 1409
    },
    {
      "epoch": 0.29283489096573206,
      "grad_norm": 0.4835333526134491,
      "learning_rate": 8e-05,
      "loss": 1.5842,
      "step": 1410
    },
    {
      "epoch": 0.29304257528556593,
      "grad_norm": 0.47053441405296326,
      "learning_rate": 8e-05,
      "loss": 1.5655,
      "step": 1411
    },
    {
      "epoch": 0.2932502596053998,
      "grad_norm": 0.4897231161594391,
      "learning_rate": 8e-05,
      "loss": 1.5427,
      "step": 1412
    },
    {
      "epoch": 0.29345794392523367,
      "grad_norm": 0.4817632734775543,
      "learning_rate": 8e-05,
      "loss": 1.5703,
      "step": 1413
    },
    {
      "epoch": 0.2936656282450675,
      "grad_norm": 0.5053437352180481,
      "learning_rate": 8e-05,
      "loss": 1.5884,
      "step": 1414
    },
    {
      "epoch": 0.29387331256490135,
      "grad_norm": 0.48642995953559875,
      "learning_rate": 8e-05,
      "loss": 1.5685,
      "step": 1415
    },
    {
      "epoch": 0.2940809968847352,
      "grad_norm": 0.4815353453159332,
      "learning_rate": 8e-05,
      "loss": 1.6158,
      "step": 1416
    },
    {
      "epoch": 0.2942886812045691,
      "grad_norm": 0.49331390857696533,
      "learning_rate": 8e-05,
      "loss": 1.6346,
      "step": 1417
    },
    {
      "epoch": 0.2944963655244029,
      "grad_norm": 0.5165848135948181,
      "learning_rate": 8e-05,
      "loss": 1.6028,
      "step": 1418
    },
    {
      "epoch": 0.29470404984423676,
      "grad_norm": 0.4790689945220947,
      "learning_rate": 8e-05,
      "loss": 1.6281,
      "step": 1419
    },
    {
      "epoch": 0.2949117341640706,
      "grad_norm": 0.46984192728996277,
      "learning_rate": 8e-05,
      "loss": 1.5487,
      "step": 1420
    },
    {
      "epoch": 0.29511941848390444,
      "grad_norm": 0.5013918876647949,
      "learning_rate": 8e-05,
      "loss": 1.5629,
      "step": 1421
    },
    {
      "epoch": 0.2953271028037383,
      "grad_norm": 0.5373260974884033,
      "learning_rate": 8e-05,
      "loss": 1.6564,
      "step": 1422
    },
    {
      "epoch": 0.2955347871235722,
      "grad_norm": 0.5395689606666565,
      "learning_rate": 8e-05,
      "loss": 1.6307,
      "step": 1423
    },
    {
      "epoch": 0.29574247144340604,
      "grad_norm": 0.5050875544548035,
      "learning_rate": 8e-05,
      "loss": 1.5577,
      "step": 1424
    },
    {
      "epoch": 0.29595015576323985,
      "grad_norm": 0.46294352412223816,
      "learning_rate": 8e-05,
      "loss": 1.5705,
      "step": 1425
    },
    {
      "epoch": 0.2961578400830737,
      "grad_norm": 0.4974973201751709,
      "learning_rate": 8e-05,
      "loss": 1.5908,
      "step": 1426
    },
    {
      "epoch": 0.2963655244029076,
      "grad_norm": 0.5147601366043091,
      "learning_rate": 8e-05,
      "loss": 1.5833,
      "step": 1427
    },
    {
      "epoch": 0.29657320872274145,
      "grad_norm": 0.5149271488189697,
      "learning_rate": 8e-05,
      "loss": 1.6029,
      "step": 1428
    },
    {
      "epoch": 0.29678089304257527,
      "grad_norm": 0.48388561606407166,
      "learning_rate": 8e-05,
      "loss": 1.5629,
      "step": 1429
    },
    {
      "epoch": 0.29698857736240913,
      "grad_norm": 0.4807972013950348,
      "learning_rate": 8e-05,
      "loss": 1.5618,
      "step": 1430
    },
    {
      "epoch": 0.297196261682243,
      "grad_norm": 0.4944983124732971,
      "learning_rate": 8e-05,
      "loss": 1.6518,
      "step": 1431
    },
    {
      "epoch": 0.29740394600207687,
      "grad_norm": 0.467385470867157,
      "learning_rate": 8e-05,
      "loss": 1.5768,
      "step": 1432
    },
    {
      "epoch": 0.2976116303219107,
      "grad_norm": 0.4817611277103424,
      "learning_rate": 8e-05,
      "loss": 1.559,
      "step": 1433
    },
    {
      "epoch": 0.29781931464174455,
      "grad_norm": 0.4777863919734955,
      "learning_rate": 8e-05,
      "loss": 1.5712,
      "step": 1434
    },
    {
      "epoch": 0.2980269989615784,
      "grad_norm": 0.46716588735580444,
      "learning_rate": 8e-05,
      "loss": 1.5301,
      "step": 1435
    },
    {
      "epoch": 0.2982346832814122,
      "grad_norm": 0.49335551261901855,
      "learning_rate": 8e-05,
      "loss": 1.633,
      "step": 1436
    },
    {
      "epoch": 0.2984423676012461,
      "grad_norm": 0.47975221276283264,
      "learning_rate": 8e-05,
      "loss": 1.5948,
      "step": 1437
    },
    {
      "epoch": 0.29865005192107996,
      "grad_norm": 0.4699002504348755,
      "learning_rate": 8e-05,
      "loss": 1.5368,
      "step": 1438
    },
    {
      "epoch": 0.29885773624091383,
      "grad_norm": 0.5026534199714661,
      "learning_rate": 8e-05,
      "loss": 1.6412,
      "step": 1439
    },
    {
      "epoch": 0.29906542056074764,
      "grad_norm": 0.4809010326862335,
      "learning_rate": 8e-05,
      "loss": 1.5511,
      "step": 1440
    },
    {
      "epoch": 0.2992731048805815,
      "grad_norm": 0.5056330561637878,
      "learning_rate": 8e-05,
      "loss": 1.6516,
      "step": 1441
    },
    {
      "epoch": 0.2994807892004154,
      "grad_norm": 0.49161332845687866,
      "learning_rate": 8e-05,
      "loss": 1.5964,
      "step": 1442
    },
    {
      "epoch": 0.29968847352024924,
      "grad_norm": 0.49896249175071716,
      "learning_rate": 8e-05,
      "loss": 1.644,
      "step": 1443
    },
    {
      "epoch": 0.29989615784008306,
      "grad_norm": 0.4822237491607666,
      "learning_rate": 8e-05,
      "loss": 1.6416,
      "step": 1444
    },
    {
      "epoch": 0.3001038421599169,
      "grad_norm": 0.48241591453552246,
      "learning_rate": 8e-05,
      "loss": 1.6328,
      "step": 1445
    },
    {
      "epoch": 0.3003115264797508,
      "grad_norm": 0.48107171058654785,
      "learning_rate": 8e-05,
      "loss": 1.6912,
      "step": 1446
    },
    {
      "epoch": 0.30051921079958466,
      "grad_norm": 0.4950486421585083,
      "learning_rate": 8e-05,
      "loss": 1.5809,
      "step": 1447
    },
    {
      "epoch": 0.30072689511941847,
      "grad_norm": 0.516384482383728,
      "learning_rate": 8e-05,
      "loss": 1.5433,
      "step": 1448
    },
    {
      "epoch": 0.30093457943925234,
      "grad_norm": 0.47670841217041016,
      "learning_rate": 8e-05,
      "loss": 1.5641,
      "step": 1449
    },
    {
      "epoch": 0.3011422637590862,
      "grad_norm": 0.4875502288341522,
      "learning_rate": 8e-05,
      "loss": 1.5905,
      "step": 1450
    },
    {
      "epoch": 0.30134994807892,
      "grad_norm": 0.4830038547515869,
      "learning_rate": 8e-05,
      "loss": 1.5338,
      "step": 1451
    },
    {
      "epoch": 0.3015576323987539,
      "grad_norm": 0.4864535331726074,
      "learning_rate": 8e-05,
      "loss": 1.5884,
      "step": 1452
    },
    {
      "epoch": 0.30176531671858775,
      "grad_norm": 0.48559343814849854,
      "learning_rate": 8e-05,
      "loss": 1.5684,
      "step": 1453
    },
    {
      "epoch": 0.3019730010384216,
      "grad_norm": 0.4923616051673889,
      "learning_rate": 8e-05,
      "loss": 1.5975,
      "step": 1454
    },
    {
      "epoch": 0.30218068535825543,
      "grad_norm": 0.4875079095363617,
      "learning_rate": 8e-05,
      "loss": 1.5574,
      "step": 1455
    },
    {
      "epoch": 0.3023883696780893,
      "grad_norm": 0.5041059851646423,
      "learning_rate": 8e-05,
      "loss": 1.566,
      "step": 1456
    },
    {
      "epoch": 0.30259605399792316,
      "grad_norm": 0.5048243403434753,
      "learning_rate": 8e-05,
      "loss": 1.5744,
      "step": 1457
    },
    {
      "epoch": 0.30280373831775703,
      "grad_norm": 0.48077791929244995,
      "learning_rate": 8e-05,
      "loss": 1.6088,
      "step": 1458
    },
    {
      "epoch": 0.30301142263759084,
      "grad_norm": 0.49609705805778503,
      "learning_rate": 8e-05,
      "loss": 1.6863,
      "step": 1459
    },
    {
      "epoch": 0.3032191069574247,
      "grad_norm": 0.5017523765563965,
      "learning_rate": 8e-05,
      "loss": 1.6013,
      "step": 1460
    },
    {
      "epoch": 0.3034267912772586,
      "grad_norm": 0.4952293634414673,
      "learning_rate": 8e-05,
      "loss": 1.6157,
      "step": 1461
    },
    {
      "epoch": 0.30363447559709245,
      "grad_norm": 0.4877483546733856,
      "learning_rate": 8e-05,
      "loss": 1.5992,
      "step": 1462
    },
    {
      "epoch": 0.30384215991692626,
      "grad_norm": 0.4839021861553192,
      "learning_rate": 8e-05,
      "loss": 1.66,
      "step": 1463
    },
    {
      "epoch": 0.3040498442367601,
      "grad_norm": 0.523611307144165,
      "learning_rate": 8e-05,
      "loss": 1.6295,
      "step": 1464
    },
    {
      "epoch": 0.304257528556594,
      "grad_norm": 0.4874498248100281,
      "learning_rate": 8e-05,
      "loss": 1.5778,
      "step": 1465
    },
    {
      "epoch": 0.3044652128764278,
      "grad_norm": 0.49205684661865234,
      "learning_rate": 8e-05,
      "loss": 1.575,
      "step": 1466
    },
    {
      "epoch": 0.30467289719626167,
      "grad_norm": 0.5001851916313171,
      "learning_rate": 8e-05,
      "loss": 1.6221,
      "step": 1467
    },
    {
      "epoch": 0.30488058151609554,
      "grad_norm": 0.5134277939796448,
      "learning_rate": 8e-05,
      "loss": 1.5791,
      "step": 1468
    },
    {
      "epoch": 0.3050882658359294,
      "grad_norm": 0.4986041784286499,
      "learning_rate": 8e-05,
      "loss": 1.6249,
      "step": 1469
    },
    {
      "epoch": 0.3052959501557632,
      "grad_norm": 0.4865819811820984,
      "learning_rate": 8e-05,
      "loss": 1.6065,
      "step": 1470
    },
    {
      "epoch": 0.3055036344755971,
      "grad_norm": 0.48902952671051025,
      "learning_rate": 8e-05,
      "loss": 1.6383,
      "step": 1471
    },
    {
      "epoch": 0.30571131879543095,
      "grad_norm": 0.5065701007843018,
      "learning_rate": 8e-05,
      "loss": 1.5784,
      "step": 1472
    },
    {
      "epoch": 0.3059190031152648,
      "grad_norm": 0.48749783635139465,
      "learning_rate": 8e-05,
      "loss": 1.5601,
      "step": 1473
    },
    {
      "epoch": 0.30612668743509863,
      "grad_norm": 0.517052948474884,
      "learning_rate": 8e-05,
      "loss": 1.6655,
      "step": 1474
    },
    {
      "epoch": 0.3063343717549325,
      "grad_norm": 0.4751560389995575,
      "learning_rate": 8e-05,
      "loss": 1.5892,
      "step": 1475
    },
    {
      "epoch": 0.30654205607476637,
      "grad_norm": 0.4731171429157257,
      "learning_rate": 8e-05,
      "loss": 1.5153,
      "step": 1476
    },
    {
      "epoch": 0.30674974039460023,
      "grad_norm": 0.5009525418281555,
      "learning_rate": 8e-05,
      "loss": 1.6547,
      "step": 1477
    },
    {
      "epoch": 0.30695742471443405,
      "grad_norm": 0.4930509626865387,
      "learning_rate": 8e-05,
      "loss": 1.513,
      "step": 1478
    },
    {
      "epoch": 0.3071651090342679,
      "grad_norm": 0.5279929041862488,
      "learning_rate": 8e-05,
      "loss": 1.6086,
      "step": 1479
    },
    {
      "epoch": 0.3073727933541018,
      "grad_norm": 0.4923204481601715,
      "learning_rate": 8e-05,
      "loss": 1.5659,
      "step": 1480
    },
    {
      "epoch": 0.3075804776739356,
      "grad_norm": 0.4844015836715698,
      "learning_rate": 8e-05,
      "loss": 1.6044,
      "step": 1481
    },
    {
      "epoch": 0.30778816199376946,
      "grad_norm": 0.5059119462966919,
      "learning_rate": 8e-05,
      "loss": 1.61,
      "step": 1482
    },
    {
      "epoch": 0.3079958463136033,
      "grad_norm": 0.5250138640403748,
      "learning_rate": 8e-05,
      "loss": 1.5921,
      "step": 1483
    },
    {
      "epoch": 0.3082035306334372,
      "grad_norm": 0.512847363948822,
      "learning_rate": 8e-05,
      "loss": 1.6694,
      "step": 1484
    },
    {
      "epoch": 0.308411214953271,
      "grad_norm": 0.5249167084693909,
      "learning_rate": 8e-05,
      "loss": 1.6558,
      "step": 1485
    },
    {
      "epoch": 0.3086188992731049,
      "grad_norm": 0.5241632461547852,
      "learning_rate": 8e-05,
      "loss": 1.6323,
      "step": 1486
    },
    {
      "epoch": 0.30882658359293874,
      "grad_norm": 0.49093738198280334,
      "learning_rate": 8e-05,
      "loss": 1.6032,
      "step": 1487
    },
    {
      "epoch": 0.3090342679127726,
      "grad_norm": 0.49326688051223755,
      "learning_rate": 8e-05,
      "loss": 1.5726,
      "step": 1488
    },
    {
      "epoch": 0.3092419522326064,
      "grad_norm": 0.49778178334236145,
      "learning_rate": 8e-05,
      "loss": 1.5792,
      "step": 1489
    },
    {
      "epoch": 0.3094496365524403,
      "grad_norm": 0.49950110912323,
      "learning_rate": 8e-05,
      "loss": 1.5706,
      "step": 1490
    },
    {
      "epoch": 0.30965732087227416,
      "grad_norm": 0.5080294013023376,
      "learning_rate": 8e-05,
      "loss": 1.64,
      "step": 1491
    },
    {
      "epoch": 0.309865005192108,
      "grad_norm": 0.5099862217903137,
      "learning_rate": 8e-05,
      "loss": 1.6631,
      "step": 1492
    },
    {
      "epoch": 0.31007268951194183,
      "grad_norm": 0.49251455068588257,
      "learning_rate": 8e-05,
      "loss": 1.5711,
      "step": 1493
    },
    {
      "epoch": 0.3102803738317757,
      "grad_norm": 0.49985072016716003,
      "learning_rate": 8e-05,
      "loss": 1.5286,
      "step": 1494
    },
    {
      "epoch": 0.31048805815160957,
      "grad_norm": 0.5015414953231812,
      "learning_rate": 8e-05,
      "loss": 1.5992,
      "step": 1495
    },
    {
      "epoch": 0.3106957424714434,
      "grad_norm": 0.5101944804191589,
      "learning_rate": 8e-05,
      "loss": 1.5845,
      "step": 1496
    },
    {
      "epoch": 0.31090342679127725,
      "grad_norm": 0.5153470635414124,
      "learning_rate": 8e-05,
      "loss": 1.7133,
      "step": 1497
    },
    {
      "epoch": 0.3111111111111111,
      "grad_norm": 0.5057701468467712,
      "learning_rate": 8e-05,
      "loss": 1.6648,
      "step": 1498
    },
    {
      "epoch": 0.311318795430945,
      "grad_norm": 0.5079687833786011,
      "learning_rate": 8e-05,
      "loss": 1.6237,
      "step": 1499
    },
    {
      "epoch": 0.3115264797507788,
      "grad_norm": 0.475410133600235,
      "learning_rate": 8e-05,
      "loss": 1.6165,
      "step": 1500
    },
    {
      "epoch": 0.31173416407061266,
      "grad_norm": 0.5307968854904175,
      "learning_rate": 8e-05,
      "loss": 1.5859,
      "step": 1501
    },
    {
      "epoch": 0.31194184839044653,
      "grad_norm": 0.47486138343811035,
      "learning_rate": 8e-05,
      "loss": 1.5676,
      "step": 1502
    },
    {
      "epoch": 0.3121495327102804,
      "grad_norm": 0.5097178220748901,
      "learning_rate": 8e-05,
      "loss": 1.5636,
      "step": 1503
    },
    {
      "epoch": 0.3123572170301142,
      "grad_norm": 0.4760870039463043,
      "learning_rate": 8e-05,
      "loss": 1.5721,
      "step": 1504
    },
    {
      "epoch": 0.3125649013499481,
      "grad_norm": 0.4842042624950409,
      "learning_rate": 8e-05,
      "loss": 1.6645,
      "step": 1505
    },
    {
      "epoch": 0.31277258566978194,
      "grad_norm": 0.5192430019378662,
      "learning_rate": 8e-05,
      "loss": 1.6495,
      "step": 1506
    },
    {
      "epoch": 0.3129802699896158,
      "grad_norm": 0.4800757169723511,
      "learning_rate": 8e-05,
      "loss": 1.5741,
      "step": 1507
    },
    {
      "epoch": 0.3131879543094496,
      "grad_norm": 0.4967311918735504,
      "learning_rate": 8e-05,
      "loss": 1.6412,
      "step": 1508
    },
    {
      "epoch": 0.3133956386292835,
      "grad_norm": 0.4981956481933594,
      "learning_rate": 8e-05,
      "loss": 1.6111,
      "step": 1509
    },
    {
      "epoch": 0.31360332294911736,
      "grad_norm": 0.5167219042778015,
      "learning_rate": 8e-05,
      "loss": 1.6406,
      "step": 1510
    },
    {
      "epoch": 0.31381100726895117,
      "grad_norm": 0.48610803484916687,
      "learning_rate": 8e-05,
      "loss": 1.5525,
      "step": 1511
    },
    {
      "epoch": 0.31401869158878504,
      "grad_norm": 0.47807812690734863,
      "learning_rate": 8e-05,
      "loss": 1.5205,
      "step": 1512
    },
    {
      "epoch": 0.3142263759086189,
      "grad_norm": 0.5001601576805115,
      "learning_rate": 8e-05,
      "loss": 1.6265,
      "step": 1513
    },
    {
      "epoch": 0.31443406022845277,
      "grad_norm": 0.5112875699996948,
      "learning_rate": 8e-05,
      "loss": 1.6357,
      "step": 1514
    },
    {
      "epoch": 0.3146417445482866,
      "grad_norm": 0.4997986853122711,
      "learning_rate": 8e-05,
      "loss": 1.6137,
      "step": 1515
    },
    {
      "epoch": 0.31484942886812045,
      "grad_norm": 0.4741184115409851,
      "learning_rate": 8e-05,
      "loss": 1.6266,
      "step": 1516
    },
    {
      "epoch": 0.3150571131879543,
      "grad_norm": 0.5144311785697937,
      "learning_rate": 8e-05,
      "loss": 1.5782,
      "step": 1517
    },
    {
      "epoch": 0.3152647975077882,
      "grad_norm": 0.5066690444946289,
      "learning_rate": 8e-05,
      "loss": 1.6677,
      "step": 1518
    },
    {
      "epoch": 0.315472481827622,
      "grad_norm": 0.5042521953582764,
      "learning_rate": 8e-05,
      "loss": 1.5916,
      "step": 1519
    },
    {
      "epoch": 0.31568016614745587,
      "grad_norm": 0.4977933466434479,
      "learning_rate": 8e-05,
      "loss": 1.5574,
      "step": 1520
    },
    {
      "epoch": 0.31588785046728973,
      "grad_norm": 0.5137097239494324,
      "learning_rate": 8e-05,
      "loss": 1.6411,
      "step": 1521
    },
    {
      "epoch": 0.3160955347871236,
      "grad_norm": 0.4740540385246277,
      "learning_rate": 8e-05,
      "loss": 1.57,
      "step": 1522
    },
    {
      "epoch": 0.3163032191069574,
      "grad_norm": 0.5093042254447937,
      "learning_rate": 8e-05,
      "loss": 1.6752,
      "step": 1523
    },
    {
      "epoch": 0.3165109034267913,
      "grad_norm": 0.47612494230270386,
      "learning_rate": 8e-05,
      "loss": 1.5722,
      "step": 1524
    },
    {
      "epoch": 0.31671858774662515,
      "grad_norm": 0.5046824812889099,
      "learning_rate": 8e-05,
      "loss": 1.649,
      "step": 1525
    },
    {
      "epoch": 0.31692627206645896,
      "grad_norm": 0.49512314796447754,
      "learning_rate": 8e-05,
      "loss": 1.5808,
      "step": 1526
    },
    {
      "epoch": 0.3171339563862928,
      "grad_norm": 0.4888535737991333,
      "learning_rate": 8e-05,
      "loss": 1.6138,
      "step": 1527
    },
    {
      "epoch": 0.3173416407061267,
      "grad_norm": 0.5067277550697327,
      "learning_rate": 8e-05,
      "loss": 1.626,
      "step": 1528
    },
    {
      "epoch": 0.31754932502596056,
      "grad_norm": 0.47953104972839355,
      "learning_rate": 8e-05,
      "loss": 1.6125,
      "step": 1529
    },
    {
      "epoch": 0.3177570093457944,
      "grad_norm": 0.5111708641052246,
      "learning_rate": 8e-05,
      "loss": 1.5847,
      "step": 1530
    },
    {
      "epoch": 0.31796469366562824,
      "grad_norm": 0.48254770040512085,
      "learning_rate": 8e-05,
      "loss": 1.5371,
      "step": 1531
    },
    {
      "epoch": 0.3181723779854621,
      "grad_norm": 0.6176688075065613,
      "learning_rate": 8e-05,
      "loss": 1.5441,
      "step": 1532
    },
    {
      "epoch": 0.318380062305296,
      "grad_norm": 0.4850156307220459,
      "learning_rate": 8e-05,
      "loss": 1.573,
      "step": 1533
    },
    {
      "epoch": 0.3185877466251298,
      "grad_norm": 0.4983154237270355,
      "learning_rate": 8e-05,
      "loss": 1.5298,
      "step": 1534
    },
    {
      "epoch": 0.31879543094496365,
      "grad_norm": 0.5193361043930054,
      "learning_rate": 8e-05,
      "loss": 1.6185,
      "step": 1535
    },
    {
      "epoch": 0.3190031152647975,
      "grad_norm": 0.483345627784729,
      "learning_rate": 8e-05,
      "loss": 1.5397,
      "step": 1536
    },
    {
      "epoch": 0.31921079958463133,
      "grad_norm": 0.48317113518714905,
      "learning_rate": 8e-05,
      "loss": 1.6233,
      "step": 1537
    },
    {
      "epoch": 0.3194184839044652,
      "grad_norm": 0.4782498776912689,
      "learning_rate": 8e-05,
      "loss": 1.5813,
      "step": 1538
    },
    {
      "epoch": 0.31962616822429907,
      "grad_norm": 0.47537708282470703,
      "learning_rate": 8e-05,
      "loss": 1.5454,
      "step": 1539
    },
    {
      "epoch": 0.31983385254413293,
      "grad_norm": 0.4868844151496887,
      "learning_rate": 8e-05,
      "loss": 1.6,
      "step": 1540
    },
    {
      "epoch": 0.32004153686396675,
      "grad_norm": 0.5081853866577148,
      "learning_rate": 8e-05,
      "loss": 1.5926,
      "step": 1541
    },
    {
      "epoch": 0.3202492211838006,
      "grad_norm": 0.48315268754959106,
      "learning_rate": 8e-05,
      "loss": 1.5898,
      "step": 1542
    },
    {
      "epoch": 0.3204569055036345,
      "grad_norm": 0.4909741282463074,
      "learning_rate": 8e-05,
      "loss": 1.5828,
      "step": 1543
    },
    {
      "epoch": 0.32066458982346835,
      "grad_norm": 0.49861210584640503,
      "learning_rate": 8e-05,
      "loss": 1.6455,
      "step": 1544
    },
    {
      "epoch": 0.32087227414330216,
      "grad_norm": 0.478344589471817,
      "learning_rate": 8e-05,
      "loss": 1.6188,
      "step": 1545
    },
    {
      "epoch": 0.32107995846313603,
      "grad_norm": 0.5060451030731201,
      "learning_rate": 8e-05,
      "loss": 1.5827,
      "step": 1546
    },
    {
      "epoch": 0.3212876427829699,
      "grad_norm": 0.48769256472587585,
      "learning_rate": 8e-05,
      "loss": 1.6526,
      "step": 1547
    },
    {
      "epoch": 0.32149532710280376,
      "grad_norm": 0.492779403924942,
      "learning_rate": 8e-05,
      "loss": 1.6032,
      "step": 1548
    },
    {
      "epoch": 0.3217030114226376,
      "grad_norm": 0.4987448453903198,
      "learning_rate": 8e-05,
      "loss": 1.5259,
      "step": 1549
    },
    {
      "epoch": 0.32191069574247144,
      "grad_norm": 0.48688334226608276,
      "learning_rate": 8e-05,
      "loss": 1.5775,
      "step": 1550
    },
    {
      "epoch": 0.3221183800623053,
      "grad_norm": 0.4970977306365967,
      "learning_rate": 8e-05,
      "loss": 1.6524,
      "step": 1551
    },
    {
      "epoch": 0.3223260643821391,
      "grad_norm": 0.5189974904060364,
      "learning_rate": 8e-05,
      "loss": 1.5758,
      "step": 1552
    },
    {
      "epoch": 0.322533748701973,
      "grad_norm": 0.48815828561782837,
      "learning_rate": 8e-05,
      "loss": 1.5439,
      "step": 1553
    },
    {
      "epoch": 0.32274143302180686,
      "grad_norm": 0.4929064214229584,
      "learning_rate": 8e-05,
      "loss": 1.545,
      "step": 1554
    },
    {
      "epoch": 0.3229491173416407,
      "grad_norm": 0.48322778940200806,
      "learning_rate": 8e-05,
      "loss": 1.5601,
      "step": 1555
    },
    {
      "epoch": 0.32315680166147454,
      "grad_norm": 0.5003882646560669,
      "learning_rate": 8e-05,
      "loss": 1.5581,
      "step": 1556
    },
    {
      "epoch": 0.3233644859813084,
      "grad_norm": 0.5089275240898132,
      "learning_rate": 8e-05,
      "loss": 1.6147,
      "step": 1557
    },
    {
      "epoch": 0.32357217030114227,
      "grad_norm": 0.48767319321632385,
      "learning_rate": 8e-05,
      "loss": 1.5953,
      "step": 1558
    },
    {
      "epoch": 0.32377985462097614,
      "grad_norm": 0.5171458125114441,
      "learning_rate": 8e-05,
      "loss": 1.6395,
      "step": 1559
    },
    {
      "epoch": 0.32398753894080995,
      "grad_norm": 0.4896124601364136,
      "learning_rate": 8e-05,
      "loss": 1.6219,
      "step": 1560
    },
    {
      "epoch": 0.3241952232606438,
      "grad_norm": 0.4953930974006653,
      "learning_rate": 8e-05,
      "loss": 1.6007,
      "step": 1561
    },
    {
      "epoch": 0.3244029075804777,
      "grad_norm": 0.5223313570022583,
      "learning_rate": 8e-05,
      "loss": 1.6687,
      "step": 1562
    },
    {
      "epoch": 0.32461059190031155,
      "grad_norm": 0.50156569480896,
      "learning_rate": 8e-05,
      "loss": 1.6407,
      "step": 1563
    },
    {
      "epoch": 0.32481827622014536,
      "grad_norm": 0.48758435249328613,
      "learning_rate": 8e-05,
      "loss": 1.5725,
      "step": 1564
    },
    {
      "epoch": 0.32502596053997923,
      "grad_norm": 0.49146777391433716,
      "learning_rate": 8e-05,
      "loss": 1.5366,
      "step": 1565
    },
    {
      "epoch": 0.3252336448598131,
      "grad_norm": 0.5082824230194092,
      "learning_rate": 8e-05,
      "loss": 1.6088,
      "step": 1566
    },
    {
      "epoch": 0.3254413291796469,
      "grad_norm": 0.5137510299682617,
      "learning_rate": 8e-05,
      "loss": 1.6697,
      "step": 1567
    },
    {
      "epoch": 0.3256490134994808,
      "grad_norm": 0.48702284693717957,
      "learning_rate": 8e-05,
      "loss": 1.5483,
      "step": 1568
    },
    {
      "epoch": 0.32585669781931464,
      "grad_norm": 0.4997518062591553,
      "learning_rate": 8e-05,
      "loss": 1.5455,
      "step": 1569
    },
    {
      "epoch": 0.3260643821391485,
      "grad_norm": 0.4797830283641815,
      "learning_rate": 8e-05,
      "loss": 1.6014,
      "step": 1570
    },
    {
      "epoch": 0.3262720664589823,
      "grad_norm": 0.4881299138069153,
      "learning_rate": 8e-05,
      "loss": 1.5972,
      "step": 1571
    },
    {
      "epoch": 0.3264797507788162,
      "grad_norm": 0.4965881407260895,
      "learning_rate": 8e-05,
      "loss": 1.5945,
      "step": 1572
    },
    {
      "epoch": 0.32668743509865006,
      "grad_norm": 0.49967294931411743,
      "learning_rate": 8e-05,
      "loss": 1.5934,
      "step": 1573
    },
    {
      "epoch": 0.3268951194184839,
      "grad_norm": 0.5094643831253052,
      "learning_rate": 8e-05,
      "loss": 1.6762,
      "step": 1574
    },
    {
      "epoch": 0.32710280373831774,
      "grad_norm": 0.4757131040096283,
      "learning_rate": 8e-05,
      "loss": 1.6471,
      "step": 1575
    },
    {
      "epoch": 0.3273104880581516,
      "grad_norm": 0.5002568364143372,
      "learning_rate": 8e-05,
      "loss": 1.538,
      "step": 1576
    },
    {
      "epoch": 0.3275181723779855,
      "grad_norm": 0.5113179087638855,
      "learning_rate": 8e-05,
      "loss": 1.5968,
      "step": 1577
    },
    {
      "epoch": 0.32772585669781934,
      "grad_norm": 0.4950655698776245,
      "learning_rate": 8e-05,
      "loss": 1.6031,
      "step": 1578
    },
    {
      "epoch": 0.32793354101765315,
      "grad_norm": 0.4921990931034088,
      "learning_rate": 8e-05,
      "loss": 1.5819,
      "step": 1579
    },
    {
      "epoch": 0.328141225337487,
      "grad_norm": 0.48755714297294617,
      "learning_rate": 8e-05,
      "loss": 1.5708,
      "step": 1580
    },
    {
      "epoch": 0.3283489096573209,
      "grad_norm": 0.49338051676750183,
      "learning_rate": 8e-05,
      "loss": 1.6067,
      "step": 1581
    },
    {
      "epoch": 0.3285565939771547,
      "grad_norm": 0.4969124495983124,
      "learning_rate": 8e-05,
      "loss": 1.5753,
      "step": 1582
    },
    {
      "epoch": 0.32876427829698857,
      "grad_norm": 0.49120715260505676,
      "learning_rate": 8e-05,
      "loss": 1.5947,
      "step": 1583
    },
    {
      "epoch": 0.32897196261682243,
      "grad_norm": 0.47977736592292786,
      "learning_rate": 8e-05,
      "loss": 1.5577,
      "step": 1584
    },
    {
      "epoch": 0.3291796469366563,
      "grad_norm": 0.49589675664901733,
      "learning_rate": 8e-05,
      "loss": 1.6218,
      "step": 1585
    },
    {
      "epoch": 0.3293873312564901,
      "grad_norm": 0.506141185760498,
      "learning_rate": 8e-05,
      "loss": 1.5925,
      "step": 1586
    },
    {
      "epoch": 0.329595015576324,
      "grad_norm": 0.5115627646446228,
      "learning_rate": 8e-05,
      "loss": 1.6107,
      "step": 1587
    },
    {
      "epoch": 0.32980269989615785,
      "grad_norm": 0.5363782644271851,
      "learning_rate": 8e-05,
      "loss": 1.6134,
      "step": 1588
    },
    {
      "epoch": 0.3300103842159917,
      "grad_norm": 0.5303878784179688,
      "learning_rate": 8e-05,
      "loss": 1.6054,
      "step": 1589
    },
    {
      "epoch": 0.3302180685358255,
      "grad_norm": 0.5335170030593872,
      "learning_rate": 8e-05,
      "loss": 1.5758,
      "step": 1590
    },
    {
      "epoch": 0.3304257528556594,
      "grad_norm": 0.49111613631248474,
      "learning_rate": 8e-05,
      "loss": 1.6672,
      "step": 1591
    },
    {
      "epoch": 0.33063343717549326,
      "grad_norm": 0.4870830774307251,
      "learning_rate": 8e-05,
      "loss": 1.6392,
      "step": 1592
    },
    {
      "epoch": 0.33084112149532713,
      "grad_norm": 0.5026964545249939,
      "learning_rate": 8e-05,
      "loss": 1.546,
      "step": 1593
    },
    {
      "epoch": 0.33104880581516094,
      "grad_norm": 0.47665566205978394,
      "learning_rate": 8e-05,
      "loss": 1.5447,
      "step": 1594
    },
    {
      "epoch": 0.3312564901349948,
      "grad_norm": 0.4723517894744873,
      "learning_rate": 8e-05,
      "loss": 1.5622,
      "step": 1595
    },
    {
      "epoch": 0.3314641744548287,
      "grad_norm": 0.48958736658096313,
      "learning_rate": 8e-05,
      "loss": 1.5539,
      "step": 1596
    },
    {
      "epoch": 0.3316718587746625,
      "grad_norm": 0.5126999616622925,
      "learning_rate": 8e-05,
      "loss": 1.6348,
      "step": 1597
    },
    {
      "epoch": 0.33187954309449635,
      "grad_norm": 0.5067579746246338,
      "learning_rate": 8e-05,
      "loss": 1.6046,
      "step": 1598
    },
    {
      "epoch": 0.3320872274143302,
      "grad_norm": 0.4940493404865265,
      "learning_rate": 8e-05,
      "loss": 1.5553,
      "step": 1599
    },
    {
      "epoch": 0.3322949117341641,
      "grad_norm": 0.5140849947929382,
      "learning_rate": 8e-05,
      "loss": 1.6741,
      "step": 1600
    },
    {
      "epoch": 0.3325025960539979,
      "grad_norm": 0.47822049260139465,
      "learning_rate": 8e-05,
      "loss": 1.5474,
      "step": 1601
    },
    {
      "epoch": 0.33271028037383177,
      "grad_norm": 0.4843752384185791,
      "learning_rate": 8e-05,
      "loss": 1.5194,
      "step": 1602
    },
    {
      "epoch": 0.33291796469366564,
      "grad_norm": 0.5137794613838196,
      "learning_rate": 8e-05,
      "loss": 1.6013,
      "step": 1603
    },
    {
      "epoch": 0.3331256490134995,
      "grad_norm": 0.4957379698753357,
      "learning_rate": 8e-05,
      "loss": 1.5426,
      "step": 1604
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.527511477470398,
      "learning_rate": 8e-05,
      "loss": 1.656,
      "step": 1605
    },
    {
      "epoch": 0.3335410176531672,
      "grad_norm": 0.4836502969264984,
      "learning_rate": 8e-05,
      "loss": 1.5422,
      "step": 1606
    },
    {
      "epoch": 0.33374870197300105,
      "grad_norm": 0.47433504462242126,
      "learning_rate": 8e-05,
      "loss": 1.5434,
      "step": 1607
    },
    {
      "epoch": 0.3339563862928349,
      "grad_norm": 0.5102041363716125,
      "learning_rate": 8e-05,
      "loss": 1.7392,
      "step": 1608
    },
    {
      "epoch": 0.33416407061266873,
      "grad_norm": 0.4976068437099457,
      "learning_rate": 8e-05,
      "loss": 1.6558,
      "step": 1609
    },
    {
      "epoch": 0.3343717549325026,
      "grad_norm": 0.49436476826667786,
      "learning_rate": 8e-05,
      "loss": 1.5691,
      "step": 1610
    },
    {
      "epoch": 0.33457943925233646,
      "grad_norm": 0.4891389310359955,
      "learning_rate": 8e-05,
      "loss": 1.6462,
      "step": 1611
    },
    {
      "epoch": 0.3347871235721703,
      "grad_norm": 0.5754526853561401,
      "learning_rate": 8e-05,
      "loss": 1.6686,
      "step": 1612
    },
    {
      "epoch": 0.33499480789200414,
      "grad_norm": 0.49320200085639954,
      "learning_rate": 8e-05,
      "loss": 1.6849,
      "step": 1613
    },
    {
      "epoch": 0.335202492211838,
      "grad_norm": 0.4995647370815277,
      "learning_rate": 8e-05,
      "loss": 1.5772,
      "step": 1614
    },
    {
      "epoch": 0.3354101765316719,
      "grad_norm": 0.48338550329208374,
      "learning_rate": 8e-05,
      "loss": 1.5217,
      "step": 1615
    },
    {
      "epoch": 0.3356178608515057,
      "grad_norm": 0.47079768776893616,
      "learning_rate": 8e-05,
      "loss": 1.5064,
      "step": 1616
    },
    {
      "epoch": 0.33582554517133956,
      "grad_norm": 0.5078921914100647,
      "learning_rate": 8e-05,
      "loss": 1.588,
      "step": 1617
    },
    {
      "epoch": 0.3360332294911734,
      "grad_norm": 0.5282722115516663,
      "learning_rate": 8e-05,
      "loss": 1.5059,
      "step": 1618
    },
    {
      "epoch": 0.3362409138110073,
      "grad_norm": 0.5180701613426208,
      "learning_rate": 8e-05,
      "loss": 1.5696,
      "step": 1619
    },
    {
      "epoch": 0.3364485981308411,
      "grad_norm": 0.5044119954109192,
      "learning_rate": 8e-05,
      "loss": 1.5619,
      "step": 1620
    },
    {
      "epoch": 0.33665628245067497,
      "grad_norm": 0.5039921402931213,
      "learning_rate": 8e-05,
      "loss": 1.6287,
      "step": 1621
    },
    {
      "epoch": 0.33686396677050884,
      "grad_norm": 0.5082249641418457,
      "learning_rate": 8e-05,
      "loss": 1.5873,
      "step": 1622
    },
    {
      "epoch": 0.3370716510903427,
      "grad_norm": 0.48252761363983154,
      "learning_rate": 8e-05,
      "loss": 1.5671,
      "step": 1623
    },
    {
      "epoch": 0.3372793354101765,
      "grad_norm": 0.5083025097846985,
      "learning_rate": 8e-05,
      "loss": 1.6022,
      "step": 1624
    },
    {
      "epoch": 0.3374870197300104,
      "grad_norm": 0.5168206095695496,
      "learning_rate": 8e-05,
      "loss": 1.6703,
      "step": 1625
    },
    {
      "epoch": 0.33769470404984425,
      "grad_norm": 0.4816582500934601,
      "learning_rate": 8e-05,
      "loss": 1.5788,
      "step": 1626
    },
    {
      "epoch": 0.33790238836967806,
      "grad_norm": 0.490945965051651,
      "learning_rate": 8e-05,
      "loss": 1.5626,
      "step": 1627
    },
    {
      "epoch": 0.33811007268951193,
      "grad_norm": 0.5070871710777283,
      "learning_rate": 8e-05,
      "loss": 1.676,
      "step": 1628
    },
    {
      "epoch": 0.3383177570093458,
      "grad_norm": 0.49754798412323,
      "learning_rate": 8e-05,
      "loss": 1.6187,
      "step": 1629
    },
    {
      "epoch": 0.33852544132917967,
      "grad_norm": 0.500566303730011,
      "learning_rate": 8e-05,
      "loss": 1.6224,
      "step": 1630
    },
    {
      "epoch": 0.3387331256490135,
      "grad_norm": 0.5073097944259644,
      "learning_rate": 8e-05,
      "loss": 1.6215,
      "step": 1631
    },
    {
      "epoch": 0.33894080996884735,
      "grad_norm": 0.4712946116924286,
      "learning_rate": 8e-05,
      "loss": 1.5831,
      "step": 1632
    },
    {
      "epoch": 0.3391484942886812,
      "grad_norm": 0.49603715538978577,
      "learning_rate": 8e-05,
      "loss": 1.6233,
      "step": 1633
    },
    {
      "epoch": 0.3393561786085151,
      "grad_norm": 0.5190393328666687,
      "learning_rate": 8e-05,
      "loss": 1.6539,
      "step": 1634
    },
    {
      "epoch": 0.3395638629283489,
      "grad_norm": 0.49072492122650146,
      "learning_rate": 8e-05,
      "loss": 1.51,
      "step": 1635
    },
    {
      "epoch": 0.33977154724818276,
      "grad_norm": 0.49581220746040344,
      "learning_rate": 8e-05,
      "loss": 1.5618,
      "step": 1636
    },
    {
      "epoch": 0.3399792315680166,
      "grad_norm": 0.4976690113544464,
      "learning_rate": 8e-05,
      "loss": 1.619,
      "step": 1637
    },
    {
      "epoch": 0.3401869158878505,
      "grad_norm": 0.4935644865036011,
      "learning_rate": 8e-05,
      "loss": 1.6651,
      "step": 1638
    },
    {
      "epoch": 0.3403946002076843,
      "grad_norm": 0.5480844974517822,
      "learning_rate": 8e-05,
      "loss": 1.7232,
      "step": 1639
    },
    {
      "epoch": 0.3406022845275182,
      "grad_norm": 0.5207673907279968,
      "learning_rate": 8e-05,
      "loss": 1.646,
      "step": 1640
    },
    {
      "epoch": 0.34080996884735204,
      "grad_norm": 0.5026153326034546,
      "learning_rate": 8e-05,
      "loss": 1.5155,
      "step": 1641
    },
    {
      "epoch": 0.34101765316718585,
      "grad_norm": 0.4776262044906616,
      "learning_rate": 8e-05,
      "loss": 1.5248,
      "step": 1642
    },
    {
      "epoch": 0.3412253374870197,
      "grad_norm": 0.5081517100334167,
      "learning_rate": 8e-05,
      "loss": 1.6524,
      "step": 1643
    },
    {
      "epoch": 0.3414330218068536,
      "grad_norm": 0.4948354661464691,
      "learning_rate": 8e-05,
      "loss": 1.6231,
      "step": 1644
    },
    {
      "epoch": 0.34164070612668745,
      "grad_norm": 0.5102015733718872,
      "learning_rate": 8e-05,
      "loss": 1.6152,
      "step": 1645
    },
    {
      "epoch": 0.34184839044652127,
      "grad_norm": 0.5125258564949036,
      "learning_rate": 8e-05,
      "loss": 1.644,
      "step": 1646
    },
    {
      "epoch": 0.34205607476635513,
      "grad_norm": 0.49975356459617615,
      "learning_rate": 8e-05,
      "loss": 1.5621,
      "step": 1647
    },
    {
      "epoch": 0.342263759086189,
      "grad_norm": 0.5100876092910767,
      "learning_rate": 8e-05,
      "loss": 1.5948,
      "step": 1648
    },
    {
      "epoch": 0.34247144340602287,
      "grad_norm": 0.4993584156036377,
      "learning_rate": 8e-05,
      "loss": 1.5826,
      "step": 1649
    },
    {
      "epoch": 0.3426791277258567,
      "grad_norm": 0.4974980652332306,
      "learning_rate": 8e-05,
      "loss": 1.5497,
      "step": 1650
    },
    {
      "epoch": 0.34288681204569055,
      "grad_norm": 0.5110526084899902,
      "learning_rate": 8e-05,
      "loss": 1.5656,
      "step": 1651
    },
    {
      "epoch": 0.3430944963655244,
      "grad_norm": 0.4917437434196472,
      "learning_rate": 8e-05,
      "loss": 1.5912,
      "step": 1652
    },
    {
      "epoch": 0.3433021806853583,
      "grad_norm": 0.5046085119247437,
      "learning_rate": 8e-05,
      "loss": 1.6313,
      "step": 1653
    },
    {
      "epoch": 0.3435098650051921,
      "grad_norm": 0.5008227229118347,
      "learning_rate": 8e-05,
      "loss": 1.5603,
      "step": 1654
    },
    {
      "epoch": 0.34371754932502596,
      "grad_norm": 0.48017212748527527,
      "learning_rate": 8e-05,
      "loss": 1.5851,
      "step": 1655
    },
    {
      "epoch": 0.34392523364485983,
      "grad_norm": 0.5130458474159241,
      "learning_rate": 8e-05,
      "loss": 1.6015,
      "step": 1656
    },
    {
      "epoch": 0.34413291796469364,
      "grad_norm": 0.5011090636253357,
      "learning_rate": 8e-05,
      "loss": 1.586,
      "step": 1657
    },
    {
      "epoch": 0.3443406022845275,
      "grad_norm": 0.4897594451904297,
      "learning_rate": 8e-05,
      "loss": 1.5021,
      "step": 1658
    },
    {
      "epoch": 0.3445482866043614,
      "grad_norm": 0.6089925765991211,
      "learning_rate": 8e-05,
      "loss": 1.6325,
      "step": 1659
    },
    {
      "epoch": 0.34475597092419524,
      "grad_norm": 0.5230244994163513,
      "learning_rate": 8e-05,
      "loss": 1.6087,
      "step": 1660
    },
    {
      "epoch": 0.34496365524402905,
      "grad_norm": 0.5000819563865662,
      "learning_rate": 8e-05,
      "loss": 1.6082,
      "step": 1661
    },
    {
      "epoch": 0.3451713395638629,
      "grad_norm": 0.5213269591331482,
      "learning_rate": 8e-05,
      "loss": 1.645,
      "step": 1662
    },
    {
      "epoch": 0.3453790238836968,
      "grad_norm": 0.4994923174381256,
      "learning_rate": 8e-05,
      "loss": 1.5727,
      "step": 1663
    },
    {
      "epoch": 0.34558670820353066,
      "grad_norm": 0.5164852142333984,
      "learning_rate": 8e-05,
      "loss": 1.6262,
      "step": 1664
    },
    {
      "epoch": 0.34579439252336447,
      "grad_norm": 0.5150447487831116,
      "learning_rate": 8e-05,
      "loss": 1.6247,
      "step": 1665
    },
    {
      "epoch": 0.34600207684319834,
      "grad_norm": 0.5243363976478577,
      "learning_rate": 8e-05,
      "loss": 1.6594,
      "step": 1666
    },
    {
      "epoch": 0.3462097611630322,
      "grad_norm": 0.48954543471336365,
      "learning_rate": 8e-05,
      "loss": 1.5796,
      "step": 1667
    },
    {
      "epoch": 0.34641744548286607,
      "grad_norm": 0.5317123532295227,
      "learning_rate": 8e-05,
      "loss": 1.6389,
      "step": 1668
    },
    {
      "epoch": 0.3466251298026999,
      "grad_norm": 0.48210567235946655,
      "learning_rate": 8e-05,
      "loss": 1.611,
      "step": 1669
    },
    {
      "epoch": 0.34683281412253375,
      "grad_norm": 0.48340731859207153,
      "learning_rate": 8e-05,
      "loss": 1.5521,
      "step": 1670
    },
    {
      "epoch": 0.3470404984423676,
      "grad_norm": 0.5049550533294678,
      "learning_rate": 8e-05,
      "loss": 1.5751,
      "step": 1671
    },
    {
      "epoch": 0.34724818276220143,
      "grad_norm": 0.5040217638015747,
      "learning_rate": 8e-05,
      "loss": 1.5263,
      "step": 1672
    },
    {
      "epoch": 0.3474558670820353,
      "grad_norm": 0.5133804678916931,
      "learning_rate": 8e-05,
      "loss": 1.651,
      "step": 1673
    },
    {
      "epoch": 0.34766355140186916,
      "grad_norm": 0.4757126569747925,
      "learning_rate": 8e-05,
      "loss": 1.5168,
      "step": 1674
    },
    {
      "epoch": 0.34787123572170303,
      "grad_norm": 0.5413657426834106,
      "learning_rate": 8e-05,
      "loss": 1.6296,
      "step": 1675
    },
    {
      "epoch": 0.34807892004153684,
      "grad_norm": 0.5213454961776733,
      "learning_rate": 8e-05,
      "loss": 1.5978,
      "step": 1676
    },
    {
      "epoch": 0.3482866043613707,
      "grad_norm": 0.4883634150028229,
      "learning_rate": 8e-05,
      "loss": 1.6488,
      "step": 1677
    },
    {
      "epoch": 0.3484942886812046,
      "grad_norm": 0.5572940111160278,
      "learning_rate": 8e-05,
      "loss": 1.5427,
      "step": 1678
    },
    {
      "epoch": 0.34870197300103845,
      "grad_norm": 0.5513511300086975,
      "learning_rate": 8e-05,
      "loss": 1.648,
      "step": 1679
    },
    {
      "epoch": 0.34890965732087226,
      "grad_norm": 0.5172803401947021,
      "learning_rate": 8e-05,
      "loss": 1.6542,
      "step": 1680
    },
    {
      "epoch": 0.3491173416407061,
      "grad_norm": 0.5045535564422607,
      "learning_rate": 8e-05,
      "loss": 1.5983,
      "step": 1681
    },
    {
      "epoch": 0.34932502596054,
      "grad_norm": 0.49136364459991455,
      "learning_rate": 8e-05,
      "loss": 1.5774,
      "step": 1682
    },
    {
      "epoch": 0.34953271028037386,
      "grad_norm": 0.4870191216468811,
      "learning_rate": 8e-05,
      "loss": 1.5807,
      "step": 1683
    },
    {
      "epoch": 0.34974039460020767,
      "grad_norm": 0.524327278137207,
      "learning_rate": 8e-05,
      "loss": 1.6471,
      "step": 1684
    },
    {
      "epoch": 0.34994807892004154,
      "grad_norm": 0.5097550749778748,
      "learning_rate": 8e-05,
      "loss": 1.5905,
      "step": 1685
    },
    {
      "epoch": 0.3501557632398754,
      "grad_norm": 0.516825795173645,
      "learning_rate": 8e-05,
      "loss": 1.6539,
      "step": 1686
    },
    {
      "epoch": 0.3503634475597092,
      "grad_norm": 0.5085751414299011,
      "learning_rate": 8e-05,
      "loss": 1.6649,
      "step": 1687
    },
    {
      "epoch": 0.3505711318795431,
      "grad_norm": 0.49749553203582764,
      "learning_rate": 8e-05,
      "loss": 1.5978,
      "step": 1688
    },
    {
      "epoch": 0.35077881619937695,
      "grad_norm": 0.49809882044792175,
      "learning_rate": 8e-05,
      "loss": 1.6228,
      "step": 1689
    },
    {
      "epoch": 0.3509865005192108,
      "grad_norm": 0.48347166180610657,
      "learning_rate": 8e-05,
      "loss": 1.5463,
      "step": 1690
    },
    {
      "epoch": 0.35119418483904463,
      "grad_norm": 0.4847400486469269,
      "learning_rate": 8e-05,
      "loss": 1.5379,
      "step": 1691
    },
    {
      "epoch": 0.3514018691588785,
      "grad_norm": 0.48331671953201294,
      "learning_rate": 8e-05,
      "loss": 1.5704,
      "step": 1692
    },
    {
      "epoch": 0.35160955347871237,
      "grad_norm": 0.49434080719947815,
      "learning_rate": 8e-05,
      "loss": 1.6055,
      "step": 1693
    },
    {
      "epoch": 0.35181723779854623,
      "grad_norm": 0.5171917676925659,
      "learning_rate": 8e-05,
      "loss": 1.6155,
      "step": 1694
    },
    {
      "epoch": 0.35202492211838005,
      "grad_norm": 0.4984574317932129,
      "learning_rate": 8e-05,
      "loss": 1.5811,
      "step": 1695
    },
    {
      "epoch": 0.3522326064382139,
      "grad_norm": 0.4882890284061432,
      "learning_rate": 8e-05,
      "loss": 1.5948,
      "step": 1696
    },
    {
      "epoch": 0.3524402907580478,
      "grad_norm": 0.48760464787483215,
      "learning_rate": 8e-05,
      "loss": 1.5775,
      "step": 1697
    },
    {
      "epoch": 0.3526479750778816,
      "grad_norm": 0.48519834876060486,
      "learning_rate": 8e-05,
      "loss": 1.5734,
      "step": 1698
    },
    {
      "epoch": 0.35285565939771546,
      "grad_norm": 0.518786609172821,
      "learning_rate": 8e-05,
      "loss": 1.5969,
      "step": 1699
    },
    {
      "epoch": 0.3530633437175493,
      "grad_norm": 0.4933086335659027,
      "learning_rate": 8e-05,
      "loss": 1.5626,
      "step": 1700
    },
    {
      "epoch": 0.3532710280373832,
      "grad_norm": 0.5093357563018799,
      "learning_rate": 8e-05,
      "loss": 1.5513,
      "step": 1701
    },
    {
      "epoch": 0.353478712357217,
      "grad_norm": 0.5171862244606018,
      "learning_rate": 8e-05,
      "loss": 1.5887,
      "step": 1702
    },
    {
      "epoch": 0.3536863966770509,
      "grad_norm": 0.5157456994056702,
      "learning_rate": 8e-05,
      "loss": 1.6256,
      "step": 1703
    },
    {
      "epoch": 0.35389408099688474,
      "grad_norm": 0.5482586622238159,
      "learning_rate": 8e-05,
      "loss": 1.6649,
      "step": 1704
    },
    {
      "epoch": 0.3541017653167186,
      "grad_norm": 0.5034733414649963,
      "learning_rate": 8e-05,
      "loss": 1.617,
      "step": 1705
    },
    {
      "epoch": 0.3543094496365524,
      "grad_norm": 0.5075942277908325,
      "learning_rate": 8e-05,
      "loss": 1.6087,
      "step": 1706
    },
    {
      "epoch": 0.3545171339563863,
      "grad_norm": 0.5159339904785156,
      "learning_rate": 8e-05,
      "loss": 1.5778,
      "step": 1707
    },
    {
      "epoch": 0.35472481827622016,
      "grad_norm": 0.5066738724708557,
      "learning_rate": 8e-05,
      "loss": 1.6548,
      "step": 1708
    },
    {
      "epoch": 0.354932502596054,
      "grad_norm": 0.48170992732048035,
      "learning_rate": 8e-05,
      "loss": 1.6448,
      "step": 1709
    },
    {
      "epoch": 0.35514018691588783,
      "grad_norm": 0.47656309604644775,
      "learning_rate": 8e-05,
      "loss": 1.5139,
      "step": 1710
    },
    {
      "epoch": 0.3553478712357217,
      "grad_norm": 0.5038301944732666,
      "learning_rate": 8e-05,
      "loss": 1.5363,
      "step": 1711
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 0.5651991963386536,
      "learning_rate": 8e-05,
      "loss": 1.6422,
      "step": 1712
    },
    {
      "epoch": 0.3557632398753894,
      "grad_norm": 0.49263623356819153,
      "learning_rate": 8e-05,
      "loss": 1.5621,
      "step": 1713
    },
    {
      "epoch": 0.35597092419522325,
      "grad_norm": 0.5081530809402466,
      "learning_rate": 8e-05,
      "loss": 1.5904,
      "step": 1714
    },
    {
      "epoch": 0.3561786085150571,
      "grad_norm": 0.5229151844978333,
      "learning_rate": 8e-05,
      "loss": 1.6046,
      "step": 1715
    },
    {
      "epoch": 0.356386292834891,
      "grad_norm": 0.4824669361114502,
      "learning_rate": 8e-05,
      "loss": 1.5738,
      "step": 1716
    },
    {
      "epoch": 0.3565939771547248,
      "grad_norm": 0.5122133493423462,
      "learning_rate": 8e-05,
      "loss": 1.635,
      "step": 1717
    },
    {
      "epoch": 0.35680166147455866,
      "grad_norm": 0.4866729974746704,
      "learning_rate": 8e-05,
      "loss": 1.5987,
      "step": 1718
    },
    {
      "epoch": 0.35700934579439253,
      "grad_norm": 0.5040173530578613,
      "learning_rate": 8e-05,
      "loss": 1.5918,
      "step": 1719
    },
    {
      "epoch": 0.3572170301142264,
      "grad_norm": 0.496620774269104,
      "learning_rate": 8e-05,
      "loss": 1.5897,
      "step": 1720
    },
    {
      "epoch": 0.3574247144340602,
      "grad_norm": 0.521233081817627,
      "learning_rate": 8e-05,
      "loss": 1.6024,
      "step": 1721
    },
    {
      "epoch": 0.3576323987538941,
      "grad_norm": 0.4817935526371002,
      "learning_rate": 8e-05,
      "loss": 1.5798,
      "step": 1722
    },
    {
      "epoch": 0.35784008307372794,
      "grad_norm": 0.5107322335243225,
      "learning_rate": 8e-05,
      "loss": 1.6237,
      "step": 1723
    },
    {
      "epoch": 0.3580477673935618,
      "grad_norm": 0.49109092354774475,
      "learning_rate": 8e-05,
      "loss": 1.5852,
      "step": 1724
    },
    {
      "epoch": 0.3582554517133956,
      "grad_norm": 0.49932169914245605,
      "learning_rate": 8e-05,
      "loss": 1.5684,
      "step": 1725
    },
    {
      "epoch": 0.3584631360332295,
      "grad_norm": 0.4933273494243622,
      "learning_rate": 8e-05,
      "loss": 1.5987,
      "step": 1726
    },
    {
      "epoch": 0.35867082035306336,
      "grad_norm": 0.5152563452720642,
      "learning_rate": 8e-05,
      "loss": 1.6229,
      "step": 1727
    },
    {
      "epoch": 0.35887850467289717,
      "grad_norm": 0.49047115445137024,
      "learning_rate": 8e-05,
      "loss": 1.5985,
      "step": 1728
    },
    {
      "epoch": 0.35908618899273104,
      "grad_norm": 0.5049684643745422,
      "learning_rate": 8e-05,
      "loss": 1.5635,
      "step": 1729
    },
    {
      "epoch": 0.3592938733125649,
      "grad_norm": 0.4843440055847168,
      "learning_rate": 8e-05,
      "loss": 1.6021,
      "step": 1730
    },
    {
      "epoch": 0.35950155763239877,
      "grad_norm": 0.48820793628692627,
      "learning_rate": 8e-05,
      "loss": 1.5598,
      "step": 1731
    },
    {
      "epoch": 0.3597092419522326,
      "grad_norm": 0.5228747725486755,
      "learning_rate": 8e-05,
      "loss": 1.5545,
      "step": 1732
    },
    {
      "epoch": 0.35991692627206645,
      "grad_norm": 0.4973585903644562,
      "learning_rate": 8e-05,
      "loss": 1.5758,
      "step": 1733
    },
    {
      "epoch": 0.3601246105919003,
      "grad_norm": 0.5158231258392334,
      "learning_rate": 8e-05,
      "loss": 1.6215,
      "step": 1734
    },
    {
      "epoch": 0.3603322949117342,
      "grad_norm": 0.4845174551010132,
      "learning_rate": 8e-05,
      "loss": 1.5932,
      "step": 1735
    },
    {
      "epoch": 0.360539979231568,
      "grad_norm": 0.505311906337738,
      "learning_rate": 8e-05,
      "loss": 1.6716,
      "step": 1736
    },
    {
      "epoch": 0.36074766355140186,
      "grad_norm": 0.5100334286689758,
      "learning_rate": 8e-05,
      "loss": 1.6133,
      "step": 1737
    },
    {
      "epoch": 0.36095534787123573,
      "grad_norm": 0.510661244392395,
      "learning_rate": 8e-05,
      "loss": 1.6705,
      "step": 1738
    },
    {
      "epoch": 0.3611630321910696,
      "grad_norm": 0.49372726678848267,
      "learning_rate": 8e-05,
      "loss": 1.6432,
      "step": 1739
    },
    {
      "epoch": 0.3613707165109034,
      "grad_norm": 0.4995138645172119,
      "learning_rate": 8e-05,
      "loss": 1.5815,
      "step": 1740
    },
    {
      "epoch": 0.3615784008307373,
      "grad_norm": 0.518836259841919,
      "learning_rate": 8e-05,
      "loss": 1.5321,
      "step": 1741
    },
    {
      "epoch": 0.36178608515057115,
      "grad_norm": 0.5100981593132019,
      "learning_rate": 8e-05,
      "loss": 1.573,
      "step": 1742
    },
    {
      "epoch": 0.36199376947040496,
      "grad_norm": 0.4908148944377899,
      "learning_rate": 8e-05,
      "loss": 1.6022,
      "step": 1743
    },
    {
      "epoch": 0.3622014537902388,
      "grad_norm": 0.5129155516624451,
      "learning_rate": 8e-05,
      "loss": 1.5825,
      "step": 1744
    },
    {
      "epoch": 0.3624091381100727,
      "grad_norm": 0.4997279942035675,
      "learning_rate": 8e-05,
      "loss": 1.672,
      "step": 1745
    },
    {
      "epoch": 0.36261682242990656,
      "grad_norm": 0.4920547902584076,
      "learning_rate": 8e-05,
      "loss": 1.5973,
      "step": 1746
    },
    {
      "epoch": 0.36282450674974037,
      "grad_norm": 0.5037687420845032,
      "learning_rate": 8e-05,
      "loss": 1.4707,
      "step": 1747
    },
    {
      "epoch": 0.36303219106957424,
      "grad_norm": 0.5036182999610901,
      "learning_rate": 8e-05,
      "loss": 1.6161,
      "step": 1748
    },
    {
      "epoch": 0.3632398753894081,
      "grad_norm": 0.5063603520393372,
      "learning_rate": 8e-05,
      "loss": 1.5124,
      "step": 1749
    },
    {
      "epoch": 0.363447559709242,
      "grad_norm": 0.5060688257217407,
      "learning_rate": 8e-05,
      "loss": 1.553,
      "step": 1750
    },
    {
      "epoch": 0.3636552440290758,
      "grad_norm": 0.4944371283054352,
      "learning_rate": 8e-05,
      "loss": 1.572,
      "step": 1751
    },
    {
      "epoch": 0.36386292834890965,
      "grad_norm": 0.5195178985595703,
      "learning_rate": 8e-05,
      "loss": 1.5257,
      "step": 1752
    },
    {
      "epoch": 0.3640706126687435,
      "grad_norm": 0.4922371208667755,
      "learning_rate": 8e-05,
      "loss": 1.5622,
      "step": 1753
    },
    {
      "epoch": 0.3642782969885774,
      "grad_norm": 0.5210584998130798,
      "learning_rate": 8e-05,
      "loss": 1.6212,
      "step": 1754
    },
    {
      "epoch": 0.3644859813084112,
      "grad_norm": 0.5490279197692871,
      "learning_rate": 8e-05,
      "loss": 1.6151,
      "step": 1755
    },
    {
      "epoch": 0.36469366562824507,
      "grad_norm": 0.4872822165489197,
      "learning_rate": 8e-05,
      "loss": 1.5555,
      "step": 1756
    },
    {
      "epoch": 0.36490134994807893,
      "grad_norm": 0.5005167126655579,
      "learning_rate": 8e-05,
      "loss": 1.5788,
      "step": 1757
    },
    {
      "epoch": 0.36510903426791275,
      "grad_norm": 0.4953737258911133,
      "learning_rate": 8e-05,
      "loss": 1.6162,
      "step": 1758
    },
    {
      "epoch": 0.3653167185877466,
      "grad_norm": 0.5173213481903076,
      "learning_rate": 8e-05,
      "loss": 1.569,
      "step": 1759
    },
    {
      "epoch": 0.3655244029075805,
      "grad_norm": 0.5120421648025513,
      "learning_rate": 8e-05,
      "loss": 1.6336,
      "step": 1760
    },
    {
      "epoch": 0.36573208722741435,
      "grad_norm": 0.5063916444778442,
      "learning_rate": 8e-05,
      "loss": 1.5695,
      "step": 1761
    },
    {
      "epoch": 0.36593977154724816,
      "grad_norm": 0.5345103144645691,
      "learning_rate": 8e-05,
      "loss": 1.5974,
      "step": 1762
    },
    {
      "epoch": 0.36614745586708203,
      "grad_norm": 0.47018736600875854,
      "learning_rate": 8e-05,
      "loss": 1.507,
      "step": 1763
    },
    {
      "epoch": 0.3663551401869159,
      "grad_norm": 0.5206469297409058,
      "learning_rate": 8e-05,
      "loss": 1.5425,
      "step": 1764
    },
    {
      "epoch": 0.36656282450674976,
      "grad_norm": 0.49119699001312256,
      "learning_rate": 8e-05,
      "loss": 1.5484,
      "step": 1765
    },
    {
      "epoch": 0.3667705088265836,
      "grad_norm": 0.48123517632484436,
      "learning_rate": 8e-05,
      "loss": 1.5174,
      "step": 1766
    },
    {
      "epoch": 0.36697819314641744,
      "grad_norm": 0.5121713280677795,
      "learning_rate": 8e-05,
      "loss": 1.6451,
      "step": 1767
    },
    {
      "epoch": 0.3671858774662513,
      "grad_norm": 0.5062276124954224,
      "learning_rate": 8e-05,
      "loss": 1.5787,
      "step": 1768
    },
    {
      "epoch": 0.3673935617860852,
      "grad_norm": 0.4850326180458069,
      "learning_rate": 8e-05,
      "loss": 1.605,
      "step": 1769
    },
    {
      "epoch": 0.367601246105919,
      "grad_norm": 0.5524799823760986,
      "learning_rate": 8e-05,
      "loss": 1.6658,
      "step": 1770
    },
    {
      "epoch": 0.36780893042575286,
      "grad_norm": 0.5042113661766052,
      "learning_rate": 8e-05,
      "loss": 1.5878,
      "step": 1771
    },
    {
      "epoch": 0.3680166147455867,
      "grad_norm": 0.5291147828102112,
      "learning_rate": 8e-05,
      "loss": 1.6076,
      "step": 1772
    },
    {
      "epoch": 0.36822429906542054,
      "grad_norm": 0.5282319188117981,
      "learning_rate": 8e-05,
      "loss": 1.6024,
      "step": 1773
    },
    {
      "epoch": 0.3684319833852544,
      "grad_norm": 0.5120516419410706,
      "learning_rate": 8e-05,
      "loss": 1.5431,
      "step": 1774
    },
    {
      "epoch": 0.36863966770508827,
      "grad_norm": 0.5158196687698364,
      "learning_rate": 8e-05,
      "loss": 1.6004,
      "step": 1775
    },
    {
      "epoch": 0.36884735202492214,
      "grad_norm": 0.50716632604599,
      "learning_rate": 8e-05,
      "loss": 1.6615,
      "step": 1776
    },
    {
      "epoch": 0.36905503634475595,
      "grad_norm": 0.5380442142486572,
      "learning_rate": 8e-05,
      "loss": 1.5219,
      "step": 1777
    },
    {
      "epoch": 0.3692627206645898,
      "grad_norm": 0.4817923903465271,
      "learning_rate": 8e-05,
      "loss": 1.5422,
      "step": 1778
    },
    {
      "epoch": 0.3694704049844237,
      "grad_norm": 0.509099006652832,
      "learning_rate": 8e-05,
      "loss": 1.5917,
      "step": 1779
    },
    {
      "epoch": 0.36967808930425755,
      "grad_norm": 0.5473061800003052,
      "learning_rate": 8e-05,
      "loss": 1.6635,
      "step": 1780
    },
    {
      "epoch": 0.36988577362409136,
      "grad_norm": 0.5307387709617615,
      "learning_rate": 8e-05,
      "loss": 1.6171,
      "step": 1781
    },
    {
      "epoch": 0.37009345794392523,
      "grad_norm": 0.5154988169670105,
      "learning_rate": 8e-05,
      "loss": 1.5596,
      "step": 1782
    },
    {
      "epoch": 0.3703011422637591,
      "grad_norm": 0.4852389097213745,
      "learning_rate": 8e-05,
      "loss": 1.5276,
      "step": 1783
    },
    {
      "epoch": 0.37050882658359297,
      "grad_norm": 0.4957145154476166,
      "learning_rate": 8e-05,
      "loss": 1.5823,
      "step": 1784
    },
    {
      "epoch": 0.3707165109034268,
      "grad_norm": 0.5096319317817688,
      "learning_rate": 8e-05,
      "loss": 1.5288,
      "step": 1785
    },
    {
      "epoch": 0.37092419522326064,
      "grad_norm": 0.5101845860481262,
      "learning_rate": 8e-05,
      "loss": 1.6088,
      "step": 1786
    },
    {
      "epoch": 0.3711318795430945,
      "grad_norm": 0.5254088640213013,
      "learning_rate": 8e-05,
      "loss": 1.6396,
      "step": 1787
    },
    {
      "epoch": 0.3713395638629283,
      "grad_norm": 0.5071054697036743,
      "learning_rate": 8e-05,
      "loss": 1.569,
      "step": 1788
    },
    {
      "epoch": 0.3715472481827622,
      "grad_norm": 0.51131272315979,
      "learning_rate": 8e-05,
      "loss": 1.6019,
      "step": 1789
    },
    {
      "epoch": 0.37175493250259606,
      "grad_norm": 0.5088433623313904,
      "learning_rate": 8e-05,
      "loss": 1.5169,
      "step": 1790
    },
    {
      "epoch": 0.3719626168224299,
      "grad_norm": 0.5493926405906677,
      "learning_rate": 8e-05,
      "loss": 1.595,
      "step": 1791
    },
    {
      "epoch": 0.37217030114226374,
      "grad_norm": 0.4982267916202545,
      "learning_rate": 8e-05,
      "loss": 1.5716,
      "step": 1792
    },
    {
      "epoch": 0.3723779854620976,
      "grad_norm": 0.5942201614379883,
      "learning_rate": 8e-05,
      "loss": 1.6425,
      "step": 1793
    },
    {
      "epoch": 0.37258566978193147,
      "grad_norm": 0.4930921196937561,
      "learning_rate": 8e-05,
      "loss": 1.5942,
      "step": 1794
    },
    {
      "epoch": 0.37279335410176534,
      "grad_norm": 0.5201108455657959,
      "learning_rate": 8e-05,
      "loss": 1.6025,
      "step": 1795
    },
    {
      "epoch": 0.37300103842159915,
      "grad_norm": 0.5413525104522705,
      "learning_rate": 8e-05,
      "loss": 1.6227,
      "step": 1796
    },
    {
      "epoch": 0.373208722741433,
      "grad_norm": 0.4801349937915802,
      "learning_rate": 8e-05,
      "loss": 1.5519,
      "step": 1797
    },
    {
      "epoch": 0.3734164070612669,
      "grad_norm": 0.545444905757904,
      "learning_rate": 8e-05,
      "loss": 1.5614,
      "step": 1798
    },
    {
      "epoch": 0.37362409138110075,
      "grad_norm": 0.5094887614250183,
      "learning_rate": 8e-05,
      "loss": 1.5961,
      "step": 1799
    },
    {
      "epoch": 0.37383177570093457,
      "grad_norm": 0.5024911761283875,
      "learning_rate": 8e-05,
      "loss": 1.6617,
      "step": 1800
    },
    {
      "epoch": 0.37403946002076843,
      "grad_norm": 0.4914880692958832,
      "learning_rate": 8e-05,
      "loss": 1.547,
      "step": 1801
    },
    {
      "epoch": 0.3742471443406023,
      "grad_norm": 0.488719642162323,
      "learning_rate": 8e-05,
      "loss": 1.5906,
      "step": 1802
    },
    {
      "epoch": 0.3744548286604361,
      "grad_norm": 0.5118622779846191,
      "learning_rate": 8e-05,
      "loss": 1.5741,
      "step": 1803
    },
    {
      "epoch": 0.37466251298027,
      "grad_norm": 0.5298313498497009,
      "learning_rate": 8e-05,
      "loss": 1.6255,
      "step": 1804
    },
    {
      "epoch": 0.37487019730010385,
      "grad_norm": 0.46859586238861084,
      "learning_rate": 8e-05,
      "loss": 1.5109,
      "step": 1805
    },
    {
      "epoch": 0.3750778816199377,
      "grad_norm": 0.5379807949066162,
      "learning_rate": 8e-05,
      "loss": 1.6255,
      "step": 1806
    },
    {
      "epoch": 0.3752855659397715,
      "grad_norm": 0.4950341284275055,
      "learning_rate": 8e-05,
      "loss": 1.659,
      "step": 1807
    },
    {
      "epoch": 0.3754932502596054,
      "grad_norm": 0.49582797288894653,
      "learning_rate": 8e-05,
      "loss": 1.6734,
      "step": 1808
    },
    {
      "epoch": 0.37570093457943926,
      "grad_norm": 0.522197961807251,
      "learning_rate": 8e-05,
      "loss": 1.5411,
      "step": 1809
    },
    {
      "epoch": 0.37590861889927313,
      "grad_norm": 0.49645760655403137,
      "learning_rate": 8e-05,
      "loss": 1.5901,
      "step": 1810
    },
    {
      "epoch": 0.37611630321910694,
      "grad_norm": 0.5847791433334351,
      "learning_rate": 8e-05,
      "loss": 1.5925,
      "step": 1811
    },
    {
      "epoch": 0.3763239875389408,
      "grad_norm": 0.4936416447162628,
      "learning_rate": 8e-05,
      "loss": 1.6004,
      "step": 1812
    },
    {
      "epoch": 0.3765316718587747,
      "grad_norm": 0.5401305556297302,
      "learning_rate": 8e-05,
      "loss": 1.6375,
      "step": 1813
    },
    {
      "epoch": 0.37673935617860854,
      "grad_norm": 0.5169132351875305,
      "learning_rate": 8e-05,
      "loss": 1.6277,
      "step": 1814
    },
    {
      "epoch": 0.37694704049844235,
      "grad_norm": 0.5267390012741089,
      "learning_rate": 8e-05,
      "loss": 1.595,
      "step": 1815
    },
    {
      "epoch": 0.3771547248182762,
      "grad_norm": 0.5062999129295349,
      "learning_rate": 8e-05,
      "loss": 1.6584,
      "step": 1816
    },
    {
      "epoch": 0.3773624091381101,
      "grad_norm": 0.4729461967945099,
      "learning_rate": 8e-05,
      "loss": 1.5783,
      "step": 1817
    },
    {
      "epoch": 0.3775700934579439,
      "grad_norm": 0.4871976375579834,
      "learning_rate": 8e-05,
      "loss": 1.6477,
      "step": 1818
    },
    {
      "epoch": 0.37777777777777777,
      "grad_norm": 0.4819161593914032,
      "learning_rate": 8e-05,
      "loss": 1.5632,
      "step": 1819
    },
    {
      "epoch": 0.37798546209761164,
      "grad_norm": 0.48835310339927673,
      "learning_rate": 8e-05,
      "loss": 1.6809,
      "step": 1820
    },
    {
      "epoch": 0.3781931464174455,
      "grad_norm": 0.5206708908081055,
      "learning_rate": 8e-05,
      "loss": 1.6919,
      "step": 1821
    },
    {
      "epoch": 0.3784008307372793,
      "grad_norm": 0.49859198927879333,
      "learning_rate": 8e-05,
      "loss": 1.624,
      "step": 1822
    },
    {
      "epoch": 0.3786085150571132,
      "grad_norm": 0.47674262523651123,
      "learning_rate": 8e-05,
      "loss": 1.4634,
      "step": 1823
    },
    {
      "epoch": 0.37881619937694705,
      "grad_norm": 0.5091409683227539,
      "learning_rate": 8e-05,
      "loss": 1.5764,
      "step": 1824
    },
    {
      "epoch": 0.3790238836967809,
      "grad_norm": 0.4878379702568054,
      "learning_rate": 8e-05,
      "loss": 1.6147,
      "step": 1825
    },
    {
      "epoch": 0.37923156801661473,
      "grad_norm": 0.4939122200012207,
      "learning_rate": 8e-05,
      "loss": 1.5762,
      "step": 1826
    },
    {
      "epoch": 0.3794392523364486,
      "grad_norm": 0.5266447067260742,
      "learning_rate": 8e-05,
      "loss": 1.6838,
      "step": 1827
    },
    {
      "epoch": 0.37964693665628246,
      "grad_norm": 0.49246594309806824,
      "learning_rate": 8e-05,
      "loss": 1.561,
      "step": 1828
    },
    {
      "epoch": 0.37985462097611633,
      "grad_norm": 0.4985910952091217,
      "learning_rate": 8e-05,
      "loss": 1.4934,
      "step": 1829
    },
    {
      "epoch": 0.38006230529595014,
      "grad_norm": 0.49930235743522644,
      "learning_rate": 8e-05,
      "loss": 1.5857,
      "step": 1830
    },
    {
      "epoch": 0.380269989615784,
      "grad_norm": 0.5117544531822205,
      "learning_rate": 8e-05,
      "loss": 1.5451,
      "step": 1831
    },
    {
      "epoch": 0.3804776739356179,
      "grad_norm": 0.5053813457489014,
      "learning_rate": 8e-05,
      "loss": 1.5836,
      "step": 1832
    },
    {
      "epoch": 0.3806853582554517,
      "grad_norm": 0.5296281576156616,
      "learning_rate": 8e-05,
      "loss": 1.5972,
      "step": 1833
    },
    {
      "epoch": 0.38089304257528556,
      "grad_norm": 0.5012361407279968,
      "learning_rate": 8e-05,
      "loss": 1.5809,
      "step": 1834
    },
    {
      "epoch": 0.3811007268951194,
      "grad_norm": 0.5064777135848999,
      "learning_rate": 8e-05,
      "loss": 1.5584,
      "step": 1835
    },
    {
      "epoch": 0.3813084112149533,
      "grad_norm": 0.5177083611488342,
      "learning_rate": 8e-05,
      "loss": 1.5851,
      "step": 1836
    },
    {
      "epoch": 0.3815160955347871,
      "grad_norm": 0.5089267492294312,
      "learning_rate": 8e-05,
      "loss": 1.5429,
      "step": 1837
    },
    {
      "epoch": 0.38172377985462097,
      "grad_norm": 0.49458834528923035,
      "learning_rate": 8e-05,
      "loss": 1.6052,
      "step": 1838
    },
    {
      "epoch": 0.38193146417445484,
      "grad_norm": 0.4878547489643097,
      "learning_rate": 8e-05,
      "loss": 1.5777,
      "step": 1839
    },
    {
      "epoch": 0.3821391484942887,
      "grad_norm": 0.5592621564865112,
      "learning_rate": 8e-05,
      "loss": 1.6285,
      "step": 1840
    },
    {
      "epoch": 0.3823468328141225,
      "grad_norm": 0.5030962824821472,
      "learning_rate": 8e-05,
      "loss": 1.5928,
      "step": 1841
    },
    {
      "epoch": 0.3825545171339564,
      "grad_norm": 0.5183880925178528,
      "learning_rate": 8e-05,
      "loss": 1.6346,
      "step": 1842
    },
    {
      "epoch": 0.38276220145379025,
      "grad_norm": 0.4975292980670929,
      "learning_rate": 8e-05,
      "loss": 1.5236,
      "step": 1843
    },
    {
      "epoch": 0.3829698857736241,
      "grad_norm": 0.5079745650291443,
      "learning_rate": 8e-05,
      "loss": 1.5622,
      "step": 1844
    },
    {
      "epoch": 0.38317757009345793,
      "grad_norm": 0.49533551931381226,
      "learning_rate": 8e-05,
      "loss": 1.563,
      "step": 1845
    },
    {
      "epoch": 0.3833852544132918,
      "grad_norm": 0.5062286257743835,
      "learning_rate": 8e-05,
      "loss": 1.6209,
      "step": 1846
    },
    {
      "epoch": 0.38359293873312567,
      "grad_norm": 0.49643903970718384,
      "learning_rate": 8e-05,
      "loss": 1.6075,
      "step": 1847
    },
    {
      "epoch": 0.3838006230529595,
      "grad_norm": 0.6108715534210205,
      "learning_rate": 8e-05,
      "loss": 1.6111,
      "step": 1848
    },
    {
      "epoch": 0.38400830737279334,
      "grad_norm": 0.5051819682121277,
      "learning_rate": 8e-05,
      "loss": 1.5039,
      "step": 1849
    },
    {
      "epoch": 0.3842159916926272,
      "grad_norm": 0.5083017945289612,
      "learning_rate": 8e-05,
      "loss": 1.5674,
      "step": 1850
    },
    {
      "epoch": 0.3844236760124611,
      "grad_norm": 0.5128228068351746,
      "learning_rate": 8e-05,
      "loss": 1.6137,
      "step": 1851
    },
    {
      "epoch": 0.3846313603322949,
      "grad_norm": 0.5230603218078613,
      "learning_rate": 8e-05,
      "loss": 1.6066,
      "step": 1852
    },
    {
      "epoch": 0.38483904465212876,
      "grad_norm": 0.527700662612915,
      "learning_rate": 8e-05,
      "loss": 1.6253,
      "step": 1853
    },
    {
      "epoch": 0.3850467289719626,
      "grad_norm": 0.567374050617218,
      "learning_rate": 8e-05,
      "loss": 1.5531,
      "step": 1854
    },
    {
      "epoch": 0.3852544132917965,
      "grad_norm": 0.4868682622909546,
      "learning_rate": 8e-05,
      "loss": 1.6318,
      "step": 1855
    },
    {
      "epoch": 0.3854620976116303,
      "grad_norm": 0.5227965116500854,
      "learning_rate": 8e-05,
      "loss": 1.5589,
      "step": 1856
    },
    {
      "epoch": 0.3856697819314642,
      "grad_norm": 0.5066382884979248,
      "learning_rate": 8e-05,
      "loss": 1.6554,
      "step": 1857
    },
    {
      "epoch": 0.38587746625129804,
      "grad_norm": 0.49491891264915466,
      "learning_rate": 8e-05,
      "loss": 1.6015,
      "step": 1858
    },
    {
      "epoch": 0.38608515057113185,
      "grad_norm": 0.5157635807991028,
      "learning_rate": 8e-05,
      "loss": 1.6216,
      "step": 1859
    },
    {
      "epoch": 0.3862928348909657,
      "grad_norm": 0.5314090847969055,
      "learning_rate": 8e-05,
      "loss": 1.5958,
      "step": 1860
    },
    {
      "epoch": 0.3865005192107996,
      "grad_norm": 0.48347991704940796,
      "learning_rate": 8e-05,
      "loss": 1.5142,
      "step": 1861
    },
    {
      "epoch": 0.38670820353063345,
      "grad_norm": 0.5229157209396362,
      "learning_rate": 8e-05,
      "loss": 1.5493,
      "step": 1862
    },
    {
      "epoch": 0.38691588785046727,
      "grad_norm": 0.5444154739379883,
      "learning_rate": 8e-05,
      "loss": 1.6436,
      "step": 1863
    },
    {
      "epoch": 0.38712357217030113,
      "grad_norm": 0.5127759575843811,
      "learning_rate": 8e-05,
      "loss": 1.6032,
      "step": 1864
    },
    {
      "epoch": 0.387331256490135,
      "grad_norm": 0.4927530884742737,
      "learning_rate": 8e-05,
      "loss": 1.6051,
      "step": 1865
    },
    {
      "epoch": 0.38753894080996887,
      "grad_norm": 0.5002699494361877,
      "learning_rate": 8e-05,
      "loss": 1.6294,
      "step": 1866
    },
    {
      "epoch": 0.3877466251298027,
      "grad_norm": 0.5165937542915344,
      "learning_rate": 8e-05,
      "loss": 1.5966,
      "step": 1867
    },
    {
      "epoch": 0.38795430944963655,
      "grad_norm": 0.5008817315101624,
      "learning_rate": 8e-05,
      "loss": 1.5651,
      "step": 1868
    },
    {
      "epoch": 0.3881619937694704,
      "grad_norm": 0.49471011757850647,
      "learning_rate": 8e-05,
      "loss": 1.5699,
      "step": 1869
    },
    {
      "epoch": 0.3883696780893043,
      "grad_norm": 0.5040736198425293,
      "learning_rate": 8e-05,
      "loss": 1.6638,
      "step": 1870
    },
    {
      "epoch": 0.3885773624091381,
      "grad_norm": 0.4965846538543701,
      "learning_rate": 8e-05,
      "loss": 1.5604,
      "step": 1871
    },
    {
      "epoch": 0.38878504672897196,
      "grad_norm": 0.48028695583343506,
      "learning_rate": 8e-05,
      "loss": 1.5656,
      "step": 1872
    },
    {
      "epoch": 0.38899273104880583,
      "grad_norm": 0.506149411201477,
      "learning_rate": 8e-05,
      "loss": 1.5405,
      "step": 1873
    },
    {
      "epoch": 0.38920041536863964,
      "grad_norm": 0.4821113049983978,
      "learning_rate": 8e-05,
      "loss": 1.6343,
      "step": 1874
    },
    {
      "epoch": 0.3894080996884735,
      "grad_norm": 0.545812726020813,
      "learning_rate": 8e-05,
      "loss": 1.6731,
      "step": 1875
    },
    {
      "epoch": 0.3896157840083074,
      "grad_norm": 0.5006311535835266,
      "learning_rate": 8e-05,
      "loss": 1.5129,
      "step": 1876
    },
    {
      "epoch": 0.38982346832814124,
      "grad_norm": 0.48582524061203003,
      "learning_rate": 8e-05,
      "loss": 1.5433,
      "step": 1877
    },
    {
      "epoch": 0.39003115264797505,
      "grad_norm": 0.5167272090911865,
      "learning_rate": 8e-05,
      "loss": 1.6357,
      "step": 1878
    },
    {
      "epoch": 0.3902388369678089,
      "grad_norm": 0.5142748951911926,
      "learning_rate": 8e-05,
      "loss": 1.6098,
      "step": 1879
    },
    {
      "epoch": 0.3904465212876428,
      "grad_norm": 0.5022456645965576,
      "learning_rate": 8e-05,
      "loss": 1.5784,
      "step": 1880
    },
    {
      "epoch": 0.39065420560747666,
      "grad_norm": 0.5111859440803528,
      "learning_rate": 8e-05,
      "loss": 1.5819,
      "step": 1881
    },
    {
      "epoch": 0.39086188992731047,
      "grad_norm": 0.49327921867370605,
      "learning_rate": 8e-05,
      "loss": 1.5932,
      "step": 1882
    },
    {
      "epoch": 0.39106957424714434,
      "grad_norm": 0.5264003872871399,
      "learning_rate": 8e-05,
      "loss": 1.615,
      "step": 1883
    },
    {
      "epoch": 0.3912772585669782,
      "grad_norm": 0.4997003376483917,
      "learning_rate": 8e-05,
      "loss": 1.5645,
      "step": 1884
    },
    {
      "epoch": 0.39148494288681207,
      "grad_norm": 0.48707181215286255,
      "learning_rate": 8e-05,
      "loss": 1.566,
      "step": 1885
    },
    {
      "epoch": 0.3916926272066459,
      "grad_norm": 0.5190602540969849,
      "learning_rate": 8e-05,
      "loss": 1.5918,
      "step": 1886
    },
    {
      "epoch": 0.39190031152647975,
      "grad_norm": 0.5049939155578613,
      "learning_rate": 8e-05,
      "loss": 1.5857,
      "step": 1887
    },
    {
      "epoch": 0.3921079958463136,
      "grad_norm": 0.5178766250610352,
      "learning_rate": 8e-05,
      "loss": 1.6534,
      "step": 1888
    },
    {
      "epoch": 0.39231568016614743,
      "grad_norm": 0.510301411151886,
      "learning_rate": 8e-05,
      "loss": 1.6123,
      "step": 1889
    },
    {
      "epoch": 0.3925233644859813,
      "grad_norm": 0.48783424496650696,
      "learning_rate": 8e-05,
      "loss": 1.51,
      "step": 1890
    },
    {
      "epoch": 0.39273104880581516,
      "grad_norm": 0.5162112712860107,
      "learning_rate": 8e-05,
      "loss": 1.5571,
      "step": 1891
    },
    {
      "epoch": 0.39293873312564903,
      "grad_norm": 0.5092341303825378,
      "learning_rate": 8e-05,
      "loss": 1.6414,
      "step": 1892
    },
    {
      "epoch": 0.39314641744548284,
      "grad_norm": 0.5020235776901245,
      "learning_rate": 8e-05,
      "loss": 1.626,
      "step": 1893
    },
    {
      "epoch": 0.3933541017653167,
      "grad_norm": 0.5041497945785522,
      "learning_rate": 8e-05,
      "loss": 1.5922,
      "step": 1894
    },
    {
      "epoch": 0.3935617860851506,
      "grad_norm": 0.5252364277839661,
      "learning_rate": 8e-05,
      "loss": 1.6367,
      "step": 1895
    },
    {
      "epoch": 0.39376947040498445,
      "grad_norm": 0.5267825126647949,
      "learning_rate": 8e-05,
      "loss": 1.6483,
      "step": 1896
    },
    {
      "epoch": 0.39397715472481826,
      "grad_norm": 0.47401174902915955,
      "learning_rate": 8e-05,
      "loss": 1.4967,
      "step": 1897
    },
    {
      "epoch": 0.3941848390446521,
      "grad_norm": 0.47972801327705383,
      "learning_rate": 8e-05,
      "loss": 1.5402,
      "step": 1898
    },
    {
      "epoch": 0.394392523364486,
      "grad_norm": 0.5107938051223755,
      "learning_rate": 8e-05,
      "loss": 1.6526,
      "step": 1899
    },
    {
      "epoch": 0.39460020768431986,
      "grad_norm": 0.4895242154598236,
      "learning_rate": 8e-05,
      "loss": 1.5815,
      "step": 1900
    },
    {
      "epoch": 0.39480789200415367,
      "grad_norm": 0.5084724426269531,
      "learning_rate": 8e-05,
      "loss": 1.5669,
      "step": 1901
    },
    {
      "epoch": 0.39501557632398754,
      "grad_norm": 0.4843336045742035,
      "learning_rate": 8e-05,
      "loss": 1.5556,
      "step": 1902
    },
    {
      "epoch": 0.3952232606438214,
      "grad_norm": 0.5088340044021606,
      "learning_rate": 8e-05,
      "loss": 1.5571,
      "step": 1903
    },
    {
      "epoch": 0.3954309449636552,
      "grad_norm": 0.5239298939704895,
      "learning_rate": 8e-05,
      "loss": 1.6224,
      "step": 1904
    },
    {
      "epoch": 0.3956386292834891,
      "grad_norm": 0.49526721239089966,
      "learning_rate": 8e-05,
      "loss": 1.6317,
      "step": 1905
    },
    {
      "epoch": 0.39584631360332295,
      "grad_norm": 0.5118904113769531,
      "learning_rate": 8e-05,
      "loss": 1.6292,
      "step": 1906
    },
    {
      "epoch": 0.3960539979231568,
      "grad_norm": 0.5064014196395874,
      "learning_rate": 8e-05,
      "loss": 1.6595,
      "step": 1907
    },
    {
      "epoch": 0.39626168224299063,
      "grad_norm": 0.5004372000694275,
      "learning_rate": 8e-05,
      "loss": 1.5685,
      "step": 1908
    },
    {
      "epoch": 0.3964693665628245,
      "grad_norm": 0.5007834434509277,
      "learning_rate": 8e-05,
      "loss": 1.5655,
      "step": 1909
    },
    {
      "epoch": 0.39667705088265837,
      "grad_norm": 0.4961642920970917,
      "learning_rate": 8e-05,
      "loss": 1.5386,
      "step": 1910
    },
    {
      "epoch": 0.39688473520249223,
      "grad_norm": 0.49714717268943787,
      "learning_rate": 8e-05,
      "loss": 1.5585,
      "step": 1911
    },
    {
      "epoch": 0.39709241952232605,
      "grad_norm": 0.5007422566413879,
      "learning_rate": 8e-05,
      "loss": 1.5251,
      "step": 1912
    },
    {
      "epoch": 0.3973001038421599,
      "grad_norm": 0.4922661781311035,
      "learning_rate": 8e-05,
      "loss": 1.5312,
      "step": 1913
    },
    {
      "epoch": 0.3975077881619938,
      "grad_norm": 0.5031492710113525,
      "learning_rate": 8e-05,
      "loss": 1.6502,
      "step": 1914
    },
    {
      "epoch": 0.39771547248182765,
      "grad_norm": 0.4995499551296234,
      "learning_rate": 8e-05,
      "loss": 1.5721,
      "step": 1915
    },
    {
      "epoch": 0.39792315680166146,
      "grad_norm": 0.5230512022972107,
      "learning_rate": 8e-05,
      "loss": 1.6209,
      "step": 1916
    },
    {
      "epoch": 0.3981308411214953,
      "grad_norm": 0.4955296516418457,
      "learning_rate": 8e-05,
      "loss": 1.6216,
      "step": 1917
    },
    {
      "epoch": 0.3983385254413292,
      "grad_norm": 0.49905237555503845,
      "learning_rate": 8e-05,
      "loss": 1.6325,
      "step": 1918
    },
    {
      "epoch": 0.398546209761163,
      "grad_norm": 0.49909424781799316,
      "learning_rate": 8e-05,
      "loss": 1.6182,
      "step": 1919
    },
    {
      "epoch": 0.3987538940809969,
      "grad_norm": 0.5071685910224915,
      "learning_rate": 8e-05,
      "loss": 1.6344,
      "step": 1920
    },
    {
      "epoch": 0.39896157840083074,
      "grad_norm": 0.4874131977558136,
      "learning_rate": 8e-05,
      "loss": 1.6058,
      "step": 1921
    },
    {
      "epoch": 0.3991692627206646,
      "grad_norm": 0.4769895374774933,
      "learning_rate": 8e-05,
      "loss": 1.4976,
      "step": 1922
    },
    {
      "epoch": 0.3993769470404984,
      "grad_norm": 0.47417256236076355,
      "learning_rate": 8e-05,
      "loss": 1.4958,
      "step": 1923
    },
    {
      "epoch": 0.3995846313603323,
      "grad_norm": 0.5021626353263855,
      "learning_rate": 8e-05,
      "loss": 1.6307,
      "step": 1924
    },
    {
      "epoch": 0.39979231568016615,
      "grad_norm": 0.48706451058387756,
      "learning_rate": 8e-05,
      "loss": 1.5546,
      "step": 1925
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.5021299719810486,
      "learning_rate": 8e-05,
      "loss": 1.6139,
      "step": 1926
    },
    {
      "epoch": 0.40020768431983383,
      "grad_norm": 0.4878026247024536,
      "learning_rate": 8e-05,
      "loss": 1.565,
      "step": 1927
    },
    {
      "epoch": 0.4004153686396677,
      "grad_norm": 0.4966025650501251,
      "learning_rate": 8e-05,
      "loss": 1.5577,
      "step": 1928
    },
    {
      "epoch": 0.40062305295950157,
      "grad_norm": 0.49730899930000305,
      "learning_rate": 8e-05,
      "loss": 1.5656,
      "step": 1929
    },
    {
      "epoch": 0.40083073727933544,
      "grad_norm": 0.5028313994407654,
      "learning_rate": 8e-05,
      "loss": 1.6599,
      "step": 1930
    },
    {
      "epoch": 0.40103842159916925,
      "grad_norm": 0.5250521898269653,
      "learning_rate": 8e-05,
      "loss": 1.5481,
      "step": 1931
    },
    {
      "epoch": 0.4012461059190031,
      "grad_norm": 0.5033625364303589,
      "learning_rate": 8e-05,
      "loss": 1.5811,
      "step": 1932
    },
    {
      "epoch": 0.401453790238837,
      "grad_norm": 0.4944176971912384,
      "learning_rate": 8e-05,
      "loss": 1.619,
      "step": 1933
    },
    {
      "epoch": 0.4016614745586708,
      "grad_norm": 0.5117709040641785,
      "learning_rate": 8e-05,
      "loss": 1.5833,
      "step": 1934
    },
    {
      "epoch": 0.40186915887850466,
      "grad_norm": 0.50387042760849,
      "learning_rate": 8e-05,
      "loss": 1.5767,
      "step": 1935
    },
    {
      "epoch": 0.40207684319833853,
      "grad_norm": 0.5241379141807556,
      "learning_rate": 8e-05,
      "loss": 1.5629,
      "step": 1936
    },
    {
      "epoch": 0.4022845275181724,
      "grad_norm": 0.5203877687454224,
      "learning_rate": 8e-05,
      "loss": 1.5862,
      "step": 1937
    },
    {
      "epoch": 0.4024922118380062,
      "grad_norm": 0.49413588643074036,
      "learning_rate": 8e-05,
      "loss": 1.5689,
      "step": 1938
    },
    {
      "epoch": 0.4026998961578401,
      "grad_norm": 0.5159587860107422,
      "learning_rate": 8e-05,
      "loss": 1.6136,
      "step": 1939
    },
    {
      "epoch": 0.40290758047767394,
      "grad_norm": 0.5072512626647949,
      "learning_rate": 8e-05,
      "loss": 1.5555,
      "step": 1940
    },
    {
      "epoch": 0.4031152647975078,
      "grad_norm": 0.500592052936554,
      "learning_rate": 8e-05,
      "loss": 1.5846,
      "step": 1941
    },
    {
      "epoch": 0.4033229491173416,
      "grad_norm": 0.4672267735004425,
      "learning_rate": 8e-05,
      "loss": 1.5187,
      "step": 1942
    },
    {
      "epoch": 0.4035306334371755,
      "grad_norm": 0.48535841703414917,
      "learning_rate": 8e-05,
      "loss": 1.6116,
      "step": 1943
    },
    {
      "epoch": 0.40373831775700936,
      "grad_norm": 0.48251843452453613,
      "learning_rate": 8e-05,
      "loss": 1.5747,
      "step": 1944
    },
    {
      "epoch": 0.4039460020768432,
      "grad_norm": 0.4987678825855255,
      "learning_rate": 8e-05,
      "loss": 1.5857,
      "step": 1945
    },
    {
      "epoch": 0.40415368639667704,
      "grad_norm": 0.5083807110786438,
      "learning_rate": 8e-05,
      "loss": 1.5437,
      "step": 1946
    },
    {
      "epoch": 0.4043613707165109,
      "grad_norm": 0.5144785642623901,
      "learning_rate": 8e-05,
      "loss": 1.6153,
      "step": 1947
    },
    {
      "epoch": 0.40456905503634477,
      "grad_norm": 0.4901933968067169,
      "learning_rate": 8e-05,
      "loss": 1.5634,
      "step": 1948
    },
    {
      "epoch": 0.4047767393561786,
      "grad_norm": 0.47113847732543945,
      "learning_rate": 8e-05,
      "loss": 1.5647,
      "step": 1949
    },
    {
      "epoch": 0.40498442367601245,
      "grad_norm": 0.4938381314277649,
      "learning_rate": 8e-05,
      "loss": 1.5668,
      "step": 1950
    },
    {
      "epoch": 0.4051921079958463,
      "grad_norm": 0.5108892321586609,
      "learning_rate": 8e-05,
      "loss": 1.6338,
      "step": 1951
    },
    {
      "epoch": 0.4053997923156802,
      "grad_norm": 0.46494653820991516,
      "learning_rate": 8e-05,
      "loss": 1.4997,
      "step": 1952
    },
    {
      "epoch": 0.405607476635514,
      "grad_norm": 0.48271963000297546,
      "learning_rate": 8e-05,
      "loss": 1.5029,
      "step": 1953
    },
    {
      "epoch": 0.40581516095534786,
      "grad_norm": 0.5015127658843994,
      "learning_rate": 8e-05,
      "loss": 1.5271,
      "step": 1954
    },
    {
      "epoch": 0.40602284527518173,
      "grad_norm": 0.5078144669532776,
      "learning_rate": 8e-05,
      "loss": 1.5529,
      "step": 1955
    },
    {
      "epoch": 0.4062305295950156,
      "grad_norm": 0.5320901870727539,
      "learning_rate": 8e-05,
      "loss": 1.6427,
      "step": 1956
    },
    {
      "epoch": 0.4064382139148494,
      "grad_norm": 0.501735270023346,
      "learning_rate": 8e-05,
      "loss": 1.5977,
      "step": 1957
    },
    {
      "epoch": 0.4066458982346833,
      "grad_norm": 0.4956776201725006,
      "learning_rate": 8e-05,
      "loss": 1.5942,
      "step": 1958
    },
    {
      "epoch": 0.40685358255451715,
      "grad_norm": 0.48016679286956787,
      "learning_rate": 8e-05,
      "loss": 1.5697,
      "step": 1959
    },
    {
      "epoch": 0.407061266874351,
      "grad_norm": 0.500160813331604,
      "learning_rate": 8e-05,
      "loss": 1.5724,
      "step": 1960
    },
    {
      "epoch": 0.4072689511941848,
      "grad_norm": 0.5022308230400085,
      "learning_rate": 8e-05,
      "loss": 1.5612,
      "step": 1961
    },
    {
      "epoch": 0.4074766355140187,
      "grad_norm": 0.5192114114761353,
      "learning_rate": 8e-05,
      "loss": 1.6205,
      "step": 1962
    },
    {
      "epoch": 0.40768431983385256,
      "grad_norm": 0.470253050327301,
      "learning_rate": 8e-05,
      "loss": 1.5219,
      "step": 1963
    },
    {
      "epoch": 0.40789200415368637,
      "grad_norm": 0.48591867089271545,
      "learning_rate": 8e-05,
      "loss": 1.6203,
      "step": 1964
    },
    {
      "epoch": 0.40809968847352024,
      "grad_norm": 0.49583899974823,
      "learning_rate": 8e-05,
      "loss": 1.548,
      "step": 1965
    },
    {
      "epoch": 0.4083073727933541,
      "grad_norm": 0.5171915292739868,
      "learning_rate": 8e-05,
      "loss": 1.6081,
      "step": 1966
    },
    {
      "epoch": 0.408515057113188,
      "grad_norm": 0.49792858958244324,
      "learning_rate": 8e-05,
      "loss": 1.6061,
      "step": 1967
    },
    {
      "epoch": 0.4087227414330218,
      "grad_norm": 0.4974616467952728,
      "learning_rate": 8e-05,
      "loss": 1.6322,
      "step": 1968
    },
    {
      "epoch": 0.40893042575285565,
      "grad_norm": 0.5105950832366943,
      "learning_rate": 8e-05,
      "loss": 1.5612,
      "step": 1969
    },
    {
      "epoch": 0.4091381100726895,
      "grad_norm": 0.49406614899635315,
      "learning_rate": 8e-05,
      "loss": 1.5887,
      "step": 1970
    },
    {
      "epoch": 0.4093457943925234,
      "grad_norm": 0.4994363784790039,
      "learning_rate": 8e-05,
      "loss": 1.6257,
      "step": 1971
    },
    {
      "epoch": 0.4095534787123572,
      "grad_norm": 0.5137128233909607,
      "learning_rate": 8e-05,
      "loss": 1.5861,
      "step": 1972
    },
    {
      "epoch": 0.40976116303219107,
      "grad_norm": 0.47833117842674255,
      "learning_rate": 8e-05,
      "loss": 1.5661,
      "step": 1973
    },
    {
      "epoch": 0.40996884735202493,
      "grad_norm": 0.5011354684829712,
      "learning_rate": 8e-05,
      "loss": 1.6508,
      "step": 1974
    },
    {
      "epoch": 0.4101765316718588,
      "grad_norm": 0.49250781536102295,
      "learning_rate": 8e-05,
      "loss": 1.587,
      "step": 1975
    },
    {
      "epoch": 0.4103842159916926,
      "grad_norm": 0.5081549882888794,
      "learning_rate": 8e-05,
      "loss": 1.5634,
      "step": 1976
    },
    {
      "epoch": 0.4105919003115265,
      "grad_norm": 0.503896176815033,
      "learning_rate": 8e-05,
      "loss": 1.5714,
      "step": 1977
    },
    {
      "epoch": 0.41079958463136035,
      "grad_norm": 0.5175586938858032,
      "learning_rate": 8e-05,
      "loss": 1.5938,
      "step": 1978
    },
    {
      "epoch": 0.41100726895119416,
      "grad_norm": 0.4877651631832123,
      "learning_rate": 8e-05,
      "loss": 1.5255,
      "step": 1979
    },
    {
      "epoch": 0.411214953271028,
      "grad_norm": 0.5077330470085144,
      "learning_rate": 8e-05,
      "loss": 1.5783,
      "step": 1980
    },
    {
      "epoch": 0.4114226375908619,
      "grad_norm": 0.48941168189048767,
      "learning_rate": 8e-05,
      "loss": 1.6246,
      "step": 1981
    },
    {
      "epoch": 0.41163032191069576,
      "grad_norm": 0.4887639582157135,
      "learning_rate": 8e-05,
      "loss": 1.581,
      "step": 1982
    },
    {
      "epoch": 0.4118380062305296,
      "grad_norm": 0.48687008023262024,
      "learning_rate": 8e-05,
      "loss": 1.5368,
      "step": 1983
    },
    {
      "epoch": 0.41204569055036344,
      "grad_norm": 0.4982350468635559,
      "learning_rate": 8e-05,
      "loss": 1.6293,
      "step": 1984
    },
    {
      "epoch": 0.4122533748701973,
      "grad_norm": 0.48353442549705505,
      "learning_rate": 8e-05,
      "loss": 1.5929,
      "step": 1985
    },
    {
      "epoch": 0.4124610591900312,
      "grad_norm": 0.47105276584625244,
      "learning_rate": 8e-05,
      "loss": 1.5741,
      "step": 1986
    },
    {
      "epoch": 0.412668743509865,
      "grad_norm": 0.49714410305023193,
      "learning_rate": 8e-05,
      "loss": 1.5607,
      "step": 1987
    },
    {
      "epoch": 0.41287642782969886,
      "grad_norm": 0.4911838173866272,
      "learning_rate": 8e-05,
      "loss": 1.5755,
      "step": 1988
    },
    {
      "epoch": 0.4130841121495327,
      "grad_norm": 0.47256773710250854,
      "learning_rate": 8e-05,
      "loss": 1.4882,
      "step": 1989
    },
    {
      "epoch": 0.4132917964693666,
      "grad_norm": 0.5255026817321777,
      "learning_rate": 8e-05,
      "loss": 1.586,
      "step": 1990
    },
    {
      "epoch": 0.4134994807892004,
      "grad_norm": 0.48072895407676697,
      "learning_rate": 8e-05,
      "loss": 1.541,
      "step": 1991
    },
    {
      "epoch": 0.41370716510903427,
      "grad_norm": 0.5017218589782715,
      "learning_rate": 8e-05,
      "loss": 1.6442,
      "step": 1992
    },
    {
      "epoch": 0.41391484942886814,
      "grad_norm": 0.49725833535194397,
      "learning_rate": 8e-05,
      "loss": 1.6044,
      "step": 1993
    },
    {
      "epoch": 0.41412253374870195,
      "grad_norm": 0.4846404790878296,
      "learning_rate": 8e-05,
      "loss": 1.5728,
      "step": 1994
    },
    {
      "epoch": 0.4143302180685358,
      "grad_norm": 0.5051259398460388,
      "learning_rate": 8e-05,
      "loss": 1.5835,
      "step": 1995
    },
    {
      "epoch": 0.4145379023883697,
      "grad_norm": 0.504957377910614,
      "learning_rate": 8e-05,
      "loss": 1.5677,
      "step": 1996
    },
    {
      "epoch": 0.41474558670820355,
      "grad_norm": 0.5060474276542664,
      "learning_rate": 8e-05,
      "loss": 1.5864,
      "step": 1997
    },
    {
      "epoch": 0.41495327102803736,
      "grad_norm": 0.5011793375015259,
      "learning_rate": 8e-05,
      "loss": 1.6073,
      "step": 1998
    },
    {
      "epoch": 0.41516095534787123,
      "grad_norm": 0.5134116411209106,
      "learning_rate": 8e-05,
      "loss": 1.59,
      "step": 1999
    },
    {
      "epoch": 0.4153686396677051,
      "grad_norm": 0.5138752460479736,
      "learning_rate": 8e-05,
      "loss": 1.5954,
      "step": 2000
    },
    {
      "epoch": 0.41557632398753896,
      "grad_norm": 0.4684045612812042,
      "learning_rate": 8e-05,
      "loss": 1.4873,
      "step": 2001
    },
    {
      "epoch": 0.4157840083073728,
      "grad_norm": 0.4988643527030945,
      "learning_rate": 8e-05,
      "loss": 1.5767,
      "step": 2002
    },
    {
      "epoch": 0.41599169262720664,
      "grad_norm": 0.4997043311595917,
      "learning_rate": 8e-05,
      "loss": 1.5523,
      "step": 2003
    },
    {
      "epoch": 0.4161993769470405,
      "grad_norm": 0.4968480169773102,
      "learning_rate": 8e-05,
      "loss": 1.5646,
      "step": 2004
    },
    {
      "epoch": 0.4164070612668744,
      "grad_norm": 0.49053269624710083,
      "learning_rate": 8e-05,
      "loss": 1.5753,
      "step": 2005
    },
    {
      "epoch": 0.4166147455867082,
      "grad_norm": 0.511942982673645,
      "learning_rate": 8e-05,
      "loss": 1.5919,
      "step": 2006
    },
    {
      "epoch": 0.41682242990654206,
      "grad_norm": 0.4872995913028717,
      "learning_rate": 8e-05,
      "loss": 1.5418,
      "step": 2007
    },
    {
      "epoch": 0.4170301142263759,
      "grad_norm": 0.4999676048755646,
      "learning_rate": 8e-05,
      "loss": 1.5057,
      "step": 2008
    },
    {
      "epoch": 0.41723779854620974,
      "grad_norm": 0.5027379989624023,
      "learning_rate": 8e-05,
      "loss": 1.5947,
      "step": 2009
    },
    {
      "epoch": 0.4174454828660436,
      "grad_norm": 0.5082154870033264,
      "learning_rate": 8e-05,
      "loss": 1.5025,
      "step": 2010
    },
    {
      "epoch": 0.41765316718587747,
      "grad_norm": 0.5095928907394409,
      "learning_rate": 8e-05,
      "loss": 1.6436,
      "step": 2011
    },
    {
      "epoch": 0.41786085150571134,
      "grad_norm": 0.5015538930892944,
      "learning_rate": 8e-05,
      "loss": 1.6064,
      "step": 2012
    },
    {
      "epoch": 0.41806853582554515,
      "grad_norm": 0.5190447568893433,
      "learning_rate": 8e-05,
      "loss": 1.6239,
      "step": 2013
    },
    {
      "epoch": 0.418276220145379,
      "grad_norm": 0.5272454023361206,
      "learning_rate": 8e-05,
      "loss": 1.5599,
      "step": 2014
    },
    {
      "epoch": 0.4184839044652129,
      "grad_norm": 0.4809977412223816,
      "learning_rate": 8e-05,
      "loss": 1.5921,
      "step": 2015
    },
    {
      "epoch": 0.41869158878504675,
      "grad_norm": 0.5302486419677734,
      "learning_rate": 8e-05,
      "loss": 1.6166,
      "step": 2016
    },
    {
      "epoch": 0.41889927310488057,
      "grad_norm": 0.5640528202056885,
      "learning_rate": 8e-05,
      "loss": 1.5801,
      "step": 2017
    },
    {
      "epoch": 0.41910695742471443,
      "grad_norm": 0.47891366481781006,
      "learning_rate": 8e-05,
      "loss": 1.5587,
      "step": 2018
    },
    {
      "epoch": 0.4193146417445483,
      "grad_norm": 0.5397331714630127,
      "learning_rate": 8e-05,
      "loss": 1.5776,
      "step": 2019
    },
    {
      "epoch": 0.4195223260643821,
      "grad_norm": 0.5956094861030579,
      "learning_rate": 8e-05,
      "loss": 1.6211,
      "step": 2020
    },
    {
      "epoch": 0.419730010384216,
      "grad_norm": 0.5107740163803101,
      "learning_rate": 8e-05,
      "loss": 1.5543,
      "step": 2021
    },
    {
      "epoch": 0.41993769470404985,
      "grad_norm": 0.5048316717147827,
      "learning_rate": 8e-05,
      "loss": 1.5948,
      "step": 2022
    },
    {
      "epoch": 0.4201453790238837,
      "grad_norm": 0.526158332824707,
      "learning_rate": 8e-05,
      "loss": 1.5892,
      "step": 2023
    },
    {
      "epoch": 0.4203530633437175,
      "grad_norm": 0.5123574733734131,
      "learning_rate": 8e-05,
      "loss": 1.5103,
      "step": 2024
    },
    {
      "epoch": 0.4205607476635514,
      "grad_norm": 0.5256283283233643,
      "learning_rate": 8e-05,
      "loss": 1.6267,
      "step": 2025
    },
    {
      "epoch": 0.42076843198338526,
      "grad_norm": 0.5114898681640625,
      "learning_rate": 8e-05,
      "loss": 1.6206,
      "step": 2026
    },
    {
      "epoch": 0.42097611630321913,
      "grad_norm": 0.5819741487503052,
      "learning_rate": 8e-05,
      "loss": 1.5611,
      "step": 2027
    },
    {
      "epoch": 0.42118380062305294,
      "grad_norm": 0.5962750315666199,
      "learning_rate": 8e-05,
      "loss": 1.5,
      "step": 2028
    },
    {
      "epoch": 0.4213914849428868,
      "grad_norm": 0.5251797437667847,
      "learning_rate": 8e-05,
      "loss": 1.6389,
      "step": 2029
    },
    {
      "epoch": 0.4215991692627207,
      "grad_norm": 0.48249077796936035,
      "learning_rate": 8e-05,
      "loss": 1.5256,
      "step": 2030
    },
    {
      "epoch": 0.42180685358255454,
      "grad_norm": 0.5865249633789062,
      "learning_rate": 8e-05,
      "loss": 1.5666,
      "step": 2031
    },
    {
      "epoch": 0.42201453790238835,
      "grad_norm": 0.5307789444923401,
      "learning_rate": 8e-05,
      "loss": 1.5807,
      "step": 2032
    },
    {
      "epoch": 0.4222222222222222,
      "grad_norm": 0.5467637777328491,
      "learning_rate": 8e-05,
      "loss": 1.6559,
      "step": 2033
    },
    {
      "epoch": 0.4224299065420561,
      "grad_norm": 0.4931851029396057,
      "learning_rate": 8e-05,
      "loss": 1.5186,
      "step": 2034
    },
    {
      "epoch": 0.4226375908618899,
      "grad_norm": 0.5111013650894165,
      "learning_rate": 8e-05,
      "loss": 1.5595,
      "step": 2035
    },
    {
      "epoch": 0.42284527518172377,
      "grad_norm": 0.5746096968650818,
      "learning_rate": 8e-05,
      "loss": 1.5242,
      "step": 2036
    },
    {
      "epoch": 0.42305295950155763,
      "grad_norm": 0.49678611755371094,
      "learning_rate": 8e-05,
      "loss": 1.5794,
      "step": 2037
    },
    {
      "epoch": 0.4232606438213915,
      "grad_norm": 0.5084017515182495,
      "learning_rate": 8e-05,
      "loss": 1.5751,
      "step": 2038
    },
    {
      "epoch": 0.4234683281412253,
      "grad_norm": 0.48849454522132874,
      "learning_rate": 8e-05,
      "loss": 1.521,
      "step": 2039
    },
    {
      "epoch": 0.4236760124610592,
      "grad_norm": 0.5270300507545471,
      "learning_rate": 8e-05,
      "loss": 1.6187,
      "step": 2040
    },
    {
      "epoch": 0.42388369678089305,
      "grad_norm": 0.5154289603233337,
      "learning_rate": 8e-05,
      "loss": 1.5474,
      "step": 2041
    },
    {
      "epoch": 0.4240913811007269,
      "grad_norm": 0.4959946870803833,
      "learning_rate": 8e-05,
      "loss": 1.5549,
      "step": 2042
    },
    {
      "epoch": 0.42429906542056073,
      "grad_norm": 0.49153828620910645,
      "learning_rate": 8e-05,
      "loss": 1.5627,
      "step": 2043
    },
    {
      "epoch": 0.4245067497403946,
      "grad_norm": 0.504295289516449,
      "learning_rate": 8e-05,
      "loss": 1.5422,
      "step": 2044
    },
    {
      "epoch": 0.42471443406022846,
      "grad_norm": 0.4918515384197235,
      "learning_rate": 8e-05,
      "loss": 1.6011,
      "step": 2045
    },
    {
      "epoch": 0.42492211838006233,
      "grad_norm": 0.5110564827919006,
      "learning_rate": 8e-05,
      "loss": 1.5277,
      "step": 2046
    },
    {
      "epoch": 0.42512980269989614,
      "grad_norm": 0.49453458189964294,
      "learning_rate": 8e-05,
      "loss": 1.5887,
      "step": 2047
    },
    {
      "epoch": 0.42533748701973,
      "grad_norm": 0.49117112159729004,
      "learning_rate": 8e-05,
      "loss": 1.5371,
      "step": 2048
    },
    {
      "epoch": 0.4255451713395639,
      "grad_norm": 0.5056717991828918,
      "learning_rate": 8e-05,
      "loss": 1.6292,
      "step": 2049
    },
    {
      "epoch": 0.4257528556593977,
      "grad_norm": 0.5276777148246765,
      "learning_rate": 8e-05,
      "loss": 1.5727,
      "step": 2050
    },
    {
      "epoch": 0.42596053997923156,
      "grad_norm": 0.4825052320957184,
      "learning_rate": 8e-05,
      "loss": 1.5985,
      "step": 2051
    },
    {
      "epoch": 0.4261682242990654,
      "grad_norm": 0.4818319082260132,
      "learning_rate": 8e-05,
      "loss": 1.5049,
      "step": 2052
    },
    {
      "epoch": 0.4263759086188993,
      "grad_norm": 0.4871183931827545,
      "learning_rate": 8e-05,
      "loss": 1.5877,
      "step": 2053
    },
    {
      "epoch": 0.4265835929387331,
      "grad_norm": 0.5131775140762329,
      "learning_rate": 8e-05,
      "loss": 1.6401,
      "step": 2054
    },
    {
      "epoch": 0.42679127725856697,
      "grad_norm": 0.5020706057548523,
      "learning_rate": 8e-05,
      "loss": 1.5895,
      "step": 2055
    },
    {
      "epoch": 0.42699896157840084,
      "grad_norm": 0.5350056886672974,
      "learning_rate": 8e-05,
      "loss": 1.584,
      "step": 2056
    },
    {
      "epoch": 0.4272066458982347,
      "grad_norm": 0.5034728646278381,
      "learning_rate": 8e-05,
      "loss": 1.581,
      "step": 2057
    },
    {
      "epoch": 0.4274143302180685,
      "grad_norm": 0.5005778670310974,
      "learning_rate": 8e-05,
      "loss": 1.5836,
      "step": 2058
    },
    {
      "epoch": 0.4276220145379024,
      "grad_norm": 0.5032508969306946,
      "learning_rate": 8e-05,
      "loss": 1.6192,
      "step": 2059
    },
    {
      "epoch": 0.42782969885773625,
      "grad_norm": 0.5014644861221313,
      "learning_rate": 8e-05,
      "loss": 1.5815,
      "step": 2060
    },
    {
      "epoch": 0.4280373831775701,
      "grad_norm": 0.48992249369621277,
      "learning_rate": 8e-05,
      "loss": 1.5504,
      "step": 2061
    },
    {
      "epoch": 0.42824506749740393,
      "grad_norm": 0.49682915210723877,
      "learning_rate": 8e-05,
      "loss": 1.5766,
      "step": 2062
    },
    {
      "epoch": 0.4284527518172378,
      "grad_norm": 0.5146527290344238,
      "learning_rate": 8e-05,
      "loss": 1.6162,
      "step": 2063
    },
    {
      "epoch": 0.42866043613707167,
      "grad_norm": 0.4989301562309265,
      "learning_rate": 8e-05,
      "loss": 1.603,
      "step": 2064
    },
    {
      "epoch": 0.4288681204569055,
      "grad_norm": 0.5287513136863708,
      "learning_rate": 8e-05,
      "loss": 1.6206,
      "step": 2065
    },
    {
      "epoch": 0.42907580477673934,
      "grad_norm": 0.5166513919830322,
      "learning_rate": 8e-05,
      "loss": 1.5177,
      "step": 2066
    },
    {
      "epoch": 0.4292834890965732,
      "grad_norm": 0.52769935131073,
      "learning_rate": 8e-05,
      "loss": 1.5699,
      "step": 2067
    },
    {
      "epoch": 0.4294911734164071,
      "grad_norm": 0.5170919895172119,
      "learning_rate": 8e-05,
      "loss": 1.6006,
      "step": 2068
    },
    {
      "epoch": 0.4296988577362409,
      "grad_norm": 0.5070621371269226,
      "learning_rate": 8e-05,
      "loss": 1.6021,
      "step": 2069
    },
    {
      "epoch": 0.42990654205607476,
      "grad_norm": 0.5316301584243774,
      "learning_rate": 8e-05,
      "loss": 1.5772,
      "step": 2070
    },
    {
      "epoch": 0.4301142263759086,
      "grad_norm": 0.4892559349536896,
      "learning_rate": 8e-05,
      "loss": 1.5981,
      "step": 2071
    },
    {
      "epoch": 0.4303219106957425,
      "grad_norm": 0.48384004831314087,
      "learning_rate": 8e-05,
      "loss": 1.5119,
      "step": 2072
    },
    {
      "epoch": 0.4305295950155763,
      "grad_norm": 0.486844539642334,
      "learning_rate": 8e-05,
      "loss": 1.5741,
      "step": 2073
    },
    {
      "epoch": 0.4307372793354102,
      "grad_norm": 0.5377826690673828,
      "learning_rate": 8e-05,
      "loss": 1.617,
      "step": 2074
    },
    {
      "epoch": 0.43094496365524404,
      "grad_norm": 0.5149186253547668,
      "learning_rate": 8e-05,
      "loss": 1.553,
      "step": 2075
    },
    {
      "epoch": 0.4311526479750779,
      "grad_norm": 0.4895002841949463,
      "learning_rate": 8e-05,
      "loss": 1.5779,
      "step": 2076
    },
    {
      "epoch": 0.4313603322949117,
      "grad_norm": 0.48911821842193604,
      "learning_rate": 8e-05,
      "loss": 1.5364,
      "step": 2077
    },
    {
      "epoch": 0.4315680166147456,
      "grad_norm": 0.49317991733551025,
      "learning_rate": 8e-05,
      "loss": 1.5906,
      "step": 2078
    },
    {
      "epoch": 0.43177570093457945,
      "grad_norm": 0.5003262162208557,
      "learning_rate": 8e-05,
      "loss": 1.5706,
      "step": 2079
    },
    {
      "epoch": 0.43198338525441327,
      "grad_norm": 0.5054447650909424,
      "learning_rate": 8e-05,
      "loss": 1.5691,
      "step": 2080
    },
    {
      "epoch": 0.43219106957424713,
      "grad_norm": 0.4890398383140564,
      "learning_rate": 8e-05,
      "loss": 1.5654,
      "step": 2081
    },
    {
      "epoch": 0.432398753894081,
      "grad_norm": 0.5025361180305481,
      "learning_rate": 8e-05,
      "loss": 1.5346,
      "step": 2082
    },
    {
      "epoch": 0.43260643821391487,
      "grad_norm": 0.5030412673950195,
      "learning_rate": 8e-05,
      "loss": 1.5995,
      "step": 2083
    },
    {
      "epoch": 0.4328141225337487,
      "grad_norm": 0.5007551908493042,
      "learning_rate": 8e-05,
      "loss": 1.6198,
      "step": 2084
    },
    {
      "epoch": 0.43302180685358255,
      "grad_norm": 0.5120386481285095,
      "learning_rate": 8e-05,
      "loss": 1.5953,
      "step": 2085
    },
    {
      "epoch": 0.4332294911734164,
      "grad_norm": 0.5004842281341553,
      "learning_rate": 8e-05,
      "loss": 1.5432,
      "step": 2086
    },
    {
      "epoch": 0.4334371754932503,
      "grad_norm": 0.521942675113678,
      "learning_rate": 8e-05,
      "loss": 1.6336,
      "step": 2087
    },
    {
      "epoch": 0.4336448598130841,
      "grad_norm": 0.494864284992218,
      "learning_rate": 8e-05,
      "loss": 1.5748,
      "step": 2088
    },
    {
      "epoch": 0.43385254413291796,
      "grad_norm": 0.49404293298721313,
      "learning_rate": 8e-05,
      "loss": 1.6049,
      "step": 2089
    },
    {
      "epoch": 0.43406022845275183,
      "grad_norm": 0.5624111294746399,
      "learning_rate": 8e-05,
      "loss": 1.5513,
      "step": 2090
    },
    {
      "epoch": 0.4342679127725857,
      "grad_norm": 0.5119521021842957,
      "learning_rate": 8e-05,
      "loss": 1.5539,
      "step": 2091
    },
    {
      "epoch": 0.4344755970924195,
      "grad_norm": 0.51494961977005,
      "learning_rate": 8e-05,
      "loss": 1.586,
      "step": 2092
    },
    {
      "epoch": 0.4346832814122534,
      "grad_norm": 0.5104636549949646,
      "learning_rate": 8e-05,
      "loss": 1.6364,
      "step": 2093
    },
    {
      "epoch": 0.43489096573208724,
      "grad_norm": 0.5149340629577637,
      "learning_rate": 8e-05,
      "loss": 1.5458,
      "step": 2094
    },
    {
      "epoch": 0.43509865005192105,
      "grad_norm": 0.5006559491157532,
      "learning_rate": 8e-05,
      "loss": 1.6001,
      "step": 2095
    },
    {
      "epoch": 0.4353063343717549,
      "grad_norm": 0.4876810908317566,
      "learning_rate": 8e-05,
      "loss": 1.5671,
      "step": 2096
    },
    {
      "epoch": 0.4355140186915888,
      "grad_norm": 0.5194754600524902,
      "learning_rate": 8e-05,
      "loss": 1.5521,
      "step": 2097
    },
    {
      "epoch": 0.43572170301142266,
      "grad_norm": 0.4904755651950836,
      "learning_rate": 8e-05,
      "loss": 1.5825,
      "step": 2098
    },
    {
      "epoch": 0.43592938733125647,
      "grad_norm": 0.49247294664382935,
      "learning_rate": 8e-05,
      "loss": 1.4871,
      "step": 2099
    },
    {
      "epoch": 0.43613707165109034,
      "grad_norm": 0.5013484358787537,
      "learning_rate": 8e-05,
      "loss": 1.5281,
      "step": 2100
    },
    {
      "epoch": 0.4363447559709242,
      "grad_norm": 0.5134440064430237,
      "learning_rate": 8e-05,
      "loss": 1.596,
      "step": 2101
    },
    {
      "epoch": 0.43655244029075807,
      "grad_norm": 0.5220896601676941,
      "learning_rate": 8e-05,
      "loss": 1.6232,
      "step": 2102
    },
    {
      "epoch": 0.4367601246105919,
      "grad_norm": 0.49945664405822754,
      "learning_rate": 8e-05,
      "loss": 1.5805,
      "step": 2103
    },
    {
      "epoch": 0.43696780893042575,
      "grad_norm": 0.5018790364265442,
      "learning_rate": 8e-05,
      "loss": 1.5935,
      "step": 2104
    },
    {
      "epoch": 0.4371754932502596,
      "grad_norm": 0.5007161498069763,
      "learning_rate": 8e-05,
      "loss": 1.6203,
      "step": 2105
    },
    {
      "epoch": 0.4373831775700935,
      "grad_norm": 0.5214696526527405,
      "learning_rate": 8e-05,
      "loss": 1.6249,
      "step": 2106
    },
    {
      "epoch": 0.4375908618899273,
      "grad_norm": 0.5219676494598389,
      "learning_rate": 8e-05,
      "loss": 1.5492,
      "step": 2107
    },
    {
      "epoch": 0.43779854620976116,
      "grad_norm": 0.5156602263450623,
      "learning_rate": 8e-05,
      "loss": 1.6083,
      "step": 2108
    },
    {
      "epoch": 0.43800623052959503,
      "grad_norm": 0.534309446811676,
      "learning_rate": 8e-05,
      "loss": 1.7035,
      "step": 2109
    },
    {
      "epoch": 0.43821391484942884,
      "grad_norm": 0.4978565275669098,
      "learning_rate": 8e-05,
      "loss": 1.5881,
      "step": 2110
    },
    {
      "epoch": 0.4384215991692627,
      "grad_norm": 0.5232987403869629,
      "learning_rate": 8e-05,
      "loss": 1.6172,
      "step": 2111
    },
    {
      "epoch": 0.4386292834890966,
      "grad_norm": 0.512504518032074,
      "learning_rate": 8e-05,
      "loss": 1.5703,
      "step": 2112
    },
    {
      "epoch": 0.43883696780893044,
      "grad_norm": 0.47827327251434326,
      "learning_rate": 8e-05,
      "loss": 1.5624,
      "step": 2113
    },
    {
      "epoch": 0.43904465212876426,
      "grad_norm": 0.49305495619773865,
      "learning_rate": 8e-05,
      "loss": 1.537,
      "step": 2114
    },
    {
      "epoch": 0.4392523364485981,
      "grad_norm": 0.4876486361026764,
      "learning_rate": 8e-05,
      "loss": 1.5634,
      "step": 2115
    },
    {
      "epoch": 0.439460020768432,
      "grad_norm": 0.5139252543449402,
      "learning_rate": 8e-05,
      "loss": 1.5554,
      "step": 2116
    },
    {
      "epoch": 0.43966770508826586,
      "grad_norm": 0.5262313485145569,
      "learning_rate": 8e-05,
      "loss": 1.5863,
      "step": 2117
    },
    {
      "epoch": 0.43987538940809967,
      "grad_norm": 0.527298092842102,
      "learning_rate": 8e-05,
      "loss": 1.5333,
      "step": 2118
    },
    {
      "epoch": 0.44008307372793354,
      "grad_norm": 0.5263831615447998,
      "learning_rate": 8e-05,
      "loss": 1.6122,
      "step": 2119
    },
    {
      "epoch": 0.4402907580477674,
      "grad_norm": 0.49422043561935425,
      "learning_rate": 8e-05,
      "loss": 1.5043,
      "step": 2120
    },
    {
      "epoch": 0.4404984423676013,
      "grad_norm": 0.4889119565486908,
      "learning_rate": 8e-05,
      "loss": 1.5843,
      "step": 2121
    },
    {
      "epoch": 0.4407061266874351,
      "grad_norm": 0.5255298614501953,
      "learning_rate": 8e-05,
      "loss": 1.6276,
      "step": 2122
    },
    {
      "epoch": 0.44091381100726895,
      "grad_norm": 0.4965208172798157,
      "learning_rate": 8e-05,
      "loss": 1.6024,
      "step": 2123
    },
    {
      "epoch": 0.4411214953271028,
      "grad_norm": 0.5215441584587097,
      "learning_rate": 8e-05,
      "loss": 1.6019,
      "step": 2124
    },
    {
      "epoch": 0.44132917964693663,
      "grad_norm": 0.50166255235672,
      "learning_rate": 8e-05,
      "loss": 1.5099,
      "step": 2125
    },
    {
      "epoch": 0.4415368639667705,
      "grad_norm": 0.48959848284721375,
      "learning_rate": 8e-05,
      "loss": 1.5324,
      "step": 2126
    },
    {
      "epoch": 0.44174454828660437,
      "grad_norm": 0.496033638715744,
      "learning_rate": 8e-05,
      "loss": 1.6065,
      "step": 2127
    },
    {
      "epoch": 0.44195223260643823,
      "grad_norm": 0.5017862319946289,
      "learning_rate": 8e-05,
      "loss": 1.5673,
      "step": 2128
    },
    {
      "epoch": 0.44215991692627205,
      "grad_norm": 0.49096164107322693,
      "learning_rate": 8e-05,
      "loss": 1.6137,
      "step": 2129
    },
    {
      "epoch": 0.4423676012461059,
      "grad_norm": 0.51192307472229,
      "learning_rate": 8e-05,
      "loss": 1.5664,
      "step": 2130
    },
    {
      "epoch": 0.4425752855659398,
      "grad_norm": 0.5064907670021057,
      "learning_rate": 8e-05,
      "loss": 1.6066,
      "step": 2131
    },
    {
      "epoch": 0.44278296988577365,
      "grad_norm": 0.4990641176700592,
      "learning_rate": 8e-05,
      "loss": 1.6383,
      "step": 2132
    },
    {
      "epoch": 0.44299065420560746,
      "grad_norm": 0.5085825324058533,
      "learning_rate": 8e-05,
      "loss": 1.5627,
      "step": 2133
    },
    {
      "epoch": 0.4431983385254413,
      "grad_norm": 0.5027173757553101,
      "learning_rate": 8e-05,
      "loss": 1.5873,
      "step": 2134
    },
    {
      "epoch": 0.4434060228452752,
      "grad_norm": 0.48605379462242126,
      "learning_rate": 8e-05,
      "loss": 1.6316,
      "step": 2135
    },
    {
      "epoch": 0.44361370716510906,
      "grad_norm": 0.5001710057258606,
      "learning_rate": 8e-05,
      "loss": 1.6056,
      "step": 2136
    },
    {
      "epoch": 0.4438213914849429,
      "grad_norm": 0.5118576884269714,
      "learning_rate": 8e-05,
      "loss": 1.5986,
      "step": 2137
    },
    {
      "epoch": 0.44402907580477674,
      "grad_norm": 0.5374376773834229,
      "learning_rate": 8e-05,
      "loss": 1.6955,
      "step": 2138
    },
    {
      "epoch": 0.4442367601246106,
      "grad_norm": 0.5199704170227051,
      "learning_rate": 8e-05,
      "loss": 1.5547,
      "step": 2139
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 0.5208665132522583,
      "learning_rate": 8e-05,
      "loss": 1.491,
      "step": 2140
    },
    {
      "epoch": 0.4446521287642783,
      "grad_norm": 0.4891216456890106,
      "learning_rate": 8e-05,
      "loss": 1.5564,
      "step": 2141
    },
    {
      "epoch": 0.44485981308411215,
      "grad_norm": 0.5020160675048828,
      "learning_rate": 8e-05,
      "loss": 1.6321,
      "step": 2142
    },
    {
      "epoch": 0.445067497403946,
      "grad_norm": 0.49476900696754456,
      "learning_rate": 8e-05,
      "loss": 1.5914,
      "step": 2143
    },
    {
      "epoch": 0.44527518172377983,
      "grad_norm": 0.49868083000183105,
      "learning_rate": 8e-05,
      "loss": 1.501,
      "step": 2144
    },
    {
      "epoch": 0.4454828660436137,
      "grad_norm": 0.5365971922874451,
      "learning_rate": 8e-05,
      "loss": 1.6427,
      "step": 2145
    },
    {
      "epoch": 0.44569055036344757,
      "grad_norm": 0.5179263353347778,
      "learning_rate": 8e-05,
      "loss": 1.5516,
      "step": 2146
    },
    {
      "epoch": 0.44589823468328144,
      "grad_norm": 0.4961378872394562,
      "learning_rate": 8e-05,
      "loss": 1.487,
      "step": 2147
    },
    {
      "epoch": 0.44610591900311525,
      "grad_norm": 0.4983571767807007,
      "learning_rate": 8e-05,
      "loss": 1.6078,
      "step": 2148
    },
    {
      "epoch": 0.4463136033229491,
      "grad_norm": 0.5751625299453735,
      "learning_rate": 8e-05,
      "loss": 1.5492,
      "step": 2149
    },
    {
      "epoch": 0.446521287642783,
      "grad_norm": 0.5159772634506226,
      "learning_rate": 8e-05,
      "loss": 1.6149,
      "step": 2150
    },
    {
      "epoch": 0.44672897196261685,
      "grad_norm": 0.48897507786750793,
      "learning_rate": 8e-05,
      "loss": 1.621,
      "step": 2151
    },
    {
      "epoch": 0.44693665628245066,
      "grad_norm": 0.544232964515686,
      "learning_rate": 8e-05,
      "loss": 1.5878,
      "step": 2152
    },
    {
      "epoch": 0.44714434060228453,
      "grad_norm": 0.5140170454978943,
      "learning_rate": 8e-05,
      "loss": 1.6066,
      "step": 2153
    },
    {
      "epoch": 0.4473520249221184,
      "grad_norm": 0.49031364917755127,
      "learning_rate": 8e-05,
      "loss": 1.5178,
      "step": 2154
    },
    {
      "epoch": 0.4475597092419522,
      "grad_norm": 0.4932582378387451,
      "learning_rate": 8e-05,
      "loss": 1.5018,
      "step": 2155
    },
    {
      "epoch": 0.4477673935617861,
      "grad_norm": 0.5137055516242981,
      "learning_rate": 8e-05,
      "loss": 1.6213,
      "step": 2156
    },
    {
      "epoch": 0.44797507788161994,
      "grad_norm": 0.5106627941131592,
      "learning_rate": 8e-05,
      "loss": 1.5928,
      "step": 2157
    },
    {
      "epoch": 0.4481827622014538,
      "grad_norm": 0.5080628395080566,
      "learning_rate": 8e-05,
      "loss": 1.5682,
      "step": 2158
    },
    {
      "epoch": 0.4483904465212876,
      "grad_norm": 0.5078219771385193,
      "learning_rate": 8e-05,
      "loss": 1.6174,
      "step": 2159
    },
    {
      "epoch": 0.4485981308411215,
      "grad_norm": 0.5262932777404785,
      "learning_rate": 8e-05,
      "loss": 1.536,
      "step": 2160
    },
    {
      "epoch": 0.44880581516095536,
      "grad_norm": 0.5089083313941956,
      "learning_rate": 8e-05,
      "loss": 1.543,
      "step": 2161
    },
    {
      "epoch": 0.4490134994807892,
      "grad_norm": 0.5383339524269104,
      "learning_rate": 8e-05,
      "loss": 1.6122,
      "step": 2162
    },
    {
      "epoch": 0.44922118380062304,
      "grad_norm": 0.5017908215522766,
      "learning_rate": 8e-05,
      "loss": 1.5862,
      "step": 2163
    },
    {
      "epoch": 0.4494288681204569,
      "grad_norm": 0.4958338439464569,
      "learning_rate": 8e-05,
      "loss": 1.5661,
      "step": 2164
    },
    {
      "epoch": 0.44963655244029077,
      "grad_norm": 0.4877220094203949,
      "learning_rate": 8e-05,
      "loss": 1.5747,
      "step": 2165
    },
    {
      "epoch": 0.44984423676012464,
      "grad_norm": 0.49807313084602356,
      "learning_rate": 8e-05,
      "loss": 1.5605,
      "step": 2166
    },
    {
      "epoch": 0.45005192107995845,
      "grad_norm": 0.5027108788490295,
      "learning_rate": 8e-05,
      "loss": 1.523,
      "step": 2167
    },
    {
      "epoch": 0.4502596053997923,
      "grad_norm": 0.492351233959198,
      "learning_rate": 8e-05,
      "loss": 1.506,
      "step": 2168
    },
    {
      "epoch": 0.4504672897196262,
      "grad_norm": 0.4846895933151245,
      "learning_rate": 8e-05,
      "loss": 1.5205,
      "step": 2169
    },
    {
      "epoch": 0.45067497403946,
      "grad_norm": 0.49987098574638367,
      "learning_rate": 8e-05,
      "loss": 1.5675,
      "step": 2170
    },
    {
      "epoch": 0.45088265835929386,
      "grad_norm": 0.4977555274963379,
      "learning_rate": 8e-05,
      "loss": 1.6083,
      "step": 2171
    },
    {
      "epoch": 0.45109034267912773,
      "grad_norm": 0.5156652927398682,
      "learning_rate": 8e-05,
      "loss": 1.6064,
      "step": 2172
    },
    {
      "epoch": 0.4512980269989616,
      "grad_norm": 0.48706451058387756,
      "learning_rate": 8e-05,
      "loss": 1.563,
      "step": 2173
    },
    {
      "epoch": 0.4515057113187954,
      "grad_norm": 0.5008767247200012,
      "learning_rate": 8e-05,
      "loss": 1.5815,
      "step": 2174
    },
    {
      "epoch": 0.4517133956386293,
      "grad_norm": 0.4886937737464905,
      "learning_rate": 8e-05,
      "loss": 1.5869,
      "step": 2175
    },
    {
      "epoch": 0.45192107995846315,
      "grad_norm": 0.5106097459793091,
      "learning_rate": 8e-05,
      "loss": 1.6086,
      "step": 2176
    },
    {
      "epoch": 0.452128764278297,
      "grad_norm": 0.5486605763435364,
      "learning_rate": 8e-05,
      "loss": 1.5641,
      "step": 2177
    },
    {
      "epoch": 0.4523364485981308,
      "grad_norm": 0.4942476749420166,
      "learning_rate": 8e-05,
      "loss": 1.5703,
      "step": 2178
    },
    {
      "epoch": 0.4525441329179647,
      "grad_norm": 0.5139703750610352,
      "learning_rate": 8e-05,
      "loss": 1.5634,
      "step": 2179
    },
    {
      "epoch": 0.45275181723779856,
      "grad_norm": 0.5057925581932068,
      "learning_rate": 8e-05,
      "loss": 1.5819,
      "step": 2180
    },
    {
      "epoch": 0.45295950155763237,
      "grad_norm": 0.49567729234695435,
      "learning_rate": 8e-05,
      "loss": 1.5186,
      "step": 2181
    },
    {
      "epoch": 0.45316718587746624,
      "grad_norm": 0.549098551273346,
      "learning_rate": 8e-05,
      "loss": 1.5975,
      "step": 2182
    },
    {
      "epoch": 0.4533748701973001,
      "grad_norm": 0.51541668176651,
      "learning_rate": 8e-05,
      "loss": 1.669,
      "step": 2183
    },
    {
      "epoch": 0.453582554517134,
      "grad_norm": 0.6035095453262329,
      "learning_rate": 8e-05,
      "loss": 1.6484,
      "step": 2184
    },
    {
      "epoch": 0.4537902388369678,
      "grad_norm": 0.5210909843444824,
      "learning_rate": 8e-05,
      "loss": 1.6159,
      "step": 2185
    },
    {
      "epoch": 0.45399792315680165,
      "grad_norm": 0.5404158234596252,
      "learning_rate": 8e-05,
      "loss": 1.4956,
      "step": 2186
    },
    {
      "epoch": 0.4542056074766355,
      "grad_norm": 0.541645884513855,
      "learning_rate": 8e-05,
      "loss": 1.5535,
      "step": 2187
    },
    {
      "epoch": 0.4544132917964694,
      "grad_norm": 0.4959733784198761,
      "learning_rate": 8e-05,
      "loss": 1.5454,
      "step": 2188
    },
    {
      "epoch": 0.4546209761163032,
      "grad_norm": 0.5585200786590576,
      "learning_rate": 8e-05,
      "loss": 1.6141,
      "step": 2189
    },
    {
      "epoch": 0.45482866043613707,
      "grad_norm": 0.5159070491790771,
      "learning_rate": 8e-05,
      "loss": 1.5757,
      "step": 2190
    },
    {
      "epoch": 0.45503634475597093,
      "grad_norm": 0.5728627443313599,
      "learning_rate": 8e-05,
      "loss": 1.6568,
      "step": 2191
    },
    {
      "epoch": 0.4552440290758048,
      "grad_norm": 0.517425000667572,
      "learning_rate": 8e-05,
      "loss": 1.483,
      "step": 2192
    },
    {
      "epoch": 0.4554517133956386,
      "grad_norm": 0.503883421421051,
      "learning_rate": 8e-05,
      "loss": 1.5684,
      "step": 2193
    },
    {
      "epoch": 0.4556593977154725,
      "grad_norm": 0.53370201587677,
      "learning_rate": 8e-05,
      "loss": 1.5402,
      "step": 2194
    },
    {
      "epoch": 0.45586708203530635,
      "grad_norm": 0.5302833318710327,
      "learning_rate": 8e-05,
      "loss": 1.5462,
      "step": 2195
    },
    {
      "epoch": 0.45607476635514016,
      "grad_norm": 0.531651496887207,
      "learning_rate": 8e-05,
      "loss": 1.5692,
      "step": 2196
    },
    {
      "epoch": 0.456282450674974,
      "grad_norm": 0.5229583978652954,
      "learning_rate": 8e-05,
      "loss": 1.5554,
      "step": 2197
    },
    {
      "epoch": 0.4564901349948079,
      "grad_norm": 0.4997648000717163,
      "learning_rate": 8e-05,
      "loss": 1.6515,
      "step": 2198
    },
    {
      "epoch": 0.45669781931464176,
      "grad_norm": 0.5124852657318115,
      "learning_rate": 8e-05,
      "loss": 1.5617,
      "step": 2199
    },
    {
      "epoch": 0.4569055036344756,
      "grad_norm": 0.5129377841949463,
      "learning_rate": 8e-05,
      "loss": 1.6057,
      "step": 2200
    },
    {
      "epoch": 0.45711318795430944,
      "grad_norm": 0.5239273905754089,
      "learning_rate": 8e-05,
      "loss": 1.5661,
      "step": 2201
    },
    {
      "epoch": 0.4573208722741433,
      "grad_norm": 0.54039466381073,
      "learning_rate": 8e-05,
      "loss": 1.526,
      "step": 2202
    },
    {
      "epoch": 0.4575285565939772,
      "grad_norm": 0.5111913681030273,
      "learning_rate": 8e-05,
      "loss": 1.6207,
      "step": 2203
    },
    {
      "epoch": 0.457736240913811,
      "grad_norm": 0.5141212940216064,
      "learning_rate": 8e-05,
      "loss": 1.5606,
      "step": 2204
    },
    {
      "epoch": 0.45794392523364486,
      "grad_norm": 0.510006844997406,
      "learning_rate": 8e-05,
      "loss": 1.559,
      "step": 2205
    },
    {
      "epoch": 0.4581516095534787,
      "grad_norm": 0.5465336441993713,
      "learning_rate": 8e-05,
      "loss": 1.5428,
      "step": 2206
    },
    {
      "epoch": 0.4583592938733126,
      "grad_norm": 0.5228673815727234,
      "learning_rate": 8e-05,
      "loss": 1.6287,
      "step": 2207
    },
    {
      "epoch": 0.4585669781931464,
      "grad_norm": 0.5198245644569397,
      "learning_rate": 8e-05,
      "loss": 1.5437,
      "step": 2208
    },
    {
      "epoch": 0.45877466251298027,
      "grad_norm": 0.5053104758262634,
      "learning_rate": 8e-05,
      "loss": 1.6151,
      "step": 2209
    },
    {
      "epoch": 0.45898234683281414,
      "grad_norm": 0.5879641771316528,
      "learning_rate": 8e-05,
      "loss": 1.5599,
      "step": 2210
    },
    {
      "epoch": 0.45919003115264795,
      "grad_norm": 0.4976397752761841,
      "learning_rate": 8e-05,
      "loss": 1.5307,
      "step": 2211
    },
    {
      "epoch": 0.4593977154724818,
      "grad_norm": 0.500487208366394,
      "learning_rate": 8e-05,
      "loss": 1.4737,
      "step": 2212
    },
    {
      "epoch": 0.4596053997923157,
      "grad_norm": 0.5141959190368652,
      "learning_rate": 8e-05,
      "loss": 1.5645,
      "step": 2213
    },
    {
      "epoch": 0.45981308411214955,
      "grad_norm": 0.5344618558883667,
      "learning_rate": 8e-05,
      "loss": 1.5177,
      "step": 2214
    },
    {
      "epoch": 0.46002076843198336,
      "grad_norm": 0.5062136650085449,
      "learning_rate": 8e-05,
      "loss": 1.5476,
      "step": 2215
    },
    {
      "epoch": 0.46022845275181723,
      "grad_norm": 0.5057580471038818,
      "learning_rate": 8e-05,
      "loss": 1.5666,
      "step": 2216
    },
    {
      "epoch": 0.4604361370716511,
      "grad_norm": 0.5169433355331421,
      "learning_rate": 8e-05,
      "loss": 1.5551,
      "step": 2217
    },
    {
      "epoch": 0.46064382139148496,
      "grad_norm": 0.5240709781646729,
      "learning_rate": 8e-05,
      "loss": 1.6269,
      "step": 2218
    },
    {
      "epoch": 0.4608515057113188,
      "grad_norm": 0.520158052444458,
      "learning_rate": 8e-05,
      "loss": 1.612,
      "step": 2219
    },
    {
      "epoch": 0.46105919003115264,
      "grad_norm": 0.5017967820167542,
      "learning_rate": 8e-05,
      "loss": 1.6067,
      "step": 2220
    },
    {
      "epoch": 0.4612668743509865,
      "grad_norm": 0.4989418685436249,
      "learning_rate": 8e-05,
      "loss": 1.5608,
      "step": 2221
    },
    {
      "epoch": 0.4614745586708204,
      "grad_norm": 0.5172755122184753,
      "learning_rate": 8e-05,
      "loss": 1.5421,
      "step": 2222
    },
    {
      "epoch": 0.4616822429906542,
      "grad_norm": 0.5159169435501099,
      "learning_rate": 8e-05,
      "loss": 1.6023,
      "step": 2223
    },
    {
      "epoch": 0.46188992731048806,
      "grad_norm": 0.5040188431739807,
      "learning_rate": 8e-05,
      "loss": 1.5828,
      "step": 2224
    },
    {
      "epoch": 0.4620976116303219,
      "grad_norm": 0.5402005314826965,
      "learning_rate": 8e-05,
      "loss": 1.6258,
      "step": 2225
    },
    {
      "epoch": 0.46230529595015574,
      "grad_norm": 0.5770882964134216,
      "learning_rate": 8e-05,
      "loss": 1.5726,
      "step": 2226
    },
    {
      "epoch": 0.4625129802699896,
      "grad_norm": 0.5088592767715454,
      "learning_rate": 8e-05,
      "loss": 1.6046,
      "step": 2227
    },
    {
      "epoch": 0.46272066458982347,
      "grad_norm": 0.5058512687683105,
      "learning_rate": 8e-05,
      "loss": 1.6188,
      "step": 2228
    },
    {
      "epoch": 0.46292834890965734,
      "grad_norm": 0.5000367760658264,
      "learning_rate": 8e-05,
      "loss": 1.5724,
      "step": 2229
    },
    {
      "epoch": 0.46313603322949115,
      "grad_norm": 0.5102644562721252,
      "learning_rate": 8e-05,
      "loss": 1.6051,
      "step": 2230
    },
    {
      "epoch": 0.463343717549325,
      "grad_norm": 0.512116014957428,
      "learning_rate": 8e-05,
      "loss": 1.5647,
      "step": 2231
    },
    {
      "epoch": 0.4635514018691589,
      "grad_norm": 0.49340537190437317,
      "learning_rate": 8e-05,
      "loss": 1.6017,
      "step": 2232
    },
    {
      "epoch": 0.46375908618899275,
      "grad_norm": 0.4973028004169464,
      "learning_rate": 8e-05,
      "loss": 1.5594,
      "step": 2233
    },
    {
      "epoch": 0.46396677050882656,
      "grad_norm": 0.5239974856376648,
      "learning_rate": 8e-05,
      "loss": 1.5847,
      "step": 2234
    },
    {
      "epoch": 0.46417445482866043,
      "grad_norm": 0.4996567666530609,
      "learning_rate": 8e-05,
      "loss": 1.5939,
      "step": 2235
    },
    {
      "epoch": 0.4643821391484943,
      "grad_norm": 0.5092349052429199,
      "learning_rate": 8e-05,
      "loss": 1.5328,
      "step": 2236
    },
    {
      "epoch": 0.46458982346832817,
      "grad_norm": 0.5141004323959351,
      "learning_rate": 8e-05,
      "loss": 1.6054,
      "step": 2237
    },
    {
      "epoch": 0.464797507788162,
      "grad_norm": 0.4950210154056549,
      "learning_rate": 8e-05,
      "loss": 1.5328,
      "step": 2238
    },
    {
      "epoch": 0.46500519210799585,
      "grad_norm": 0.499602347612381,
      "learning_rate": 8e-05,
      "loss": 1.555,
      "step": 2239
    },
    {
      "epoch": 0.4652128764278297,
      "grad_norm": 0.5248525142669678,
      "learning_rate": 8e-05,
      "loss": 1.6277,
      "step": 2240
    },
    {
      "epoch": 0.4654205607476635,
      "grad_norm": 0.5104266405105591,
      "learning_rate": 8e-05,
      "loss": 1.5733,
      "step": 2241
    },
    {
      "epoch": 0.4656282450674974,
      "grad_norm": 0.4947991669178009,
      "learning_rate": 8e-05,
      "loss": 1.5984,
      "step": 2242
    },
    {
      "epoch": 0.46583592938733126,
      "grad_norm": 0.519220232963562,
      "learning_rate": 8e-05,
      "loss": 1.6088,
      "step": 2243
    },
    {
      "epoch": 0.4660436137071651,
      "grad_norm": 0.4922308921813965,
      "learning_rate": 8e-05,
      "loss": 1.5143,
      "step": 2244
    },
    {
      "epoch": 0.46625129802699894,
      "grad_norm": 0.5001751184463501,
      "learning_rate": 8e-05,
      "loss": 1.5317,
      "step": 2245
    },
    {
      "epoch": 0.4664589823468328,
      "grad_norm": 0.5132100582122803,
      "learning_rate": 8e-05,
      "loss": 1.5765,
      "step": 2246
    },
    {
      "epoch": 0.4666666666666667,
      "grad_norm": 0.4972493350505829,
      "learning_rate": 8e-05,
      "loss": 1.5546,
      "step": 2247
    },
    {
      "epoch": 0.46687435098650054,
      "grad_norm": 0.5021941661834717,
      "learning_rate": 8e-05,
      "loss": 1.6312,
      "step": 2248
    },
    {
      "epoch": 0.46708203530633435,
      "grad_norm": 0.48572269082069397,
      "learning_rate": 8e-05,
      "loss": 1.5355,
      "step": 2249
    },
    {
      "epoch": 0.4672897196261682,
      "grad_norm": 0.5086265206336975,
      "learning_rate": 8e-05,
      "loss": 1.5432,
      "step": 2250
    },
    {
      "epoch": 0.4674974039460021,
      "grad_norm": 0.5098671913146973,
      "learning_rate": 8e-05,
      "loss": 1.5045,
      "step": 2251
    },
    {
      "epoch": 0.46770508826583596,
      "grad_norm": 0.5196540951728821,
      "learning_rate": 8e-05,
      "loss": 1.5743,
      "step": 2252
    },
    {
      "epoch": 0.46791277258566977,
      "grad_norm": 0.5160411596298218,
      "learning_rate": 8e-05,
      "loss": 1.6001,
      "step": 2253
    },
    {
      "epoch": 0.46812045690550363,
      "grad_norm": 0.5127061605453491,
      "learning_rate": 8e-05,
      "loss": 1.534,
      "step": 2254
    },
    {
      "epoch": 0.4683281412253375,
      "grad_norm": 0.5143935084342957,
      "learning_rate": 8e-05,
      "loss": 1.5723,
      "step": 2255
    },
    {
      "epoch": 0.4685358255451713,
      "grad_norm": 0.5065985918045044,
      "learning_rate": 8e-05,
      "loss": 1.5099,
      "step": 2256
    },
    {
      "epoch": 0.4687435098650052,
      "grad_norm": 0.5203616619110107,
      "learning_rate": 8e-05,
      "loss": 1.5427,
      "step": 2257
    },
    {
      "epoch": 0.46895119418483905,
      "grad_norm": 0.5806429386138916,
      "learning_rate": 8e-05,
      "loss": 1.6105,
      "step": 2258
    },
    {
      "epoch": 0.4691588785046729,
      "grad_norm": 0.4942633807659149,
      "learning_rate": 8e-05,
      "loss": 1.5873,
      "step": 2259
    },
    {
      "epoch": 0.46936656282450673,
      "grad_norm": 0.5223998427391052,
      "learning_rate": 8e-05,
      "loss": 1.6349,
      "step": 2260
    },
    {
      "epoch": 0.4695742471443406,
      "grad_norm": 0.5162349343299866,
      "learning_rate": 8e-05,
      "loss": 1.5431,
      "step": 2261
    },
    {
      "epoch": 0.46978193146417446,
      "grad_norm": 0.4979266822338104,
      "learning_rate": 8e-05,
      "loss": 1.5814,
      "step": 2262
    },
    {
      "epoch": 0.46998961578400833,
      "grad_norm": 0.5179408192634583,
      "learning_rate": 8e-05,
      "loss": 1.6749,
      "step": 2263
    },
    {
      "epoch": 0.47019730010384214,
      "grad_norm": 0.5000909566879272,
      "learning_rate": 8e-05,
      "loss": 1.5196,
      "step": 2264
    },
    {
      "epoch": 0.470404984423676,
      "grad_norm": 0.4916440546512604,
      "learning_rate": 8e-05,
      "loss": 1.5558,
      "step": 2265
    },
    {
      "epoch": 0.4706126687435099,
      "grad_norm": 0.5041160583496094,
      "learning_rate": 8e-05,
      "loss": 1.6098,
      "step": 2266
    },
    {
      "epoch": 0.47082035306334374,
      "grad_norm": 0.4847199320793152,
      "learning_rate": 8e-05,
      "loss": 1.5478,
      "step": 2267
    },
    {
      "epoch": 0.47102803738317756,
      "grad_norm": 0.5013338327407837,
      "learning_rate": 8e-05,
      "loss": 1.5531,
      "step": 2268
    },
    {
      "epoch": 0.4712357217030114,
      "grad_norm": 0.49625855684280396,
      "learning_rate": 8e-05,
      "loss": 1.5985,
      "step": 2269
    },
    {
      "epoch": 0.4714434060228453,
      "grad_norm": 0.5018194317817688,
      "learning_rate": 8e-05,
      "loss": 1.5717,
      "step": 2270
    },
    {
      "epoch": 0.4716510903426791,
      "grad_norm": 0.5168567299842834,
      "learning_rate": 8e-05,
      "loss": 1.5784,
      "step": 2271
    },
    {
      "epoch": 0.47185877466251297,
      "grad_norm": 0.49007734656333923,
      "learning_rate": 8e-05,
      "loss": 1.5197,
      "step": 2272
    },
    {
      "epoch": 0.47206645898234684,
      "grad_norm": 0.4975670576095581,
      "learning_rate": 8e-05,
      "loss": 1.6233,
      "step": 2273
    },
    {
      "epoch": 0.4722741433021807,
      "grad_norm": 0.5049093961715698,
      "learning_rate": 8e-05,
      "loss": 1.5514,
      "step": 2274
    },
    {
      "epoch": 0.4724818276220145,
      "grad_norm": 0.5157124996185303,
      "learning_rate": 8e-05,
      "loss": 1.6597,
      "step": 2275
    },
    {
      "epoch": 0.4726895119418484,
      "grad_norm": 0.5225481986999512,
      "learning_rate": 8e-05,
      "loss": 1.6546,
      "step": 2276
    },
    {
      "epoch": 0.47289719626168225,
      "grad_norm": 0.507124125957489,
      "learning_rate": 8e-05,
      "loss": 1.6082,
      "step": 2277
    },
    {
      "epoch": 0.4731048805815161,
      "grad_norm": 0.495510071516037,
      "learning_rate": 8e-05,
      "loss": 1.5596,
      "step": 2278
    },
    {
      "epoch": 0.47331256490134993,
      "grad_norm": 0.5406957864761353,
      "learning_rate": 8e-05,
      "loss": 1.5302,
      "step": 2279
    },
    {
      "epoch": 0.4735202492211838,
      "grad_norm": 0.5365445613861084,
      "learning_rate": 8e-05,
      "loss": 1.5712,
      "step": 2280
    },
    {
      "epoch": 0.47372793354101767,
      "grad_norm": 0.4859418272972107,
      "learning_rate": 8e-05,
      "loss": 1.6174,
      "step": 2281
    },
    {
      "epoch": 0.47393561786085153,
      "grad_norm": 0.502739429473877,
      "learning_rate": 8e-05,
      "loss": 1.5976,
      "step": 2282
    },
    {
      "epoch": 0.47414330218068534,
      "grad_norm": 0.5150384902954102,
      "learning_rate": 8e-05,
      "loss": 1.5256,
      "step": 2283
    },
    {
      "epoch": 0.4743509865005192,
      "grad_norm": 0.48385000228881836,
      "learning_rate": 8e-05,
      "loss": 1.5797,
      "step": 2284
    },
    {
      "epoch": 0.4745586708203531,
      "grad_norm": 0.5246078372001648,
      "learning_rate": 8e-05,
      "loss": 1.6164,
      "step": 2285
    },
    {
      "epoch": 0.4747663551401869,
      "grad_norm": 0.5253199934959412,
      "learning_rate": 8e-05,
      "loss": 1.5422,
      "step": 2286
    },
    {
      "epoch": 0.47497403946002076,
      "grad_norm": 0.5063892006874084,
      "learning_rate": 8e-05,
      "loss": 1.5868,
      "step": 2287
    },
    {
      "epoch": 0.4751817237798546,
      "grad_norm": 0.5025436878204346,
      "learning_rate": 8e-05,
      "loss": 1.5405,
      "step": 2288
    },
    {
      "epoch": 0.4753894080996885,
      "grad_norm": 0.48187509179115295,
      "learning_rate": 8e-05,
      "loss": 1.4724,
      "step": 2289
    },
    {
      "epoch": 0.4755970924195223,
      "grad_norm": 0.5353221893310547,
      "learning_rate": 8e-05,
      "loss": 1.5803,
      "step": 2290
    },
    {
      "epoch": 0.47580477673935617,
      "grad_norm": 0.5231779217720032,
      "learning_rate": 8e-05,
      "loss": 1.6353,
      "step": 2291
    },
    {
      "epoch": 0.47601246105919004,
      "grad_norm": 0.5240789651870728,
      "learning_rate": 8e-05,
      "loss": 1.5412,
      "step": 2292
    },
    {
      "epoch": 0.4762201453790239,
      "grad_norm": 0.4947764277458191,
      "learning_rate": 8e-05,
      "loss": 1.5188,
      "step": 2293
    },
    {
      "epoch": 0.4764278296988577,
      "grad_norm": 0.4879968464374542,
      "learning_rate": 8e-05,
      "loss": 1.5719,
      "step": 2294
    },
    {
      "epoch": 0.4766355140186916,
      "grad_norm": 0.49523064494132996,
      "learning_rate": 8e-05,
      "loss": 1.5927,
      "step": 2295
    },
    {
      "epoch": 0.47684319833852545,
      "grad_norm": 0.5027074217796326,
      "learning_rate": 8e-05,
      "loss": 1.5833,
      "step": 2296
    },
    {
      "epoch": 0.4770508826583593,
      "grad_norm": 0.5084169507026672,
      "learning_rate": 8e-05,
      "loss": 1.596,
      "step": 2297
    },
    {
      "epoch": 0.47725856697819313,
      "grad_norm": 0.51845782995224,
      "learning_rate": 8e-05,
      "loss": 1.5654,
      "step": 2298
    },
    {
      "epoch": 0.477466251298027,
      "grad_norm": 0.49563729763031006,
      "learning_rate": 8e-05,
      "loss": 1.6486,
      "step": 2299
    },
    {
      "epoch": 0.47767393561786087,
      "grad_norm": 0.510164737701416,
      "learning_rate": 8e-05,
      "loss": 1.6042,
      "step": 2300
    },
    {
      "epoch": 0.4778816199376947,
      "grad_norm": 0.5058431029319763,
      "learning_rate": 8e-05,
      "loss": 1.5206,
      "step": 2301
    },
    {
      "epoch": 0.47808930425752855,
      "grad_norm": 0.49267512559890747,
      "learning_rate": 8e-05,
      "loss": 1.5328,
      "step": 2302
    },
    {
      "epoch": 0.4782969885773624,
      "grad_norm": 0.5097163915634155,
      "learning_rate": 8e-05,
      "loss": 1.5723,
      "step": 2303
    },
    {
      "epoch": 0.4785046728971963,
      "grad_norm": 0.5137566328048706,
      "learning_rate": 8e-05,
      "loss": 1.5595,
      "step": 2304
    },
    {
      "epoch": 0.4787123572170301,
      "grad_norm": 0.511559009552002,
      "learning_rate": 8e-05,
      "loss": 1.6153,
      "step": 2305
    },
    {
      "epoch": 0.47892004153686396,
      "grad_norm": 0.49793705344200134,
      "learning_rate": 8e-05,
      "loss": 1.5774,
      "step": 2306
    },
    {
      "epoch": 0.47912772585669783,
      "grad_norm": 0.5060765743255615,
      "learning_rate": 8e-05,
      "loss": 1.5523,
      "step": 2307
    },
    {
      "epoch": 0.4793354101765317,
      "grad_norm": 0.49947306513786316,
      "learning_rate": 8e-05,
      "loss": 1.6676,
      "step": 2308
    },
    {
      "epoch": 0.4795430944963655,
      "grad_norm": 0.49617889523506165,
      "learning_rate": 8e-05,
      "loss": 1.5915,
      "step": 2309
    },
    {
      "epoch": 0.4797507788161994,
      "grad_norm": 0.5186249613761902,
      "learning_rate": 8e-05,
      "loss": 1.5919,
      "step": 2310
    },
    {
      "epoch": 0.47995846313603324,
      "grad_norm": 0.5017900466918945,
      "learning_rate": 8e-05,
      "loss": 1.5352,
      "step": 2311
    },
    {
      "epoch": 0.4801661474558671,
      "grad_norm": 0.5195544362068176,
      "learning_rate": 8e-05,
      "loss": 1.5829,
      "step": 2312
    },
    {
      "epoch": 0.4803738317757009,
      "grad_norm": 0.5048125386238098,
      "learning_rate": 8e-05,
      "loss": 1.6061,
      "step": 2313
    },
    {
      "epoch": 0.4805815160955348,
      "grad_norm": 0.5114527940750122,
      "learning_rate": 8e-05,
      "loss": 1.6474,
      "step": 2314
    },
    {
      "epoch": 0.48078920041536866,
      "grad_norm": 0.50144362449646,
      "learning_rate": 8e-05,
      "loss": 1.6303,
      "step": 2315
    },
    {
      "epoch": 0.48099688473520247,
      "grad_norm": 0.5169421434402466,
      "learning_rate": 8e-05,
      "loss": 1.5849,
      "step": 2316
    },
    {
      "epoch": 0.48120456905503634,
      "grad_norm": 0.4925006031990051,
      "learning_rate": 8e-05,
      "loss": 1.5671,
      "step": 2317
    },
    {
      "epoch": 0.4814122533748702,
      "grad_norm": 0.5097764134407043,
      "learning_rate": 8e-05,
      "loss": 1.4887,
      "step": 2318
    },
    {
      "epoch": 0.48161993769470407,
      "grad_norm": 0.4904961585998535,
      "learning_rate": 8e-05,
      "loss": 1.5444,
      "step": 2319
    },
    {
      "epoch": 0.4818276220145379,
      "grad_norm": 0.5130316615104675,
      "learning_rate": 8e-05,
      "loss": 1.6072,
      "step": 2320
    },
    {
      "epoch": 0.48203530633437175,
      "grad_norm": 0.4950331747531891,
      "learning_rate": 8e-05,
      "loss": 1.5635,
      "step": 2321
    },
    {
      "epoch": 0.4822429906542056,
      "grad_norm": 0.5047598481178284,
      "learning_rate": 8e-05,
      "loss": 1.5922,
      "step": 2322
    },
    {
      "epoch": 0.4824506749740395,
      "grad_norm": 0.5059360265731812,
      "learning_rate": 8e-05,
      "loss": 1.5522,
      "step": 2323
    },
    {
      "epoch": 0.4826583592938733,
      "grad_norm": 0.5029763579368591,
      "learning_rate": 8e-05,
      "loss": 1.6158,
      "step": 2324
    },
    {
      "epoch": 0.48286604361370716,
      "grad_norm": 0.518887460231781,
      "learning_rate": 8e-05,
      "loss": 1.5916,
      "step": 2325
    },
    {
      "epoch": 0.48307372793354103,
      "grad_norm": 0.5052504539489746,
      "learning_rate": 8e-05,
      "loss": 1.5139,
      "step": 2326
    },
    {
      "epoch": 0.4832814122533749,
      "grad_norm": 0.5410659909248352,
      "learning_rate": 8e-05,
      "loss": 1.6304,
      "step": 2327
    },
    {
      "epoch": 0.4834890965732087,
      "grad_norm": 0.5123754739761353,
      "learning_rate": 8e-05,
      "loss": 1.5425,
      "step": 2328
    },
    {
      "epoch": 0.4836967808930426,
      "grad_norm": 0.5117024779319763,
      "learning_rate": 8e-05,
      "loss": 1.5976,
      "step": 2329
    },
    {
      "epoch": 0.48390446521287644,
      "grad_norm": 0.5014856457710266,
      "learning_rate": 8e-05,
      "loss": 1.5196,
      "step": 2330
    },
    {
      "epoch": 0.48411214953271026,
      "grad_norm": 0.5097591280937195,
      "learning_rate": 8e-05,
      "loss": 1.6369,
      "step": 2331
    },
    {
      "epoch": 0.4843198338525441,
      "grad_norm": 0.5165501832962036,
      "learning_rate": 8e-05,
      "loss": 1.6051,
      "step": 2332
    },
    {
      "epoch": 0.484527518172378,
      "grad_norm": 0.4933048486709595,
      "learning_rate": 8e-05,
      "loss": 1.5319,
      "step": 2333
    },
    {
      "epoch": 0.48473520249221186,
      "grad_norm": 0.5414435267448425,
      "learning_rate": 8e-05,
      "loss": 1.5775,
      "step": 2334
    },
    {
      "epoch": 0.48494288681204567,
      "grad_norm": 0.5097305178642273,
      "learning_rate": 8e-05,
      "loss": 1.5904,
      "step": 2335
    },
    {
      "epoch": 0.48515057113187954,
      "grad_norm": 0.5298729538917542,
      "learning_rate": 8e-05,
      "loss": 1.5748,
      "step": 2336
    },
    {
      "epoch": 0.4853582554517134,
      "grad_norm": 0.496655136346817,
      "learning_rate": 8e-05,
      "loss": 1.5137,
      "step": 2337
    },
    {
      "epoch": 0.4855659397715473,
      "grad_norm": 0.575095534324646,
      "learning_rate": 8e-05,
      "loss": 1.5802,
      "step": 2338
    },
    {
      "epoch": 0.4857736240913811,
      "grad_norm": 0.55838942527771,
      "learning_rate": 8e-05,
      "loss": 1.6481,
      "step": 2339
    },
    {
      "epoch": 0.48598130841121495,
      "grad_norm": 0.5484670400619507,
      "learning_rate": 8e-05,
      "loss": 1.5421,
      "step": 2340
    },
    {
      "epoch": 0.4861889927310488,
      "grad_norm": 0.49879464507102966,
      "learning_rate": 8e-05,
      "loss": 1.5479,
      "step": 2341
    },
    {
      "epoch": 0.48639667705088263,
      "grad_norm": 0.514241099357605,
      "learning_rate": 8e-05,
      "loss": 1.6005,
      "step": 2342
    },
    {
      "epoch": 0.4866043613707165,
      "grad_norm": 0.5414352416992188,
      "learning_rate": 8e-05,
      "loss": 1.5833,
      "step": 2343
    },
    {
      "epoch": 0.48681204569055037,
      "grad_norm": 0.5201272368431091,
      "learning_rate": 8e-05,
      "loss": 1.6411,
      "step": 2344
    },
    {
      "epoch": 0.48701973001038423,
      "grad_norm": 0.5231587290763855,
      "learning_rate": 8e-05,
      "loss": 1.5735,
      "step": 2345
    },
    {
      "epoch": 0.48722741433021804,
      "grad_norm": 0.5916883945465088,
      "learning_rate": 8e-05,
      "loss": 1.654,
      "step": 2346
    },
    {
      "epoch": 0.4874350986500519,
      "grad_norm": 0.4834756553173065,
      "learning_rate": 8e-05,
      "loss": 1.5024,
      "step": 2347
    },
    {
      "epoch": 0.4876427829698858,
      "grad_norm": 0.5379783511161804,
      "learning_rate": 8e-05,
      "loss": 1.5899,
      "step": 2348
    },
    {
      "epoch": 0.48785046728971965,
      "grad_norm": 0.5307522416114807,
      "learning_rate": 8e-05,
      "loss": 1.6059,
      "step": 2349
    },
    {
      "epoch": 0.48805815160955346,
      "grad_norm": 0.5545095205307007,
      "learning_rate": 8e-05,
      "loss": 1.6131,
      "step": 2350
    },
    {
      "epoch": 0.4882658359293873,
      "grad_norm": 0.5184333920478821,
      "learning_rate": 8e-05,
      "loss": 1.5489,
      "step": 2351
    },
    {
      "epoch": 0.4884735202492212,
      "grad_norm": 0.5013905763626099,
      "learning_rate": 8e-05,
      "loss": 1.5485,
      "step": 2352
    },
    {
      "epoch": 0.48868120456905506,
      "grad_norm": 0.5201506614685059,
      "learning_rate": 8e-05,
      "loss": 1.5989,
      "step": 2353
    },
    {
      "epoch": 0.4888888888888889,
      "grad_norm": 0.6300234198570251,
      "learning_rate": 8e-05,
      "loss": 1.6314,
      "step": 2354
    },
    {
      "epoch": 0.48909657320872274,
      "grad_norm": 0.5080524682998657,
      "learning_rate": 8e-05,
      "loss": 1.5786,
      "step": 2355
    },
    {
      "epoch": 0.4893042575285566,
      "grad_norm": 0.5111988186836243,
      "learning_rate": 8e-05,
      "loss": 1.5494,
      "step": 2356
    },
    {
      "epoch": 0.4895119418483904,
      "grad_norm": 0.48312148451805115,
      "learning_rate": 8e-05,
      "loss": 1.5011,
      "step": 2357
    },
    {
      "epoch": 0.4897196261682243,
      "grad_norm": 0.5628437995910645,
      "learning_rate": 8e-05,
      "loss": 1.6562,
      "step": 2358
    },
    {
      "epoch": 0.48992731048805815,
      "grad_norm": 0.4936966598033905,
      "learning_rate": 8e-05,
      "loss": 1.5238,
      "step": 2359
    },
    {
      "epoch": 0.490134994807892,
      "grad_norm": 0.5063262581825256,
      "learning_rate": 8e-05,
      "loss": 1.5278,
      "step": 2360
    },
    {
      "epoch": 0.49034267912772583,
      "grad_norm": 0.5016118884086609,
      "learning_rate": 8e-05,
      "loss": 1.5497,
      "step": 2361
    },
    {
      "epoch": 0.4905503634475597,
      "grad_norm": 0.5366820096969604,
      "learning_rate": 8e-05,
      "loss": 1.5932,
      "step": 2362
    },
    {
      "epoch": 0.49075804776739357,
      "grad_norm": 0.5510308146476746,
      "learning_rate": 8e-05,
      "loss": 1.5999,
      "step": 2363
    },
    {
      "epoch": 0.49096573208722744,
      "grad_norm": 0.5173181891441345,
      "learning_rate": 8e-05,
      "loss": 1.5276,
      "step": 2364
    },
    {
      "epoch": 0.49117341640706125,
      "grad_norm": 0.5081064105033875,
      "learning_rate": 8e-05,
      "loss": 1.5833,
      "step": 2365
    },
    {
      "epoch": 0.4913811007268951,
      "grad_norm": 0.49907955527305603,
      "learning_rate": 8e-05,
      "loss": 1.5682,
      "step": 2366
    },
    {
      "epoch": 0.491588785046729,
      "grad_norm": 0.501561164855957,
      "learning_rate": 8e-05,
      "loss": 1.6015,
      "step": 2367
    },
    {
      "epoch": 0.49179646936656285,
      "grad_norm": 0.5394383668899536,
      "learning_rate": 8e-05,
      "loss": 1.5138,
      "step": 2368
    },
    {
      "epoch": 0.49200415368639666,
      "grad_norm": 0.5115194320678711,
      "learning_rate": 8e-05,
      "loss": 1.5979,
      "step": 2369
    },
    {
      "epoch": 0.49221183800623053,
      "grad_norm": 0.5246303081512451,
      "learning_rate": 8e-05,
      "loss": 1.6148,
      "step": 2370
    },
    {
      "epoch": 0.4924195223260644,
      "grad_norm": 0.491183340549469,
      "learning_rate": 8e-05,
      "loss": 1.5987,
      "step": 2371
    },
    {
      "epoch": 0.4926272066458982,
      "grad_norm": 0.5178416967391968,
      "learning_rate": 8e-05,
      "loss": 1.6278,
      "step": 2372
    },
    {
      "epoch": 0.4928348909657321,
      "grad_norm": 0.5169252753257751,
      "learning_rate": 8e-05,
      "loss": 1.5663,
      "step": 2373
    },
    {
      "epoch": 0.49304257528556594,
      "grad_norm": 0.48223575949668884,
      "learning_rate": 8e-05,
      "loss": 1.5656,
      "step": 2374
    },
    {
      "epoch": 0.4932502596053998,
      "grad_norm": 0.5244314670562744,
      "learning_rate": 8e-05,
      "loss": 1.5936,
      "step": 2375
    },
    {
      "epoch": 0.4934579439252336,
      "grad_norm": 0.5011839866638184,
      "learning_rate": 8e-05,
      "loss": 1.5836,
      "step": 2376
    },
    {
      "epoch": 0.4936656282450675,
      "grad_norm": 0.5012759566307068,
      "learning_rate": 8e-05,
      "loss": 1.4722,
      "step": 2377
    },
    {
      "epoch": 0.49387331256490136,
      "grad_norm": 0.5099151134490967,
      "learning_rate": 8e-05,
      "loss": 1.5995,
      "step": 2378
    },
    {
      "epoch": 0.4940809968847352,
      "grad_norm": 0.5069243311882019,
      "learning_rate": 8e-05,
      "loss": 1.5511,
      "step": 2379
    },
    {
      "epoch": 0.49428868120456904,
      "grad_norm": 0.48993173241615295,
      "learning_rate": 8e-05,
      "loss": 1.5201,
      "step": 2380
    },
    {
      "epoch": 0.4944963655244029,
      "grad_norm": 0.5133856534957886,
      "learning_rate": 8e-05,
      "loss": 1.6163,
      "step": 2381
    },
    {
      "epoch": 0.49470404984423677,
      "grad_norm": 0.48842230439186096,
      "learning_rate": 8e-05,
      "loss": 1.4761,
      "step": 2382
    },
    {
      "epoch": 0.49491173416407064,
      "grad_norm": 0.5133230090141296,
      "learning_rate": 8e-05,
      "loss": 1.607,
      "step": 2383
    },
    {
      "epoch": 0.49511941848390445,
      "grad_norm": 0.5300762057304382,
      "learning_rate": 8e-05,
      "loss": 1.6077,
      "step": 2384
    },
    {
      "epoch": 0.4953271028037383,
      "grad_norm": 0.5309485793113708,
      "learning_rate": 8e-05,
      "loss": 1.6195,
      "step": 2385
    },
    {
      "epoch": 0.4955347871235722,
      "grad_norm": 0.5198656320571899,
      "learning_rate": 8e-05,
      "loss": 1.586,
      "step": 2386
    },
    {
      "epoch": 0.495742471443406,
      "grad_norm": 0.5134472846984863,
      "learning_rate": 8e-05,
      "loss": 1.5686,
      "step": 2387
    },
    {
      "epoch": 0.49595015576323986,
      "grad_norm": 0.5401621460914612,
      "learning_rate": 8e-05,
      "loss": 1.5476,
      "step": 2388
    },
    {
      "epoch": 0.49615784008307373,
      "grad_norm": 0.5133628249168396,
      "learning_rate": 8e-05,
      "loss": 1.5707,
      "step": 2389
    },
    {
      "epoch": 0.4963655244029076,
      "grad_norm": 0.5219677686691284,
      "learning_rate": 8e-05,
      "loss": 1.5748,
      "step": 2390
    },
    {
      "epoch": 0.4965732087227414,
      "grad_norm": 0.5132991671562195,
      "learning_rate": 8e-05,
      "loss": 1.5675,
      "step": 2391
    },
    {
      "epoch": 0.4967808930425753,
      "grad_norm": 0.5127162933349609,
      "learning_rate": 8e-05,
      "loss": 1.5678,
      "step": 2392
    },
    {
      "epoch": 0.49698857736240915,
      "grad_norm": 0.5062135457992554,
      "learning_rate": 8e-05,
      "loss": 1.5377,
      "step": 2393
    },
    {
      "epoch": 0.497196261682243,
      "grad_norm": 0.5313796997070312,
      "learning_rate": 8e-05,
      "loss": 1.6169,
      "step": 2394
    },
    {
      "epoch": 0.4974039460020768,
      "grad_norm": 0.5403634309768677,
      "learning_rate": 8e-05,
      "loss": 1.5679,
      "step": 2395
    },
    {
      "epoch": 0.4976116303219107,
      "grad_norm": 0.5055514574050903,
      "learning_rate": 8e-05,
      "loss": 1.5609,
      "step": 2396
    },
    {
      "epoch": 0.49781931464174456,
      "grad_norm": 0.4901556372642517,
      "learning_rate": 8e-05,
      "loss": 1.5681,
      "step": 2397
    },
    {
      "epoch": 0.4980269989615784,
      "grad_norm": 0.5094218254089355,
      "learning_rate": 8e-05,
      "loss": 1.5921,
      "step": 2398
    },
    {
      "epoch": 0.49823468328141224,
      "grad_norm": 0.5202116370201111,
      "learning_rate": 8e-05,
      "loss": 1.604,
      "step": 2399
    },
    {
      "epoch": 0.4984423676012461,
      "grad_norm": 0.4881117343902588,
      "learning_rate": 8e-05,
      "loss": 1.4603,
      "step": 2400
    },
    {
      "epoch": 0.49865005192108,
      "grad_norm": 0.5098052620887756,
      "learning_rate": 8e-05,
      "loss": 1.525,
      "step": 2401
    },
    {
      "epoch": 0.4988577362409138,
      "grad_norm": 0.5244508385658264,
      "learning_rate": 8e-05,
      "loss": 1.632,
      "step": 2402
    },
    {
      "epoch": 0.49906542056074765,
      "grad_norm": 0.5339807271957397,
      "learning_rate": 8e-05,
      "loss": 1.7082,
      "step": 2403
    },
    {
      "epoch": 0.4992731048805815,
      "grad_norm": 0.5100862979888916,
      "learning_rate": 8e-05,
      "loss": 1.5351,
      "step": 2404
    },
    {
      "epoch": 0.4994807892004154,
      "grad_norm": 0.5070978999137878,
      "learning_rate": 8e-05,
      "loss": 1.5406,
      "step": 2405
    },
    {
      "epoch": 0.4996884735202492,
      "grad_norm": 0.514641523361206,
      "learning_rate": 8e-05,
      "loss": 1.6159,
      "step": 2406
    },
    {
      "epoch": 0.49989615784008307,
      "grad_norm": 0.5069220662117004,
      "learning_rate": 8e-05,
      "loss": 1.6052,
      "step": 2407
    },
    {
      "epoch": 0.5001038421599169,
      "grad_norm": 0.526537299156189,
      "learning_rate": 8e-05,
      "loss": 1.5635,
      "step": 2408
    },
    {
      "epoch": 0.5003115264797507,
      "grad_norm": 0.5100792050361633,
      "learning_rate": 8e-05,
      "loss": 1.5645,
      "step": 2409
    },
    {
      "epoch": 0.5005192107995846,
      "grad_norm": 0.4965640902519226,
      "learning_rate": 8e-05,
      "loss": 1.4884,
      "step": 2410
    },
    {
      "epoch": 0.5007268951194185,
      "grad_norm": 0.5005423426628113,
      "learning_rate": 8e-05,
      "loss": 1.5408,
      "step": 2411
    },
    {
      "epoch": 0.5009345794392523,
      "grad_norm": 0.5040135979652405,
      "learning_rate": 8e-05,
      "loss": 1.5669,
      "step": 2412
    },
    {
      "epoch": 0.5011422637590862,
      "grad_norm": 0.4898592233657837,
      "learning_rate": 8e-05,
      "loss": 1.505,
      "step": 2413
    },
    {
      "epoch": 0.5013499480789201,
      "grad_norm": 0.5353949666023254,
      "learning_rate": 8e-05,
      "loss": 1.5671,
      "step": 2414
    },
    {
      "epoch": 0.5015576323987538,
      "grad_norm": 0.5377967953681946,
      "learning_rate": 8e-05,
      "loss": 1.6208,
      "step": 2415
    },
    {
      "epoch": 0.5017653167185877,
      "grad_norm": 0.5099054574966431,
      "learning_rate": 8e-05,
      "loss": 1.5954,
      "step": 2416
    },
    {
      "epoch": 0.5019730010384216,
      "grad_norm": 0.5304335355758667,
      "learning_rate": 8e-05,
      "loss": 1.5279,
      "step": 2417
    },
    {
      "epoch": 0.5021806853582554,
      "grad_norm": 0.5465752482414246,
      "learning_rate": 8e-05,
      "loss": 1.595,
      "step": 2418
    },
    {
      "epoch": 0.5023883696780893,
      "grad_norm": 0.49918806552886963,
      "learning_rate": 8e-05,
      "loss": 1.5669,
      "step": 2419
    },
    {
      "epoch": 0.5025960539979232,
      "grad_norm": 0.5252723693847656,
      "learning_rate": 8e-05,
      "loss": 1.527,
      "step": 2420
    },
    {
      "epoch": 0.502803738317757,
      "grad_norm": 0.5622281432151794,
      "learning_rate": 8e-05,
      "loss": 1.6202,
      "step": 2421
    },
    {
      "epoch": 0.5030114226375909,
      "grad_norm": 0.5915151834487915,
      "learning_rate": 8e-05,
      "loss": 1.6003,
      "step": 2422
    },
    {
      "epoch": 0.5032191069574247,
      "grad_norm": 0.5138158202171326,
      "learning_rate": 8e-05,
      "loss": 1.5566,
      "step": 2423
    },
    {
      "epoch": 0.5034267912772585,
      "grad_norm": 0.5394647121429443,
      "learning_rate": 8e-05,
      "loss": 1.5688,
      "step": 2424
    },
    {
      "epoch": 0.5036344755970924,
      "grad_norm": 0.5186442732810974,
      "learning_rate": 8e-05,
      "loss": 1.5793,
      "step": 2425
    },
    {
      "epoch": 0.5038421599169263,
      "grad_norm": 0.5293716192245483,
      "learning_rate": 8e-05,
      "loss": 1.5513,
      "step": 2426
    },
    {
      "epoch": 0.5040498442367601,
      "grad_norm": 0.524944007396698,
      "learning_rate": 8e-05,
      "loss": 1.5179,
      "step": 2427
    },
    {
      "epoch": 0.504257528556594,
      "grad_norm": 0.5945055484771729,
      "learning_rate": 8e-05,
      "loss": 1.5778,
      "step": 2428
    },
    {
      "epoch": 0.5044652128764279,
      "grad_norm": 0.5441060662269592,
      "learning_rate": 8e-05,
      "loss": 1.5757,
      "step": 2429
    },
    {
      "epoch": 0.5046728971962616,
      "grad_norm": 0.5250440239906311,
      "learning_rate": 8e-05,
      "loss": 1.5304,
      "step": 2430
    },
    {
      "epoch": 0.5048805815160955,
      "grad_norm": 0.5178036689758301,
      "learning_rate": 8e-05,
      "loss": 1.5371,
      "step": 2431
    },
    {
      "epoch": 0.5050882658359294,
      "grad_norm": 0.5340495705604553,
      "learning_rate": 8e-05,
      "loss": 1.5492,
      "step": 2432
    },
    {
      "epoch": 0.5052959501557632,
      "grad_norm": 0.5054112076759338,
      "learning_rate": 8e-05,
      "loss": 1.5761,
      "step": 2433
    },
    {
      "epoch": 0.5055036344755971,
      "grad_norm": 0.5301451086997986,
      "learning_rate": 8e-05,
      "loss": 1.599,
      "step": 2434
    },
    {
      "epoch": 0.505711318795431,
      "grad_norm": 0.5750793814659119,
      "learning_rate": 8e-05,
      "loss": 1.6637,
      "step": 2435
    },
    {
      "epoch": 0.5059190031152648,
      "grad_norm": 0.5496008992195129,
      "learning_rate": 8e-05,
      "loss": 1.5936,
      "step": 2436
    },
    {
      "epoch": 0.5061266874350987,
      "grad_norm": 0.5835946202278137,
      "learning_rate": 8e-05,
      "loss": 1.5705,
      "step": 2437
    },
    {
      "epoch": 0.5063343717549325,
      "grad_norm": 0.5562037825584412,
      "learning_rate": 8e-05,
      "loss": 1.5426,
      "step": 2438
    },
    {
      "epoch": 0.5065420560747663,
      "grad_norm": 0.49563848972320557,
      "learning_rate": 8e-05,
      "loss": 1.563,
      "step": 2439
    },
    {
      "epoch": 0.5067497403946002,
      "grad_norm": 0.533659815788269,
      "learning_rate": 8e-05,
      "loss": 1.5853,
      "step": 2440
    },
    {
      "epoch": 0.5069574247144341,
      "grad_norm": 0.5142748355865479,
      "learning_rate": 8e-05,
      "loss": 1.4993,
      "step": 2441
    },
    {
      "epoch": 0.5071651090342679,
      "grad_norm": 0.5077499747276306,
      "learning_rate": 8e-05,
      "loss": 1.5803,
      "step": 2442
    },
    {
      "epoch": 0.5073727933541018,
      "grad_norm": 0.5426058769226074,
      "learning_rate": 8e-05,
      "loss": 1.59,
      "step": 2443
    },
    {
      "epoch": 0.5075804776739357,
      "grad_norm": 0.4893079996109009,
      "learning_rate": 8e-05,
      "loss": 1.4893,
      "step": 2444
    },
    {
      "epoch": 0.5077881619937694,
      "grad_norm": 0.5184099078178406,
      "learning_rate": 8e-05,
      "loss": 1.5245,
      "step": 2445
    },
    {
      "epoch": 0.5079958463136033,
      "grad_norm": 0.5037515759468079,
      "learning_rate": 8e-05,
      "loss": 1.482,
      "step": 2446
    },
    {
      "epoch": 0.5082035306334372,
      "grad_norm": 0.5105049014091492,
      "learning_rate": 8e-05,
      "loss": 1.5913,
      "step": 2447
    },
    {
      "epoch": 0.508411214953271,
      "grad_norm": 0.513841450214386,
      "learning_rate": 8e-05,
      "loss": 1.5031,
      "step": 2448
    },
    {
      "epoch": 0.5086188992731049,
      "grad_norm": 0.5362356305122375,
      "learning_rate": 8e-05,
      "loss": 1.5876,
      "step": 2449
    },
    {
      "epoch": 0.5088265835929388,
      "grad_norm": 0.5155436396598816,
      "learning_rate": 8e-05,
      "loss": 1.5723,
      "step": 2450
    },
    {
      "epoch": 0.5090342679127726,
      "grad_norm": 0.5344365239143372,
      "learning_rate": 8e-05,
      "loss": 1.5596,
      "step": 2451
    },
    {
      "epoch": 0.5092419522326065,
      "grad_norm": 0.5077399611473083,
      "learning_rate": 8e-05,
      "loss": 1.6037,
      "step": 2452
    },
    {
      "epoch": 0.5094496365524402,
      "grad_norm": 0.5176429748535156,
      "learning_rate": 8e-05,
      "loss": 1.5707,
      "step": 2453
    },
    {
      "epoch": 0.5096573208722741,
      "grad_norm": 0.5297008156776428,
      "learning_rate": 8e-05,
      "loss": 1.5247,
      "step": 2454
    },
    {
      "epoch": 0.509865005192108,
      "grad_norm": 0.5335763096809387,
      "learning_rate": 8e-05,
      "loss": 1.5811,
      "step": 2455
    },
    {
      "epoch": 0.5100726895119418,
      "grad_norm": 0.5081517100334167,
      "learning_rate": 8e-05,
      "loss": 1.5559,
      "step": 2456
    },
    {
      "epoch": 0.5102803738317757,
      "grad_norm": 0.5149271488189697,
      "learning_rate": 8e-05,
      "loss": 1.5795,
      "step": 2457
    },
    {
      "epoch": 0.5104880581516096,
      "grad_norm": 0.49932461977005005,
      "learning_rate": 8e-05,
      "loss": 1.5799,
      "step": 2458
    },
    {
      "epoch": 0.5106957424714434,
      "grad_norm": 0.5233448147773743,
      "learning_rate": 8e-05,
      "loss": 1.5389,
      "step": 2459
    },
    {
      "epoch": 0.5109034267912772,
      "grad_norm": 0.53509521484375,
      "learning_rate": 8e-05,
      "loss": 1.6398,
      "step": 2460
    },
    {
      "epoch": 0.5111111111111111,
      "grad_norm": 0.551110565662384,
      "learning_rate": 8e-05,
      "loss": 1.535,
      "step": 2461
    },
    {
      "epoch": 0.5113187954309449,
      "grad_norm": 0.5153088569641113,
      "learning_rate": 8e-05,
      "loss": 1.5462,
      "step": 2462
    },
    {
      "epoch": 0.5115264797507788,
      "grad_norm": 0.5195600390434265,
      "learning_rate": 8e-05,
      "loss": 1.5478,
      "step": 2463
    },
    {
      "epoch": 0.5117341640706127,
      "grad_norm": 0.5105565786361694,
      "learning_rate": 8e-05,
      "loss": 1.5332,
      "step": 2464
    },
    {
      "epoch": 0.5119418483904465,
      "grad_norm": 0.5178881287574768,
      "learning_rate": 8e-05,
      "loss": 1.5375,
      "step": 2465
    },
    {
      "epoch": 0.5121495327102804,
      "grad_norm": 0.5123568773269653,
      "learning_rate": 8e-05,
      "loss": 1.6039,
      "step": 2466
    },
    {
      "epoch": 0.5123572170301143,
      "grad_norm": 0.512981653213501,
      "learning_rate": 8e-05,
      "loss": 1.5822,
      "step": 2467
    },
    {
      "epoch": 0.512564901349948,
      "grad_norm": 0.5098233819007874,
      "learning_rate": 8e-05,
      "loss": 1.6099,
      "step": 2468
    },
    {
      "epoch": 0.5127725856697819,
      "grad_norm": 0.5267816781997681,
      "learning_rate": 8e-05,
      "loss": 1.5593,
      "step": 2469
    },
    {
      "epoch": 0.5129802699896158,
      "grad_norm": 0.517514705657959,
      "learning_rate": 8e-05,
      "loss": 1.5874,
      "step": 2470
    },
    {
      "epoch": 0.5131879543094496,
      "grad_norm": 0.49412310123443604,
      "learning_rate": 8e-05,
      "loss": 1.5829,
      "step": 2471
    },
    {
      "epoch": 0.5133956386292835,
      "grad_norm": 0.49666720628738403,
      "learning_rate": 8e-05,
      "loss": 1.5223,
      "step": 2472
    },
    {
      "epoch": 0.5136033229491174,
      "grad_norm": 0.511103630065918,
      "learning_rate": 8e-05,
      "loss": 1.5971,
      "step": 2473
    },
    {
      "epoch": 0.5138110072689512,
      "grad_norm": 0.49603182077407837,
      "learning_rate": 8e-05,
      "loss": 1.6282,
      "step": 2474
    },
    {
      "epoch": 0.514018691588785,
      "grad_norm": 0.4902198016643524,
      "learning_rate": 8e-05,
      "loss": 1.6366,
      "step": 2475
    },
    {
      "epoch": 0.5142263759086189,
      "grad_norm": 0.5027957558631897,
      "learning_rate": 8e-05,
      "loss": 1.5777,
      "step": 2476
    },
    {
      "epoch": 0.5144340602284527,
      "grad_norm": 0.5001282095909119,
      "learning_rate": 8e-05,
      "loss": 1.4579,
      "step": 2477
    },
    {
      "epoch": 0.5146417445482866,
      "grad_norm": 0.49658671021461487,
      "learning_rate": 8e-05,
      "loss": 1.5798,
      "step": 2478
    },
    {
      "epoch": 0.5148494288681205,
      "grad_norm": 0.5023085474967957,
      "learning_rate": 8e-05,
      "loss": 1.544,
      "step": 2479
    },
    {
      "epoch": 0.5150571131879543,
      "grad_norm": 0.5538403391838074,
      "learning_rate": 8e-05,
      "loss": 1.5892,
      "step": 2480
    },
    {
      "epoch": 0.5152647975077882,
      "grad_norm": 0.5351791381835938,
      "learning_rate": 8e-05,
      "loss": 1.6152,
      "step": 2481
    },
    {
      "epoch": 0.5154724818276221,
      "grad_norm": 0.5126028060913086,
      "learning_rate": 8e-05,
      "loss": 1.5706,
      "step": 2482
    },
    {
      "epoch": 0.5156801661474558,
      "grad_norm": 0.509198009967804,
      "learning_rate": 8e-05,
      "loss": 1.5492,
      "step": 2483
    },
    {
      "epoch": 0.5158878504672897,
      "grad_norm": 0.5237105488777161,
      "learning_rate": 8e-05,
      "loss": 1.5066,
      "step": 2484
    },
    {
      "epoch": 0.5160955347871236,
      "grad_norm": 0.49990713596343994,
      "learning_rate": 8e-05,
      "loss": 1.5941,
      "step": 2485
    },
    {
      "epoch": 0.5163032191069574,
      "grad_norm": 0.5106729865074158,
      "learning_rate": 8e-05,
      "loss": 1.5571,
      "step": 2486
    },
    {
      "epoch": 0.5165109034267913,
      "grad_norm": 0.5249480605125427,
      "learning_rate": 8e-05,
      "loss": 1.5523,
      "step": 2487
    },
    {
      "epoch": 0.5167185877466252,
      "grad_norm": 0.49067258834838867,
      "learning_rate": 8e-05,
      "loss": 1.6248,
      "step": 2488
    },
    {
      "epoch": 0.516926272066459,
      "grad_norm": 0.5179197192192078,
      "learning_rate": 8e-05,
      "loss": 1.5817,
      "step": 2489
    },
    {
      "epoch": 0.5171339563862928,
      "grad_norm": 0.5394155979156494,
      "learning_rate": 8e-05,
      "loss": 1.5518,
      "step": 2490
    },
    {
      "epoch": 0.5173416407061266,
      "grad_norm": 0.49925529956817627,
      "learning_rate": 8e-05,
      "loss": 1.5646,
      "step": 2491
    },
    {
      "epoch": 0.5175493250259605,
      "grad_norm": 0.5597730278968811,
      "learning_rate": 8e-05,
      "loss": 1.6009,
      "step": 2492
    },
    {
      "epoch": 0.5177570093457944,
      "grad_norm": 0.5028907060623169,
      "learning_rate": 8e-05,
      "loss": 1.55,
      "step": 2493
    },
    {
      "epoch": 0.5179646936656283,
      "grad_norm": 0.5173121690750122,
      "learning_rate": 8e-05,
      "loss": 1.5156,
      "step": 2494
    },
    {
      "epoch": 0.5181723779854621,
      "grad_norm": 0.5130355954170227,
      "learning_rate": 8e-05,
      "loss": 1.6179,
      "step": 2495
    },
    {
      "epoch": 0.518380062305296,
      "grad_norm": 0.492567241191864,
      "learning_rate": 8e-05,
      "loss": 1.5069,
      "step": 2496
    },
    {
      "epoch": 0.5185877466251299,
      "grad_norm": 0.5323511958122253,
      "learning_rate": 8e-05,
      "loss": 1.5661,
      "step": 2497
    },
    {
      "epoch": 0.5187954309449636,
      "grad_norm": 0.5106134414672852,
      "learning_rate": 8e-05,
      "loss": 1.5629,
      "step": 2498
    },
    {
      "epoch": 0.5190031152647975,
      "grad_norm": 0.5243561863899231,
      "learning_rate": 8e-05,
      "loss": 1.5809,
      "step": 2499
    },
    {
      "epoch": 0.5192107995846313,
      "grad_norm": 0.5350177884101868,
      "learning_rate": 8e-05,
      "loss": 1.5558,
      "step": 2500
    },
    {
      "epoch": 0.5194184839044652,
      "grad_norm": 0.4913942217826843,
      "learning_rate": 8e-05,
      "loss": 1.5466,
      "step": 2501
    },
    {
      "epoch": 0.5196261682242991,
      "grad_norm": 0.5095742344856262,
      "learning_rate": 8e-05,
      "loss": 1.5512,
      "step": 2502
    },
    {
      "epoch": 0.519833852544133,
      "grad_norm": 0.5109645128250122,
      "learning_rate": 8e-05,
      "loss": 1.5481,
      "step": 2503
    },
    {
      "epoch": 0.5200415368639668,
      "grad_norm": 0.5052945017814636,
      "learning_rate": 8e-05,
      "loss": 1.5722,
      "step": 2504
    },
    {
      "epoch": 0.5202492211838006,
      "grad_norm": 0.5314079523086548,
      "learning_rate": 8e-05,
      "loss": 1.5591,
      "step": 2505
    },
    {
      "epoch": 0.5204569055036344,
      "grad_norm": 0.5043820738792419,
      "learning_rate": 8e-05,
      "loss": 1.5327,
      "step": 2506
    },
    {
      "epoch": 0.5206645898234683,
      "grad_norm": 0.5031879544258118,
      "learning_rate": 8e-05,
      "loss": 1.4327,
      "step": 2507
    },
    {
      "epoch": 0.5208722741433022,
      "grad_norm": 0.5129706859588623,
      "learning_rate": 8e-05,
      "loss": 1.5805,
      "step": 2508
    },
    {
      "epoch": 0.521079958463136,
      "grad_norm": 0.5188937187194824,
      "learning_rate": 8e-05,
      "loss": 1.5256,
      "step": 2509
    },
    {
      "epoch": 0.5212876427829699,
      "grad_norm": 0.5003663897514343,
      "learning_rate": 8e-05,
      "loss": 1.536,
      "step": 2510
    },
    {
      "epoch": 0.5214953271028038,
      "grad_norm": 0.5461739897727966,
      "learning_rate": 8e-05,
      "loss": 1.5779,
      "step": 2511
    },
    {
      "epoch": 0.5217030114226376,
      "grad_norm": 0.5200935006141663,
      "learning_rate": 8e-05,
      "loss": 1.6121,
      "step": 2512
    },
    {
      "epoch": 0.5219106957424714,
      "grad_norm": 0.5046533942222595,
      "learning_rate": 8e-05,
      "loss": 1.5157,
      "step": 2513
    },
    {
      "epoch": 0.5221183800623053,
      "grad_norm": 0.5248345136642456,
      "learning_rate": 8e-05,
      "loss": 1.6406,
      "step": 2514
    },
    {
      "epoch": 0.5223260643821391,
      "grad_norm": 0.5107331871986389,
      "learning_rate": 8e-05,
      "loss": 1.5285,
      "step": 2515
    },
    {
      "epoch": 0.522533748701973,
      "grad_norm": 0.532707691192627,
      "learning_rate": 8e-05,
      "loss": 1.5672,
      "step": 2516
    },
    {
      "epoch": 0.5227414330218069,
      "grad_norm": 0.5015535354614258,
      "learning_rate": 8e-05,
      "loss": 1.6222,
      "step": 2517
    },
    {
      "epoch": 0.5229491173416407,
      "grad_norm": 0.49384546279907227,
      "learning_rate": 8e-05,
      "loss": 1.5419,
      "step": 2518
    },
    {
      "epoch": 0.5231568016614746,
      "grad_norm": 0.5301195383071899,
      "learning_rate": 8e-05,
      "loss": 1.5981,
      "step": 2519
    },
    {
      "epoch": 0.5233644859813084,
      "grad_norm": 0.5169888138771057,
      "learning_rate": 8e-05,
      "loss": 1.5472,
      "step": 2520
    },
    {
      "epoch": 0.5235721703011422,
      "grad_norm": 0.5029063820838928,
      "learning_rate": 8e-05,
      "loss": 1.588,
      "step": 2521
    },
    {
      "epoch": 0.5237798546209761,
      "grad_norm": 0.5058553814888,
      "learning_rate": 8e-05,
      "loss": 1.5353,
      "step": 2522
    },
    {
      "epoch": 0.52398753894081,
      "grad_norm": 0.5162726640701294,
      "learning_rate": 8e-05,
      "loss": 1.5503,
      "step": 2523
    },
    {
      "epoch": 0.5241952232606438,
      "grad_norm": 0.526066780090332,
      "learning_rate": 8e-05,
      "loss": 1.5831,
      "step": 2524
    },
    {
      "epoch": 0.5244029075804777,
      "grad_norm": 0.5261390805244446,
      "learning_rate": 8e-05,
      "loss": 1.5507,
      "step": 2525
    },
    {
      "epoch": 0.5246105919003116,
      "grad_norm": 0.515093207359314,
      "learning_rate": 8e-05,
      "loss": 1.539,
      "step": 2526
    },
    {
      "epoch": 0.5248182762201454,
      "grad_norm": 0.5058615803718567,
      "learning_rate": 8e-05,
      "loss": 1.6147,
      "step": 2527
    },
    {
      "epoch": 0.5250259605399792,
      "grad_norm": 0.5275096297264099,
      "learning_rate": 8e-05,
      "loss": 1.5747,
      "step": 2528
    },
    {
      "epoch": 0.525233644859813,
      "grad_norm": 0.521035373210907,
      "learning_rate": 8e-05,
      "loss": 1.531,
      "step": 2529
    },
    {
      "epoch": 0.5254413291796469,
      "grad_norm": 0.525314211845398,
      "learning_rate": 8e-05,
      "loss": 1.5806,
      "step": 2530
    },
    {
      "epoch": 0.5256490134994808,
      "grad_norm": 0.5049483180046082,
      "learning_rate": 8e-05,
      "loss": 1.5496,
      "step": 2531
    },
    {
      "epoch": 0.5258566978193147,
      "grad_norm": 0.5101779103279114,
      "learning_rate": 8e-05,
      "loss": 1.5434,
      "step": 2532
    },
    {
      "epoch": 0.5260643821391485,
      "grad_norm": 0.5049511194229126,
      "learning_rate": 8e-05,
      "loss": 1.5017,
      "step": 2533
    },
    {
      "epoch": 0.5262720664589824,
      "grad_norm": 0.5197418332099915,
      "learning_rate": 8e-05,
      "loss": 1.5678,
      "step": 2534
    },
    {
      "epoch": 0.5264797507788161,
      "grad_norm": 0.5398335456848145,
      "learning_rate": 8e-05,
      "loss": 1.6633,
      "step": 2535
    },
    {
      "epoch": 0.52668743509865,
      "grad_norm": 0.506309986114502,
      "learning_rate": 8e-05,
      "loss": 1.4991,
      "step": 2536
    },
    {
      "epoch": 0.5268951194184839,
      "grad_norm": 0.515005886554718,
      "learning_rate": 8e-05,
      "loss": 1.5842,
      "step": 2537
    },
    {
      "epoch": 0.5271028037383177,
      "grad_norm": 0.49161919951438904,
      "learning_rate": 8e-05,
      "loss": 1.5945,
      "step": 2538
    },
    {
      "epoch": 0.5273104880581516,
      "grad_norm": 0.4888719320297241,
      "learning_rate": 8e-05,
      "loss": 1.5212,
      "step": 2539
    },
    {
      "epoch": 0.5275181723779855,
      "grad_norm": 0.4999425709247589,
      "learning_rate": 8e-05,
      "loss": 1.5424,
      "step": 2540
    },
    {
      "epoch": 0.5277258566978194,
      "grad_norm": 0.5267025232315063,
      "learning_rate": 8e-05,
      "loss": 1.5701,
      "step": 2541
    },
    {
      "epoch": 0.5279335410176532,
      "grad_norm": 0.5228183269500732,
      "learning_rate": 8e-05,
      "loss": 1.5789,
      "step": 2542
    },
    {
      "epoch": 0.528141225337487,
      "grad_norm": 0.5197039842605591,
      "learning_rate": 8e-05,
      "loss": 1.592,
      "step": 2543
    },
    {
      "epoch": 0.5283489096573208,
      "grad_norm": 0.5103591084480286,
      "learning_rate": 8e-05,
      "loss": 1.5852,
      "step": 2544
    },
    {
      "epoch": 0.5285565939771547,
      "grad_norm": 0.4849054515361786,
      "learning_rate": 8e-05,
      "loss": 1.5137,
      "step": 2545
    },
    {
      "epoch": 0.5287642782969886,
      "grad_norm": 0.5198196172714233,
      "learning_rate": 8e-05,
      "loss": 1.5192,
      "step": 2546
    },
    {
      "epoch": 0.5289719626168224,
      "grad_norm": 0.49438804388046265,
      "learning_rate": 8e-05,
      "loss": 1.5662,
      "step": 2547
    },
    {
      "epoch": 0.5291796469366563,
      "grad_norm": 0.5067175626754761,
      "learning_rate": 8e-05,
      "loss": 1.5503,
      "step": 2548
    },
    {
      "epoch": 0.5293873312564902,
      "grad_norm": 0.5209910273551941,
      "learning_rate": 8e-05,
      "loss": 1.5551,
      "step": 2549
    },
    {
      "epoch": 0.5295950155763239,
      "grad_norm": 0.5711840391159058,
      "learning_rate": 8e-05,
      "loss": 1.6114,
      "step": 2550
    },
    {
      "epoch": 0.5298026998961578,
      "grad_norm": 0.5297765135765076,
      "learning_rate": 8e-05,
      "loss": 1.4573,
      "step": 2551
    },
    {
      "epoch": 0.5300103842159917,
      "grad_norm": 0.5275728106498718,
      "learning_rate": 8e-05,
      "loss": 1.5492,
      "step": 2552
    },
    {
      "epoch": 0.5302180685358255,
      "grad_norm": 0.513297438621521,
      "learning_rate": 8e-05,
      "loss": 1.5549,
      "step": 2553
    },
    {
      "epoch": 0.5304257528556594,
      "grad_norm": 0.5029946565628052,
      "learning_rate": 8e-05,
      "loss": 1.5527,
      "step": 2554
    },
    {
      "epoch": 0.5306334371754933,
      "grad_norm": 0.56758713722229,
      "learning_rate": 8e-05,
      "loss": 1.6427,
      "step": 2555
    },
    {
      "epoch": 0.5308411214953271,
      "grad_norm": 0.5123136639595032,
      "learning_rate": 8e-05,
      "loss": 1.561,
      "step": 2556
    },
    {
      "epoch": 0.531048805815161,
      "grad_norm": 0.5428341627120972,
      "learning_rate": 8e-05,
      "loss": 1.6549,
      "step": 2557
    },
    {
      "epoch": 0.5312564901349948,
      "grad_norm": 0.5145912766456604,
      "learning_rate": 8e-05,
      "loss": 1.6259,
      "step": 2558
    },
    {
      "epoch": 0.5314641744548286,
      "grad_norm": 0.5414952635765076,
      "learning_rate": 8e-05,
      "loss": 1.5668,
      "step": 2559
    },
    {
      "epoch": 0.5316718587746625,
      "grad_norm": 0.5508964657783508,
      "learning_rate": 8e-05,
      "loss": 1.4993,
      "step": 2560
    },
    {
      "epoch": 0.5318795430944964,
      "grad_norm": 0.501332700252533,
      "learning_rate": 8e-05,
      "loss": 1.5632,
      "step": 2561
    },
    {
      "epoch": 0.5320872274143302,
      "grad_norm": 0.4879615902900696,
      "learning_rate": 8e-05,
      "loss": 1.4787,
      "step": 2562
    },
    {
      "epoch": 0.5322949117341641,
      "grad_norm": 0.5425028800964355,
      "learning_rate": 8e-05,
      "loss": 1.5681,
      "step": 2563
    },
    {
      "epoch": 0.532502596053998,
      "grad_norm": 0.5095462203025818,
      "learning_rate": 8e-05,
      "loss": 1.543,
      "step": 2564
    },
    {
      "epoch": 0.5327102803738317,
      "grad_norm": 0.5352701544761658,
      "learning_rate": 8e-05,
      "loss": 1.6086,
      "step": 2565
    },
    {
      "epoch": 0.5329179646936656,
      "grad_norm": 0.4908084571361542,
      "learning_rate": 8e-05,
      "loss": 1.5646,
      "step": 2566
    },
    {
      "epoch": 0.5331256490134995,
      "grad_norm": 0.551490306854248,
      "learning_rate": 8e-05,
      "loss": 1.5477,
      "step": 2567
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 0.5218336582183838,
      "learning_rate": 8e-05,
      "loss": 1.5666,
      "step": 2568
    },
    {
      "epoch": 0.5335410176531672,
      "grad_norm": 0.5142163634300232,
      "learning_rate": 8e-05,
      "loss": 1.6095,
      "step": 2569
    },
    {
      "epoch": 0.5337487019730011,
      "grad_norm": 0.5136362910270691,
      "learning_rate": 8e-05,
      "loss": 1.5997,
      "step": 2570
    },
    {
      "epoch": 0.5339563862928349,
      "grad_norm": 0.539950966835022,
      "learning_rate": 8e-05,
      "loss": 1.6219,
      "step": 2571
    },
    {
      "epoch": 0.5341640706126688,
      "grad_norm": 0.5029425621032715,
      "learning_rate": 8e-05,
      "loss": 1.5646,
      "step": 2572
    },
    {
      "epoch": 0.5343717549325026,
      "grad_norm": 0.562980055809021,
      "learning_rate": 8e-05,
      "loss": 1.5901,
      "step": 2573
    },
    {
      "epoch": 0.5345794392523364,
      "grad_norm": 0.5117045044898987,
      "learning_rate": 8e-05,
      "loss": 1.5533,
      "step": 2574
    },
    {
      "epoch": 0.5347871235721703,
      "grad_norm": 0.5418522953987122,
      "learning_rate": 8e-05,
      "loss": 1.5846,
      "step": 2575
    },
    {
      "epoch": 0.5349948078920042,
      "grad_norm": 0.5144622921943665,
      "learning_rate": 8e-05,
      "loss": 1.6386,
      "step": 2576
    },
    {
      "epoch": 0.535202492211838,
      "grad_norm": 0.5126820206642151,
      "learning_rate": 8e-05,
      "loss": 1.524,
      "step": 2577
    },
    {
      "epoch": 0.5354101765316719,
      "grad_norm": 0.5278242826461792,
      "learning_rate": 8e-05,
      "loss": 1.6653,
      "step": 2578
    },
    {
      "epoch": 0.5356178608515058,
      "grad_norm": 0.5213025212287903,
      "learning_rate": 8e-05,
      "loss": 1.5981,
      "step": 2579
    },
    {
      "epoch": 0.5358255451713395,
      "grad_norm": 0.5076565146446228,
      "learning_rate": 8e-05,
      "loss": 1.5918,
      "step": 2580
    },
    {
      "epoch": 0.5360332294911734,
      "grad_norm": 0.5242885947227478,
      "learning_rate": 8e-05,
      "loss": 1.5806,
      "step": 2581
    },
    {
      "epoch": 0.5362409138110072,
      "grad_norm": 0.5121769905090332,
      "learning_rate": 8e-05,
      "loss": 1.4982,
      "step": 2582
    },
    {
      "epoch": 0.5364485981308411,
      "grad_norm": 0.5083255767822266,
      "learning_rate": 8e-05,
      "loss": 1.5928,
      "step": 2583
    },
    {
      "epoch": 0.536656282450675,
      "grad_norm": 0.4839821755886078,
      "learning_rate": 8e-05,
      "loss": 1.5292,
      "step": 2584
    },
    {
      "epoch": 0.5368639667705088,
      "grad_norm": 0.5013673901557922,
      "learning_rate": 8e-05,
      "loss": 1.5556,
      "step": 2585
    },
    {
      "epoch": 0.5370716510903427,
      "grad_norm": 0.5166984796524048,
      "learning_rate": 8e-05,
      "loss": 1.5705,
      "step": 2586
    },
    {
      "epoch": 0.5372793354101766,
      "grad_norm": 0.5307886004447937,
      "learning_rate": 8e-05,
      "loss": 1.5807,
      "step": 2587
    },
    {
      "epoch": 0.5374870197300103,
      "grad_norm": 0.5106934309005737,
      "learning_rate": 8e-05,
      "loss": 1.5918,
      "step": 2588
    },
    {
      "epoch": 0.5376947040498442,
      "grad_norm": 0.5113598704338074,
      "learning_rate": 8e-05,
      "loss": 1.5927,
      "step": 2589
    },
    {
      "epoch": 0.5379023883696781,
      "grad_norm": 0.5320143103599548,
      "learning_rate": 8e-05,
      "loss": 1.615,
      "step": 2590
    },
    {
      "epoch": 0.5381100726895119,
      "grad_norm": 0.5382052063941956,
      "learning_rate": 8e-05,
      "loss": 1.5541,
      "step": 2591
    },
    {
      "epoch": 0.5383177570093458,
      "grad_norm": 0.5445559620857239,
      "learning_rate": 8e-05,
      "loss": 1.5962,
      "step": 2592
    },
    {
      "epoch": 0.5385254413291797,
      "grad_norm": 0.49818602204322815,
      "learning_rate": 8e-05,
      "loss": 1.5125,
      "step": 2593
    },
    {
      "epoch": 0.5387331256490135,
      "grad_norm": 0.5041065812110901,
      "learning_rate": 8e-05,
      "loss": 1.5459,
      "step": 2594
    },
    {
      "epoch": 0.5389408099688473,
      "grad_norm": 0.5130652189254761,
      "learning_rate": 8e-05,
      "loss": 1.5832,
      "step": 2595
    },
    {
      "epoch": 0.5391484942886812,
      "grad_norm": 0.5492714047431946,
      "learning_rate": 8e-05,
      "loss": 1.5877,
      "step": 2596
    },
    {
      "epoch": 0.539356178608515,
      "grad_norm": 0.4994000196456909,
      "learning_rate": 8e-05,
      "loss": 1.4821,
      "step": 2597
    },
    {
      "epoch": 0.5395638629283489,
      "grad_norm": 0.529475748538971,
      "learning_rate": 8e-05,
      "loss": 1.6487,
      "step": 2598
    },
    {
      "epoch": 0.5397715472481828,
      "grad_norm": 0.5280352830886841,
      "learning_rate": 8e-05,
      "loss": 1.5581,
      "step": 2599
    },
    {
      "epoch": 0.5399792315680166,
      "grad_norm": 0.5195119380950928,
      "learning_rate": 8e-05,
      "loss": 1.5381,
      "step": 2600
    },
    {
      "epoch": 0.5401869158878505,
      "grad_norm": 0.5004687905311584,
      "learning_rate": 8e-05,
      "loss": 1.4899,
      "step": 2601
    },
    {
      "epoch": 0.5403946002076844,
      "grad_norm": 0.5243152976036072,
      "learning_rate": 8e-05,
      "loss": 1.6188,
      "step": 2602
    },
    {
      "epoch": 0.5406022845275181,
      "grad_norm": 0.49481499195098877,
      "learning_rate": 8e-05,
      "loss": 1.5984,
      "step": 2603
    },
    {
      "epoch": 0.540809968847352,
      "grad_norm": 0.4946075975894928,
      "learning_rate": 8e-05,
      "loss": 1.5462,
      "step": 2604
    },
    {
      "epoch": 0.5410176531671859,
      "grad_norm": 0.5581692457199097,
      "learning_rate": 8e-05,
      "loss": 1.5007,
      "step": 2605
    },
    {
      "epoch": 0.5412253374870197,
      "grad_norm": 0.514069676399231,
      "learning_rate": 8e-05,
      "loss": 1.5571,
      "step": 2606
    },
    {
      "epoch": 0.5414330218068536,
      "grad_norm": 0.5100674629211426,
      "learning_rate": 8e-05,
      "loss": 1.5248,
      "step": 2607
    },
    {
      "epoch": 0.5416407061266875,
      "grad_norm": 0.5168145895004272,
      "learning_rate": 8e-05,
      "loss": 1.5596,
      "step": 2608
    },
    {
      "epoch": 0.5418483904465213,
      "grad_norm": 0.5216502547264099,
      "learning_rate": 8e-05,
      "loss": 1.5369,
      "step": 2609
    },
    {
      "epoch": 0.5420560747663551,
      "grad_norm": 0.5220406651496887,
      "learning_rate": 8e-05,
      "loss": 1.6143,
      "step": 2610
    },
    {
      "epoch": 0.542263759086189,
      "grad_norm": 0.5106246471405029,
      "learning_rate": 8e-05,
      "loss": 1.5871,
      "step": 2611
    },
    {
      "epoch": 0.5424714434060228,
      "grad_norm": 0.5054118633270264,
      "learning_rate": 8e-05,
      "loss": 1.4816,
      "step": 2612
    },
    {
      "epoch": 0.5426791277258567,
      "grad_norm": 0.5267699956893921,
      "learning_rate": 8e-05,
      "loss": 1.5261,
      "step": 2613
    },
    {
      "epoch": 0.5428868120456906,
      "grad_norm": 0.5472517609596252,
      "learning_rate": 8e-05,
      "loss": 1.6014,
      "step": 2614
    },
    {
      "epoch": 0.5430944963655244,
      "grad_norm": 0.5042798519134521,
      "learning_rate": 8e-05,
      "loss": 1.5761,
      "step": 2615
    },
    {
      "epoch": 0.5433021806853583,
      "grad_norm": 0.5117297172546387,
      "learning_rate": 8e-05,
      "loss": 1.5245,
      "step": 2616
    },
    {
      "epoch": 0.5435098650051922,
      "grad_norm": 0.6044188737869263,
      "learning_rate": 8e-05,
      "loss": 1.571,
      "step": 2617
    },
    {
      "epoch": 0.5437175493250259,
      "grad_norm": 0.5296552777290344,
      "learning_rate": 8e-05,
      "loss": 1.5735,
      "step": 2618
    },
    {
      "epoch": 0.5439252336448598,
      "grad_norm": 0.5351176261901855,
      "learning_rate": 8e-05,
      "loss": 1.5473,
      "step": 2619
    },
    {
      "epoch": 0.5441329179646937,
      "grad_norm": 0.5096343755722046,
      "learning_rate": 8e-05,
      "loss": 1.549,
      "step": 2620
    },
    {
      "epoch": 0.5443406022845275,
      "grad_norm": 0.6275840401649475,
      "learning_rate": 8e-05,
      "loss": 1.5542,
      "step": 2621
    },
    {
      "epoch": 0.5445482866043614,
      "grad_norm": 0.5917977094650269,
      "learning_rate": 8e-05,
      "loss": 1.5259,
      "step": 2622
    },
    {
      "epoch": 0.5447559709241953,
      "grad_norm": 0.5169813632965088,
      "learning_rate": 8e-05,
      "loss": 1.5327,
      "step": 2623
    },
    {
      "epoch": 0.5449636552440291,
      "grad_norm": 0.5194140076637268,
      "learning_rate": 8e-05,
      "loss": 1.6058,
      "step": 2624
    },
    {
      "epoch": 0.5451713395638629,
      "grad_norm": 0.5479727387428284,
      "learning_rate": 8e-05,
      "loss": 1.5855,
      "step": 2625
    },
    {
      "epoch": 0.5453790238836967,
      "grad_norm": 0.5423098206520081,
      "learning_rate": 8e-05,
      "loss": 1.58,
      "step": 2626
    },
    {
      "epoch": 0.5455867082035306,
      "grad_norm": 0.5254849195480347,
      "learning_rate": 8e-05,
      "loss": 1.5466,
      "step": 2627
    },
    {
      "epoch": 0.5457943925233645,
      "grad_norm": 0.5386984944343567,
      "learning_rate": 8e-05,
      "loss": 1.5583,
      "step": 2628
    },
    {
      "epoch": 0.5460020768431983,
      "grad_norm": 0.5074415802955627,
      "learning_rate": 8e-05,
      "loss": 1.5341,
      "step": 2629
    },
    {
      "epoch": 0.5462097611630322,
      "grad_norm": 0.5718663334846497,
      "learning_rate": 8e-05,
      "loss": 1.6086,
      "step": 2630
    },
    {
      "epoch": 0.5464174454828661,
      "grad_norm": 0.5871424674987793,
      "learning_rate": 8e-05,
      "loss": 1.5826,
      "step": 2631
    },
    {
      "epoch": 0.5466251298027,
      "grad_norm": 0.5480154156684875,
      "learning_rate": 8e-05,
      "loss": 1.6565,
      "step": 2632
    },
    {
      "epoch": 0.5468328141225337,
      "grad_norm": 0.48676350712776184,
      "learning_rate": 8e-05,
      "loss": 1.5143,
      "step": 2633
    },
    {
      "epoch": 0.5470404984423676,
      "grad_norm": 0.5588537454605103,
      "learning_rate": 8e-05,
      "loss": 1.5561,
      "step": 2634
    },
    {
      "epoch": 0.5472481827622014,
      "grad_norm": 0.5413640141487122,
      "learning_rate": 8e-05,
      "loss": 1.6585,
      "step": 2635
    },
    {
      "epoch": 0.5474558670820353,
      "grad_norm": 0.49471515417099,
      "learning_rate": 8e-05,
      "loss": 1.4957,
      "step": 2636
    },
    {
      "epoch": 0.5476635514018692,
      "grad_norm": 0.5229868292808533,
      "learning_rate": 8e-05,
      "loss": 1.5581,
      "step": 2637
    },
    {
      "epoch": 0.547871235721703,
      "grad_norm": 0.5099173188209534,
      "learning_rate": 8e-05,
      "loss": 1.5943,
      "step": 2638
    },
    {
      "epoch": 0.5480789200415369,
      "grad_norm": 0.5230522751808167,
      "learning_rate": 8e-05,
      "loss": 1.5662,
      "step": 2639
    },
    {
      "epoch": 0.5482866043613707,
      "grad_norm": 0.5060945749282837,
      "learning_rate": 8e-05,
      "loss": 1.5591,
      "step": 2640
    },
    {
      "epoch": 0.5484942886812045,
      "grad_norm": 0.5048191547393799,
      "learning_rate": 8e-05,
      "loss": 1.5547,
      "step": 2641
    },
    {
      "epoch": 0.5487019730010384,
      "grad_norm": 0.49480634927749634,
      "learning_rate": 8e-05,
      "loss": 1.5281,
      "step": 2642
    },
    {
      "epoch": 0.5489096573208723,
      "grad_norm": 0.503041684627533,
      "learning_rate": 8e-05,
      "loss": 1.5378,
      "step": 2643
    },
    {
      "epoch": 0.5491173416407061,
      "grad_norm": 0.5225034952163696,
      "learning_rate": 8e-05,
      "loss": 1.4993,
      "step": 2644
    },
    {
      "epoch": 0.54932502596054,
      "grad_norm": 0.5275830030441284,
      "learning_rate": 8e-05,
      "loss": 1.6099,
      "step": 2645
    },
    {
      "epoch": 0.5495327102803739,
      "grad_norm": 0.5186808109283447,
      "learning_rate": 8e-05,
      "loss": 1.605,
      "step": 2646
    },
    {
      "epoch": 0.5497403946002077,
      "grad_norm": 0.5072364211082458,
      "learning_rate": 8e-05,
      "loss": 1.5667,
      "step": 2647
    },
    {
      "epoch": 0.5499480789200415,
      "grad_norm": 0.5087660551071167,
      "learning_rate": 8e-05,
      "loss": 1.5476,
      "step": 2648
    },
    {
      "epoch": 0.5501557632398754,
      "grad_norm": 0.5886751413345337,
      "learning_rate": 8e-05,
      "loss": 1.6084,
      "step": 2649
    },
    {
      "epoch": 0.5503634475597092,
      "grad_norm": 0.5000873804092407,
      "learning_rate": 8e-05,
      "loss": 1.5246,
      "step": 2650
    },
    {
      "epoch": 0.5505711318795431,
      "grad_norm": 0.5032468438148499,
      "learning_rate": 8e-05,
      "loss": 1.5161,
      "step": 2651
    },
    {
      "epoch": 0.550778816199377,
      "grad_norm": 0.5816187262535095,
      "learning_rate": 8e-05,
      "loss": 1.6107,
      "step": 2652
    },
    {
      "epoch": 0.5509865005192108,
      "grad_norm": 0.5264134407043457,
      "learning_rate": 8e-05,
      "loss": 1.6337,
      "step": 2653
    },
    {
      "epoch": 0.5511941848390447,
      "grad_norm": 0.5257280468940735,
      "learning_rate": 8e-05,
      "loss": 1.6287,
      "step": 2654
    },
    {
      "epoch": 0.5514018691588785,
      "grad_norm": 0.47970205545425415,
      "learning_rate": 8e-05,
      "loss": 1.4825,
      "step": 2655
    },
    {
      "epoch": 0.5516095534787123,
      "grad_norm": 0.4971425235271454,
      "learning_rate": 8e-05,
      "loss": 1.5492,
      "step": 2656
    },
    {
      "epoch": 0.5518172377985462,
      "grad_norm": 0.5874940752983093,
      "learning_rate": 8e-05,
      "loss": 1.5363,
      "step": 2657
    },
    {
      "epoch": 0.5520249221183801,
      "grad_norm": 0.5532630085945129,
      "learning_rate": 8e-05,
      "loss": 1.5289,
      "step": 2658
    },
    {
      "epoch": 0.5522326064382139,
      "grad_norm": 0.5012029409408569,
      "learning_rate": 8e-05,
      "loss": 1.592,
      "step": 2659
    },
    {
      "epoch": 0.5524402907580478,
      "grad_norm": 0.5123198628425598,
      "learning_rate": 8e-05,
      "loss": 1.5494,
      "step": 2660
    },
    {
      "epoch": 0.5526479750778817,
      "grad_norm": 0.521674394607544,
      "learning_rate": 8e-05,
      "loss": 1.5383,
      "step": 2661
    },
    {
      "epoch": 0.5528556593977155,
      "grad_norm": 0.5289007425308228,
      "learning_rate": 8e-05,
      "loss": 1.542,
      "step": 2662
    },
    {
      "epoch": 0.5530633437175493,
      "grad_norm": 0.5011623501777649,
      "learning_rate": 8e-05,
      "loss": 1.5637,
      "step": 2663
    },
    {
      "epoch": 0.5532710280373832,
      "grad_norm": 0.5280729532241821,
      "learning_rate": 8e-05,
      "loss": 1.5499,
      "step": 2664
    },
    {
      "epoch": 0.553478712357217,
      "grad_norm": 0.5130934715270996,
      "learning_rate": 8e-05,
      "loss": 1.596,
      "step": 2665
    },
    {
      "epoch": 0.5536863966770509,
      "grad_norm": 0.5125265717506409,
      "learning_rate": 8e-05,
      "loss": 1.5102,
      "step": 2666
    },
    {
      "epoch": 0.5538940809968848,
      "grad_norm": 0.5373044610023499,
      "learning_rate": 8e-05,
      "loss": 1.5319,
      "step": 2667
    },
    {
      "epoch": 0.5541017653167186,
      "grad_norm": 0.5123081207275391,
      "learning_rate": 8e-05,
      "loss": 1.5356,
      "step": 2668
    },
    {
      "epoch": 0.5543094496365525,
      "grad_norm": 0.5117352604866028,
      "learning_rate": 8e-05,
      "loss": 1.5098,
      "step": 2669
    },
    {
      "epoch": 0.5545171339563862,
      "grad_norm": 0.5064887404441833,
      "learning_rate": 8e-05,
      "loss": 1.5513,
      "step": 2670
    },
    {
      "epoch": 0.5547248182762201,
      "grad_norm": 0.5283700823783875,
      "learning_rate": 8e-05,
      "loss": 1.6139,
      "step": 2671
    },
    {
      "epoch": 0.554932502596054,
      "grad_norm": 0.5440184473991394,
      "learning_rate": 8e-05,
      "loss": 1.5557,
      "step": 2672
    },
    {
      "epoch": 0.5551401869158878,
      "grad_norm": 0.5373353958129883,
      "learning_rate": 8e-05,
      "loss": 1.608,
      "step": 2673
    },
    {
      "epoch": 0.5553478712357217,
      "grad_norm": 0.5456499457359314,
      "learning_rate": 8e-05,
      "loss": 1.6078,
      "step": 2674
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.5141017436981201,
      "learning_rate": 8e-05,
      "loss": 1.5639,
      "step": 2675
    },
    {
      "epoch": 0.5557632398753894,
      "grad_norm": 0.5070343613624573,
      "learning_rate": 8e-05,
      "loss": 1.5033,
      "step": 2676
    },
    {
      "epoch": 0.5559709241952233,
      "grad_norm": 0.5327128171920776,
      "learning_rate": 8e-05,
      "loss": 1.5481,
      "step": 2677
    },
    {
      "epoch": 0.5561786085150571,
      "grad_norm": 0.6137487888336182,
      "learning_rate": 8e-05,
      "loss": 1.6048,
      "step": 2678
    },
    {
      "epoch": 0.5563862928348909,
      "grad_norm": 0.5100077390670776,
      "learning_rate": 8e-05,
      "loss": 1.4818,
      "step": 2679
    },
    {
      "epoch": 0.5565939771547248,
      "grad_norm": 0.5046921372413635,
      "learning_rate": 8e-05,
      "loss": 1.5365,
      "step": 2680
    },
    {
      "epoch": 0.5568016614745587,
      "grad_norm": 0.5135619044303894,
      "learning_rate": 8e-05,
      "loss": 1.5347,
      "step": 2681
    },
    {
      "epoch": 0.5570093457943925,
      "grad_norm": 0.5596857666969299,
      "learning_rate": 8e-05,
      "loss": 1.5836,
      "step": 2682
    },
    {
      "epoch": 0.5572170301142264,
      "grad_norm": 0.5241439938545227,
      "learning_rate": 8e-05,
      "loss": 1.6102,
      "step": 2683
    },
    {
      "epoch": 0.5574247144340603,
      "grad_norm": 0.5305902361869812,
      "learning_rate": 8e-05,
      "loss": 1.5581,
      "step": 2684
    },
    {
      "epoch": 0.557632398753894,
      "grad_norm": 0.4992052912712097,
      "learning_rate": 8e-05,
      "loss": 1.5119,
      "step": 2685
    },
    {
      "epoch": 0.5578400830737279,
      "grad_norm": 0.5193141102790833,
      "learning_rate": 8e-05,
      "loss": 1.5624,
      "step": 2686
    },
    {
      "epoch": 0.5580477673935618,
      "grad_norm": 0.5143951773643494,
      "learning_rate": 8e-05,
      "loss": 1.5473,
      "step": 2687
    },
    {
      "epoch": 0.5582554517133956,
      "grad_norm": 0.5419348478317261,
      "learning_rate": 8e-05,
      "loss": 1.5579,
      "step": 2688
    },
    {
      "epoch": 0.5584631360332295,
      "grad_norm": 0.5398904085159302,
      "learning_rate": 8e-05,
      "loss": 1.5888,
      "step": 2689
    },
    {
      "epoch": 0.5586708203530634,
      "grad_norm": 0.5264995098114014,
      "learning_rate": 8e-05,
      "loss": 1.5997,
      "step": 2690
    },
    {
      "epoch": 0.5588785046728972,
      "grad_norm": 0.5043335556983948,
      "learning_rate": 8e-05,
      "loss": 1.5461,
      "step": 2691
    },
    {
      "epoch": 0.5590861889927311,
      "grad_norm": 0.5361884236335754,
      "learning_rate": 8e-05,
      "loss": 1.5636,
      "step": 2692
    },
    {
      "epoch": 0.5592938733125649,
      "grad_norm": 0.5052306056022644,
      "learning_rate": 8e-05,
      "loss": 1.5238,
      "step": 2693
    },
    {
      "epoch": 0.5595015576323987,
      "grad_norm": 0.49911513924598694,
      "learning_rate": 8e-05,
      "loss": 1.537,
      "step": 2694
    },
    {
      "epoch": 0.5597092419522326,
      "grad_norm": 0.5349092483520508,
      "learning_rate": 8e-05,
      "loss": 1.5804,
      "step": 2695
    },
    {
      "epoch": 0.5599169262720665,
      "grad_norm": 0.5042306780815125,
      "learning_rate": 8e-05,
      "loss": 1.4703,
      "step": 2696
    },
    {
      "epoch": 0.5601246105919003,
      "grad_norm": 0.5325839519500732,
      "learning_rate": 8e-05,
      "loss": 1.5393,
      "step": 2697
    },
    {
      "epoch": 0.5603322949117342,
      "grad_norm": 0.5401809811592102,
      "learning_rate": 8e-05,
      "loss": 1.6033,
      "step": 2698
    },
    {
      "epoch": 0.5605399792315681,
      "grad_norm": 0.5372620224952698,
      "learning_rate": 8e-05,
      "loss": 1.4942,
      "step": 2699
    },
    {
      "epoch": 0.5607476635514018,
      "grad_norm": 0.5419166684150696,
      "learning_rate": 8e-05,
      "loss": 1.5295,
      "step": 2700
    },
    {
      "epoch": 0.5609553478712357,
      "grad_norm": 0.5134370923042297,
      "learning_rate": 8e-05,
      "loss": 1.4819,
      "step": 2701
    },
    {
      "epoch": 0.5611630321910696,
      "grad_norm": 0.5202261209487915,
      "learning_rate": 8e-05,
      "loss": 1.5344,
      "step": 2702
    },
    {
      "epoch": 0.5613707165109034,
      "grad_norm": 0.5219168066978455,
      "learning_rate": 8e-05,
      "loss": 1.5119,
      "step": 2703
    },
    {
      "epoch": 0.5615784008307373,
      "grad_norm": 0.522109866142273,
      "learning_rate": 8e-05,
      "loss": 1.5409,
      "step": 2704
    },
    {
      "epoch": 0.5617860851505712,
      "grad_norm": 0.5673719644546509,
      "learning_rate": 8e-05,
      "loss": 1.6408,
      "step": 2705
    },
    {
      "epoch": 0.561993769470405,
      "grad_norm": 0.5494527816772461,
      "learning_rate": 8e-05,
      "loss": 1.5707,
      "step": 2706
    },
    {
      "epoch": 0.5622014537902389,
      "grad_norm": 0.5136489272117615,
      "learning_rate": 8e-05,
      "loss": 1.5566,
      "step": 2707
    },
    {
      "epoch": 0.5624091381100726,
      "grad_norm": 0.5121893286705017,
      "learning_rate": 8e-05,
      "loss": 1.5256,
      "step": 2708
    },
    {
      "epoch": 0.5626168224299065,
      "grad_norm": 0.49312853813171387,
      "learning_rate": 8e-05,
      "loss": 1.5215,
      "step": 2709
    },
    {
      "epoch": 0.5628245067497404,
      "grad_norm": 0.5472443103790283,
      "learning_rate": 8e-05,
      "loss": 1.5127,
      "step": 2710
    },
    {
      "epoch": 0.5630321910695743,
      "grad_norm": 0.5221959352493286,
      "learning_rate": 8e-05,
      "loss": 1.5957,
      "step": 2711
    },
    {
      "epoch": 0.5632398753894081,
      "grad_norm": 0.5420714020729065,
      "learning_rate": 8e-05,
      "loss": 1.5427,
      "step": 2712
    },
    {
      "epoch": 0.563447559709242,
      "grad_norm": 0.5317118763923645,
      "learning_rate": 8e-05,
      "loss": 1.6281,
      "step": 2713
    },
    {
      "epoch": 0.5636552440290759,
      "grad_norm": 0.531112790107727,
      "learning_rate": 8e-05,
      "loss": 1.5929,
      "step": 2714
    },
    {
      "epoch": 0.5638629283489096,
      "grad_norm": 0.5138751864433289,
      "learning_rate": 8e-05,
      "loss": 1.5962,
      "step": 2715
    },
    {
      "epoch": 0.5640706126687435,
      "grad_norm": 0.493773877620697,
      "learning_rate": 8e-05,
      "loss": 1.4706,
      "step": 2716
    },
    {
      "epoch": 0.5642782969885773,
      "grad_norm": 0.5258544087409973,
      "learning_rate": 8e-05,
      "loss": 1.5494,
      "step": 2717
    },
    {
      "epoch": 0.5644859813084112,
      "grad_norm": 0.5253569483757019,
      "learning_rate": 8e-05,
      "loss": 1.5538,
      "step": 2718
    },
    {
      "epoch": 0.5646936656282451,
      "grad_norm": 0.5212979912757874,
      "learning_rate": 8e-05,
      "loss": 1.4899,
      "step": 2719
    },
    {
      "epoch": 0.564901349948079,
      "grad_norm": 0.5010308027267456,
      "learning_rate": 8e-05,
      "loss": 1.5258,
      "step": 2720
    },
    {
      "epoch": 0.5651090342679128,
      "grad_norm": 0.5242239236831665,
      "learning_rate": 8e-05,
      "loss": 1.5427,
      "step": 2721
    },
    {
      "epoch": 0.5653167185877467,
      "grad_norm": 0.4979899227619171,
      "learning_rate": 8e-05,
      "loss": 1.5751,
      "step": 2722
    },
    {
      "epoch": 0.5655244029075804,
      "grad_norm": 0.5082122087478638,
      "learning_rate": 8e-05,
      "loss": 1.5637,
      "step": 2723
    },
    {
      "epoch": 0.5657320872274143,
      "grad_norm": 0.5392714738845825,
      "learning_rate": 8e-05,
      "loss": 1.6375,
      "step": 2724
    },
    {
      "epoch": 0.5659397715472482,
      "grad_norm": 0.502173900604248,
      "learning_rate": 8e-05,
      "loss": 1.5923,
      "step": 2725
    },
    {
      "epoch": 0.566147455867082,
      "grad_norm": 0.4920365810394287,
      "learning_rate": 8e-05,
      "loss": 1.5303,
      "step": 2726
    },
    {
      "epoch": 0.5663551401869159,
      "grad_norm": 0.4974154233932495,
      "learning_rate": 8e-05,
      "loss": 1.5815,
      "step": 2727
    },
    {
      "epoch": 0.5665628245067498,
      "grad_norm": 0.5457067489624023,
      "learning_rate": 8e-05,
      "loss": 1.5774,
      "step": 2728
    },
    {
      "epoch": 0.5667705088265836,
      "grad_norm": 0.5052133202552795,
      "learning_rate": 8e-05,
      "loss": 1.508,
      "step": 2729
    },
    {
      "epoch": 0.5669781931464174,
      "grad_norm": 0.5542793869972229,
      "learning_rate": 8e-05,
      "loss": 1.5994,
      "step": 2730
    },
    {
      "epoch": 0.5671858774662513,
      "grad_norm": 0.5140535235404968,
      "learning_rate": 8e-05,
      "loss": 1.5062,
      "step": 2731
    },
    {
      "epoch": 0.5673935617860851,
      "grad_norm": 0.5263282060623169,
      "learning_rate": 8e-05,
      "loss": 1.5932,
      "step": 2732
    },
    {
      "epoch": 0.567601246105919,
      "grad_norm": 0.5152779817581177,
      "learning_rate": 8e-05,
      "loss": 1.4961,
      "step": 2733
    },
    {
      "epoch": 0.5678089304257529,
      "grad_norm": 0.5263432860374451,
      "learning_rate": 8e-05,
      "loss": 1.5839,
      "step": 2734
    },
    {
      "epoch": 0.5680166147455867,
      "grad_norm": 0.4997681677341461,
      "learning_rate": 8e-05,
      "loss": 1.5555,
      "step": 2735
    },
    {
      "epoch": 0.5682242990654206,
      "grad_norm": 0.5276387929916382,
      "learning_rate": 8e-05,
      "loss": 1.5914,
      "step": 2736
    },
    {
      "epoch": 0.5684319833852544,
      "grad_norm": 0.4941411316394806,
      "learning_rate": 8e-05,
      "loss": 1.5112,
      "step": 2737
    },
    {
      "epoch": 0.5686396677050882,
      "grad_norm": 0.5030642151832581,
      "learning_rate": 8e-05,
      "loss": 1.4287,
      "step": 2738
    },
    {
      "epoch": 0.5688473520249221,
      "grad_norm": 0.47869259119033813,
      "learning_rate": 8e-05,
      "loss": 1.5055,
      "step": 2739
    },
    {
      "epoch": 0.569055036344756,
      "grad_norm": 0.5265507102012634,
      "learning_rate": 8e-05,
      "loss": 1.5973,
      "step": 2740
    },
    {
      "epoch": 0.5692627206645898,
      "grad_norm": 0.528171956539154,
      "learning_rate": 8e-05,
      "loss": 1.5892,
      "step": 2741
    },
    {
      "epoch": 0.5694704049844237,
      "grad_norm": 0.5014118552207947,
      "learning_rate": 8e-05,
      "loss": 1.5203,
      "step": 2742
    },
    {
      "epoch": 0.5696780893042576,
      "grad_norm": 0.5167685747146606,
      "learning_rate": 8e-05,
      "loss": 1.564,
      "step": 2743
    },
    {
      "epoch": 0.5698857736240914,
      "grad_norm": 0.5222561359405518,
      "learning_rate": 8e-05,
      "loss": 1.5243,
      "step": 2744
    },
    {
      "epoch": 0.5700934579439252,
      "grad_norm": 0.5133991241455078,
      "learning_rate": 8e-05,
      "loss": 1.5403,
      "step": 2745
    },
    {
      "epoch": 0.570301142263759,
      "grad_norm": 0.5201916098594666,
      "learning_rate": 8e-05,
      "loss": 1.5732,
      "step": 2746
    },
    {
      "epoch": 0.5705088265835929,
      "grad_norm": 0.5097384452819824,
      "learning_rate": 8e-05,
      "loss": 1.5958,
      "step": 2747
    },
    {
      "epoch": 0.5707165109034268,
      "grad_norm": 0.5123527646064758,
      "learning_rate": 8e-05,
      "loss": 1.5869,
      "step": 2748
    },
    {
      "epoch": 0.5709241952232607,
      "grad_norm": 0.5189564228057861,
      "learning_rate": 8e-05,
      "loss": 1.5932,
      "step": 2749
    },
    {
      "epoch": 0.5711318795430945,
      "grad_norm": 0.5192084312438965,
      "learning_rate": 8e-05,
      "loss": 1.5284,
      "step": 2750
    },
    {
      "epoch": 0.5713395638629284,
      "grad_norm": 0.5026838183403015,
      "learning_rate": 8e-05,
      "loss": 1.5058,
      "step": 2751
    },
    {
      "epoch": 0.5715472481827621,
      "grad_norm": 0.5398958325386047,
      "learning_rate": 8e-05,
      "loss": 1.6339,
      "step": 2752
    },
    {
      "epoch": 0.571754932502596,
      "grad_norm": 0.5269039273262024,
      "learning_rate": 8e-05,
      "loss": 1.5635,
      "step": 2753
    },
    {
      "epoch": 0.5719626168224299,
      "grad_norm": 0.5309609174728394,
      "learning_rate": 8e-05,
      "loss": 1.4882,
      "step": 2754
    },
    {
      "epoch": 0.5721703011422637,
      "grad_norm": 0.5078904032707214,
      "learning_rate": 8e-05,
      "loss": 1.6141,
      "step": 2755
    },
    {
      "epoch": 0.5723779854620976,
      "grad_norm": 0.5262061357498169,
      "learning_rate": 8e-05,
      "loss": 1.608,
      "step": 2756
    },
    {
      "epoch": 0.5725856697819315,
      "grad_norm": 0.5232616662979126,
      "learning_rate": 8e-05,
      "loss": 1.5799,
      "step": 2757
    },
    {
      "epoch": 0.5727933541017654,
      "grad_norm": 0.49849748611450195,
      "learning_rate": 8e-05,
      "loss": 1.5073,
      "step": 2758
    },
    {
      "epoch": 0.5730010384215992,
      "grad_norm": 0.5224372744560242,
      "learning_rate": 8e-05,
      "loss": 1.6051,
      "step": 2759
    },
    {
      "epoch": 0.573208722741433,
      "grad_norm": 0.518502414226532,
      "learning_rate": 8e-05,
      "loss": 1.5905,
      "step": 2760
    },
    {
      "epoch": 0.5734164070612668,
      "grad_norm": 0.511958658695221,
      "learning_rate": 8e-05,
      "loss": 1.4851,
      "step": 2761
    },
    {
      "epoch": 0.5736240913811007,
      "grad_norm": 0.5087369680404663,
      "learning_rate": 8e-05,
      "loss": 1.5221,
      "step": 2762
    },
    {
      "epoch": 0.5738317757009346,
      "grad_norm": 0.7710164785385132,
      "learning_rate": 8e-05,
      "loss": 1.5034,
      "step": 2763
    },
    {
      "epoch": 0.5740394600207684,
      "grad_norm": 0.5106395483016968,
      "learning_rate": 8e-05,
      "loss": 1.5988,
      "step": 2764
    },
    {
      "epoch": 0.5742471443406023,
      "grad_norm": 0.5100325345993042,
      "learning_rate": 8e-05,
      "loss": 1.5096,
      "step": 2765
    },
    {
      "epoch": 0.5744548286604362,
      "grad_norm": 0.49486199021339417,
      "learning_rate": 8e-05,
      "loss": 1.4733,
      "step": 2766
    },
    {
      "epoch": 0.5746625129802699,
      "grad_norm": 0.5689396262168884,
      "learning_rate": 8e-05,
      "loss": 1.65,
      "step": 2767
    },
    {
      "epoch": 0.5748701973001038,
      "grad_norm": 0.5266242623329163,
      "learning_rate": 8e-05,
      "loss": 1.5513,
      "step": 2768
    },
    {
      "epoch": 0.5750778816199377,
      "grad_norm": 0.5084609985351562,
      "learning_rate": 8e-05,
      "loss": 1.5034,
      "step": 2769
    },
    {
      "epoch": 0.5752855659397715,
      "grad_norm": 0.5007745027542114,
      "learning_rate": 8e-05,
      "loss": 1.5009,
      "step": 2770
    },
    {
      "epoch": 0.5754932502596054,
      "grad_norm": 0.520416796207428,
      "learning_rate": 8e-05,
      "loss": 1.5231,
      "step": 2771
    },
    {
      "epoch": 0.5757009345794393,
      "grad_norm": 0.5064281225204468,
      "learning_rate": 8e-05,
      "loss": 1.5356,
      "step": 2772
    },
    {
      "epoch": 0.5759086188992731,
      "grad_norm": 0.5355567336082458,
      "learning_rate": 8e-05,
      "loss": 1.5831,
      "step": 2773
    },
    {
      "epoch": 0.576116303219107,
      "grad_norm": 0.5114003419876099,
      "learning_rate": 8e-05,
      "loss": 1.5855,
      "step": 2774
    },
    {
      "epoch": 0.5763239875389408,
      "grad_norm": 0.5105564594268799,
      "learning_rate": 8e-05,
      "loss": 1.5277,
      "step": 2775
    },
    {
      "epoch": 0.5765316718587746,
      "grad_norm": 0.5268367528915405,
      "learning_rate": 8e-05,
      "loss": 1.607,
      "step": 2776
    },
    {
      "epoch": 0.5767393561786085,
      "grad_norm": 0.5006731748580933,
      "learning_rate": 8e-05,
      "loss": 1.6212,
      "step": 2777
    },
    {
      "epoch": 0.5769470404984424,
      "grad_norm": 0.5249113440513611,
      "learning_rate": 8e-05,
      "loss": 1.5704,
      "step": 2778
    },
    {
      "epoch": 0.5771547248182762,
      "grad_norm": 0.5069587230682373,
      "learning_rate": 8e-05,
      "loss": 1.5597,
      "step": 2779
    },
    {
      "epoch": 0.5773624091381101,
      "grad_norm": 0.5126855373382568,
      "learning_rate": 8e-05,
      "loss": 1.5091,
      "step": 2780
    },
    {
      "epoch": 0.577570093457944,
      "grad_norm": 0.5261822938919067,
      "learning_rate": 8e-05,
      "loss": 1.4847,
      "step": 2781
    },
    {
      "epoch": 0.5777777777777777,
      "grad_norm": 0.5000608563423157,
      "learning_rate": 8e-05,
      "loss": 1.5287,
      "step": 2782
    },
    {
      "epoch": 0.5779854620976116,
      "grad_norm": 0.5185624361038208,
      "learning_rate": 8e-05,
      "loss": 1.5692,
      "step": 2783
    },
    {
      "epoch": 0.5781931464174455,
      "grad_norm": 0.5299122929573059,
      "learning_rate": 8e-05,
      "loss": 1.5857,
      "step": 2784
    },
    {
      "epoch": 0.5784008307372793,
      "grad_norm": 0.5279645323753357,
      "learning_rate": 8e-05,
      "loss": 1.586,
      "step": 2785
    },
    {
      "epoch": 0.5786085150571132,
      "grad_norm": 0.5176546573638916,
      "learning_rate": 8e-05,
      "loss": 1.6287,
      "step": 2786
    },
    {
      "epoch": 0.5788161993769471,
      "grad_norm": 0.512985348701477,
      "learning_rate": 8e-05,
      "loss": 1.5642,
      "step": 2787
    },
    {
      "epoch": 0.5790238836967809,
      "grad_norm": 0.5028579831123352,
      "learning_rate": 8e-05,
      "loss": 1.4601,
      "step": 2788
    },
    {
      "epoch": 0.5792315680166148,
      "grad_norm": 0.5143795609474182,
      "learning_rate": 8e-05,
      "loss": 1.5745,
      "step": 2789
    },
    {
      "epoch": 0.5794392523364486,
      "grad_norm": 0.5350067019462585,
      "learning_rate": 8e-05,
      "loss": 1.6075,
      "step": 2790
    },
    {
      "epoch": 0.5796469366562824,
      "grad_norm": 0.5133538842201233,
      "learning_rate": 8e-05,
      "loss": 1.6325,
      "step": 2791
    },
    {
      "epoch": 0.5798546209761163,
      "grad_norm": 0.49931225180625916,
      "learning_rate": 8e-05,
      "loss": 1.5322,
      "step": 2792
    },
    {
      "epoch": 0.5800623052959502,
      "grad_norm": 0.5204224586486816,
      "learning_rate": 8e-05,
      "loss": 1.5818,
      "step": 2793
    },
    {
      "epoch": 0.580269989615784,
      "grad_norm": 0.5103965997695923,
      "learning_rate": 8e-05,
      "loss": 1.518,
      "step": 2794
    },
    {
      "epoch": 0.5804776739356179,
      "grad_norm": 0.5115767121315002,
      "learning_rate": 8e-05,
      "loss": 1.5504,
      "step": 2795
    },
    {
      "epoch": 0.5806853582554518,
      "grad_norm": 0.5340980291366577,
      "learning_rate": 8e-05,
      "loss": 1.5652,
      "step": 2796
    },
    {
      "epoch": 0.5808930425752855,
      "grad_norm": 0.535073459148407,
      "learning_rate": 8e-05,
      "loss": 1.5424,
      "step": 2797
    },
    {
      "epoch": 0.5811007268951194,
      "grad_norm": 0.5498014092445374,
      "learning_rate": 8e-05,
      "loss": 1.6218,
      "step": 2798
    },
    {
      "epoch": 0.5813084112149532,
      "grad_norm": 0.5135794281959534,
      "learning_rate": 8e-05,
      "loss": 1.5992,
      "step": 2799
    },
    {
      "epoch": 0.5815160955347871,
      "grad_norm": 0.5113550424575806,
      "learning_rate": 8e-05,
      "loss": 1.5281,
      "step": 2800
    },
    {
      "epoch": 0.581723779854621,
      "grad_norm": 0.5498125553131104,
      "learning_rate": 8e-05,
      "loss": 1.5895,
      "step": 2801
    },
    {
      "epoch": 0.5819314641744548,
      "grad_norm": 0.5104403495788574,
      "learning_rate": 8e-05,
      "loss": 1.4963,
      "step": 2802
    },
    {
      "epoch": 0.5821391484942887,
      "grad_norm": 0.5074317455291748,
      "learning_rate": 8e-05,
      "loss": 1.5134,
      "step": 2803
    },
    {
      "epoch": 0.5823468328141226,
      "grad_norm": 0.5222622156143188,
      "learning_rate": 8e-05,
      "loss": 1.525,
      "step": 2804
    },
    {
      "epoch": 0.5825545171339563,
      "grad_norm": 0.5213282108306885,
      "learning_rate": 8e-05,
      "loss": 1.5767,
      "step": 2805
    },
    {
      "epoch": 0.5827622014537902,
      "grad_norm": 0.5274381637573242,
      "learning_rate": 8e-05,
      "loss": 1.6413,
      "step": 2806
    },
    {
      "epoch": 0.5829698857736241,
      "grad_norm": 0.5199068784713745,
      "learning_rate": 8e-05,
      "loss": 1.5654,
      "step": 2807
    },
    {
      "epoch": 0.5831775700934579,
      "grad_norm": 0.5022305846214294,
      "learning_rate": 8e-05,
      "loss": 1.5406,
      "step": 2808
    },
    {
      "epoch": 0.5833852544132918,
      "grad_norm": 0.5012241005897522,
      "learning_rate": 8e-05,
      "loss": 1.5701,
      "step": 2809
    },
    {
      "epoch": 0.5835929387331257,
      "grad_norm": 0.5507093667984009,
      "learning_rate": 8e-05,
      "loss": 1.5963,
      "step": 2810
    },
    {
      "epoch": 0.5838006230529595,
      "grad_norm": 0.5283317565917969,
      "learning_rate": 8e-05,
      "loss": 1.5609,
      "step": 2811
    },
    {
      "epoch": 0.5840083073727933,
      "grad_norm": 0.5261059403419495,
      "learning_rate": 8e-05,
      "loss": 1.5373,
      "step": 2812
    },
    {
      "epoch": 0.5842159916926272,
      "grad_norm": 0.5131132006645203,
      "learning_rate": 8e-05,
      "loss": 1.5572,
      "step": 2813
    },
    {
      "epoch": 0.584423676012461,
      "grad_norm": 0.5188307762145996,
      "learning_rate": 8e-05,
      "loss": 1.5868,
      "step": 2814
    },
    {
      "epoch": 0.5846313603322949,
      "grad_norm": 0.514680802822113,
      "learning_rate": 8e-05,
      "loss": 1.5561,
      "step": 2815
    },
    {
      "epoch": 0.5848390446521288,
      "grad_norm": 0.5244585275650024,
      "learning_rate": 8e-05,
      "loss": 1.6002,
      "step": 2816
    },
    {
      "epoch": 0.5850467289719626,
      "grad_norm": 0.5245557427406311,
      "learning_rate": 8e-05,
      "loss": 1.6044,
      "step": 2817
    },
    {
      "epoch": 0.5852544132917965,
      "grad_norm": 0.51822429895401,
      "learning_rate": 8e-05,
      "loss": 1.5826,
      "step": 2818
    },
    {
      "epoch": 0.5854620976116304,
      "grad_norm": 0.5252143144607544,
      "learning_rate": 8e-05,
      "loss": 1.5347,
      "step": 2819
    },
    {
      "epoch": 0.5856697819314641,
      "grad_norm": 0.5217255353927612,
      "learning_rate": 8e-05,
      "loss": 1.5964,
      "step": 2820
    },
    {
      "epoch": 0.585877466251298,
      "grad_norm": 0.5508246421813965,
      "learning_rate": 8e-05,
      "loss": 1.6135,
      "step": 2821
    },
    {
      "epoch": 0.5860851505711319,
      "grad_norm": 0.5165861248970032,
      "learning_rate": 8e-05,
      "loss": 1.5907,
      "step": 2822
    },
    {
      "epoch": 0.5862928348909657,
      "grad_norm": 0.5337631702423096,
      "learning_rate": 8e-05,
      "loss": 1.5683,
      "step": 2823
    },
    {
      "epoch": 0.5865005192107996,
      "grad_norm": 0.5257221460342407,
      "learning_rate": 8e-05,
      "loss": 1.6359,
      "step": 2824
    },
    {
      "epoch": 0.5867082035306335,
      "grad_norm": 0.5221315622329712,
      "learning_rate": 8e-05,
      "loss": 1.5157,
      "step": 2825
    },
    {
      "epoch": 0.5869158878504673,
      "grad_norm": 0.5321351289749146,
      "learning_rate": 8e-05,
      "loss": 1.5844,
      "step": 2826
    },
    {
      "epoch": 0.5871235721703011,
      "grad_norm": 0.5378949046134949,
      "learning_rate": 8e-05,
      "loss": 1.5158,
      "step": 2827
    },
    {
      "epoch": 0.587331256490135,
      "grad_norm": 0.5337066650390625,
      "learning_rate": 8e-05,
      "loss": 1.537,
      "step": 2828
    },
    {
      "epoch": 0.5875389408099688,
      "grad_norm": 0.5137457251548767,
      "learning_rate": 8e-05,
      "loss": 1.5569,
      "step": 2829
    },
    {
      "epoch": 0.5877466251298027,
      "grad_norm": 0.5192795395851135,
      "learning_rate": 8e-05,
      "loss": 1.5328,
      "step": 2830
    },
    {
      "epoch": 0.5879543094496366,
      "grad_norm": 0.5277906656265259,
      "learning_rate": 8e-05,
      "loss": 1.5183,
      "step": 2831
    },
    {
      "epoch": 0.5881619937694704,
      "grad_norm": 0.550859272480011,
      "learning_rate": 8e-05,
      "loss": 1.6268,
      "step": 2832
    },
    {
      "epoch": 0.5883696780893043,
      "grad_norm": 0.518708348274231,
      "learning_rate": 8e-05,
      "loss": 1.4964,
      "step": 2833
    },
    {
      "epoch": 0.5885773624091382,
      "grad_norm": 0.5725675225257874,
      "learning_rate": 8e-05,
      "loss": 1.6776,
      "step": 2834
    },
    {
      "epoch": 0.5887850467289719,
      "grad_norm": 0.512139081954956,
      "learning_rate": 8e-05,
      "loss": 1.6053,
      "step": 2835
    },
    {
      "epoch": 0.5889927310488058,
      "grad_norm": 0.5205104947090149,
      "learning_rate": 8e-05,
      "loss": 1.5682,
      "step": 2836
    },
    {
      "epoch": 0.5892004153686397,
      "grad_norm": 0.5213608741760254,
      "learning_rate": 8e-05,
      "loss": 1.5272,
      "step": 2837
    },
    {
      "epoch": 0.5894080996884735,
      "grad_norm": 0.51758873462677,
      "learning_rate": 8e-05,
      "loss": 1.6262,
      "step": 2838
    },
    {
      "epoch": 0.5896157840083074,
      "grad_norm": 0.5142424702644348,
      "learning_rate": 8e-05,
      "loss": 1.59,
      "step": 2839
    },
    {
      "epoch": 0.5898234683281413,
      "grad_norm": 0.5190350413322449,
      "learning_rate": 8e-05,
      "loss": 1.586,
      "step": 2840
    },
    {
      "epoch": 0.5900311526479751,
      "grad_norm": 0.5234148502349854,
      "learning_rate": 8e-05,
      "loss": 1.5719,
      "step": 2841
    },
    {
      "epoch": 0.5902388369678089,
      "grad_norm": 0.5219846367835999,
      "learning_rate": 8e-05,
      "loss": 1.5988,
      "step": 2842
    },
    {
      "epoch": 0.5904465212876427,
      "grad_norm": 0.5159264206886292,
      "learning_rate": 8e-05,
      "loss": 1.458,
      "step": 2843
    },
    {
      "epoch": 0.5906542056074766,
      "grad_norm": 0.5145145058631897,
      "learning_rate": 8e-05,
      "loss": 1.5651,
      "step": 2844
    },
    {
      "epoch": 0.5908618899273105,
      "grad_norm": 0.5303623676300049,
      "learning_rate": 8e-05,
      "loss": 1.5558,
      "step": 2845
    },
    {
      "epoch": 0.5910695742471443,
      "grad_norm": 0.5262109041213989,
      "learning_rate": 8e-05,
      "loss": 1.5433,
      "step": 2846
    },
    {
      "epoch": 0.5912772585669782,
      "grad_norm": 0.5659428834915161,
      "learning_rate": 8e-05,
      "loss": 1.5229,
      "step": 2847
    },
    {
      "epoch": 0.5914849428868121,
      "grad_norm": 0.5190215110778809,
      "learning_rate": 8e-05,
      "loss": 1.5649,
      "step": 2848
    },
    {
      "epoch": 0.591692627206646,
      "grad_norm": 0.5212072134017944,
      "learning_rate": 8e-05,
      "loss": 1.4807,
      "step": 2849
    },
    {
      "epoch": 0.5919003115264797,
      "grad_norm": 0.5107869505882263,
      "learning_rate": 8e-05,
      "loss": 1.5145,
      "step": 2850
    },
    {
      "epoch": 0.5921079958463136,
      "grad_norm": 0.4874977469444275,
      "learning_rate": 8e-05,
      "loss": 1.5159,
      "step": 2851
    },
    {
      "epoch": 0.5923156801661474,
      "grad_norm": 0.5429779291152954,
      "learning_rate": 8e-05,
      "loss": 1.5825,
      "step": 2852
    },
    {
      "epoch": 0.5925233644859813,
      "grad_norm": 0.5432573556900024,
      "learning_rate": 8e-05,
      "loss": 1.6157,
      "step": 2853
    },
    {
      "epoch": 0.5927310488058152,
      "grad_norm": 0.5516713261604309,
      "learning_rate": 8e-05,
      "loss": 1.5427,
      "step": 2854
    },
    {
      "epoch": 0.592938733125649,
      "grad_norm": 0.5550795197486877,
      "learning_rate": 8e-05,
      "loss": 1.6004,
      "step": 2855
    },
    {
      "epoch": 0.5931464174454829,
      "grad_norm": 0.5827699899673462,
      "learning_rate": 8e-05,
      "loss": 1.5963,
      "step": 2856
    },
    {
      "epoch": 0.5933541017653167,
      "grad_norm": 0.5084548592567444,
      "learning_rate": 8e-05,
      "loss": 1.5318,
      "step": 2857
    },
    {
      "epoch": 0.5935617860851505,
      "grad_norm": 0.49697956442832947,
      "learning_rate": 8e-05,
      "loss": 1.5825,
      "step": 2858
    },
    {
      "epoch": 0.5937694704049844,
      "grad_norm": 0.5428705215454102,
      "learning_rate": 8e-05,
      "loss": 1.5933,
      "step": 2859
    },
    {
      "epoch": 0.5939771547248183,
      "grad_norm": 0.5335211157798767,
      "learning_rate": 8e-05,
      "loss": 1.5817,
      "step": 2860
    },
    {
      "epoch": 0.5941848390446521,
      "grad_norm": 0.524454653263092,
      "learning_rate": 8e-05,
      "loss": 1.5548,
      "step": 2861
    },
    {
      "epoch": 0.594392523364486,
      "grad_norm": 0.5151386857032776,
      "learning_rate": 8e-05,
      "loss": 1.5608,
      "step": 2862
    },
    {
      "epoch": 0.5946002076843199,
      "grad_norm": 0.5568569898605347,
      "learning_rate": 8e-05,
      "loss": 1.6079,
      "step": 2863
    },
    {
      "epoch": 0.5948078920041537,
      "grad_norm": 0.5202062726020813,
      "learning_rate": 8e-05,
      "loss": 1.5789,
      "step": 2864
    },
    {
      "epoch": 0.5950155763239875,
      "grad_norm": 0.5226384997367859,
      "learning_rate": 8e-05,
      "loss": 1.6255,
      "step": 2865
    },
    {
      "epoch": 0.5952232606438214,
      "grad_norm": 0.5193942189216614,
      "learning_rate": 8e-05,
      "loss": 1.5521,
      "step": 2866
    },
    {
      "epoch": 0.5954309449636552,
      "grad_norm": 0.5268439650535583,
      "learning_rate": 8e-05,
      "loss": 1.5577,
      "step": 2867
    },
    {
      "epoch": 0.5956386292834891,
      "grad_norm": 0.5028793811798096,
      "learning_rate": 8e-05,
      "loss": 1.5542,
      "step": 2868
    },
    {
      "epoch": 0.595846313603323,
      "grad_norm": 0.5032245516777039,
      "learning_rate": 8e-05,
      "loss": 1.5401,
      "step": 2869
    },
    {
      "epoch": 0.5960539979231568,
      "grad_norm": 0.5338500738143921,
      "learning_rate": 8e-05,
      "loss": 1.6173,
      "step": 2870
    },
    {
      "epoch": 0.5962616822429907,
      "grad_norm": 0.5496775507926941,
      "learning_rate": 8e-05,
      "loss": 1.5215,
      "step": 2871
    },
    {
      "epoch": 0.5964693665628245,
      "grad_norm": 0.5108234882354736,
      "learning_rate": 8e-05,
      "loss": 1.5494,
      "step": 2872
    },
    {
      "epoch": 0.5966770508826583,
      "grad_norm": 0.5167278051376343,
      "learning_rate": 8e-05,
      "loss": 1.4923,
      "step": 2873
    },
    {
      "epoch": 0.5968847352024922,
      "grad_norm": 0.5146141052246094,
      "learning_rate": 8e-05,
      "loss": 1.5076,
      "step": 2874
    },
    {
      "epoch": 0.5970924195223261,
      "grad_norm": 0.5203628540039062,
      "learning_rate": 8e-05,
      "loss": 1.5583,
      "step": 2875
    },
    {
      "epoch": 0.5973001038421599,
      "grad_norm": 0.5065957903862,
      "learning_rate": 8e-05,
      "loss": 1.5551,
      "step": 2876
    },
    {
      "epoch": 0.5975077881619938,
      "grad_norm": 0.5048226118087769,
      "learning_rate": 8e-05,
      "loss": 1.4814,
      "step": 2877
    },
    {
      "epoch": 0.5977154724818277,
      "grad_norm": 0.5416272282600403,
      "learning_rate": 8e-05,
      "loss": 1.6222,
      "step": 2878
    },
    {
      "epoch": 0.5979231568016615,
      "grad_norm": 0.5329111814498901,
      "learning_rate": 8e-05,
      "loss": 1.5685,
      "step": 2879
    },
    {
      "epoch": 0.5981308411214953,
      "grad_norm": 0.5529356002807617,
      "learning_rate": 8e-05,
      "loss": 1.6551,
      "step": 2880
    },
    {
      "epoch": 0.5983385254413291,
      "grad_norm": 0.5571921467781067,
      "learning_rate": 8e-05,
      "loss": 1.6435,
      "step": 2881
    },
    {
      "epoch": 0.598546209761163,
      "grad_norm": 0.5245065093040466,
      "learning_rate": 8e-05,
      "loss": 1.5972,
      "step": 2882
    },
    {
      "epoch": 0.5987538940809969,
      "grad_norm": 0.5337019562721252,
      "learning_rate": 8e-05,
      "loss": 1.5105,
      "step": 2883
    },
    {
      "epoch": 0.5989615784008308,
      "grad_norm": 0.5535258650779724,
      "learning_rate": 8e-05,
      "loss": 1.6439,
      "step": 2884
    },
    {
      "epoch": 0.5991692627206646,
      "grad_norm": 0.510281503200531,
      "learning_rate": 8e-05,
      "loss": 1.5489,
      "step": 2885
    },
    {
      "epoch": 0.5993769470404985,
      "grad_norm": 0.5224695801734924,
      "learning_rate": 8e-05,
      "loss": 1.5011,
      "step": 2886
    },
    {
      "epoch": 0.5995846313603322,
      "grad_norm": 0.533809244632721,
      "learning_rate": 8e-05,
      "loss": 1.6423,
      "step": 2887
    },
    {
      "epoch": 0.5997923156801661,
      "grad_norm": 0.574894368648529,
      "learning_rate": 8e-05,
      "loss": 1.5208,
      "step": 2888
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.5299583673477173,
      "learning_rate": 8e-05,
      "loss": 1.5983,
      "step": 2889
    },
    {
      "epoch": 0.6002076843198338,
      "grad_norm": 0.5273470878601074,
      "learning_rate": 8e-05,
      "loss": 1.5455,
      "step": 2890
    },
    {
      "epoch": 0.6004153686396677,
      "grad_norm": 0.5376371145248413,
      "learning_rate": 8e-05,
      "loss": 1.5338,
      "step": 2891
    },
    {
      "epoch": 0.6006230529595016,
      "grad_norm": 0.5445219278335571,
      "learning_rate": 8e-05,
      "loss": 1.6045,
      "step": 2892
    },
    {
      "epoch": 0.6008307372793354,
      "grad_norm": 0.5201619267463684,
      "learning_rate": 8e-05,
      "loss": 1.5215,
      "step": 2893
    },
    {
      "epoch": 0.6010384215991693,
      "grad_norm": 0.5154725909233093,
      "learning_rate": 8e-05,
      "loss": 1.5279,
      "step": 2894
    },
    {
      "epoch": 0.6012461059190031,
      "grad_norm": 0.5375370979309082,
      "learning_rate": 8e-05,
      "loss": 1.5252,
      "step": 2895
    },
    {
      "epoch": 0.6014537902388369,
      "grad_norm": 0.5726920366287231,
      "learning_rate": 8e-05,
      "loss": 1.5772,
      "step": 2896
    },
    {
      "epoch": 0.6016614745586708,
      "grad_norm": 0.5069287419319153,
      "learning_rate": 8e-05,
      "loss": 1.5117,
      "step": 2897
    },
    {
      "epoch": 0.6018691588785047,
      "grad_norm": 0.5421019792556763,
      "learning_rate": 8e-05,
      "loss": 1.5698,
      "step": 2898
    },
    {
      "epoch": 0.6020768431983385,
      "grad_norm": 0.5770023465156555,
      "learning_rate": 8e-05,
      "loss": 1.6081,
      "step": 2899
    },
    {
      "epoch": 0.6022845275181724,
      "grad_norm": 0.5031346678733826,
      "learning_rate": 8e-05,
      "loss": 1.5555,
      "step": 2900
    },
    {
      "epoch": 0.6024922118380063,
      "grad_norm": 0.5268486142158508,
      "learning_rate": 8e-05,
      "loss": 1.5491,
      "step": 2901
    },
    {
      "epoch": 0.60269989615784,
      "grad_norm": 0.5094104409217834,
      "learning_rate": 8e-05,
      "loss": 1.546,
      "step": 2902
    },
    {
      "epoch": 0.6029075804776739,
      "grad_norm": 0.5445466637611389,
      "learning_rate": 8e-05,
      "loss": 1.5345,
      "step": 2903
    },
    {
      "epoch": 0.6031152647975078,
      "grad_norm": 0.5537915229797363,
      "learning_rate": 8e-05,
      "loss": 1.5728,
      "step": 2904
    },
    {
      "epoch": 0.6033229491173416,
      "grad_norm": 0.5232887268066406,
      "learning_rate": 8e-05,
      "loss": 1.5328,
      "step": 2905
    },
    {
      "epoch": 0.6035306334371755,
      "grad_norm": 0.5250490307807922,
      "learning_rate": 8e-05,
      "loss": 1.6072,
      "step": 2906
    },
    {
      "epoch": 0.6037383177570094,
      "grad_norm": 0.5186933279037476,
      "learning_rate": 8e-05,
      "loss": 1.5354,
      "step": 2907
    },
    {
      "epoch": 0.6039460020768432,
      "grad_norm": 0.5262655019760132,
      "learning_rate": 8e-05,
      "loss": 1.5914,
      "step": 2908
    },
    {
      "epoch": 0.6041536863966771,
      "grad_norm": 0.5019351243972778,
      "learning_rate": 8e-05,
      "loss": 1.5235,
      "step": 2909
    },
    {
      "epoch": 0.6043613707165109,
      "grad_norm": 0.4954097270965576,
      "learning_rate": 8e-05,
      "loss": 1.5543,
      "step": 2910
    },
    {
      "epoch": 0.6045690550363447,
      "grad_norm": 0.512431800365448,
      "learning_rate": 8e-05,
      "loss": 1.5578,
      "step": 2911
    },
    {
      "epoch": 0.6047767393561786,
      "grad_norm": 0.5076355934143066,
      "learning_rate": 8e-05,
      "loss": 1.561,
      "step": 2912
    },
    {
      "epoch": 0.6049844236760125,
      "grad_norm": 0.5237712860107422,
      "learning_rate": 8e-05,
      "loss": 1.5584,
      "step": 2913
    },
    {
      "epoch": 0.6051921079958463,
      "grad_norm": 0.5152472257614136,
      "learning_rate": 8e-05,
      "loss": 1.5429,
      "step": 2914
    },
    {
      "epoch": 0.6053997923156802,
      "grad_norm": 0.5119134783744812,
      "learning_rate": 8e-05,
      "loss": 1.5631,
      "step": 2915
    },
    {
      "epoch": 0.6056074766355141,
      "grad_norm": 0.5183678865432739,
      "learning_rate": 8e-05,
      "loss": 1.5855,
      "step": 2916
    },
    {
      "epoch": 0.6058151609553478,
      "grad_norm": 0.49895596504211426,
      "learning_rate": 8e-05,
      "loss": 1.5152,
      "step": 2917
    },
    {
      "epoch": 0.6060228452751817,
      "grad_norm": 0.5047574043273926,
      "learning_rate": 8e-05,
      "loss": 1.5017,
      "step": 2918
    },
    {
      "epoch": 0.6062305295950156,
      "grad_norm": 0.5099537372589111,
      "learning_rate": 8e-05,
      "loss": 1.6106,
      "step": 2919
    },
    {
      "epoch": 0.6064382139148494,
      "grad_norm": 0.5520958304405212,
      "learning_rate": 8e-05,
      "loss": 1.6264,
      "step": 2920
    },
    {
      "epoch": 0.6066458982346833,
      "grad_norm": 0.5420238375663757,
      "learning_rate": 8e-05,
      "loss": 1.4995,
      "step": 2921
    },
    {
      "epoch": 0.6068535825545172,
      "grad_norm": 0.5576627850532532,
      "learning_rate": 8e-05,
      "loss": 1.559,
      "step": 2922
    },
    {
      "epoch": 0.607061266874351,
      "grad_norm": 0.5126776695251465,
      "learning_rate": 8e-05,
      "loss": 1.5922,
      "step": 2923
    },
    {
      "epoch": 0.6072689511941849,
      "grad_norm": 0.525304913520813,
      "learning_rate": 8e-05,
      "loss": 1.6135,
      "step": 2924
    },
    {
      "epoch": 0.6074766355140186,
      "grad_norm": 0.5227261781692505,
      "learning_rate": 8e-05,
      "loss": 1.6031,
      "step": 2925
    },
    {
      "epoch": 0.6076843198338525,
      "grad_norm": 0.5197954773902893,
      "learning_rate": 8e-05,
      "loss": 1.5591,
      "step": 2926
    },
    {
      "epoch": 0.6078920041536864,
      "grad_norm": 0.5202577114105225,
      "learning_rate": 8e-05,
      "loss": 1.5878,
      "step": 2927
    },
    {
      "epoch": 0.6080996884735203,
      "grad_norm": 0.5209299921989441,
      "learning_rate": 8e-05,
      "loss": 1.5847,
      "step": 2928
    },
    {
      "epoch": 0.6083073727933541,
      "grad_norm": 0.5529150366783142,
      "learning_rate": 8e-05,
      "loss": 1.5966,
      "step": 2929
    },
    {
      "epoch": 0.608515057113188,
      "grad_norm": 0.5113958120346069,
      "learning_rate": 8e-05,
      "loss": 1.5736,
      "step": 2930
    },
    {
      "epoch": 0.6087227414330219,
      "grad_norm": 0.5247915983200073,
      "learning_rate": 8e-05,
      "loss": 1.5975,
      "step": 2931
    },
    {
      "epoch": 0.6089304257528556,
      "grad_norm": 0.5492368340492249,
      "learning_rate": 8e-05,
      "loss": 1.5366,
      "step": 2932
    },
    {
      "epoch": 0.6091381100726895,
      "grad_norm": 0.5445862412452698,
      "learning_rate": 8e-05,
      "loss": 1.5235,
      "step": 2933
    },
    {
      "epoch": 0.6093457943925233,
      "grad_norm": 0.5135022401809692,
      "learning_rate": 8e-05,
      "loss": 1.5669,
      "step": 2934
    },
    {
      "epoch": 0.6095534787123572,
      "grad_norm": 0.5535476803779602,
      "learning_rate": 8e-05,
      "loss": 1.5846,
      "step": 2935
    },
    {
      "epoch": 0.6097611630321911,
      "grad_norm": 0.5978537797927856,
      "learning_rate": 8e-05,
      "loss": 1.5923,
      "step": 2936
    },
    {
      "epoch": 0.609968847352025,
      "grad_norm": 0.5011223554611206,
      "learning_rate": 8e-05,
      "loss": 1.5336,
      "step": 2937
    },
    {
      "epoch": 0.6101765316718588,
      "grad_norm": 0.5410469770431519,
      "learning_rate": 8e-05,
      "loss": 1.5075,
      "step": 2938
    },
    {
      "epoch": 0.6103842159916927,
      "grad_norm": 0.5120070576667786,
      "learning_rate": 8e-05,
      "loss": 1.5891,
      "step": 2939
    },
    {
      "epoch": 0.6105919003115264,
      "grad_norm": 0.6055616140365601,
      "learning_rate": 8e-05,
      "loss": 1.4815,
      "step": 2940
    },
    {
      "epoch": 0.6107995846313603,
      "grad_norm": 0.5197583436965942,
      "learning_rate": 8e-05,
      "loss": 1.5151,
      "step": 2941
    },
    {
      "epoch": 0.6110072689511942,
      "grad_norm": 0.5293586850166321,
      "learning_rate": 8e-05,
      "loss": 1.6045,
      "step": 2942
    },
    {
      "epoch": 0.611214953271028,
      "grad_norm": 0.5343866348266602,
      "learning_rate": 8e-05,
      "loss": 1.5406,
      "step": 2943
    },
    {
      "epoch": 0.6114226375908619,
      "grad_norm": 0.5327885746955872,
      "learning_rate": 8e-05,
      "loss": 1.5581,
      "step": 2944
    },
    {
      "epoch": 0.6116303219106958,
      "grad_norm": 0.5412554740905762,
      "learning_rate": 8e-05,
      "loss": 1.6321,
      "step": 2945
    },
    {
      "epoch": 0.6118380062305296,
      "grad_norm": 0.5238969922065735,
      "learning_rate": 8e-05,
      "loss": 1.611,
      "step": 2946
    },
    {
      "epoch": 0.6120456905503634,
      "grad_norm": 0.5044591426849365,
      "learning_rate": 8e-05,
      "loss": 1.5551,
      "step": 2947
    },
    {
      "epoch": 0.6122533748701973,
      "grad_norm": 0.6135985255241394,
      "learning_rate": 8e-05,
      "loss": 1.5487,
      "step": 2948
    },
    {
      "epoch": 0.6124610591900311,
      "grad_norm": 0.5363000631332397,
      "learning_rate": 8e-05,
      "loss": 1.5965,
      "step": 2949
    },
    {
      "epoch": 0.612668743509865,
      "grad_norm": 0.5959219932556152,
      "learning_rate": 8e-05,
      "loss": 1.642,
      "step": 2950
    },
    {
      "epoch": 0.6128764278296989,
      "grad_norm": 0.5022416710853577,
      "learning_rate": 8e-05,
      "loss": 1.5019,
      "step": 2951
    },
    {
      "epoch": 0.6130841121495327,
      "grad_norm": 0.5241201519966125,
      "learning_rate": 8e-05,
      "loss": 1.5078,
      "step": 2952
    },
    {
      "epoch": 0.6132917964693666,
      "grad_norm": 0.5270079970359802,
      "learning_rate": 8e-05,
      "loss": 1.5387,
      "step": 2953
    },
    {
      "epoch": 0.6134994807892005,
      "grad_norm": 0.548295795917511,
      "learning_rate": 8e-05,
      "loss": 1.6071,
      "step": 2954
    },
    {
      "epoch": 0.6137071651090342,
      "grad_norm": 0.5182678699493408,
      "learning_rate": 8e-05,
      "loss": 1.5863,
      "step": 2955
    },
    {
      "epoch": 0.6139148494288681,
      "grad_norm": 0.5253292918205261,
      "learning_rate": 8e-05,
      "loss": 1.5464,
      "step": 2956
    },
    {
      "epoch": 0.614122533748702,
      "grad_norm": 0.492260605096817,
      "learning_rate": 8e-05,
      "loss": 1.4529,
      "step": 2957
    },
    {
      "epoch": 0.6143302180685358,
      "grad_norm": 0.6125956773757935,
      "learning_rate": 8e-05,
      "loss": 1.5692,
      "step": 2958
    },
    {
      "epoch": 0.6145379023883697,
      "grad_norm": 0.6347879767417908,
      "learning_rate": 8e-05,
      "loss": 1.6548,
      "step": 2959
    },
    {
      "epoch": 0.6147455867082036,
      "grad_norm": 0.5303540825843811,
      "learning_rate": 8e-05,
      "loss": 1.6188,
      "step": 2960
    },
    {
      "epoch": 0.6149532710280374,
      "grad_norm": 0.5319226384162903,
      "learning_rate": 8e-05,
      "loss": 1.497,
      "step": 2961
    },
    {
      "epoch": 0.6151609553478712,
      "grad_norm": 0.5587071180343628,
      "learning_rate": 8e-05,
      "loss": 1.6561,
      "step": 2962
    },
    {
      "epoch": 0.615368639667705,
      "grad_norm": 0.5441991686820984,
      "learning_rate": 8e-05,
      "loss": 1.5086,
      "step": 2963
    },
    {
      "epoch": 0.6155763239875389,
      "grad_norm": 0.4938202500343323,
      "learning_rate": 8e-05,
      "loss": 1.5147,
      "step": 2964
    },
    {
      "epoch": 0.6157840083073728,
      "grad_norm": 0.5410319566726685,
      "learning_rate": 8e-05,
      "loss": 1.6568,
      "step": 2965
    },
    {
      "epoch": 0.6159916926272067,
      "grad_norm": 0.531709611415863,
      "learning_rate": 8e-05,
      "loss": 1.5243,
      "step": 2966
    },
    {
      "epoch": 0.6161993769470405,
      "grad_norm": 0.5373491644859314,
      "learning_rate": 8e-05,
      "loss": 1.581,
      "step": 2967
    },
    {
      "epoch": 0.6164070612668744,
      "grad_norm": 0.5017946362495422,
      "learning_rate": 8e-05,
      "loss": 1.5224,
      "step": 2968
    },
    {
      "epoch": 0.6166147455867083,
      "grad_norm": 0.510314404964447,
      "learning_rate": 8e-05,
      "loss": 1.6072,
      "step": 2969
    },
    {
      "epoch": 0.616822429906542,
      "grad_norm": 0.519645631313324,
      "learning_rate": 8e-05,
      "loss": 1.5428,
      "step": 2970
    },
    {
      "epoch": 0.6170301142263759,
      "grad_norm": 0.5236020088195801,
      "learning_rate": 8e-05,
      "loss": 1.5758,
      "step": 2971
    },
    {
      "epoch": 0.6172377985462097,
      "grad_norm": 0.521712064743042,
      "learning_rate": 8e-05,
      "loss": 1.5461,
      "step": 2972
    },
    {
      "epoch": 0.6174454828660436,
      "grad_norm": 0.5110123157501221,
      "learning_rate": 8e-05,
      "loss": 1.5284,
      "step": 2973
    },
    {
      "epoch": 0.6176531671858775,
      "grad_norm": 0.5233707427978516,
      "learning_rate": 8e-05,
      "loss": 1.5502,
      "step": 2974
    },
    {
      "epoch": 0.6178608515057114,
      "grad_norm": 0.532986581325531,
      "learning_rate": 8e-05,
      "loss": 1.5442,
      "step": 2975
    },
    {
      "epoch": 0.6180685358255452,
      "grad_norm": 0.5192633867263794,
      "learning_rate": 8e-05,
      "loss": 1.5694,
      "step": 2976
    },
    {
      "epoch": 0.618276220145379,
      "grad_norm": 0.5456683039665222,
      "learning_rate": 8e-05,
      "loss": 1.545,
      "step": 2977
    },
    {
      "epoch": 0.6184839044652128,
      "grad_norm": 0.5262755751609802,
      "learning_rate": 8e-05,
      "loss": 1.5764,
      "step": 2978
    },
    {
      "epoch": 0.6186915887850467,
      "grad_norm": 0.5116017460823059,
      "learning_rate": 8e-05,
      "loss": 1.6291,
      "step": 2979
    },
    {
      "epoch": 0.6188992731048806,
      "grad_norm": 0.5196945667266846,
      "learning_rate": 8e-05,
      "loss": 1.5285,
      "step": 2980
    },
    {
      "epoch": 0.6191069574247144,
      "grad_norm": 0.525067925453186,
      "learning_rate": 8e-05,
      "loss": 1.5704,
      "step": 2981
    },
    {
      "epoch": 0.6193146417445483,
      "grad_norm": 0.5199265480041504,
      "learning_rate": 8e-05,
      "loss": 1.5554,
      "step": 2982
    },
    {
      "epoch": 0.6195223260643822,
      "grad_norm": 0.5034192204475403,
      "learning_rate": 8e-05,
      "loss": 1.5135,
      "step": 2983
    },
    {
      "epoch": 0.619730010384216,
      "grad_norm": 0.5294010043144226,
      "learning_rate": 8e-05,
      "loss": 1.5838,
      "step": 2984
    },
    {
      "epoch": 0.6199376947040498,
      "grad_norm": 0.5039417743682861,
      "learning_rate": 8e-05,
      "loss": 1.5084,
      "step": 2985
    },
    {
      "epoch": 0.6201453790238837,
      "grad_norm": 0.5068845152854919,
      "learning_rate": 8e-05,
      "loss": 1.5872,
      "step": 2986
    },
    {
      "epoch": 0.6203530633437175,
      "grad_norm": 0.521058201789856,
      "learning_rate": 8e-05,
      "loss": 1.4993,
      "step": 2987
    },
    {
      "epoch": 0.6205607476635514,
      "grad_norm": 0.5185069441795349,
      "learning_rate": 8e-05,
      "loss": 1.54,
      "step": 2988
    },
    {
      "epoch": 0.6207684319833853,
      "grad_norm": 0.5288187265396118,
      "learning_rate": 8e-05,
      "loss": 1.5946,
      "step": 2989
    },
    {
      "epoch": 0.6209761163032191,
      "grad_norm": 0.5338004231452942,
      "learning_rate": 8e-05,
      "loss": 1.5141,
      "step": 2990
    },
    {
      "epoch": 0.621183800623053,
      "grad_norm": 0.5099069476127625,
      "learning_rate": 8e-05,
      "loss": 1.5778,
      "step": 2991
    },
    {
      "epoch": 0.6213914849428868,
      "grad_norm": 0.5432291626930237,
      "learning_rate": 8e-05,
      "loss": 1.5671,
      "step": 2992
    },
    {
      "epoch": 0.6215991692627206,
      "grad_norm": 0.5076735019683838,
      "learning_rate": 8e-05,
      "loss": 1.5615,
      "step": 2993
    },
    {
      "epoch": 0.6218068535825545,
      "grad_norm": 0.5396541357040405,
      "learning_rate": 8e-05,
      "loss": 1.5829,
      "step": 2994
    },
    {
      "epoch": 0.6220145379023884,
      "grad_norm": 0.5078719854354858,
      "learning_rate": 8e-05,
      "loss": 1.5176,
      "step": 2995
    },
    {
      "epoch": 0.6222222222222222,
      "grad_norm": 0.503608226776123,
      "learning_rate": 8e-05,
      "loss": 1.531,
      "step": 2996
    },
    {
      "epoch": 0.6224299065420561,
      "grad_norm": 0.5179398655891418,
      "learning_rate": 8e-05,
      "loss": 1.6097,
      "step": 2997
    },
    {
      "epoch": 0.62263759086189,
      "grad_norm": 0.5163557529449463,
      "learning_rate": 8e-05,
      "loss": 1.6221,
      "step": 2998
    },
    {
      "epoch": 0.6228452751817238,
      "grad_norm": 0.5109874606132507,
      "learning_rate": 8e-05,
      "loss": 1.5046,
      "step": 2999
    },
    {
      "epoch": 0.6230529595015576,
      "grad_norm": 0.5169478058815002,
      "learning_rate": 8e-05,
      "loss": 1.5841,
      "step": 3000
    },
    {
      "epoch": 0.6232606438213915,
      "grad_norm": 0.5242609977722168,
      "learning_rate": 8e-05,
      "loss": 1.5208,
      "step": 3001
    },
    {
      "epoch": 0.6234683281412253,
      "grad_norm": 0.516659140586853,
      "learning_rate": 8e-05,
      "loss": 1.544,
      "step": 3002
    },
    {
      "epoch": 0.6236760124610592,
      "grad_norm": 0.5018888115882874,
      "learning_rate": 8e-05,
      "loss": 1.5253,
      "step": 3003
    },
    {
      "epoch": 0.6238836967808931,
      "grad_norm": 0.496075302362442,
      "learning_rate": 8e-05,
      "loss": 1.502,
      "step": 3004
    },
    {
      "epoch": 0.6240913811007269,
      "grad_norm": 0.4930664300918579,
      "learning_rate": 8e-05,
      "loss": 1.5183,
      "step": 3005
    },
    {
      "epoch": 0.6242990654205608,
      "grad_norm": 0.5343693494796753,
      "learning_rate": 8e-05,
      "loss": 1.5741,
      "step": 3006
    },
    {
      "epoch": 0.6245067497403946,
      "grad_norm": 0.5333400964736938,
      "learning_rate": 8e-05,
      "loss": 1.5834,
      "step": 3007
    },
    {
      "epoch": 0.6247144340602284,
      "grad_norm": 0.5007697939872742,
      "learning_rate": 8e-05,
      "loss": 1.4562,
      "step": 3008
    },
    {
      "epoch": 0.6249221183800623,
      "grad_norm": 0.5132413506507874,
      "learning_rate": 8e-05,
      "loss": 1.5022,
      "step": 3009
    },
    {
      "epoch": 0.6251298026998962,
      "grad_norm": 0.5218618512153625,
      "learning_rate": 8e-05,
      "loss": 1.5203,
      "step": 3010
    },
    {
      "epoch": 0.62533748701973,
      "grad_norm": 0.5128524303436279,
      "learning_rate": 8e-05,
      "loss": 1.6179,
      "step": 3011
    },
    {
      "epoch": 0.6255451713395639,
      "grad_norm": 0.5142005681991577,
      "learning_rate": 8e-05,
      "loss": 1.5895,
      "step": 3012
    },
    {
      "epoch": 0.6257528556593978,
      "grad_norm": 0.5448899865150452,
      "learning_rate": 8e-05,
      "loss": 1.5996,
      "step": 3013
    },
    {
      "epoch": 0.6259605399792316,
      "grad_norm": 0.5296397805213928,
      "learning_rate": 8e-05,
      "loss": 1.5697,
      "step": 3014
    },
    {
      "epoch": 0.6261682242990654,
      "grad_norm": 0.5109992623329163,
      "learning_rate": 8e-05,
      "loss": 1.5157,
      "step": 3015
    },
    {
      "epoch": 0.6263759086188992,
      "grad_norm": 0.5397012829780579,
      "learning_rate": 8e-05,
      "loss": 1.5967,
      "step": 3016
    },
    {
      "epoch": 0.6265835929387331,
      "grad_norm": 0.5059266090393066,
      "learning_rate": 8e-05,
      "loss": 1.5431,
      "step": 3017
    },
    {
      "epoch": 0.626791277258567,
      "grad_norm": 0.49859824776649475,
      "learning_rate": 8e-05,
      "loss": 1.4803,
      "step": 3018
    },
    {
      "epoch": 0.6269989615784008,
      "grad_norm": 0.5274714827537537,
      "learning_rate": 8e-05,
      "loss": 1.5688,
      "step": 3019
    },
    {
      "epoch": 0.6272066458982347,
      "grad_norm": 0.5198060274124146,
      "learning_rate": 8e-05,
      "loss": 1.497,
      "step": 3020
    },
    {
      "epoch": 0.6274143302180686,
      "grad_norm": 0.5041719675064087,
      "learning_rate": 8e-05,
      "loss": 1.4922,
      "step": 3021
    },
    {
      "epoch": 0.6276220145379023,
      "grad_norm": 0.5605435967445374,
      "learning_rate": 8e-05,
      "loss": 1.5527,
      "step": 3022
    },
    {
      "epoch": 0.6278296988577362,
      "grad_norm": 0.5161044597625732,
      "learning_rate": 8e-05,
      "loss": 1.5863,
      "step": 3023
    },
    {
      "epoch": 0.6280373831775701,
      "grad_norm": 0.5125625133514404,
      "learning_rate": 8e-05,
      "loss": 1.53,
      "step": 3024
    },
    {
      "epoch": 0.6282450674974039,
      "grad_norm": 0.5024579167366028,
      "learning_rate": 8e-05,
      "loss": 1.557,
      "step": 3025
    },
    {
      "epoch": 0.6284527518172378,
      "grad_norm": 0.5236228108406067,
      "learning_rate": 8e-05,
      "loss": 1.5145,
      "step": 3026
    },
    {
      "epoch": 0.6286604361370717,
      "grad_norm": 0.5330371260643005,
      "learning_rate": 8e-05,
      "loss": 1.5926,
      "step": 3027
    },
    {
      "epoch": 0.6288681204569055,
      "grad_norm": 0.5498360991477966,
      "learning_rate": 8e-05,
      "loss": 1.6321,
      "step": 3028
    },
    {
      "epoch": 0.6290758047767394,
      "grad_norm": 0.5137662887573242,
      "learning_rate": 8e-05,
      "loss": 1.5616,
      "step": 3029
    },
    {
      "epoch": 0.6292834890965732,
      "grad_norm": 0.528255820274353,
      "learning_rate": 8e-05,
      "loss": 1.5993,
      "step": 3030
    },
    {
      "epoch": 0.629491173416407,
      "grad_norm": 0.5130833387374878,
      "learning_rate": 8e-05,
      "loss": 1.5434,
      "step": 3031
    },
    {
      "epoch": 0.6296988577362409,
      "grad_norm": 0.4864673614501953,
      "learning_rate": 8e-05,
      "loss": 1.4912,
      "step": 3032
    },
    {
      "epoch": 0.6299065420560748,
      "grad_norm": 0.5418030023574829,
      "learning_rate": 8e-05,
      "loss": 1.5674,
      "step": 3033
    },
    {
      "epoch": 0.6301142263759086,
      "grad_norm": 0.532538115978241,
      "learning_rate": 8e-05,
      "loss": 1.6535,
      "step": 3034
    },
    {
      "epoch": 0.6303219106957425,
      "grad_norm": 0.5369309186935425,
      "learning_rate": 8e-05,
      "loss": 1.5989,
      "step": 3035
    },
    {
      "epoch": 0.6305295950155764,
      "grad_norm": 0.5634323358535767,
      "learning_rate": 8e-05,
      "loss": 1.5401,
      "step": 3036
    },
    {
      "epoch": 0.6307372793354101,
      "grad_norm": 0.5228213667869568,
      "learning_rate": 8e-05,
      "loss": 1.5792,
      "step": 3037
    },
    {
      "epoch": 0.630944963655244,
      "grad_norm": 0.5049849152565002,
      "learning_rate": 8e-05,
      "loss": 1.4955,
      "step": 3038
    },
    {
      "epoch": 0.6311526479750779,
      "grad_norm": 0.5268635153770447,
      "learning_rate": 8e-05,
      "loss": 1.6085,
      "step": 3039
    },
    {
      "epoch": 0.6313603322949117,
      "grad_norm": 0.5162904858589172,
      "learning_rate": 8e-05,
      "loss": 1.579,
      "step": 3040
    },
    {
      "epoch": 0.6315680166147456,
      "grad_norm": 0.5065432190895081,
      "learning_rate": 8e-05,
      "loss": 1.4774,
      "step": 3041
    },
    {
      "epoch": 0.6317757009345795,
      "grad_norm": 0.4926864802837372,
      "learning_rate": 8e-05,
      "loss": 1.4674,
      "step": 3042
    },
    {
      "epoch": 0.6319833852544133,
      "grad_norm": 0.520024299621582,
      "learning_rate": 8e-05,
      "loss": 1.5642,
      "step": 3043
    },
    {
      "epoch": 0.6321910695742472,
      "grad_norm": 0.511763334274292,
      "learning_rate": 8e-05,
      "loss": 1.5539,
      "step": 3044
    },
    {
      "epoch": 0.632398753894081,
      "grad_norm": 0.5275795459747314,
      "learning_rate": 8e-05,
      "loss": 1.5929,
      "step": 3045
    },
    {
      "epoch": 0.6326064382139148,
      "grad_norm": 0.5092682838439941,
      "learning_rate": 8e-05,
      "loss": 1.5041,
      "step": 3046
    },
    {
      "epoch": 0.6328141225337487,
      "grad_norm": 0.5073072910308838,
      "learning_rate": 8e-05,
      "loss": 1.5718,
      "step": 3047
    },
    {
      "epoch": 0.6330218068535826,
      "grad_norm": 0.5382453203201294,
      "learning_rate": 8e-05,
      "loss": 1.5627,
      "step": 3048
    },
    {
      "epoch": 0.6332294911734164,
      "grad_norm": 0.5363137722015381,
      "learning_rate": 8e-05,
      "loss": 1.5487,
      "step": 3049
    },
    {
      "epoch": 0.6334371754932503,
      "grad_norm": 0.5139375329017639,
      "learning_rate": 8e-05,
      "loss": 1.5292,
      "step": 3050
    },
    {
      "epoch": 0.6336448598130842,
      "grad_norm": 0.5390216112136841,
      "learning_rate": 8e-05,
      "loss": 1.5909,
      "step": 3051
    },
    {
      "epoch": 0.6338525441329179,
      "grad_norm": 0.515113115310669,
      "learning_rate": 8e-05,
      "loss": 1.5489,
      "step": 3052
    },
    {
      "epoch": 0.6340602284527518,
      "grad_norm": 0.6222773194313049,
      "learning_rate": 8e-05,
      "loss": 1.5496,
      "step": 3053
    },
    {
      "epoch": 0.6342679127725857,
      "grad_norm": 0.5217164158821106,
      "learning_rate": 8e-05,
      "loss": 1.6267,
      "step": 3054
    },
    {
      "epoch": 0.6344755970924195,
      "grad_norm": 0.5437771677970886,
      "learning_rate": 8e-05,
      "loss": 1.6263,
      "step": 3055
    },
    {
      "epoch": 0.6346832814122534,
      "grad_norm": 0.5235287547111511,
      "learning_rate": 8e-05,
      "loss": 1.5464,
      "step": 3056
    },
    {
      "epoch": 0.6348909657320873,
      "grad_norm": 0.5067916512489319,
      "learning_rate": 8e-05,
      "loss": 1.5267,
      "step": 3057
    },
    {
      "epoch": 0.6350986500519211,
      "grad_norm": 0.5758751034736633,
      "learning_rate": 8e-05,
      "loss": 1.5189,
      "step": 3058
    },
    {
      "epoch": 0.6353063343717549,
      "grad_norm": 0.5157259106636047,
      "learning_rate": 8e-05,
      "loss": 1.4527,
      "step": 3059
    },
    {
      "epoch": 0.6355140186915887,
      "grad_norm": 0.5036389231681824,
      "learning_rate": 8e-05,
      "loss": 1.5362,
      "step": 3060
    },
    {
      "epoch": 0.6357217030114226,
      "grad_norm": 0.5562075972557068,
      "learning_rate": 8e-05,
      "loss": 1.6016,
      "step": 3061
    },
    {
      "epoch": 0.6359293873312565,
      "grad_norm": 0.537006139755249,
      "learning_rate": 8e-05,
      "loss": 1.5393,
      "step": 3062
    },
    {
      "epoch": 0.6361370716510903,
      "grad_norm": 0.5111969709396362,
      "learning_rate": 8e-05,
      "loss": 1.5731,
      "step": 3063
    },
    {
      "epoch": 0.6363447559709242,
      "grad_norm": 0.6285061240196228,
      "learning_rate": 8e-05,
      "loss": 1.6481,
      "step": 3064
    },
    {
      "epoch": 0.6365524402907581,
      "grad_norm": 0.5460458993911743,
      "learning_rate": 8e-05,
      "loss": 1.4854,
      "step": 3065
    },
    {
      "epoch": 0.636760124610592,
      "grad_norm": 0.5169022679328918,
      "learning_rate": 8e-05,
      "loss": 1.5382,
      "step": 3066
    },
    {
      "epoch": 0.6369678089304257,
      "grad_norm": 0.5269701480865479,
      "learning_rate": 8e-05,
      "loss": 1.5205,
      "step": 3067
    },
    {
      "epoch": 0.6371754932502596,
      "grad_norm": 0.5026315450668335,
      "learning_rate": 8e-05,
      "loss": 1.5508,
      "step": 3068
    },
    {
      "epoch": 0.6373831775700934,
      "grad_norm": 0.5309653282165527,
      "learning_rate": 8e-05,
      "loss": 1.5766,
      "step": 3069
    },
    {
      "epoch": 0.6375908618899273,
      "grad_norm": 0.5116264224052429,
      "learning_rate": 8e-05,
      "loss": 1.5679,
      "step": 3070
    },
    {
      "epoch": 0.6377985462097612,
      "grad_norm": 0.5210251808166504,
      "learning_rate": 8e-05,
      "loss": 1.5174,
      "step": 3071
    },
    {
      "epoch": 0.638006230529595,
      "grad_norm": 0.4994380474090576,
      "learning_rate": 8e-05,
      "loss": 1.5593,
      "step": 3072
    },
    {
      "epoch": 0.6382139148494289,
      "grad_norm": 0.5246351361274719,
      "learning_rate": 8e-05,
      "loss": 1.5237,
      "step": 3073
    },
    {
      "epoch": 0.6384215991692627,
      "grad_norm": 0.5221892595291138,
      "learning_rate": 8e-05,
      "loss": 1.51,
      "step": 3074
    },
    {
      "epoch": 0.6386292834890965,
      "grad_norm": 0.5222132802009583,
      "learning_rate": 8e-05,
      "loss": 1.5578,
      "step": 3075
    },
    {
      "epoch": 0.6388369678089304,
      "grad_norm": 0.503761351108551,
      "learning_rate": 8e-05,
      "loss": 1.5699,
      "step": 3076
    },
    {
      "epoch": 0.6390446521287643,
      "grad_norm": 0.5134929418563843,
      "learning_rate": 8e-05,
      "loss": 1.5239,
      "step": 3077
    },
    {
      "epoch": 0.6392523364485981,
      "grad_norm": 0.5154242515563965,
      "learning_rate": 8e-05,
      "loss": 1.5821,
      "step": 3078
    },
    {
      "epoch": 0.639460020768432,
      "grad_norm": 0.5819329619407654,
      "learning_rate": 8e-05,
      "loss": 1.5649,
      "step": 3079
    },
    {
      "epoch": 0.6396677050882659,
      "grad_norm": 0.5277559161186218,
      "learning_rate": 8e-05,
      "loss": 1.5431,
      "step": 3080
    },
    {
      "epoch": 0.6398753894080997,
      "grad_norm": 0.5070793628692627,
      "learning_rate": 8e-05,
      "loss": 1.5583,
      "step": 3081
    },
    {
      "epoch": 0.6400830737279335,
      "grad_norm": 0.5349634289741516,
      "learning_rate": 8e-05,
      "loss": 1.5111,
      "step": 3082
    },
    {
      "epoch": 0.6402907580477674,
      "grad_norm": 0.5691190958023071,
      "learning_rate": 8e-05,
      "loss": 1.555,
      "step": 3083
    },
    {
      "epoch": 0.6404984423676012,
      "grad_norm": 0.528887927532196,
      "learning_rate": 8e-05,
      "loss": 1.5236,
      "step": 3084
    },
    {
      "epoch": 0.6407061266874351,
      "grad_norm": 0.5441225171089172,
      "learning_rate": 8e-05,
      "loss": 1.536,
      "step": 3085
    },
    {
      "epoch": 0.640913811007269,
      "grad_norm": 0.5341313481330872,
      "learning_rate": 8e-05,
      "loss": 1.6596,
      "step": 3086
    },
    {
      "epoch": 0.6411214953271028,
      "grad_norm": 0.5154920220375061,
      "learning_rate": 8e-05,
      "loss": 1.5356,
      "step": 3087
    },
    {
      "epoch": 0.6413291796469367,
      "grad_norm": 0.5456720590591431,
      "learning_rate": 8e-05,
      "loss": 1.5147,
      "step": 3088
    },
    {
      "epoch": 0.6415368639667705,
      "grad_norm": 0.5055673122406006,
      "learning_rate": 8e-05,
      "loss": 1.4346,
      "step": 3089
    },
    {
      "epoch": 0.6417445482866043,
      "grad_norm": 0.5346052050590515,
      "learning_rate": 8e-05,
      "loss": 1.5669,
      "step": 3090
    },
    {
      "epoch": 0.6419522326064382,
      "grad_norm": 0.5550487637519836,
      "learning_rate": 8e-05,
      "loss": 1.5449,
      "step": 3091
    },
    {
      "epoch": 0.6421599169262721,
      "grad_norm": 0.5220410227775574,
      "learning_rate": 8e-05,
      "loss": 1.546,
      "step": 3092
    },
    {
      "epoch": 0.6423676012461059,
      "grad_norm": 0.5404638648033142,
      "learning_rate": 8e-05,
      "loss": 1.6079,
      "step": 3093
    },
    {
      "epoch": 0.6425752855659398,
      "grad_norm": 0.538062572479248,
      "learning_rate": 8e-05,
      "loss": 1.589,
      "step": 3094
    },
    {
      "epoch": 0.6427829698857737,
      "grad_norm": 0.5388756990432739,
      "learning_rate": 8e-05,
      "loss": 1.6408,
      "step": 3095
    },
    {
      "epoch": 0.6429906542056075,
      "grad_norm": 0.5816310048103333,
      "learning_rate": 8e-05,
      "loss": 1.4533,
      "step": 3096
    },
    {
      "epoch": 0.6431983385254413,
      "grad_norm": 0.5416650176048279,
      "learning_rate": 8e-05,
      "loss": 1.5235,
      "step": 3097
    },
    {
      "epoch": 0.6434060228452751,
      "grad_norm": 0.5160452723503113,
      "learning_rate": 8e-05,
      "loss": 1.48,
      "step": 3098
    },
    {
      "epoch": 0.643613707165109,
      "grad_norm": 0.5227875709533691,
      "learning_rate": 8e-05,
      "loss": 1.5908,
      "step": 3099
    },
    {
      "epoch": 0.6438213914849429,
      "grad_norm": 0.5626046657562256,
      "learning_rate": 8e-05,
      "loss": 1.5309,
      "step": 3100
    },
    {
      "epoch": 0.6440290758047768,
      "grad_norm": 0.5371845960617065,
      "learning_rate": 8e-05,
      "loss": 1.5393,
      "step": 3101
    },
    {
      "epoch": 0.6442367601246106,
      "grad_norm": 0.5423967838287354,
      "learning_rate": 8e-05,
      "loss": 1.5133,
      "step": 3102
    },
    {
      "epoch": 0.6444444444444445,
      "grad_norm": 0.5547438859939575,
      "learning_rate": 8e-05,
      "loss": 1.5604,
      "step": 3103
    },
    {
      "epoch": 0.6446521287642782,
      "grad_norm": 0.5284349322319031,
      "learning_rate": 8e-05,
      "loss": 1.5978,
      "step": 3104
    },
    {
      "epoch": 0.6448598130841121,
      "grad_norm": 0.5673092007637024,
      "learning_rate": 8e-05,
      "loss": 1.5656,
      "step": 3105
    },
    {
      "epoch": 0.645067497403946,
      "grad_norm": 0.5218328237533569,
      "learning_rate": 8e-05,
      "loss": 1.5677,
      "step": 3106
    },
    {
      "epoch": 0.6452751817237798,
      "grad_norm": 0.5550040006637573,
      "learning_rate": 8e-05,
      "loss": 1.5784,
      "step": 3107
    },
    {
      "epoch": 0.6454828660436137,
      "grad_norm": 0.5519328117370605,
      "learning_rate": 8e-05,
      "loss": 1.6386,
      "step": 3108
    },
    {
      "epoch": 0.6456905503634476,
      "grad_norm": 0.5180282592773438,
      "learning_rate": 8e-05,
      "loss": 1.6313,
      "step": 3109
    },
    {
      "epoch": 0.6458982346832814,
      "grad_norm": 0.5330683588981628,
      "learning_rate": 8e-05,
      "loss": 1.5093,
      "step": 3110
    },
    {
      "epoch": 0.6461059190031153,
      "grad_norm": 0.529720664024353,
      "learning_rate": 8e-05,
      "loss": 1.5887,
      "step": 3111
    },
    {
      "epoch": 0.6463136033229491,
      "grad_norm": 0.5654529929161072,
      "learning_rate": 8e-05,
      "loss": 1.5392,
      "step": 3112
    },
    {
      "epoch": 0.6465212876427829,
      "grad_norm": 0.5645838975906372,
      "learning_rate": 8e-05,
      "loss": 1.5394,
      "step": 3113
    },
    {
      "epoch": 0.6467289719626168,
      "grad_norm": 0.5112038850784302,
      "learning_rate": 8e-05,
      "loss": 1.5154,
      "step": 3114
    },
    {
      "epoch": 0.6469366562824507,
      "grad_norm": 0.5384823083877563,
      "learning_rate": 8e-05,
      "loss": 1.6108,
      "step": 3115
    },
    {
      "epoch": 0.6471443406022845,
      "grad_norm": 0.5259283185005188,
      "learning_rate": 8e-05,
      "loss": 1.6125,
      "step": 3116
    },
    {
      "epoch": 0.6473520249221184,
      "grad_norm": 0.5012771487236023,
      "learning_rate": 8e-05,
      "loss": 1.5311,
      "step": 3117
    },
    {
      "epoch": 0.6475597092419523,
      "grad_norm": 0.5418326258659363,
      "learning_rate": 8e-05,
      "loss": 1.5536,
      "step": 3118
    },
    {
      "epoch": 0.647767393561786,
      "grad_norm": 0.5814041495323181,
      "learning_rate": 8e-05,
      "loss": 1.5429,
      "step": 3119
    },
    {
      "epoch": 0.6479750778816199,
      "grad_norm": 0.5095088481903076,
      "learning_rate": 8e-05,
      "loss": 1.5717,
      "step": 3120
    },
    {
      "epoch": 0.6481827622014538,
      "grad_norm": 0.5700481534004211,
      "learning_rate": 8e-05,
      "loss": 1.5612,
      "step": 3121
    },
    {
      "epoch": 0.6483904465212876,
      "grad_norm": 0.5317157506942749,
      "learning_rate": 8e-05,
      "loss": 1.542,
      "step": 3122
    },
    {
      "epoch": 0.6485981308411215,
      "grad_norm": 0.513659656047821,
      "learning_rate": 8e-05,
      "loss": 1.4478,
      "step": 3123
    },
    {
      "epoch": 0.6488058151609554,
      "grad_norm": 0.5258849859237671,
      "learning_rate": 8e-05,
      "loss": 1.6313,
      "step": 3124
    },
    {
      "epoch": 0.6490134994807892,
      "grad_norm": 0.5217114090919495,
      "learning_rate": 8e-05,
      "loss": 1.5514,
      "step": 3125
    },
    {
      "epoch": 0.6492211838006231,
      "grad_norm": 0.5313328504562378,
      "learning_rate": 8e-05,
      "loss": 1.6513,
      "step": 3126
    },
    {
      "epoch": 0.6494288681204569,
      "grad_norm": 0.5158863663673401,
      "learning_rate": 8e-05,
      "loss": 1.5513,
      "step": 3127
    },
    {
      "epoch": 0.6496365524402907,
      "grad_norm": 0.5265626311302185,
      "learning_rate": 8e-05,
      "loss": 1.5185,
      "step": 3128
    },
    {
      "epoch": 0.6498442367601246,
      "grad_norm": 0.5074301958084106,
      "learning_rate": 8e-05,
      "loss": 1.4582,
      "step": 3129
    },
    {
      "epoch": 0.6500519210799585,
      "grad_norm": 0.5295581817626953,
      "learning_rate": 8e-05,
      "loss": 1.5707,
      "step": 3130
    },
    {
      "epoch": 0.6502596053997923,
      "grad_norm": 0.5183844566345215,
      "learning_rate": 8e-05,
      "loss": 1.5866,
      "step": 3131
    },
    {
      "epoch": 0.6504672897196262,
      "grad_norm": 0.5013550519943237,
      "learning_rate": 8e-05,
      "loss": 1.5236,
      "step": 3132
    },
    {
      "epoch": 0.6506749740394601,
      "grad_norm": 0.5204567909240723,
      "learning_rate": 8e-05,
      "loss": 1.5503,
      "step": 3133
    },
    {
      "epoch": 0.6508826583592938,
      "grad_norm": 0.5222823619842529,
      "learning_rate": 8e-05,
      "loss": 1.6236,
      "step": 3134
    },
    {
      "epoch": 0.6510903426791277,
      "grad_norm": 0.5108862519264221,
      "learning_rate": 8e-05,
      "loss": 1.5109,
      "step": 3135
    },
    {
      "epoch": 0.6512980269989616,
      "grad_norm": 0.5182685852050781,
      "learning_rate": 8e-05,
      "loss": 1.5123,
      "step": 3136
    },
    {
      "epoch": 0.6515057113187954,
      "grad_norm": 0.5151105523109436,
      "learning_rate": 8e-05,
      "loss": 1.5209,
      "step": 3137
    },
    {
      "epoch": 0.6517133956386293,
      "grad_norm": 0.5476807355880737,
      "learning_rate": 8e-05,
      "loss": 1.5842,
      "step": 3138
    },
    {
      "epoch": 0.6519210799584632,
      "grad_norm": 0.5228112936019897,
      "learning_rate": 8e-05,
      "loss": 1.5134,
      "step": 3139
    },
    {
      "epoch": 0.652128764278297,
      "grad_norm": 0.5262482166290283,
      "learning_rate": 8e-05,
      "loss": 1.5241,
      "step": 3140
    },
    {
      "epoch": 0.6523364485981309,
      "grad_norm": 0.5675541758537292,
      "learning_rate": 8e-05,
      "loss": 1.5663,
      "step": 3141
    },
    {
      "epoch": 0.6525441329179646,
      "grad_norm": 0.5227878093719482,
      "learning_rate": 8e-05,
      "loss": 1.5537,
      "step": 3142
    },
    {
      "epoch": 0.6527518172377985,
      "grad_norm": 0.5449708700180054,
      "learning_rate": 8e-05,
      "loss": 1.5471,
      "step": 3143
    },
    {
      "epoch": 0.6529595015576324,
      "grad_norm": 0.5156728625297546,
      "learning_rate": 8e-05,
      "loss": 1.5383,
      "step": 3144
    },
    {
      "epoch": 0.6531671858774662,
      "grad_norm": 0.5271214842796326,
      "learning_rate": 8e-05,
      "loss": 1.6015,
      "step": 3145
    },
    {
      "epoch": 0.6533748701973001,
      "grad_norm": 0.5095561742782593,
      "learning_rate": 8e-05,
      "loss": 1.5097,
      "step": 3146
    },
    {
      "epoch": 0.653582554517134,
      "grad_norm": 0.5529608726501465,
      "learning_rate": 8e-05,
      "loss": 1.5382,
      "step": 3147
    },
    {
      "epoch": 0.6537902388369679,
      "grad_norm": 0.5545696020126343,
      "learning_rate": 8e-05,
      "loss": 1.5303,
      "step": 3148
    },
    {
      "epoch": 0.6539979231568016,
      "grad_norm": 0.6470621824264526,
      "learning_rate": 8e-05,
      "loss": 1.5581,
      "step": 3149
    },
    {
      "epoch": 0.6542056074766355,
      "grad_norm": 0.521135151386261,
      "learning_rate": 8e-05,
      "loss": 1.5239,
      "step": 3150
    },
    {
      "epoch": 0.6544132917964693,
      "grad_norm": 0.5598577857017517,
      "learning_rate": 8e-05,
      "loss": 1.6454,
      "step": 3151
    },
    {
      "epoch": 0.6546209761163032,
      "grad_norm": 0.5060534477233887,
      "learning_rate": 8e-05,
      "loss": 1.5953,
      "step": 3152
    },
    {
      "epoch": 0.6548286604361371,
      "grad_norm": 0.5501383543014526,
      "learning_rate": 8e-05,
      "loss": 1.5727,
      "step": 3153
    },
    {
      "epoch": 0.655036344755971,
      "grad_norm": 0.5838757753372192,
      "learning_rate": 8e-05,
      "loss": 1.5794,
      "step": 3154
    },
    {
      "epoch": 0.6552440290758048,
      "grad_norm": 0.5301080942153931,
      "learning_rate": 8e-05,
      "loss": 1.5297,
      "step": 3155
    },
    {
      "epoch": 0.6554517133956387,
      "grad_norm": 0.5095718502998352,
      "learning_rate": 8e-05,
      "loss": 1.4335,
      "step": 3156
    },
    {
      "epoch": 0.6556593977154724,
      "grad_norm": 0.5119460225105286,
      "learning_rate": 8e-05,
      "loss": 1.5001,
      "step": 3157
    },
    {
      "epoch": 0.6558670820353063,
      "grad_norm": 0.5048596858978271,
      "learning_rate": 8e-05,
      "loss": 1.5568,
      "step": 3158
    },
    {
      "epoch": 0.6560747663551402,
      "grad_norm": 0.5434015989303589,
      "learning_rate": 8e-05,
      "loss": 1.6469,
      "step": 3159
    },
    {
      "epoch": 0.656282450674974,
      "grad_norm": 0.5216411352157593,
      "learning_rate": 8e-05,
      "loss": 1.5764,
      "step": 3160
    },
    {
      "epoch": 0.6564901349948079,
      "grad_norm": 0.5377340316772461,
      "learning_rate": 8e-05,
      "loss": 1.5973,
      "step": 3161
    },
    {
      "epoch": 0.6566978193146418,
      "grad_norm": 0.5488721132278442,
      "learning_rate": 8e-05,
      "loss": 1.6415,
      "step": 3162
    },
    {
      "epoch": 0.6569055036344756,
      "grad_norm": 0.5382291078567505,
      "learning_rate": 8e-05,
      "loss": 1.607,
      "step": 3163
    },
    {
      "epoch": 0.6571131879543094,
      "grad_norm": 0.5220791697502136,
      "learning_rate": 8e-05,
      "loss": 1.5486,
      "step": 3164
    },
    {
      "epoch": 0.6573208722741433,
      "grad_norm": 0.5148298740386963,
      "learning_rate": 8e-05,
      "loss": 1.6163,
      "step": 3165
    },
    {
      "epoch": 0.6575285565939771,
      "grad_norm": 0.5793923139572144,
      "learning_rate": 8e-05,
      "loss": 1.5041,
      "step": 3166
    },
    {
      "epoch": 0.657736240913811,
      "grad_norm": 0.5293853878974915,
      "learning_rate": 8e-05,
      "loss": 1.5577,
      "step": 3167
    },
    {
      "epoch": 0.6579439252336449,
      "grad_norm": 0.5176268219947815,
      "learning_rate": 8e-05,
      "loss": 1.5799,
      "step": 3168
    },
    {
      "epoch": 0.6581516095534787,
      "grad_norm": 0.5291829109191895,
      "learning_rate": 8e-05,
      "loss": 1.5476,
      "step": 3169
    },
    {
      "epoch": 0.6583592938733126,
      "grad_norm": 0.5412970185279846,
      "learning_rate": 8e-05,
      "loss": 1.5753,
      "step": 3170
    },
    {
      "epoch": 0.6585669781931465,
      "grad_norm": 0.508479118347168,
      "learning_rate": 8e-05,
      "loss": 1.4658,
      "step": 3171
    },
    {
      "epoch": 0.6587746625129802,
      "grad_norm": 0.5168586373329163,
      "learning_rate": 8e-05,
      "loss": 1.5315,
      "step": 3172
    },
    {
      "epoch": 0.6589823468328141,
      "grad_norm": 0.5382446646690369,
      "learning_rate": 8e-05,
      "loss": 1.6064,
      "step": 3173
    },
    {
      "epoch": 0.659190031152648,
      "grad_norm": 0.511417806148529,
      "learning_rate": 8e-05,
      "loss": 1.5242,
      "step": 3174
    },
    {
      "epoch": 0.6593977154724818,
      "grad_norm": 0.5325655341148376,
      "learning_rate": 8e-05,
      "loss": 1.5464,
      "step": 3175
    },
    {
      "epoch": 0.6596053997923157,
      "grad_norm": 0.5209563374519348,
      "learning_rate": 8e-05,
      "loss": 1.5294,
      "step": 3176
    },
    {
      "epoch": 0.6598130841121496,
      "grad_norm": 0.5548104047775269,
      "learning_rate": 8e-05,
      "loss": 1.5274,
      "step": 3177
    },
    {
      "epoch": 0.6600207684319834,
      "grad_norm": 0.5407391786575317,
      "learning_rate": 8e-05,
      "loss": 1.5503,
      "step": 3178
    },
    {
      "epoch": 0.6602284527518172,
      "grad_norm": 0.5083649158477783,
      "learning_rate": 8e-05,
      "loss": 1.5226,
      "step": 3179
    },
    {
      "epoch": 0.660436137071651,
      "grad_norm": 0.5150434970855713,
      "learning_rate": 8e-05,
      "loss": 1.5005,
      "step": 3180
    },
    {
      "epoch": 0.6606438213914849,
      "grad_norm": 0.5460854768753052,
      "learning_rate": 8e-05,
      "loss": 1.5559,
      "step": 3181
    },
    {
      "epoch": 0.6608515057113188,
      "grad_norm": 0.5881128311157227,
      "learning_rate": 8e-05,
      "loss": 1.6327,
      "step": 3182
    },
    {
      "epoch": 0.6610591900311527,
      "grad_norm": 0.539505124092102,
      "learning_rate": 8e-05,
      "loss": 1.579,
      "step": 3183
    },
    {
      "epoch": 0.6612668743509865,
      "grad_norm": 0.5233002305030823,
      "learning_rate": 8e-05,
      "loss": 1.5846,
      "step": 3184
    },
    {
      "epoch": 0.6614745586708204,
      "grad_norm": 0.5348186492919922,
      "learning_rate": 8e-05,
      "loss": 1.5692,
      "step": 3185
    },
    {
      "epoch": 0.6616822429906543,
      "grad_norm": 0.5318605303764343,
      "learning_rate": 8e-05,
      "loss": 1.5905,
      "step": 3186
    },
    {
      "epoch": 0.661889927310488,
      "grad_norm": 0.5109202861785889,
      "learning_rate": 8e-05,
      "loss": 1.5779,
      "step": 3187
    },
    {
      "epoch": 0.6620976116303219,
      "grad_norm": 0.5168127417564392,
      "learning_rate": 8e-05,
      "loss": 1.5567,
      "step": 3188
    },
    {
      "epoch": 0.6623052959501557,
      "grad_norm": 0.5081555843353271,
      "learning_rate": 8e-05,
      "loss": 1.5226,
      "step": 3189
    },
    {
      "epoch": 0.6625129802699896,
      "grad_norm": 0.5576549768447876,
      "learning_rate": 8e-05,
      "loss": 1.5654,
      "step": 3190
    },
    {
      "epoch": 0.6627206645898235,
      "grad_norm": 0.5441120862960815,
      "learning_rate": 8e-05,
      "loss": 1.5623,
      "step": 3191
    },
    {
      "epoch": 0.6629283489096574,
      "grad_norm": 0.5081104636192322,
      "learning_rate": 8e-05,
      "loss": 1.5209,
      "step": 3192
    },
    {
      "epoch": 0.6631360332294912,
      "grad_norm": 0.5548802614212036,
      "learning_rate": 8e-05,
      "loss": 1.5604,
      "step": 3193
    },
    {
      "epoch": 0.663343717549325,
      "grad_norm": 0.5266252756118774,
      "learning_rate": 8e-05,
      "loss": 1.5169,
      "step": 3194
    },
    {
      "epoch": 0.6635514018691588,
      "grad_norm": 0.5315518975257874,
      "learning_rate": 8e-05,
      "loss": 1.488,
      "step": 3195
    },
    {
      "epoch": 0.6637590861889927,
      "grad_norm": 0.5006552338600159,
      "learning_rate": 8e-05,
      "loss": 1.5222,
      "step": 3196
    },
    {
      "epoch": 0.6639667705088266,
      "grad_norm": 0.5342662334442139,
      "learning_rate": 8e-05,
      "loss": 1.6111,
      "step": 3197
    },
    {
      "epoch": 0.6641744548286604,
      "grad_norm": 0.5125207901000977,
      "learning_rate": 8e-05,
      "loss": 1.4599,
      "step": 3198
    },
    {
      "epoch": 0.6643821391484943,
      "grad_norm": 0.5058197975158691,
      "learning_rate": 8e-05,
      "loss": 1.4905,
      "step": 3199
    },
    {
      "epoch": 0.6645898234683282,
      "grad_norm": 0.5297023057937622,
      "learning_rate": 8e-05,
      "loss": 1.5195,
      "step": 3200
    },
    {
      "epoch": 0.664797507788162,
      "grad_norm": 0.5354586243629456,
      "learning_rate": 8e-05,
      "loss": 1.5768,
      "step": 3201
    },
    {
      "epoch": 0.6650051921079958,
      "grad_norm": 0.5730552077293396,
      "learning_rate": 8e-05,
      "loss": 1.5301,
      "step": 3202
    },
    {
      "epoch": 0.6652128764278297,
      "grad_norm": 0.5455592274665833,
      "learning_rate": 8e-05,
      "loss": 1.5187,
      "step": 3203
    },
    {
      "epoch": 0.6654205607476635,
      "grad_norm": 0.5239001512527466,
      "learning_rate": 8e-05,
      "loss": 1.539,
      "step": 3204
    },
    {
      "epoch": 0.6656282450674974,
      "grad_norm": 0.5362431406974792,
      "learning_rate": 8e-05,
      "loss": 1.6012,
      "step": 3205
    },
    {
      "epoch": 0.6658359293873313,
      "grad_norm": 0.5227289199829102,
      "learning_rate": 8e-05,
      "loss": 1.5184,
      "step": 3206
    },
    {
      "epoch": 0.6660436137071651,
      "grad_norm": 0.616928219795227,
      "learning_rate": 8e-05,
      "loss": 1.5842,
      "step": 3207
    },
    {
      "epoch": 0.666251298026999,
      "grad_norm": 0.5601013898849487,
      "learning_rate": 8e-05,
      "loss": 1.5736,
      "step": 3208
    },
    {
      "epoch": 0.6664589823468328,
      "grad_norm": 0.5242985486984253,
      "learning_rate": 8e-05,
      "loss": 1.5856,
      "step": 3209
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.5099378228187561,
      "learning_rate": 8e-05,
      "loss": 1.5528,
      "step": 3210
    },
    {
      "epoch": 0.6668743509865005,
      "grad_norm": 0.5576867461204529,
      "learning_rate": 8e-05,
      "loss": 1.532,
      "step": 3211
    },
    {
      "epoch": 0.6670820353063344,
      "grad_norm": 0.5440665483474731,
      "learning_rate": 8e-05,
      "loss": 1.5894,
      "step": 3212
    },
    {
      "epoch": 0.6672897196261682,
      "grad_norm": 0.5238298773765564,
      "learning_rate": 8e-05,
      "loss": 1.5715,
      "step": 3213
    },
    {
      "epoch": 0.6674974039460021,
      "grad_norm": 0.5310534834861755,
      "learning_rate": 8e-05,
      "loss": 1.5487,
      "step": 3214
    },
    {
      "epoch": 0.667705088265836,
      "grad_norm": 0.5108256936073303,
      "learning_rate": 8e-05,
      "loss": 1.5673,
      "step": 3215
    },
    {
      "epoch": 0.6679127725856698,
      "grad_norm": 0.6176890134811401,
      "learning_rate": 8e-05,
      "loss": 1.5717,
      "step": 3216
    },
    {
      "epoch": 0.6681204569055036,
      "grad_norm": 0.5360254645347595,
      "learning_rate": 8e-05,
      "loss": 1.5478,
      "step": 3217
    },
    {
      "epoch": 0.6683281412253375,
      "grad_norm": 0.5275375247001648,
      "learning_rate": 8e-05,
      "loss": 1.542,
      "step": 3218
    },
    {
      "epoch": 0.6685358255451713,
      "grad_norm": 0.518865168094635,
      "learning_rate": 8e-05,
      "loss": 1.5489,
      "step": 3219
    },
    {
      "epoch": 0.6687435098650052,
      "grad_norm": 0.5309369564056396,
      "learning_rate": 8e-05,
      "loss": 1.5362,
      "step": 3220
    },
    {
      "epoch": 0.6689511941848391,
      "grad_norm": 0.5379794836044312,
      "learning_rate": 8e-05,
      "loss": 1.5644,
      "step": 3221
    },
    {
      "epoch": 0.6691588785046729,
      "grad_norm": 0.5286773443222046,
      "learning_rate": 8e-05,
      "loss": 1.4702,
      "step": 3222
    },
    {
      "epoch": 0.6693665628245068,
      "grad_norm": 0.5246649980545044,
      "learning_rate": 8e-05,
      "loss": 1.5295,
      "step": 3223
    },
    {
      "epoch": 0.6695742471443406,
      "grad_norm": 0.5348759293556213,
      "learning_rate": 8e-05,
      "loss": 1.5171,
      "step": 3224
    },
    {
      "epoch": 0.6697819314641744,
      "grad_norm": 0.5250929594039917,
      "learning_rate": 8e-05,
      "loss": 1.4747,
      "step": 3225
    },
    {
      "epoch": 0.6699896157840083,
      "grad_norm": 0.5414816737174988,
      "learning_rate": 8e-05,
      "loss": 1.6321,
      "step": 3226
    },
    {
      "epoch": 0.6701973001038422,
      "grad_norm": 0.5267562866210938,
      "learning_rate": 8e-05,
      "loss": 1.5539,
      "step": 3227
    },
    {
      "epoch": 0.670404984423676,
      "grad_norm": 0.5297579765319824,
      "learning_rate": 8e-05,
      "loss": 1.5155,
      "step": 3228
    },
    {
      "epoch": 0.6706126687435099,
      "grad_norm": 0.5101528763771057,
      "learning_rate": 8e-05,
      "loss": 1.5102,
      "step": 3229
    },
    {
      "epoch": 0.6708203530633438,
      "grad_norm": 0.5926492810249329,
      "learning_rate": 8e-05,
      "loss": 1.5326,
      "step": 3230
    },
    {
      "epoch": 0.6710280373831776,
      "grad_norm": 0.5332081913948059,
      "learning_rate": 8e-05,
      "loss": 1.5585,
      "step": 3231
    },
    {
      "epoch": 0.6712357217030114,
      "grad_norm": 0.5265396237373352,
      "learning_rate": 8e-05,
      "loss": 1.5597,
      "step": 3232
    },
    {
      "epoch": 0.6714434060228452,
      "grad_norm": 0.6264560222625732,
      "learning_rate": 8e-05,
      "loss": 1.5334,
      "step": 3233
    },
    {
      "epoch": 0.6716510903426791,
      "grad_norm": 0.5279672145843506,
      "learning_rate": 8e-05,
      "loss": 1.5398,
      "step": 3234
    },
    {
      "epoch": 0.671858774662513,
      "grad_norm": 0.5178937911987305,
      "learning_rate": 8e-05,
      "loss": 1.6034,
      "step": 3235
    },
    {
      "epoch": 0.6720664589823468,
      "grad_norm": 0.5612587928771973,
      "learning_rate": 8e-05,
      "loss": 1.5386,
      "step": 3236
    },
    {
      "epoch": 0.6722741433021807,
      "grad_norm": 0.5288073420524597,
      "learning_rate": 8e-05,
      "loss": 1.5814,
      "step": 3237
    },
    {
      "epoch": 0.6724818276220146,
      "grad_norm": 0.5133301019668579,
      "learning_rate": 8e-05,
      "loss": 1.4975,
      "step": 3238
    },
    {
      "epoch": 0.6726895119418483,
      "grad_norm": 0.5571306347846985,
      "learning_rate": 8e-05,
      "loss": 1.5231,
      "step": 3239
    },
    {
      "epoch": 0.6728971962616822,
      "grad_norm": 0.6173458099365234,
      "learning_rate": 8e-05,
      "loss": 1.5723,
      "step": 3240
    },
    {
      "epoch": 0.6731048805815161,
      "grad_norm": 0.5417127013206482,
      "learning_rate": 8e-05,
      "loss": 1.5721,
      "step": 3241
    },
    {
      "epoch": 0.6733125649013499,
      "grad_norm": 0.5144616365432739,
      "learning_rate": 8e-05,
      "loss": 1.4901,
      "step": 3242
    },
    {
      "epoch": 0.6735202492211838,
      "grad_norm": 0.5067961812019348,
      "learning_rate": 8e-05,
      "loss": 1.58,
      "step": 3243
    },
    {
      "epoch": 0.6737279335410177,
      "grad_norm": 0.5312129259109497,
      "learning_rate": 8e-05,
      "loss": 1.5204,
      "step": 3244
    },
    {
      "epoch": 0.6739356178608515,
      "grad_norm": 0.5913172364234924,
      "learning_rate": 8e-05,
      "loss": 1.4948,
      "step": 3245
    },
    {
      "epoch": 0.6741433021806854,
      "grad_norm": 0.5416056513786316,
      "learning_rate": 8e-05,
      "loss": 1.5915,
      "step": 3246
    },
    {
      "epoch": 0.6743509865005192,
      "grad_norm": 0.5223923921585083,
      "learning_rate": 8e-05,
      "loss": 1.5105,
      "step": 3247
    },
    {
      "epoch": 0.674558670820353,
      "grad_norm": 0.5623733401298523,
      "learning_rate": 8e-05,
      "loss": 1.4877,
      "step": 3248
    },
    {
      "epoch": 0.6747663551401869,
      "grad_norm": 0.6014411449432373,
      "learning_rate": 8e-05,
      "loss": 1.5893,
      "step": 3249
    },
    {
      "epoch": 0.6749740394600208,
      "grad_norm": 0.5220860242843628,
      "learning_rate": 8e-05,
      "loss": 1.5067,
      "step": 3250
    },
    {
      "epoch": 0.6751817237798546,
      "grad_norm": 0.5102582573890686,
      "learning_rate": 8e-05,
      "loss": 1.5064,
      "step": 3251
    },
    {
      "epoch": 0.6753894080996885,
      "grad_norm": 0.5487357974052429,
      "learning_rate": 8e-05,
      "loss": 1.5261,
      "step": 3252
    },
    {
      "epoch": 0.6755970924195224,
      "grad_norm": 0.5346094369888306,
      "learning_rate": 8e-05,
      "loss": 1.615,
      "step": 3253
    },
    {
      "epoch": 0.6758047767393561,
      "grad_norm": 0.5015448331832886,
      "learning_rate": 8e-05,
      "loss": 1.494,
      "step": 3254
    },
    {
      "epoch": 0.67601246105919,
      "grad_norm": 0.5249627232551575,
      "learning_rate": 8e-05,
      "loss": 1.5223,
      "step": 3255
    },
    {
      "epoch": 0.6762201453790239,
      "grad_norm": 0.5087130665779114,
      "learning_rate": 8e-05,
      "loss": 1.5167,
      "step": 3256
    },
    {
      "epoch": 0.6764278296988577,
      "grad_norm": 0.5213695764541626,
      "learning_rate": 8e-05,
      "loss": 1.5587,
      "step": 3257
    },
    {
      "epoch": 0.6766355140186916,
      "grad_norm": 0.5180026888847351,
      "learning_rate": 8e-05,
      "loss": 1.5414,
      "step": 3258
    },
    {
      "epoch": 0.6768431983385255,
      "grad_norm": 0.5281468033790588,
      "learning_rate": 8e-05,
      "loss": 1.5922,
      "step": 3259
    },
    {
      "epoch": 0.6770508826583593,
      "grad_norm": 0.5434516668319702,
      "learning_rate": 8e-05,
      "loss": 1.6039,
      "step": 3260
    },
    {
      "epoch": 0.6772585669781932,
      "grad_norm": 0.5349191427230835,
      "learning_rate": 8e-05,
      "loss": 1.4738,
      "step": 3261
    },
    {
      "epoch": 0.677466251298027,
      "grad_norm": 0.5366538166999817,
      "learning_rate": 8e-05,
      "loss": 1.4773,
      "step": 3262
    },
    {
      "epoch": 0.6776739356178608,
      "grad_norm": 0.5260535478591919,
      "learning_rate": 8e-05,
      "loss": 1.5494,
      "step": 3263
    },
    {
      "epoch": 0.6778816199376947,
      "grad_norm": 0.5320129990577698,
      "learning_rate": 8e-05,
      "loss": 1.5731,
      "step": 3264
    },
    {
      "epoch": 0.6780893042575286,
      "grad_norm": 0.5303055644035339,
      "learning_rate": 8e-05,
      "loss": 1.6089,
      "step": 3265
    },
    {
      "epoch": 0.6782969885773624,
      "grad_norm": 0.5409517884254456,
      "learning_rate": 8e-05,
      "loss": 1.5991,
      "step": 3266
    },
    {
      "epoch": 0.6785046728971963,
      "grad_norm": 0.5185012817382812,
      "learning_rate": 8e-05,
      "loss": 1.536,
      "step": 3267
    },
    {
      "epoch": 0.6787123572170302,
      "grad_norm": 0.5497683882713318,
      "learning_rate": 8e-05,
      "loss": 1.5379,
      "step": 3268
    },
    {
      "epoch": 0.6789200415368639,
      "grad_norm": 0.5022315979003906,
      "learning_rate": 8e-05,
      "loss": 1.456,
      "step": 3269
    },
    {
      "epoch": 0.6791277258566978,
      "grad_norm": 0.5274485945701599,
      "learning_rate": 8e-05,
      "loss": 1.5626,
      "step": 3270
    },
    {
      "epoch": 0.6793354101765317,
      "grad_norm": 0.532622754573822,
      "learning_rate": 8e-05,
      "loss": 1.5313,
      "step": 3271
    },
    {
      "epoch": 0.6795430944963655,
      "grad_norm": 0.5456002354621887,
      "learning_rate": 8e-05,
      "loss": 1.5341,
      "step": 3272
    },
    {
      "epoch": 0.6797507788161994,
      "grad_norm": 0.5423003435134888,
      "learning_rate": 8e-05,
      "loss": 1.5688,
      "step": 3273
    },
    {
      "epoch": 0.6799584631360333,
      "grad_norm": 0.5303918123245239,
      "learning_rate": 8e-05,
      "loss": 1.5804,
      "step": 3274
    },
    {
      "epoch": 0.6801661474558671,
      "grad_norm": 0.5585711002349854,
      "learning_rate": 8e-05,
      "loss": 1.5244,
      "step": 3275
    },
    {
      "epoch": 0.680373831775701,
      "grad_norm": 0.5514115691184998,
      "learning_rate": 8e-05,
      "loss": 1.6032,
      "step": 3276
    },
    {
      "epoch": 0.6805815160955347,
      "grad_norm": 0.4991404414176941,
      "learning_rate": 8e-05,
      "loss": 1.5286,
      "step": 3277
    },
    {
      "epoch": 0.6807892004153686,
      "grad_norm": 0.5174127221107483,
      "learning_rate": 8e-05,
      "loss": 1.5762,
      "step": 3278
    },
    {
      "epoch": 0.6809968847352025,
      "grad_norm": 0.5118331909179688,
      "learning_rate": 8e-05,
      "loss": 1.5116,
      "step": 3279
    },
    {
      "epoch": 0.6812045690550363,
      "grad_norm": 0.5663711428642273,
      "learning_rate": 8e-05,
      "loss": 1.5568,
      "step": 3280
    },
    {
      "epoch": 0.6814122533748702,
      "grad_norm": 0.5848709344863892,
      "learning_rate": 8e-05,
      "loss": 1.5824,
      "step": 3281
    },
    {
      "epoch": 0.6816199376947041,
      "grad_norm": 0.5493605732917786,
      "learning_rate": 8e-05,
      "loss": 1.5113,
      "step": 3282
    },
    {
      "epoch": 0.681827622014538,
      "grad_norm": 0.549248218536377,
      "learning_rate": 8e-05,
      "loss": 1.6642,
      "step": 3283
    },
    {
      "epoch": 0.6820353063343717,
      "grad_norm": 0.5828871726989746,
      "learning_rate": 8e-05,
      "loss": 1.5632,
      "step": 3284
    },
    {
      "epoch": 0.6822429906542056,
      "grad_norm": 0.5965055227279663,
      "learning_rate": 8e-05,
      "loss": 1.4636,
      "step": 3285
    },
    {
      "epoch": 0.6824506749740394,
      "grad_norm": 0.5404725670814514,
      "learning_rate": 8e-05,
      "loss": 1.6031,
      "step": 3286
    },
    {
      "epoch": 0.6826583592938733,
      "grad_norm": 0.5237371921539307,
      "learning_rate": 8e-05,
      "loss": 1.4944,
      "step": 3287
    },
    {
      "epoch": 0.6828660436137072,
      "grad_norm": 0.5903368592262268,
      "learning_rate": 8e-05,
      "loss": 1.5349,
      "step": 3288
    },
    {
      "epoch": 0.683073727933541,
      "grad_norm": 0.59480881690979,
      "learning_rate": 8e-05,
      "loss": 1.5609,
      "step": 3289
    },
    {
      "epoch": 0.6832814122533749,
      "grad_norm": 0.5141411423683167,
      "learning_rate": 8e-05,
      "loss": 1.5243,
      "step": 3290
    },
    {
      "epoch": 0.6834890965732088,
      "grad_norm": 0.5714674592018127,
      "learning_rate": 8e-05,
      "loss": 1.5703,
      "step": 3291
    },
    {
      "epoch": 0.6836967808930425,
      "grad_norm": 0.5407187342643738,
      "learning_rate": 8e-05,
      "loss": 1.5659,
      "step": 3292
    },
    {
      "epoch": 0.6839044652128764,
      "grad_norm": 0.544731080532074,
      "learning_rate": 8e-05,
      "loss": 1.6949,
      "step": 3293
    },
    {
      "epoch": 0.6841121495327103,
      "grad_norm": 0.6666676998138428,
      "learning_rate": 8e-05,
      "loss": 1.546,
      "step": 3294
    },
    {
      "epoch": 0.6843198338525441,
      "grad_norm": 0.5029351711273193,
      "learning_rate": 8e-05,
      "loss": 1.5557,
      "step": 3295
    },
    {
      "epoch": 0.684527518172378,
      "grad_norm": 0.5221892595291138,
      "learning_rate": 8e-05,
      "loss": 1.5276,
      "step": 3296
    },
    {
      "epoch": 0.6847352024922119,
      "grad_norm": 0.5402727127075195,
      "learning_rate": 8e-05,
      "loss": 1.5479,
      "step": 3297
    },
    {
      "epoch": 0.6849428868120457,
      "grad_norm": 0.5681512951850891,
      "learning_rate": 8e-05,
      "loss": 1.546,
      "step": 3298
    },
    {
      "epoch": 0.6851505711318795,
      "grad_norm": 0.5732626914978027,
      "learning_rate": 8e-05,
      "loss": 1.5092,
      "step": 3299
    },
    {
      "epoch": 0.6853582554517134,
      "grad_norm": 0.5508812069892883,
      "learning_rate": 8e-05,
      "loss": 1.5732,
      "step": 3300
    },
    {
      "epoch": 0.6855659397715472,
      "grad_norm": 0.5706160068511963,
      "learning_rate": 8e-05,
      "loss": 1.5312,
      "step": 3301
    },
    {
      "epoch": 0.6857736240913811,
      "grad_norm": 0.6041212677955627,
      "learning_rate": 8e-05,
      "loss": 1.5596,
      "step": 3302
    },
    {
      "epoch": 0.685981308411215,
      "grad_norm": 0.6013287305831909,
      "learning_rate": 8e-05,
      "loss": 1.546,
      "step": 3303
    },
    {
      "epoch": 0.6861889927310488,
      "grad_norm": 0.5200188159942627,
      "learning_rate": 8e-05,
      "loss": 1.445,
      "step": 3304
    },
    {
      "epoch": 0.6863966770508827,
      "grad_norm": 0.5601400136947632,
      "learning_rate": 8e-05,
      "loss": 1.5563,
      "step": 3305
    },
    {
      "epoch": 0.6866043613707166,
      "grad_norm": 0.5410693287849426,
      "learning_rate": 8e-05,
      "loss": 1.5221,
      "step": 3306
    },
    {
      "epoch": 0.6868120456905503,
      "grad_norm": 0.5613017678260803,
      "learning_rate": 8e-05,
      "loss": 1.5079,
      "step": 3307
    },
    {
      "epoch": 0.6870197300103842,
      "grad_norm": 0.5355778932571411,
      "learning_rate": 8e-05,
      "loss": 1.528,
      "step": 3308
    },
    {
      "epoch": 0.6872274143302181,
      "grad_norm": 0.5654533505439758,
      "learning_rate": 8e-05,
      "loss": 1.5704,
      "step": 3309
    },
    {
      "epoch": 0.6874350986500519,
      "grad_norm": 0.5657114386558533,
      "learning_rate": 8e-05,
      "loss": 1.5184,
      "step": 3310
    },
    {
      "epoch": 0.6876427829698858,
      "grad_norm": 0.4926314949989319,
      "learning_rate": 8e-05,
      "loss": 1.4594,
      "step": 3311
    },
    {
      "epoch": 0.6878504672897197,
      "grad_norm": 0.5354579091072083,
      "learning_rate": 8e-05,
      "loss": 1.5626,
      "step": 3312
    },
    {
      "epoch": 0.6880581516095535,
      "grad_norm": 0.5261551737785339,
      "learning_rate": 8e-05,
      "loss": 1.5768,
      "step": 3313
    },
    {
      "epoch": 0.6882658359293873,
      "grad_norm": 0.5045207738876343,
      "learning_rate": 8e-05,
      "loss": 1.5481,
      "step": 3314
    },
    {
      "epoch": 0.6884735202492211,
      "grad_norm": 0.5383133888244629,
      "learning_rate": 8e-05,
      "loss": 1.551,
      "step": 3315
    },
    {
      "epoch": 0.688681204569055,
      "grad_norm": 0.5161110162734985,
      "learning_rate": 8e-05,
      "loss": 1.5339,
      "step": 3316
    },
    {
      "epoch": 0.6888888888888889,
      "grad_norm": 0.5171722769737244,
      "learning_rate": 8e-05,
      "loss": 1.4895,
      "step": 3317
    },
    {
      "epoch": 0.6890965732087228,
      "grad_norm": 0.5459824204444885,
      "learning_rate": 8e-05,
      "loss": 1.6121,
      "step": 3318
    },
    {
      "epoch": 0.6893042575285566,
      "grad_norm": 0.5495988726615906,
      "learning_rate": 8e-05,
      "loss": 1.5394,
      "step": 3319
    },
    {
      "epoch": 0.6895119418483905,
      "grad_norm": 0.5487689971923828,
      "learning_rate": 8e-05,
      "loss": 1.6297,
      "step": 3320
    },
    {
      "epoch": 0.6897196261682244,
      "grad_norm": 0.5556051135063171,
      "learning_rate": 8e-05,
      "loss": 1.5393,
      "step": 3321
    },
    {
      "epoch": 0.6899273104880581,
      "grad_norm": 0.5246308445930481,
      "learning_rate": 8e-05,
      "loss": 1.5349,
      "step": 3322
    },
    {
      "epoch": 0.690134994807892,
      "grad_norm": 0.5490745306015015,
      "learning_rate": 8e-05,
      "loss": 1.5846,
      "step": 3323
    },
    {
      "epoch": 0.6903426791277258,
      "grad_norm": 0.5153035521507263,
      "learning_rate": 8e-05,
      "loss": 1.6171,
      "step": 3324
    },
    {
      "epoch": 0.6905503634475597,
      "grad_norm": 0.5387035012245178,
      "learning_rate": 8e-05,
      "loss": 1.6408,
      "step": 3325
    },
    {
      "epoch": 0.6907580477673936,
      "grad_norm": 0.5278893709182739,
      "learning_rate": 8e-05,
      "loss": 1.5207,
      "step": 3326
    },
    {
      "epoch": 0.6909657320872274,
      "grad_norm": 0.5582882165908813,
      "learning_rate": 8e-05,
      "loss": 1.5923,
      "step": 3327
    },
    {
      "epoch": 0.6911734164070613,
      "grad_norm": 0.5358442068099976,
      "learning_rate": 8e-05,
      "loss": 1.6358,
      "step": 3328
    },
    {
      "epoch": 0.6913811007268951,
      "grad_norm": 0.5252634286880493,
      "learning_rate": 8e-05,
      "loss": 1.4895,
      "step": 3329
    },
    {
      "epoch": 0.6915887850467289,
      "grad_norm": 0.5190280675888062,
      "learning_rate": 8e-05,
      "loss": 1.5396,
      "step": 3330
    },
    {
      "epoch": 0.6917964693665628,
      "grad_norm": 0.5335167050361633,
      "learning_rate": 8e-05,
      "loss": 1.5118,
      "step": 3331
    },
    {
      "epoch": 0.6920041536863967,
      "grad_norm": 0.5237466096878052,
      "learning_rate": 8e-05,
      "loss": 1.5362,
      "step": 3332
    },
    {
      "epoch": 0.6922118380062305,
      "grad_norm": 0.5078586935997009,
      "learning_rate": 8e-05,
      "loss": 1.4633,
      "step": 3333
    },
    {
      "epoch": 0.6924195223260644,
      "grad_norm": 0.5379639267921448,
      "learning_rate": 8e-05,
      "loss": 1.5639,
      "step": 3334
    },
    {
      "epoch": 0.6926272066458983,
      "grad_norm": 0.5186348557472229,
      "learning_rate": 8e-05,
      "loss": 1.6112,
      "step": 3335
    },
    {
      "epoch": 0.6928348909657321,
      "grad_norm": 0.5529906749725342,
      "learning_rate": 8e-05,
      "loss": 1.5547,
      "step": 3336
    },
    {
      "epoch": 0.6930425752855659,
      "grad_norm": 0.5259935259819031,
      "learning_rate": 8e-05,
      "loss": 1.5081,
      "step": 3337
    },
    {
      "epoch": 0.6932502596053998,
      "grad_norm": 0.5370574593544006,
      "learning_rate": 8e-05,
      "loss": 1.5138,
      "step": 3338
    },
    {
      "epoch": 0.6934579439252336,
      "grad_norm": 0.5026061534881592,
      "learning_rate": 8e-05,
      "loss": 1.5112,
      "step": 3339
    },
    {
      "epoch": 0.6936656282450675,
      "grad_norm": 0.5202752351760864,
      "learning_rate": 8e-05,
      "loss": 1.5931,
      "step": 3340
    },
    {
      "epoch": 0.6938733125649014,
      "grad_norm": 0.5417784452438354,
      "learning_rate": 8e-05,
      "loss": 1.5807,
      "step": 3341
    },
    {
      "epoch": 0.6940809968847352,
      "grad_norm": 0.5327773690223694,
      "learning_rate": 8e-05,
      "loss": 1.4861,
      "step": 3342
    },
    {
      "epoch": 0.6942886812045691,
      "grad_norm": 0.5095700621604919,
      "learning_rate": 8e-05,
      "loss": 1.4801,
      "step": 3343
    },
    {
      "epoch": 0.6944963655244029,
      "grad_norm": 0.5191102027893066,
      "learning_rate": 8e-05,
      "loss": 1.5028,
      "step": 3344
    },
    {
      "epoch": 0.6947040498442367,
      "grad_norm": 0.5580428242683411,
      "learning_rate": 8e-05,
      "loss": 1.5094,
      "step": 3345
    },
    {
      "epoch": 0.6949117341640706,
      "grad_norm": 0.5500212907791138,
      "learning_rate": 8e-05,
      "loss": 1.5521,
      "step": 3346
    },
    {
      "epoch": 0.6951194184839045,
      "grad_norm": 0.5216303467750549,
      "learning_rate": 8e-05,
      "loss": 1.6048,
      "step": 3347
    },
    {
      "epoch": 0.6953271028037383,
      "grad_norm": 0.4946167767047882,
      "learning_rate": 8e-05,
      "loss": 1.4028,
      "step": 3348
    },
    {
      "epoch": 0.6955347871235722,
      "grad_norm": 0.5921928286552429,
      "learning_rate": 8e-05,
      "loss": 1.5024,
      "step": 3349
    },
    {
      "epoch": 0.6957424714434061,
      "grad_norm": 0.5284919142723083,
      "learning_rate": 8e-05,
      "loss": 1.5762,
      "step": 3350
    },
    {
      "epoch": 0.6959501557632399,
      "grad_norm": 0.5295422077178955,
      "learning_rate": 8e-05,
      "loss": 1.6271,
      "step": 3351
    },
    {
      "epoch": 0.6961578400830737,
      "grad_norm": 0.5277194976806641,
      "learning_rate": 8e-05,
      "loss": 1.5262,
      "step": 3352
    },
    {
      "epoch": 0.6963655244029076,
      "grad_norm": 0.5392433404922485,
      "learning_rate": 8e-05,
      "loss": 1.6063,
      "step": 3353
    },
    {
      "epoch": 0.6965732087227414,
      "grad_norm": 0.5105103254318237,
      "learning_rate": 8e-05,
      "loss": 1.5345,
      "step": 3354
    },
    {
      "epoch": 0.6967808930425753,
      "grad_norm": 0.533957839012146,
      "learning_rate": 8e-05,
      "loss": 1.5114,
      "step": 3355
    },
    {
      "epoch": 0.6969885773624092,
      "grad_norm": 0.533530592918396,
      "learning_rate": 8e-05,
      "loss": 1.5849,
      "step": 3356
    },
    {
      "epoch": 0.697196261682243,
      "grad_norm": 0.5333319902420044,
      "learning_rate": 8e-05,
      "loss": 1.5065,
      "step": 3357
    },
    {
      "epoch": 0.6974039460020769,
      "grad_norm": 0.5846842527389526,
      "learning_rate": 8e-05,
      "loss": 1.5499,
      "step": 3358
    },
    {
      "epoch": 0.6976116303219106,
      "grad_norm": 0.5376784205436707,
      "learning_rate": 8e-05,
      "loss": 1.4858,
      "step": 3359
    },
    {
      "epoch": 0.6978193146417445,
      "grad_norm": 0.5476508736610413,
      "learning_rate": 8e-05,
      "loss": 1.5555,
      "step": 3360
    },
    {
      "epoch": 0.6980269989615784,
      "grad_norm": 0.5344243049621582,
      "learning_rate": 8e-05,
      "loss": 1.5771,
      "step": 3361
    },
    {
      "epoch": 0.6982346832814122,
      "grad_norm": 0.548423171043396,
      "learning_rate": 8e-05,
      "loss": 1.5967,
      "step": 3362
    },
    {
      "epoch": 0.6984423676012461,
      "grad_norm": 0.5320685505867004,
      "learning_rate": 8e-05,
      "loss": 1.5057,
      "step": 3363
    },
    {
      "epoch": 0.69865005192108,
      "grad_norm": 0.5275939702987671,
      "learning_rate": 8e-05,
      "loss": 1.6339,
      "step": 3364
    },
    {
      "epoch": 0.6988577362409139,
      "grad_norm": 0.5213542580604553,
      "learning_rate": 8e-05,
      "loss": 1.5467,
      "step": 3365
    },
    {
      "epoch": 0.6990654205607477,
      "grad_norm": 0.6275697350502014,
      "learning_rate": 8e-05,
      "loss": 1.5509,
      "step": 3366
    },
    {
      "epoch": 0.6992731048805815,
      "grad_norm": 0.5778084397315979,
      "learning_rate": 8e-05,
      "loss": 1.4989,
      "step": 3367
    },
    {
      "epoch": 0.6994807892004153,
      "grad_norm": 0.525407612323761,
      "learning_rate": 8e-05,
      "loss": 1.5143,
      "step": 3368
    },
    {
      "epoch": 0.6996884735202492,
      "grad_norm": 0.6687942147254944,
      "learning_rate": 8e-05,
      "loss": 1.5324,
      "step": 3369
    },
    {
      "epoch": 0.6998961578400831,
      "grad_norm": 0.5476462244987488,
      "learning_rate": 8e-05,
      "loss": 1.5724,
      "step": 3370
    },
    {
      "epoch": 0.700103842159917,
      "grad_norm": 0.5212531685829163,
      "learning_rate": 8e-05,
      "loss": 1.5626,
      "step": 3371
    },
    {
      "epoch": 0.7003115264797508,
      "grad_norm": 0.5417208075523376,
      "learning_rate": 8e-05,
      "loss": 1.5214,
      "step": 3372
    },
    {
      "epoch": 0.7005192107995847,
      "grad_norm": 0.5677985548973083,
      "learning_rate": 8e-05,
      "loss": 1.5237,
      "step": 3373
    },
    {
      "epoch": 0.7007268951194184,
      "grad_norm": 0.5420865416526794,
      "learning_rate": 8e-05,
      "loss": 1.4964,
      "step": 3374
    },
    {
      "epoch": 0.7009345794392523,
      "grad_norm": 0.5534452795982361,
      "learning_rate": 8e-05,
      "loss": 1.5377,
      "step": 3375
    },
    {
      "epoch": 0.7011422637590862,
      "grad_norm": 0.6092537641525269,
      "learning_rate": 8e-05,
      "loss": 1.5212,
      "step": 3376
    },
    {
      "epoch": 0.70134994807892,
      "grad_norm": 0.5454957485198975,
      "learning_rate": 8e-05,
      "loss": 1.6043,
      "step": 3377
    },
    {
      "epoch": 0.7015576323987539,
      "grad_norm": 0.5490703582763672,
      "learning_rate": 8e-05,
      "loss": 1.5724,
      "step": 3378
    },
    {
      "epoch": 0.7017653167185878,
      "grad_norm": 0.5254939198493958,
      "learning_rate": 8e-05,
      "loss": 1.4922,
      "step": 3379
    },
    {
      "epoch": 0.7019730010384216,
      "grad_norm": 0.5790081024169922,
      "learning_rate": 8e-05,
      "loss": 1.6087,
      "step": 3380
    },
    {
      "epoch": 0.7021806853582554,
      "grad_norm": 0.5370410680770874,
      "learning_rate": 8e-05,
      "loss": 1.5913,
      "step": 3381
    },
    {
      "epoch": 0.7023883696780893,
      "grad_norm": 0.5259770154953003,
      "learning_rate": 8e-05,
      "loss": 1.5024,
      "step": 3382
    },
    {
      "epoch": 0.7025960539979231,
      "grad_norm": 0.5853316783905029,
      "learning_rate": 8e-05,
      "loss": 1.4998,
      "step": 3383
    },
    {
      "epoch": 0.702803738317757,
      "grad_norm": 0.5315910577774048,
      "learning_rate": 8e-05,
      "loss": 1.5471,
      "step": 3384
    },
    {
      "epoch": 0.7030114226375909,
      "grad_norm": 0.5402783751487732,
      "learning_rate": 8e-05,
      "loss": 1.5827,
      "step": 3385
    },
    {
      "epoch": 0.7032191069574247,
      "grad_norm": 0.5240859985351562,
      "learning_rate": 8e-05,
      "loss": 1.4901,
      "step": 3386
    },
    {
      "epoch": 0.7034267912772586,
      "grad_norm": 0.5147637128829956,
      "learning_rate": 8e-05,
      "loss": 1.5351,
      "step": 3387
    },
    {
      "epoch": 0.7036344755970925,
      "grad_norm": 0.5533263683319092,
      "learning_rate": 8e-05,
      "loss": 1.5742,
      "step": 3388
    },
    {
      "epoch": 0.7038421599169262,
      "grad_norm": 0.5614248514175415,
      "learning_rate": 8e-05,
      "loss": 1.5711,
      "step": 3389
    },
    {
      "epoch": 0.7040498442367601,
      "grad_norm": 0.5665902495384216,
      "learning_rate": 8e-05,
      "loss": 1.5691,
      "step": 3390
    },
    {
      "epoch": 0.704257528556594,
      "grad_norm": 0.5120455026626587,
      "learning_rate": 8e-05,
      "loss": 1.5091,
      "step": 3391
    },
    {
      "epoch": 0.7044652128764278,
      "grad_norm": 0.5080638527870178,
      "learning_rate": 8e-05,
      "loss": 1.4478,
      "step": 3392
    },
    {
      "epoch": 0.7046728971962617,
      "grad_norm": 0.5795406103134155,
      "learning_rate": 8e-05,
      "loss": 1.6067,
      "step": 3393
    },
    {
      "epoch": 0.7048805815160956,
      "grad_norm": 0.5283856391906738,
      "learning_rate": 8e-05,
      "loss": 1.5486,
      "step": 3394
    },
    {
      "epoch": 0.7050882658359294,
      "grad_norm": 0.5504375696182251,
      "learning_rate": 8e-05,
      "loss": 1.517,
      "step": 3395
    },
    {
      "epoch": 0.7052959501557632,
      "grad_norm": 0.5622807741165161,
      "learning_rate": 8e-05,
      "loss": 1.5626,
      "step": 3396
    },
    {
      "epoch": 0.705503634475597,
      "grad_norm": 0.5102107524871826,
      "learning_rate": 8e-05,
      "loss": 1.4042,
      "step": 3397
    },
    {
      "epoch": 0.7057113187954309,
      "grad_norm": 0.524326741695404,
      "learning_rate": 8e-05,
      "loss": 1.5609,
      "step": 3398
    },
    {
      "epoch": 0.7059190031152648,
      "grad_norm": 0.5115851759910583,
      "learning_rate": 8e-05,
      "loss": 1.5759,
      "step": 3399
    },
    {
      "epoch": 0.7061266874350987,
      "grad_norm": 0.5784327387809753,
      "learning_rate": 8e-05,
      "loss": 1.5562,
      "step": 3400
    },
    {
      "epoch": 0.7063343717549325,
      "grad_norm": 0.5472557544708252,
      "learning_rate": 8e-05,
      "loss": 1.4959,
      "step": 3401
    },
    {
      "epoch": 0.7065420560747664,
      "grad_norm": 0.5197425484657288,
      "learning_rate": 8e-05,
      "loss": 1.5754,
      "step": 3402
    },
    {
      "epoch": 0.7067497403946003,
      "grad_norm": 0.5759093761444092,
      "learning_rate": 8e-05,
      "loss": 1.6056,
      "step": 3403
    },
    {
      "epoch": 0.706957424714434,
      "grad_norm": 0.5086911916732788,
      "learning_rate": 8e-05,
      "loss": 1.5365,
      "step": 3404
    },
    {
      "epoch": 0.7071651090342679,
      "grad_norm": 0.5257661938667297,
      "learning_rate": 8e-05,
      "loss": 1.5421,
      "step": 3405
    },
    {
      "epoch": 0.7073727933541017,
      "grad_norm": 0.5204954743385315,
      "learning_rate": 8e-05,
      "loss": 1.5383,
      "step": 3406
    },
    {
      "epoch": 0.7075804776739356,
      "grad_norm": 0.5520181655883789,
      "learning_rate": 8e-05,
      "loss": 1.5114,
      "step": 3407
    },
    {
      "epoch": 0.7077881619937695,
      "grad_norm": 0.5500684976577759,
      "learning_rate": 8e-05,
      "loss": 1.579,
      "step": 3408
    },
    {
      "epoch": 0.7079958463136033,
      "grad_norm": 0.5092450380325317,
      "learning_rate": 8e-05,
      "loss": 1.5278,
      "step": 3409
    },
    {
      "epoch": 0.7082035306334372,
      "grad_norm": 0.5043307542800903,
      "learning_rate": 8e-05,
      "loss": 1.5389,
      "step": 3410
    },
    {
      "epoch": 0.708411214953271,
      "grad_norm": 0.520301103591919,
      "learning_rate": 8e-05,
      "loss": 1.5292,
      "step": 3411
    },
    {
      "epoch": 0.7086188992731048,
      "grad_norm": 0.5313974022865295,
      "learning_rate": 8e-05,
      "loss": 1.5535,
      "step": 3412
    },
    {
      "epoch": 0.7088265835929387,
      "grad_norm": 0.5289424657821655,
      "learning_rate": 8e-05,
      "loss": 1.5561,
      "step": 3413
    },
    {
      "epoch": 0.7090342679127726,
      "grad_norm": 0.5287513136863708,
      "learning_rate": 8e-05,
      "loss": 1.5357,
      "step": 3414
    },
    {
      "epoch": 0.7092419522326064,
      "grad_norm": 0.5026020407676697,
      "learning_rate": 8e-05,
      "loss": 1.5512,
      "step": 3415
    },
    {
      "epoch": 0.7094496365524403,
      "grad_norm": 0.5290111899375916,
      "learning_rate": 8e-05,
      "loss": 1.4993,
      "step": 3416
    },
    {
      "epoch": 0.7096573208722742,
      "grad_norm": 0.5144578814506531,
      "learning_rate": 8e-05,
      "loss": 1.532,
      "step": 3417
    },
    {
      "epoch": 0.709865005192108,
      "grad_norm": 0.524332582950592,
      "learning_rate": 8e-05,
      "loss": 1.5254,
      "step": 3418
    },
    {
      "epoch": 0.7100726895119418,
      "grad_norm": 0.5344679355621338,
      "learning_rate": 8e-05,
      "loss": 1.5618,
      "step": 3419
    },
    {
      "epoch": 0.7102803738317757,
      "grad_norm": 0.5045977830886841,
      "learning_rate": 8e-05,
      "loss": 1.5508,
      "step": 3420
    },
    {
      "epoch": 0.7104880581516095,
      "grad_norm": 0.6042292714118958,
      "learning_rate": 8e-05,
      "loss": 1.634,
      "step": 3421
    },
    {
      "epoch": 0.7106957424714434,
      "grad_norm": 0.5266607403755188,
      "learning_rate": 8e-05,
      "loss": 1.5272,
      "step": 3422
    },
    {
      "epoch": 0.7109034267912773,
      "grad_norm": 0.5095388889312744,
      "learning_rate": 8e-05,
      "loss": 1.5081,
      "step": 3423
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 0.520750105381012,
      "learning_rate": 8e-05,
      "loss": 1.5497,
      "step": 3424
    },
    {
      "epoch": 0.711318795430945,
      "grad_norm": 0.5199542045593262,
      "learning_rate": 8e-05,
      "loss": 1.5266,
      "step": 3425
    },
    {
      "epoch": 0.7115264797507788,
      "grad_norm": 0.5272617936134338,
      "learning_rate": 8e-05,
      "loss": 1.5398,
      "step": 3426
    },
    {
      "epoch": 0.7117341640706126,
      "grad_norm": 0.5266063809394836,
      "learning_rate": 8e-05,
      "loss": 1.6223,
      "step": 3427
    },
    {
      "epoch": 0.7119418483904465,
      "grad_norm": 0.5335801243782043,
      "learning_rate": 8e-05,
      "loss": 1.6319,
      "step": 3428
    },
    {
      "epoch": 0.7121495327102804,
      "grad_norm": 0.5172614455223083,
      "learning_rate": 8e-05,
      "loss": 1.5712,
      "step": 3429
    },
    {
      "epoch": 0.7123572170301142,
      "grad_norm": 0.530680239200592,
      "learning_rate": 8e-05,
      "loss": 1.4593,
      "step": 3430
    },
    {
      "epoch": 0.7125649013499481,
      "grad_norm": 0.5231680274009705,
      "learning_rate": 8e-05,
      "loss": 1.5627,
      "step": 3431
    },
    {
      "epoch": 0.712772585669782,
      "grad_norm": 0.5158873796463013,
      "learning_rate": 8e-05,
      "loss": 1.5036,
      "step": 3432
    },
    {
      "epoch": 0.7129802699896158,
      "grad_norm": 0.5409302115440369,
      "learning_rate": 8e-05,
      "loss": 1.5711,
      "step": 3433
    },
    {
      "epoch": 0.7131879543094496,
      "grad_norm": 0.5092711448669434,
      "learning_rate": 8e-05,
      "loss": 1.4867,
      "step": 3434
    },
    {
      "epoch": 0.7133956386292835,
      "grad_norm": 0.5376372933387756,
      "learning_rate": 8e-05,
      "loss": 1.4828,
      "step": 3435
    },
    {
      "epoch": 0.7136033229491173,
      "grad_norm": 0.5479236841201782,
      "learning_rate": 8e-05,
      "loss": 1.5426,
      "step": 3436
    },
    {
      "epoch": 0.7138110072689512,
      "grad_norm": 0.5100098848342896,
      "learning_rate": 8e-05,
      "loss": 1.5041,
      "step": 3437
    },
    {
      "epoch": 0.7140186915887851,
      "grad_norm": 0.5238848328590393,
      "learning_rate": 8e-05,
      "loss": 1.5349,
      "step": 3438
    },
    {
      "epoch": 0.7142263759086189,
      "grad_norm": 0.5074477195739746,
      "learning_rate": 8e-05,
      "loss": 1.6004,
      "step": 3439
    },
    {
      "epoch": 0.7144340602284528,
      "grad_norm": 0.5064161419868469,
      "learning_rate": 8e-05,
      "loss": 1.5853,
      "step": 3440
    },
    {
      "epoch": 0.7146417445482866,
      "grad_norm": 0.5056182742118835,
      "learning_rate": 8e-05,
      "loss": 1.5264,
      "step": 3441
    },
    {
      "epoch": 0.7148494288681204,
      "grad_norm": 0.5369742512702942,
      "learning_rate": 8e-05,
      "loss": 1.6015,
      "step": 3442
    },
    {
      "epoch": 0.7150571131879543,
      "grad_norm": 0.5019688606262207,
      "learning_rate": 8e-05,
      "loss": 1.5191,
      "step": 3443
    },
    {
      "epoch": 0.7152647975077882,
      "grad_norm": 0.5375534892082214,
      "learning_rate": 8e-05,
      "loss": 1.5675,
      "step": 3444
    },
    {
      "epoch": 0.715472481827622,
      "grad_norm": 0.5510736703872681,
      "learning_rate": 8e-05,
      "loss": 1.4996,
      "step": 3445
    },
    {
      "epoch": 0.7156801661474559,
      "grad_norm": 0.5127111077308655,
      "learning_rate": 8e-05,
      "loss": 1.5464,
      "step": 3446
    },
    {
      "epoch": 0.7158878504672898,
      "grad_norm": 0.5190873742103577,
      "learning_rate": 8e-05,
      "loss": 1.5222,
      "step": 3447
    },
    {
      "epoch": 0.7160955347871236,
      "grad_norm": 0.5694050192832947,
      "learning_rate": 8e-05,
      "loss": 1.5938,
      "step": 3448
    },
    {
      "epoch": 0.7163032191069574,
      "grad_norm": 0.5129214525222778,
      "learning_rate": 8e-05,
      "loss": 1.5368,
      "step": 3449
    },
    {
      "epoch": 0.7165109034267912,
      "grad_norm": 0.5361027121543884,
      "learning_rate": 8e-05,
      "loss": 1.5752,
      "step": 3450
    },
    {
      "epoch": 0.7167185877466251,
      "grad_norm": 0.5257657766342163,
      "learning_rate": 8e-05,
      "loss": 1.6174,
      "step": 3451
    },
    {
      "epoch": 0.716926272066459,
      "grad_norm": 0.5467077493667603,
      "learning_rate": 8e-05,
      "loss": 1.5806,
      "step": 3452
    },
    {
      "epoch": 0.7171339563862928,
      "grad_norm": 0.6113228797912598,
      "learning_rate": 8e-05,
      "loss": 1.5835,
      "step": 3453
    },
    {
      "epoch": 0.7173416407061267,
      "grad_norm": 0.5408405065536499,
      "learning_rate": 8e-05,
      "loss": 1.5268,
      "step": 3454
    },
    {
      "epoch": 0.7175493250259606,
      "grad_norm": 0.5440720915794373,
      "learning_rate": 8e-05,
      "loss": 1.5581,
      "step": 3455
    },
    {
      "epoch": 0.7177570093457943,
      "grad_norm": 0.5272147059440613,
      "learning_rate": 8e-05,
      "loss": 1.5071,
      "step": 3456
    },
    {
      "epoch": 0.7179646936656282,
      "grad_norm": 0.7017815709114075,
      "learning_rate": 8e-05,
      "loss": 1.5573,
      "step": 3457
    },
    {
      "epoch": 0.7181723779854621,
      "grad_norm": 0.5151212811470032,
      "learning_rate": 8e-05,
      "loss": 1.5104,
      "step": 3458
    },
    {
      "epoch": 0.7183800623052959,
      "grad_norm": 0.5468786358833313,
      "learning_rate": 8e-05,
      "loss": 1.5893,
      "step": 3459
    },
    {
      "epoch": 0.7185877466251298,
      "grad_norm": 0.5950132012367249,
      "learning_rate": 8e-05,
      "loss": 1.5387,
      "step": 3460
    },
    {
      "epoch": 0.7187954309449637,
      "grad_norm": 0.6163037419319153,
      "learning_rate": 8e-05,
      "loss": 1.5532,
      "step": 3461
    },
    {
      "epoch": 0.7190031152647975,
      "grad_norm": 0.5301946401596069,
      "learning_rate": 8e-05,
      "loss": 1.5504,
      "step": 3462
    },
    {
      "epoch": 0.7192107995846314,
      "grad_norm": 0.6254122257232666,
      "learning_rate": 8e-05,
      "loss": 1.5152,
      "step": 3463
    },
    {
      "epoch": 0.7194184839044652,
      "grad_norm": 0.5170111656188965,
      "learning_rate": 8e-05,
      "loss": 1.4381,
      "step": 3464
    },
    {
      "epoch": 0.719626168224299,
      "grad_norm": 0.5274315476417542,
      "learning_rate": 8e-05,
      "loss": 1.5424,
      "step": 3465
    },
    {
      "epoch": 0.7198338525441329,
      "grad_norm": 0.5312260985374451,
      "learning_rate": 8e-05,
      "loss": 1.54,
      "step": 3466
    },
    {
      "epoch": 0.7200415368639668,
      "grad_norm": 0.535598874092102,
      "learning_rate": 8e-05,
      "loss": 1.5218,
      "step": 3467
    },
    {
      "epoch": 0.7202492211838006,
      "grad_norm": 0.5161129832267761,
      "learning_rate": 8e-05,
      "loss": 1.4629,
      "step": 3468
    },
    {
      "epoch": 0.7204569055036345,
      "grad_norm": 0.5363750457763672,
      "learning_rate": 8e-05,
      "loss": 1.5512,
      "step": 3469
    },
    {
      "epoch": 0.7206645898234684,
      "grad_norm": 0.5508981347084045,
      "learning_rate": 8e-05,
      "loss": 1.5606,
      "step": 3470
    },
    {
      "epoch": 0.7208722741433021,
      "grad_norm": 0.5215038657188416,
      "learning_rate": 8e-05,
      "loss": 1.5769,
      "step": 3471
    },
    {
      "epoch": 0.721079958463136,
      "grad_norm": 0.5061600804328918,
      "learning_rate": 8e-05,
      "loss": 1.5505,
      "step": 3472
    },
    {
      "epoch": 0.7212876427829699,
      "grad_norm": 0.5280421376228333,
      "learning_rate": 8e-05,
      "loss": 1.5607,
      "step": 3473
    },
    {
      "epoch": 0.7214953271028037,
      "grad_norm": 0.5401023626327515,
      "learning_rate": 8e-05,
      "loss": 1.542,
      "step": 3474
    },
    {
      "epoch": 0.7217030114226376,
      "grad_norm": 0.5061230659484863,
      "learning_rate": 8e-05,
      "loss": 1.5816,
      "step": 3475
    },
    {
      "epoch": 0.7219106957424715,
      "grad_norm": 0.5174618363380432,
      "learning_rate": 8e-05,
      "loss": 1.4707,
      "step": 3476
    },
    {
      "epoch": 0.7221183800623053,
      "grad_norm": 0.5627274513244629,
      "learning_rate": 8e-05,
      "loss": 1.4916,
      "step": 3477
    },
    {
      "epoch": 0.7223260643821392,
      "grad_norm": 0.5398745536804199,
      "learning_rate": 8e-05,
      "loss": 1.5452,
      "step": 3478
    },
    {
      "epoch": 0.722533748701973,
      "grad_norm": 0.6112679839134216,
      "learning_rate": 8e-05,
      "loss": 1.5498,
      "step": 3479
    },
    {
      "epoch": 0.7227414330218068,
      "grad_norm": 0.558030366897583,
      "learning_rate": 8e-05,
      "loss": 1.5909,
      "step": 3480
    },
    {
      "epoch": 0.7229491173416407,
      "grad_norm": 0.5367770195007324,
      "learning_rate": 8e-05,
      "loss": 1.5875,
      "step": 3481
    },
    {
      "epoch": 0.7231568016614746,
      "grad_norm": 0.5455817580223083,
      "learning_rate": 8e-05,
      "loss": 1.5298,
      "step": 3482
    },
    {
      "epoch": 0.7233644859813084,
      "grad_norm": 0.5508899688720703,
      "learning_rate": 8e-05,
      "loss": 1.5809,
      "step": 3483
    },
    {
      "epoch": 0.7235721703011423,
      "grad_norm": 0.5574979782104492,
      "learning_rate": 8e-05,
      "loss": 1.5312,
      "step": 3484
    },
    {
      "epoch": 0.7237798546209762,
      "grad_norm": 0.5332745313644409,
      "learning_rate": 8e-05,
      "loss": 1.5433,
      "step": 3485
    },
    {
      "epoch": 0.7239875389408099,
      "grad_norm": 0.5174273252487183,
      "learning_rate": 8e-05,
      "loss": 1.4944,
      "step": 3486
    },
    {
      "epoch": 0.7241952232606438,
      "grad_norm": 0.5466638803482056,
      "learning_rate": 8e-05,
      "loss": 1.574,
      "step": 3487
    },
    {
      "epoch": 0.7244029075804777,
      "grad_norm": 0.5374346375465393,
      "learning_rate": 8e-05,
      "loss": 1.5367,
      "step": 3488
    },
    {
      "epoch": 0.7246105919003115,
      "grad_norm": 0.5599860548973083,
      "learning_rate": 8e-05,
      "loss": 1.5834,
      "step": 3489
    },
    {
      "epoch": 0.7248182762201454,
      "grad_norm": 0.5291121602058411,
      "learning_rate": 8e-05,
      "loss": 1.5585,
      "step": 3490
    },
    {
      "epoch": 0.7250259605399793,
      "grad_norm": 0.5371955037117004,
      "learning_rate": 8e-05,
      "loss": 1.5339,
      "step": 3491
    },
    {
      "epoch": 0.7252336448598131,
      "grad_norm": 0.5767238140106201,
      "learning_rate": 8e-05,
      "loss": 1.5604,
      "step": 3492
    },
    {
      "epoch": 0.725441329179647,
      "grad_norm": 0.5335049629211426,
      "learning_rate": 8e-05,
      "loss": 1.4911,
      "step": 3493
    },
    {
      "epoch": 0.7256490134994807,
      "grad_norm": 0.5382176637649536,
      "learning_rate": 8e-05,
      "loss": 1.5721,
      "step": 3494
    },
    {
      "epoch": 0.7258566978193146,
      "grad_norm": 0.5417324900627136,
      "learning_rate": 8e-05,
      "loss": 1.5321,
      "step": 3495
    },
    {
      "epoch": 0.7260643821391485,
      "grad_norm": 0.5267302393913269,
      "learning_rate": 8e-05,
      "loss": 1.5867,
      "step": 3496
    },
    {
      "epoch": 0.7262720664589823,
      "grad_norm": 0.5352759957313538,
      "learning_rate": 8e-05,
      "loss": 1.549,
      "step": 3497
    },
    {
      "epoch": 0.7264797507788162,
      "grad_norm": 0.5570414662361145,
      "learning_rate": 8e-05,
      "loss": 1.5718,
      "step": 3498
    },
    {
      "epoch": 0.7266874350986501,
      "grad_norm": 0.5303803086280823,
      "learning_rate": 8e-05,
      "loss": 1.5508,
      "step": 3499
    },
    {
      "epoch": 0.726895119418484,
      "grad_norm": 0.5424216389656067,
      "learning_rate": 8e-05,
      "loss": 1.5587,
      "step": 3500
    },
    {
      "epoch": 0.7271028037383177,
      "grad_norm": 0.5421119332313538,
      "learning_rate": 8e-05,
      "loss": 1.4842,
      "step": 3501
    },
    {
      "epoch": 0.7273104880581516,
      "grad_norm": 0.5182427763938904,
      "learning_rate": 8e-05,
      "loss": 1.5051,
      "step": 3502
    },
    {
      "epoch": 0.7275181723779854,
      "grad_norm": 0.5280840396881104,
      "learning_rate": 8e-05,
      "loss": 1.5393,
      "step": 3503
    },
    {
      "epoch": 0.7277258566978193,
      "grad_norm": 0.5481265783309937,
      "learning_rate": 8e-05,
      "loss": 1.5867,
      "step": 3504
    },
    {
      "epoch": 0.7279335410176532,
      "grad_norm": 0.5149214863777161,
      "learning_rate": 8e-05,
      "loss": 1.5183,
      "step": 3505
    },
    {
      "epoch": 0.728141225337487,
      "grad_norm": 0.5356025099754333,
      "learning_rate": 8e-05,
      "loss": 1.5413,
      "step": 3506
    },
    {
      "epoch": 0.7283489096573209,
      "grad_norm": 0.5142056941986084,
      "learning_rate": 8e-05,
      "loss": 1.4075,
      "step": 3507
    },
    {
      "epoch": 0.7285565939771548,
      "grad_norm": 0.5186562538146973,
      "learning_rate": 8e-05,
      "loss": 1.5592,
      "step": 3508
    },
    {
      "epoch": 0.7287642782969885,
      "grad_norm": 0.5602302551269531,
      "learning_rate": 8e-05,
      "loss": 1.5396,
      "step": 3509
    },
    {
      "epoch": 0.7289719626168224,
      "grad_norm": 0.5718480348587036,
      "learning_rate": 8e-05,
      "loss": 1.5244,
      "step": 3510
    },
    {
      "epoch": 0.7291796469366563,
      "grad_norm": 0.4969232380390167,
      "learning_rate": 8e-05,
      "loss": 1.563,
      "step": 3511
    },
    {
      "epoch": 0.7293873312564901,
      "grad_norm": 0.5076689720153809,
      "learning_rate": 8e-05,
      "loss": 1.5113,
      "step": 3512
    },
    {
      "epoch": 0.729595015576324,
      "grad_norm": 0.5111002326011658,
      "learning_rate": 8e-05,
      "loss": 1.5443,
      "step": 3513
    },
    {
      "epoch": 0.7298026998961579,
      "grad_norm": 0.5244272947311401,
      "learning_rate": 8e-05,
      "loss": 1.5657,
      "step": 3514
    },
    {
      "epoch": 0.7300103842159917,
      "grad_norm": 0.49697282910346985,
      "learning_rate": 8e-05,
      "loss": 1.4675,
      "step": 3515
    },
    {
      "epoch": 0.7302180685358255,
      "grad_norm": 0.5233168005943298,
      "learning_rate": 8e-05,
      "loss": 1.5859,
      "step": 3516
    },
    {
      "epoch": 0.7304257528556594,
      "grad_norm": 0.5185022950172424,
      "learning_rate": 8e-05,
      "loss": 1.4999,
      "step": 3517
    },
    {
      "epoch": 0.7306334371754932,
      "grad_norm": 0.5441354513168335,
      "learning_rate": 8e-05,
      "loss": 1.5924,
      "step": 3518
    },
    {
      "epoch": 0.7308411214953271,
      "grad_norm": 0.5033422112464905,
      "learning_rate": 8e-05,
      "loss": 1.5096,
      "step": 3519
    },
    {
      "epoch": 0.731048805815161,
      "grad_norm": 0.4997793734073639,
      "learning_rate": 8e-05,
      "loss": 1.5173,
      "step": 3520
    },
    {
      "epoch": 0.7312564901349948,
      "grad_norm": 0.5255282521247864,
      "learning_rate": 8e-05,
      "loss": 1.4691,
      "step": 3521
    },
    {
      "epoch": 0.7314641744548287,
      "grad_norm": 0.5342074036598206,
      "learning_rate": 8e-05,
      "loss": 1.4995,
      "step": 3522
    },
    {
      "epoch": 0.7316718587746626,
      "grad_norm": 0.5344077944755554,
      "learning_rate": 8e-05,
      "loss": 1.5349,
      "step": 3523
    },
    {
      "epoch": 0.7318795430944963,
      "grad_norm": 0.5318564176559448,
      "learning_rate": 8e-05,
      "loss": 1.5616,
      "step": 3524
    },
    {
      "epoch": 0.7320872274143302,
      "grad_norm": 0.5484447479248047,
      "learning_rate": 8e-05,
      "loss": 1.5303,
      "step": 3525
    },
    {
      "epoch": 0.7322949117341641,
      "grad_norm": 0.5155816078186035,
      "learning_rate": 8e-05,
      "loss": 1.5218,
      "step": 3526
    },
    {
      "epoch": 0.7325025960539979,
      "grad_norm": 0.5102131366729736,
      "learning_rate": 8e-05,
      "loss": 1.5057,
      "step": 3527
    },
    {
      "epoch": 0.7327102803738318,
      "grad_norm": 0.526218593120575,
      "learning_rate": 8e-05,
      "loss": 1.5454,
      "step": 3528
    },
    {
      "epoch": 0.7329179646936657,
      "grad_norm": 0.5744272470474243,
      "learning_rate": 8e-05,
      "loss": 1.6486,
      "step": 3529
    },
    {
      "epoch": 0.7331256490134995,
      "grad_norm": 0.5435904860496521,
      "learning_rate": 8e-05,
      "loss": 1.5996,
      "step": 3530
    },
    {
      "epoch": 0.7333333333333333,
      "grad_norm": 0.6104902029037476,
      "learning_rate": 8e-05,
      "loss": 1.5851,
      "step": 3531
    },
    {
      "epoch": 0.7335410176531671,
      "grad_norm": 0.5694754719734192,
      "learning_rate": 8e-05,
      "loss": 1.5397,
      "step": 3532
    },
    {
      "epoch": 0.733748701973001,
      "grad_norm": 0.5344653725624084,
      "learning_rate": 8e-05,
      "loss": 1.6037,
      "step": 3533
    },
    {
      "epoch": 0.7339563862928349,
      "grad_norm": 0.521511971950531,
      "learning_rate": 8e-05,
      "loss": 1.6081,
      "step": 3534
    },
    {
      "epoch": 0.7341640706126688,
      "grad_norm": 0.6099985241889954,
      "learning_rate": 8e-05,
      "loss": 1.5246,
      "step": 3535
    },
    {
      "epoch": 0.7343717549325026,
      "grad_norm": 0.5827126502990723,
      "learning_rate": 8e-05,
      "loss": 1.5082,
      "step": 3536
    },
    {
      "epoch": 0.7345794392523365,
      "grad_norm": 0.504467785358429,
      "learning_rate": 8e-05,
      "loss": 1.4979,
      "step": 3537
    },
    {
      "epoch": 0.7347871235721704,
      "grad_norm": 0.5229232907295227,
      "learning_rate": 8e-05,
      "loss": 1.5072,
      "step": 3538
    },
    {
      "epoch": 0.7349948078920041,
      "grad_norm": 0.5311305522918701,
      "learning_rate": 8e-05,
      "loss": 1.5319,
      "step": 3539
    },
    {
      "epoch": 0.735202492211838,
      "grad_norm": 0.5311620831489563,
      "learning_rate": 8e-05,
      "loss": 1.5488,
      "step": 3540
    },
    {
      "epoch": 0.7354101765316718,
      "grad_norm": 0.5268400311470032,
      "learning_rate": 8e-05,
      "loss": 1.5681,
      "step": 3541
    },
    {
      "epoch": 0.7356178608515057,
      "grad_norm": 0.5212923884391785,
      "learning_rate": 8e-05,
      "loss": 1.5745,
      "step": 3542
    },
    {
      "epoch": 0.7358255451713396,
      "grad_norm": 0.5616365671157837,
      "learning_rate": 8e-05,
      "loss": 1.5217,
      "step": 3543
    },
    {
      "epoch": 0.7360332294911734,
      "grad_norm": 0.5554832816123962,
      "learning_rate": 8e-05,
      "loss": 1.4965,
      "step": 3544
    },
    {
      "epoch": 0.7362409138110073,
      "grad_norm": 0.5522962808609009,
      "learning_rate": 8e-05,
      "loss": 1.5919,
      "step": 3545
    },
    {
      "epoch": 0.7364485981308411,
      "grad_norm": 0.5040736198425293,
      "learning_rate": 8e-05,
      "loss": 1.454,
      "step": 3546
    },
    {
      "epoch": 0.7366562824506749,
      "grad_norm": 0.5797344446182251,
      "learning_rate": 8e-05,
      "loss": 1.5454,
      "step": 3547
    },
    {
      "epoch": 0.7368639667705088,
      "grad_norm": 0.5217843651771545,
      "learning_rate": 8e-05,
      "loss": 1.5157,
      "step": 3548
    },
    {
      "epoch": 0.7370716510903427,
      "grad_norm": 0.5438790917396545,
      "learning_rate": 8e-05,
      "loss": 1.5654,
      "step": 3549
    },
    {
      "epoch": 0.7372793354101765,
      "grad_norm": 0.5103795528411865,
      "learning_rate": 8e-05,
      "loss": 1.45,
      "step": 3550
    },
    {
      "epoch": 0.7374870197300104,
      "grad_norm": 0.5335983633995056,
      "learning_rate": 8e-05,
      "loss": 1.4975,
      "step": 3551
    },
    {
      "epoch": 0.7376947040498443,
      "grad_norm": 0.5823286175727844,
      "learning_rate": 8e-05,
      "loss": 1.5197,
      "step": 3552
    },
    {
      "epoch": 0.7379023883696781,
      "grad_norm": 0.5598112344741821,
      "learning_rate": 8e-05,
      "loss": 1.5423,
      "step": 3553
    },
    {
      "epoch": 0.7381100726895119,
      "grad_norm": 0.5465172529220581,
      "learning_rate": 8e-05,
      "loss": 1.5409,
      "step": 3554
    },
    {
      "epoch": 0.7383177570093458,
      "grad_norm": 0.5616025328636169,
      "learning_rate": 8e-05,
      "loss": 1.6016,
      "step": 3555
    },
    {
      "epoch": 0.7385254413291796,
      "grad_norm": 0.5810492634773254,
      "learning_rate": 8e-05,
      "loss": 1.6343,
      "step": 3556
    },
    {
      "epoch": 0.7387331256490135,
      "grad_norm": 0.5739856958389282,
      "learning_rate": 8e-05,
      "loss": 1.5924,
      "step": 3557
    },
    {
      "epoch": 0.7389408099688474,
      "grad_norm": 0.5240557193756104,
      "learning_rate": 8e-05,
      "loss": 1.5295,
      "step": 3558
    },
    {
      "epoch": 0.7391484942886812,
      "grad_norm": 0.5313481092453003,
      "learning_rate": 8e-05,
      "loss": 1.5626,
      "step": 3559
    },
    {
      "epoch": 0.7393561786085151,
      "grad_norm": 0.6045335531234741,
      "learning_rate": 8e-05,
      "loss": 1.5439,
      "step": 3560
    },
    {
      "epoch": 0.7395638629283489,
      "grad_norm": 0.5207858681678772,
      "learning_rate": 8e-05,
      "loss": 1.5277,
      "step": 3561
    },
    {
      "epoch": 0.7397715472481827,
      "grad_norm": 0.5392525792121887,
      "learning_rate": 8e-05,
      "loss": 1.5702,
      "step": 3562
    },
    {
      "epoch": 0.7399792315680166,
      "grad_norm": 0.5182170271873474,
      "learning_rate": 8e-05,
      "loss": 1.4937,
      "step": 3563
    },
    {
      "epoch": 0.7401869158878505,
      "grad_norm": 0.5690430998802185,
      "learning_rate": 8e-05,
      "loss": 1.569,
      "step": 3564
    },
    {
      "epoch": 0.7403946002076843,
      "grad_norm": 0.5354436635971069,
      "learning_rate": 8e-05,
      "loss": 1.5953,
      "step": 3565
    },
    {
      "epoch": 0.7406022845275182,
      "grad_norm": 0.5204501748085022,
      "learning_rate": 8e-05,
      "loss": 1.5314,
      "step": 3566
    },
    {
      "epoch": 0.7408099688473521,
      "grad_norm": 0.5107585787773132,
      "learning_rate": 8e-05,
      "loss": 1.5173,
      "step": 3567
    },
    {
      "epoch": 0.7410176531671859,
      "grad_norm": 0.650890052318573,
      "learning_rate": 8e-05,
      "loss": 1.5377,
      "step": 3568
    },
    {
      "epoch": 0.7412253374870197,
      "grad_norm": 0.5243777632713318,
      "learning_rate": 8e-05,
      "loss": 1.5691,
      "step": 3569
    },
    {
      "epoch": 0.7414330218068536,
      "grad_norm": 0.5239023566246033,
      "learning_rate": 8e-05,
      "loss": 1.5388,
      "step": 3570
    },
    {
      "epoch": 0.7416407061266874,
      "grad_norm": 0.5219586491584778,
      "learning_rate": 8e-05,
      "loss": 1.5199,
      "step": 3571
    },
    {
      "epoch": 0.7418483904465213,
      "grad_norm": 0.5232170820236206,
      "learning_rate": 8e-05,
      "loss": 1.5444,
      "step": 3572
    },
    {
      "epoch": 0.7420560747663552,
      "grad_norm": 0.5388113856315613,
      "learning_rate": 8e-05,
      "loss": 1.5467,
      "step": 3573
    },
    {
      "epoch": 0.742263759086189,
      "grad_norm": 0.536497175693512,
      "learning_rate": 8e-05,
      "loss": 1.6033,
      "step": 3574
    },
    {
      "epoch": 0.7424714434060229,
      "grad_norm": 0.5292633771896362,
      "learning_rate": 8e-05,
      "loss": 1.5542,
      "step": 3575
    },
    {
      "epoch": 0.7426791277258566,
      "grad_norm": 0.5561937093734741,
      "learning_rate": 8e-05,
      "loss": 1.5639,
      "step": 3576
    },
    {
      "epoch": 0.7428868120456905,
      "grad_norm": 0.5604279041290283,
      "learning_rate": 8e-05,
      "loss": 1.5771,
      "step": 3577
    },
    {
      "epoch": 0.7430944963655244,
      "grad_norm": 0.5260542631149292,
      "learning_rate": 8e-05,
      "loss": 1.4796,
      "step": 3578
    },
    {
      "epoch": 0.7433021806853582,
      "grad_norm": 0.5196414589881897,
      "learning_rate": 8e-05,
      "loss": 1.4981,
      "step": 3579
    },
    {
      "epoch": 0.7435098650051921,
      "grad_norm": 0.5542535781860352,
      "learning_rate": 8e-05,
      "loss": 1.5655,
      "step": 3580
    },
    {
      "epoch": 0.743717549325026,
      "grad_norm": 0.5323366522789001,
      "learning_rate": 8e-05,
      "loss": 1.5435,
      "step": 3581
    },
    {
      "epoch": 0.7439252336448599,
      "grad_norm": 0.5446087718009949,
      "learning_rate": 8e-05,
      "loss": 1.4949,
      "step": 3582
    },
    {
      "epoch": 0.7441329179646937,
      "grad_norm": 0.5530855655670166,
      "learning_rate": 8e-05,
      "loss": 1.5354,
      "step": 3583
    },
    {
      "epoch": 0.7443406022845275,
      "grad_norm": 0.5483283996582031,
      "learning_rate": 8e-05,
      "loss": 1.5414,
      "step": 3584
    },
    {
      "epoch": 0.7445482866043613,
      "grad_norm": 0.5852934122085571,
      "learning_rate": 8e-05,
      "loss": 1.6045,
      "step": 3585
    },
    {
      "epoch": 0.7447559709241952,
      "grad_norm": 0.5104108452796936,
      "learning_rate": 8e-05,
      "loss": 1.4721,
      "step": 3586
    },
    {
      "epoch": 0.7449636552440291,
      "grad_norm": 0.5348018407821655,
      "learning_rate": 8e-05,
      "loss": 1.5866,
      "step": 3587
    },
    {
      "epoch": 0.7451713395638629,
      "grad_norm": 0.5467799305915833,
      "learning_rate": 8e-05,
      "loss": 1.4879,
      "step": 3588
    },
    {
      "epoch": 0.7453790238836968,
      "grad_norm": 0.5091169476509094,
      "learning_rate": 8e-05,
      "loss": 1.4809,
      "step": 3589
    },
    {
      "epoch": 0.7455867082035307,
      "grad_norm": 0.5519127249717712,
      "learning_rate": 8e-05,
      "loss": 1.5843,
      "step": 3590
    },
    {
      "epoch": 0.7457943925233644,
      "grad_norm": 0.5352176427841187,
      "learning_rate": 8e-05,
      "loss": 1.4491,
      "step": 3591
    },
    {
      "epoch": 0.7460020768431983,
      "grad_norm": 0.5244325399398804,
      "learning_rate": 8e-05,
      "loss": 1.4607,
      "step": 3592
    },
    {
      "epoch": 0.7462097611630322,
      "grad_norm": 0.536231279373169,
      "learning_rate": 8e-05,
      "loss": 1.5724,
      "step": 3593
    },
    {
      "epoch": 0.746417445482866,
      "grad_norm": 0.529789388179779,
      "learning_rate": 8e-05,
      "loss": 1.5177,
      "step": 3594
    },
    {
      "epoch": 0.7466251298026999,
      "grad_norm": 0.52439284324646,
      "learning_rate": 8e-05,
      "loss": 1.5126,
      "step": 3595
    },
    {
      "epoch": 0.7468328141225338,
      "grad_norm": 0.5214512348175049,
      "learning_rate": 8e-05,
      "loss": 1.5621,
      "step": 3596
    },
    {
      "epoch": 0.7470404984423676,
      "grad_norm": 0.5600927472114563,
      "learning_rate": 8e-05,
      "loss": 1.6236,
      "step": 3597
    },
    {
      "epoch": 0.7472481827622015,
      "grad_norm": 0.5270238518714905,
      "learning_rate": 8e-05,
      "loss": 1.5313,
      "step": 3598
    },
    {
      "epoch": 0.7474558670820353,
      "grad_norm": 0.5473811626434326,
      "learning_rate": 8e-05,
      "loss": 1.5617,
      "step": 3599
    },
    {
      "epoch": 0.7476635514018691,
      "grad_norm": 0.5100362300872803,
      "learning_rate": 8e-05,
      "loss": 1.4536,
      "step": 3600
    },
    {
      "epoch": 0.747871235721703,
      "grad_norm": 0.5378039479255676,
      "learning_rate": 8e-05,
      "loss": 1.5351,
      "step": 3601
    },
    {
      "epoch": 0.7480789200415369,
      "grad_norm": 0.5332158207893372,
      "learning_rate": 8e-05,
      "loss": 1.5725,
      "step": 3602
    },
    {
      "epoch": 0.7482866043613707,
      "grad_norm": 0.551750123500824,
      "learning_rate": 8e-05,
      "loss": 1.5355,
      "step": 3603
    },
    {
      "epoch": 0.7484942886812046,
      "grad_norm": 0.5231562256813049,
      "learning_rate": 8e-05,
      "loss": 1.532,
      "step": 3604
    },
    {
      "epoch": 0.7487019730010385,
      "grad_norm": 0.551157534122467,
      "learning_rate": 8e-05,
      "loss": 1.5723,
      "step": 3605
    },
    {
      "epoch": 0.7489096573208722,
      "grad_norm": 0.5363884568214417,
      "learning_rate": 8e-05,
      "loss": 1.5189,
      "step": 3606
    },
    {
      "epoch": 0.7491173416407061,
      "grad_norm": 0.5688973069190979,
      "learning_rate": 8e-05,
      "loss": 1.5926,
      "step": 3607
    },
    {
      "epoch": 0.74932502596054,
      "grad_norm": 0.5524569153785706,
      "learning_rate": 8e-05,
      "loss": 1.6058,
      "step": 3608
    },
    {
      "epoch": 0.7495327102803738,
      "grad_norm": 0.5445985794067383,
      "learning_rate": 8e-05,
      "loss": 1.5726,
      "step": 3609
    },
    {
      "epoch": 0.7497403946002077,
      "grad_norm": 0.5331879258155823,
      "learning_rate": 8e-05,
      "loss": 1.5344,
      "step": 3610
    },
    {
      "epoch": 0.7499480789200416,
      "grad_norm": 0.5841154456138611,
      "learning_rate": 8e-05,
      "loss": 1.5436,
      "step": 3611
    },
    {
      "epoch": 0.7501557632398754,
      "grad_norm": 0.6170110106468201,
      "learning_rate": 8e-05,
      "loss": 1.5712,
      "step": 3612
    },
    {
      "epoch": 0.7503634475597093,
      "grad_norm": 0.5854488015174866,
      "learning_rate": 8e-05,
      "loss": 1.4997,
      "step": 3613
    },
    {
      "epoch": 0.750571131879543,
      "grad_norm": 0.5285217761993408,
      "learning_rate": 8e-05,
      "loss": 1.6019,
      "step": 3614
    },
    {
      "epoch": 0.7507788161993769,
      "grad_norm": 0.5762574672698975,
      "learning_rate": 8e-05,
      "loss": 1.5872,
      "step": 3615
    },
    {
      "epoch": 0.7509865005192108,
      "grad_norm": 0.6581540703773499,
      "learning_rate": 8e-05,
      "loss": 1.5756,
      "step": 3616
    },
    {
      "epoch": 0.7511941848390447,
      "grad_norm": 0.6120259761810303,
      "learning_rate": 8e-05,
      "loss": 1.4895,
      "step": 3617
    },
    {
      "epoch": 0.7514018691588785,
      "grad_norm": 0.5296320915222168,
      "learning_rate": 8e-05,
      "loss": 1.5619,
      "step": 3618
    },
    {
      "epoch": 0.7516095534787124,
      "grad_norm": 0.5763751268386841,
      "learning_rate": 8e-05,
      "loss": 1.5127,
      "step": 3619
    },
    {
      "epoch": 0.7518172377985463,
      "grad_norm": 0.598600447177887,
      "learning_rate": 8e-05,
      "loss": 1.5114,
      "step": 3620
    },
    {
      "epoch": 0.75202492211838,
      "grad_norm": 0.5320046544075012,
      "learning_rate": 8e-05,
      "loss": 1.5635,
      "step": 3621
    },
    {
      "epoch": 0.7522326064382139,
      "grad_norm": 0.530717134475708,
      "learning_rate": 8e-05,
      "loss": 1.5821,
      "step": 3622
    },
    {
      "epoch": 0.7524402907580477,
      "grad_norm": 0.5376477837562561,
      "learning_rate": 8e-05,
      "loss": 1.6225,
      "step": 3623
    },
    {
      "epoch": 0.7526479750778816,
      "grad_norm": 0.5327107310295105,
      "learning_rate": 8e-05,
      "loss": 1.5769,
      "step": 3624
    },
    {
      "epoch": 0.7528556593977155,
      "grad_norm": 0.5061706900596619,
      "learning_rate": 8e-05,
      "loss": 1.5424,
      "step": 3625
    },
    {
      "epoch": 0.7530633437175493,
      "grad_norm": 0.525826632976532,
      "learning_rate": 8e-05,
      "loss": 1.602,
      "step": 3626
    },
    {
      "epoch": 0.7532710280373832,
      "grad_norm": 0.5364912748336792,
      "learning_rate": 8e-05,
      "loss": 1.5436,
      "step": 3627
    },
    {
      "epoch": 0.7534787123572171,
      "grad_norm": 0.5414999723434448,
      "learning_rate": 8e-05,
      "loss": 1.5878,
      "step": 3628
    },
    {
      "epoch": 0.7536863966770508,
      "grad_norm": 0.5492070317268372,
      "learning_rate": 8e-05,
      "loss": 1.5317,
      "step": 3629
    },
    {
      "epoch": 0.7538940809968847,
      "grad_norm": 0.5089728236198425,
      "learning_rate": 8e-05,
      "loss": 1.5096,
      "step": 3630
    },
    {
      "epoch": 0.7541017653167186,
      "grad_norm": 0.5130892395973206,
      "learning_rate": 8e-05,
      "loss": 1.567,
      "step": 3631
    },
    {
      "epoch": 0.7543094496365524,
      "grad_norm": 0.5085206627845764,
      "learning_rate": 8e-05,
      "loss": 1.4721,
      "step": 3632
    },
    {
      "epoch": 0.7545171339563863,
      "grad_norm": 0.5374481081962585,
      "learning_rate": 8e-05,
      "loss": 1.5467,
      "step": 3633
    },
    {
      "epoch": 0.7547248182762202,
      "grad_norm": 0.5519620776176453,
      "learning_rate": 8e-05,
      "loss": 1.521,
      "step": 3634
    },
    {
      "epoch": 0.754932502596054,
      "grad_norm": 0.5374706983566284,
      "learning_rate": 8e-05,
      "loss": 1.6377,
      "step": 3635
    },
    {
      "epoch": 0.7551401869158878,
      "grad_norm": 0.5524253249168396,
      "learning_rate": 8e-05,
      "loss": 1.5908,
      "step": 3636
    },
    {
      "epoch": 0.7553478712357217,
      "grad_norm": 0.5147302150726318,
      "learning_rate": 8e-05,
      "loss": 1.5239,
      "step": 3637
    },
    {
      "epoch": 0.7555555555555555,
      "grad_norm": 0.5289570689201355,
      "learning_rate": 8e-05,
      "loss": 1.4757,
      "step": 3638
    },
    {
      "epoch": 0.7557632398753894,
      "grad_norm": 0.530564546585083,
      "learning_rate": 8e-05,
      "loss": 1.5465,
      "step": 3639
    },
    {
      "epoch": 0.7559709241952233,
      "grad_norm": 0.5244117379188538,
      "learning_rate": 8e-05,
      "loss": 1.5932,
      "step": 3640
    },
    {
      "epoch": 0.7561786085150571,
      "grad_norm": 0.5700787901878357,
      "learning_rate": 8e-05,
      "loss": 1.4732,
      "step": 3641
    },
    {
      "epoch": 0.756386292834891,
      "grad_norm": 0.5700501799583435,
      "learning_rate": 8e-05,
      "loss": 1.5769,
      "step": 3642
    },
    {
      "epoch": 0.7565939771547249,
      "grad_norm": 0.5256038308143616,
      "learning_rate": 8e-05,
      "loss": 1.5362,
      "step": 3643
    },
    {
      "epoch": 0.7568016614745586,
      "grad_norm": 0.529216468334198,
      "learning_rate": 8e-05,
      "loss": 1.5283,
      "step": 3644
    },
    {
      "epoch": 0.7570093457943925,
      "grad_norm": 0.6143906116485596,
      "learning_rate": 8e-05,
      "loss": 1.5868,
      "step": 3645
    },
    {
      "epoch": 0.7572170301142264,
      "grad_norm": 0.5543007254600525,
      "learning_rate": 8e-05,
      "loss": 1.5385,
      "step": 3646
    },
    {
      "epoch": 0.7574247144340602,
      "grad_norm": 0.5231019854545593,
      "learning_rate": 8e-05,
      "loss": 1.492,
      "step": 3647
    },
    {
      "epoch": 0.7576323987538941,
      "grad_norm": 0.5208254456520081,
      "learning_rate": 8e-05,
      "loss": 1.5503,
      "step": 3648
    },
    {
      "epoch": 0.757840083073728,
      "grad_norm": 0.5446212887763977,
      "learning_rate": 8e-05,
      "loss": 1.5481,
      "step": 3649
    },
    {
      "epoch": 0.7580477673935618,
      "grad_norm": 0.5497376322746277,
      "learning_rate": 8e-05,
      "loss": 1.534,
      "step": 3650
    },
    {
      "epoch": 0.7582554517133956,
      "grad_norm": 0.506187379360199,
      "learning_rate": 8e-05,
      "loss": 1.3956,
      "step": 3651
    },
    {
      "epoch": 0.7584631360332295,
      "grad_norm": 0.5366476774215698,
      "learning_rate": 8e-05,
      "loss": 1.5138,
      "step": 3652
    },
    {
      "epoch": 0.7586708203530633,
      "grad_norm": 0.543053925037384,
      "learning_rate": 8e-05,
      "loss": 1.5754,
      "step": 3653
    },
    {
      "epoch": 0.7588785046728972,
      "grad_norm": 0.5126652121543884,
      "learning_rate": 8e-05,
      "loss": 1.5224,
      "step": 3654
    },
    {
      "epoch": 0.7590861889927311,
      "grad_norm": 0.5708200931549072,
      "learning_rate": 8e-05,
      "loss": 1.5735,
      "step": 3655
    },
    {
      "epoch": 0.7592938733125649,
      "grad_norm": 0.5376601815223694,
      "learning_rate": 8e-05,
      "loss": 1.5838,
      "step": 3656
    },
    {
      "epoch": 0.7595015576323988,
      "grad_norm": 0.5308980345726013,
      "learning_rate": 8e-05,
      "loss": 1.5478,
      "step": 3657
    },
    {
      "epoch": 0.7597092419522327,
      "grad_norm": 0.5364058017730713,
      "learning_rate": 8e-05,
      "loss": 1.5234,
      "step": 3658
    },
    {
      "epoch": 0.7599169262720664,
      "grad_norm": 0.5030051469802856,
      "learning_rate": 8e-05,
      "loss": 1.5507,
      "step": 3659
    },
    {
      "epoch": 0.7601246105919003,
      "grad_norm": 0.5164998769760132,
      "learning_rate": 8e-05,
      "loss": 1.5643,
      "step": 3660
    },
    {
      "epoch": 0.7603322949117342,
      "grad_norm": 0.5217921137809753,
      "learning_rate": 8e-05,
      "loss": 1.5219,
      "step": 3661
    },
    {
      "epoch": 0.760539979231568,
      "grad_norm": 0.5244473218917847,
      "learning_rate": 8e-05,
      "loss": 1.5788,
      "step": 3662
    },
    {
      "epoch": 0.7607476635514019,
      "grad_norm": 0.5254426598548889,
      "learning_rate": 8e-05,
      "loss": 1.564,
      "step": 3663
    },
    {
      "epoch": 0.7609553478712358,
      "grad_norm": 0.5221173763275146,
      "learning_rate": 8e-05,
      "loss": 1.5711,
      "step": 3664
    },
    {
      "epoch": 0.7611630321910696,
      "grad_norm": 0.5219637155532837,
      "learning_rate": 8e-05,
      "loss": 1.519,
      "step": 3665
    },
    {
      "epoch": 0.7613707165109034,
      "grad_norm": 0.5310571193695068,
      "learning_rate": 8e-05,
      "loss": 1.557,
      "step": 3666
    },
    {
      "epoch": 0.7615784008307372,
      "grad_norm": 0.5286462306976318,
      "learning_rate": 8e-05,
      "loss": 1.5303,
      "step": 3667
    },
    {
      "epoch": 0.7617860851505711,
      "grad_norm": 0.5410661101341248,
      "learning_rate": 8e-05,
      "loss": 1.5578,
      "step": 3668
    },
    {
      "epoch": 0.761993769470405,
      "grad_norm": 0.5603257417678833,
      "learning_rate": 8e-05,
      "loss": 1.5243,
      "step": 3669
    },
    {
      "epoch": 0.7622014537902388,
      "grad_norm": 0.5272408127784729,
      "learning_rate": 8e-05,
      "loss": 1.5781,
      "step": 3670
    },
    {
      "epoch": 0.7624091381100727,
      "grad_norm": 0.5292373299598694,
      "learning_rate": 8e-05,
      "loss": 1.5167,
      "step": 3671
    },
    {
      "epoch": 0.7626168224299066,
      "grad_norm": 0.5567618608474731,
      "learning_rate": 8e-05,
      "loss": 1.6142,
      "step": 3672
    },
    {
      "epoch": 0.7628245067497404,
      "grad_norm": 0.530898928642273,
      "learning_rate": 8e-05,
      "loss": 1.5782,
      "step": 3673
    },
    {
      "epoch": 0.7630321910695742,
      "grad_norm": 0.5554488301277161,
      "learning_rate": 8e-05,
      "loss": 1.5639,
      "step": 3674
    },
    {
      "epoch": 0.7632398753894081,
      "grad_norm": 0.5173704624176025,
      "learning_rate": 8e-05,
      "loss": 1.449,
      "step": 3675
    },
    {
      "epoch": 0.7634475597092419,
      "grad_norm": 0.5154054164886475,
      "learning_rate": 8e-05,
      "loss": 1.5573,
      "step": 3676
    },
    {
      "epoch": 0.7636552440290758,
      "grad_norm": 0.5558220744132996,
      "learning_rate": 8e-05,
      "loss": 1.4849,
      "step": 3677
    },
    {
      "epoch": 0.7638629283489097,
      "grad_norm": 0.5283823013305664,
      "learning_rate": 8e-05,
      "loss": 1.5467,
      "step": 3678
    },
    {
      "epoch": 0.7640706126687435,
      "grad_norm": 0.5086648464202881,
      "learning_rate": 8e-05,
      "loss": 1.5313,
      "step": 3679
    },
    {
      "epoch": 0.7642782969885774,
      "grad_norm": 0.5476326942443848,
      "learning_rate": 8e-05,
      "loss": 1.5565,
      "step": 3680
    },
    {
      "epoch": 0.7644859813084112,
      "grad_norm": 0.5187767148017883,
      "learning_rate": 8e-05,
      "loss": 1.5503,
      "step": 3681
    },
    {
      "epoch": 0.764693665628245,
      "grad_norm": 0.5209642648696899,
      "learning_rate": 8e-05,
      "loss": 1.521,
      "step": 3682
    },
    {
      "epoch": 0.7649013499480789,
      "grad_norm": 0.6158143281936646,
      "learning_rate": 8e-05,
      "loss": 1.5699,
      "step": 3683
    },
    {
      "epoch": 0.7651090342679128,
      "grad_norm": 0.533125102519989,
      "learning_rate": 8e-05,
      "loss": 1.5247,
      "step": 3684
    },
    {
      "epoch": 0.7653167185877466,
      "grad_norm": 0.5476651191711426,
      "learning_rate": 8e-05,
      "loss": 1.5817,
      "step": 3685
    },
    {
      "epoch": 0.7655244029075805,
      "grad_norm": 0.5406033992767334,
      "learning_rate": 8e-05,
      "loss": 1.5268,
      "step": 3686
    },
    {
      "epoch": 0.7657320872274144,
      "grad_norm": 0.5116484761238098,
      "learning_rate": 8e-05,
      "loss": 1.5659,
      "step": 3687
    },
    {
      "epoch": 0.7659397715472482,
      "grad_norm": 0.5250674486160278,
      "learning_rate": 8e-05,
      "loss": 1.5302,
      "step": 3688
    },
    {
      "epoch": 0.766147455867082,
      "grad_norm": 0.5246085524559021,
      "learning_rate": 8e-05,
      "loss": 1.5411,
      "step": 3689
    },
    {
      "epoch": 0.7663551401869159,
      "grad_norm": 0.551929771900177,
      "learning_rate": 8e-05,
      "loss": 1.5537,
      "step": 3690
    },
    {
      "epoch": 0.7665628245067497,
      "grad_norm": 0.5516889691352844,
      "learning_rate": 8e-05,
      "loss": 1.5546,
      "step": 3691
    },
    {
      "epoch": 0.7667705088265836,
      "grad_norm": 0.5642858743667603,
      "learning_rate": 8e-05,
      "loss": 1.5497,
      "step": 3692
    },
    {
      "epoch": 0.7669781931464175,
      "grad_norm": 0.5495569109916687,
      "learning_rate": 8e-05,
      "loss": 1.5542,
      "step": 3693
    },
    {
      "epoch": 0.7671858774662513,
      "grad_norm": 0.5332452058792114,
      "learning_rate": 8e-05,
      "loss": 1.604,
      "step": 3694
    },
    {
      "epoch": 0.7673935617860852,
      "grad_norm": 0.5165169835090637,
      "learning_rate": 8e-05,
      "loss": 1.5048,
      "step": 3695
    },
    {
      "epoch": 0.767601246105919,
      "grad_norm": 0.5231041312217712,
      "learning_rate": 8e-05,
      "loss": 1.5654,
      "step": 3696
    },
    {
      "epoch": 0.7678089304257528,
      "grad_norm": 0.5719845294952393,
      "learning_rate": 8e-05,
      "loss": 1.5162,
      "step": 3697
    },
    {
      "epoch": 0.7680166147455867,
      "grad_norm": 0.5225372314453125,
      "learning_rate": 8e-05,
      "loss": 1.4841,
      "step": 3698
    },
    {
      "epoch": 0.7682242990654206,
      "grad_norm": 0.5113576054573059,
      "learning_rate": 8e-05,
      "loss": 1.5891,
      "step": 3699
    },
    {
      "epoch": 0.7684319833852544,
      "grad_norm": 0.5561226606369019,
      "learning_rate": 8e-05,
      "loss": 1.5505,
      "step": 3700
    },
    {
      "epoch": 0.7686396677050883,
      "grad_norm": 0.6226199269294739,
      "learning_rate": 8e-05,
      "loss": 1.5935,
      "step": 3701
    },
    {
      "epoch": 0.7688473520249222,
      "grad_norm": 0.5137208700180054,
      "learning_rate": 8e-05,
      "loss": 1.522,
      "step": 3702
    },
    {
      "epoch": 0.7690550363447559,
      "grad_norm": 0.5757611393928528,
      "learning_rate": 8e-05,
      "loss": 1.5695,
      "step": 3703
    },
    {
      "epoch": 0.7692627206645898,
      "grad_norm": 0.5572495460510254,
      "learning_rate": 8e-05,
      "loss": 1.5331,
      "step": 3704
    },
    {
      "epoch": 0.7694704049844237,
      "grad_norm": 0.5161001682281494,
      "learning_rate": 8e-05,
      "loss": 1.4906,
      "step": 3705
    },
    {
      "epoch": 0.7696780893042575,
      "grad_norm": 0.5393316149711609,
      "learning_rate": 8e-05,
      "loss": 1.5211,
      "step": 3706
    },
    {
      "epoch": 0.7698857736240914,
      "grad_norm": 0.5212880373001099,
      "learning_rate": 8e-05,
      "loss": 1.5536,
      "step": 3707
    },
    {
      "epoch": 0.7700934579439253,
      "grad_norm": 0.5416380167007446,
      "learning_rate": 8e-05,
      "loss": 1.5221,
      "step": 3708
    },
    {
      "epoch": 0.7703011422637591,
      "grad_norm": 0.5313342213630676,
      "learning_rate": 8e-05,
      "loss": 1.5787,
      "step": 3709
    },
    {
      "epoch": 0.770508826583593,
      "grad_norm": 0.5286886692047119,
      "learning_rate": 8e-05,
      "loss": 1.4609,
      "step": 3710
    },
    {
      "epoch": 0.7707165109034267,
      "grad_norm": 0.5460977554321289,
      "learning_rate": 8e-05,
      "loss": 1.4522,
      "step": 3711
    },
    {
      "epoch": 0.7709241952232606,
      "grad_norm": 0.5295391082763672,
      "learning_rate": 8e-05,
      "loss": 1.5344,
      "step": 3712
    },
    {
      "epoch": 0.7711318795430945,
      "grad_norm": 0.519713282585144,
      "learning_rate": 8e-05,
      "loss": 1.4611,
      "step": 3713
    },
    {
      "epoch": 0.7713395638629283,
      "grad_norm": 0.5278452634811401,
      "learning_rate": 8e-05,
      "loss": 1.4816,
      "step": 3714
    },
    {
      "epoch": 0.7715472481827622,
      "grad_norm": 0.539854109287262,
      "learning_rate": 8e-05,
      "loss": 1.5747,
      "step": 3715
    },
    {
      "epoch": 0.7717549325025961,
      "grad_norm": 0.5963377952575684,
      "learning_rate": 8e-05,
      "loss": 1.5974,
      "step": 3716
    },
    {
      "epoch": 0.77196261682243,
      "grad_norm": 0.5380481481552124,
      "learning_rate": 8e-05,
      "loss": 1.5527,
      "step": 3717
    },
    {
      "epoch": 0.7721703011422637,
      "grad_norm": 0.5103070139884949,
      "learning_rate": 8e-05,
      "loss": 1.5822,
      "step": 3718
    },
    {
      "epoch": 0.7723779854620976,
      "grad_norm": 0.503212571144104,
      "learning_rate": 8e-05,
      "loss": 1.4821,
      "step": 3719
    },
    {
      "epoch": 0.7725856697819314,
      "grad_norm": 0.5585311055183411,
      "learning_rate": 8e-05,
      "loss": 1.5638,
      "step": 3720
    },
    {
      "epoch": 0.7727933541017653,
      "grad_norm": 0.5344582796096802,
      "learning_rate": 8e-05,
      "loss": 1.5107,
      "step": 3721
    },
    {
      "epoch": 0.7730010384215992,
      "grad_norm": 0.5314329266548157,
      "learning_rate": 8e-05,
      "loss": 1.5135,
      "step": 3722
    },
    {
      "epoch": 0.773208722741433,
      "grad_norm": 0.5369237065315247,
      "learning_rate": 8e-05,
      "loss": 1.5282,
      "step": 3723
    },
    {
      "epoch": 0.7734164070612669,
      "grad_norm": 0.5226271152496338,
      "learning_rate": 8e-05,
      "loss": 1.5969,
      "step": 3724
    },
    {
      "epoch": 0.7736240913811008,
      "grad_norm": 0.5536156296730042,
      "learning_rate": 8e-05,
      "loss": 1.4831,
      "step": 3725
    },
    {
      "epoch": 0.7738317757009345,
      "grad_norm": 0.5462344884872437,
      "learning_rate": 8e-05,
      "loss": 1.4605,
      "step": 3726
    },
    {
      "epoch": 0.7740394600207684,
      "grad_norm": 0.5441974997520447,
      "learning_rate": 8e-05,
      "loss": 1.5823,
      "step": 3727
    },
    {
      "epoch": 0.7742471443406023,
      "grad_norm": 0.6400229930877686,
      "learning_rate": 8e-05,
      "loss": 1.5086,
      "step": 3728
    },
    {
      "epoch": 0.7744548286604361,
      "grad_norm": 0.5908493399620056,
      "learning_rate": 8e-05,
      "loss": 1.5211,
      "step": 3729
    },
    {
      "epoch": 0.77466251298027,
      "grad_norm": 0.5397082567214966,
      "learning_rate": 8e-05,
      "loss": 1.569,
      "step": 3730
    },
    {
      "epoch": 0.7748701973001039,
      "grad_norm": 0.5884261131286621,
      "learning_rate": 8e-05,
      "loss": 1.5036,
      "step": 3731
    },
    {
      "epoch": 0.7750778816199377,
      "grad_norm": 0.5432841181755066,
      "learning_rate": 8e-05,
      "loss": 1.5318,
      "step": 3732
    },
    {
      "epoch": 0.7752855659397715,
      "grad_norm": 0.5402141809463501,
      "learning_rate": 8e-05,
      "loss": 1.4901,
      "step": 3733
    },
    {
      "epoch": 0.7754932502596054,
      "grad_norm": 0.503089189529419,
      "learning_rate": 8e-05,
      "loss": 1.425,
      "step": 3734
    },
    {
      "epoch": 0.7757009345794392,
      "grad_norm": 0.5395124554634094,
      "learning_rate": 8e-05,
      "loss": 1.5326,
      "step": 3735
    },
    {
      "epoch": 0.7759086188992731,
      "grad_norm": 0.5114360451698303,
      "learning_rate": 8e-05,
      "loss": 1.5063,
      "step": 3736
    },
    {
      "epoch": 0.776116303219107,
      "grad_norm": 0.5176555514335632,
      "learning_rate": 8e-05,
      "loss": 1.5408,
      "step": 3737
    },
    {
      "epoch": 0.7763239875389408,
      "grad_norm": 0.5622590780258179,
      "learning_rate": 8e-05,
      "loss": 1.5739,
      "step": 3738
    },
    {
      "epoch": 0.7765316718587747,
      "grad_norm": 0.5745481252670288,
      "learning_rate": 8e-05,
      "loss": 1.6205,
      "step": 3739
    },
    {
      "epoch": 0.7767393561786086,
      "grad_norm": 0.5247728824615479,
      "learning_rate": 8e-05,
      "loss": 1.5518,
      "step": 3740
    },
    {
      "epoch": 0.7769470404984423,
      "grad_norm": 0.5253990888595581,
      "learning_rate": 8e-05,
      "loss": 1.5576,
      "step": 3741
    },
    {
      "epoch": 0.7771547248182762,
      "grad_norm": 0.5572378635406494,
      "learning_rate": 8e-05,
      "loss": 1.6626,
      "step": 3742
    },
    {
      "epoch": 0.77736240913811,
      "grad_norm": 0.5535411834716797,
      "learning_rate": 8e-05,
      "loss": 1.5649,
      "step": 3743
    },
    {
      "epoch": 0.7775700934579439,
      "grad_norm": 0.5490415096282959,
      "learning_rate": 8e-05,
      "loss": 1.6202,
      "step": 3744
    },
    {
      "epoch": 0.7777777777777778,
      "grad_norm": 0.5346564650535583,
      "learning_rate": 8e-05,
      "loss": 1.4885,
      "step": 3745
    },
    {
      "epoch": 0.7779854620976117,
      "grad_norm": 0.5394704341888428,
      "learning_rate": 8e-05,
      "loss": 1.4872,
      "step": 3746
    },
    {
      "epoch": 0.7781931464174455,
      "grad_norm": 0.559691309928894,
      "learning_rate": 8e-05,
      "loss": 1.5719,
      "step": 3747
    },
    {
      "epoch": 0.7784008307372793,
      "grad_norm": 0.5369914770126343,
      "learning_rate": 8e-05,
      "loss": 1.5087,
      "step": 3748
    },
    {
      "epoch": 0.7786085150571131,
      "grad_norm": 0.5158330202102661,
      "learning_rate": 8e-05,
      "loss": 1.5148,
      "step": 3749
    },
    {
      "epoch": 0.778816199376947,
      "grad_norm": 0.5457910895347595,
      "learning_rate": 8e-05,
      "loss": 1.5982,
      "step": 3750
    },
    {
      "epoch": 0.7790238836967809,
      "grad_norm": 0.5563228130340576,
      "learning_rate": 8e-05,
      "loss": 1.5853,
      "step": 3751
    },
    {
      "epoch": 0.7792315680166148,
      "grad_norm": 0.5679547786712646,
      "learning_rate": 8e-05,
      "loss": 1.54,
      "step": 3752
    },
    {
      "epoch": 0.7794392523364486,
      "grad_norm": 0.5261079668998718,
      "learning_rate": 8e-05,
      "loss": 1.494,
      "step": 3753
    },
    {
      "epoch": 0.7796469366562825,
      "grad_norm": 0.5446485280990601,
      "learning_rate": 8e-05,
      "loss": 1.5572,
      "step": 3754
    },
    {
      "epoch": 0.7798546209761164,
      "grad_norm": 0.564666748046875,
      "learning_rate": 8e-05,
      "loss": 1.6029,
      "step": 3755
    },
    {
      "epoch": 0.7800623052959501,
      "grad_norm": 0.5414217114448547,
      "learning_rate": 8e-05,
      "loss": 1.5387,
      "step": 3756
    },
    {
      "epoch": 0.780269989615784,
      "grad_norm": 0.5821365714073181,
      "learning_rate": 8e-05,
      "loss": 1.536,
      "step": 3757
    },
    {
      "epoch": 0.7804776739356178,
      "grad_norm": 0.5483496785163879,
      "learning_rate": 8e-05,
      "loss": 1.5154,
      "step": 3758
    },
    {
      "epoch": 0.7806853582554517,
      "grad_norm": 0.536778450012207,
      "learning_rate": 8e-05,
      "loss": 1.5255,
      "step": 3759
    },
    {
      "epoch": 0.7808930425752856,
      "grad_norm": 0.5337021350860596,
      "learning_rate": 8e-05,
      "loss": 1.558,
      "step": 3760
    },
    {
      "epoch": 0.7811007268951194,
      "grad_norm": 0.5149087309837341,
      "learning_rate": 8e-05,
      "loss": 1.5522,
      "step": 3761
    },
    {
      "epoch": 0.7813084112149533,
      "grad_norm": 0.5480493307113647,
      "learning_rate": 8e-05,
      "loss": 1.5485,
      "step": 3762
    },
    {
      "epoch": 0.7815160955347871,
      "grad_norm": 0.5506593585014343,
      "learning_rate": 8e-05,
      "loss": 1.5485,
      "step": 3763
    },
    {
      "epoch": 0.7817237798546209,
      "grad_norm": 0.5390865802764893,
      "learning_rate": 8e-05,
      "loss": 1.4992,
      "step": 3764
    },
    {
      "epoch": 0.7819314641744548,
      "grad_norm": 0.5540247559547424,
      "learning_rate": 8e-05,
      "loss": 1.4694,
      "step": 3765
    },
    {
      "epoch": 0.7821391484942887,
      "grad_norm": 0.5953584313392639,
      "learning_rate": 8e-05,
      "loss": 1.5755,
      "step": 3766
    },
    {
      "epoch": 0.7823468328141225,
      "grad_norm": 0.5392823219299316,
      "learning_rate": 8e-05,
      "loss": 1.5343,
      "step": 3767
    },
    {
      "epoch": 0.7825545171339564,
      "grad_norm": 0.5309379696846008,
      "learning_rate": 8e-05,
      "loss": 1.5243,
      "step": 3768
    },
    {
      "epoch": 0.7827622014537903,
      "grad_norm": 0.5200415253639221,
      "learning_rate": 8e-05,
      "loss": 1.5701,
      "step": 3769
    },
    {
      "epoch": 0.7829698857736241,
      "grad_norm": 0.5380063652992249,
      "learning_rate": 8e-05,
      "loss": 1.5039,
      "step": 3770
    },
    {
      "epoch": 0.7831775700934579,
      "grad_norm": 0.5777196884155273,
      "learning_rate": 8e-05,
      "loss": 1.5382,
      "step": 3771
    },
    {
      "epoch": 0.7833852544132918,
      "grad_norm": 0.5390869379043579,
      "learning_rate": 8e-05,
      "loss": 1.6057,
      "step": 3772
    },
    {
      "epoch": 0.7835929387331256,
      "grad_norm": 0.5991033315658569,
      "learning_rate": 8e-05,
      "loss": 1.5773,
      "step": 3773
    },
    {
      "epoch": 0.7838006230529595,
      "grad_norm": 0.5758764743804932,
      "learning_rate": 8e-05,
      "loss": 1.538,
      "step": 3774
    },
    {
      "epoch": 0.7840083073727934,
      "grad_norm": 0.5281041860580444,
      "learning_rate": 8e-05,
      "loss": 1.4606,
      "step": 3775
    },
    {
      "epoch": 0.7842159916926272,
      "grad_norm": 0.5381190180778503,
      "learning_rate": 8e-05,
      "loss": 1.4627,
      "step": 3776
    },
    {
      "epoch": 0.7844236760124611,
      "grad_norm": 0.6352688670158386,
      "learning_rate": 8e-05,
      "loss": 1.53,
      "step": 3777
    },
    {
      "epoch": 0.7846313603322949,
      "grad_norm": 0.5391112565994263,
      "learning_rate": 8e-05,
      "loss": 1.5821,
      "step": 3778
    },
    {
      "epoch": 0.7848390446521287,
      "grad_norm": 0.5541571974754333,
      "learning_rate": 8e-05,
      "loss": 1.5405,
      "step": 3779
    },
    {
      "epoch": 0.7850467289719626,
      "grad_norm": 0.5539214015007019,
      "learning_rate": 8e-05,
      "loss": 1.5434,
      "step": 3780
    },
    {
      "epoch": 0.7852544132917965,
      "grad_norm": 0.6011430621147156,
      "learning_rate": 8e-05,
      "loss": 1.5797,
      "step": 3781
    },
    {
      "epoch": 0.7854620976116303,
      "grad_norm": 0.5402131676673889,
      "learning_rate": 8e-05,
      "loss": 1.4782,
      "step": 3782
    },
    {
      "epoch": 0.7856697819314642,
      "grad_norm": 0.5159510970115662,
      "learning_rate": 8e-05,
      "loss": 1.5968,
      "step": 3783
    },
    {
      "epoch": 0.7858774662512981,
      "grad_norm": 0.6394030451774597,
      "learning_rate": 8e-05,
      "loss": 1.4903,
      "step": 3784
    },
    {
      "epoch": 0.7860851505711319,
      "grad_norm": 0.527470588684082,
      "learning_rate": 8e-05,
      "loss": 1.5388,
      "step": 3785
    },
    {
      "epoch": 0.7862928348909657,
      "grad_norm": 0.5422571897506714,
      "learning_rate": 8e-05,
      "loss": 1.5523,
      "step": 3786
    },
    {
      "epoch": 0.7865005192107996,
      "grad_norm": 0.5498519539833069,
      "learning_rate": 8e-05,
      "loss": 1.5383,
      "step": 3787
    },
    {
      "epoch": 0.7867082035306334,
      "grad_norm": 0.5266768932342529,
      "learning_rate": 8e-05,
      "loss": 1.5432,
      "step": 3788
    },
    {
      "epoch": 0.7869158878504673,
      "grad_norm": 0.5560333728790283,
      "learning_rate": 8e-05,
      "loss": 1.5156,
      "step": 3789
    },
    {
      "epoch": 0.7871235721703012,
      "grad_norm": 0.5451468825340271,
      "learning_rate": 8e-05,
      "loss": 1.5328,
      "step": 3790
    },
    {
      "epoch": 0.787331256490135,
      "grad_norm": 0.5401631593704224,
      "learning_rate": 8e-05,
      "loss": 1.5332,
      "step": 3791
    },
    {
      "epoch": 0.7875389408099689,
      "grad_norm": 0.5429092049598694,
      "learning_rate": 8e-05,
      "loss": 1.6132,
      "step": 3792
    },
    {
      "epoch": 0.7877466251298026,
      "grad_norm": 0.626421332359314,
      "learning_rate": 8e-05,
      "loss": 1.5967,
      "step": 3793
    },
    {
      "epoch": 0.7879543094496365,
      "grad_norm": 0.5097197890281677,
      "learning_rate": 8e-05,
      "loss": 1.4764,
      "step": 3794
    },
    {
      "epoch": 0.7881619937694704,
      "grad_norm": 0.5446649193763733,
      "learning_rate": 8e-05,
      "loss": 1.5701,
      "step": 3795
    },
    {
      "epoch": 0.7883696780893042,
      "grad_norm": 0.5284068584442139,
      "learning_rate": 8e-05,
      "loss": 1.563,
      "step": 3796
    },
    {
      "epoch": 0.7885773624091381,
      "grad_norm": 0.538684606552124,
      "learning_rate": 8e-05,
      "loss": 1.5517,
      "step": 3797
    },
    {
      "epoch": 0.788785046728972,
      "grad_norm": 0.5488366484642029,
      "learning_rate": 8e-05,
      "loss": 1.5658,
      "step": 3798
    },
    {
      "epoch": 0.7889927310488059,
      "grad_norm": 0.5305287837982178,
      "learning_rate": 8e-05,
      "loss": 1.5848,
      "step": 3799
    },
    {
      "epoch": 0.7892004153686397,
      "grad_norm": 0.549168050289154,
      "learning_rate": 8e-05,
      "loss": 1.5821,
      "step": 3800
    },
    {
      "epoch": 0.7894080996884735,
      "grad_norm": 0.5640422105789185,
      "learning_rate": 8e-05,
      "loss": 1.5503,
      "step": 3801
    },
    {
      "epoch": 0.7896157840083073,
      "grad_norm": 0.5144663453102112,
      "learning_rate": 8e-05,
      "loss": 1.4981,
      "step": 3802
    },
    {
      "epoch": 0.7898234683281412,
      "grad_norm": 0.5397584438323975,
      "learning_rate": 8e-05,
      "loss": 1.5111,
      "step": 3803
    },
    {
      "epoch": 0.7900311526479751,
      "grad_norm": 0.5236941576004028,
      "learning_rate": 8e-05,
      "loss": 1.5957,
      "step": 3804
    },
    {
      "epoch": 0.7902388369678089,
      "grad_norm": 0.5418887138366699,
      "learning_rate": 8e-05,
      "loss": 1.5201,
      "step": 3805
    },
    {
      "epoch": 0.7904465212876428,
      "grad_norm": 0.52094966173172,
      "learning_rate": 8e-05,
      "loss": 1.5714,
      "step": 3806
    },
    {
      "epoch": 0.7906542056074767,
      "grad_norm": 0.5204759240150452,
      "learning_rate": 8e-05,
      "loss": 1.5828,
      "step": 3807
    },
    {
      "epoch": 0.7908618899273104,
      "grad_norm": 0.5478829741477966,
      "learning_rate": 8e-05,
      "loss": 1.5298,
      "step": 3808
    },
    {
      "epoch": 0.7910695742471443,
      "grad_norm": 0.5409246683120728,
      "learning_rate": 8e-05,
      "loss": 1.5374,
      "step": 3809
    },
    {
      "epoch": 0.7912772585669782,
      "grad_norm": 0.5874788761138916,
      "learning_rate": 8e-05,
      "loss": 1.4324,
      "step": 3810
    },
    {
      "epoch": 0.791484942886812,
      "grad_norm": 0.5431894063949585,
      "learning_rate": 8e-05,
      "loss": 1.5114,
      "step": 3811
    },
    {
      "epoch": 0.7916926272066459,
      "grad_norm": 0.5161185264587402,
      "learning_rate": 8e-05,
      "loss": 1.4868,
      "step": 3812
    },
    {
      "epoch": 0.7919003115264798,
      "grad_norm": 0.5417200922966003,
      "learning_rate": 8e-05,
      "loss": 1.4905,
      "step": 3813
    },
    {
      "epoch": 0.7921079958463136,
      "grad_norm": 0.6025152206420898,
      "learning_rate": 8e-05,
      "loss": 1.5265,
      "step": 3814
    },
    {
      "epoch": 0.7923156801661475,
      "grad_norm": 0.5396698713302612,
      "learning_rate": 8e-05,
      "loss": 1.57,
      "step": 3815
    },
    {
      "epoch": 0.7925233644859813,
      "grad_norm": 0.5804276466369629,
      "learning_rate": 8e-05,
      "loss": 1.5333,
      "step": 3816
    },
    {
      "epoch": 0.7927310488058151,
      "grad_norm": 0.5589690208435059,
      "learning_rate": 8e-05,
      "loss": 1.5731,
      "step": 3817
    },
    {
      "epoch": 0.792938733125649,
      "grad_norm": 0.5346149802207947,
      "learning_rate": 8e-05,
      "loss": 1.5591,
      "step": 3818
    },
    {
      "epoch": 0.7931464174454829,
      "grad_norm": 0.5690224766731262,
      "learning_rate": 8e-05,
      "loss": 1.5287,
      "step": 3819
    },
    {
      "epoch": 0.7933541017653167,
      "grad_norm": 0.5319130420684814,
      "learning_rate": 8e-05,
      "loss": 1.5573,
      "step": 3820
    },
    {
      "epoch": 0.7935617860851506,
      "grad_norm": 0.5262956619262695,
      "learning_rate": 8e-05,
      "loss": 1.5282,
      "step": 3821
    },
    {
      "epoch": 0.7937694704049845,
      "grad_norm": 0.5359975695610046,
      "learning_rate": 8e-05,
      "loss": 1.4986,
      "step": 3822
    },
    {
      "epoch": 0.7939771547248182,
      "grad_norm": 0.546958863735199,
      "learning_rate": 8e-05,
      "loss": 1.4408,
      "step": 3823
    },
    {
      "epoch": 0.7941848390446521,
      "grad_norm": 0.5572725534439087,
      "learning_rate": 8e-05,
      "loss": 1.5488,
      "step": 3824
    },
    {
      "epoch": 0.794392523364486,
      "grad_norm": 0.5235803127288818,
      "learning_rate": 8e-05,
      "loss": 1.5324,
      "step": 3825
    },
    {
      "epoch": 0.7946002076843198,
      "grad_norm": 0.6095070838928223,
      "learning_rate": 8e-05,
      "loss": 1.5498,
      "step": 3826
    },
    {
      "epoch": 0.7948078920041537,
      "grad_norm": 0.5295899510383606,
      "learning_rate": 8e-05,
      "loss": 1.5111,
      "step": 3827
    },
    {
      "epoch": 0.7950155763239876,
      "grad_norm": 0.5198343992233276,
      "learning_rate": 8e-05,
      "loss": 1.4896,
      "step": 3828
    },
    {
      "epoch": 0.7952232606438214,
      "grad_norm": 0.588471531867981,
      "learning_rate": 8e-05,
      "loss": 1.5409,
      "step": 3829
    },
    {
      "epoch": 0.7954309449636553,
      "grad_norm": 0.5903350710868835,
      "learning_rate": 8e-05,
      "loss": 1.5569,
      "step": 3830
    },
    {
      "epoch": 0.795638629283489,
      "grad_norm": 0.5204113125801086,
      "learning_rate": 8e-05,
      "loss": 1.4548,
      "step": 3831
    },
    {
      "epoch": 0.7958463136033229,
      "grad_norm": 0.53733229637146,
      "learning_rate": 8e-05,
      "loss": 1.5114,
      "step": 3832
    },
    {
      "epoch": 0.7960539979231568,
      "grad_norm": 0.5347501635551453,
      "learning_rate": 8e-05,
      "loss": 1.5253,
      "step": 3833
    },
    {
      "epoch": 0.7962616822429907,
      "grad_norm": 0.5387030839920044,
      "learning_rate": 8e-05,
      "loss": 1.4934,
      "step": 3834
    },
    {
      "epoch": 0.7964693665628245,
      "grad_norm": 0.5264984965324402,
      "learning_rate": 8e-05,
      "loss": 1.5689,
      "step": 3835
    },
    {
      "epoch": 0.7966770508826584,
      "grad_norm": 0.5262815952301025,
      "learning_rate": 8e-05,
      "loss": 1.5178,
      "step": 3836
    },
    {
      "epoch": 0.7968847352024923,
      "grad_norm": 0.5308522582054138,
      "learning_rate": 8e-05,
      "loss": 1.4885,
      "step": 3837
    },
    {
      "epoch": 0.797092419522326,
      "grad_norm": 0.5495749115943909,
      "learning_rate": 8e-05,
      "loss": 1.5334,
      "step": 3838
    },
    {
      "epoch": 0.7973001038421599,
      "grad_norm": 0.5428898930549622,
      "learning_rate": 8e-05,
      "loss": 1.5138,
      "step": 3839
    },
    {
      "epoch": 0.7975077881619937,
      "grad_norm": 0.5243093967437744,
      "learning_rate": 8e-05,
      "loss": 1.5552,
      "step": 3840
    },
    {
      "epoch": 0.7977154724818276,
      "grad_norm": 0.5269336104393005,
      "learning_rate": 8e-05,
      "loss": 1.6203,
      "step": 3841
    },
    {
      "epoch": 0.7979231568016615,
      "grad_norm": 0.541847825050354,
      "learning_rate": 8e-05,
      "loss": 1.5833,
      "step": 3842
    },
    {
      "epoch": 0.7981308411214953,
      "grad_norm": 0.5268791317939758,
      "learning_rate": 8e-05,
      "loss": 1.4895,
      "step": 3843
    },
    {
      "epoch": 0.7983385254413292,
      "grad_norm": 0.5529650449752808,
      "learning_rate": 8e-05,
      "loss": 1.595,
      "step": 3844
    },
    {
      "epoch": 0.7985462097611631,
      "grad_norm": 0.5543302893638611,
      "learning_rate": 8e-05,
      "loss": 1.5282,
      "step": 3845
    },
    {
      "epoch": 0.7987538940809968,
      "grad_norm": 0.5363597869873047,
      "learning_rate": 8e-05,
      "loss": 1.5711,
      "step": 3846
    },
    {
      "epoch": 0.7989615784008307,
      "grad_norm": 0.5447390675544739,
      "learning_rate": 8e-05,
      "loss": 1.5641,
      "step": 3847
    },
    {
      "epoch": 0.7991692627206646,
      "grad_norm": 0.5374597907066345,
      "learning_rate": 8e-05,
      "loss": 1.5513,
      "step": 3848
    },
    {
      "epoch": 0.7993769470404984,
      "grad_norm": 0.5956964492797852,
      "learning_rate": 8e-05,
      "loss": 1.5628,
      "step": 3849
    },
    {
      "epoch": 0.7995846313603323,
      "grad_norm": 0.5426408648490906,
      "learning_rate": 8e-05,
      "loss": 1.5458,
      "step": 3850
    },
    {
      "epoch": 0.7997923156801662,
      "grad_norm": 0.5439097285270691,
      "learning_rate": 8e-05,
      "loss": 1.5257,
      "step": 3851
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.5443709492683411,
      "learning_rate": 8e-05,
      "loss": 1.4881,
      "step": 3852
    },
    {
      "epoch": 0.8002076843198338,
      "grad_norm": 0.5784890651702881,
      "learning_rate": 8e-05,
      "loss": 1.4945,
      "step": 3853
    },
    {
      "epoch": 0.8004153686396677,
      "grad_norm": 0.5435579419136047,
      "learning_rate": 8e-05,
      "loss": 1.5352,
      "step": 3854
    },
    {
      "epoch": 0.8006230529595015,
      "grad_norm": 0.5404057502746582,
      "learning_rate": 8e-05,
      "loss": 1.5313,
      "step": 3855
    },
    {
      "epoch": 0.8008307372793354,
      "grad_norm": 0.5357280373573303,
      "learning_rate": 8e-05,
      "loss": 1.5272,
      "step": 3856
    },
    {
      "epoch": 0.8010384215991693,
      "grad_norm": 0.5426300764083862,
      "learning_rate": 8e-05,
      "loss": 1.5312,
      "step": 3857
    },
    {
      "epoch": 0.8012461059190031,
      "grad_norm": 0.5688260197639465,
      "learning_rate": 8e-05,
      "loss": 1.5624,
      "step": 3858
    },
    {
      "epoch": 0.801453790238837,
      "grad_norm": 0.5829392075538635,
      "learning_rate": 8e-05,
      "loss": 1.5552,
      "step": 3859
    },
    {
      "epoch": 0.8016614745586709,
      "grad_norm": 0.5540609359741211,
      "learning_rate": 8e-05,
      "loss": 1.4807,
      "step": 3860
    },
    {
      "epoch": 0.8018691588785046,
      "grad_norm": 0.5700802803039551,
      "learning_rate": 8e-05,
      "loss": 1.5688,
      "step": 3861
    },
    {
      "epoch": 0.8020768431983385,
      "grad_norm": 0.5573641657829285,
      "learning_rate": 8e-05,
      "loss": 1.5094,
      "step": 3862
    },
    {
      "epoch": 0.8022845275181724,
      "grad_norm": 0.5363988280296326,
      "learning_rate": 8e-05,
      "loss": 1.5671,
      "step": 3863
    },
    {
      "epoch": 0.8024922118380062,
      "grad_norm": 0.5173343420028687,
      "learning_rate": 8e-05,
      "loss": 1.4531,
      "step": 3864
    },
    {
      "epoch": 0.8026998961578401,
      "grad_norm": 0.49960580468177795,
      "learning_rate": 8e-05,
      "loss": 1.4694,
      "step": 3865
    },
    {
      "epoch": 0.802907580477674,
      "grad_norm": 0.51841801404953,
      "learning_rate": 8e-05,
      "loss": 1.4853,
      "step": 3866
    },
    {
      "epoch": 0.8031152647975078,
      "grad_norm": 0.5884705185890198,
      "learning_rate": 8e-05,
      "loss": 1.5636,
      "step": 3867
    },
    {
      "epoch": 0.8033229491173416,
      "grad_norm": 0.5620328187942505,
      "learning_rate": 8e-05,
      "loss": 1.5385,
      "step": 3868
    },
    {
      "epoch": 0.8035306334371755,
      "grad_norm": 0.5107353329658508,
      "learning_rate": 8e-05,
      "loss": 1.5377,
      "step": 3869
    },
    {
      "epoch": 0.8037383177570093,
      "grad_norm": 0.5952506065368652,
      "learning_rate": 8e-05,
      "loss": 1.6012,
      "step": 3870
    },
    {
      "epoch": 0.8039460020768432,
      "grad_norm": 0.49518468976020813,
      "learning_rate": 8e-05,
      "loss": 1.5132,
      "step": 3871
    },
    {
      "epoch": 0.8041536863966771,
      "grad_norm": 0.518493115901947,
      "learning_rate": 8e-05,
      "loss": 1.448,
      "step": 3872
    },
    {
      "epoch": 0.8043613707165109,
      "grad_norm": 0.5589504837989807,
      "learning_rate": 8e-05,
      "loss": 1.587,
      "step": 3873
    },
    {
      "epoch": 0.8045690550363448,
      "grad_norm": 0.5282282829284668,
      "learning_rate": 8e-05,
      "loss": 1.5367,
      "step": 3874
    },
    {
      "epoch": 0.8047767393561787,
      "grad_norm": 0.5434615612030029,
      "learning_rate": 8e-05,
      "loss": 1.4948,
      "step": 3875
    },
    {
      "epoch": 0.8049844236760124,
      "grad_norm": 0.6129417419433594,
      "learning_rate": 8e-05,
      "loss": 1.5627,
      "step": 3876
    },
    {
      "epoch": 0.8051921079958463,
      "grad_norm": 0.5177566409111023,
      "learning_rate": 8e-05,
      "loss": 1.5175,
      "step": 3877
    },
    {
      "epoch": 0.8053997923156802,
      "grad_norm": 0.55455482006073,
      "learning_rate": 8e-05,
      "loss": 1.5278,
      "step": 3878
    },
    {
      "epoch": 0.805607476635514,
      "grad_norm": 0.5107983350753784,
      "learning_rate": 8e-05,
      "loss": 1.5035,
      "step": 3879
    },
    {
      "epoch": 0.8058151609553479,
      "grad_norm": 0.5506031513214111,
      "learning_rate": 8e-05,
      "loss": 1.6251,
      "step": 3880
    },
    {
      "epoch": 0.8060228452751818,
      "grad_norm": 0.5219122767448425,
      "learning_rate": 8e-05,
      "loss": 1.5214,
      "step": 3881
    },
    {
      "epoch": 0.8062305295950156,
      "grad_norm": 0.512962281703949,
      "learning_rate": 8e-05,
      "loss": 1.4916,
      "step": 3882
    },
    {
      "epoch": 0.8064382139148494,
      "grad_norm": 0.5244505405426025,
      "learning_rate": 8e-05,
      "loss": 1.5047,
      "step": 3883
    },
    {
      "epoch": 0.8066458982346832,
      "grad_norm": 0.531764566898346,
      "learning_rate": 8e-05,
      "loss": 1.6211,
      "step": 3884
    },
    {
      "epoch": 0.8068535825545171,
      "grad_norm": 0.5583875775337219,
      "learning_rate": 8e-05,
      "loss": 1.5204,
      "step": 3885
    },
    {
      "epoch": 0.807061266874351,
      "grad_norm": 0.5349747538566589,
      "learning_rate": 8e-05,
      "loss": 1.5784,
      "step": 3886
    },
    {
      "epoch": 0.8072689511941848,
      "grad_norm": 0.550213634967804,
      "learning_rate": 8e-05,
      "loss": 1.5372,
      "step": 3887
    },
    {
      "epoch": 0.8074766355140187,
      "grad_norm": 0.565680980682373,
      "learning_rate": 8e-05,
      "loss": 1.6243,
      "step": 3888
    },
    {
      "epoch": 0.8076843198338526,
      "grad_norm": 0.5179294347763062,
      "learning_rate": 8e-05,
      "loss": 1.4506,
      "step": 3889
    },
    {
      "epoch": 0.8078920041536864,
      "grad_norm": 0.5388593673706055,
      "learning_rate": 8e-05,
      "loss": 1.4828,
      "step": 3890
    },
    {
      "epoch": 0.8080996884735202,
      "grad_norm": 0.6075422763824463,
      "learning_rate": 8e-05,
      "loss": 1.5359,
      "step": 3891
    },
    {
      "epoch": 0.8083073727933541,
      "grad_norm": 0.5533674359321594,
      "learning_rate": 8e-05,
      "loss": 1.5306,
      "step": 3892
    },
    {
      "epoch": 0.8085150571131879,
      "grad_norm": 0.5397309064865112,
      "learning_rate": 8e-05,
      "loss": 1.6229,
      "step": 3893
    },
    {
      "epoch": 0.8087227414330218,
      "grad_norm": 0.5120832324028015,
      "learning_rate": 8e-05,
      "loss": 1.5472,
      "step": 3894
    },
    {
      "epoch": 0.8089304257528557,
      "grad_norm": 0.5653396248817444,
      "learning_rate": 8e-05,
      "loss": 1.539,
      "step": 3895
    },
    {
      "epoch": 0.8091381100726895,
      "grad_norm": 0.5369875431060791,
      "learning_rate": 8e-05,
      "loss": 1.5268,
      "step": 3896
    },
    {
      "epoch": 0.8093457943925234,
      "grad_norm": 0.5352119207382202,
      "learning_rate": 8e-05,
      "loss": 1.5817,
      "step": 3897
    },
    {
      "epoch": 0.8095534787123572,
      "grad_norm": 0.5234546065330505,
      "learning_rate": 8e-05,
      "loss": 1.4817,
      "step": 3898
    },
    {
      "epoch": 0.809761163032191,
      "grad_norm": 0.5159884095191956,
      "learning_rate": 8e-05,
      "loss": 1.522,
      "step": 3899
    },
    {
      "epoch": 0.8099688473520249,
      "grad_norm": 0.5263754725456238,
      "learning_rate": 8e-05,
      "loss": 1.5193,
      "step": 3900
    },
    {
      "epoch": 0.8101765316718588,
      "grad_norm": 0.540711522102356,
      "learning_rate": 8e-05,
      "loss": 1.5111,
      "step": 3901
    },
    {
      "epoch": 0.8103842159916926,
      "grad_norm": 0.5038042664527893,
      "learning_rate": 8e-05,
      "loss": 1.5555,
      "step": 3902
    },
    {
      "epoch": 0.8105919003115265,
      "grad_norm": 0.5215405821800232,
      "learning_rate": 8e-05,
      "loss": 1.4944,
      "step": 3903
    },
    {
      "epoch": 0.8107995846313604,
      "grad_norm": 0.5543336272239685,
      "learning_rate": 8e-05,
      "loss": 1.4342,
      "step": 3904
    },
    {
      "epoch": 0.8110072689511942,
      "grad_norm": 0.5466092824935913,
      "learning_rate": 8e-05,
      "loss": 1.5057,
      "step": 3905
    },
    {
      "epoch": 0.811214953271028,
      "grad_norm": 0.5810185074806213,
      "learning_rate": 8e-05,
      "loss": 1.534,
      "step": 3906
    },
    {
      "epoch": 0.8114226375908619,
      "grad_norm": 0.5618957877159119,
      "learning_rate": 8e-05,
      "loss": 1.4649,
      "step": 3907
    },
    {
      "epoch": 0.8116303219106957,
      "grad_norm": 0.5785271525382996,
      "learning_rate": 8e-05,
      "loss": 1.55,
      "step": 3908
    },
    {
      "epoch": 0.8118380062305296,
      "grad_norm": 0.554823637008667,
      "learning_rate": 8e-05,
      "loss": 1.557,
      "step": 3909
    },
    {
      "epoch": 0.8120456905503635,
      "grad_norm": 0.5068426132202148,
      "learning_rate": 8e-05,
      "loss": 1.4651,
      "step": 3910
    },
    {
      "epoch": 0.8122533748701973,
      "grad_norm": 0.5867827534675598,
      "learning_rate": 8e-05,
      "loss": 1.5452,
      "step": 3911
    },
    {
      "epoch": 0.8124610591900312,
      "grad_norm": 0.5372866988182068,
      "learning_rate": 8e-05,
      "loss": 1.4772,
      "step": 3912
    },
    {
      "epoch": 0.812668743509865,
      "grad_norm": 0.5329793095588684,
      "learning_rate": 8e-05,
      "loss": 1.4761,
      "step": 3913
    },
    {
      "epoch": 0.8128764278296988,
      "grad_norm": 0.5528751611709595,
      "learning_rate": 8e-05,
      "loss": 1.5436,
      "step": 3914
    },
    {
      "epoch": 0.8130841121495327,
      "grad_norm": 0.5616529583930969,
      "learning_rate": 8e-05,
      "loss": 1.5041,
      "step": 3915
    },
    {
      "epoch": 0.8132917964693666,
      "grad_norm": 0.5282413363456726,
      "learning_rate": 8e-05,
      "loss": 1.5101,
      "step": 3916
    },
    {
      "epoch": 0.8134994807892004,
      "grad_norm": 0.5532453060150146,
      "learning_rate": 8e-05,
      "loss": 1.5149,
      "step": 3917
    },
    {
      "epoch": 0.8137071651090343,
      "grad_norm": 0.5570338368415833,
      "learning_rate": 8e-05,
      "loss": 1.5577,
      "step": 3918
    },
    {
      "epoch": 0.8139148494288682,
      "grad_norm": 0.5608822107315063,
      "learning_rate": 8e-05,
      "loss": 1.5488,
      "step": 3919
    },
    {
      "epoch": 0.814122533748702,
      "grad_norm": 0.5342090725898743,
      "learning_rate": 8e-05,
      "loss": 1.5654,
      "step": 3920
    },
    {
      "epoch": 0.8143302180685358,
      "grad_norm": 0.5562800765037537,
      "learning_rate": 8e-05,
      "loss": 1.4966,
      "step": 3921
    },
    {
      "epoch": 0.8145379023883697,
      "grad_norm": 0.5398460626602173,
      "learning_rate": 8e-05,
      "loss": 1.4866,
      "step": 3922
    },
    {
      "epoch": 0.8147455867082035,
      "grad_norm": 0.5428885817527771,
      "learning_rate": 8e-05,
      "loss": 1.6067,
      "step": 3923
    },
    {
      "epoch": 0.8149532710280374,
      "grad_norm": 0.5578614473342896,
      "learning_rate": 8e-05,
      "loss": 1.5412,
      "step": 3924
    },
    {
      "epoch": 0.8151609553478713,
      "grad_norm": 0.5321469902992249,
      "learning_rate": 8e-05,
      "loss": 1.4906,
      "step": 3925
    },
    {
      "epoch": 0.8153686396677051,
      "grad_norm": 0.5549134612083435,
      "learning_rate": 8e-05,
      "loss": 1.5318,
      "step": 3926
    },
    {
      "epoch": 0.815576323987539,
      "grad_norm": 0.5271235704421997,
      "learning_rate": 8e-05,
      "loss": 1.4706,
      "step": 3927
    },
    {
      "epoch": 0.8157840083073727,
      "grad_norm": 0.5550113320350647,
      "learning_rate": 8e-05,
      "loss": 1.5441,
      "step": 3928
    },
    {
      "epoch": 0.8159916926272066,
      "grad_norm": 0.5361166000366211,
      "learning_rate": 8e-05,
      "loss": 1.5285,
      "step": 3929
    },
    {
      "epoch": 0.8161993769470405,
      "grad_norm": 0.5583904981613159,
      "learning_rate": 8e-05,
      "loss": 1.5635,
      "step": 3930
    },
    {
      "epoch": 0.8164070612668743,
      "grad_norm": 0.5698258876800537,
      "learning_rate": 8e-05,
      "loss": 1.5428,
      "step": 3931
    },
    {
      "epoch": 0.8166147455867082,
      "grad_norm": 0.5614790916442871,
      "learning_rate": 8e-05,
      "loss": 1.4356,
      "step": 3932
    },
    {
      "epoch": 0.8168224299065421,
      "grad_norm": 0.5800708532333374,
      "learning_rate": 8e-05,
      "loss": 1.6096,
      "step": 3933
    },
    {
      "epoch": 0.817030114226376,
      "grad_norm": 0.5604009628295898,
      "learning_rate": 8e-05,
      "loss": 1.5978,
      "step": 3934
    },
    {
      "epoch": 0.8172377985462098,
      "grad_norm": 0.5288925766944885,
      "learning_rate": 8e-05,
      "loss": 1.5725,
      "step": 3935
    },
    {
      "epoch": 0.8174454828660436,
      "grad_norm": 0.5271170735359192,
      "learning_rate": 8e-05,
      "loss": 1.4998,
      "step": 3936
    },
    {
      "epoch": 0.8176531671858774,
      "grad_norm": 0.5243780612945557,
      "learning_rate": 8e-05,
      "loss": 1.4966,
      "step": 3937
    },
    {
      "epoch": 0.8178608515057113,
      "grad_norm": 0.6347025632858276,
      "learning_rate": 8e-05,
      "loss": 1.4989,
      "step": 3938
    },
    {
      "epoch": 0.8180685358255452,
      "grad_norm": 0.5280063152313232,
      "learning_rate": 8e-05,
      "loss": 1.4773,
      "step": 3939
    },
    {
      "epoch": 0.818276220145379,
      "grad_norm": 0.5614491701126099,
      "learning_rate": 8e-05,
      "loss": 1.5039,
      "step": 3940
    },
    {
      "epoch": 0.8184839044652129,
      "grad_norm": 0.5759320259094238,
      "learning_rate": 8e-05,
      "loss": 1.5506,
      "step": 3941
    },
    {
      "epoch": 0.8186915887850468,
      "grad_norm": 0.5403061509132385,
      "learning_rate": 8e-05,
      "loss": 1.5672,
      "step": 3942
    },
    {
      "epoch": 0.8188992731048805,
      "grad_norm": 0.5220894813537598,
      "learning_rate": 8e-05,
      "loss": 1.5138,
      "step": 3943
    },
    {
      "epoch": 0.8191069574247144,
      "grad_norm": 0.5523865222930908,
      "learning_rate": 8e-05,
      "loss": 1.5213,
      "step": 3944
    },
    {
      "epoch": 0.8193146417445483,
      "grad_norm": 0.5424647927284241,
      "learning_rate": 8e-05,
      "loss": 1.5696,
      "step": 3945
    },
    {
      "epoch": 0.8195223260643821,
      "grad_norm": 0.5446579456329346,
      "learning_rate": 8e-05,
      "loss": 1.5575,
      "step": 3946
    },
    {
      "epoch": 0.819730010384216,
      "grad_norm": 0.5582678914070129,
      "learning_rate": 8e-05,
      "loss": 1.5539,
      "step": 3947
    },
    {
      "epoch": 0.8199376947040499,
      "grad_norm": 0.5966761708259583,
      "learning_rate": 8e-05,
      "loss": 1.6009,
      "step": 3948
    },
    {
      "epoch": 0.8201453790238837,
      "grad_norm": 0.524156928062439,
      "learning_rate": 8e-05,
      "loss": 1.4805,
      "step": 3949
    },
    {
      "epoch": 0.8203530633437176,
      "grad_norm": 0.540702223777771,
      "learning_rate": 8e-05,
      "loss": 1.5251,
      "step": 3950
    },
    {
      "epoch": 0.8205607476635514,
      "grad_norm": 0.5291163325309753,
      "learning_rate": 8e-05,
      "loss": 1.5485,
      "step": 3951
    },
    {
      "epoch": 0.8207684319833852,
      "grad_norm": 0.5278533697128296,
      "learning_rate": 8e-05,
      "loss": 1.4565,
      "step": 3952
    },
    {
      "epoch": 0.8209761163032191,
      "grad_norm": 0.5291278958320618,
      "learning_rate": 8e-05,
      "loss": 1.5018,
      "step": 3953
    },
    {
      "epoch": 0.821183800623053,
      "grad_norm": 0.5107881426811218,
      "learning_rate": 8e-05,
      "loss": 1.5047,
      "step": 3954
    },
    {
      "epoch": 0.8213914849428868,
      "grad_norm": 0.5312121510505676,
      "learning_rate": 8e-05,
      "loss": 1.5325,
      "step": 3955
    },
    {
      "epoch": 0.8215991692627207,
      "grad_norm": 0.5497841835021973,
      "learning_rate": 8e-05,
      "loss": 1.5908,
      "step": 3956
    },
    {
      "epoch": 0.8218068535825546,
      "grad_norm": 0.5752426385879517,
      "learning_rate": 8e-05,
      "loss": 1.436,
      "step": 3957
    },
    {
      "epoch": 0.8220145379023883,
      "grad_norm": 0.5783313512802124,
      "learning_rate": 8e-05,
      "loss": 1.57,
      "step": 3958
    },
    {
      "epoch": 0.8222222222222222,
      "grad_norm": 0.5411394238471985,
      "learning_rate": 8e-05,
      "loss": 1.5915,
      "step": 3959
    },
    {
      "epoch": 0.822429906542056,
      "grad_norm": 0.5626519918441772,
      "learning_rate": 8e-05,
      "loss": 1.5643,
      "step": 3960
    },
    {
      "epoch": 0.8226375908618899,
      "grad_norm": 0.546779215335846,
      "learning_rate": 8e-05,
      "loss": 1.4926,
      "step": 3961
    },
    {
      "epoch": 0.8228452751817238,
      "grad_norm": 0.5330975651741028,
      "learning_rate": 8e-05,
      "loss": 1.5392,
      "step": 3962
    },
    {
      "epoch": 0.8230529595015577,
      "grad_norm": 0.5250078439712524,
      "learning_rate": 8e-05,
      "loss": 1.4762,
      "step": 3963
    },
    {
      "epoch": 0.8232606438213915,
      "grad_norm": 0.530977189540863,
      "learning_rate": 8e-05,
      "loss": 1.4968,
      "step": 3964
    },
    {
      "epoch": 0.8234683281412254,
      "grad_norm": 0.5293608903884888,
      "learning_rate": 8e-05,
      "loss": 1.476,
      "step": 3965
    },
    {
      "epoch": 0.8236760124610591,
      "grad_norm": 0.5713688135147095,
      "learning_rate": 8e-05,
      "loss": 1.4945,
      "step": 3966
    },
    {
      "epoch": 0.823883696780893,
      "grad_norm": 0.5358543395996094,
      "learning_rate": 8e-05,
      "loss": 1.5153,
      "step": 3967
    },
    {
      "epoch": 0.8240913811007269,
      "grad_norm": 0.5338706970214844,
      "learning_rate": 8e-05,
      "loss": 1.5646,
      "step": 3968
    },
    {
      "epoch": 0.8242990654205608,
      "grad_norm": 0.6816396713256836,
      "learning_rate": 8e-05,
      "loss": 1.4351,
      "step": 3969
    },
    {
      "epoch": 0.8245067497403946,
      "grad_norm": 0.537402331829071,
      "learning_rate": 8e-05,
      "loss": 1.5386,
      "step": 3970
    },
    {
      "epoch": 0.8247144340602285,
      "grad_norm": 0.5510597825050354,
      "learning_rate": 8e-05,
      "loss": 1.5877,
      "step": 3971
    },
    {
      "epoch": 0.8249221183800624,
      "grad_norm": 0.571393609046936,
      "learning_rate": 8e-05,
      "loss": 1.5536,
      "step": 3972
    },
    {
      "epoch": 0.8251298026998961,
      "grad_norm": 0.5510338544845581,
      "learning_rate": 8e-05,
      "loss": 1.5102,
      "step": 3973
    },
    {
      "epoch": 0.82533748701973,
      "grad_norm": 0.552266538143158,
      "learning_rate": 8e-05,
      "loss": 1.5868,
      "step": 3974
    },
    {
      "epoch": 0.8255451713395638,
      "grad_norm": 0.5477768778800964,
      "learning_rate": 8e-05,
      "loss": 1.5652,
      "step": 3975
    },
    {
      "epoch": 0.8257528556593977,
      "grad_norm": 0.5507093071937561,
      "learning_rate": 8e-05,
      "loss": 1.4214,
      "step": 3976
    },
    {
      "epoch": 0.8259605399792316,
      "grad_norm": 0.5357552766799927,
      "learning_rate": 8e-05,
      "loss": 1.5058,
      "step": 3977
    },
    {
      "epoch": 0.8261682242990654,
      "grad_norm": 0.522476315498352,
      "learning_rate": 8e-05,
      "loss": 1.5471,
      "step": 3978
    },
    {
      "epoch": 0.8263759086188993,
      "grad_norm": 0.6307571530342102,
      "learning_rate": 8e-05,
      "loss": 1.6253,
      "step": 3979
    },
    {
      "epoch": 0.8265835929387332,
      "grad_norm": 0.5378223657608032,
      "learning_rate": 8e-05,
      "loss": 1.5494,
      "step": 3980
    },
    {
      "epoch": 0.8267912772585669,
      "grad_norm": 0.5201992988586426,
      "learning_rate": 8e-05,
      "loss": 1.5177,
      "step": 3981
    },
    {
      "epoch": 0.8269989615784008,
      "grad_norm": 0.5398663878440857,
      "learning_rate": 8e-05,
      "loss": 1.5006,
      "step": 3982
    },
    {
      "epoch": 0.8272066458982347,
      "grad_norm": 0.5968499779701233,
      "learning_rate": 8e-05,
      "loss": 1.5273,
      "step": 3983
    },
    {
      "epoch": 0.8274143302180685,
      "grad_norm": 0.5356617569923401,
      "learning_rate": 8e-05,
      "loss": 1.5076,
      "step": 3984
    },
    {
      "epoch": 0.8276220145379024,
      "grad_norm": 0.5770037770271301,
      "learning_rate": 8e-05,
      "loss": 1.5679,
      "step": 3985
    },
    {
      "epoch": 0.8278296988577363,
      "grad_norm": 0.5473908185958862,
      "learning_rate": 8e-05,
      "loss": 1.5546,
      "step": 3986
    },
    {
      "epoch": 0.8280373831775701,
      "grad_norm": 0.5516231060028076,
      "learning_rate": 8e-05,
      "loss": 1.5228,
      "step": 3987
    },
    {
      "epoch": 0.8282450674974039,
      "grad_norm": 0.5564890503883362,
      "learning_rate": 8e-05,
      "loss": 1.5043,
      "step": 3988
    },
    {
      "epoch": 0.8284527518172378,
      "grad_norm": 0.5442187190055847,
      "learning_rate": 8e-05,
      "loss": 1.4937,
      "step": 3989
    },
    {
      "epoch": 0.8286604361370716,
      "grad_norm": 0.5638508796691895,
      "learning_rate": 8e-05,
      "loss": 1.5975,
      "step": 3990
    },
    {
      "epoch": 0.8288681204569055,
      "grad_norm": 0.5704795122146606,
      "learning_rate": 8e-05,
      "loss": 1.4575,
      "step": 3991
    },
    {
      "epoch": 0.8290758047767394,
      "grad_norm": 0.5389898419380188,
      "learning_rate": 8e-05,
      "loss": 1.4632,
      "step": 3992
    },
    {
      "epoch": 0.8292834890965732,
      "grad_norm": 0.5508019924163818,
      "learning_rate": 8e-05,
      "loss": 1.5582,
      "step": 3993
    },
    {
      "epoch": 0.8294911734164071,
      "grad_norm": 0.574353039264679,
      "learning_rate": 8e-05,
      "loss": 1.5927,
      "step": 3994
    },
    {
      "epoch": 0.829698857736241,
      "grad_norm": 0.5273694396018982,
      "learning_rate": 8e-05,
      "loss": 1.5911,
      "step": 3995
    },
    {
      "epoch": 0.8299065420560747,
      "grad_norm": 0.5783042907714844,
      "learning_rate": 8e-05,
      "loss": 1.5944,
      "step": 3996
    },
    {
      "epoch": 0.8301142263759086,
      "grad_norm": 0.5857273936271667,
      "learning_rate": 8e-05,
      "loss": 1.5803,
      "step": 3997
    },
    {
      "epoch": 0.8303219106957425,
      "grad_norm": 0.5762231349945068,
      "learning_rate": 8e-05,
      "loss": 1.5302,
      "step": 3998
    },
    {
      "epoch": 0.8305295950155763,
      "grad_norm": 0.5323285460472107,
      "learning_rate": 8e-05,
      "loss": 1.4999,
      "step": 3999
    },
    {
      "epoch": 0.8307372793354102,
      "grad_norm": 0.6111394762992859,
      "learning_rate": 8e-05,
      "loss": 1.536,
      "step": 4000
    },
    {
      "epoch": 0.8309449636552441,
      "grad_norm": 0.5615333914756775,
      "learning_rate": 8e-05,
      "loss": 1.5163,
      "step": 4001
    },
    {
      "epoch": 0.8311526479750779,
      "grad_norm": 0.5138621926307678,
      "learning_rate": 8e-05,
      "loss": 1.5611,
      "step": 4002
    },
    {
      "epoch": 0.8313603322949117,
      "grad_norm": 0.5488884449005127,
      "learning_rate": 8e-05,
      "loss": 1.5781,
      "step": 4003
    },
    {
      "epoch": 0.8315680166147456,
      "grad_norm": 0.5374660491943359,
      "learning_rate": 8e-05,
      "loss": 1.5182,
      "step": 4004
    },
    {
      "epoch": 0.8317757009345794,
      "grad_norm": 0.509480893611908,
      "learning_rate": 8e-05,
      "loss": 1.5279,
      "step": 4005
    },
    {
      "epoch": 0.8319833852544133,
      "grad_norm": 0.5221795439720154,
      "learning_rate": 8e-05,
      "loss": 1.5139,
      "step": 4006
    },
    {
      "epoch": 0.8321910695742472,
      "grad_norm": 0.5480198860168457,
      "learning_rate": 8e-05,
      "loss": 1.565,
      "step": 4007
    },
    {
      "epoch": 0.832398753894081,
      "grad_norm": 0.561700701713562,
      "learning_rate": 8e-05,
      "loss": 1.5546,
      "step": 4008
    },
    {
      "epoch": 0.8326064382139149,
      "grad_norm": 0.5475932359695435,
      "learning_rate": 8e-05,
      "loss": 1.5772,
      "step": 4009
    },
    {
      "epoch": 0.8328141225337488,
      "grad_norm": 0.5535194277763367,
      "learning_rate": 8e-05,
      "loss": 1.6336,
      "step": 4010
    },
    {
      "epoch": 0.8330218068535825,
      "grad_norm": 0.5214248895645142,
      "learning_rate": 8e-05,
      "loss": 1.5448,
      "step": 4011
    },
    {
      "epoch": 0.8332294911734164,
      "grad_norm": 0.5418645143508911,
      "learning_rate": 8e-05,
      "loss": 1.5457,
      "step": 4012
    },
    {
      "epoch": 0.8334371754932502,
      "grad_norm": 0.5370267033576965,
      "learning_rate": 8e-05,
      "loss": 1.5454,
      "step": 4013
    },
    {
      "epoch": 0.8336448598130841,
      "grad_norm": 0.5209023952484131,
      "learning_rate": 8e-05,
      "loss": 1.461,
      "step": 4014
    },
    {
      "epoch": 0.833852544132918,
      "grad_norm": 0.5714237093925476,
      "learning_rate": 8e-05,
      "loss": 1.502,
      "step": 4015
    },
    {
      "epoch": 0.8340602284527519,
      "grad_norm": 0.5499559640884399,
      "learning_rate": 8e-05,
      "loss": 1.5103,
      "step": 4016
    },
    {
      "epoch": 0.8342679127725857,
      "grad_norm": 0.5502960681915283,
      "learning_rate": 8e-05,
      "loss": 1.5795,
      "step": 4017
    },
    {
      "epoch": 0.8344755970924195,
      "grad_norm": 0.5501389503479004,
      "learning_rate": 8e-05,
      "loss": 1.539,
      "step": 4018
    },
    {
      "epoch": 0.8346832814122533,
      "grad_norm": 0.6214225888252258,
      "learning_rate": 8e-05,
      "loss": 1.5297,
      "step": 4019
    },
    {
      "epoch": 0.8348909657320872,
      "grad_norm": 0.5427936911582947,
      "learning_rate": 8e-05,
      "loss": 1.5851,
      "step": 4020
    },
    {
      "epoch": 0.8350986500519211,
      "grad_norm": 0.5430470705032349,
      "learning_rate": 8e-05,
      "loss": 1.5527,
      "step": 4021
    },
    {
      "epoch": 0.8353063343717549,
      "grad_norm": 0.5648637413978577,
      "learning_rate": 8e-05,
      "loss": 1.5856,
      "step": 4022
    },
    {
      "epoch": 0.8355140186915888,
      "grad_norm": 0.5404366850852966,
      "learning_rate": 8e-05,
      "loss": 1.5243,
      "step": 4023
    },
    {
      "epoch": 0.8357217030114227,
      "grad_norm": 0.563101589679718,
      "learning_rate": 8e-05,
      "loss": 1.4633,
      "step": 4024
    },
    {
      "epoch": 0.8359293873312564,
      "grad_norm": 0.5742359757423401,
      "learning_rate": 8e-05,
      "loss": 1.5419,
      "step": 4025
    },
    {
      "epoch": 0.8361370716510903,
      "grad_norm": 0.5941342115402222,
      "learning_rate": 8e-05,
      "loss": 1.6013,
      "step": 4026
    },
    {
      "epoch": 0.8363447559709242,
      "grad_norm": 0.563202977180481,
      "learning_rate": 8e-05,
      "loss": 1.5644,
      "step": 4027
    },
    {
      "epoch": 0.836552440290758,
      "grad_norm": 0.5732769966125488,
      "learning_rate": 8e-05,
      "loss": 1.5705,
      "step": 4028
    },
    {
      "epoch": 0.8367601246105919,
      "grad_norm": 0.547235369682312,
      "learning_rate": 8e-05,
      "loss": 1.5609,
      "step": 4029
    },
    {
      "epoch": 0.8369678089304258,
      "grad_norm": 0.5238963961601257,
      "learning_rate": 8e-05,
      "loss": 1.5115,
      "step": 4030
    },
    {
      "epoch": 0.8371754932502596,
      "grad_norm": 0.5442341566085815,
      "learning_rate": 8e-05,
      "loss": 1.5145,
      "step": 4031
    },
    {
      "epoch": 0.8373831775700935,
      "grad_norm": 0.5489323139190674,
      "learning_rate": 8e-05,
      "loss": 1.4579,
      "step": 4032
    },
    {
      "epoch": 0.8375908618899273,
      "grad_norm": 0.5066717863082886,
      "learning_rate": 8e-05,
      "loss": 1.4825,
      "step": 4033
    },
    {
      "epoch": 0.8377985462097611,
      "grad_norm": 0.5124364495277405,
      "learning_rate": 8e-05,
      "loss": 1.5355,
      "step": 4034
    },
    {
      "epoch": 0.838006230529595,
      "grad_norm": 0.5398315787315369,
      "learning_rate": 8e-05,
      "loss": 1.4965,
      "step": 4035
    },
    {
      "epoch": 0.8382139148494289,
      "grad_norm": 0.5319242477416992,
      "learning_rate": 8e-05,
      "loss": 1.5422,
      "step": 4036
    },
    {
      "epoch": 0.8384215991692627,
      "grad_norm": 0.5388317704200745,
      "learning_rate": 8e-05,
      "loss": 1.5583,
      "step": 4037
    },
    {
      "epoch": 0.8386292834890966,
      "grad_norm": 0.5179262757301331,
      "learning_rate": 8e-05,
      "loss": 1.5398,
      "step": 4038
    },
    {
      "epoch": 0.8388369678089305,
      "grad_norm": 0.5979048013687134,
      "learning_rate": 8e-05,
      "loss": 1.507,
      "step": 4039
    },
    {
      "epoch": 0.8390446521287642,
      "grad_norm": 0.531641960144043,
      "learning_rate": 8e-05,
      "loss": 1.6137,
      "step": 4040
    },
    {
      "epoch": 0.8392523364485981,
      "grad_norm": 0.5186541676521301,
      "learning_rate": 8e-05,
      "loss": 1.5281,
      "step": 4041
    },
    {
      "epoch": 0.839460020768432,
      "grad_norm": 0.5275489091873169,
      "learning_rate": 8e-05,
      "loss": 1.5155,
      "step": 4042
    },
    {
      "epoch": 0.8396677050882658,
      "grad_norm": 0.5519757866859436,
      "learning_rate": 8e-05,
      "loss": 1.5466,
      "step": 4043
    },
    {
      "epoch": 0.8398753894080997,
      "grad_norm": 0.5254910588264465,
      "learning_rate": 8e-05,
      "loss": 1.526,
      "step": 4044
    },
    {
      "epoch": 0.8400830737279336,
      "grad_norm": 0.5400928258895874,
      "learning_rate": 8e-05,
      "loss": 1.5047,
      "step": 4045
    },
    {
      "epoch": 0.8402907580477674,
      "grad_norm": 0.5544579029083252,
      "learning_rate": 8e-05,
      "loss": 1.4869,
      "step": 4046
    },
    {
      "epoch": 0.8404984423676013,
      "grad_norm": 0.5367680191993713,
      "learning_rate": 8e-05,
      "loss": 1.5418,
      "step": 4047
    },
    {
      "epoch": 0.840706126687435,
      "grad_norm": 0.5273227095603943,
      "learning_rate": 8e-05,
      "loss": 1.5576,
      "step": 4048
    },
    {
      "epoch": 0.8409138110072689,
      "grad_norm": 0.5547626614570618,
      "learning_rate": 8e-05,
      "loss": 1.5993,
      "step": 4049
    },
    {
      "epoch": 0.8411214953271028,
      "grad_norm": 0.5311830043792725,
      "learning_rate": 8e-05,
      "loss": 1.5481,
      "step": 4050
    },
    {
      "epoch": 0.8413291796469367,
      "grad_norm": 0.5458120703697205,
      "learning_rate": 8e-05,
      "loss": 1.5601,
      "step": 4051
    },
    {
      "epoch": 0.8415368639667705,
      "grad_norm": 0.5379188060760498,
      "learning_rate": 8e-05,
      "loss": 1.4611,
      "step": 4052
    },
    {
      "epoch": 0.8417445482866044,
      "grad_norm": 0.5371125936508179,
      "learning_rate": 8e-05,
      "loss": 1.5203,
      "step": 4053
    },
    {
      "epoch": 0.8419522326064383,
      "grad_norm": 0.5373204350471497,
      "learning_rate": 8e-05,
      "loss": 1.5543,
      "step": 4054
    },
    {
      "epoch": 0.842159916926272,
      "grad_norm": 0.5394884943962097,
      "learning_rate": 8e-05,
      "loss": 1.523,
      "step": 4055
    },
    {
      "epoch": 0.8423676012461059,
      "grad_norm": 0.5670446753501892,
      "learning_rate": 8e-05,
      "loss": 1.59,
      "step": 4056
    },
    {
      "epoch": 0.8425752855659397,
      "grad_norm": 0.5369330644607544,
      "learning_rate": 8e-05,
      "loss": 1.47,
      "step": 4057
    },
    {
      "epoch": 0.8427829698857736,
      "grad_norm": 0.5214245915412903,
      "learning_rate": 8e-05,
      "loss": 1.4661,
      "step": 4058
    },
    {
      "epoch": 0.8429906542056075,
      "grad_norm": 0.5466499328613281,
      "learning_rate": 8e-05,
      "loss": 1.4723,
      "step": 4059
    },
    {
      "epoch": 0.8431983385254413,
      "grad_norm": 0.5338225960731506,
      "learning_rate": 8e-05,
      "loss": 1.5003,
      "step": 4060
    },
    {
      "epoch": 0.8434060228452752,
      "grad_norm": 0.5210661292076111,
      "learning_rate": 8e-05,
      "loss": 1.5034,
      "step": 4061
    },
    {
      "epoch": 0.8436137071651091,
      "grad_norm": 0.5492287278175354,
      "learning_rate": 8e-05,
      "loss": 1.4983,
      "step": 4062
    },
    {
      "epoch": 0.8438213914849428,
      "grad_norm": 0.5308985114097595,
      "learning_rate": 8e-05,
      "loss": 1.5363,
      "step": 4063
    },
    {
      "epoch": 0.8440290758047767,
      "grad_norm": 0.5165321230888367,
      "learning_rate": 8e-05,
      "loss": 1.4852,
      "step": 4064
    },
    {
      "epoch": 0.8442367601246106,
      "grad_norm": 0.5440815091133118,
      "learning_rate": 8e-05,
      "loss": 1.527,
      "step": 4065
    },
    {
      "epoch": 0.8444444444444444,
      "grad_norm": 0.5439822673797607,
      "learning_rate": 8e-05,
      "loss": 1.5513,
      "step": 4066
    },
    {
      "epoch": 0.8446521287642783,
      "grad_norm": 0.5333788990974426,
      "learning_rate": 8e-05,
      "loss": 1.5351,
      "step": 4067
    },
    {
      "epoch": 0.8448598130841122,
      "grad_norm": 0.5439633131027222,
      "learning_rate": 8e-05,
      "loss": 1.5441,
      "step": 4068
    },
    {
      "epoch": 0.845067497403946,
      "grad_norm": 0.576550304889679,
      "learning_rate": 8e-05,
      "loss": 1.4042,
      "step": 4069
    },
    {
      "epoch": 0.8452751817237798,
      "grad_norm": 0.5670768618583679,
      "learning_rate": 8e-05,
      "loss": 1.5748,
      "step": 4070
    },
    {
      "epoch": 0.8454828660436137,
      "grad_norm": 0.570607602596283,
      "learning_rate": 8e-05,
      "loss": 1.648,
      "step": 4071
    },
    {
      "epoch": 0.8456905503634475,
      "grad_norm": 0.532333254814148,
      "learning_rate": 8e-05,
      "loss": 1.5168,
      "step": 4072
    },
    {
      "epoch": 0.8458982346832814,
      "grad_norm": 0.539936900138855,
      "learning_rate": 8e-05,
      "loss": 1.5527,
      "step": 4073
    },
    {
      "epoch": 0.8461059190031153,
      "grad_norm": 0.5199199914932251,
      "learning_rate": 8e-05,
      "loss": 1.488,
      "step": 4074
    },
    {
      "epoch": 0.8463136033229491,
      "grad_norm": 0.5832904577255249,
      "learning_rate": 8e-05,
      "loss": 1.549,
      "step": 4075
    },
    {
      "epoch": 0.846521287642783,
      "grad_norm": 0.5443639159202576,
      "learning_rate": 8e-05,
      "loss": 1.5784,
      "step": 4076
    },
    {
      "epoch": 0.8467289719626169,
      "grad_norm": 0.5204474329948425,
      "learning_rate": 8e-05,
      "loss": 1.4697,
      "step": 4077
    },
    {
      "epoch": 0.8469366562824506,
      "grad_norm": 0.5725163221359253,
      "learning_rate": 8e-05,
      "loss": 1.5172,
      "step": 4078
    },
    {
      "epoch": 0.8471443406022845,
      "grad_norm": 0.6486271023750305,
      "learning_rate": 8e-05,
      "loss": 1.5617,
      "step": 4079
    },
    {
      "epoch": 0.8473520249221184,
      "grad_norm": 0.5463476777076721,
      "learning_rate": 8e-05,
      "loss": 1.4938,
      "step": 4080
    },
    {
      "epoch": 0.8475597092419522,
      "grad_norm": 0.5705676674842834,
      "learning_rate": 8e-05,
      "loss": 1.5153,
      "step": 4081
    },
    {
      "epoch": 0.8477673935617861,
      "grad_norm": 0.5501434803009033,
      "learning_rate": 8e-05,
      "loss": 1.4762,
      "step": 4082
    },
    {
      "epoch": 0.84797507788162,
      "grad_norm": 0.5594474673271179,
      "learning_rate": 8e-05,
      "loss": 1.5516,
      "step": 4083
    },
    {
      "epoch": 0.8481827622014538,
      "grad_norm": 0.5547730326652527,
      "learning_rate": 8e-05,
      "loss": 1.4954,
      "step": 4084
    },
    {
      "epoch": 0.8483904465212876,
      "grad_norm": 0.5516170859336853,
      "learning_rate": 8e-05,
      "loss": 1.5449,
      "step": 4085
    },
    {
      "epoch": 0.8485981308411215,
      "grad_norm": 0.5502596497535706,
      "learning_rate": 8e-05,
      "loss": 1.5865,
      "step": 4086
    },
    {
      "epoch": 0.8488058151609553,
      "grad_norm": 0.5618826746940613,
      "learning_rate": 8e-05,
      "loss": 1.555,
      "step": 4087
    },
    {
      "epoch": 0.8490134994807892,
      "grad_norm": 0.560907244682312,
      "learning_rate": 8e-05,
      "loss": 1.5375,
      "step": 4088
    },
    {
      "epoch": 0.8492211838006231,
      "grad_norm": 0.5508263111114502,
      "learning_rate": 8e-05,
      "loss": 1.5525,
      "step": 4089
    },
    {
      "epoch": 0.8494288681204569,
      "grad_norm": 0.5411702990531921,
      "learning_rate": 8e-05,
      "loss": 1.5573,
      "step": 4090
    },
    {
      "epoch": 0.8496365524402908,
      "grad_norm": 0.6111897230148315,
      "learning_rate": 8e-05,
      "loss": 1.5279,
      "step": 4091
    },
    {
      "epoch": 0.8498442367601247,
      "grad_norm": 0.5265816450119019,
      "learning_rate": 8e-05,
      "loss": 1.5321,
      "step": 4092
    },
    {
      "epoch": 0.8500519210799584,
      "grad_norm": 0.6112878918647766,
      "learning_rate": 8e-05,
      "loss": 1.5968,
      "step": 4093
    },
    {
      "epoch": 0.8502596053997923,
      "grad_norm": 0.6115054488182068,
      "learning_rate": 8e-05,
      "loss": 1.5069,
      "step": 4094
    },
    {
      "epoch": 0.8504672897196262,
      "grad_norm": 0.5600935816764832,
      "learning_rate": 8e-05,
      "loss": 1.4423,
      "step": 4095
    },
    {
      "epoch": 0.85067497403946,
      "grad_norm": 0.6327739953994751,
      "learning_rate": 8e-05,
      "loss": 1.4989,
      "step": 4096
    },
    {
      "epoch": 0.8508826583592939,
      "grad_norm": 0.5668690800666809,
      "learning_rate": 8e-05,
      "loss": 1.4625,
      "step": 4097
    },
    {
      "epoch": 0.8510903426791278,
      "grad_norm": 0.6276752948760986,
      "learning_rate": 8e-05,
      "loss": 1.5993,
      "step": 4098
    },
    {
      "epoch": 0.8512980269989616,
      "grad_norm": 0.5278996825218201,
      "learning_rate": 8e-05,
      "loss": 1.4278,
      "step": 4099
    },
    {
      "epoch": 0.8515057113187954,
      "grad_norm": 0.5873367786407471,
      "learning_rate": 8e-05,
      "loss": 1.5804,
      "step": 4100
    },
    {
      "epoch": 0.8517133956386292,
      "grad_norm": 0.6634363532066345,
      "learning_rate": 8e-05,
      "loss": 1.5709,
      "step": 4101
    },
    {
      "epoch": 0.8519210799584631,
      "grad_norm": 0.5260589718818665,
      "learning_rate": 8e-05,
      "loss": 1.4825,
      "step": 4102
    },
    {
      "epoch": 0.852128764278297,
      "grad_norm": 0.5167551040649414,
      "learning_rate": 8e-05,
      "loss": 1.4967,
      "step": 4103
    },
    {
      "epoch": 0.8523364485981308,
      "grad_norm": 0.5843436121940613,
      "learning_rate": 8e-05,
      "loss": 1.5591,
      "step": 4104
    },
    {
      "epoch": 0.8525441329179647,
      "grad_norm": 0.5058687329292297,
      "learning_rate": 8e-05,
      "loss": 1.5016,
      "step": 4105
    },
    {
      "epoch": 0.8527518172377986,
      "grad_norm": 0.5123895406723022,
      "learning_rate": 8e-05,
      "loss": 1.5028,
      "step": 4106
    },
    {
      "epoch": 0.8529595015576324,
      "grad_norm": 0.5638048052787781,
      "learning_rate": 8e-05,
      "loss": 1.5927,
      "step": 4107
    },
    {
      "epoch": 0.8531671858774662,
      "grad_norm": 0.548582911491394,
      "learning_rate": 8e-05,
      "loss": 1.5276,
      "step": 4108
    },
    {
      "epoch": 0.8533748701973001,
      "grad_norm": 0.5390303730964661,
      "learning_rate": 8e-05,
      "loss": 1.4694,
      "step": 4109
    },
    {
      "epoch": 0.8535825545171339,
      "grad_norm": 0.5299291014671326,
      "learning_rate": 8e-05,
      "loss": 1.5197,
      "step": 4110
    },
    {
      "epoch": 0.8537902388369678,
      "grad_norm": 0.571440577507019,
      "learning_rate": 8e-05,
      "loss": 1.5485,
      "step": 4111
    },
    {
      "epoch": 0.8539979231568017,
      "grad_norm": 0.552693247795105,
      "learning_rate": 8e-05,
      "loss": 1.5302,
      "step": 4112
    },
    {
      "epoch": 0.8542056074766355,
      "grad_norm": 0.5545945167541504,
      "learning_rate": 8e-05,
      "loss": 1.4984,
      "step": 4113
    },
    {
      "epoch": 0.8544132917964694,
      "grad_norm": 0.5322532653808594,
      "learning_rate": 8e-05,
      "loss": 1.625,
      "step": 4114
    },
    {
      "epoch": 0.8546209761163032,
      "grad_norm": 0.5357773900032043,
      "learning_rate": 8e-05,
      "loss": 1.5154,
      "step": 4115
    },
    {
      "epoch": 0.854828660436137,
      "grad_norm": 0.5189059972763062,
      "learning_rate": 8e-05,
      "loss": 1.5895,
      "step": 4116
    },
    {
      "epoch": 0.8550363447559709,
      "grad_norm": 0.6186309456825256,
      "learning_rate": 8e-05,
      "loss": 1.6693,
      "step": 4117
    },
    {
      "epoch": 0.8552440290758048,
      "grad_norm": 0.5378537774085999,
      "learning_rate": 8e-05,
      "loss": 1.622,
      "step": 4118
    },
    {
      "epoch": 0.8554517133956386,
      "grad_norm": 0.504642903804779,
      "learning_rate": 8e-05,
      "loss": 1.5311,
      "step": 4119
    },
    {
      "epoch": 0.8556593977154725,
      "grad_norm": 0.553855299949646,
      "learning_rate": 8e-05,
      "loss": 1.5937,
      "step": 4120
    },
    {
      "epoch": 0.8558670820353064,
      "grad_norm": 0.6109062433242798,
      "learning_rate": 8e-05,
      "loss": 1.5049,
      "step": 4121
    },
    {
      "epoch": 0.8560747663551402,
      "grad_norm": 0.5732825994491577,
      "learning_rate": 8e-05,
      "loss": 1.5901,
      "step": 4122
    },
    {
      "epoch": 0.856282450674974,
      "grad_norm": 0.5736696124076843,
      "learning_rate": 8e-05,
      "loss": 1.5426,
      "step": 4123
    },
    {
      "epoch": 0.8564901349948079,
      "grad_norm": 0.535771369934082,
      "learning_rate": 8e-05,
      "loss": 1.4976,
      "step": 4124
    },
    {
      "epoch": 0.8566978193146417,
      "grad_norm": 0.5533987283706665,
      "learning_rate": 8e-05,
      "loss": 1.4919,
      "step": 4125
    },
    {
      "epoch": 0.8569055036344756,
      "grad_norm": 0.5147508978843689,
      "learning_rate": 8e-05,
      "loss": 1.5408,
      "step": 4126
    },
    {
      "epoch": 0.8571131879543095,
      "grad_norm": 0.5331518650054932,
      "learning_rate": 8e-05,
      "loss": 1.5127,
      "step": 4127
    },
    {
      "epoch": 0.8573208722741433,
      "grad_norm": 0.5425363779067993,
      "learning_rate": 8e-05,
      "loss": 1.5183,
      "step": 4128
    },
    {
      "epoch": 0.8575285565939772,
      "grad_norm": 0.5604209303855896,
      "learning_rate": 8e-05,
      "loss": 1.5791,
      "step": 4129
    },
    {
      "epoch": 0.857736240913811,
      "grad_norm": 0.5611351132392883,
      "learning_rate": 8e-05,
      "loss": 1.5154,
      "step": 4130
    },
    {
      "epoch": 0.8579439252336448,
      "grad_norm": 0.5452185273170471,
      "learning_rate": 8e-05,
      "loss": 1.6407,
      "step": 4131
    },
    {
      "epoch": 0.8581516095534787,
      "grad_norm": 0.5583484172821045,
      "learning_rate": 8e-05,
      "loss": 1.603,
      "step": 4132
    },
    {
      "epoch": 0.8583592938733126,
      "grad_norm": 0.5416769981384277,
      "learning_rate": 8e-05,
      "loss": 1.4556,
      "step": 4133
    },
    {
      "epoch": 0.8585669781931464,
      "grad_norm": 0.5185193419456482,
      "learning_rate": 8e-05,
      "loss": 1.4573,
      "step": 4134
    },
    {
      "epoch": 0.8587746625129803,
      "grad_norm": 0.547526478767395,
      "learning_rate": 8e-05,
      "loss": 1.5,
      "step": 4135
    },
    {
      "epoch": 0.8589823468328142,
      "grad_norm": 0.5527027249336243,
      "learning_rate": 8e-05,
      "loss": 1.5441,
      "step": 4136
    },
    {
      "epoch": 0.859190031152648,
      "grad_norm": 0.5204476714134216,
      "learning_rate": 8e-05,
      "loss": 1.4425,
      "step": 4137
    },
    {
      "epoch": 0.8593977154724818,
      "grad_norm": 0.5701795220375061,
      "learning_rate": 8e-05,
      "loss": 1.5139,
      "step": 4138
    },
    {
      "epoch": 0.8596053997923156,
      "grad_norm": 0.5447306036949158,
      "learning_rate": 8e-05,
      "loss": 1.5449,
      "step": 4139
    },
    {
      "epoch": 0.8598130841121495,
      "grad_norm": 0.5561647415161133,
      "learning_rate": 8e-05,
      "loss": 1.5452,
      "step": 4140
    },
    {
      "epoch": 0.8600207684319834,
      "grad_norm": 0.5332565903663635,
      "learning_rate": 8e-05,
      "loss": 1.5483,
      "step": 4141
    },
    {
      "epoch": 0.8602284527518173,
      "grad_norm": 0.5357800722122192,
      "learning_rate": 8e-05,
      "loss": 1.5216,
      "step": 4142
    },
    {
      "epoch": 0.8604361370716511,
      "grad_norm": 0.540325403213501,
      "learning_rate": 8e-05,
      "loss": 1.4729,
      "step": 4143
    },
    {
      "epoch": 0.860643821391485,
      "grad_norm": 0.5466431379318237,
      "learning_rate": 8e-05,
      "loss": 1.5546,
      "step": 4144
    },
    {
      "epoch": 0.8608515057113187,
      "grad_norm": 0.5825132131576538,
      "learning_rate": 8e-05,
      "loss": 1.5442,
      "step": 4145
    },
    {
      "epoch": 0.8610591900311526,
      "grad_norm": 0.5389710664749146,
      "learning_rate": 8e-05,
      "loss": 1.5691,
      "step": 4146
    },
    {
      "epoch": 0.8612668743509865,
      "grad_norm": 0.5395361185073853,
      "learning_rate": 8e-05,
      "loss": 1.5278,
      "step": 4147
    },
    {
      "epoch": 0.8614745586708203,
      "grad_norm": 0.5261792540550232,
      "learning_rate": 8e-05,
      "loss": 1.5347,
      "step": 4148
    },
    {
      "epoch": 0.8616822429906542,
      "grad_norm": 0.5786598324775696,
      "learning_rate": 8e-05,
      "loss": 1.5219,
      "step": 4149
    },
    {
      "epoch": 0.8618899273104881,
      "grad_norm": 0.5467092990875244,
      "learning_rate": 8e-05,
      "loss": 1.5439,
      "step": 4150
    },
    {
      "epoch": 0.862097611630322,
      "grad_norm": 0.5433300733566284,
      "learning_rate": 8e-05,
      "loss": 1.4693,
      "step": 4151
    },
    {
      "epoch": 0.8623052959501558,
      "grad_norm": 0.5423396825790405,
      "learning_rate": 8e-05,
      "loss": 1.5343,
      "step": 4152
    },
    {
      "epoch": 0.8625129802699896,
      "grad_norm": 0.5553238987922668,
      "learning_rate": 8e-05,
      "loss": 1.5644,
      "step": 4153
    },
    {
      "epoch": 0.8627206645898234,
      "grad_norm": 0.5624662637710571,
      "learning_rate": 8e-05,
      "loss": 1.6371,
      "step": 4154
    },
    {
      "epoch": 0.8629283489096573,
      "grad_norm": 0.5311103463172913,
      "learning_rate": 8e-05,
      "loss": 1.5042,
      "step": 4155
    },
    {
      "epoch": 0.8631360332294912,
      "grad_norm": 0.5343470573425293,
      "learning_rate": 8e-05,
      "loss": 1.5188,
      "step": 4156
    },
    {
      "epoch": 0.863343717549325,
      "grad_norm": 0.5321030020713806,
      "learning_rate": 8e-05,
      "loss": 1.6287,
      "step": 4157
    },
    {
      "epoch": 0.8635514018691589,
      "grad_norm": 0.5414921641349792,
      "learning_rate": 8e-05,
      "loss": 1.5063,
      "step": 4158
    },
    {
      "epoch": 0.8637590861889928,
      "grad_norm": 0.5469291806221008,
      "learning_rate": 8e-05,
      "loss": 1.5306,
      "step": 4159
    },
    {
      "epoch": 0.8639667705088265,
      "grad_norm": 0.5274113416671753,
      "learning_rate": 8e-05,
      "loss": 1.5226,
      "step": 4160
    },
    {
      "epoch": 0.8641744548286604,
      "grad_norm": 0.5480718612670898,
      "learning_rate": 8e-05,
      "loss": 1.4959,
      "step": 4161
    },
    {
      "epoch": 0.8643821391484943,
      "grad_norm": 0.5283731818199158,
      "learning_rate": 8e-05,
      "loss": 1.4995,
      "step": 4162
    },
    {
      "epoch": 0.8645898234683281,
      "grad_norm": 0.5445735454559326,
      "learning_rate": 8e-05,
      "loss": 1.5749,
      "step": 4163
    },
    {
      "epoch": 0.864797507788162,
      "grad_norm": 0.5314701795578003,
      "learning_rate": 8e-05,
      "loss": 1.5484,
      "step": 4164
    },
    {
      "epoch": 0.8650051921079959,
      "grad_norm": 0.5795189738273621,
      "learning_rate": 8e-05,
      "loss": 1.5256,
      "step": 4165
    },
    {
      "epoch": 0.8652128764278297,
      "grad_norm": 0.6255568861961365,
      "learning_rate": 8e-05,
      "loss": 1.4949,
      "step": 4166
    },
    {
      "epoch": 0.8654205607476636,
      "grad_norm": 0.5383212566375732,
      "learning_rate": 8e-05,
      "loss": 1.6105,
      "step": 4167
    },
    {
      "epoch": 0.8656282450674974,
      "grad_norm": 0.559436559677124,
      "learning_rate": 8e-05,
      "loss": 1.579,
      "step": 4168
    },
    {
      "epoch": 0.8658359293873312,
      "grad_norm": 0.5349182486534119,
      "learning_rate": 8e-05,
      "loss": 1.5326,
      "step": 4169
    },
    {
      "epoch": 0.8660436137071651,
      "grad_norm": 0.5406621098518372,
      "learning_rate": 8e-05,
      "loss": 1.462,
      "step": 4170
    },
    {
      "epoch": 0.866251298026999,
      "grad_norm": 0.5969509482383728,
      "learning_rate": 8e-05,
      "loss": 1.6169,
      "step": 4171
    },
    {
      "epoch": 0.8664589823468328,
      "grad_norm": 0.5488429069519043,
      "learning_rate": 8e-05,
      "loss": 1.4823,
      "step": 4172
    },
    {
      "epoch": 0.8666666666666667,
      "grad_norm": 0.5515419840812683,
      "learning_rate": 8e-05,
      "loss": 1.5472,
      "step": 4173
    },
    {
      "epoch": 0.8668743509865006,
      "grad_norm": 0.5823199152946472,
      "learning_rate": 8e-05,
      "loss": 1.5949,
      "step": 4174
    },
    {
      "epoch": 0.8670820353063343,
      "grad_norm": 0.5642107129096985,
      "learning_rate": 8e-05,
      "loss": 1.5532,
      "step": 4175
    },
    {
      "epoch": 0.8672897196261682,
      "grad_norm": 0.5283724665641785,
      "learning_rate": 8e-05,
      "loss": 1.5477,
      "step": 4176
    },
    {
      "epoch": 0.867497403946002,
      "grad_norm": 0.5420722961425781,
      "learning_rate": 8e-05,
      "loss": 1.5874,
      "step": 4177
    },
    {
      "epoch": 0.8677050882658359,
      "grad_norm": 0.6576200127601624,
      "learning_rate": 8e-05,
      "loss": 1.5418,
      "step": 4178
    },
    {
      "epoch": 0.8679127725856698,
      "grad_norm": 0.6050857305526733,
      "learning_rate": 8e-05,
      "loss": 1.5485,
      "step": 4179
    },
    {
      "epoch": 0.8681204569055037,
      "grad_norm": 0.5492483973503113,
      "learning_rate": 8e-05,
      "loss": 1.5297,
      "step": 4180
    },
    {
      "epoch": 0.8683281412253375,
      "grad_norm": 0.5656388401985168,
      "learning_rate": 8e-05,
      "loss": 1.5638,
      "step": 4181
    },
    {
      "epoch": 0.8685358255451714,
      "grad_norm": 0.5260201692581177,
      "learning_rate": 8e-05,
      "loss": 1.5321,
      "step": 4182
    },
    {
      "epoch": 0.8687435098650051,
      "grad_norm": 0.5901237726211548,
      "learning_rate": 8e-05,
      "loss": 1.5074,
      "step": 4183
    },
    {
      "epoch": 0.868951194184839,
      "grad_norm": 0.5367937088012695,
      "learning_rate": 8e-05,
      "loss": 1.5577,
      "step": 4184
    },
    {
      "epoch": 0.8691588785046729,
      "grad_norm": 0.5272131562232971,
      "learning_rate": 8e-05,
      "loss": 1.4882,
      "step": 4185
    },
    {
      "epoch": 0.8693665628245068,
      "grad_norm": 0.6219227313995361,
      "learning_rate": 8e-05,
      "loss": 1.5004,
      "step": 4186
    },
    {
      "epoch": 0.8695742471443406,
      "grad_norm": 0.5530415177345276,
      "learning_rate": 8e-05,
      "loss": 1.4826,
      "step": 4187
    },
    {
      "epoch": 0.8697819314641745,
      "grad_norm": 0.5585103631019592,
      "learning_rate": 8e-05,
      "loss": 1.5432,
      "step": 4188
    },
    {
      "epoch": 0.8699896157840084,
      "grad_norm": 0.56638503074646,
      "learning_rate": 8e-05,
      "loss": 1.5302,
      "step": 4189
    },
    {
      "epoch": 0.8701973001038421,
      "grad_norm": 0.5242372155189514,
      "learning_rate": 8e-05,
      "loss": 1.5069,
      "step": 4190
    },
    {
      "epoch": 0.870404984423676,
      "grad_norm": 0.5555446743965149,
      "learning_rate": 8e-05,
      "loss": 1.5152,
      "step": 4191
    },
    {
      "epoch": 0.8706126687435098,
      "grad_norm": 0.5374391078948975,
      "learning_rate": 8e-05,
      "loss": 1.504,
      "step": 4192
    },
    {
      "epoch": 0.8708203530633437,
      "grad_norm": 0.5152431130409241,
      "learning_rate": 8e-05,
      "loss": 1.4678,
      "step": 4193
    },
    {
      "epoch": 0.8710280373831776,
      "grad_norm": 0.5511677265167236,
      "learning_rate": 8e-05,
      "loss": 1.5986,
      "step": 4194
    },
    {
      "epoch": 0.8712357217030114,
      "grad_norm": 0.6084768772125244,
      "learning_rate": 8e-05,
      "loss": 1.4864,
      "step": 4195
    },
    {
      "epoch": 0.8714434060228453,
      "grad_norm": 0.5424408912658691,
      "learning_rate": 8e-05,
      "loss": 1.4909,
      "step": 4196
    },
    {
      "epoch": 0.8716510903426792,
      "grad_norm": 0.5213330388069153,
      "learning_rate": 8e-05,
      "loss": 1.482,
      "step": 4197
    },
    {
      "epoch": 0.8718587746625129,
      "grad_norm": 0.6206569671630859,
      "learning_rate": 8e-05,
      "loss": 1.4389,
      "step": 4198
    },
    {
      "epoch": 0.8720664589823468,
      "grad_norm": 0.5570772886276245,
      "learning_rate": 8e-05,
      "loss": 1.5605,
      "step": 4199
    },
    {
      "epoch": 0.8722741433021807,
      "grad_norm": 0.5366504788398743,
      "learning_rate": 8e-05,
      "loss": 1.4966,
      "step": 4200
    },
    {
      "epoch": 0.8724818276220145,
      "grad_norm": 0.5707666873931885,
      "learning_rate": 8e-05,
      "loss": 1.5839,
      "step": 4201
    },
    {
      "epoch": 0.8726895119418484,
      "grad_norm": 0.6071116924285889,
      "learning_rate": 8e-05,
      "loss": 1.5061,
      "step": 4202
    },
    {
      "epoch": 0.8728971962616823,
      "grad_norm": 0.5423369407653809,
      "learning_rate": 8e-05,
      "loss": 1.5127,
      "step": 4203
    },
    {
      "epoch": 0.8731048805815161,
      "grad_norm": 0.5461960434913635,
      "learning_rate": 8e-05,
      "loss": 1.5176,
      "step": 4204
    },
    {
      "epoch": 0.8733125649013499,
      "grad_norm": 0.5446351170539856,
      "learning_rate": 8e-05,
      "loss": 1.532,
      "step": 4205
    },
    {
      "epoch": 0.8735202492211838,
      "grad_norm": 0.5574485659599304,
      "learning_rate": 8e-05,
      "loss": 1.5578,
      "step": 4206
    },
    {
      "epoch": 0.8737279335410176,
      "grad_norm": 0.5535857081413269,
      "learning_rate": 8e-05,
      "loss": 1.5385,
      "step": 4207
    },
    {
      "epoch": 0.8739356178608515,
      "grad_norm": 0.5574408769607544,
      "learning_rate": 8e-05,
      "loss": 1.5432,
      "step": 4208
    },
    {
      "epoch": 0.8741433021806854,
      "grad_norm": 0.5765231251716614,
      "learning_rate": 8e-05,
      "loss": 1.5138,
      "step": 4209
    },
    {
      "epoch": 0.8743509865005192,
      "grad_norm": 0.5485952496528625,
      "learning_rate": 8e-05,
      "loss": 1.514,
      "step": 4210
    },
    {
      "epoch": 0.8745586708203531,
      "grad_norm": 0.5387266278266907,
      "learning_rate": 8e-05,
      "loss": 1.4594,
      "step": 4211
    },
    {
      "epoch": 0.874766355140187,
      "grad_norm": 0.6751468181610107,
      "learning_rate": 8e-05,
      "loss": 1.5761,
      "step": 4212
    },
    {
      "epoch": 0.8749740394600207,
      "grad_norm": 0.5510568618774414,
      "learning_rate": 8e-05,
      "loss": 1.5937,
      "step": 4213
    },
    {
      "epoch": 0.8751817237798546,
      "grad_norm": 0.5636010766029358,
      "learning_rate": 8e-05,
      "loss": 1.5433,
      "step": 4214
    },
    {
      "epoch": 0.8753894080996885,
      "grad_norm": 0.5983996987342834,
      "learning_rate": 8e-05,
      "loss": 1.5033,
      "step": 4215
    },
    {
      "epoch": 0.8755970924195223,
      "grad_norm": 0.5609733462333679,
      "learning_rate": 8e-05,
      "loss": 1.5789,
      "step": 4216
    },
    {
      "epoch": 0.8758047767393562,
      "grad_norm": 0.5443954467773438,
      "learning_rate": 8e-05,
      "loss": 1.5424,
      "step": 4217
    },
    {
      "epoch": 0.8760124610591901,
      "grad_norm": 0.5640705227851868,
      "learning_rate": 8e-05,
      "loss": 1.5457,
      "step": 4218
    },
    {
      "epoch": 0.8762201453790239,
      "grad_norm": 0.5597903728485107,
      "learning_rate": 8e-05,
      "loss": 1.4801,
      "step": 4219
    },
    {
      "epoch": 0.8764278296988577,
      "grad_norm": 0.5559795498847961,
      "learning_rate": 8e-05,
      "loss": 1.5102,
      "step": 4220
    },
    {
      "epoch": 0.8766355140186916,
      "grad_norm": 0.5658156871795654,
      "learning_rate": 8e-05,
      "loss": 1.5588,
      "step": 4221
    },
    {
      "epoch": 0.8768431983385254,
      "grad_norm": 0.5939041376113892,
      "learning_rate": 8e-05,
      "loss": 1.6088,
      "step": 4222
    },
    {
      "epoch": 0.8770508826583593,
      "grad_norm": 0.5389169454574585,
      "learning_rate": 8e-05,
      "loss": 1.6063,
      "step": 4223
    },
    {
      "epoch": 0.8772585669781932,
      "grad_norm": 0.5385518074035645,
      "learning_rate": 8e-05,
      "loss": 1.5561,
      "step": 4224
    },
    {
      "epoch": 0.877466251298027,
      "grad_norm": 0.5546438694000244,
      "learning_rate": 8e-05,
      "loss": 1.4984,
      "step": 4225
    },
    {
      "epoch": 0.8776739356178609,
      "grad_norm": 0.5732212066650391,
      "learning_rate": 8e-05,
      "loss": 1.6017,
      "step": 4226
    },
    {
      "epoch": 0.8778816199376948,
      "grad_norm": 0.5223263502120972,
      "learning_rate": 8e-05,
      "loss": 1.5341,
      "step": 4227
    },
    {
      "epoch": 0.8780893042575285,
      "grad_norm": 0.5600264668464661,
      "learning_rate": 8e-05,
      "loss": 1.5408,
      "step": 4228
    },
    {
      "epoch": 0.8782969885773624,
      "grad_norm": 0.5329475402832031,
      "learning_rate": 8e-05,
      "loss": 1.4759,
      "step": 4229
    },
    {
      "epoch": 0.8785046728971962,
      "grad_norm": 0.5668915510177612,
      "learning_rate": 8e-05,
      "loss": 1.493,
      "step": 4230
    },
    {
      "epoch": 0.8787123572170301,
      "grad_norm": 0.5543871521949768,
      "learning_rate": 8e-05,
      "loss": 1.5784,
      "step": 4231
    },
    {
      "epoch": 0.878920041536864,
      "grad_norm": 0.5443382859230042,
      "learning_rate": 8e-05,
      "loss": 1.5131,
      "step": 4232
    },
    {
      "epoch": 0.8791277258566979,
      "grad_norm": 0.5522803664207458,
      "learning_rate": 8e-05,
      "loss": 1.5253,
      "step": 4233
    },
    {
      "epoch": 0.8793354101765317,
      "grad_norm": 0.5921510457992554,
      "learning_rate": 8e-05,
      "loss": 1.576,
      "step": 4234
    },
    {
      "epoch": 0.8795430944963655,
      "grad_norm": 0.5092579126358032,
      "learning_rate": 8e-05,
      "loss": 1.5031,
      "step": 4235
    },
    {
      "epoch": 0.8797507788161993,
      "grad_norm": 0.5595648884773254,
      "learning_rate": 8e-05,
      "loss": 1.5555,
      "step": 4236
    },
    {
      "epoch": 0.8799584631360332,
      "grad_norm": 0.52916419506073,
      "learning_rate": 8e-05,
      "loss": 1.5463,
      "step": 4237
    },
    {
      "epoch": 0.8801661474558671,
      "grad_norm": 0.5369413495063782,
      "learning_rate": 8e-05,
      "loss": 1.5365,
      "step": 4238
    },
    {
      "epoch": 0.8803738317757009,
      "grad_norm": 0.519568145275116,
      "learning_rate": 8e-05,
      "loss": 1.4533,
      "step": 4239
    },
    {
      "epoch": 0.8805815160955348,
      "grad_norm": 0.5537676811218262,
      "learning_rate": 8e-05,
      "loss": 1.5725,
      "step": 4240
    },
    {
      "epoch": 0.8807892004153687,
      "grad_norm": 0.519368052482605,
      "learning_rate": 8e-05,
      "loss": 1.5131,
      "step": 4241
    },
    {
      "epoch": 0.8809968847352025,
      "grad_norm": 0.5874369740486145,
      "learning_rate": 8e-05,
      "loss": 1.481,
      "step": 4242
    },
    {
      "epoch": 0.8812045690550363,
      "grad_norm": 0.5312866568565369,
      "learning_rate": 8e-05,
      "loss": 1.5407,
      "step": 4243
    },
    {
      "epoch": 0.8814122533748702,
      "grad_norm": 0.5592029690742493,
      "learning_rate": 8e-05,
      "loss": 1.48,
      "step": 4244
    },
    {
      "epoch": 0.881619937694704,
      "grad_norm": 0.5118473172187805,
      "learning_rate": 8e-05,
      "loss": 1.5192,
      "step": 4245
    },
    {
      "epoch": 0.8818276220145379,
      "grad_norm": 0.6147160530090332,
      "learning_rate": 8e-05,
      "loss": 1.6088,
      "step": 4246
    },
    {
      "epoch": 0.8820353063343718,
      "grad_norm": 0.5198656916618347,
      "learning_rate": 8e-05,
      "loss": 1.5005,
      "step": 4247
    },
    {
      "epoch": 0.8822429906542056,
      "grad_norm": 0.5508028864860535,
      "learning_rate": 8e-05,
      "loss": 1.5482,
      "step": 4248
    },
    {
      "epoch": 0.8824506749740395,
      "grad_norm": 0.5273780822753906,
      "learning_rate": 8e-05,
      "loss": 1.5133,
      "step": 4249
    },
    {
      "epoch": 0.8826583592938733,
      "grad_norm": 0.6182386875152588,
      "learning_rate": 8e-05,
      "loss": 1.5676,
      "step": 4250
    },
    {
      "epoch": 0.8828660436137071,
      "grad_norm": 0.5553271174430847,
      "learning_rate": 8e-05,
      "loss": 1.5883,
      "step": 4251
    },
    {
      "epoch": 0.883073727933541,
      "grad_norm": 0.5344847440719604,
      "learning_rate": 8e-05,
      "loss": 1.4964,
      "step": 4252
    },
    {
      "epoch": 0.8832814122533749,
      "grad_norm": 0.546604573726654,
      "learning_rate": 8e-05,
      "loss": 1.4613,
      "step": 4253
    },
    {
      "epoch": 0.8834890965732087,
      "grad_norm": 0.5706937313079834,
      "learning_rate": 8e-05,
      "loss": 1.5463,
      "step": 4254
    },
    {
      "epoch": 0.8836967808930426,
      "grad_norm": 0.6246913075447083,
      "learning_rate": 8e-05,
      "loss": 1.5014,
      "step": 4255
    },
    {
      "epoch": 0.8839044652128765,
      "grad_norm": 0.5374740958213806,
      "learning_rate": 8e-05,
      "loss": 1.4677,
      "step": 4256
    },
    {
      "epoch": 0.8841121495327103,
      "grad_norm": 0.5699962377548218,
      "learning_rate": 8e-05,
      "loss": 1.583,
      "step": 4257
    },
    {
      "epoch": 0.8843198338525441,
      "grad_norm": 0.553851842880249,
      "learning_rate": 8e-05,
      "loss": 1.6081,
      "step": 4258
    },
    {
      "epoch": 0.884527518172378,
      "grad_norm": 0.5520926117897034,
      "learning_rate": 8e-05,
      "loss": 1.547,
      "step": 4259
    },
    {
      "epoch": 0.8847352024922118,
      "grad_norm": 0.5466866493225098,
      "learning_rate": 8e-05,
      "loss": 1.4926,
      "step": 4260
    },
    {
      "epoch": 0.8849428868120457,
      "grad_norm": 0.5163272023200989,
      "learning_rate": 8e-05,
      "loss": 1.4419,
      "step": 4261
    },
    {
      "epoch": 0.8851505711318796,
      "grad_norm": 0.5326411128044128,
      "learning_rate": 8e-05,
      "loss": 1.5197,
      "step": 4262
    },
    {
      "epoch": 0.8853582554517134,
      "grad_norm": 0.5553964972496033,
      "learning_rate": 8e-05,
      "loss": 1.5841,
      "step": 4263
    },
    {
      "epoch": 0.8855659397715473,
      "grad_norm": 0.5232428908348083,
      "learning_rate": 8e-05,
      "loss": 1.5577,
      "step": 4264
    },
    {
      "epoch": 0.885773624091381,
      "grad_norm": 0.5510069727897644,
      "learning_rate": 8e-05,
      "loss": 1.5891,
      "step": 4265
    },
    {
      "epoch": 0.8859813084112149,
      "grad_norm": 0.5180588960647583,
      "learning_rate": 8e-05,
      "loss": 1.494,
      "step": 4266
    },
    {
      "epoch": 0.8861889927310488,
      "grad_norm": 0.5884924530982971,
      "learning_rate": 8e-05,
      "loss": 1.5679,
      "step": 4267
    },
    {
      "epoch": 0.8863966770508827,
      "grad_norm": 0.5629843473434448,
      "learning_rate": 8e-05,
      "loss": 1.5348,
      "step": 4268
    },
    {
      "epoch": 0.8866043613707165,
      "grad_norm": 0.5547498464584351,
      "learning_rate": 8e-05,
      "loss": 1.5635,
      "step": 4269
    },
    {
      "epoch": 0.8868120456905504,
      "grad_norm": 0.5301056504249573,
      "learning_rate": 8e-05,
      "loss": 1.5287,
      "step": 4270
    },
    {
      "epoch": 0.8870197300103843,
      "grad_norm": 0.5747231245040894,
      "learning_rate": 8e-05,
      "loss": 1.5762,
      "step": 4271
    },
    {
      "epoch": 0.8872274143302181,
      "grad_norm": 0.5300017595291138,
      "learning_rate": 8e-05,
      "loss": 1.5334,
      "step": 4272
    },
    {
      "epoch": 0.8874350986500519,
      "grad_norm": 0.5446390509605408,
      "learning_rate": 8e-05,
      "loss": 1.5835,
      "step": 4273
    },
    {
      "epoch": 0.8876427829698857,
      "grad_norm": 0.5481982231140137,
      "learning_rate": 8e-05,
      "loss": 1.4938,
      "step": 4274
    },
    {
      "epoch": 0.8878504672897196,
      "grad_norm": 0.5434316396713257,
      "learning_rate": 8e-05,
      "loss": 1.568,
      "step": 4275
    },
    {
      "epoch": 0.8880581516095535,
      "grad_norm": 0.5895819067955017,
      "learning_rate": 8e-05,
      "loss": 1.5886,
      "step": 4276
    },
    {
      "epoch": 0.8882658359293873,
      "grad_norm": 0.5486096739768982,
      "learning_rate": 8e-05,
      "loss": 1.5532,
      "step": 4277
    },
    {
      "epoch": 0.8884735202492212,
      "grad_norm": 0.5380463004112244,
      "learning_rate": 8e-05,
      "loss": 1.546,
      "step": 4278
    },
    {
      "epoch": 0.8886812045690551,
      "grad_norm": 0.5782397985458374,
      "learning_rate": 8e-05,
      "loss": 1.5303,
      "step": 4279
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.5500957369804382,
      "learning_rate": 8e-05,
      "loss": 1.4846,
      "step": 4280
    },
    {
      "epoch": 0.8890965732087227,
      "grad_norm": 0.5288175344467163,
      "learning_rate": 8e-05,
      "loss": 1.5499,
      "step": 4281
    },
    {
      "epoch": 0.8893042575285566,
      "grad_norm": 0.5673590302467346,
      "learning_rate": 8e-05,
      "loss": 1.5393,
      "step": 4282
    },
    {
      "epoch": 0.8895119418483904,
      "grad_norm": 0.5695676803588867,
      "learning_rate": 8e-05,
      "loss": 1.4794,
      "step": 4283
    },
    {
      "epoch": 0.8897196261682243,
      "grad_norm": 0.553641140460968,
      "learning_rate": 8e-05,
      "loss": 1.5186,
      "step": 4284
    },
    {
      "epoch": 0.8899273104880582,
      "grad_norm": 0.5462049841880798,
      "learning_rate": 8e-05,
      "loss": 1.4833,
      "step": 4285
    },
    {
      "epoch": 0.890134994807892,
      "grad_norm": 0.5689340233802795,
      "learning_rate": 8e-05,
      "loss": 1.4984,
      "step": 4286
    },
    {
      "epoch": 0.8903426791277259,
      "grad_norm": 0.5379559397697449,
      "learning_rate": 8e-05,
      "loss": 1.5158,
      "step": 4287
    },
    {
      "epoch": 0.8905503634475597,
      "grad_norm": 0.5357993841171265,
      "learning_rate": 8e-05,
      "loss": 1.5095,
      "step": 4288
    },
    {
      "epoch": 0.8907580477673935,
      "grad_norm": 0.5592495203018188,
      "learning_rate": 8e-05,
      "loss": 1.5655,
      "step": 4289
    },
    {
      "epoch": 0.8909657320872274,
      "grad_norm": 0.5552611947059631,
      "learning_rate": 8e-05,
      "loss": 1.5061,
      "step": 4290
    },
    {
      "epoch": 0.8911734164070613,
      "grad_norm": 0.5433019399642944,
      "learning_rate": 8e-05,
      "loss": 1.5101,
      "step": 4291
    },
    {
      "epoch": 0.8913811007268951,
      "grad_norm": 0.5456854701042175,
      "learning_rate": 8e-05,
      "loss": 1.4473,
      "step": 4292
    },
    {
      "epoch": 0.891588785046729,
      "grad_norm": 0.526128351688385,
      "learning_rate": 8e-05,
      "loss": 1.477,
      "step": 4293
    },
    {
      "epoch": 0.8917964693665629,
      "grad_norm": 0.5620971918106079,
      "learning_rate": 8e-05,
      "loss": 1.4975,
      "step": 4294
    },
    {
      "epoch": 0.8920041536863966,
      "grad_norm": 0.5467966198921204,
      "learning_rate": 8e-05,
      "loss": 1.5119,
      "step": 4295
    },
    {
      "epoch": 0.8922118380062305,
      "grad_norm": 0.5574969053268433,
      "learning_rate": 8e-05,
      "loss": 1.4973,
      "step": 4296
    },
    {
      "epoch": 0.8924195223260644,
      "grad_norm": 0.5514340400695801,
      "learning_rate": 8e-05,
      "loss": 1.5406,
      "step": 4297
    },
    {
      "epoch": 0.8926272066458982,
      "grad_norm": 0.5266357660293579,
      "learning_rate": 8e-05,
      "loss": 1.504,
      "step": 4298
    },
    {
      "epoch": 0.8928348909657321,
      "grad_norm": 0.6000077724456787,
      "learning_rate": 8e-05,
      "loss": 1.551,
      "step": 4299
    },
    {
      "epoch": 0.893042575285566,
      "grad_norm": 0.5272939205169678,
      "learning_rate": 8e-05,
      "loss": 1.5418,
      "step": 4300
    },
    {
      "epoch": 0.8932502596053998,
      "grad_norm": 0.5443248152732849,
      "learning_rate": 8e-05,
      "loss": 1.5443,
      "step": 4301
    },
    {
      "epoch": 0.8934579439252337,
      "grad_norm": 0.5493087768554688,
      "learning_rate": 8e-05,
      "loss": 1.535,
      "step": 4302
    },
    {
      "epoch": 0.8936656282450675,
      "grad_norm": 0.532324492931366,
      "learning_rate": 8e-05,
      "loss": 1.5727,
      "step": 4303
    },
    {
      "epoch": 0.8938733125649013,
      "grad_norm": 0.5252280235290527,
      "learning_rate": 8e-05,
      "loss": 1.5031,
      "step": 4304
    },
    {
      "epoch": 0.8940809968847352,
      "grad_norm": 0.5459213256835938,
      "learning_rate": 8e-05,
      "loss": 1.5191,
      "step": 4305
    },
    {
      "epoch": 0.8942886812045691,
      "grad_norm": 0.5310253500938416,
      "learning_rate": 8e-05,
      "loss": 1.5602,
      "step": 4306
    },
    {
      "epoch": 0.8944963655244029,
      "grad_norm": 0.5570583343505859,
      "learning_rate": 8e-05,
      "loss": 1.4707,
      "step": 4307
    },
    {
      "epoch": 0.8947040498442368,
      "grad_norm": 0.5441780090332031,
      "learning_rate": 8e-05,
      "loss": 1.4637,
      "step": 4308
    },
    {
      "epoch": 0.8949117341640707,
      "grad_norm": 0.6435619592666626,
      "learning_rate": 8e-05,
      "loss": 1.5897,
      "step": 4309
    },
    {
      "epoch": 0.8951194184839044,
      "grad_norm": 0.5852838754653931,
      "learning_rate": 8e-05,
      "loss": 1.5826,
      "step": 4310
    },
    {
      "epoch": 0.8953271028037383,
      "grad_norm": 0.5541663765907288,
      "learning_rate": 8e-05,
      "loss": 1.4662,
      "step": 4311
    },
    {
      "epoch": 0.8955347871235722,
      "grad_norm": 0.5126922130584717,
      "learning_rate": 8e-05,
      "loss": 1.4674,
      "step": 4312
    },
    {
      "epoch": 0.895742471443406,
      "grad_norm": 0.5479088425636292,
      "learning_rate": 8e-05,
      "loss": 1.521,
      "step": 4313
    },
    {
      "epoch": 0.8959501557632399,
      "grad_norm": 0.5621117949485779,
      "learning_rate": 8e-05,
      "loss": 1.5566,
      "step": 4314
    },
    {
      "epoch": 0.8961578400830738,
      "grad_norm": 0.543610692024231,
      "learning_rate": 8e-05,
      "loss": 1.5447,
      "step": 4315
    },
    {
      "epoch": 0.8963655244029076,
      "grad_norm": 0.5378540754318237,
      "learning_rate": 8e-05,
      "loss": 1.4123,
      "step": 4316
    },
    {
      "epoch": 0.8965732087227415,
      "grad_norm": 0.6271756887435913,
      "learning_rate": 8e-05,
      "loss": 1.5062,
      "step": 4317
    },
    {
      "epoch": 0.8967808930425752,
      "grad_norm": 0.5898131728172302,
      "learning_rate": 8e-05,
      "loss": 1.5094,
      "step": 4318
    },
    {
      "epoch": 0.8969885773624091,
      "grad_norm": 0.551611065864563,
      "learning_rate": 8e-05,
      "loss": 1.5412,
      "step": 4319
    },
    {
      "epoch": 0.897196261682243,
      "grad_norm": 0.5873730778694153,
      "learning_rate": 8e-05,
      "loss": 1.4668,
      "step": 4320
    },
    {
      "epoch": 0.8974039460020768,
      "grad_norm": 0.5961907505989075,
      "learning_rate": 8e-05,
      "loss": 1.5735,
      "step": 4321
    },
    {
      "epoch": 0.8976116303219107,
      "grad_norm": 0.5339446663856506,
      "learning_rate": 8e-05,
      "loss": 1.539,
      "step": 4322
    },
    {
      "epoch": 0.8978193146417446,
      "grad_norm": 0.5802297592163086,
      "learning_rate": 8e-05,
      "loss": 1.499,
      "step": 4323
    },
    {
      "epoch": 0.8980269989615784,
      "grad_norm": 0.5793024301528931,
      "learning_rate": 8e-05,
      "loss": 1.529,
      "step": 4324
    },
    {
      "epoch": 0.8982346832814122,
      "grad_norm": 0.5670417547225952,
      "learning_rate": 8e-05,
      "loss": 1.5418,
      "step": 4325
    },
    {
      "epoch": 0.8984423676012461,
      "grad_norm": 0.5319920778274536,
      "learning_rate": 8e-05,
      "loss": 1.5454,
      "step": 4326
    },
    {
      "epoch": 0.8986500519210799,
      "grad_norm": 0.5314199328422546,
      "learning_rate": 8e-05,
      "loss": 1.4856,
      "step": 4327
    },
    {
      "epoch": 0.8988577362409138,
      "grad_norm": 0.5784634947776794,
      "learning_rate": 8e-05,
      "loss": 1.5624,
      "step": 4328
    },
    {
      "epoch": 0.8990654205607477,
      "grad_norm": 0.5497854351997375,
      "learning_rate": 8e-05,
      "loss": 1.566,
      "step": 4329
    },
    {
      "epoch": 0.8992731048805815,
      "grad_norm": 0.5359938740730286,
      "learning_rate": 8e-05,
      "loss": 1.6024,
      "step": 4330
    },
    {
      "epoch": 0.8994807892004154,
      "grad_norm": 0.5193989276885986,
      "learning_rate": 8e-05,
      "loss": 1.4721,
      "step": 4331
    },
    {
      "epoch": 0.8996884735202493,
      "grad_norm": 0.5629150867462158,
      "learning_rate": 8e-05,
      "loss": 1.4769,
      "step": 4332
    },
    {
      "epoch": 0.899896157840083,
      "grad_norm": 0.5602332353591919,
      "learning_rate": 8e-05,
      "loss": 1.5702,
      "step": 4333
    },
    {
      "epoch": 0.9001038421599169,
      "grad_norm": 0.5657452344894409,
      "learning_rate": 8e-05,
      "loss": 1.5434,
      "step": 4334
    },
    {
      "epoch": 0.9003115264797508,
      "grad_norm": 0.5679588317871094,
      "learning_rate": 8e-05,
      "loss": 1.5983,
      "step": 4335
    },
    {
      "epoch": 0.9005192107995846,
      "grad_norm": 0.5438083410263062,
      "learning_rate": 8e-05,
      "loss": 1.4774,
      "step": 4336
    },
    {
      "epoch": 0.9007268951194185,
      "grad_norm": 0.5463825464248657,
      "learning_rate": 8e-05,
      "loss": 1.5478,
      "step": 4337
    },
    {
      "epoch": 0.9009345794392524,
      "grad_norm": 0.5970432162284851,
      "learning_rate": 8e-05,
      "loss": 1.5752,
      "step": 4338
    },
    {
      "epoch": 0.9011422637590862,
      "grad_norm": 0.5761677622795105,
      "learning_rate": 8e-05,
      "loss": 1.5506,
      "step": 4339
    },
    {
      "epoch": 0.90134994807892,
      "grad_norm": 0.6018775701522827,
      "learning_rate": 8e-05,
      "loss": 1.5708,
      "step": 4340
    },
    {
      "epoch": 0.9015576323987539,
      "grad_norm": 0.5392941832542419,
      "learning_rate": 8e-05,
      "loss": 1.5244,
      "step": 4341
    },
    {
      "epoch": 0.9017653167185877,
      "grad_norm": 0.6014353632926941,
      "learning_rate": 8e-05,
      "loss": 1.5635,
      "step": 4342
    },
    {
      "epoch": 0.9019730010384216,
      "grad_norm": 0.5207732319831848,
      "learning_rate": 8e-05,
      "loss": 1.517,
      "step": 4343
    },
    {
      "epoch": 0.9021806853582555,
      "grad_norm": 0.5324587225914001,
      "learning_rate": 8e-05,
      "loss": 1.5257,
      "step": 4344
    },
    {
      "epoch": 0.9023883696780893,
      "grad_norm": 0.562929093837738,
      "learning_rate": 8e-05,
      "loss": 1.547,
      "step": 4345
    },
    {
      "epoch": 0.9025960539979232,
      "grad_norm": 0.5457765460014343,
      "learning_rate": 8e-05,
      "loss": 1.5431,
      "step": 4346
    },
    {
      "epoch": 0.902803738317757,
      "grad_norm": 0.5520152449607849,
      "learning_rate": 8e-05,
      "loss": 1.5533,
      "step": 4347
    },
    {
      "epoch": 0.9030114226375908,
      "grad_norm": 0.5124415755271912,
      "learning_rate": 8e-05,
      "loss": 1.4612,
      "step": 4348
    },
    {
      "epoch": 0.9032191069574247,
      "grad_norm": 0.5637764930725098,
      "learning_rate": 8e-05,
      "loss": 1.5402,
      "step": 4349
    },
    {
      "epoch": 0.9034267912772586,
      "grad_norm": 0.58651202917099,
      "learning_rate": 8e-05,
      "loss": 1.5316,
      "step": 4350
    },
    {
      "epoch": 0.9036344755970924,
      "grad_norm": 0.5313969254493713,
      "learning_rate": 8e-05,
      "loss": 1.4415,
      "step": 4351
    },
    {
      "epoch": 0.9038421599169263,
      "grad_norm": 0.5260809659957886,
      "learning_rate": 8e-05,
      "loss": 1.4873,
      "step": 4352
    },
    {
      "epoch": 0.9040498442367602,
      "grad_norm": 0.5297077298164368,
      "learning_rate": 8e-05,
      "loss": 1.4806,
      "step": 4353
    },
    {
      "epoch": 0.904257528556594,
      "grad_norm": 0.6230749487876892,
      "learning_rate": 8e-05,
      "loss": 1.5616,
      "step": 4354
    },
    {
      "epoch": 0.9044652128764278,
      "grad_norm": 0.5617753267288208,
      "learning_rate": 8e-05,
      "loss": 1.4778,
      "step": 4355
    },
    {
      "epoch": 0.9046728971962616,
      "grad_norm": 0.5925421714782715,
      "learning_rate": 8e-05,
      "loss": 1.5476,
      "step": 4356
    },
    {
      "epoch": 0.9048805815160955,
      "grad_norm": 0.5179497599601746,
      "learning_rate": 8e-05,
      "loss": 1.4606,
      "step": 4357
    },
    {
      "epoch": 0.9050882658359294,
      "grad_norm": 0.5413150191307068,
      "learning_rate": 8e-05,
      "loss": 1.5119,
      "step": 4358
    },
    {
      "epoch": 0.9052959501557633,
      "grad_norm": 0.5487344264984131,
      "learning_rate": 8e-05,
      "loss": 1.4666,
      "step": 4359
    },
    {
      "epoch": 0.9055036344755971,
      "grad_norm": 0.5395278930664062,
      "learning_rate": 8e-05,
      "loss": 1.5207,
      "step": 4360
    },
    {
      "epoch": 0.905711318795431,
      "grad_norm": 0.5592114925384521,
      "learning_rate": 8e-05,
      "loss": 1.4212,
      "step": 4361
    },
    {
      "epoch": 0.9059190031152647,
      "grad_norm": 0.5624213218688965,
      "learning_rate": 8e-05,
      "loss": 1.5176,
      "step": 4362
    },
    {
      "epoch": 0.9061266874350986,
      "grad_norm": 0.5603268146514893,
      "learning_rate": 8e-05,
      "loss": 1.5094,
      "step": 4363
    },
    {
      "epoch": 0.9063343717549325,
      "grad_norm": 0.5356675982475281,
      "learning_rate": 8e-05,
      "loss": 1.5585,
      "step": 4364
    },
    {
      "epoch": 0.9065420560747663,
      "grad_norm": 0.5393280982971191,
      "learning_rate": 8e-05,
      "loss": 1.5073,
      "step": 4365
    },
    {
      "epoch": 0.9067497403946002,
      "grad_norm": 0.6184099316596985,
      "learning_rate": 8e-05,
      "loss": 1.5244,
      "step": 4366
    },
    {
      "epoch": 0.9069574247144341,
      "grad_norm": 0.5404500961303711,
      "learning_rate": 8e-05,
      "loss": 1.4774,
      "step": 4367
    },
    {
      "epoch": 0.907165109034268,
      "grad_norm": 0.5636690258979797,
      "learning_rate": 8e-05,
      "loss": 1.5598,
      "step": 4368
    },
    {
      "epoch": 0.9073727933541018,
      "grad_norm": 0.5397518277168274,
      "learning_rate": 8e-05,
      "loss": 1.5497,
      "step": 4369
    },
    {
      "epoch": 0.9075804776739356,
      "grad_norm": 0.5227951407432556,
      "learning_rate": 8e-05,
      "loss": 1.5628,
      "step": 4370
    },
    {
      "epoch": 0.9077881619937694,
      "grad_norm": 0.5551480650901794,
      "learning_rate": 8e-05,
      "loss": 1.5394,
      "step": 4371
    },
    {
      "epoch": 0.9079958463136033,
      "grad_norm": 0.5420199632644653,
      "learning_rate": 8e-05,
      "loss": 1.5732,
      "step": 4372
    },
    {
      "epoch": 0.9082035306334372,
      "grad_norm": 0.5501382946968079,
      "learning_rate": 8e-05,
      "loss": 1.53,
      "step": 4373
    },
    {
      "epoch": 0.908411214953271,
      "grad_norm": 0.5001388192176819,
      "learning_rate": 8e-05,
      "loss": 1.4737,
      "step": 4374
    },
    {
      "epoch": 0.9086188992731049,
      "grad_norm": 0.5779982805252075,
      "learning_rate": 8e-05,
      "loss": 1.5618,
      "step": 4375
    },
    {
      "epoch": 0.9088265835929388,
      "grad_norm": 0.5158848762512207,
      "learning_rate": 8e-05,
      "loss": 1.5257,
      "step": 4376
    },
    {
      "epoch": 0.9090342679127725,
      "grad_norm": 0.5292873382568359,
      "learning_rate": 8e-05,
      "loss": 1.5944,
      "step": 4377
    },
    {
      "epoch": 0.9092419522326064,
      "grad_norm": 0.5428850054740906,
      "learning_rate": 8e-05,
      "loss": 1.4838,
      "step": 4378
    },
    {
      "epoch": 0.9094496365524403,
      "grad_norm": 0.5470462441444397,
      "learning_rate": 8e-05,
      "loss": 1.5521,
      "step": 4379
    },
    {
      "epoch": 0.9096573208722741,
      "grad_norm": 0.573563814163208,
      "learning_rate": 8e-05,
      "loss": 1.556,
      "step": 4380
    },
    {
      "epoch": 0.909865005192108,
      "grad_norm": 0.5446181297302246,
      "learning_rate": 8e-05,
      "loss": 1.5504,
      "step": 4381
    },
    {
      "epoch": 0.9100726895119419,
      "grad_norm": 0.5328667759895325,
      "learning_rate": 8e-05,
      "loss": 1.4478,
      "step": 4382
    },
    {
      "epoch": 0.9102803738317757,
      "grad_norm": 0.5424851179122925,
      "learning_rate": 8e-05,
      "loss": 1.5631,
      "step": 4383
    },
    {
      "epoch": 0.9104880581516096,
      "grad_norm": 0.5432864427566528,
      "learning_rate": 8e-05,
      "loss": 1.5958,
      "step": 4384
    },
    {
      "epoch": 0.9106957424714434,
      "grad_norm": 0.5275771021842957,
      "learning_rate": 8e-05,
      "loss": 1.4609,
      "step": 4385
    },
    {
      "epoch": 0.9109034267912772,
      "grad_norm": 0.5476359128952026,
      "learning_rate": 8e-05,
      "loss": 1.5765,
      "step": 4386
    },
    {
      "epoch": 0.9111111111111111,
      "grad_norm": 0.54507976770401,
      "learning_rate": 8e-05,
      "loss": 1.5941,
      "step": 4387
    },
    {
      "epoch": 0.911318795430945,
      "grad_norm": 0.5382817983627319,
      "learning_rate": 8e-05,
      "loss": 1.5333,
      "step": 4388
    },
    {
      "epoch": 0.9115264797507788,
      "grad_norm": 0.5487110614776611,
      "learning_rate": 8e-05,
      "loss": 1.5397,
      "step": 4389
    },
    {
      "epoch": 0.9117341640706127,
      "grad_norm": 0.5642266869544983,
      "learning_rate": 8e-05,
      "loss": 1.5318,
      "step": 4390
    },
    {
      "epoch": 0.9119418483904466,
      "grad_norm": 0.5253781676292419,
      "learning_rate": 8e-05,
      "loss": 1.468,
      "step": 4391
    },
    {
      "epoch": 0.9121495327102803,
      "grad_norm": 0.5295657515525818,
      "learning_rate": 8e-05,
      "loss": 1.5042,
      "step": 4392
    },
    {
      "epoch": 0.9123572170301142,
      "grad_norm": 0.5437835454940796,
      "learning_rate": 8e-05,
      "loss": 1.5544,
      "step": 4393
    },
    {
      "epoch": 0.912564901349948,
      "grad_norm": 0.5340899229049683,
      "learning_rate": 8e-05,
      "loss": 1.4975,
      "step": 4394
    },
    {
      "epoch": 0.9127725856697819,
      "grad_norm": 0.535540759563446,
      "learning_rate": 8e-05,
      "loss": 1.504,
      "step": 4395
    },
    {
      "epoch": 0.9129802699896158,
      "grad_norm": 0.5558109283447266,
      "learning_rate": 8e-05,
      "loss": 1.5031,
      "step": 4396
    },
    {
      "epoch": 0.9131879543094497,
      "grad_norm": 0.5340763330459595,
      "learning_rate": 8e-05,
      "loss": 1.5445,
      "step": 4397
    },
    {
      "epoch": 0.9133956386292835,
      "grad_norm": 0.5924686789512634,
      "learning_rate": 8e-05,
      "loss": 1.5207,
      "step": 4398
    },
    {
      "epoch": 0.9136033229491174,
      "grad_norm": 0.5298249125480652,
      "learning_rate": 8e-05,
      "loss": 1.525,
      "step": 4399
    },
    {
      "epoch": 0.9138110072689511,
      "grad_norm": 0.5403388738632202,
      "learning_rate": 8e-05,
      "loss": 1.5024,
      "step": 4400
    },
    {
      "epoch": 0.914018691588785,
      "grad_norm": 0.5582671761512756,
      "learning_rate": 8e-05,
      "loss": 1.4987,
      "step": 4401
    },
    {
      "epoch": 0.9142263759086189,
      "grad_norm": 0.5745201110839844,
      "learning_rate": 8e-05,
      "loss": 1.6007,
      "step": 4402
    },
    {
      "epoch": 0.9144340602284527,
      "grad_norm": 0.5658930540084839,
      "learning_rate": 8e-05,
      "loss": 1.5337,
      "step": 4403
    },
    {
      "epoch": 0.9146417445482866,
      "grad_norm": 0.5523620247840881,
      "learning_rate": 8e-05,
      "loss": 1.5264,
      "step": 4404
    },
    {
      "epoch": 0.9148494288681205,
      "grad_norm": 0.6002260446548462,
      "learning_rate": 8e-05,
      "loss": 1.5587,
      "step": 4405
    },
    {
      "epoch": 0.9150571131879544,
      "grad_norm": 0.5515719056129456,
      "learning_rate": 8e-05,
      "loss": 1.5558,
      "step": 4406
    },
    {
      "epoch": 0.9152647975077881,
      "grad_norm": 0.5414637327194214,
      "learning_rate": 8e-05,
      "loss": 1.4874,
      "step": 4407
    },
    {
      "epoch": 0.915472481827622,
      "grad_norm": 0.5455270409584045,
      "learning_rate": 8e-05,
      "loss": 1.5753,
      "step": 4408
    },
    {
      "epoch": 0.9156801661474558,
      "grad_norm": 0.5329412221908569,
      "learning_rate": 8e-05,
      "loss": 1.5308,
      "step": 4409
    },
    {
      "epoch": 0.9158878504672897,
      "grad_norm": 0.5615268349647522,
      "learning_rate": 8e-05,
      "loss": 1.5608,
      "step": 4410
    },
    {
      "epoch": 0.9160955347871236,
      "grad_norm": 0.546230673789978,
      "learning_rate": 8e-05,
      "loss": 1.5059,
      "step": 4411
    },
    {
      "epoch": 0.9163032191069574,
      "grad_norm": 0.5376216173171997,
      "learning_rate": 8e-05,
      "loss": 1.5425,
      "step": 4412
    },
    {
      "epoch": 0.9165109034267913,
      "grad_norm": 0.5437529683113098,
      "learning_rate": 8e-05,
      "loss": 1.5276,
      "step": 4413
    },
    {
      "epoch": 0.9167185877466252,
      "grad_norm": 0.5437432527542114,
      "learning_rate": 8e-05,
      "loss": 1.5052,
      "step": 4414
    },
    {
      "epoch": 0.9169262720664589,
      "grad_norm": 0.5434216260910034,
      "learning_rate": 8e-05,
      "loss": 1.5452,
      "step": 4415
    },
    {
      "epoch": 0.9171339563862928,
      "grad_norm": 0.5337106585502625,
      "learning_rate": 8e-05,
      "loss": 1.5684,
      "step": 4416
    },
    {
      "epoch": 0.9173416407061267,
      "grad_norm": 0.5374414920806885,
      "learning_rate": 8e-05,
      "loss": 1.5376,
      "step": 4417
    },
    {
      "epoch": 0.9175493250259605,
      "grad_norm": 0.5668291449546814,
      "learning_rate": 8e-05,
      "loss": 1.502,
      "step": 4418
    },
    {
      "epoch": 0.9177570093457944,
      "grad_norm": 0.5484527945518494,
      "learning_rate": 8e-05,
      "loss": 1.575,
      "step": 4419
    },
    {
      "epoch": 0.9179646936656283,
      "grad_norm": 0.5591867566108704,
      "learning_rate": 8e-05,
      "loss": 1.4816,
      "step": 4420
    },
    {
      "epoch": 0.9181723779854621,
      "grad_norm": 0.5619500279426575,
      "learning_rate": 8e-05,
      "loss": 1.5321,
      "step": 4421
    },
    {
      "epoch": 0.9183800623052959,
      "grad_norm": 0.5561618208885193,
      "learning_rate": 8e-05,
      "loss": 1.4795,
      "step": 4422
    },
    {
      "epoch": 0.9185877466251298,
      "grad_norm": 0.5708007216453552,
      "learning_rate": 8e-05,
      "loss": 1.5205,
      "step": 4423
    },
    {
      "epoch": 0.9187954309449636,
      "grad_norm": 0.542363166809082,
      "learning_rate": 8e-05,
      "loss": 1.5682,
      "step": 4424
    },
    {
      "epoch": 0.9190031152647975,
      "grad_norm": 0.5808282494544983,
      "learning_rate": 8e-05,
      "loss": 1.5741,
      "step": 4425
    },
    {
      "epoch": 0.9192107995846314,
      "grad_norm": 0.5346624255180359,
      "learning_rate": 8e-05,
      "loss": 1.504,
      "step": 4426
    },
    {
      "epoch": 0.9194184839044652,
      "grad_norm": 0.5349721312522888,
      "learning_rate": 8e-05,
      "loss": 1.5264,
      "step": 4427
    },
    {
      "epoch": 0.9196261682242991,
      "grad_norm": 0.5734345316886902,
      "learning_rate": 8e-05,
      "loss": 1.5313,
      "step": 4428
    },
    {
      "epoch": 0.919833852544133,
      "grad_norm": 0.5072407722473145,
      "learning_rate": 8e-05,
      "loss": 1.5234,
      "step": 4429
    },
    {
      "epoch": 0.9200415368639667,
      "grad_norm": 0.554547131061554,
      "learning_rate": 8e-05,
      "loss": 1.5703,
      "step": 4430
    },
    {
      "epoch": 0.9202492211838006,
      "grad_norm": 0.5368897914886475,
      "learning_rate": 8e-05,
      "loss": 1.5646,
      "step": 4431
    },
    {
      "epoch": 0.9204569055036345,
      "grad_norm": 0.543056070804596,
      "learning_rate": 8e-05,
      "loss": 1.4451,
      "step": 4432
    },
    {
      "epoch": 0.9206645898234683,
      "grad_norm": 0.5691911578178406,
      "learning_rate": 8e-05,
      "loss": 1.5259,
      "step": 4433
    },
    {
      "epoch": 0.9208722741433022,
      "grad_norm": 0.5333191156387329,
      "learning_rate": 8e-05,
      "loss": 1.5741,
      "step": 4434
    },
    {
      "epoch": 0.9210799584631361,
      "grad_norm": 0.5495170950889587,
      "learning_rate": 8e-05,
      "loss": 1.5076,
      "step": 4435
    },
    {
      "epoch": 0.9212876427829699,
      "grad_norm": 0.5519574284553528,
      "learning_rate": 8e-05,
      "loss": 1.5244,
      "step": 4436
    },
    {
      "epoch": 0.9214953271028037,
      "grad_norm": 0.54426509141922,
      "learning_rate": 8e-05,
      "loss": 1.4975,
      "step": 4437
    },
    {
      "epoch": 0.9217030114226376,
      "grad_norm": 0.5710655450820923,
      "learning_rate": 8e-05,
      "loss": 1.6009,
      "step": 4438
    },
    {
      "epoch": 0.9219106957424714,
      "grad_norm": 0.592400312423706,
      "learning_rate": 8e-05,
      "loss": 1.5599,
      "step": 4439
    },
    {
      "epoch": 0.9221183800623053,
      "grad_norm": 0.5609350204467773,
      "learning_rate": 8e-05,
      "loss": 1.454,
      "step": 4440
    },
    {
      "epoch": 0.9223260643821392,
      "grad_norm": 0.6109335422515869,
      "learning_rate": 8e-05,
      "loss": 1.5065,
      "step": 4441
    },
    {
      "epoch": 0.922533748701973,
      "grad_norm": 0.5496782660484314,
      "learning_rate": 8e-05,
      "loss": 1.519,
      "step": 4442
    },
    {
      "epoch": 0.9227414330218069,
      "grad_norm": 0.519262433052063,
      "learning_rate": 8e-05,
      "loss": 1.4934,
      "step": 4443
    },
    {
      "epoch": 0.9229491173416408,
      "grad_norm": 0.5399003028869629,
      "learning_rate": 8e-05,
      "loss": 1.5021,
      "step": 4444
    },
    {
      "epoch": 0.9231568016614745,
      "grad_norm": 0.5618257522583008,
      "learning_rate": 8e-05,
      "loss": 1.5281,
      "step": 4445
    },
    {
      "epoch": 0.9233644859813084,
      "grad_norm": 0.5414952635765076,
      "learning_rate": 8e-05,
      "loss": 1.5209,
      "step": 4446
    },
    {
      "epoch": 0.9235721703011422,
      "grad_norm": 0.5440877079963684,
      "learning_rate": 8e-05,
      "loss": 1.5034,
      "step": 4447
    },
    {
      "epoch": 0.9237798546209761,
      "grad_norm": 0.537289023399353,
      "learning_rate": 8e-05,
      "loss": 1.5123,
      "step": 4448
    },
    {
      "epoch": 0.92398753894081,
      "grad_norm": 0.5339873433113098,
      "learning_rate": 8e-05,
      "loss": 1.5216,
      "step": 4449
    },
    {
      "epoch": 0.9241952232606438,
      "grad_norm": 0.6100795865058899,
      "learning_rate": 8e-05,
      "loss": 1.481,
      "step": 4450
    },
    {
      "epoch": 0.9244029075804777,
      "grad_norm": 0.5847253799438477,
      "learning_rate": 8e-05,
      "loss": 1.5843,
      "step": 4451
    },
    {
      "epoch": 0.9246105919003115,
      "grad_norm": 0.5441491007804871,
      "learning_rate": 8e-05,
      "loss": 1.5868,
      "step": 4452
    },
    {
      "epoch": 0.9248182762201453,
      "grad_norm": 0.5502142906188965,
      "learning_rate": 8e-05,
      "loss": 1.513,
      "step": 4453
    },
    {
      "epoch": 0.9250259605399792,
      "grad_norm": 0.6175906658172607,
      "learning_rate": 8e-05,
      "loss": 1.5046,
      "step": 4454
    },
    {
      "epoch": 0.9252336448598131,
      "grad_norm": 0.5425466895103455,
      "learning_rate": 8e-05,
      "loss": 1.5054,
      "step": 4455
    },
    {
      "epoch": 0.9254413291796469,
      "grad_norm": 0.5472467541694641,
      "learning_rate": 8e-05,
      "loss": 1.527,
      "step": 4456
    },
    {
      "epoch": 0.9256490134994808,
      "grad_norm": 0.5165514945983887,
      "learning_rate": 8e-05,
      "loss": 1.4357,
      "step": 4457
    },
    {
      "epoch": 0.9258566978193147,
      "grad_norm": 0.5451850295066833,
      "learning_rate": 8e-05,
      "loss": 1.5286,
      "step": 4458
    },
    {
      "epoch": 0.9260643821391485,
      "grad_norm": 0.5393959879875183,
      "learning_rate": 8e-05,
      "loss": 1.4989,
      "step": 4459
    },
    {
      "epoch": 0.9262720664589823,
      "grad_norm": 0.5985507965087891,
      "learning_rate": 8e-05,
      "loss": 1.4766,
      "step": 4460
    },
    {
      "epoch": 0.9264797507788162,
      "grad_norm": 0.5382755994796753,
      "learning_rate": 8e-05,
      "loss": 1.4557,
      "step": 4461
    },
    {
      "epoch": 0.92668743509865,
      "grad_norm": 0.5683870315551758,
      "learning_rate": 8e-05,
      "loss": 1.5942,
      "step": 4462
    },
    {
      "epoch": 0.9268951194184839,
      "grad_norm": 0.542346715927124,
      "learning_rate": 8e-05,
      "loss": 1.4652,
      "step": 4463
    },
    {
      "epoch": 0.9271028037383178,
      "grad_norm": 0.5361062288284302,
      "learning_rate": 8e-05,
      "loss": 1.4873,
      "step": 4464
    },
    {
      "epoch": 0.9273104880581516,
      "grad_norm": 0.5513811707496643,
      "learning_rate": 8e-05,
      "loss": 1.5715,
      "step": 4465
    },
    {
      "epoch": 0.9275181723779855,
      "grad_norm": 0.5389137268066406,
      "learning_rate": 8e-05,
      "loss": 1.5352,
      "step": 4466
    },
    {
      "epoch": 0.9277258566978193,
      "grad_norm": 0.5426903963088989,
      "learning_rate": 8e-05,
      "loss": 1.5711,
      "step": 4467
    },
    {
      "epoch": 0.9279335410176531,
      "grad_norm": 0.516876757144928,
      "learning_rate": 8e-05,
      "loss": 1.4457,
      "step": 4468
    },
    {
      "epoch": 0.928141225337487,
      "grad_norm": 0.5738803744316101,
      "learning_rate": 8e-05,
      "loss": 1.4831,
      "step": 4469
    },
    {
      "epoch": 0.9283489096573209,
      "grad_norm": 0.6255513429641724,
      "learning_rate": 8e-05,
      "loss": 1.4949,
      "step": 4470
    },
    {
      "epoch": 0.9285565939771547,
      "grad_norm": 0.5377402305603027,
      "learning_rate": 8e-05,
      "loss": 1.5458,
      "step": 4471
    },
    {
      "epoch": 0.9287642782969886,
      "grad_norm": 0.5525277256965637,
      "learning_rate": 8e-05,
      "loss": 1.5709,
      "step": 4472
    },
    {
      "epoch": 0.9289719626168225,
      "grad_norm": 0.6273601651191711,
      "learning_rate": 8e-05,
      "loss": 1.5867,
      "step": 4473
    },
    {
      "epoch": 0.9291796469366563,
      "grad_norm": 0.567503035068512,
      "learning_rate": 8e-05,
      "loss": 1.4733,
      "step": 4474
    },
    {
      "epoch": 0.9293873312564901,
      "grad_norm": 0.5550807118415833,
      "learning_rate": 8e-05,
      "loss": 1.5185,
      "step": 4475
    },
    {
      "epoch": 0.929595015576324,
      "grad_norm": 0.5302478671073914,
      "learning_rate": 8e-05,
      "loss": 1.5215,
      "step": 4476
    },
    {
      "epoch": 0.9298026998961578,
      "grad_norm": 0.557488203048706,
      "learning_rate": 8e-05,
      "loss": 1.4496,
      "step": 4477
    },
    {
      "epoch": 0.9300103842159917,
      "grad_norm": 0.5561457872390747,
      "learning_rate": 8e-05,
      "loss": 1.5619,
      "step": 4478
    },
    {
      "epoch": 0.9302180685358256,
      "grad_norm": 0.5618305802345276,
      "learning_rate": 8e-05,
      "loss": 1.5685,
      "step": 4479
    },
    {
      "epoch": 0.9304257528556594,
      "grad_norm": 0.5437638759613037,
      "learning_rate": 8e-05,
      "loss": 1.5646,
      "step": 4480
    },
    {
      "epoch": 0.9306334371754933,
      "grad_norm": 0.5747361183166504,
      "learning_rate": 8e-05,
      "loss": 1.5585,
      "step": 4481
    },
    {
      "epoch": 0.930841121495327,
      "grad_norm": 0.548672080039978,
      "learning_rate": 8e-05,
      "loss": 1.5361,
      "step": 4482
    },
    {
      "epoch": 0.9310488058151609,
      "grad_norm": 0.5425190329551697,
      "learning_rate": 8e-05,
      "loss": 1.5355,
      "step": 4483
    },
    {
      "epoch": 0.9312564901349948,
      "grad_norm": 0.5269076824188232,
      "learning_rate": 8e-05,
      "loss": 1.527,
      "step": 4484
    },
    {
      "epoch": 0.9314641744548287,
      "grad_norm": 0.5408453941345215,
      "learning_rate": 8e-05,
      "loss": 1.5095,
      "step": 4485
    },
    {
      "epoch": 0.9316718587746625,
      "grad_norm": 0.5515931844711304,
      "learning_rate": 8e-05,
      "loss": 1.4304,
      "step": 4486
    },
    {
      "epoch": 0.9318795430944964,
      "grad_norm": 0.5591109991073608,
      "learning_rate": 8e-05,
      "loss": 1.5628,
      "step": 4487
    },
    {
      "epoch": 0.9320872274143303,
      "grad_norm": 0.5728244781494141,
      "learning_rate": 8e-05,
      "loss": 1.5472,
      "step": 4488
    },
    {
      "epoch": 0.9322949117341641,
      "grad_norm": 0.5686746835708618,
      "learning_rate": 8e-05,
      "loss": 1.4806,
      "step": 4489
    },
    {
      "epoch": 0.9325025960539979,
      "grad_norm": 0.6569135189056396,
      "learning_rate": 8e-05,
      "loss": 1.572,
      "step": 4490
    },
    {
      "epoch": 0.9327102803738317,
      "grad_norm": 0.5510392189025879,
      "learning_rate": 8e-05,
      "loss": 1.4913,
      "step": 4491
    },
    {
      "epoch": 0.9329179646936656,
      "grad_norm": 0.5780843496322632,
      "learning_rate": 8e-05,
      "loss": 1.5374,
      "step": 4492
    },
    {
      "epoch": 0.9331256490134995,
      "grad_norm": 0.655059814453125,
      "learning_rate": 8e-05,
      "loss": 1.5369,
      "step": 4493
    },
    {
      "epoch": 0.9333333333333333,
      "grad_norm": 0.5716407299041748,
      "learning_rate": 8e-05,
      "loss": 1.5187,
      "step": 4494
    },
    {
      "epoch": 0.9335410176531672,
      "grad_norm": 0.5765836238861084,
      "learning_rate": 8e-05,
      "loss": 1.5184,
      "step": 4495
    },
    {
      "epoch": 0.9337487019730011,
      "grad_norm": 0.5645720362663269,
      "learning_rate": 8e-05,
      "loss": 1.5138,
      "step": 4496
    },
    {
      "epoch": 0.9339563862928348,
      "grad_norm": 0.6273457407951355,
      "learning_rate": 8e-05,
      "loss": 1.5191,
      "step": 4497
    },
    {
      "epoch": 0.9341640706126687,
      "grad_norm": 0.6928898692131042,
      "learning_rate": 8e-05,
      "loss": 1.5446,
      "step": 4498
    },
    {
      "epoch": 0.9343717549325026,
      "grad_norm": 0.571147620677948,
      "learning_rate": 8e-05,
      "loss": 1.5265,
      "step": 4499
    },
    {
      "epoch": 0.9345794392523364,
      "grad_norm": 0.5683586597442627,
      "learning_rate": 8e-05,
      "loss": 1.5299,
      "step": 4500
    },
    {
      "epoch": 0.9347871235721703,
      "grad_norm": 0.6003159880638123,
      "learning_rate": 8e-05,
      "loss": 1.5022,
      "step": 4501
    },
    {
      "epoch": 0.9349948078920042,
      "grad_norm": 0.5484350323677063,
      "learning_rate": 8e-05,
      "loss": 1.5154,
      "step": 4502
    },
    {
      "epoch": 0.935202492211838,
      "grad_norm": 0.5624589920043945,
      "learning_rate": 8e-05,
      "loss": 1.6002,
      "step": 4503
    },
    {
      "epoch": 0.9354101765316719,
      "grad_norm": 0.552061140537262,
      "learning_rate": 8e-05,
      "loss": 1.5362,
      "step": 4504
    },
    {
      "epoch": 0.9356178608515057,
      "grad_norm": 0.5790058374404907,
      "learning_rate": 8e-05,
      "loss": 1.578,
      "step": 4505
    },
    {
      "epoch": 0.9358255451713395,
      "grad_norm": 0.6513620615005493,
      "learning_rate": 8e-05,
      "loss": 1.5291,
      "step": 4506
    },
    {
      "epoch": 0.9360332294911734,
      "grad_norm": 0.548109233379364,
      "learning_rate": 8e-05,
      "loss": 1.5462,
      "step": 4507
    },
    {
      "epoch": 0.9362409138110073,
      "grad_norm": 0.5415007472038269,
      "learning_rate": 8e-05,
      "loss": 1.5293,
      "step": 4508
    },
    {
      "epoch": 0.9364485981308411,
      "grad_norm": 0.5607280135154724,
      "learning_rate": 8e-05,
      "loss": 1.5294,
      "step": 4509
    },
    {
      "epoch": 0.936656282450675,
      "grad_norm": 0.5666238069534302,
      "learning_rate": 8e-05,
      "loss": 1.5276,
      "step": 4510
    },
    {
      "epoch": 0.9368639667705089,
      "grad_norm": 0.5420122146606445,
      "learning_rate": 8e-05,
      "loss": 1.5848,
      "step": 4511
    },
    {
      "epoch": 0.9370716510903426,
      "grad_norm": 0.6130264401435852,
      "learning_rate": 8e-05,
      "loss": 1.4893,
      "step": 4512
    },
    {
      "epoch": 0.9372793354101765,
      "grad_norm": 0.5671524405479431,
      "learning_rate": 8e-05,
      "loss": 1.5968,
      "step": 4513
    },
    {
      "epoch": 0.9374870197300104,
      "grad_norm": 0.5798551440238953,
      "learning_rate": 8e-05,
      "loss": 1.4488,
      "step": 4514
    },
    {
      "epoch": 0.9376947040498442,
      "grad_norm": 0.5973544716835022,
      "learning_rate": 8e-05,
      "loss": 1.4635,
      "step": 4515
    },
    {
      "epoch": 0.9379023883696781,
      "grad_norm": 0.6140970587730408,
      "learning_rate": 8e-05,
      "loss": 1.5589,
      "step": 4516
    },
    {
      "epoch": 0.938110072689512,
      "grad_norm": 0.5505404472351074,
      "learning_rate": 8e-05,
      "loss": 1.5056,
      "step": 4517
    },
    {
      "epoch": 0.9383177570093458,
      "grad_norm": 0.6027393937110901,
      "learning_rate": 8e-05,
      "loss": 1.4741,
      "step": 4518
    },
    {
      "epoch": 0.9385254413291797,
      "grad_norm": 0.593338131904602,
      "learning_rate": 8e-05,
      "loss": 1.4517,
      "step": 4519
    },
    {
      "epoch": 0.9387331256490135,
      "grad_norm": 0.5911040306091309,
      "learning_rate": 8e-05,
      "loss": 1.5108,
      "step": 4520
    },
    {
      "epoch": 0.9389408099688473,
      "grad_norm": 0.5398374199867249,
      "learning_rate": 8e-05,
      "loss": 1.5037,
      "step": 4521
    },
    {
      "epoch": 0.9391484942886812,
      "grad_norm": 0.5704691410064697,
      "learning_rate": 8e-05,
      "loss": 1.4943,
      "step": 4522
    },
    {
      "epoch": 0.9393561786085151,
      "grad_norm": 0.5514391660690308,
      "learning_rate": 8e-05,
      "loss": 1.5292,
      "step": 4523
    },
    {
      "epoch": 0.9395638629283489,
      "grad_norm": 0.527151882648468,
      "learning_rate": 8e-05,
      "loss": 1.478,
      "step": 4524
    },
    {
      "epoch": 0.9397715472481828,
      "grad_norm": 0.5545868873596191,
      "learning_rate": 8e-05,
      "loss": 1.5428,
      "step": 4525
    },
    {
      "epoch": 0.9399792315680167,
      "grad_norm": 0.5245705842971802,
      "learning_rate": 8e-05,
      "loss": 1.5215,
      "step": 4526
    },
    {
      "epoch": 0.9401869158878504,
      "grad_norm": 0.5282145738601685,
      "learning_rate": 8e-05,
      "loss": 1.4492,
      "step": 4527
    },
    {
      "epoch": 0.9403946002076843,
      "grad_norm": 0.6142186522483826,
      "learning_rate": 8e-05,
      "loss": 1.4863,
      "step": 4528
    },
    {
      "epoch": 0.9406022845275182,
      "grad_norm": 0.6031098365783691,
      "learning_rate": 8e-05,
      "loss": 1.4711,
      "step": 4529
    },
    {
      "epoch": 0.940809968847352,
      "grad_norm": 0.5368979573249817,
      "learning_rate": 8e-05,
      "loss": 1.4869,
      "step": 4530
    },
    {
      "epoch": 0.9410176531671859,
      "grad_norm": 0.5586937665939331,
      "learning_rate": 8e-05,
      "loss": 1.509,
      "step": 4531
    },
    {
      "epoch": 0.9412253374870198,
      "grad_norm": 0.588105320930481,
      "learning_rate": 8e-05,
      "loss": 1.571,
      "step": 4532
    },
    {
      "epoch": 0.9414330218068536,
      "grad_norm": 0.5683290958404541,
      "learning_rate": 8e-05,
      "loss": 1.5029,
      "step": 4533
    },
    {
      "epoch": 0.9416407061266875,
      "grad_norm": 0.5634755492210388,
      "learning_rate": 8e-05,
      "loss": 1.5357,
      "step": 4534
    },
    {
      "epoch": 0.9418483904465212,
      "grad_norm": 0.5526583790779114,
      "learning_rate": 8e-05,
      "loss": 1.6255,
      "step": 4535
    },
    {
      "epoch": 0.9420560747663551,
      "grad_norm": 0.5321868658065796,
      "learning_rate": 8e-05,
      "loss": 1.5576,
      "step": 4536
    },
    {
      "epoch": 0.942263759086189,
      "grad_norm": 0.5295841693878174,
      "learning_rate": 8e-05,
      "loss": 1.5685,
      "step": 4537
    },
    {
      "epoch": 0.9424714434060228,
      "grad_norm": 0.5981540083885193,
      "learning_rate": 8e-05,
      "loss": 1.5532,
      "step": 4538
    },
    {
      "epoch": 0.9426791277258567,
      "grad_norm": 0.5772187113761902,
      "learning_rate": 8e-05,
      "loss": 1.5262,
      "step": 4539
    },
    {
      "epoch": 0.9428868120456906,
      "grad_norm": 0.5469943881034851,
      "learning_rate": 8e-05,
      "loss": 1.4661,
      "step": 4540
    },
    {
      "epoch": 0.9430944963655244,
      "grad_norm": 0.5601049661636353,
      "learning_rate": 8e-05,
      "loss": 1.4734,
      "step": 4541
    },
    {
      "epoch": 0.9433021806853582,
      "grad_norm": 0.6580184698104858,
      "learning_rate": 8e-05,
      "loss": 1.5003,
      "step": 4542
    },
    {
      "epoch": 0.9435098650051921,
      "grad_norm": 0.5327724814414978,
      "learning_rate": 8e-05,
      "loss": 1.4434,
      "step": 4543
    },
    {
      "epoch": 0.9437175493250259,
      "grad_norm": 0.5281966328620911,
      "learning_rate": 8e-05,
      "loss": 1.5175,
      "step": 4544
    },
    {
      "epoch": 0.9439252336448598,
      "grad_norm": 0.5242224931716919,
      "learning_rate": 8e-05,
      "loss": 1.5698,
      "step": 4545
    },
    {
      "epoch": 0.9441329179646937,
      "grad_norm": 0.5960102081298828,
      "learning_rate": 8e-05,
      "loss": 1.5564,
      "step": 4546
    },
    {
      "epoch": 0.9443406022845275,
      "grad_norm": 0.5489022731781006,
      "learning_rate": 8e-05,
      "loss": 1.5211,
      "step": 4547
    },
    {
      "epoch": 0.9445482866043614,
      "grad_norm": 0.542113721370697,
      "learning_rate": 8e-05,
      "loss": 1.5183,
      "step": 4548
    },
    {
      "epoch": 0.9447559709241953,
      "grad_norm": 0.5541703104972839,
      "learning_rate": 8e-05,
      "loss": 1.4987,
      "step": 4549
    },
    {
      "epoch": 0.944963655244029,
      "grad_norm": 0.5534771084785461,
      "learning_rate": 8e-05,
      "loss": 1.4645,
      "step": 4550
    },
    {
      "epoch": 0.9451713395638629,
      "grad_norm": 0.5465671420097351,
      "learning_rate": 8e-05,
      "loss": 1.564,
      "step": 4551
    },
    {
      "epoch": 0.9453790238836968,
      "grad_norm": 0.5781581401824951,
      "learning_rate": 8e-05,
      "loss": 1.5331,
      "step": 4552
    },
    {
      "epoch": 0.9455867082035306,
      "grad_norm": 0.5479032397270203,
      "learning_rate": 8e-05,
      "loss": 1.6254,
      "step": 4553
    },
    {
      "epoch": 0.9457943925233645,
      "grad_norm": 0.5651295185089111,
      "learning_rate": 8e-05,
      "loss": 1.4843,
      "step": 4554
    },
    {
      "epoch": 0.9460020768431984,
      "grad_norm": 0.5894498229026794,
      "learning_rate": 8e-05,
      "loss": 1.5327,
      "step": 4555
    },
    {
      "epoch": 0.9462097611630322,
      "grad_norm": 0.5756381154060364,
      "learning_rate": 8e-05,
      "loss": 1.608,
      "step": 4556
    },
    {
      "epoch": 0.946417445482866,
      "grad_norm": 0.5424879193305969,
      "learning_rate": 8e-05,
      "loss": 1.5355,
      "step": 4557
    },
    {
      "epoch": 0.9466251298026999,
      "grad_norm": 0.5302530527114868,
      "learning_rate": 8e-05,
      "loss": 1.4851,
      "step": 4558
    },
    {
      "epoch": 0.9468328141225337,
      "grad_norm": 0.5629392266273499,
      "learning_rate": 8e-05,
      "loss": 1.5498,
      "step": 4559
    },
    {
      "epoch": 0.9470404984423676,
      "grad_norm": 0.557964563369751,
      "learning_rate": 8e-05,
      "loss": 1.5461,
      "step": 4560
    },
    {
      "epoch": 0.9472481827622015,
      "grad_norm": 0.5579253435134888,
      "learning_rate": 8e-05,
      "loss": 1.5257,
      "step": 4561
    },
    {
      "epoch": 0.9474558670820353,
      "grad_norm": 0.5240344405174255,
      "learning_rate": 8e-05,
      "loss": 1.5048,
      "step": 4562
    },
    {
      "epoch": 0.9476635514018692,
      "grad_norm": 0.5323419570922852,
      "learning_rate": 8e-05,
      "loss": 1.5354,
      "step": 4563
    },
    {
      "epoch": 0.9478712357217031,
      "grad_norm": 0.5428701639175415,
      "learning_rate": 8e-05,
      "loss": 1.5025,
      "step": 4564
    },
    {
      "epoch": 0.9480789200415368,
      "grad_norm": 0.5244649648666382,
      "learning_rate": 8e-05,
      "loss": 1.5005,
      "step": 4565
    },
    {
      "epoch": 0.9482866043613707,
      "grad_norm": 0.5624781847000122,
      "learning_rate": 8e-05,
      "loss": 1.5323,
      "step": 4566
    },
    {
      "epoch": 0.9484942886812046,
      "grad_norm": 0.5645930171012878,
      "learning_rate": 8e-05,
      "loss": 1.5791,
      "step": 4567
    },
    {
      "epoch": 0.9487019730010384,
      "grad_norm": 0.5441578030586243,
      "learning_rate": 8e-05,
      "loss": 1.5371,
      "step": 4568
    },
    {
      "epoch": 0.9489096573208723,
      "grad_norm": 0.5477387309074402,
      "learning_rate": 8e-05,
      "loss": 1.4983,
      "step": 4569
    },
    {
      "epoch": 0.9491173416407062,
      "grad_norm": 0.5662717223167419,
      "learning_rate": 8e-05,
      "loss": 1.5482,
      "step": 4570
    },
    {
      "epoch": 0.94932502596054,
      "grad_norm": 0.5493006706237793,
      "learning_rate": 8e-05,
      "loss": 1.5535,
      "step": 4571
    },
    {
      "epoch": 0.9495327102803738,
      "grad_norm": 0.554337739944458,
      "learning_rate": 8e-05,
      "loss": 1.5147,
      "step": 4572
    },
    {
      "epoch": 0.9497403946002076,
      "grad_norm": 0.5474506616592407,
      "learning_rate": 8e-05,
      "loss": 1.4468,
      "step": 4573
    },
    {
      "epoch": 0.9499480789200415,
      "grad_norm": 0.5508801937103271,
      "learning_rate": 8e-05,
      "loss": 1.5426,
      "step": 4574
    },
    {
      "epoch": 0.9501557632398754,
      "grad_norm": 0.5581814050674438,
      "learning_rate": 8e-05,
      "loss": 1.5092,
      "step": 4575
    },
    {
      "epoch": 0.9503634475597093,
      "grad_norm": 0.5851505398750305,
      "learning_rate": 8e-05,
      "loss": 1.5833,
      "step": 4576
    },
    {
      "epoch": 0.9505711318795431,
      "grad_norm": 0.5425330400466919,
      "learning_rate": 8e-05,
      "loss": 1.4815,
      "step": 4577
    },
    {
      "epoch": 0.950778816199377,
      "grad_norm": 0.5892306566238403,
      "learning_rate": 8e-05,
      "loss": 1.5234,
      "step": 4578
    },
    {
      "epoch": 0.9509865005192109,
      "grad_norm": 0.5555946230888367,
      "learning_rate": 8e-05,
      "loss": 1.5211,
      "step": 4579
    },
    {
      "epoch": 0.9511941848390446,
      "grad_norm": 0.5682836174964905,
      "learning_rate": 8e-05,
      "loss": 1.5057,
      "step": 4580
    },
    {
      "epoch": 0.9514018691588785,
      "grad_norm": 0.5369886755943298,
      "learning_rate": 8e-05,
      "loss": 1.5753,
      "step": 4581
    },
    {
      "epoch": 0.9516095534787123,
      "grad_norm": 0.5447344779968262,
      "learning_rate": 8e-05,
      "loss": 1.5113,
      "step": 4582
    },
    {
      "epoch": 0.9518172377985462,
      "grad_norm": 0.5620760321617126,
      "learning_rate": 8e-05,
      "loss": 1.5926,
      "step": 4583
    },
    {
      "epoch": 0.9520249221183801,
      "grad_norm": 0.5205639600753784,
      "learning_rate": 8e-05,
      "loss": 1.4993,
      "step": 4584
    },
    {
      "epoch": 0.952232606438214,
      "grad_norm": 0.5516989827156067,
      "learning_rate": 8e-05,
      "loss": 1.5532,
      "step": 4585
    },
    {
      "epoch": 0.9524402907580478,
      "grad_norm": 0.5953296422958374,
      "learning_rate": 8e-05,
      "loss": 1.5927,
      "step": 4586
    },
    {
      "epoch": 0.9526479750778816,
      "grad_norm": 0.5737614035606384,
      "learning_rate": 8e-05,
      "loss": 1.5741,
      "step": 4587
    },
    {
      "epoch": 0.9528556593977154,
      "grad_norm": 0.5736824870109558,
      "learning_rate": 8e-05,
      "loss": 1.5254,
      "step": 4588
    },
    {
      "epoch": 0.9530633437175493,
      "grad_norm": 0.6801788210868835,
      "learning_rate": 8e-05,
      "loss": 1.5045,
      "step": 4589
    },
    {
      "epoch": 0.9532710280373832,
      "grad_norm": 0.5624447464942932,
      "learning_rate": 8e-05,
      "loss": 1.4792,
      "step": 4590
    },
    {
      "epoch": 0.953478712357217,
      "grad_norm": 0.533266544342041,
      "learning_rate": 8e-05,
      "loss": 1.466,
      "step": 4591
    },
    {
      "epoch": 0.9536863966770509,
      "grad_norm": 0.6132524013519287,
      "learning_rate": 8e-05,
      "loss": 1.5328,
      "step": 4592
    },
    {
      "epoch": 0.9538940809968848,
      "grad_norm": 0.5733515620231628,
      "learning_rate": 8e-05,
      "loss": 1.4923,
      "step": 4593
    },
    {
      "epoch": 0.9541017653167186,
      "grad_norm": 0.576765775680542,
      "learning_rate": 8e-05,
      "loss": 1.592,
      "step": 4594
    },
    {
      "epoch": 0.9543094496365524,
      "grad_norm": 0.5176913738250732,
      "learning_rate": 8e-05,
      "loss": 1.4767,
      "step": 4595
    },
    {
      "epoch": 0.9545171339563863,
      "grad_norm": 0.5590279698371887,
      "learning_rate": 8e-05,
      "loss": 1.4966,
      "step": 4596
    },
    {
      "epoch": 0.9547248182762201,
      "grad_norm": 0.6225258111953735,
      "learning_rate": 8e-05,
      "loss": 1.5701,
      "step": 4597
    },
    {
      "epoch": 0.954932502596054,
      "grad_norm": 0.5392278432846069,
      "learning_rate": 8e-05,
      "loss": 1.4582,
      "step": 4598
    },
    {
      "epoch": 0.9551401869158879,
      "grad_norm": 0.5552417635917664,
      "learning_rate": 8e-05,
      "loss": 1.5608,
      "step": 4599
    },
    {
      "epoch": 0.9553478712357217,
      "grad_norm": 0.5463722944259644,
      "learning_rate": 8e-05,
      "loss": 1.5261,
      "step": 4600
    },
    {
      "epoch": 0.9555555555555556,
      "grad_norm": 0.5532234311103821,
      "learning_rate": 8e-05,
      "loss": 1.5028,
      "step": 4601
    },
    {
      "epoch": 0.9557632398753894,
      "grad_norm": 0.5900952219963074,
      "learning_rate": 8e-05,
      "loss": 1.5405,
      "step": 4602
    },
    {
      "epoch": 0.9559709241952232,
      "grad_norm": 0.5493490099906921,
      "learning_rate": 8e-05,
      "loss": 1.5245,
      "step": 4603
    },
    {
      "epoch": 0.9561786085150571,
      "grad_norm": 0.539199948310852,
      "learning_rate": 8e-05,
      "loss": 1.5012,
      "step": 4604
    },
    {
      "epoch": 0.956386292834891,
      "grad_norm": 0.6017288565635681,
      "learning_rate": 8e-05,
      "loss": 1.6017,
      "step": 4605
    },
    {
      "epoch": 0.9565939771547248,
      "grad_norm": 0.5687580108642578,
      "learning_rate": 8e-05,
      "loss": 1.5919,
      "step": 4606
    },
    {
      "epoch": 0.9568016614745587,
      "grad_norm": 0.5285506844520569,
      "learning_rate": 8e-05,
      "loss": 1.484,
      "step": 4607
    },
    {
      "epoch": 0.9570093457943926,
      "grad_norm": 0.5265766382217407,
      "learning_rate": 8e-05,
      "loss": 1.5157,
      "step": 4608
    },
    {
      "epoch": 0.9572170301142264,
      "grad_norm": 0.539530336856842,
      "learning_rate": 8e-05,
      "loss": 1.4869,
      "step": 4609
    },
    {
      "epoch": 0.9574247144340602,
      "grad_norm": 0.6017909049987793,
      "learning_rate": 8e-05,
      "loss": 1.5435,
      "step": 4610
    },
    {
      "epoch": 0.957632398753894,
      "grad_norm": 0.5629965662956238,
      "learning_rate": 8e-05,
      "loss": 1.5041,
      "step": 4611
    },
    {
      "epoch": 0.9578400830737279,
      "grad_norm": 0.5457871556282043,
      "learning_rate": 8e-05,
      "loss": 1.5096,
      "step": 4612
    },
    {
      "epoch": 0.9580477673935618,
      "grad_norm": 0.5930299758911133,
      "learning_rate": 8e-05,
      "loss": 1.5687,
      "step": 4613
    },
    {
      "epoch": 0.9582554517133957,
      "grad_norm": 0.5691019296646118,
      "learning_rate": 8e-05,
      "loss": 1.4932,
      "step": 4614
    },
    {
      "epoch": 0.9584631360332295,
      "grad_norm": 0.5549120903015137,
      "learning_rate": 8e-05,
      "loss": 1.4996,
      "step": 4615
    },
    {
      "epoch": 0.9586708203530634,
      "grad_norm": 0.5707115530967712,
      "learning_rate": 8e-05,
      "loss": 1.5067,
      "step": 4616
    },
    {
      "epoch": 0.9588785046728971,
      "grad_norm": 0.5605545043945312,
      "learning_rate": 8e-05,
      "loss": 1.564,
      "step": 4617
    },
    {
      "epoch": 0.959086188992731,
      "grad_norm": 0.5548222064971924,
      "learning_rate": 8e-05,
      "loss": 1.4879,
      "step": 4618
    },
    {
      "epoch": 0.9592938733125649,
      "grad_norm": 0.5707319974899292,
      "learning_rate": 8e-05,
      "loss": 1.504,
      "step": 4619
    },
    {
      "epoch": 0.9595015576323987,
      "grad_norm": 0.5498775243759155,
      "learning_rate": 8e-05,
      "loss": 1.5439,
      "step": 4620
    },
    {
      "epoch": 0.9597092419522326,
      "grad_norm": 0.5459156632423401,
      "learning_rate": 8e-05,
      "loss": 1.4602,
      "step": 4621
    },
    {
      "epoch": 0.9599169262720665,
      "grad_norm": 0.5805017352104187,
      "learning_rate": 8e-05,
      "loss": 1.548,
      "step": 4622
    },
    {
      "epoch": 0.9601246105919004,
      "grad_norm": 0.5641297698020935,
      "learning_rate": 8e-05,
      "loss": 1.4584,
      "step": 4623
    },
    {
      "epoch": 0.9603322949117342,
      "grad_norm": 0.5321623086929321,
      "learning_rate": 8e-05,
      "loss": 1.4287,
      "step": 4624
    },
    {
      "epoch": 0.960539979231568,
      "grad_norm": 0.547888457775116,
      "learning_rate": 8e-05,
      "loss": 1.5716,
      "step": 4625
    },
    {
      "epoch": 0.9607476635514018,
      "grad_norm": 0.546263575553894,
      "learning_rate": 8e-05,
      "loss": 1.4446,
      "step": 4626
    },
    {
      "epoch": 0.9609553478712357,
      "grad_norm": 0.5693860650062561,
      "learning_rate": 8e-05,
      "loss": 1.5564,
      "step": 4627
    },
    {
      "epoch": 0.9611630321910696,
      "grad_norm": 0.5444315671920776,
      "learning_rate": 8e-05,
      "loss": 1.523,
      "step": 4628
    },
    {
      "epoch": 0.9613707165109034,
      "grad_norm": 0.5783362984657288,
      "learning_rate": 8e-05,
      "loss": 1.4887,
      "step": 4629
    },
    {
      "epoch": 0.9615784008307373,
      "grad_norm": 0.5789819955825806,
      "learning_rate": 8e-05,
      "loss": 1.5055,
      "step": 4630
    },
    {
      "epoch": 0.9617860851505712,
      "grad_norm": 0.5203589797019958,
      "learning_rate": 8e-05,
      "loss": 1.44,
      "step": 4631
    },
    {
      "epoch": 0.9619937694704049,
      "grad_norm": 0.5812758803367615,
      "learning_rate": 8e-05,
      "loss": 1.5315,
      "step": 4632
    },
    {
      "epoch": 0.9622014537902388,
      "grad_norm": 0.5280937552452087,
      "learning_rate": 8e-05,
      "loss": 1.44,
      "step": 4633
    },
    {
      "epoch": 0.9624091381100727,
      "grad_norm": 0.5395685434341431,
      "learning_rate": 8e-05,
      "loss": 1.494,
      "step": 4634
    },
    {
      "epoch": 0.9626168224299065,
      "grad_norm": 0.5728075504302979,
      "learning_rate": 8e-05,
      "loss": 1.516,
      "step": 4635
    },
    {
      "epoch": 0.9628245067497404,
      "grad_norm": 0.5547771453857422,
      "learning_rate": 8e-05,
      "loss": 1.4463,
      "step": 4636
    },
    {
      "epoch": 0.9630321910695743,
      "grad_norm": 0.5561551451683044,
      "learning_rate": 8e-05,
      "loss": 1.4583,
      "step": 4637
    },
    {
      "epoch": 0.9632398753894081,
      "grad_norm": 0.5641451478004456,
      "learning_rate": 8e-05,
      "loss": 1.5258,
      "step": 4638
    },
    {
      "epoch": 0.963447559709242,
      "grad_norm": 0.5466979742050171,
      "learning_rate": 8e-05,
      "loss": 1.4964,
      "step": 4639
    },
    {
      "epoch": 0.9636552440290758,
      "grad_norm": 0.5529464483261108,
      "learning_rate": 8e-05,
      "loss": 1.4855,
      "step": 4640
    },
    {
      "epoch": 0.9638629283489096,
      "grad_norm": 0.5496307015419006,
      "learning_rate": 8e-05,
      "loss": 1.5591,
      "step": 4641
    },
    {
      "epoch": 0.9640706126687435,
      "grad_norm": 0.5581245422363281,
      "learning_rate": 8e-05,
      "loss": 1.5316,
      "step": 4642
    },
    {
      "epoch": 0.9642782969885774,
      "grad_norm": 0.5732630491256714,
      "learning_rate": 8e-05,
      "loss": 1.5269,
      "step": 4643
    },
    {
      "epoch": 0.9644859813084112,
      "grad_norm": 0.5870099067687988,
      "learning_rate": 8e-05,
      "loss": 1.5671,
      "step": 4644
    },
    {
      "epoch": 0.9646936656282451,
      "grad_norm": 0.5670344233512878,
      "learning_rate": 8e-05,
      "loss": 1.5692,
      "step": 4645
    },
    {
      "epoch": 0.964901349948079,
      "grad_norm": 0.5183953642845154,
      "learning_rate": 8e-05,
      "loss": 1.5139,
      "step": 4646
    },
    {
      "epoch": 0.9651090342679127,
      "grad_norm": 0.564356803894043,
      "learning_rate": 8e-05,
      "loss": 1.4599,
      "step": 4647
    },
    {
      "epoch": 0.9653167185877466,
      "grad_norm": 0.5357316136360168,
      "learning_rate": 8e-05,
      "loss": 1.4418,
      "step": 4648
    },
    {
      "epoch": 0.9655244029075805,
      "grad_norm": 0.5574904084205627,
      "learning_rate": 8e-05,
      "loss": 1.5288,
      "step": 4649
    },
    {
      "epoch": 0.9657320872274143,
      "grad_norm": 0.5518751740455627,
      "learning_rate": 8e-05,
      "loss": 1.5345,
      "step": 4650
    },
    {
      "epoch": 0.9659397715472482,
      "grad_norm": 0.5475333333015442,
      "learning_rate": 8e-05,
      "loss": 1.4866,
      "step": 4651
    },
    {
      "epoch": 0.9661474558670821,
      "grad_norm": 0.5753294825553894,
      "learning_rate": 8e-05,
      "loss": 1.4754,
      "step": 4652
    },
    {
      "epoch": 0.9663551401869159,
      "grad_norm": 0.5737543106079102,
      "learning_rate": 8e-05,
      "loss": 1.5574,
      "step": 4653
    },
    {
      "epoch": 0.9665628245067498,
      "grad_norm": 0.5426458716392517,
      "learning_rate": 8e-05,
      "loss": 1.5492,
      "step": 4654
    },
    {
      "epoch": 0.9667705088265836,
      "grad_norm": 0.5765553116798401,
      "learning_rate": 8e-05,
      "loss": 1.5528,
      "step": 4655
    },
    {
      "epoch": 0.9669781931464174,
      "grad_norm": 0.61495041847229,
      "learning_rate": 8e-05,
      "loss": 1.5649,
      "step": 4656
    },
    {
      "epoch": 0.9671858774662513,
      "grad_norm": 0.5621324777603149,
      "learning_rate": 8e-05,
      "loss": 1.5242,
      "step": 4657
    },
    {
      "epoch": 0.9673935617860852,
      "grad_norm": 0.5682986974716187,
      "learning_rate": 8e-05,
      "loss": 1.539,
      "step": 4658
    },
    {
      "epoch": 0.967601246105919,
      "grad_norm": 0.5477182865142822,
      "learning_rate": 8e-05,
      "loss": 1.4798,
      "step": 4659
    },
    {
      "epoch": 0.9678089304257529,
      "grad_norm": 0.5914707183837891,
      "learning_rate": 8e-05,
      "loss": 1.5529,
      "step": 4660
    },
    {
      "epoch": 0.9680166147455868,
      "grad_norm": 0.5926550626754761,
      "learning_rate": 8e-05,
      "loss": 1.5862,
      "step": 4661
    },
    {
      "epoch": 0.9682242990654205,
      "grad_norm": 0.5551722049713135,
      "learning_rate": 8e-05,
      "loss": 1.5539,
      "step": 4662
    },
    {
      "epoch": 0.9684319833852544,
      "grad_norm": 0.5900006294250488,
      "learning_rate": 8e-05,
      "loss": 1.5002,
      "step": 4663
    },
    {
      "epoch": 0.9686396677050882,
      "grad_norm": 0.5730093717575073,
      "learning_rate": 8e-05,
      "loss": 1.5863,
      "step": 4664
    },
    {
      "epoch": 0.9688473520249221,
      "grad_norm": 0.5292286276817322,
      "learning_rate": 8e-05,
      "loss": 1.4275,
      "step": 4665
    },
    {
      "epoch": 0.969055036344756,
      "grad_norm": 0.6158426403999329,
      "learning_rate": 8e-05,
      "loss": 1.5641,
      "step": 4666
    },
    {
      "epoch": 0.9692627206645898,
      "grad_norm": 0.5952554941177368,
      "learning_rate": 8e-05,
      "loss": 1.5676,
      "step": 4667
    },
    {
      "epoch": 0.9694704049844237,
      "grad_norm": 0.5356010794639587,
      "learning_rate": 8e-05,
      "loss": 1.552,
      "step": 4668
    },
    {
      "epoch": 0.9696780893042575,
      "grad_norm": 0.5584197640419006,
      "learning_rate": 8e-05,
      "loss": 1.5362,
      "step": 4669
    },
    {
      "epoch": 0.9698857736240913,
      "grad_norm": 0.612156867980957,
      "learning_rate": 8e-05,
      "loss": 1.5587,
      "step": 4670
    },
    {
      "epoch": 0.9700934579439252,
      "grad_norm": 0.5209654569625854,
      "learning_rate": 8e-05,
      "loss": 1.445,
      "step": 4671
    },
    {
      "epoch": 0.9703011422637591,
      "grad_norm": 0.5556756258010864,
      "learning_rate": 8e-05,
      "loss": 1.5024,
      "step": 4672
    },
    {
      "epoch": 0.9705088265835929,
      "grad_norm": 0.5425582528114319,
      "learning_rate": 8e-05,
      "loss": 1.5166,
      "step": 4673
    },
    {
      "epoch": 0.9707165109034268,
      "grad_norm": 0.641613781452179,
      "learning_rate": 8e-05,
      "loss": 1.52,
      "step": 4674
    },
    {
      "epoch": 0.9709241952232607,
      "grad_norm": 0.5480392575263977,
      "learning_rate": 8e-05,
      "loss": 1.5774,
      "step": 4675
    },
    {
      "epoch": 0.9711318795430945,
      "grad_norm": 0.5259415507316589,
      "learning_rate": 8e-05,
      "loss": 1.5365,
      "step": 4676
    },
    {
      "epoch": 0.9713395638629283,
      "grad_norm": 0.6312845349311829,
      "learning_rate": 8e-05,
      "loss": 1.5239,
      "step": 4677
    },
    {
      "epoch": 0.9715472481827622,
      "grad_norm": 0.5325696468353271,
      "learning_rate": 8e-05,
      "loss": 1.5048,
      "step": 4678
    },
    {
      "epoch": 0.971754932502596,
      "grad_norm": 0.5516330599784851,
      "learning_rate": 8e-05,
      "loss": 1.5588,
      "step": 4679
    },
    {
      "epoch": 0.9719626168224299,
      "grad_norm": 0.5896750092506409,
      "learning_rate": 8e-05,
      "loss": 1.502,
      "step": 4680
    },
    {
      "epoch": 0.9721703011422638,
      "grad_norm": 0.5593307614326477,
      "learning_rate": 8e-05,
      "loss": 1.5417,
      "step": 4681
    },
    {
      "epoch": 0.9723779854620976,
      "grad_norm": 0.5611069202423096,
      "learning_rate": 8e-05,
      "loss": 1.5364,
      "step": 4682
    },
    {
      "epoch": 0.9725856697819315,
      "grad_norm": 0.5368297696113586,
      "learning_rate": 8e-05,
      "loss": 1.4593,
      "step": 4683
    },
    {
      "epoch": 0.9727933541017653,
      "grad_norm": 0.5230823755264282,
      "learning_rate": 8e-05,
      "loss": 1.4835,
      "step": 4684
    },
    {
      "epoch": 0.9730010384215991,
      "grad_norm": 0.5531055331230164,
      "learning_rate": 8e-05,
      "loss": 1.5056,
      "step": 4685
    },
    {
      "epoch": 0.973208722741433,
      "grad_norm": 0.5491374731063843,
      "learning_rate": 8e-05,
      "loss": 1.5407,
      "step": 4686
    },
    {
      "epoch": 0.9734164070612669,
      "grad_norm": 0.5461668968200684,
      "learning_rate": 8e-05,
      "loss": 1.5817,
      "step": 4687
    },
    {
      "epoch": 0.9736240913811007,
      "grad_norm": 0.5577301979064941,
      "learning_rate": 8e-05,
      "loss": 1.4687,
      "step": 4688
    },
    {
      "epoch": 0.9738317757009346,
      "grad_norm": 0.5358908772468567,
      "learning_rate": 8e-05,
      "loss": 1.4973,
      "step": 4689
    },
    {
      "epoch": 0.9740394600207685,
      "grad_norm": 0.55990070104599,
      "learning_rate": 8e-05,
      "loss": 1.5309,
      "step": 4690
    },
    {
      "epoch": 0.9742471443406023,
      "grad_norm": 0.5322283506393433,
      "learning_rate": 8e-05,
      "loss": 1.5173,
      "step": 4691
    },
    {
      "epoch": 0.9744548286604361,
      "grad_norm": 0.5635113716125488,
      "learning_rate": 8e-05,
      "loss": 1.4632,
      "step": 4692
    },
    {
      "epoch": 0.97466251298027,
      "grad_norm": 0.5296542644500732,
      "learning_rate": 8e-05,
      "loss": 1.5417,
      "step": 4693
    },
    {
      "epoch": 0.9748701973001038,
      "grad_norm": 0.552404522895813,
      "learning_rate": 8e-05,
      "loss": 1.5436,
      "step": 4694
    },
    {
      "epoch": 0.9750778816199377,
      "grad_norm": 0.5425599217414856,
      "learning_rate": 8e-05,
      "loss": 1.5144,
      "step": 4695
    },
    {
      "epoch": 0.9752855659397716,
      "grad_norm": 0.5408025979995728,
      "learning_rate": 8e-05,
      "loss": 1.5343,
      "step": 4696
    },
    {
      "epoch": 0.9754932502596054,
      "grad_norm": 0.5457563996315002,
      "learning_rate": 8e-05,
      "loss": 1.5341,
      "step": 4697
    },
    {
      "epoch": 0.9757009345794393,
      "grad_norm": 0.5345210433006287,
      "learning_rate": 8e-05,
      "loss": 1.481,
      "step": 4698
    },
    {
      "epoch": 0.975908618899273,
      "grad_norm": 0.573657214641571,
      "learning_rate": 8e-05,
      "loss": 1.5549,
      "step": 4699
    },
    {
      "epoch": 0.9761163032191069,
      "grad_norm": 0.5282204151153564,
      "learning_rate": 8e-05,
      "loss": 1.6358,
      "step": 4700
    },
    {
      "epoch": 0.9763239875389408,
      "grad_norm": 0.5454461574554443,
      "learning_rate": 8e-05,
      "loss": 1.615,
      "step": 4701
    },
    {
      "epoch": 0.9765316718587747,
      "grad_norm": 0.5407221913337708,
      "learning_rate": 8e-05,
      "loss": 1.493,
      "step": 4702
    },
    {
      "epoch": 0.9767393561786085,
      "grad_norm": 0.5548648834228516,
      "learning_rate": 8e-05,
      "loss": 1.5638,
      "step": 4703
    },
    {
      "epoch": 0.9769470404984424,
      "grad_norm": 0.5290051102638245,
      "learning_rate": 8e-05,
      "loss": 1.5366,
      "step": 4704
    },
    {
      "epoch": 0.9771547248182763,
      "grad_norm": 0.562443733215332,
      "learning_rate": 8e-05,
      "loss": 1.5962,
      "step": 4705
    },
    {
      "epoch": 0.9773624091381101,
      "grad_norm": 0.5909368991851807,
      "learning_rate": 8e-05,
      "loss": 1.5206,
      "step": 4706
    },
    {
      "epoch": 0.9775700934579439,
      "grad_norm": 0.5472081899642944,
      "learning_rate": 8e-05,
      "loss": 1.4945,
      "step": 4707
    },
    {
      "epoch": 0.9777777777777777,
      "grad_norm": 0.5259490609169006,
      "learning_rate": 8e-05,
      "loss": 1.4956,
      "step": 4708
    },
    {
      "epoch": 0.9779854620976116,
      "grad_norm": 0.5571253895759583,
      "learning_rate": 8e-05,
      "loss": 1.4978,
      "step": 4709
    },
    {
      "epoch": 0.9781931464174455,
      "grad_norm": 0.5329480767250061,
      "learning_rate": 8e-05,
      "loss": 1.5094,
      "step": 4710
    },
    {
      "epoch": 0.9784008307372793,
      "grad_norm": 0.551841139793396,
      "learning_rate": 8e-05,
      "loss": 1.6054,
      "step": 4711
    },
    {
      "epoch": 0.9786085150571132,
      "grad_norm": 0.5530389547348022,
      "learning_rate": 8e-05,
      "loss": 1.5277,
      "step": 4712
    },
    {
      "epoch": 0.9788161993769471,
      "grad_norm": 0.5720009803771973,
      "learning_rate": 8e-05,
      "loss": 1.5624,
      "step": 4713
    },
    {
      "epoch": 0.9790238836967808,
      "grad_norm": 0.5356488823890686,
      "learning_rate": 8e-05,
      "loss": 1.5297,
      "step": 4714
    },
    {
      "epoch": 0.9792315680166147,
      "grad_norm": 0.5629680156707764,
      "learning_rate": 8e-05,
      "loss": 1.5077,
      "step": 4715
    },
    {
      "epoch": 0.9794392523364486,
      "grad_norm": 0.5734153389930725,
      "learning_rate": 8e-05,
      "loss": 1.4279,
      "step": 4716
    },
    {
      "epoch": 0.9796469366562824,
      "grad_norm": 0.5761595964431763,
      "learning_rate": 8e-05,
      "loss": 1.5534,
      "step": 4717
    },
    {
      "epoch": 0.9798546209761163,
      "grad_norm": 0.5506202578544617,
      "learning_rate": 8e-05,
      "loss": 1.5882,
      "step": 4718
    },
    {
      "epoch": 0.9800623052959502,
      "grad_norm": 0.555680513381958,
      "learning_rate": 8e-05,
      "loss": 1.535,
      "step": 4719
    },
    {
      "epoch": 0.980269989615784,
      "grad_norm": 0.5411162376403809,
      "learning_rate": 8e-05,
      "loss": 1.5882,
      "step": 4720
    },
    {
      "epoch": 0.9804776739356179,
      "grad_norm": 0.5144530534744263,
      "learning_rate": 8e-05,
      "loss": 1.4875,
      "step": 4721
    },
    {
      "epoch": 0.9806853582554517,
      "grad_norm": 0.5955888032913208,
      "learning_rate": 8e-05,
      "loss": 1.5642,
      "step": 4722
    },
    {
      "epoch": 0.9808930425752855,
      "grad_norm": 0.5640544891357422,
      "learning_rate": 8e-05,
      "loss": 1.546,
      "step": 4723
    },
    {
      "epoch": 0.9811007268951194,
      "grad_norm": 0.6010296940803528,
      "learning_rate": 8e-05,
      "loss": 1.4795,
      "step": 4724
    },
    {
      "epoch": 0.9813084112149533,
      "grad_norm": 0.5588179230690002,
      "learning_rate": 8e-05,
      "loss": 1.4657,
      "step": 4725
    },
    {
      "epoch": 0.9815160955347871,
      "grad_norm": 0.5601072311401367,
      "learning_rate": 8e-05,
      "loss": 1.4798,
      "step": 4726
    },
    {
      "epoch": 0.981723779854621,
      "grad_norm": 0.5728517174720764,
      "learning_rate": 8e-05,
      "loss": 1.5665,
      "step": 4727
    },
    {
      "epoch": 0.9819314641744549,
      "grad_norm": 0.5644830465316772,
      "learning_rate": 8e-05,
      "loss": 1.5097,
      "step": 4728
    },
    {
      "epoch": 0.9821391484942886,
      "grad_norm": 0.5460310578346252,
      "learning_rate": 8e-05,
      "loss": 1.53,
      "step": 4729
    },
    {
      "epoch": 0.9823468328141225,
      "grad_norm": 0.5375919938087463,
      "learning_rate": 8e-05,
      "loss": 1.5434,
      "step": 4730
    },
    {
      "epoch": 0.9825545171339564,
      "grad_norm": 0.5816050171852112,
      "learning_rate": 8e-05,
      "loss": 1.5153,
      "step": 4731
    },
    {
      "epoch": 0.9827622014537902,
      "grad_norm": 0.5501047968864441,
      "learning_rate": 8e-05,
      "loss": 1.4192,
      "step": 4732
    },
    {
      "epoch": 0.9829698857736241,
      "grad_norm": 0.5344328284263611,
      "learning_rate": 8e-05,
      "loss": 1.4665,
      "step": 4733
    },
    {
      "epoch": 0.983177570093458,
      "grad_norm": 0.548666775226593,
      "learning_rate": 8e-05,
      "loss": 1.456,
      "step": 4734
    },
    {
      "epoch": 0.9833852544132918,
      "grad_norm": 0.5399496555328369,
      "learning_rate": 8e-05,
      "loss": 1.5142,
      "step": 4735
    },
    {
      "epoch": 0.9835929387331257,
      "grad_norm": 0.5277336835861206,
      "learning_rate": 8e-05,
      "loss": 1.4639,
      "step": 4736
    },
    {
      "epoch": 0.9838006230529595,
      "grad_norm": 0.5542486906051636,
      "learning_rate": 8e-05,
      "loss": 1.5298,
      "step": 4737
    },
    {
      "epoch": 0.9840083073727933,
      "grad_norm": 0.5799738764762878,
      "learning_rate": 8e-05,
      "loss": 1.5033,
      "step": 4738
    },
    {
      "epoch": 0.9842159916926272,
      "grad_norm": 0.5936886668205261,
      "learning_rate": 8e-05,
      "loss": 1.4834,
      "step": 4739
    },
    {
      "epoch": 0.9844236760124611,
      "grad_norm": 0.5801473259925842,
      "learning_rate": 8e-05,
      "loss": 1.5231,
      "step": 4740
    },
    {
      "epoch": 0.9846313603322949,
      "grad_norm": 0.5430822968482971,
      "learning_rate": 8e-05,
      "loss": 1.5166,
      "step": 4741
    },
    {
      "epoch": 0.9848390446521288,
      "grad_norm": 0.5598301291465759,
      "learning_rate": 8e-05,
      "loss": 1.5482,
      "step": 4742
    },
    {
      "epoch": 0.9850467289719627,
      "grad_norm": 0.5544182658195496,
      "learning_rate": 8e-05,
      "loss": 1.5061,
      "step": 4743
    },
    {
      "epoch": 0.9852544132917964,
      "grad_norm": 0.5487730503082275,
      "learning_rate": 8e-05,
      "loss": 1.5246,
      "step": 4744
    },
    {
      "epoch": 0.9854620976116303,
      "grad_norm": 0.5663031339645386,
      "learning_rate": 8e-05,
      "loss": 1.5604,
      "step": 4745
    },
    {
      "epoch": 0.9856697819314642,
      "grad_norm": 0.5447244644165039,
      "learning_rate": 8e-05,
      "loss": 1.5448,
      "step": 4746
    },
    {
      "epoch": 0.985877466251298,
      "grad_norm": 0.5563861727714539,
      "learning_rate": 8e-05,
      "loss": 1.5408,
      "step": 4747
    },
    {
      "epoch": 0.9860851505711319,
      "grad_norm": 0.5548373460769653,
      "learning_rate": 8e-05,
      "loss": 1.5779,
      "step": 4748
    },
    {
      "epoch": 0.9862928348909658,
      "grad_norm": 0.5432301163673401,
      "learning_rate": 8e-05,
      "loss": 1.5606,
      "step": 4749
    },
    {
      "epoch": 0.9865005192107996,
      "grad_norm": 0.6209354996681213,
      "learning_rate": 8e-05,
      "loss": 1.5207,
      "step": 4750
    },
    {
      "epoch": 0.9867082035306335,
      "grad_norm": 0.6065150499343872,
      "learning_rate": 8e-05,
      "loss": 1.4341,
      "step": 4751
    },
    {
      "epoch": 0.9869158878504672,
      "grad_norm": 0.5605137944221497,
      "learning_rate": 8e-05,
      "loss": 1.6266,
      "step": 4752
    },
    {
      "epoch": 0.9871235721703011,
      "grad_norm": 0.5662351846694946,
      "learning_rate": 8e-05,
      "loss": 1.446,
      "step": 4753
    },
    {
      "epoch": 0.987331256490135,
      "grad_norm": 0.5438563227653503,
      "learning_rate": 8e-05,
      "loss": 1.4473,
      "step": 4754
    },
    {
      "epoch": 0.9875389408099688,
      "grad_norm": 0.5519151091575623,
      "learning_rate": 8e-05,
      "loss": 1.5047,
      "step": 4755
    },
    {
      "epoch": 0.9877466251298027,
      "grad_norm": 0.6175982356071472,
      "learning_rate": 8e-05,
      "loss": 1.5972,
      "step": 4756
    },
    {
      "epoch": 0.9879543094496366,
      "grad_norm": 0.5414287447929382,
      "learning_rate": 8e-05,
      "loss": 1.5777,
      "step": 4757
    },
    {
      "epoch": 0.9881619937694704,
      "grad_norm": 0.5720454454421997,
      "learning_rate": 8e-05,
      "loss": 1.5452,
      "step": 4758
    },
    {
      "epoch": 0.9883696780893042,
      "grad_norm": 0.5543997883796692,
      "learning_rate": 8e-05,
      "loss": 1.5112,
      "step": 4759
    },
    {
      "epoch": 0.9885773624091381,
      "grad_norm": 0.5498349070549011,
      "learning_rate": 8e-05,
      "loss": 1.5652,
      "step": 4760
    },
    {
      "epoch": 0.9887850467289719,
      "grad_norm": 0.6565308570861816,
      "learning_rate": 8e-05,
      "loss": 1.5968,
      "step": 4761
    },
    {
      "epoch": 0.9889927310488058,
      "grad_norm": 0.555107057094574,
      "learning_rate": 8e-05,
      "loss": 1.4821,
      "step": 4762
    },
    {
      "epoch": 0.9892004153686397,
      "grad_norm": 0.5678965449333191,
      "learning_rate": 8e-05,
      "loss": 1.4879,
      "step": 4763
    },
    {
      "epoch": 0.9894080996884735,
      "grad_norm": 0.5877130031585693,
      "learning_rate": 8e-05,
      "loss": 1.5111,
      "step": 4764
    },
    {
      "epoch": 0.9896157840083074,
      "grad_norm": 0.612705409526825,
      "learning_rate": 8e-05,
      "loss": 1.5168,
      "step": 4765
    },
    {
      "epoch": 0.9898234683281413,
      "grad_norm": 0.5680394768714905,
      "learning_rate": 8e-05,
      "loss": 1.6056,
      "step": 4766
    },
    {
      "epoch": 0.990031152647975,
      "grad_norm": 0.5172173976898193,
      "learning_rate": 8e-05,
      "loss": 1.483,
      "step": 4767
    },
    {
      "epoch": 0.9902388369678089,
      "grad_norm": 0.579006016254425,
      "learning_rate": 8e-05,
      "loss": 1.5171,
      "step": 4768
    },
    {
      "epoch": 0.9904465212876428,
      "grad_norm": 0.5441842079162598,
      "learning_rate": 8e-05,
      "loss": 1.4655,
      "step": 4769
    },
    {
      "epoch": 0.9906542056074766,
      "grad_norm": 0.5395541191101074,
      "learning_rate": 8e-05,
      "loss": 1.5368,
      "step": 4770
    },
    {
      "epoch": 0.9908618899273105,
      "grad_norm": 0.580875813961029,
      "learning_rate": 8e-05,
      "loss": 1.547,
      "step": 4771
    },
    {
      "epoch": 0.9910695742471444,
      "grad_norm": 0.5491443276405334,
      "learning_rate": 8e-05,
      "loss": 1.5477,
      "step": 4772
    },
    {
      "epoch": 0.9912772585669782,
      "grad_norm": 0.5579102635383606,
      "learning_rate": 8e-05,
      "loss": 1.5435,
      "step": 4773
    },
    {
      "epoch": 0.991484942886812,
      "grad_norm": 0.5571985244750977,
      "learning_rate": 8e-05,
      "loss": 1.5173,
      "step": 4774
    },
    {
      "epoch": 0.9916926272066459,
      "grad_norm": 0.5357962250709534,
      "learning_rate": 8e-05,
      "loss": 1.531,
      "step": 4775
    },
    {
      "epoch": 0.9919003115264797,
      "grad_norm": 0.5589263439178467,
      "learning_rate": 8e-05,
      "loss": 1.5042,
      "step": 4776
    },
    {
      "epoch": 0.9921079958463136,
      "grad_norm": 0.5493354797363281,
      "learning_rate": 8e-05,
      "loss": 1.5249,
      "step": 4777
    },
    {
      "epoch": 0.9923156801661475,
      "grad_norm": 0.5731258392333984,
      "learning_rate": 8e-05,
      "loss": 1.5522,
      "step": 4778
    },
    {
      "epoch": 0.9925233644859813,
      "grad_norm": 0.5368343591690063,
      "learning_rate": 8e-05,
      "loss": 1.4908,
      "step": 4779
    },
    {
      "epoch": 0.9927310488058152,
      "grad_norm": 0.5405492186546326,
      "learning_rate": 8e-05,
      "loss": 1.5066,
      "step": 4780
    },
    {
      "epoch": 0.9929387331256491,
      "grad_norm": 0.6304225325584412,
      "learning_rate": 8e-05,
      "loss": 1.4995,
      "step": 4781
    },
    {
      "epoch": 0.9931464174454828,
      "grad_norm": 0.6054435968399048,
      "learning_rate": 8e-05,
      "loss": 1.5022,
      "step": 4782
    },
    {
      "epoch": 0.9933541017653167,
      "grad_norm": 0.607144832611084,
      "learning_rate": 8e-05,
      "loss": 1.548,
      "step": 4783
    },
    {
      "epoch": 0.9935617860851506,
      "grad_norm": 0.5454983711242676,
      "learning_rate": 8e-05,
      "loss": 1.4925,
      "step": 4784
    },
    {
      "epoch": 0.9937694704049844,
      "grad_norm": 0.6379164457321167,
      "learning_rate": 8e-05,
      "loss": 1.527,
      "step": 4785
    },
    {
      "epoch": 0.9939771547248183,
      "grad_norm": 0.5677633881568909,
      "learning_rate": 8e-05,
      "loss": 1.4768,
      "step": 4786
    },
    {
      "epoch": 0.9941848390446522,
      "grad_norm": 0.5792461037635803,
      "learning_rate": 8e-05,
      "loss": 1.4832,
      "step": 4787
    },
    {
      "epoch": 0.994392523364486,
      "grad_norm": 0.5529457926750183,
      "learning_rate": 8e-05,
      "loss": 1.6114,
      "step": 4788
    },
    {
      "epoch": 0.9946002076843198,
      "grad_norm": 0.6206566095352173,
      "learning_rate": 8e-05,
      "loss": 1.5268,
      "step": 4789
    },
    {
      "epoch": 0.9948078920041536,
      "grad_norm": 0.5420348644256592,
      "learning_rate": 8e-05,
      "loss": 1.5646,
      "step": 4790
    },
    {
      "epoch": 0.9950155763239875,
      "grad_norm": 0.5410664081573486,
      "learning_rate": 8e-05,
      "loss": 1.5479,
      "step": 4791
    },
    {
      "epoch": 0.9952232606438214,
      "grad_norm": 0.5613983273506165,
      "learning_rate": 8e-05,
      "loss": 1.4779,
      "step": 4792
    },
    {
      "epoch": 0.9954309449636553,
      "grad_norm": 0.5694583654403687,
      "learning_rate": 8e-05,
      "loss": 1.5234,
      "step": 4793
    },
    {
      "epoch": 0.9956386292834891,
      "grad_norm": 0.5669589042663574,
      "learning_rate": 8e-05,
      "loss": 1.5469,
      "step": 4794
    },
    {
      "epoch": 0.995846313603323,
      "grad_norm": 0.5912125706672668,
      "learning_rate": 8e-05,
      "loss": 1.6053,
      "step": 4795
    },
    {
      "epoch": 0.9960539979231569,
      "grad_norm": 0.5180086493492126,
      "learning_rate": 8e-05,
      "loss": 1.5066,
      "step": 4796
    },
    {
      "epoch": 0.9962616822429906,
      "grad_norm": 0.5145615339279175,
      "learning_rate": 8e-05,
      "loss": 1.5982,
      "step": 4797
    },
    {
      "epoch": 0.9964693665628245,
      "grad_norm": 0.5497413277626038,
      "learning_rate": 8e-05,
      "loss": 1.5046,
      "step": 4798
    },
    {
      "epoch": 0.9966770508826583,
      "grad_norm": 0.5785250067710876,
      "learning_rate": 8e-05,
      "loss": 1.5337,
      "step": 4799
    },
    {
      "epoch": 0.9968847352024922,
      "grad_norm": 0.5751317739486694,
      "learning_rate": 8e-05,
      "loss": 1.5636,
      "step": 4800
    },
    {
      "epoch": 0.9970924195223261,
      "grad_norm": 0.5439082980155945,
      "learning_rate": 8e-05,
      "loss": 1.5119,
      "step": 4801
    },
    {
      "epoch": 0.99730010384216,
      "grad_norm": 0.5442352294921875,
      "learning_rate": 8e-05,
      "loss": 1.4841,
      "step": 4802
    },
    {
      "epoch": 0.9975077881619938,
      "grad_norm": 0.5484969019889832,
      "learning_rate": 8e-05,
      "loss": 1.476,
      "step": 4803
    },
    {
      "epoch": 0.9977154724818276,
      "grad_norm": 0.5395694375038147,
      "learning_rate": 8e-05,
      "loss": 1.4249,
      "step": 4804
    },
    {
      "epoch": 0.9979231568016614,
      "grad_norm": 0.5270029306411743,
      "learning_rate": 8e-05,
      "loss": 1.4841,
      "step": 4805
    },
    {
      "epoch": 0.9981308411214953,
      "grad_norm": 0.5689374208450317,
      "learning_rate": 8e-05,
      "loss": 1.4984,
      "step": 4806
    },
    {
      "epoch": 0.9983385254413292,
      "grad_norm": 0.6203802227973938,
      "learning_rate": 8e-05,
      "loss": 1.4805,
      "step": 4807
    },
    {
      "epoch": 0.998546209761163,
      "grad_norm": 0.6458123326301575,
      "learning_rate": 8e-05,
      "loss": 1.5656,
      "step": 4808
    },
    {
      "epoch": 0.9987538940809969,
      "grad_norm": 0.580292820930481,
      "learning_rate": 8e-05,
      "loss": 1.5652,
      "step": 4809
    },
    {
      "epoch": 0.9989615784008308,
      "grad_norm": 0.5393871068954468,
      "learning_rate": 8e-05,
      "loss": 1.5021,
      "step": 4810
    },
    {
      "epoch": 0.9991692627206646,
      "grad_norm": 0.7075796723365784,
      "learning_rate": 8e-05,
      "loss": 1.4968,
      "step": 4811
    },
    {
      "epoch": 0.9993769470404984,
      "grad_norm": 0.6596865057945251,
      "learning_rate": 8e-05,
      "loss": 1.4469,
      "step": 4812
    },
    {
      "epoch": 0.9995846313603323,
      "grad_norm": 0.5620818138122559,
      "learning_rate": 8e-05,
      "loss": 1.4847,
      "step": 4813
    },
    {
      "epoch": 0.9997923156801661,
      "grad_norm": 0.5666627883911133,
      "learning_rate": 8e-05,
      "loss": 1.5554,
      "step": 4814
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.568848729133606,
      "learning_rate": 8e-05,
      "loss": 1.5858,
      "step": 4815
    }
  ],
  "logging_steps": 1,
  "max_steps": 4815,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.7333812723814236e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}