{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.998691670300916,
  "eval_steps": 500,
  "global_step": 1719,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 5.910665035247803,
      "learning_rate": 3.846153846153847e-07,
      "loss": 0.5662,
      "step": 1
    },
    {
      "epoch": 0.0,
      "grad_norm": 5.771595478057861,
      "learning_rate": 7.692307692307694e-07,
      "loss": 0.5619,
      "step": 2
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.088587760925293,
      "learning_rate": 1.153846153846154e-06,
      "loss": 0.5729,
      "step": 3
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.8097639083862305,
      "learning_rate": 1.5384615384615387e-06,
      "loss": 0.5561,
      "step": 4
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.116614818572998,
      "learning_rate": 1.9230769230769234e-06,
      "loss": 0.549,
      "step": 5
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.842957973480225,
      "learning_rate": 2.307692307692308e-06,
      "loss": 0.5381,
      "step": 6
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.623682498931885,
      "learning_rate": 2.6923076923076923e-06,
      "loss": 0.5215,
      "step": 7
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.316253185272217,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 0.5164,
      "step": 8
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.980376243591309,
      "learning_rate": 3.4615384615384617e-06,
      "loss": 0.4633,
      "step": 9
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.2344512939453125,
      "learning_rate": 3.846153846153847e-06,
      "loss": 0.419,
      "step": 10
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.3573031425476074,
      "learning_rate": 4.230769230769231e-06,
      "loss": 0.3872,
      "step": 11
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.3249635696411133,
      "learning_rate": 4.615384615384616e-06,
      "loss": 0.3175,
      "step": 12
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.547247052192688,
      "learning_rate": 5e-06,
      "loss": 0.2779,
      "step": 13
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.139501690864563,
      "learning_rate": 5.384615384615385e-06,
      "loss": 0.2572,
      "step": 14
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.8542943596839905,
      "learning_rate": 5.769230769230769e-06,
      "loss": 0.2495,
      "step": 15
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.6551614999771118,
      "learning_rate": 6.153846153846155e-06,
      "loss": 0.2171,
      "step": 16
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.6285273432731628,
      "learning_rate": 6.538461538461539e-06,
      "loss": 0.2075,
      "step": 17
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.4663761556148529,
      "learning_rate": 6.923076923076923e-06,
      "loss": 0.2052,
      "step": 18
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.4270327389240265,
      "learning_rate": 7.307692307692308e-06,
      "loss": 0.1868,
      "step": 19
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.46944189071655273,
      "learning_rate": 7.692307692307694e-06,
      "loss": 0.1921,
      "step": 20
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.40482595562934875,
      "learning_rate": 8.076923076923077e-06,
      "loss": 0.1785,
      "step": 21
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.3452730178833008,
      "learning_rate": 8.461538461538462e-06,
      "loss": 0.1865,
      "step": 22
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.3096699118614197,
      "learning_rate": 8.846153846153847e-06,
      "loss": 0.1821,
      "step": 23
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.2846967577934265,
      "learning_rate": 9.230769230769232e-06,
      "loss": 0.1646,
      "step": 24
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.20479772984981537,
      "learning_rate": 9.615384615384616e-06,
      "loss": 0.1754,
      "step": 25
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.1983906775712967,
      "learning_rate": 1e-05,
      "loss": 0.1599,
      "step": 26
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.16849859058856964,
      "learning_rate": 1.0384615384615386e-05,
      "loss": 0.168,
      "step": 27
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.18651075661182404,
      "learning_rate": 1.076923076923077e-05,
      "loss": 0.1574,
      "step": 28
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.19659897685050964,
      "learning_rate": 1.1153846153846154e-05,
      "loss": 0.1615,
      "step": 29
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.32734742760658264,
      "learning_rate": 1.1538461538461538e-05,
      "loss": 0.1676,
      "step": 30
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.17453128099441528,
      "learning_rate": 1.1923076923076925e-05,
      "loss": 0.1579,
      "step": 31
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.19505834579467773,
      "learning_rate": 1.230769230769231e-05,
      "loss": 0.1531,
      "step": 32
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.16609768569469452,
      "learning_rate": 1.2692307692307693e-05,
      "loss": 0.1563,
      "step": 33
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1682271659374237,
      "learning_rate": 1.3076923076923078e-05,
      "loss": 0.1605,
      "step": 34
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.30195435881614685,
      "learning_rate": 1.3461538461538463e-05,
      "loss": 0.162,
      "step": 35
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.15065325796604156,
      "learning_rate": 1.3846153846153847e-05,
      "loss": 0.1554,
      "step": 36
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1473916620016098,
      "learning_rate": 1.4230769230769232e-05,
      "loss": 0.1563,
      "step": 37
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.12049812823534012,
      "learning_rate": 1.4615384615384615e-05,
      "loss": 0.1536,
      "step": 38
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.15077902376651764,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.1531,
      "step": 39
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1712457239627838,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 0.1531,
      "step": 40
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.22883470356464386,
      "learning_rate": 1.576923076923077e-05,
      "loss": 0.1454,
      "step": 41
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.2349240928888321,
      "learning_rate": 1.6153846153846154e-05,
      "loss": 0.1625,
      "step": 42
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.16796810925006866,
      "learning_rate": 1.653846153846154e-05,
      "loss": 0.1496,
      "step": 43
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.14424921572208405,
      "learning_rate": 1.6923076923076924e-05,
      "loss": 0.1568,
      "step": 44
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.19721318781375885,
      "learning_rate": 1.730769230769231e-05,
      "loss": 0.1622,
      "step": 45
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.13727277517318726,
      "learning_rate": 1.7692307692307694e-05,
      "loss": 0.1478,
      "step": 46
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.1533319503068924,
      "learning_rate": 1.807692307692308e-05,
      "loss": 0.1562,
      "step": 47
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.18232691287994385,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.1557,
      "step": 48
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.11424045264720917,
      "learning_rate": 1.8846153846153846e-05,
      "loss": 0.1521,
      "step": 49
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.16621725261211395,
      "learning_rate": 1.923076923076923e-05,
      "loss": 0.1593,
      "step": 50
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.15255121886730194,
      "learning_rate": 1.9615384615384617e-05,
      "loss": 0.1581,
      "step": 51
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.17464229464530945,
      "learning_rate": 2e-05,
      "loss": 0.1509,
      "step": 52
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.16042707860469818,
      "learning_rate": 1.999998224182132e-05,
      "loss": 0.1496,
      "step": 53
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1732797771692276,
      "learning_rate": 1.9999928967348344e-05,
      "loss": 0.1555,
      "step": 54
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.1283779889345169,
      "learning_rate": 1.9999840176770285e-05,
      "loss": 0.1538,
      "step": 55
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.13336089253425598,
      "learning_rate": 1.9999715870402497e-05,
      "loss": 0.155,
      "step": 56
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.17042049765586853,
      "learning_rate": 1.9999556048686468e-05,
      "loss": 0.1513,
      "step": 57
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.186001718044281,
      "learning_rate": 1.999936071218983e-05,
      "loss": 0.1476,
      "step": 58
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.15404030680656433,
      "learning_rate": 1.9999129861606345e-05,
      "loss": 0.1509,
      "step": 59
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.21596437692642212,
      "learning_rate": 1.9998863497755908e-05,
      "loss": 0.1582,
      "step": 60
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.13059711456298828,
      "learning_rate": 1.999856162158455e-05,
      "loss": 0.1464,
      "step": 61
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.1580839455127716,
      "learning_rate": 1.9998224234164423e-05,
      "loss": 0.1513,
      "step": 62
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.12865480780601501,
      "learning_rate": 1.9997851336693804e-05,
      "loss": 0.1503,
      "step": 63
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.17513297498226166,
      "learning_rate": 1.999744293049709e-05,
      "loss": 0.1473,
      "step": 64
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.11717430502176285,
      "learning_rate": 1.9996999017024795e-05,
      "loss": 0.1536,
      "step": 65
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.11595330387353897,
      "learning_rate": 1.9996519597853533e-05,
      "loss": 0.1458,
      "step": 66
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.13235943019390106,
      "learning_rate": 1.9996004674686026e-05,
      "loss": 0.147,
      "step": 67
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.12007725238800049,
      "learning_rate": 1.9995454249351093e-05,
      "loss": 0.1482,
      "step": 68
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.10564625263214111,
      "learning_rate": 1.999486832380365e-05,
      "loss": 0.135,
      "step": 69
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.1476900577545166,
      "learning_rate": 1.9994246900124686e-05,
      "loss": 0.1533,
      "step": 70
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.15184570848941803,
      "learning_rate": 1.999358998052127e-05,
      "loss": 0.1536,
      "step": 71
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.14482644200325012,
      "learning_rate": 1.999289756732655e-05,
      "loss": 0.1418,
      "step": 72
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.16501332819461823,
      "learning_rate": 1.9992169662999715e-05,
      "loss": 0.1488,
      "step": 73
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.1715027540922165,
      "learning_rate": 1.999140627012602e-05,
      "loss": 0.1547,
      "step": 74
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.13953837752342224,
      "learning_rate": 1.9990607391416758e-05,
      "loss": 0.1476,
      "step": 75
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.12739704549312592,
      "learning_rate": 1.998977302970926e-05,
      "loss": 0.1609,
      "step": 76
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.15448978543281555,
      "learning_rate": 1.9988903187966866e-05,
      "loss": 0.1527,
      "step": 77
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.1813410222530365,
      "learning_rate": 1.9987997869278947e-05,
      "loss": 0.1458,
      "step": 78
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.12222440540790558,
      "learning_rate": 1.9987057076860853e-05,
      "loss": 0.144,
      "step": 79
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.13381321728229523,
      "learning_rate": 1.998608081405395e-05,
      "loss": 0.147,
      "step": 80
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.16005554795265198,
      "learning_rate": 1.9985069084325557e-05,
      "loss": 0.1374,
      "step": 81
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.24670375883579254,
      "learning_rate": 1.9984021891268973e-05,
      "loss": 0.1494,
      "step": 82
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.18134163320064545,
      "learning_rate": 1.9982939238603447e-05,
      "loss": 0.1405,
      "step": 83
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.12725010514259338,
      "learning_rate": 1.9981821130174168e-05,
      "loss": 0.1425,
      "step": 84
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.16447272896766663,
      "learning_rate": 1.998066756995225e-05,
      "loss": 0.1512,
      "step": 85
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.14608418941497803,
      "learning_rate": 1.9979478562034713e-05,
      "loss": 0.1458,
      "step": 86
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.14944171905517578,
      "learning_rate": 1.9978254110644487e-05,
      "loss": 0.1417,
      "step": 87
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.18665340542793274,
      "learning_rate": 1.9976994220130374e-05,
      "loss": 0.147,
      "step": 88
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.14529865980148315,
      "learning_rate": 1.9975698894967044e-05,
      "loss": 0.1524,
      "step": 89
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.1614844799041748,
      "learning_rate": 1.9974368139755028e-05,
      "loss": 0.1397,
      "step": 90
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.12654133141040802,
      "learning_rate": 1.997300195922067e-05,
      "loss": 0.1425,
      "step": 91
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.12260783463716507,
      "learning_rate": 1.997160035821616e-05,
      "loss": 0.1431,
      "step": 92
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.14651092886924744,
      "learning_rate": 1.997016334171947e-05,
      "loss": 0.1344,
      "step": 93
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.15125717222690582,
      "learning_rate": 1.9968690914834353e-05,
      "loss": 0.1467,
      "step": 94
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.14826516807079315,
      "learning_rate": 1.9967183082790343e-05,
      "loss": 0.1354,
      "step": 95
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.1300920695066452,
      "learning_rate": 1.99656398509427e-05,
      "loss": 0.1413,
      "step": 96
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.20764034986495972,
      "learning_rate": 1.9964061224772428e-05,
      "loss": 0.1436,
      "step": 97
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.2112439125776291,
      "learning_rate": 1.9962447209886225e-05,
      "loss": 0.137,
      "step": 98
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.1658688187599182,
      "learning_rate": 1.9960797812016496e-05,
      "loss": 0.1563,
      "step": 99
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.12187421321868896,
      "learning_rate": 1.995911303702129e-05,
      "loss": 0.1374,
      "step": 100
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.24928458034992218,
      "learning_rate": 1.9957392890884322e-05,
      "loss": 0.1462,
      "step": 101
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.15588997304439545,
      "learning_rate": 1.995563737971492e-05,
      "loss": 0.1482,
      "step": 102
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.24570755660533905,
      "learning_rate": 1.9953846509748017e-05,
      "loss": 0.1493,
      "step": 103
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.2456379234790802,
      "learning_rate": 1.995202028734414e-05,
      "loss": 0.1368,
      "step": 104
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.12289536744356155,
      "learning_rate": 1.9950158718989358e-05,
      "loss": 0.14,
      "step": 105
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.14341500401496887,
      "learning_rate": 1.994826181129529e-05,
      "loss": 0.1448,
      "step": 106
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.15434548258781433,
      "learning_rate": 1.9946329570999054e-05,
      "loss": 0.1471,
      "step": 107
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.12091369181871414,
      "learning_rate": 1.994436200496327e-05,
      "loss": 0.139,
      "step": 108
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.13661761581897736,
      "learning_rate": 1.994235912017601e-05,
      "loss": 0.1418,
      "step": 109
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.2033718079328537,
      "learning_rate": 1.9940320923750796e-05,
      "loss": 0.156,
      "step": 110
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.15100596845149994,
      "learning_rate": 1.9938247422926556e-05,
      "loss": 0.146,
      "step": 111
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.13823716342449188,
      "learning_rate": 1.9936138625067612e-05,
      "loss": 0.1436,
      "step": 112
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.13788972795009613,
      "learning_rate": 1.993399453766365e-05,
      "loss": 0.1393,
      "step": 113
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.16080260276794434,
      "learning_rate": 1.9931815168329674e-05,
      "loss": 0.1394,
      "step": 114
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.2249675989151001,
      "learning_rate": 1.992960052480602e-05,
      "loss": 0.155,
      "step": 115
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.13301348686218262,
      "learning_rate": 1.99273506149583e-05,
      "loss": 0.138,
      "step": 116
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.17179261147975922,
      "learning_rate": 1.9925065446777365e-05,
      "loss": 0.1561,
      "step": 117
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.13693870604038239,
      "learning_rate": 1.99227450283793e-05,
      "loss": 0.1433,
      "step": 118
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.10760874301195145,
      "learning_rate": 1.992038936800539e-05,
      "loss": 0.1362,
      "step": 119
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.16395406424999237,
      "learning_rate": 1.9917998474022083e-05,
      "loss": 0.1488,
      "step": 120
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.13689997792243958,
      "learning_rate": 1.991557235492096e-05,
      "loss": 0.135,
      "step": 121
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.17047208547592163,
      "learning_rate": 1.9913111019318715e-05,
      "loss": 0.1471,
      "step": 122
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.16670551896095276,
      "learning_rate": 1.9910614475957114e-05,
      "loss": 0.1448,
      "step": 123
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.14589165151119232,
      "learning_rate": 1.9908082733702972e-05,
      "loss": 0.1449,
      "step": 124
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.1511642038822174,
      "learning_rate": 1.990551580154811e-05,
      "loss": 0.143,
      "step": 125
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.1669401377439499,
      "learning_rate": 1.990291368860934e-05,
      "loss": 0.149,
      "step": 126
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.12463987618684769,
      "learning_rate": 1.990027640412842e-05,
      "loss": 0.1395,
      "step": 127
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.17946122586727142,
      "learning_rate": 1.9897603957472022e-05,
      "loss": 0.142,
      "step": 128
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.14758679270744324,
      "learning_rate": 1.9894896358131707e-05,
      "loss": 0.1515,
      "step": 129
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.12469500303268433,
      "learning_rate": 1.9892153615723873e-05,
      "loss": 0.1401,
      "step": 130
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.12411573529243469,
      "learning_rate": 1.988937573998975e-05,
      "loss": 0.1434,
      "step": 131
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.1325671225786209,
      "learning_rate": 1.9886562740795337e-05,
      "loss": 0.1421,
      "step": 132
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.14690901339054108,
      "learning_rate": 1.9883714628131382e-05,
      "loss": 0.1486,
      "step": 133
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.15927645564079285,
      "learning_rate": 1.9880831412113346e-05,
      "loss": 0.1384,
      "step": 134
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.1822102814912796,
      "learning_rate": 1.987791310298136e-05,
      "loss": 0.143,
      "step": 135
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.13575796782970428,
      "learning_rate": 1.98749597111002e-05,
      "loss": 0.1501,
      "step": 136
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.14061860740184784,
      "learning_rate": 1.9871971246959228e-05,
      "loss": 0.1412,
      "step": 137
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.1346607357263565,
      "learning_rate": 1.986894772117239e-05,
      "loss": 0.1365,
      "step": 138
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.1513783037662506,
      "learning_rate": 1.9865889144478144e-05,
      "loss": 0.1442,
      "step": 139
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.13938775658607483,
      "learning_rate": 1.986279552773944e-05,
      "loss": 0.1463,
      "step": 140
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.14437884092330933,
      "learning_rate": 1.9859666881943684e-05,
      "loss": 0.1411,
      "step": 141
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.1324140429496765,
      "learning_rate": 1.9856503218202674e-05,
      "loss": 0.1368,
      "step": 142
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.14142128825187683,
      "learning_rate": 1.98533045477526e-05,
      "loss": 0.147,
      "step": 143
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.1271912157535553,
      "learning_rate": 1.9850070881953974e-05,
      "loss": 0.1393,
      "step": 144
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.19452053308486938,
      "learning_rate": 1.9846802232291593e-05,
      "loss": 0.1463,
      "step": 145
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.15622642636299133,
      "learning_rate": 1.984349861037452e-05,
      "loss": 0.1489,
      "step": 146
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.1186322346329689,
      "learning_rate": 1.9840160027936005e-05,
      "loss": 0.141,
      "step": 147
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.14990271627902985,
      "learning_rate": 1.9836786496833485e-05,
      "loss": 0.1456,
      "step": 148
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.14955374598503113,
      "learning_rate": 1.9833378029048512e-05,
      "loss": 0.1524,
      "step": 149
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.1300581991672516,
      "learning_rate": 1.9829934636686722e-05,
      "loss": 0.1395,
      "step": 150
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.14859917759895325,
      "learning_rate": 1.9826456331977786e-05,
      "loss": 0.1504,
      "step": 151
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.12377198785543442,
      "learning_rate": 1.982294312727538e-05,
      "loss": 0.1427,
      "step": 152
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.12188497930765152,
      "learning_rate": 1.981939503505713e-05,
      "loss": 0.1372,
      "step": 153
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.12545304000377655,
      "learning_rate": 1.9815812067924562e-05,
      "loss": 0.1344,
      "step": 154
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.1244107335805893,
      "learning_rate": 1.981219423860307e-05,
      "loss": 0.1452,
      "step": 155
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.16565024852752686,
      "learning_rate": 1.9808541559941864e-05,
      "loss": 0.1325,
      "step": 156
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.12382091581821442,
      "learning_rate": 1.9804854044913934e-05,
      "loss": 0.1443,
      "step": 157
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.12897789478302002,
      "learning_rate": 1.980113170661599e-05,
      "loss": 0.1531,
      "step": 158
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.16778744757175446,
      "learning_rate": 1.9797374558268418e-05,
      "loss": 0.1496,
      "step": 159
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.11507518589496613,
      "learning_rate": 1.9793582613215242e-05,
      "loss": 0.132,
      "step": 160
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.18453346192836761,
      "learning_rate": 1.9789755884924067e-05,
      "loss": 0.1403,
      "step": 161
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.13673938810825348,
      "learning_rate": 1.9785894386986043e-05,
      "loss": 0.1434,
      "step": 162
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1111239343881607,
      "learning_rate": 1.9781998133115797e-05,
      "loss": 0.1292,
      "step": 163
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.24351872503757477,
      "learning_rate": 1.977806713715141e-05,
      "loss": 0.1581,
      "step": 164
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.1638852208852768,
      "learning_rate": 1.9774101413054347e-05,
      "loss": 0.1429,
      "step": 165
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.12461619079113007,
      "learning_rate": 1.9770100974909412e-05,
      "loss": 0.1357,
      "step": 166
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.17899064719676971,
      "learning_rate": 1.9766065836924705e-05,
      "loss": 0.1512,
      "step": 167
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.12398865818977356,
      "learning_rate": 1.976199601343157e-05,
      "loss": 0.1434,
      "step": 168
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.17089152336120605,
      "learning_rate": 1.9757891518884535e-05,
      "loss": 0.1364,
      "step": 169
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.13913339376449585,
      "learning_rate": 1.9753752367861265e-05,
      "loss": 0.1446,
      "step": 170
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.12167846411466599,
      "learning_rate": 1.9749578575062524e-05,
      "loss": 0.1376,
      "step": 171
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.13391132652759552,
      "learning_rate": 1.97453701553121e-05,
      "loss": 0.1413,
      "step": 172
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.12007936090230942,
      "learning_rate": 1.974112712355677e-05,
      "loss": 0.1372,
      "step": 173
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.14309321343898773,
      "learning_rate": 1.9736849494866234e-05,
      "loss": 0.1421,
      "step": 174
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.12029058486223221,
      "learning_rate": 1.973253728443307e-05,
      "loss": 0.1383,
      "step": 175
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.12588417530059814,
      "learning_rate": 1.9728190507572677e-05,
      "loss": 0.1451,
      "step": 176
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.13425788283348083,
      "learning_rate": 1.972380917972323e-05,
      "loss": 0.1395,
      "step": 177
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.12714862823486328,
      "learning_rate": 1.9719393316445607e-05,
      "loss": 0.1316,
      "step": 178
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.1295824944972992,
      "learning_rate": 1.9714942933423343e-05,
      "loss": 0.1397,
      "step": 179
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.13883793354034424,
      "learning_rate": 1.9710458046462585e-05,
      "loss": 0.1346,
      "step": 180
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.1370517909526825,
      "learning_rate": 1.9705938671492007e-05,
      "loss": 0.1393,
      "step": 181
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.14999105036258698,
      "learning_rate": 1.9701384824562786e-05,
      "loss": 0.1376,
      "step": 182
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.15495701134204865,
      "learning_rate": 1.9696796521848532e-05,
      "loss": 0.1343,
      "step": 183
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.15792758762836456,
      "learning_rate": 1.9692173779645222e-05,
      "loss": 0.1324,
      "step": 184
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.1866873800754547,
      "learning_rate": 1.968751661437115e-05,
      "loss": 0.1386,
      "step": 185
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.19024105370044708,
      "learning_rate": 1.9682825042566874e-05,
      "loss": 0.1391,
      "step": 186
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.1566675305366516,
      "learning_rate": 1.967809908089515e-05,
      "loss": 0.1362,
      "step": 187
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.1956055611371994,
      "learning_rate": 1.9673338746140868e-05,
      "loss": 0.143,
      "step": 188
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.15400373935699463,
      "learning_rate": 1.9668544055211004e-05,
      "loss": 0.1429,
      "step": 189
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.17495127022266388,
      "learning_rate": 1.9663715025134555e-05,
      "loss": 0.1504,
      "step": 190
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.21118001639842987,
      "learning_rate": 1.9658851673062475e-05,
      "loss": 0.1385,
      "step": 191
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.16880832612514496,
      "learning_rate": 1.9653954016267623e-05,
      "loss": 0.153,
      "step": 192
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.3060968518257141,
      "learning_rate": 1.9649022072144684e-05,
      "loss": 0.1373,
      "step": 193
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.17289908230304718,
      "learning_rate": 1.9644055858210137e-05,
      "loss": 0.1386,
      "step": 194
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.16520214080810547,
      "learning_rate": 1.9639055392102156e-05,
      "loss": 0.1349,
      "step": 195
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.18059268593788147,
      "learning_rate": 1.963402069158058e-05,
      "loss": 0.1461,
      "step": 196
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.16904336214065552,
      "learning_rate": 1.962895177452683e-05,
      "loss": 0.1468,
      "step": 197
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.17283155024051666,
      "learning_rate": 1.9623848658943855e-05,
      "loss": 0.1371,
      "step": 198
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.13215090334415436,
      "learning_rate": 1.9618711362956056e-05,
      "loss": 0.1401,
      "step": 199
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1454171985387802,
      "learning_rate": 1.961353990480925e-05,
      "loss": 0.1391,
      "step": 200
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.14165633916854858,
      "learning_rate": 1.960833430287056e-05,
      "loss": 0.1488,
      "step": 201
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1664198637008667,
      "learning_rate": 1.960309457562839e-05,
      "loss": 0.1516,
      "step": 202
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.13381633162498474,
      "learning_rate": 1.9597820741692347e-05,
      "loss": 0.1405,
      "step": 203
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.16438031196594238,
      "learning_rate": 1.9592512819793166e-05,
      "loss": 0.1447,
      "step": 204
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.14394235610961914,
      "learning_rate": 1.9587170828782652e-05,
      "loss": 0.1299,
      "step": 205
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.13756124675273895,
      "learning_rate": 1.958179478763361e-05,
      "loss": 0.1388,
      "step": 206
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.12917117774486542,
      "learning_rate": 1.957638471543978e-05,
      "loss": 0.1344,
      "step": 207
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.13050352036952972,
      "learning_rate": 1.9570940631415773e-05,
      "loss": 0.1346,
      "step": 208
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.1411200910806656,
      "learning_rate": 1.9565462554896986e-05,
      "loss": 0.1289,
      "step": 209
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.15984003245830536,
      "learning_rate": 1.9559950505339554e-05,
      "loss": 0.1523,
      "step": 210
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.13131478428840637,
      "learning_rate": 1.9554404502320266e-05,
      "loss": 0.1281,
      "step": 211
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.16635118424892426,
      "learning_rate": 1.954882456553651e-05,
      "loss": 0.1422,
      "step": 212
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.1810072362422943,
      "learning_rate": 1.9543210714806188e-05,
      "loss": 0.1465,
      "step": 213
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.22101140022277832,
      "learning_rate": 1.9537562970067644e-05,
      "loss": 0.1283,
      "step": 214
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.14121511578559875,
      "learning_rate": 1.9531881351379624e-05,
      "loss": 0.138,
      "step": 215
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.12035953253507614,
      "learning_rate": 1.952616587892116e-05,
      "loss": 0.1323,
      "step": 216
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.15852952003479004,
      "learning_rate": 1.9520416572991525e-05,
      "loss": 0.1367,
      "step": 217
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.17982420325279236,
      "learning_rate": 1.951463345401017e-05,
      "loss": 0.1454,
      "step": 218
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.25477373600006104,
      "learning_rate": 1.9508816542516613e-05,
      "loss": 0.1422,
      "step": 219
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.12057405710220337,
      "learning_rate": 1.9502965859170418e-05,
      "loss": 0.1344,
      "step": 220
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.15752074122428894,
      "learning_rate": 1.9497081424751073e-05,
      "loss": 0.1405,
      "step": 221
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.17521244287490845,
      "learning_rate": 1.949116326015795e-05,
      "loss": 0.1411,
      "step": 222
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.19864840805530548,
      "learning_rate": 1.948521138641021e-05,
      "loss": 0.1421,
      "step": 223
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.1920522302389145,
      "learning_rate": 1.947922582464674e-05,
      "loss": 0.1388,
      "step": 224
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.11913024634122849,
      "learning_rate": 1.9473206596126083e-05,
      "loss": 0.1328,
      "step": 225
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.12546971440315247,
      "learning_rate": 1.946715372222634e-05,
      "loss": 0.1353,
      "step": 226
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1619074046611786,
      "learning_rate": 1.9461067224445112e-05,
      "loss": 0.1379,
      "step": 227
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.22584852576255798,
      "learning_rate": 1.9454947124399424e-05,
      "loss": 0.1512,
      "step": 228
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.20896130800247192,
      "learning_rate": 1.9448793443825647e-05,
      "loss": 0.1463,
      "step": 229
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.17055071890354156,
      "learning_rate": 1.9442606204579408e-05,
      "loss": 0.135,
      "step": 230
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.15285927057266235,
      "learning_rate": 1.9436385428635525e-05,
      "loss": 0.1407,
      "step": 231
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.16020944714546204,
      "learning_rate": 1.943013113808793e-05,
      "loss": 0.1417,
      "step": 232
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.21106746792793274,
      "learning_rate": 1.9423843355149588e-05,
      "loss": 0.1468,
      "step": 233
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.14515185356140137,
      "learning_rate": 1.941752210215241e-05,
      "loss": 0.1416,
      "step": 234
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.19377292692661285,
      "learning_rate": 1.9411167401547188e-05,
      "loss": 0.1403,
      "step": 235
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.15671144425868988,
      "learning_rate": 1.94047792759035e-05,
      "loss": 0.1442,
      "step": 236
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.13283459842205048,
      "learning_rate": 1.939835774790964e-05,
      "loss": 0.138,
      "step": 237
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.23299051821231842,
      "learning_rate": 1.939190284037254e-05,
      "loss": 0.1416,
      "step": 238
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.23310522735118866,
      "learning_rate": 1.938541457621768e-05,
      "loss": 0.1399,
      "step": 239
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.21059724688529968,
      "learning_rate": 1.937889297848901e-05,
      "loss": 0.1444,
      "step": 240
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.17735469341278076,
      "learning_rate": 1.937233807034887e-05,
      "loss": 0.1381,
      "step": 241
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1174931600689888,
      "learning_rate": 1.93657498750779e-05,
      "loss": 0.1386,
      "step": 242
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.18651504814624786,
      "learning_rate": 1.9359128416074983e-05,
      "loss": 0.1359,
      "step": 243
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.12041015923023224,
      "learning_rate": 1.9352473716857117e-05,
      "loss": 0.1382,
      "step": 244
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.2746945023536682,
      "learning_rate": 1.9345785801059376e-05,
      "loss": 0.1441,
      "step": 245
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.17444874346256256,
      "learning_rate": 1.93390646924348e-05,
      "loss": 0.1346,
      "step": 246
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.14390631020069122,
      "learning_rate": 1.9332310414854313e-05,
      "loss": 0.1394,
      "step": 247
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1742342710494995,
      "learning_rate": 1.9325522992306657e-05,
      "loss": 0.1336,
      "step": 248
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1286180168390274,
      "learning_rate": 1.9318702448898277e-05,
      "loss": 0.1318,
      "step": 249
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.11181923002004623,
      "learning_rate": 1.9311848808853267e-05,
      "loss": 0.1261,
      "step": 250
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.24050500988960266,
      "learning_rate": 1.930496209651325e-05,
      "loss": 0.1424,
      "step": 251
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1454314887523651,
      "learning_rate": 1.9298042336337327e-05,
      "loss": 0.1358,
      "step": 252
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1317042112350464,
      "learning_rate": 1.9291089552901965e-05,
      "loss": 0.1335,
      "step": 253
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.13461628556251526,
      "learning_rate": 1.9284103770900917e-05,
      "loss": 0.1341,
      "step": 254
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.14653219282627106,
      "learning_rate": 1.9277085015145133e-05,
      "loss": 0.139,
      "step": 255
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.16250993311405182,
      "learning_rate": 1.927003331056268e-05,
      "loss": 0.1365,
      "step": 256
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.16071082651615143,
      "learning_rate": 1.9262948682198645e-05,
      "loss": 0.1472,
      "step": 257
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.19748622179031372,
      "learning_rate": 1.925583115521504e-05,
      "loss": 0.1359,
      "step": 258
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1450420767068863,
      "learning_rate": 1.9248680754890738e-05,
      "loss": 0.1485,
      "step": 259
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1335296630859375,
      "learning_rate": 1.9241497506621353e-05,
      "loss": 0.1392,
      "step": 260
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.15764756500720978,
      "learning_rate": 1.9234281435919165e-05,
      "loss": 0.1367,
      "step": 261
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.16038714349269867,
      "learning_rate": 1.9227032568413028e-05,
      "loss": 0.1439,
      "step": 262
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.25381746888160706,
      "learning_rate": 1.921975092984828e-05,
      "loss": 0.1449,
      "step": 263
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.17594526708126068,
      "learning_rate": 1.921243654608665e-05,
      "loss": 0.1401,
      "step": 264
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.14588139951229095,
      "learning_rate": 1.920508944310616e-05,
      "loss": 0.1364,
      "step": 265
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.1688345968723297,
      "learning_rate": 1.9197709647001047e-05,
      "loss": 0.1298,
      "step": 266
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.15290652215480804,
      "learning_rate": 1.919029718398166e-05,
      "loss": 0.1379,
      "step": 267
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.16612355411052704,
      "learning_rate": 1.9182852080374364e-05,
      "loss": 0.1357,
      "step": 268
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1405448168516159,
      "learning_rate": 1.917537436262146e-05,
      "loss": 0.1397,
      "step": 269
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1392367035150528,
      "learning_rate": 1.9167864057281072e-05,
      "loss": 0.1392,
      "step": 270
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.14967630803585052,
      "learning_rate": 1.9160321191027073e-05,
      "loss": 0.1448,
      "step": 271
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.09828483313322067,
      "learning_rate": 1.9152745790648973e-05,
      "loss": 0.1378,
      "step": 272
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.16568328440189362,
      "learning_rate": 1.914513788305184e-05,
      "loss": 0.1437,
      "step": 273
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.18456247448921204,
      "learning_rate": 1.9137497495256183e-05,
      "loss": 0.1547,
      "step": 274
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.12843498587608337,
      "learning_rate": 1.9129824654397882e-05,
      "loss": 0.1405,
      "step": 275
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1240093782544136,
      "learning_rate": 1.912211938772807e-05,
      "loss": 0.1377,
      "step": 276
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.12029384076595306,
      "learning_rate": 1.911438172261305e-05,
      "loss": 0.1405,
      "step": 277
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.14429883658885956,
      "learning_rate": 1.9106611686534188e-05,
      "loss": 0.1453,
      "step": 278
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.11662904173135757,
      "learning_rate": 1.9098809307087823e-05,
      "loss": 0.1336,
      "step": 279
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.15285204350948334,
      "learning_rate": 1.9090974611985163e-05,
      "loss": 0.1368,
      "step": 280
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.12474565207958221,
      "learning_rate": 1.9083107629052192e-05,
      "loss": 0.1426,
      "step": 281
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.1581178903579712,
      "learning_rate": 1.9075208386229567e-05,
      "loss": 0.1391,
      "step": 282
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.16635560989379883,
      "learning_rate": 1.906727691157252e-05,
      "loss": 0.1408,
      "step": 283
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.10616085678339005,
      "learning_rate": 1.9059313233250767e-05,
      "loss": 0.1315,
      "step": 284
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.14377376437187195,
      "learning_rate": 1.9051317379548387e-05,
      "loss": 0.1399,
      "step": 285
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.17061512172222137,
      "learning_rate": 1.9043289378863737e-05,
      "loss": 0.1345,
      "step": 286
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1545940488576889,
      "learning_rate": 1.9035229259709354e-05,
      "loss": 0.1379,
      "step": 287
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.13912220299243927,
      "learning_rate": 1.9027137050711846e-05,
      "loss": 0.1367,
      "step": 288
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.17505572736263275,
      "learning_rate": 1.9019012780611793e-05,
      "loss": 0.1434,
      "step": 289
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.1628369241952896,
      "learning_rate": 1.9010856478263636e-05,
      "loss": 0.1414,
      "step": 290
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.15574808418750763,
      "learning_rate": 1.90026681726356e-05,
      "loss": 0.1406,
      "step": 291
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.145838662981987,
      "learning_rate": 1.8994447892809556e-05,
      "loss": 0.1424,
      "step": 292
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.22621150314807892,
      "learning_rate": 1.8986195667980947e-05,
      "loss": 0.1441,
      "step": 293
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.1623389720916748,
      "learning_rate": 1.897791152745867e-05,
      "loss": 0.1392,
      "step": 294
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.16793805360794067,
      "learning_rate": 1.896959550066497e-05,
      "loss": 0.1384,
      "step": 295
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.15599815547466278,
      "learning_rate": 1.8961247617135353e-05,
      "loss": 0.1416,
      "step": 296
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.19707897305488586,
      "learning_rate": 1.8952867906518454e-05,
      "loss": 0.1329,
      "step": 297
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.13621143996715546,
      "learning_rate": 1.8944456398575955e-05,
      "loss": 0.1356,
      "step": 298
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.141517773270607,
      "learning_rate": 1.893601312318247e-05,
      "loss": 0.1373,
      "step": 299
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.17788486182689667,
      "learning_rate": 1.892753811032543e-05,
      "loss": 0.1476,
      "step": 300
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1389748603105545,
      "learning_rate": 1.8919031390105003e-05,
      "loss": 0.1354,
      "step": 301
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.18589496612548828,
      "learning_rate": 1.8910492992733957e-05,
      "loss": 0.1406,
      "step": 302
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.11037491261959076,
      "learning_rate": 1.8901922948537567e-05,
      "loss": 0.122,
      "step": 303
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.15758061408996582,
      "learning_rate": 1.889332128795351e-05,
      "loss": 0.1402,
      "step": 304
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.13788828253746033,
      "learning_rate": 1.8884688041531753e-05,
      "loss": 0.1378,
      "step": 305
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.16533860564231873,
      "learning_rate": 1.8876023239934435e-05,
      "loss": 0.14,
      "step": 306
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1480337232351303,
      "learning_rate": 1.8867326913935786e-05,
      "loss": 0.135,
      "step": 307
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.2024344503879547,
      "learning_rate": 1.885859909442198e-05,
      "loss": 0.1351,
      "step": 308
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.177449032664299,
      "learning_rate": 1.884983981239106e-05,
      "loss": 0.132,
      "step": 309
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.15997964143753052,
      "learning_rate": 1.8841049098952797e-05,
      "loss": 0.1338,
      "step": 310
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.16571328043937683,
      "learning_rate": 1.8832226985328606e-05,
      "loss": 0.1371,
      "step": 311
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1679169237613678,
      "learning_rate": 1.8823373502851422e-05,
      "loss": 0.1382,
      "step": 312
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1807669848203659,
      "learning_rate": 1.881448868296559e-05,
      "loss": 0.1417,
      "step": 313
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.164134219288826,
      "learning_rate": 1.8805572557226755e-05,
      "loss": 0.1407,
      "step": 314
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1522090882062912,
      "learning_rate": 1.8796625157301744e-05,
      "loss": 0.129,
      "step": 315
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.17090144753456116,
      "learning_rate": 1.8787646514968465e-05,
      "loss": 0.1413,
      "step": 316
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.16151976585388184,
      "learning_rate": 1.8778636662115787e-05,
      "loss": 0.1364,
      "step": 317
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.19610628485679626,
      "learning_rate": 1.8769595630743417e-05,
      "loss": 0.1353,
      "step": 318
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.17941401898860931,
      "learning_rate": 1.8760523452961813e-05,
      "loss": 0.1389,
      "step": 319
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.16495488584041595,
      "learning_rate": 1.8751420160992045e-05,
      "loss": 0.1383,
      "step": 320
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.14701516926288605,
      "learning_rate": 1.8742285787165686e-05,
      "loss": 0.1352,
      "step": 321
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.18544933199882507,
      "learning_rate": 1.8733120363924707e-05,
      "loss": 0.139,
      "step": 322
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.1587456911802292,
      "learning_rate": 1.8723923923821358e-05,
      "loss": 0.1326,
      "step": 323
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.13041561841964722,
      "learning_rate": 1.8714696499518034e-05,
      "loss": 0.1335,
      "step": 324
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.22151540219783783,
      "learning_rate": 1.8705438123787193e-05,
      "loss": 0.1414,
      "step": 325
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.152569979429245,
      "learning_rate": 1.869614882951121e-05,
      "loss": 0.1379,
      "step": 326
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.12596142292022705,
      "learning_rate": 1.8686828649682272e-05,
      "loss": 0.1301,
      "step": 327
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.14535236358642578,
      "learning_rate": 1.867747761740227e-05,
      "loss": 0.1528,
      "step": 328
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.27292972803115845,
      "learning_rate": 1.8668095765882656e-05,
      "loss": 0.1416,
      "step": 329
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.23864008486270905,
      "learning_rate": 1.8658683128444352e-05,
      "loss": 0.1378,
      "step": 330
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.19546228647232056,
      "learning_rate": 1.8649239738517623e-05,
      "loss": 0.1373,
      "step": 331
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.17039385437965393,
      "learning_rate": 1.8639765629641943e-05,
      "loss": 0.1321,
      "step": 332
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.1484154611825943,
      "learning_rate": 1.86302608354659e-05,
      "loss": 0.128,
      "step": 333
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.29283076524734497,
      "learning_rate": 1.862072538974706e-05,
      "loss": 0.1362,
      "step": 334
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.2925666868686676,
      "learning_rate": 1.8611159326351847e-05,
      "loss": 0.1405,
      "step": 335
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.2313055843114853,
      "learning_rate": 1.860156267925544e-05,
      "loss": 0.1451,
      "step": 336
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.2219887524843216,
      "learning_rate": 1.8591935482541636e-05,
      "loss": 0.1403,
      "step": 337
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.20466695725917816,
      "learning_rate": 1.8582277770402727e-05,
      "loss": 0.1456,
      "step": 338
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.23283225297927856,
      "learning_rate": 1.8572589577139382e-05,
      "loss": 0.1481,
      "step": 339
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.13946789503097534,
      "learning_rate": 1.8562870937160545e-05,
      "loss": 0.1395,
      "step": 340
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.22407369315624237,
      "learning_rate": 1.8553121884983284e-05,
      "loss": 0.1367,
      "step": 341
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.12007815390825272,
      "learning_rate": 1.8543342455232674e-05,
      "loss": 0.1292,
      "step": 342
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.15134385228157043,
      "learning_rate": 1.8533532682641693e-05,
      "loss": 0.141,
      "step": 343
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.15021508932113647,
      "learning_rate": 1.8523692602051078e-05,
      "loss": 0.1378,
      "step": 344
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.1870955526828766,
      "learning_rate": 1.8513822248409213e-05,
      "loss": 0.1451,
      "step": 345
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.18127886950969696,
      "learning_rate": 1.8503921656771995e-05,
      "loss": 0.1428,
      "step": 346
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.1412736475467682,
      "learning_rate": 1.849399086230272e-05,
      "loss": 0.1353,
      "step": 347
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.126445010304451,
      "learning_rate": 1.8484029900271953e-05,
      "loss": 0.1331,
      "step": 348
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.1492401361465454,
      "learning_rate": 1.8474038806057407e-05,
      "loss": 0.1329,
      "step": 349
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.1417461335659027,
      "learning_rate": 1.8464017615143805e-05,
      "loss": 0.1431,
      "step": 350
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.10406292974948883,
      "learning_rate": 1.8453966363122764e-05,
      "loss": 0.1328,
      "step": 351
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.13288991153240204,
      "learning_rate": 1.8443885085692674e-05,
      "loss": 0.134,
      "step": 352
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.15323688089847565,
      "learning_rate": 1.843377381865856e-05,
      "loss": 0.1306,
      "step": 353
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.13318227231502533,
      "learning_rate": 1.8423632597931957e-05,
      "loss": 0.1281,
      "step": 354
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.12018125504255295,
      "learning_rate": 1.8413461459530792e-05,
      "loss": 0.1333,
      "step": 355
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.1724357008934021,
      "learning_rate": 1.8403260439579238e-05,
      "loss": 0.1365,
      "step": 356
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.18728823959827423,
      "learning_rate": 1.8393029574307607e-05,
      "loss": 0.1429,
      "step": 357
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.16602103412151337,
      "learning_rate": 1.83827689000522e-05,
      "loss": 0.133,
      "step": 358
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.15758636593818665,
      "learning_rate": 1.8372478453255194e-05,
      "loss": 0.1416,
      "step": 359
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.20182034373283386,
      "learning_rate": 1.8362158270464515e-05,
      "loss": 0.1347,
      "step": 360
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.15463927388191223,
      "learning_rate": 1.835180838833369e-05,
      "loss": 0.1395,
      "step": 361
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.20479172468185425,
      "learning_rate": 1.834142884362173e-05,
      "loss": 0.1291,
      "step": 362
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.14601053297519684,
      "learning_rate": 1.833101967319299e-05,
      "loss": 0.1287,
      "step": 363
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.15544813871383667,
      "learning_rate": 1.8320580914017062e-05,
      "loss": 0.1364,
      "step": 364
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.15212176740169525,
      "learning_rate": 1.8310112603168613e-05,
      "loss": 0.1426,
      "step": 365
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.17145603895187378,
      "learning_rate": 1.8299614777827267e-05,
      "loss": 0.138,
      "step": 366
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.1746693253517151,
      "learning_rate": 1.828908747527748e-05,
      "loss": 0.1344,
      "step": 367
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.21381957828998566,
      "learning_rate": 1.827853073290839e-05,
      "loss": 0.1501,
      "step": 368
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.1626751869916916,
      "learning_rate": 1.8267944588213704e-05,
      "loss": 0.1409,
      "step": 369
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.16132423281669617,
      "learning_rate": 1.8257329078791555e-05,
      "loss": 0.139,
      "step": 370
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.15548644959926605,
      "learning_rate": 1.8246684242344358e-05,
      "loss": 0.1281,
      "step": 371
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1603497713804245,
      "learning_rate": 1.82360101166787e-05,
      "loss": 0.1342,
      "step": 372
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1884692758321762,
      "learning_rate": 1.8225306739705185e-05,
      "loss": 0.1274,
      "step": 373
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1739269495010376,
      "learning_rate": 1.8214574149438315e-05,
      "loss": 0.138,
      "step": 374
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.15011073648929596,
      "learning_rate": 1.820381238399633e-05,
      "loss": 0.1268,
      "step": 375
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1428864598274231,
      "learning_rate": 1.8193021481601112e-05,
      "loss": 0.1351,
      "step": 376
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.14912478625774384,
      "learning_rate": 1.818220148057801e-05,
      "loss": 0.1329,
      "step": 377
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.2448536455631256,
      "learning_rate": 1.8171352419355725e-05,
      "loss": 0.1372,
      "step": 378
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.21458327770233154,
      "learning_rate": 1.816047433646617e-05,
      "loss": 0.1347,
      "step": 379
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1864161491394043,
      "learning_rate": 1.8149567270544342e-05,
      "loss": 0.1428,
      "step": 380
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.17744745314121246,
      "learning_rate": 1.8138631260328154e-05,
      "loss": 0.1305,
      "step": 381
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.19454380869865417,
      "learning_rate": 1.8127666344658336e-05,
      "loss": 0.139,
      "step": 382
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.17242197692394257,
      "learning_rate": 1.811667256247828e-05,
      "loss": 0.133,
      "step": 383
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.24748246371746063,
      "learning_rate": 1.8105649952833883e-05,
      "loss": 0.1331,
      "step": 384
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.2282569259405136,
      "learning_rate": 1.809459855487345e-05,
      "loss": 0.1376,
      "step": 385
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.19509662687778473,
      "learning_rate": 1.808351840784752e-05,
      "loss": 0.1326,
      "step": 386
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.1997249871492386,
      "learning_rate": 1.8072409551108734e-05,
      "loss": 0.1272,
      "step": 387
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.17559269070625305,
      "learning_rate": 1.8061272024111707e-05,
      "loss": 0.1222,
      "step": 388
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.25800377130508423,
      "learning_rate": 1.8050105866412877e-05,
      "loss": 0.1389,
      "step": 389
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2314135581254959,
      "learning_rate": 1.803891111767037e-05,
      "loss": 0.1347,
      "step": 390
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.15699076652526855,
      "learning_rate": 1.8027687817643856e-05,
      "loss": 0.1295,
      "step": 391
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.1685732752084732,
      "learning_rate": 1.8016436006194407e-05,
      "loss": 0.132,
      "step": 392
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.18155637383460999,
      "learning_rate": 1.8005155723284363e-05,
      "loss": 0.1406,
      "step": 393
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.20918649435043335,
      "learning_rate": 1.7993847008977176e-05,
      "loss": 0.133,
      "step": 394
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.24005411565303802,
      "learning_rate": 1.7982509903437275e-05,
      "loss": 0.1339,
      "step": 395
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.1847439706325531,
      "learning_rate": 1.797114444692994e-05,
      "loss": 0.132,
      "step": 396
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.17285490036010742,
      "learning_rate": 1.7959750679821125e-05,
      "loss": 0.1314,
      "step": 397
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.21208739280700684,
      "learning_rate": 1.7948328642577343e-05,
      "loss": 0.1552,
      "step": 398
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.2574717104434967,
      "learning_rate": 1.7936878375765507e-05,
      "loss": 0.1515,
      "step": 399
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.21646028757095337,
      "learning_rate": 1.7925399920052798e-05,
      "loss": 0.1396,
      "step": 400
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.23925833404064178,
      "learning_rate": 1.791389331620651e-05,
      "loss": 0.1365,
      "step": 401
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.23673556745052338,
      "learning_rate": 1.79023586050939e-05,
      "loss": 0.148,
      "step": 402
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.16269028186798096,
      "learning_rate": 1.7890795827682067e-05,
      "loss": 0.1308,
      "step": 403
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.21490103006362915,
      "learning_rate": 1.787920502503779e-05,
      "loss": 0.1372,
      "step": 404
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.16239553689956665,
      "learning_rate": 1.7867586238327366e-05,
      "loss": 0.1384,
      "step": 405
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.2009923905134201,
      "learning_rate": 1.78559395088165e-05,
      "loss": 0.1369,
      "step": 406
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.21127651631832123,
      "learning_rate": 1.7844264877870126e-05,
      "loss": 0.1408,
      "step": 407
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.165143683552742,
      "learning_rate": 1.7832562386952293e-05,
      "loss": 0.1356,
      "step": 408
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.18612289428710938,
      "learning_rate": 1.7820832077625977e-05,
      "loss": 0.1457,
      "step": 409
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.13980364799499512,
      "learning_rate": 1.780907399155296e-05,
      "loss": 0.1224,
      "step": 410
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1685197651386261,
      "learning_rate": 1.7797288170493692e-05,
      "loss": 0.1246,
      "step": 411
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.22625800967216492,
      "learning_rate": 1.7785474656307108e-05,
      "loss": 0.1307,
      "step": 412
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.3002471923828125,
      "learning_rate": 1.777363349095051e-05,
      "loss": 0.152,
      "step": 413
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1669785976409912,
      "learning_rate": 1.7761764716479405e-05,
      "loss": 0.1237,
      "step": 414
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.237270787358284,
      "learning_rate": 1.774986837504735e-05,
      "loss": 0.1351,
      "step": 415
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.18258631229400635,
      "learning_rate": 1.773794450890583e-05,
      "loss": 0.1291,
      "step": 416
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.1599574089050293,
      "learning_rate": 1.772599316040406e-05,
      "loss": 0.1311,
      "step": 417
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.19871650636196136,
      "learning_rate": 1.7714014371988883e-05,
      "loss": 0.1407,
      "step": 418
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.18687580525875092,
      "learning_rate": 1.770200818620459e-05,
      "loss": 0.1296,
      "step": 419
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.16156288981437683,
      "learning_rate": 1.7689974645692786e-05,
      "loss": 0.1301,
      "step": 420
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.2020213007926941,
      "learning_rate": 1.767791379319221e-05,
      "loss": 0.1341,
      "step": 421
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.15818969905376434,
      "learning_rate": 1.7665825671538628e-05,
      "loss": 0.13,
      "step": 422
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.17172771692276,
      "learning_rate": 1.7653710323664643e-05,
      "loss": 0.1357,
      "step": 423
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1800169050693512,
      "learning_rate": 1.7641567792599558e-05,
      "loss": 0.1385,
      "step": 424
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.18685565888881683,
      "learning_rate": 1.7629398121469216e-05,
      "loss": 0.1297,
      "step": 425
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.17401066422462463,
      "learning_rate": 1.7617201353495855e-05,
      "loss": 0.1276,
      "step": 426
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.23841150104999542,
      "learning_rate": 1.760497753199796e-05,
      "loss": 0.1321,
      "step": 427
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.22171254456043243,
      "learning_rate": 1.7592726700390087e-05,
      "loss": 0.1389,
      "step": 428
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.18666167557239532,
      "learning_rate": 1.7580448902182724e-05,
      "loss": 0.1295,
      "step": 429
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.20041997730731964,
      "learning_rate": 1.7568144180982148e-05,
      "loss": 0.1436,
      "step": 430
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.20260149240493774,
      "learning_rate": 1.7555812580490234e-05,
      "loss": 0.1343,
      "step": 431
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.21036387979984283,
      "learning_rate": 1.754345414450435e-05,
      "loss": 0.1344,
      "step": 432
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.2078201323747635,
      "learning_rate": 1.753106891691714e-05,
      "loss": 0.1263,
      "step": 433
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.18322327733039856,
      "learning_rate": 1.751865694171644e-05,
      "loss": 0.1335,
      "step": 434
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.18306390941143036,
      "learning_rate": 1.750621826298505e-05,
      "loss": 0.1356,
      "step": 435
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.22829410433769226,
      "learning_rate": 1.7493752924900632e-05,
      "loss": 0.1386,
      "step": 436
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.14326487481594086,
      "learning_rate": 1.748126097173553e-05,
      "loss": 0.126,
      "step": 437
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.1849300116300583,
      "learning_rate": 1.74687424478566e-05,
      "loss": 0.1297,
      "step": 438
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.2141926884651184,
      "learning_rate": 1.7456197397725092e-05,
      "loss": 0.1302,
      "step": 439
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.18119148910045624,
      "learning_rate": 1.7443625865896447e-05,
      "loss": 0.1383,
      "step": 440
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.18320521712303162,
      "learning_rate": 1.7431027897020167e-05,
      "loss": 0.1316,
      "step": 441
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.196560800075531,
      "learning_rate": 1.7418403535839653e-05,
      "loss": 0.1346,
      "step": 442
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.16851496696472168,
      "learning_rate": 1.7405752827192028e-05,
      "loss": 0.1386,
      "step": 443
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.18364140391349792,
      "learning_rate": 1.739307581600801e-05,
      "loss": 0.1271,
      "step": 444
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.19678296148777008,
      "learning_rate": 1.7380372547311722e-05,
      "loss": 0.1288,
      "step": 445
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.19194363057613373,
      "learning_rate": 1.736764306622055e-05,
      "loss": 0.1439,
      "step": 446
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.17510387301445007,
      "learning_rate": 1.7354887417944962e-05,
      "loss": 0.1312,
      "step": 447
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.19436456263065338,
      "learning_rate": 1.734210564778839e-05,
      "loss": 0.1235,
      "step": 448
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.18059295415878296,
      "learning_rate": 1.732929780114702e-05,
      "loss": 0.1317,
      "step": 449
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.2032422572374344,
      "learning_rate": 1.731646392350965e-05,
      "loss": 0.1369,
      "step": 450
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.18783380091190338,
      "learning_rate": 1.7303604060457545e-05,
      "loss": 0.1402,
      "step": 451
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.19288314878940582,
      "learning_rate": 1.7290718257664258e-05,
      "loss": 0.1341,
      "step": 452
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.22946001589298248,
      "learning_rate": 1.7277806560895466e-05,
      "loss": 0.1379,
      "step": 453
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.19840770959854126,
      "learning_rate": 1.7264869016008804e-05,
      "loss": 0.1242,
      "step": 454
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.20633384585380554,
      "learning_rate": 1.7251905668953727e-05,
      "loss": 0.1357,
      "step": 455
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.19114214181900024,
      "learning_rate": 1.7238916565771316e-05,
      "loss": 0.13,
      "step": 456
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.2529546618461609,
      "learning_rate": 1.7225901752594137e-05,
      "loss": 0.1294,
      "step": 457
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.2007884979248047,
      "learning_rate": 1.7212861275646065e-05,
      "loss": 0.1363,
      "step": 458
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.2559095621109009,
      "learning_rate": 1.719979518124213e-05,
      "loss": 0.1423,
      "step": 459
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.18447914719581604,
      "learning_rate": 1.7186703515788327e-05,
      "loss": 0.135,
      "step": 460
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.1905982494354248,
      "learning_rate": 1.7173586325781495e-05,
      "loss": 0.1292,
      "step": 461
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.20021343231201172,
      "learning_rate": 1.716044365780911e-05,
      "loss": 0.1352,
      "step": 462
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.17757542431354523,
      "learning_rate": 1.714727555854914e-05,
      "loss": 0.1398,
      "step": 463
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.24342171847820282,
      "learning_rate": 1.713408207476988e-05,
      "loss": 0.124,
      "step": 464
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.2015884518623352,
      "learning_rate": 1.7120863253329773e-05,
      "loss": 0.1424,
      "step": 465
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.1732528656721115,
      "learning_rate": 1.7107619141177262e-05,
      "loss": 0.1333,
      "step": 466
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.18846891820430756,
      "learning_rate": 1.709434978535061e-05,
      "loss": 0.1354,
      "step": 467
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.15533560514450073,
      "learning_rate": 1.7081055232977735e-05,
      "loss": 0.1288,
      "step": 468
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.1695532202720642,
      "learning_rate": 1.706773553127604e-05,
      "loss": 0.1303,
      "step": 469
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.23584920167922974,
      "learning_rate": 1.705439072755226e-05,
      "loss": 0.1407,
      "step": 470
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.20207595825195312,
      "learning_rate": 1.7041020869202266e-05,
      "loss": 0.1343,
      "step": 471
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.16381163895130157,
      "learning_rate": 1.7027626003710937e-05,
      "loss": 0.1307,
      "step": 472
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.1604747623205185,
      "learning_rate": 1.7014206178651948e-05,
      "loss": 0.1315,
      "step": 473
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.1837093085050583,
      "learning_rate": 1.7000761441687636e-05,
      "loss": 0.1315,
      "step": 474
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.15001989901065826,
      "learning_rate": 1.6987291840568803e-05,
      "loss": 0.127,
      "step": 475
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.17914797365665436,
      "learning_rate": 1.697379742313457e-05,
      "loss": 0.126,
      "step": 476
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.24492678046226501,
      "learning_rate": 1.6960278237312188e-05,
      "loss": 0.1357,
      "step": 477
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.1759316772222519,
      "learning_rate": 1.6946734331116886e-05,
      "loss": 0.1366,
      "step": 478
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.18484148383140564,
      "learning_rate": 1.6933165752651678e-05,
      "loss": 0.1358,
      "step": 479
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.18875397741794586,
      "learning_rate": 1.6919572550107218e-05,
      "loss": 0.1316,
      "step": 480
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.22361242771148682,
      "learning_rate": 1.690595477176161e-05,
      "loss": 0.1423,
      "step": 481
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.18579301238059998,
      "learning_rate": 1.6892312465980238e-05,
      "loss": 0.1275,
      "step": 482
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.1770302802324295,
      "learning_rate": 1.687864568121561e-05,
      "loss": 0.1238,
      "step": 483
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.2037944197654724,
      "learning_rate": 1.6864954466007156e-05,
      "loss": 0.1328,
      "step": 484
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.17485260963439941,
      "learning_rate": 1.6851238868981095e-05,
      "loss": 0.1262,
      "step": 485
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1860603392124176,
      "learning_rate": 1.6837498938850228e-05,
      "loss": 0.1353,
      "step": 486
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.22289137542247772,
      "learning_rate": 1.6823734724413777e-05,
      "loss": 0.1278,
      "step": 487
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.19846822321414948,
      "learning_rate": 1.6809946274557228e-05,
      "loss": 0.1262,
      "step": 488
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.21490567922592163,
      "learning_rate": 1.6796133638252126e-05,
      "loss": 0.1373,
      "step": 489
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.172812357544899,
      "learning_rate": 1.6782296864555924e-05,
      "loss": 0.1275,
      "step": 490
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.27832940220832825,
      "learning_rate": 1.6768436002611808e-05,
      "loss": 0.1401,
      "step": 491
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.2117912471294403,
      "learning_rate": 1.6754551101648497e-05,
      "loss": 0.1317,
      "step": 492
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.19993631541728973,
      "learning_rate": 1.6740642210980117e-05,
      "loss": 0.1428,
      "step": 493
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.24337910115718842,
      "learning_rate": 1.6726709380005968e-05,
      "loss": 0.1362,
      "step": 494
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.23150412738323212,
      "learning_rate": 1.67127526582104e-05,
      "loss": 0.1349,
      "step": 495
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.24470743536949158,
      "learning_rate": 1.66987720951626e-05,
      "loss": 0.1305,
      "step": 496
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.20445090532302856,
      "learning_rate": 1.668476774051644e-05,
      "loss": 0.1371,
      "step": 497
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.25253817439079285,
      "learning_rate": 1.6670739644010276e-05,
      "loss": 0.1354,
      "step": 498
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.2221703976392746,
      "learning_rate": 1.6656687855466807e-05,
      "loss": 0.1338,
      "step": 499
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.17464248836040497,
      "learning_rate": 1.664261242479287e-05,
      "loss": 0.1291,
      "step": 500
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.27430105209350586,
      "learning_rate": 1.6628513401979252e-05,
      "loss": 0.1459,
      "step": 501
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.22063203155994415,
      "learning_rate": 1.6614390837100565e-05,
      "loss": 0.14,
      "step": 502
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.30432701110839844,
      "learning_rate": 1.6600244780315e-05,
      "loss": 0.1363,
      "step": 503
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.20645496249198914,
      "learning_rate": 1.6586075281864205e-05,
      "loss": 0.1272,
      "step": 504
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.20736758410930634,
      "learning_rate": 1.6571882392073076e-05,
      "loss": 0.1291,
      "step": 505
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.21969826519489288,
      "learning_rate": 1.655766616134959e-05,
      "loss": 0.1474,
      "step": 506
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.2293238639831543,
      "learning_rate": 1.6543426640184618e-05,
      "loss": 0.1334,
      "step": 507
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.2698013484477997,
      "learning_rate": 1.6529163879151752e-05,
      "loss": 0.1358,
      "step": 508
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.2137279212474823,
      "learning_rate": 1.6514877928907122e-05,
      "loss": 0.132,
      "step": 509
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.2193555384874344,
      "learning_rate": 1.650056884018922e-05,
      "loss": 0.1474,
      "step": 510
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.17913977801799774,
      "learning_rate": 1.6486236663818718e-05,
      "loss": 0.1361,
      "step": 511
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.19582532346248627,
      "learning_rate": 1.647188145069829e-05,
      "loss": 0.1273,
      "step": 512
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.22956804931163788,
      "learning_rate": 1.6457503251812417e-05,
      "loss": 0.1291,
      "step": 513
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.2083774358034134,
      "learning_rate": 1.6443102118227226e-05,
      "loss": 0.128,
      "step": 514
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.21824759244918823,
      "learning_rate": 1.64286781010903e-05,
      "loss": 0.1306,
      "step": 515
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1954454779624939,
      "learning_rate": 1.641423125163049e-05,
      "loss": 0.1277,
      "step": 516
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1896297186613083,
      "learning_rate": 1.6399761621157746e-05,
      "loss": 0.1314,
      "step": 517
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.18475739657878876,
      "learning_rate": 1.6385269261062923e-05,
      "loss": 0.1294,
      "step": 518
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.22530777752399445,
      "learning_rate": 1.6370754222817607e-05,
      "loss": 0.1219,
      "step": 519
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.25659915804862976,
      "learning_rate": 1.6356216557973926e-05,
      "loss": 0.1331,
      "step": 520
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.22420406341552734,
      "learning_rate": 1.634165631816437e-05,
      "loss": 0.1361,
      "step": 521
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.3164658844470978,
      "learning_rate": 1.63270735551016e-05,
      "loss": 0.132,
      "step": 522
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.2204064577817917,
      "learning_rate": 1.631246832057829e-05,
      "loss": 0.1271,
      "step": 523
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.19449515640735626,
      "learning_rate": 1.629784066646691e-05,
      "loss": 0.1228,
      "step": 524
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.29996275901794434,
      "learning_rate": 1.628319064471955e-05,
      "loss": 0.1414,
      "step": 525
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.2644004821777344,
      "learning_rate": 1.6268518307367763e-05,
      "loss": 0.1483,
      "step": 526
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1952865719795227,
      "learning_rate": 1.6253823706522334e-05,
      "loss": 0.1194,
      "step": 527
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.3027459681034088,
      "learning_rate": 1.6239106894373146e-05,
      "loss": 0.1439,
      "step": 528
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.2112572342157364,
      "learning_rate": 1.6224367923188947e-05,
      "loss": 0.1286,
      "step": 529
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.27474159002304077,
      "learning_rate": 1.6209606845317195e-05,
      "loss": 0.1291,
      "step": 530
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.2539788782596588,
      "learning_rate": 1.619482371318386e-05,
      "loss": 0.1345,
      "step": 531
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.26599258184432983,
      "learning_rate": 1.6180018579293246e-05,
      "loss": 0.133,
      "step": 532
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.2407587766647339,
      "learning_rate": 1.6165191496227795e-05,
      "loss": 0.1316,
      "step": 533
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.2652303874492645,
      "learning_rate": 1.6150342516647904e-05,
      "loss": 0.1356,
      "step": 534
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.2322882115840912,
      "learning_rate": 1.6135471693291736e-05,
      "loss": 0.1272,
      "step": 535
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.22672323882579803,
      "learning_rate": 1.6120579078975047e-05,
      "loss": 0.1307,
      "step": 536
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.2555321156978607,
      "learning_rate": 1.610566472659097e-05,
      "loss": 0.1241,
      "step": 537
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.3590485453605652,
      "learning_rate": 1.609072868910986e-05,
      "loss": 0.1306,
      "step": 538
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.20348507165908813,
      "learning_rate": 1.6075771019579072e-05,
      "loss": 0.1256,
      "step": 539
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.23639464378356934,
      "learning_rate": 1.6060791771122802e-05,
      "loss": 0.1245,
      "step": 540
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.22655493021011353,
      "learning_rate": 1.6045790996941893e-05,
      "loss": 0.1225,
      "step": 541
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.24429163336753845,
      "learning_rate": 1.6030768750313624e-05,
      "loss": 0.1293,
      "step": 542
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.23465780913829803,
      "learning_rate": 1.601572508459154e-05,
      "loss": 0.1211,
      "step": 543
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.21169759333133698,
      "learning_rate": 1.6000660053205265e-05,
      "loss": 0.1263,
      "step": 544
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.2709573805332184,
      "learning_rate": 1.5985573709660305e-05,
      "loss": 0.1333,
      "step": 545
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.24457389116287231,
      "learning_rate": 1.5970466107537852e-05,
      "loss": 0.1272,
      "step": 546
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.2838245630264282,
      "learning_rate": 1.5955337300494612e-05,
      "loss": 0.1359,
      "step": 547
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.3450663983821869,
      "learning_rate": 1.594018734226259e-05,
      "loss": 0.1313,
      "step": 548
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.2677037715911865,
      "learning_rate": 1.5925016286648924e-05,
      "loss": 0.1393,
      "step": 549
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.21865220367908478,
      "learning_rate": 1.5909824187535673e-05,
      "loss": 0.1429,
      "step": 550
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.22864189743995667,
      "learning_rate": 1.5894611098879643e-05,
      "loss": 0.1323,
      "step": 551
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.26998138427734375,
      "learning_rate": 1.5879377074712185e-05,
      "loss": 0.1311,
      "step": 552
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.22603292763233185,
      "learning_rate": 1.586412216913899e-05,
      "loss": 0.1307,
      "step": 553
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.18835076689720154,
      "learning_rate": 1.5848846436339945e-05,
      "loss": 0.1257,
      "step": 554
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.2411699891090393,
      "learning_rate": 1.583354993056888e-05,
      "loss": 0.1265,
      "step": 555
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.2197343409061432,
      "learning_rate": 1.581823270615341e-05,
      "loss": 0.1214,
      "step": 556
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.19162169098854065,
      "learning_rate": 1.5802894817494738e-05,
      "loss": 0.1351,
      "step": 557
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.19216634333133698,
      "learning_rate": 1.5787536319067454e-05,
      "loss": 0.1309,
      "step": 558
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.23687519133090973,
      "learning_rate": 1.5772157265419357e-05,
      "loss": 0.1331,
      "step": 559
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.2712344825267792,
      "learning_rate": 1.575675771117124e-05,
      "loss": 0.1345,
      "step": 560
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.25521332025527954,
      "learning_rate": 1.5741337711016707e-05,
      "loss": 0.144,
      "step": 561
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.2198556512594223,
      "learning_rate": 1.5725897319721988e-05,
      "loss": 0.1322,
      "step": 562
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.22969600558280945,
      "learning_rate": 1.5710436592125724e-05,
      "loss": 0.1342,
      "step": 563
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.24502252042293549,
      "learning_rate": 1.569495558313879e-05,
      "loss": 0.1385,
      "step": 564
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.27333834767341614,
      "learning_rate": 1.5679454347744086e-05,
      "loss": 0.1331,
      "step": 565
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.22713664174079895,
      "learning_rate": 1.566393294099636e-05,
      "loss": 0.1277,
      "step": 566
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.24380964040756226,
      "learning_rate": 1.564839141802199e-05,
      "loss": 0.1393,
      "step": 567
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.2457427680492401,
      "learning_rate": 1.5632829834018808e-05,
      "loss": 0.1326,
      "step": 568
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.30003249645233154,
      "learning_rate": 1.561724824425589e-05,
      "loss": 0.1311,
      "step": 569
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.29111576080322266,
      "learning_rate": 1.5601646704073368e-05,
      "loss": 0.1281,
      "step": 570
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.2141617238521576,
      "learning_rate": 1.5586025268882225e-05,
      "loss": 0.1288,
      "step": 571
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.2665215730667114,
      "learning_rate": 1.5570383994164115e-05,
      "loss": 0.1388,
      "step": 572
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.18581970036029816,
      "learning_rate": 1.5554722935471148e-05,
      "loss": 0.1173,
      "step": 573
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.22161142528057098,
      "learning_rate": 1.5539042148425695e-05,
      "loss": 0.1277,
      "step": 574
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.2058258205652237,
      "learning_rate": 1.5523341688720202e-05,
      "loss": 0.1276,
      "step": 575
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.2367851287126541,
      "learning_rate": 1.5507621612116987e-05,
      "loss": 0.134,
      "step": 576
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.2826822102069855,
      "learning_rate": 1.5491881974448025e-05,
      "loss": 0.1347,
      "step": 577
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.22101634740829468,
      "learning_rate": 1.5476122831614785e-05,
      "loss": 0.1239,
      "step": 578
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.2560238242149353,
      "learning_rate": 1.5460344239587998e-05,
      "loss": 0.1275,
      "step": 579
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.2069055140018463,
      "learning_rate": 1.5444546254407477e-05,
      "loss": 0.1233,
      "step": 580
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.2712881565093994,
      "learning_rate": 1.542872893218191e-05,
      "loss": 0.1342,
      "step": 581
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.28810566663742065,
      "learning_rate": 1.541289232908867e-05,
      "loss": 0.1363,
      "step": 582
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.2699289321899414,
      "learning_rate": 1.5397036501373588e-05,
      "loss": 0.1424,
      "step": 583
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.24033977091312408,
      "learning_rate": 1.53811615053508e-05,
      "loss": 0.1363,
      "step": 584
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.19432562589645386,
      "learning_rate": 1.536526739740251e-05,
      "loss": 0.1356,
      "step": 585
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.17603319883346558,
      "learning_rate": 1.534935423397879e-05,
      "loss": 0.128,
      "step": 586
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.26333582401275635,
      "learning_rate": 1.533342207159741e-05,
      "loss": 0.1358,
      "step": 587
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.2180451601743698,
      "learning_rate": 1.53174709668436e-05,
      "loss": 0.1324,
      "step": 588
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.23847247660160065,
      "learning_rate": 1.5301500976369873e-05,
      "loss": 0.1342,
      "step": 589
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.2516377568244934,
      "learning_rate": 1.528551215689583e-05,
      "loss": 0.1245,
      "step": 590
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.2832164764404297,
      "learning_rate": 1.5269504565207922e-05,
      "loss": 0.1392,
      "step": 591
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.2872714400291443,
      "learning_rate": 1.5253478258159287e-05,
      "loss": 0.1359,
      "step": 592
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.18683849275112152,
      "learning_rate": 1.5237433292669527e-05,
      "loss": 0.1183,
      "step": 593
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.2903664708137512,
      "learning_rate": 1.5221369725724522e-05,
      "loss": 0.1347,
      "step": 594
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.21880270540714264,
      "learning_rate": 1.5205287614376201e-05,
      "loss": 0.1209,
      "step": 595
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.21589195728302002,
      "learning_rate": 1.5189187015742373e-05,
      "loss": 0.1273,
      "step": 596
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.23598885536193848,
      "learning_rate": 1.5173067987006497e-05,
      "loss": 0.1305,
      "step": 597
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.257012277841568,
      "learning_rate": 1.5156930585417492e-05,
      "loss": 0.1263,
      "step": 598
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.23764438927173615,
      "learning_rate": 1.514077486828953e-05,
      "loss": 0.1312,
      "step": 599
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.22329208254814148,
      "learning_rate": 1.5124600893001831e-05,
      "loss": 0.1367,
      "step": 600
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.2199406623840332,
      "learning_rate": 1.5108408716998465e-05,
      "loss": 0.1329,
      "step": 601
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.21324387192726135,
      "learning_rate": 1.5092198397788145e-05,
      "loss": 0.1314,
      "step": 602
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.2264215350151062,
      "learning_rate": 1.5075969992944018e-05,
      "loss": 0.1284,
      "step": 603
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.20683738589286804,
      "learning_rate": 1.5059723560103466e-05,
      "loss": 0.1266,
      "step": 604
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.23490509390830994,
      "learning_rate": 1.5043459156967903e-05,
      "loss": 0.1313,
      "step": 605
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.2654430866241455,
      "learning_rate": 1.5027176841302563e-05,
      "loss": 0.1343,
      "step": 606
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.26405930519104004,
      "learning_rate": 1.5010876670936297e-05,
      "loss": 0.1281,
      "step": 607
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.23888154327869415,
      "learning_rate": 1.4994558703761377e-05,
      "loss": 0.1316,
      "step": 608
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.18824058771133423,
      "learning_rate": 1.4978222997733278e-05,
      "loss": 0.1261,
      "step": 609
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.22767682373523712,
      "learning_rate": 1.4961869610870477e-05,
      "loss": 0.1346,
      "step": 610
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.21578750014305115,
      "learning_rate": 1.4945498601254244e-05,
      "loss": 0.1262,
      "step": 611
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.18218088150024414,
      "learning_rate": 1.4929110027028448e-05,
      "loss": 0.1295,
      "step": 612
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.17604951560497284,
      "learning_rate": 1.4912703946399328e-05,
      "loss": 0.1251,
      "step": 613
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.259363055229187,
      "learning_rate": 1.489628041763531e-05,
      "loss": 0.1338,
      "step": 614
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.27487215399742126,
      "learning_rate": 1.4879839499066786e-05,
      "loss": 0.1363,
      "step": 615
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.24068598449230194,
      "learning_rate": 1.4863381249085909e-05,
      "loss": 0.1248,
      "step": 616
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.23120303452014923,
      "learning_rate": 1.4846905726146388e-05,
      "loss": 0.1212,
      "step": 617
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.29931211471557617,
      "learning_rate": 1.4830412988763278e-05,
      "loss": 0.1337,
      "step": 618
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.25536733865737915,
      "learning_rate": 1.4813903095512774e-05,
      "loss": 0.1259,
      "step": 619
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.2542603313922882,
      "learning_rate": 1.4797376105032007e-05,
      "loss": 0.1211,
      "step": 620
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.3510386347770691,
      "learning_rate": 1.4780832076018824e-05,
      "loss": 0.1397,
      "step": 621
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.37623947858810425,
      "learning_rate": 1.4764271067231591e-05,
      "loss": 0.1428,
      "step": 622
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.2595818042755127,
      "learning_rate": 1.4747693137488974e-05,
      "loss": 0.1288,
      "step": 623
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.2907295525074005,
      "learning_rate": 1.4731098345669748e-05,
      "loss": 0.1368,
      "step": 624
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.29793843626976013,
      "learning_rate": 1.471448675071256e-05,
      "loss": 0.1249,
      "step": 625
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.2708103060722351,
      "learning_rate": 1.4697858411615753e-05,
      "loss": 0.1363,
      "step": 626
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.2683532238006592,
      "learning_rate": 1.4681213387437125e-05,
      "loss": 0.1298,
      "step": 627
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.24503564834594727,
      "learning_rate": 1.4664551737293742e-05,
      "loss": 0.131,
      "step": 628
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.2394891381263733,
      "learning_rate": 1.4647873520361713e-05,
      "loss": 0.1287,
      "step": 629
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.2935561537742615,
      "learning_rate": 1.4631178795875994e-05,
      "loss": 0.1298,
      "step": 630
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.28654900193214417,
      "learning_rate": 1.4614467623130155e-05,
      "loss": 0.1355,
      "step": 631
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.25191953778266907,
      "learning_rate": 1.4597740061476207e-05,
      "loss": 0.1252,
      "step": 632
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.2829602360725403,
      "learning_rate": 1.4580996170324347e-05,
      "loss": 0.1427,
      "step": 633
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.27054595947265625,
      "learning_rate": 1.4564236009142781e-05,
      "loss": 0.1267,
      "step": 634
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.28524667024612427,
      "learning_rate": 1.4547459637457494e-05,
      "loss": 0.1294,
      "step": 635
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.33464595675468445,
      "learning_rate": 1.4530667114852051e-05,
      "loss": 0.1361,
      "step": 636
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.25658556818962097,
      "learning_rate": 1.451385850096737e-05,
      "loss": 0.1168,
      "step": 637
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.25373998284339905,
      "learning_rate": 1.449703385550153e-05,
      "loss": 0.1245,
      "step": 638
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.26443809270858765,
      "learning_rate": 1.448019323820954e-05,
      "loss": 0.1293,
      "step": 639
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.3036229908466339,
      "learning_rate": 1.4463336708903142e-05,
      "loss": 0.1283,
      "step": 640
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.299350380897522,
      "learning_rate": 1.4446464327450582e-05,
      "loss": 0.1378,
      "step": 641
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.22419188916683197,
      "learning_rate": 1.4429576153776417e-05,
      "loss": 0.1284,
      "step": 642
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.2839418947696686,
      "learning_rate": 1.4412672247861285e-05,
      "loss": 0.1325,
      "step": 643
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.3462979197502136,
      "learning_rate": 1.4395752669741705e-05,
      "loss": 0.1284,
      "step": 644
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.25637516379356384,
      "learning_rate": 1.4378817479509854e-05,
      "loss": 0.1276,
      "step": 645
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.2482435405254364,
      "learning_rate": 1.4361866737313362e-05,
      "loss": 0.123,
      "step": 646
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.2543370723724365,
      "learning_rate": 1.4344900503355086e-05,
      "loss": 0.1395,
      "step": 647
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.2602517008781433,
      "learning_rate": 1.4327918837892911e-05,
      "loss": 0.1265,
      "step": 648
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.26915252208709717,
      "learning_rate": 1.431092180123953e-05,
      "loss": 0.1234,
      "step": 649
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.24584051966667175,
      "learning_rate": 1.429390945376222e-05,
      "loss": 0.1258,
      "step": 650
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.31423231959342957,
      "learning_rate": 1.4276881855882648e-05,
      "loss": 0.1467,
      "step": 651
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.2718120813369751,
      "learning_rate": 1.4259839068076628e-05,
      "loss": 0.1325,
      "step": 652
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.23196832835674286,
      "learning_rate": 1.424278115087395e-05,
      "loss": 0.1183,
      "step": 653
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.2673061490058899,
      "learning_rate": 1.4225708164858114e-05,
      "loss": 0.1247,
      "step": 654
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.2473834902048111,
      "learning_rate": 1.4208620170666146e-05,
      "loss": 0.1313,
      "step": 655
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.22686772048473358,
      "learning_rate": 1.419151722898838e-05,
      "loss": 0.134,
      "step": 656
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.22457249462604523,
      "learning_rate": 1.4174399400568233e-05,
      "loss": 0.128,
      "step": 657
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.2764395475387573,
      "learning_rate": 1.4157266746201996e-05,
      "loss": 0.1279,
      "step": 658
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.23162925243377686,
      "learning_rate": 1.414011932673862e-05,
      "loss": 0.1211,
      "step": 659
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.2702430486679077,
      "learning_rate": 1.4122957203079489e-05,
      "loss": 0.1345,
      "step": 660
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.25539204478263855,
      "learning_rate": 1.4105780436178212e-05,
      "loss": 0.119,
      "step": 661
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.2659144699573517,
      "learning_rate": 1.4088589087040414e-05,
      "loss": 0.1341,
      "step": 662
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.24564450979232788,
      "learning_rate": 1.4071383216723504e-05,
      "loss": 0.12,
      "step": 663
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.2496199607849121,
      "learning_rate": 1.4054162886336466e-05,
      "loss": 0.1214,
      "step": 664
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.295223206281662,
      "learning_rate": 1.4036928157039638e-05,
      "loss": 0.131,
      "step": 665
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.3617158532142639,
      "learning_rate": 1.4019679090044503e-05,
      "loss": 0.1333,
      "step": 666
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.22648681700229645,
      "learning_rate": 1.4002415746613462e-05,
      "loss": 0.1181,
      "step": 667
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.2870987355709076,
      "learning_rate": 1.3985138188059622e-05,
      "loss": 0.1261,
      "step": 668
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.2604856491088867,
      "learning_rate": 1.3967846475746578e-05,
      "loss": 0.1248,
      "step": 669
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.29645872116088867,
      "learning_rate": 1.3950540671088198e-05,
      "loss": 0.1283,
      "step": 670
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.33818647265434265,
      "learning_rate": 1.393322083554839e-05,
      "loss": 0.13,
      "step": 671
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.2933283746242523,
      "learning_rate": 1.3915887030640905e-05,
      "loss": 0.1291,
      "step": 672
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.2696940004825592,
      "learning_rate": 1.3898539317929101e-05,
      "loss": 0.1344,
      "step": 673
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.2504938840866089,
      "learning_rate": 1.3881177759025732e-05,
      "loss": 0.1196,
      "step": 674
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.27981093525886536,
      "learning_rate": 1.3863802415592738e-05,
      "loss": 0.1196,
      "step": 675
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.31877508759498596,
      "learning_rate": 1.3846413349341006e-05,
      "loss": 0.1195,
      "step": 676
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.23064351081848145,
      "learning_rate": 1.3829010622030166e-05,
      "loss": 0.1275,
      "step": 677
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.355947345495224,
      "learning_rate": 1.3811594295468365e-05,
      "loss": 0.1272,
      "step": 678
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.3581264317035675,
      "learning_rate": 1.3794164431512052e-05,
      "loss": 0.1234,
      "step": 679
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.2725159823894501,
      "learning_rate": 1.377672109206575e-05,
      "loss": 0.1243,
      "step": 680
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.26085466146469116,
      "learning_rate": 1.3759264339081855e-05,
      "loss": 0.1278,
      "step": 681
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.28182634711265564,
      "learning_rate": 1.374179423456039e-05,
      "loss": 0.1331,
      "step": 682
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.29763472080230713,
      "learning_rate": 1.3724310840548804e-05,
      "loss": 0.1222,
      "step": 683
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.3194727301597595,
      "learning_rate": 1.3706814219141742e-05,
      "loss": 0.1303,
      "step": 684
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.3084741234779358,
      "learning_rate": 1.3689304432480832e-05,
      "loss": 0.1308,
      "step": 685
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.29822906851768494,
      "learning_rate": 1.3671781542754455e-05,
      "loss": 0.1294,
      "step": 686
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.3116920590400696,
      "learning_rate": 1.3654245612197534e-05,
      "loss": 0.1312,
      "step": 687
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.26224011182785034,
      "learning_rate": 1.363669670309131e-05,
      "loss": 0.1241,
      "step": 688
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.2982347905635834,
      "learning_rate": 1.3619134877763112e-05,
      "loss": 0.1285,
      "step": 689
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.2621435523033142,
      "learning_rate": 1.3601560198586145e-05,
      "loss": 0.1317,
      "step": 690
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.3401296138763428,
      "learning_rate": 1.3583972727979274e-05,
      "loss": 0.1297,
      "step": 691
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.2582213878631592,
      "learning_rate": 1.3566372528406777e-05,
      "loss": 0.121,
      "step": 692
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.252268522977829,
      "learning_rate": 1.3548759662378163e-05,
      "loss": 0.1288,
      "step": 693
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.25547274947166443,
      "learning_rate": 1.3531134192447913e-05,
      "loss": 0.1343,
      "step": 694
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.2503824234008789,
      "learning_rate": 1.3513496181215273e-05,
      "loss": 0.1295,
      "step": 695
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.2730405926704407,
      "learning_rate": 1.349584569132404e-05,
      "loss": 0.1297,
      "step": 696
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.2788792550563812,
      "learning_rate": 1.3478182785462317e-05,
      "loss": 0.1367,
      "step": 697
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.30198124051094055,
      "learning_rate": 1.3460507526362311e-05,
      "loss": 0.1229,
      "step": 698
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.29159095883369446,
      "learning_rate": 1.3442819976800114e-05,
      "loss": 0.1221,
      "step": 699
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.3546750545501709,
      "learning_rate": 1.3425120199595455e-05,
      "loss": 0.1275,
      "step": 700
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.2808311879634857,
      "learning_rate": 1.340740825761149e-05,
      "loss": 0.1362,
      "step": 701
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.31112587451934814,
      "learning_rate": 1.338968421375459e-05,
      "loss": 0.1234,
      "step": 702
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.28536108136177063,
      "learning_rate": 1.33719481309741e-05,
      "loss": 0.1328,
      "step": 703
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.29962673783302307,
      "learning_rate": 1.3354200072262124e-05,
      "loss": 0.118,
      "step": 704
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.24461856484413147,
      "learning_rate": 1.3336440100653309e-05,
      "loss": 0.1229,
      "step": 705
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.24503368139266968,
      "learning_rate": 1.3318668279224598e-05,
      "loss": 0.1219,
      "step": 706
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.2203916311264038,
      "learning_rate": 1.3300884671095032e-05,
      "loss": 0.121,
      "step": 707
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.25306275486946106,
      "learning_rate": 1.3283089339425504e-05,
      "loss": 0.1314,
      "step": 708
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.24175699055194855,
      "learning_rate": 1.3265282347418551e-05,
      "loss": 0.1246,
      "step": 709
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.24566194415092468,
      "learning_rate": 1.3247463758318122e-05,
      "loss": 0.1267,
      "step": 710
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.28256329894065857,
      "learning_rate": 1.3229633635409358e-05,
      "loss": 0.12,
      "step": 711
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.2806382179260254,
      "learning_rate": 1.3211792042018361e-05,
      "loss": 0.1302,
      "step": 712
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.30707529187202454,
      "learning_rate": 1.319393904151197e-05,
      "loss": 0.1231,
      "step": 713
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.352463036775589,
      "learning_rate": 1.3176074697297536e-05,
      "loss": 0.125,
      "step": 714
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.40101900696754456,
      "learning_rate": 1.3158199072822707e-05,
      "loss": 0.1238,
      "step": 715
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.37553203105926514,
      "learning_rate": 1.3140312231575186e-05,
      "loss": 0.1265,
      "step": 716
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.3193184733390808,
      "learning_rate": 1.3122414237082521e-05,
      "loss": 0.1292,
      "step": 717
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.3477811813354492,
      "learning_rate": 1.3104505152911871e-05,
      "loss": 0.1319,
      "step": 718
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.3007843494415283,
      "learning_rate": 1.3086585042669773e-05,
      "loss": 0.1374,
      "step": 719
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.2771112024784088,
      "learning_rate": 1.3068653970001935e-05,
      "loss": 0.1236,
      "step": 720
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.28000134229660034,
      "learning_rate": 1.3050711998592993e-05,
      "loss": 0.1379,
      "step": 721
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.3654780089855194,
      "learning_rate": 1.3032759192166293e-05,
      "loss": 0.1207,
      "step": 722
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.23554198443889618,
      "learning_rate": 1.3014795614483671e-05,
      "loss": 0.1244,
      "step": 723
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.30108940601348877,
      "learning_rate": 1.2996821329345204e-05,
      "loss": 0.1175,
      "step": 724
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.3085353970527649,
      "learning_rate": 1.2978836400589008e-05,
      "loss": 0.1338,
      "step": 725
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.23512117564678192,
      "learning_rate": 1.2960840892090998e-05,
      "loss": 0.1203,
      "step": 726
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.29902541637420654,
      "learning_rate": 1.2942834867764669e-05,
      "loss": 0.1373,
      "step": 727
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.292229026556015,
      "learning_rate": 1.292481839156085e-05,
      "loss": 0.1324,
      "step": 728
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.34417927265167236,
      "learning_rate": 1.2906791527467517e-05,
      "loss": 0.1299,
      "step": 729
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.2394486516714096,
      "learning_rate": 1.2888754339509513e-05,
      "loss": 0.1194,
      "step": 730
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.2529979646205902,
      "learning_rate": 1.2870706891748364e-05,
      "loss": 0.1316,
      "step": 731
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.36386552453041077,
      "learning_rate": 1.285264924828203e-05,
      "loss": 0.1406,
      "step": 732
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.25270214676856995,
      "learning_rate": 1.2834581473244682e-05,
      "loss": 0.1264,
      "step": 733
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.2924668490886688,
      "learning_rate": 1.2816503630806474e-05,
      "loss": 0.132,
      "step": 734
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.3351571559906006,
      "learning_rate": 1.279841578517332e-05,
      "loss": 0.1382,
      "step": 735
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.25560733675956726,
      "learning_rate": 1.2780318000586658e-05,
      "loss": 0.1338,
      "step": 736
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.2729761004447937,
      "learning_rate": 1.2762210341323227e-05,
      "loss": 0.1178,
      "step": 737
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.2758296728134155,
      "learning_rate": 1.2744092871694837e-05,
      "loss": 0.1367,
      "step": 738
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.2934372127056122,
      "learning_rate": 1.2725965656048138e-05,
      "loss": 0.1336,
      "step": 739
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.2869187891483307,
      "learning_rate": 1.27078287587644e-05,
      "loss": 0.127,
      "step": 740
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.300844669342041,
      "learning_rate": 1.2689682244259274e-05,
      "loss": 0.1227,
      "step": 741
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.286884605884552,
      "learning_rate": 1.2671526176982567e-05,
      "loss": 0.127,
      "step": 742
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.30446118116378784,
      "learning_rate": 1.2653360621418023e-05,
      "loss": 0.1248,
      "step": 743
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.2709696888923645,
      "learning_rate": 1.2635185642083073e-05,
      "loss": 0.1278,
      "step": 744
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.3100476861000061,
      "learning_rate": 1.2617001303528621e-05,
      "loss": 0.1335,
      "step": 745
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.28725165128707886,
      "learning_rate": 1.2598807670338821e-05,
      "loss": 0.1166,
      "step": 746
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.38413098454475403,
      "learning_rate": 1.2580604807130828e-05,
      "loss": 0.1375,
      "step": 747
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.29943931102752686,
      "learning_rate": 1.2562392778554577e-05,
      "loss": 0.1286,
      "step": 748
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.30691230297088623,
      "learning_rate": 1.2544171649292567e-05,
      "loss": 0.1124,
      "step": 749
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.28991368412971497,
      "learning_rate": 1.2525941484059606e-05,
      "loss": 0.1219,
      "step": 750
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.30914247035980225,
      "learning_rate": 1.2507702347602597e-05,
      "loss": 0.1141,
      "step": 751
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.3269689083099365,
      "learning_rate": 1.248945430470032e-05,
      "loss": 0.1366,
      "step": 752
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.2920926809310913,
      "learning_rate": 1.2471197420163169e-05,
      "loss": 0.1252,
      "step": 753
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.26600247621536255,
      "learning_rate": 1.2452931758832947e-05,
      "loss": 0.1208,
      "step": 754
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.35574987530708313,
      "learning_rate": 1.2434657385582634e-05,
      "loss": 0.12,
      "step": 755
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.3068075180053711,
      "learning_rate": 1.2416374365316141e-05,
      "loss": 0.1166,
      "step": 756
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.3685455918312073,
      "learning_rate": 1.2398082762968102e-05,
      "loss": 0.1193,
      "step": 757
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.36364829540252686,
      "learning_rate": 1.2379782643503622e-05,
      "loss": 0.1245,
      "step": 758
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.37386736273765564,
      "learning_rate": 1.236147407191806e-05,
      "loss": 0.1167,
      "step": 759
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.3883266746997833,
      "learning_rate": 1.2343157113236795e-05,
      "loss": 0.1279,
      "step": 760
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.34696564078330994,
      "learning_rate": 1.232483183251499e-05,
      "loss": 0.12,
      "step": 761
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.4503585398197174,
      "learning_rate": 1.2306498294837366e-05,
      "loss": 0.1304,
      "step": 762
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.3366601765155792,
      "learning_rate": 1.2288156565317975e-05,
      "loss": 0.1383,
      "step": 763
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.3837183713912964,
      "learning_rate": 1.2269806709099954e-05,
      "loss": 0.132,
      "step": 764
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.3730437159538269,
      "learning_rate": 1.2251448791355309e-05,
      "loss": 0.1258,
      "step": 765
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.36076053977012634,
      "learning_rate": 1.2233082877284679e-05,
      "loss": 0.1289,
      "step": 766
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.39444661140441895,
      "learning_rate": 1.22147090321171e-05,
      "loss": 0.128,
      "step": 767
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.33830878138542175,
      "learning_rate": 1.2196327321109778e-05,
      "loss": 0.1281,
      "step": 768
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.3627125322818756,
      "learning_rate": 1.217793780954785e-05,
      "loss": 0.128,
      "step": 769
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.2689242959022522,
      "learning_rate": 1.2159540562744169e-05,
      "loss": 0.1185,
      "step": 770
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.31704840064048767,
      "learning_rate": 1.2141135646039048e-05,
      "loss": 0.1274,
      "step": 771
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.2958633303642273,
      "learning_rate": 1.2122723124800054e-05,
      "loss": 0.1201,
      "step": 772
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.26740583777427673,
      "learning_rate": 1.210430306442175e-05,
      "loss": 0.1281,
      "step": 773
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.23514264822006226,
      "learning_rate": 1.2085875530325482e-05,
      "loss": 0.1215,
      "step": 774
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.2842799723148346,
      "learning_rate": 1.2067440587959138e-05,
      "loss": 0.1271,
      "step": 775
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.30431070923805237,
      "learning_rate": 1.204899830279692e-05,
      "loss": 0.1246,
      "step": 776
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.2661820352077484,
      "learning_rate": 1.2030548740339102e-05,
      "loss": 0.1227,
      "step": 777
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.3203689754009247,
      "learning_rate": 1.2012091966111816e-05,
      "loss": 0.1317,
      "step": 778
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.28192949295043945,
      "learning_rate": 1.1993628045666799e-05,
      "loss": 0.1251,
      "step": 779
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.31617242097854614,
      "learning_rate": 1.197515704458117e-05,
      "loss": 0.1183,
      "step": 780
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.34724292159080505,
      "learning_rate": 1.1956679028457196e-05,
      "loss": 0.1191,
      "step": 781
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.38423895835876465,
      "learning_rate": 1.193819406292206e-05,
      "loss": 0.1252,
      "step": 782
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.31843841075897217,
      "learning_rate": 1.1919702213627624e-05,
      "loss": 0.1179,
      "step": 783
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.299985408782959,
      "learning_rate": 1.1901203546250205e-05,
      "loss": 0.1213,
      "step": 784
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.36920154094696045,
      "learning_rate": 1.1882698126490327e-05,
      "loss": 0.1251,
      "step": 785
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.30652371048927307,
      "learning_rate": 1.1864186020072503e-05,
      "loss": 0.1257,
      "step": 786
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.37738490104675293,
      "learning_rate": 1.184566729274499e-05,
      "loss": 0.1202,
      "step": 787
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.3405987620353699,
      "learning_rate": 1.1827142010279567e-05,
      "loss": 0.1282,
      "step": 788
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.32858145236968994,
      "learning_rate": 1.1808610238471279e-05,
      "loss": 0.1249,
      "step": 789
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.3222396969795227,
      "learning_rate": 1.1790072043138237e-05,
      "loss": 0.1302,
      "step": 790
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.31144025921821594,
      "learning_rate": 1.1771527490121356e-05,
      "loss": 0.1347,
      "step": 791
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.35553333163261414,
      "learning_rate": 1.1752976645284133e-05,
      "loss": 0.1229,
      "step": 792
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.2680851221084595,
      "learning_rate": 1.173441957451241e-05,
      "loss": 0.1196,
      "step": 793
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.28336870670318604,
      "learning_rate": 1.1715856343714148e-05,
      "loss": 0.1173,
      "step": 794
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.3197232484817505,
      "learning_rate": 1.1697287018819173e-05,
      "loss": 0.1242,
      "step": 795
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.35233789682388306,
      "learning_rate": 1.1678711665778968e-05,
      "loss": 0.1282,
      "step": 796
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.35812243819236755,
      "learning_rate": 1.1660130350566419e-05,
      "loss": 0.1365,
      "step": 797
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.30994343757629395,
      "learning_rate": 1.164154313917559e-05,
      "loss": 0.1226,
      "step": 798
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.33946338295936584,
      "learning_rate": 1.1622950097621488e-05,
      "loss": 0.121,
      "step": 799
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.24299083650112152,
      "learning_rate": 1.160435129193982e-05,
      "loss": 0.1201,
      "step": 800
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.27806708216667175,
      "learning_rate": 1.1585746788186765e-05,
      "loss": 0.1191,
      "step": 801
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.30519163608551025,
      "learning_rate": 1.1567136652438752e-05,
      "loss": 0.1217,
      "step": 802
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.38272616267204285,
      "learning_rate": 1.15485209507922e-05,
      "loss": 0.1334,
      "step": 803
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.3146188259124756,
      "learning_rate": 1.1529899749363303e-05,
      "loss": 0.1215,
      "step": 804
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.35263314843177795,
      "learning_rate": 1.151127311428778e-05,
      "loss": 0.1171,
      "step": 805
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.3798489570617676,
      "learning_rate": 1.1492641111720657e-05,
      "loss": 0.132,
      "step": 806
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.315523624420166,
      "learning_rate": 1.1474003807836019e-05,
      "loss": 0.119,
      "step": 807
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.32863014936447144,
      "learning_rate": 1.145536126882678e-05,
      "loss": 0.1207,
      "step": 808
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.30183953046798706,
      "learning_rate": 1.1436713560904456e-05,
      "loss": 0.1166,
      "step": 809
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.3046282231807709,
      "learning_rate": 1.14180607502989e-05,
      "loss": 0.1198,
      "step": 810
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.32344409823417664,
      "learning_rate": 1.139940290325811e-05,
      "loss": 0.1267,
      "step": 811
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.3685789406299591,
      "learning_rate": 1.1380740086047958e-05,
      "loss": 0.1343,
      "step": 812
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.3002364933490753,
      "learning_rate": 1.1362072364951971e-05,
      "loss": 0.1225,
      "step": 813
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.31314989924430847,
      "learning_rate": 1.13433998062711e-05,
      "loss": 0.121,
      "step": 814
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.3336435556411743,
      "learning_rate": 1.1324722476323469e-05,
      "loss": 0.1138,
      "step": 815
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.35002535581588745,
      "learning_rate": 1.130604044144415e-05,
      "loss": 0.1274,
      "step": 816
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.406056672334671,
      "learning_rate": 1.1287353767984925e-05,
      "loss": 0.1321,
      "step": 817
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.3289960026741028,
      "learning_rate": 1.1268662522314053e-05,
      "loss": 0.1142,
      "step": 818
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.34095779061317444,
      "learning_rate": 1.1249966770816028e-05,
      "loss": 0.1187,
      "step": 819
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.30711236596107483,
      "learning_rate": 1.1231266579891352e-05,
      "loss": 0.117,
      "step": 820
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.34369957447052,
      "learning_rate": 1.121256201595629e-05,
      "loss": 0.1315,
      "step": 821
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.3208088278770447,
      "learning_rate": 1.1193853145442638e-05,
      "loss": 0.1238,
      "step": 822
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.3982299864292145,
      "learning_rate": 1.1175140034797491e-05,
      "loss": 0.1281,
      "step": 823
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.3457778990268707,
      "learning_rate": 1.1156422750483003e-05,
      "loss": 0.1254,
      "step": 824
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.308923602104187,
      "learning_rate": 1.1137701358976142e-05,
      "loss": 0.1247,
      "step": 825
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.3407731056213379,
      "learning_rate": 1.1118975926768483e-05,
      "loss": 0.1156,
      "step": 826
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.33312690258026123,
      "learning_rate": 1.1100246520365935e-05,
      "loss": 0.1242,
      "step": 827
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.26486337184906006,
      "learning_rate": 1.1081513206288521e-05,
      "loss": 0.1107,
      "step": 828
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.32099536061286926,
      "learning_rate": 1.1062776051070159e-05,
      "loss": 0.1246,
      "step": 829
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.3835618197917938,
      "learning_rate": 1.1044035121258391e-05,
      "loss": 0.1283,
      "step": 830
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.3495025932788849,
      "learning_rate": 1.102529048341418e-05,
      "loss": 0.1142,
      "step": 831
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.34656041860580444,
      "learning_rate": 1.1006542204111644e-05,
      "loss": 0.122,
      "step": 832
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.33667826652526855,
      "learning_rate": 1.0987790349937846e-05,
      "loss": 0.1295,
      "step": 833
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.3694384694099426,
      "learning_rate": 1.0969034987492542e-05,
      "loss": 0.1293,
      "step": 834
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.2614348232746124,
      "learning_rate": 1.0950276183387948e-05,
      "loss": 0.1197,
      "step": 835
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.2672187387943268,
      "learning_rate": 1.09315140042485e-05,
      "loss": 0.1172,
      "step": 836
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.33795592188835144,
      "learning_rate": 1.0912748516710626e-05,
      "loss": 0.1239,
      "step": 837
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.40587592124938965,
      "learning_rate": 1.0893979787422502e-05,
      "loss": 0.1203,
      "step": 838
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.38800135254859924,
      "learning_rate": 1.0875207883043817e-05,
      "loss": 0.1218,
      "step": 839
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.4661693274974823,
      "learning_rate": 1.0856432870245537e-05,
      "loss": 0.1138,
      "step": 840
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.3663930892944336,
      "learning_rate": 1.0837654815709668e-05,
      "loss": 0.1194,
      "step": 841
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.39449039101600647,
      "learning_rate": 1.0818873786129022e-05,
      "loss": 0.1236,
      "step": 842
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.37344130873680115,
      "learning_rate": 1.0800089848206976e-05,
      "loss": 0.1242,
      "step": 843
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.2919343113899231,
      "learning_rate": 1.0781303068657232e-05,
      "loss": 0.1096,
      "step": 844
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.3285546898841858,
      "learning_rate": 1.0762513514203588e-05,
      "loss": 0.1177,
      "step": 845
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.33259719610214233,
      "learning_rate": 1.0743721251579694e-05,
      "loss": 0.1119,
      "step": 846
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.4538424015045166,
      "learning_rate": 1.072492634752883e-05,
      "loss": 0.1369,
      "step": 847
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.3444289267063141,
      "learning_rate": 1.070612886880364e-05,
      "loss": 0.1332,
      "step": 848
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.310960978269577,
      "learning_rate": 1.0687328882165928e-05,
      "loss": 0.1333,
      "step": 849
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.39196282625198364,
      "learning_rate": 1.0668526454386392e-05,
      "loss": 0.1287,
      "step": 850
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.3444955348968506,
      "learning_rate": 1.064972165224441e-05,
      "loss": 0.1294,
      "step": 851
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.3038831651210785,
      "learning_rate": 1.0630914542527793e-05,
      "loss": 0.1176,
      "step": 852
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.3133896589279175,
      "learning_rate": 1.0612105192032539e-05,
      "loss": 0.1128,
      "step": 853
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.38025397062301636,
      "learning_rate": 1.059329366756261e-05,
      "loss": 0.113,
      "step": 854
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.30050504207611084,
      "learning_rate": 1.0574480035929688e-05,
      "loss": 0.1297,
      "step": 855
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.3363340198993683,
      "learning_rate": 1.055566436395294e-05,
      "loss": 0.1143,
      "step": 856
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.28416895866394043,
      "learning_rate": 1.053684671845878e-05,
      "loss": 0.1156,
      "step": 857
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.3077041506767273,
      "learning_rate": 1.0518027166280632e-05,
      "loss": 0.1261,
      "step": 858
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.3824397623538971,
      "learning_rate": 1.0499205774258685e-05,
      "loss": 0.1262,
      "step": 859
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.333523154258728,
      "learning_rate": 1.0480382609239672e-05,
      "loss": 0.1183,
      "step": 860
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.30974864959716797,
      "learning_rate": 1.0461557738076615e-05,
      "loss": 0.1155,
      "step": 861
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.33961594104766846,
      "learning_rate": 1.0442731227628602e-05,
      "loss": 0.1177,
      "step": 862
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.36659467220306396,
      "learning_rate": 1.0423903144760538e-05,
      "loss": 0.1314,
      "step": 863
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.31679567694664,
      "learning_rate": 1.040507355634292e-05,
      "loss": 0.1286,
      "step": 864
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.35801607370376587,
      "learning_rate": 1.0386242529251582e-05,
      "loss": 0.1292,
      "step": 865
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.2850901782512665,
      "learning_rate": 1.0367410130367472e-05,
      "loss": 0.122,
      "step": 866
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.3095703125,
      "learning_rate": 1.0348576426576416e-05,
      "loss": 0.1274,
      "step": 867
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.36840081214904785,
      "learning_rate": 1.0329741484768862e-05,
      "loss": 0.1265,
      "step": 868
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.32150399684906006,
      "learning_rate": 1.0310905371839673e-05,
      "loss": 0.1158,
      "step": 869
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.355345219373703,
      "learning_rate": 1.029206815468785e-05,
      "loss": 0.1191,
      "step": 870
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.27587589621543884,
      "learning_rate": 1.0273229900216335e-05,
      "loss": 0.1164,
      "step": 871
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.350415974855423,
      "learning_rate": 1.025439067533174e-05,
      "loss": 0.1221,
      "step": 872
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.4711579382419586,
      "learning_rate": 1.0235550546944132e-05,
      "loss": 0.1182,
      "step": 873
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.4131545126438141,
      "learning_rate": 1.0216709581966781e-05,
      "loss": 0.1312,
      "step": 874
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.3326095640659332,
      "learning_rate": 1.0197867847315937e-05,
      "loss": 0.1284,
      "step": 875
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.3458001911640167,
      "learning_rate": 1.0179025409910575e-05,
      "loss": 0.1209,
      "step": 876
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.3523850739002228,
      "learning_rate": 1.0160182336672171e-05,
      "loss": 0.1184,
      "step": 877
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.4152713418006897,
      "learning_rate": 1.0141338694524455e-05,
      "loss": 0.1201,
      "step": 878
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.414760023355484,
      "learning_rate": 1.0122494550393178e-05,
      "loss": 0.1255,
      "step": 879
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.317658931016922,
      "learning_rate": 1.0103649971205878e-05,
      "loss": 0.1158,
      "step": 880
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.34657546877861023,
      "learning_rate": 1.0084805023891639e-05,
      "loss": 0.1359,
      "step": 881
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.28743842244148254,
      "learning_rate": 1.0065959775380847e-05,
      "loss": 0.1273,
      "step": 882
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.3483695387840271,
      "learning_rate": 1.0047114292604957e-05,
      "loss": 0.1231,
      "step": 883
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.31169822812080383,
      "learning_rate": 1.0028268642496264e-05,
      "loss": 0.1095,
      "step": 884
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.26510486006736755,
      "learning_rate": 1.000942289198765e-05,
      "loss": 0.1159,
      "step": 885
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.254341721534729,
      "learning_rate": 9.990577108012353e-06,
      "loss": 0.119,
      "step": 886
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.3478366732597351,
      "learning_rate": 9.971731357503741e-06,
      "loss": 0.1287,
      "step": 887
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.3898210823535919,
      "learning_rate": 9.952885707395044e-06,
      "loss": 0.1208,
      "step": 888
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.35076501965522766,
      "learning_rate": 9.934040224619154e-06,
      "loss": 0.1259,
      "step": 889
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.3294999301433563,
      "learning_rate": 9.915194976108363e-06,
      "loss": 0.1235,
      "step": 890
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.38058730959892273,
      "learning_rate": 9.896350028794122e-06,
      "loss": 0.1255,
      "step": 891
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.29607054591178894,
      "learning_rate": 9.877505449606827e-06,
      "loss": 0.1218,
      "step": 892
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.3143492341041565,
      "learning_rate": 9.85866130547555e-06,
      "loss": 0.1266,
      "step": 893
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.2735792100429535,
      "learning_rate": 9.839817663327835e-06,
      "loss": 0.1128,
      "step": 894
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.3311600387096405,
      "learning_rate": 9.820974590089428e-06,
      "loss": 0.1302,
      "step": 895
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.3816298544406891,
      "learning_rate": 9.802132152684063e-06,
      "loss": 0.1352,
      "step": 896
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.3320065438747406,
      "learning_rate": 9.783290418033222e-06,
      "loss": 0.1207,
      "step": 897
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.3045239746570587,
      "learning_rate": 9.76444945305587e-06,
      "loss": 0.1211,
      "step": 898
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.31413891911506653,
      "learning_rate": 9.745609324668266e-06,
      "loss": 0.1131,
      "step": 899
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.43259865045547485,
      "learning_rate": 9.726770099783669e-06,
      "loss": 0.1226,
      "step": 900
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.44788745045661926,
      "learning_rate": 9.70793184531215e-06,
      "loss": 0.1287,
      "step": 901
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.40162384510040283,
      "learning_rate": 9.689094628160332e-06,
      "loss": 0.1287,
      "step": 902
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.2954537570476532,
      "learning_rate": 9.670258515231137e-06,
      "loss": 0.1225,
      "step": 903
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.3706112504005432,
      "learning_rate": 9.651423573423589e-06,
      "loss": 0.1204,
      "step": 904
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.2926884591579437,
      "learning_rate": 9.63258986963253e-06,
      "loss": 0.1177,
      "step": 905
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.3163410425186157,
      "learning_rate": 9.613757470748423e-06,
      "loss": 0.1193,
      "step": 906
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.29135626554489136,
      "learning_rate": 9.594926443657084e-06,
      "loss": 0.1236,
      "step": 907
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.31334686279296875,
      "learning_rate": 9.57609685523946e-06,
      "loss": 0.1191,
      "step": 908
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.35619404911994934,
      "learning_rate": 9.5572687723714e-06,
      "loss": 0.1247,
      "step": 909
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.31272345781326294,
      "learning_rate": 9.538442261923387e-06,
      "loss": 0.1189,
      "step": 910
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.3580284118652344,
      "learning_rate": 9.519617390760333e-06,
      "loss": 0.1121,
      "step": 911
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.35814324021339417,
      "learning_rate": 9.500794225741318e-06,
      "loss": 0.1141,
      "step": 912
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.339386522769928,
      "learning_rate": 9.481972833719373e-06,
      "loss": 0.1095,
      "step": 913
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.3677462041378021,
      "learning_rate": 9.463153281541223e-06,
      "loss": 0.1217,
      "step": 914
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.37574324011802673,
      "learning_rate": 9.444335636047062e-06,
      "loss": 0.1258,
      "step": 915
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.3995684087276459,
      "learning_rate": 9.425519964070317e-06,
      "loss": 0.1109,
      "step": 916
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.5380546450614929,
      "learning_rate": 9.406706332437395e-06,
      "loss": 0.133,
      "step": 917
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.38268473744392395,
      "learning_rate": 9.387894807967466e-06,
      "loss": 0.1214,
      "step": 918
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.38451096415519714,
      "learning_rate": 9.36908545747221e-06,
      "loss": 0.1137,
      "step": 919
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.43880900740623474,
      "learning_rate": 9.350278347755591e-06,
      "loss": 0.1356,
      "step": 920
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.40110471844673157,
      "learning_rate": 9.331473545613611e-06,
      "loss": 0.1248,
      "step": 921
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.43446022272109985,
      "learning_rate": 9.312671117834076e-06,
      "loss": 0.1187,
      "step": 922
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.4405592083930969,
      "learning_rate": 9.293871131196364e-06,
      "loss": 0.1241,
      "step": 923
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.3944067060947418,
      "learning_rate": 9.275073652471173e-06,
      "loss": 0.1118,
      "step": 924
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.3871515691280365,
      "learning_rate": 9.256278748420307e-06,
      "loss": 0.114,
      "step": 925
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.42183035612106323,
      "learning_rate": 9.237486485796417e-06,
      "loss": 0.1261,
      "step": 926
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.46369391679763794,
      "learning_rate": 9.218696931342773e-06,
      "loss": 0.1264,
      "step": 927
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.3826935589313507,
      "learning_rate": 9.199910151793026e-06,
      "loss": 0.1246,
      "step": 928
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.3553304076194763,
      "learning_rate": 9.18112621387098e-06,
      "loss": 0.1096,
      "step": 929
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.391335666179657,
      "learning_rate": 9.162345184290334e-06,
      "loss": 0.1087,
      "step": 930
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.3531568944454193,
      "learning_rate": 9.143567129754465e-06,
      "loss": 0.1235,
      "step": 931
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.4913727343082428,
      "learning_rate": 9.124792116956186e-06,
      "loss": 0.1313,
      "step": 932
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.38267356157302856,
      "learning_rate": 9.106020212577501e-06,
      "loss": 0.1306,
      "step": 933
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.4557020664215088,
      "learning_rate": 9.087251483289376e-06,
      "loss": 0.1257,
      "step": 934
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.37805190682411194,
      "learning_rate": 9.068485995751504e-06,
      "loss": 0.1081,
      "step": 935
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.42168304324150085,
      "learning_rate": 9.049723816612056e-06,
      "loss": 0.1261,
      "step": 936
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.3669287860393524,
      "learning_rate": 9.030965012507461e-06,
      "loss": 0.1217,
      "step": 937
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.3708566129207611,
      "learning_rate": 9.012209650062157e-06,
      "loss": 0.1189,
      "step": 938
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.4192459285259247,
      "learning_rate": 8.993457795888358e-06,
      "loss": 0.1245,
      "step": 939
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.3058549165725708,
      "learning_rate": 8.974709516585825e-06,
      "loss": 0.1259,
      "step": 940
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.3823601305484772,
      "learning_rate": 8.95596487874161e-06,
      "loss": 0.126,
      "step": 941
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.3646482229232788,
      "learning_rate": 8.937223948929845e-06,
      "loss": 0.1163,
      "step": 942
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.37197476625442505,
      "learning_rate": 8.91848679371148e-06,
      "loss": 0.1243,
      "step": 943
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.3767468333244324,
      "learning_rate": 8.899753479634067e-06,
      "loss": 0.1278,
      "step": 944
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.31196027994155884,
      "learning_rate": 8.88102407323152e-06,
      "loss": 0.1162,
      "step": 945
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.3199111819267273,
      "learning_rate": 8.862298641023858e-06,
      "loss": 0.1179,
      "step": 946
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.3048057556152344,
      "learning_rate": 8.843577249517004e-06,
      "loss": 0.1257,
      "step": 947
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.42693763971328735,
      "learning_rate": 8.82485996520251e-06,
      "loss": 0.1328,
      "step": 948
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.43489277362823486,
      "learning_rate": 8.806146854557364e-06,
      "loss": 0.1184,
      "step": 949
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.3449189066886902,
      "learning_rate": 8.787437984043714e-06,
      "loss": 0.1137,
      "step": 950
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.3741619288921356,
      "learning_rate": 8.76873342010865e-06,
      "loss": 0.1337,
      "step": 951
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.3315809369087219,
      "learning_rate": 8.750033229183976e-06,
      "loss": 0.1062,
      "step": 952
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.340667188167572,
      "learning_rate": 8.73133747768595e-06,
      "loss": 0.1168,
      "step": 953
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.3843550980091095,
      "learning_rate": 8.71264623201508e-06,
      "loss": 0.1206,
      "step": 954
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.3667396008968353,
      "learning_rate": 8.693959558555853e-06,
      "loss": 0.1348,
      "step": 955
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.3556472957134247,
      "learning_rate": 8.675277523676532e-06,
      "loss": 0.1267,
      "step": 956
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.27359670400619507,
      "learning_rate": 8.656600193728903e-06,
      "loss": 0.1222,
      "step": 957
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.30881252884864807,
      "learning_rate": 8.637927635048029e-06,
      "loss": 0.1182,
      "step": 958
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.34924551844596863,
      "learning_rate": 8.619259913952047e-06,
      "loss": 0.1288,
      "step": 959
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.3397780656814575,
      "learning_rate": 8.600597096741893e-06,
      "loss": 0.1214,
      "step": 960
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.3601126968860626,
      "learning_rate": 8.581939249701105e-06,
      "loss": 0.1116,
      "step": 961
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.4026654362678528,
      "learning_rate": 8.563286439095549e-06,
      "loss": 0.115,
      "step": 962
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.34074801206588745,
      "learning_rate": 8.544638731173218e-06,
      "loss": 0.1192,
      "step": 963
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.37266552448272705,
      "learning_rate": 8.525996192163985e-06,
      "loss": 0.1212,
      "step": 964
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.41644564270973206,
      "learning_rate": 8.507358888279346e-06,
      "loss": 0.1215,
      "step": 965
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.414080947637558,
      "learning_rate": 8.488726885712226e-06,
      "loss": 0.1345,
      "step": 966
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.3494729995727539,
      "learning_rate": 8.4701002506367e-06,
      "loss": 0.1232,
      "step": 967
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.3789871633052826,
      "learning_rate": 8.451479049207801e-06,
      "loss": 0.1085,
      "step": 968
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.31436386704444885,
      "learning_rate": 8.432863347561251e-06,
      "loss": 0.127,
      "step": 969
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.3709452748298645,
      "learning_rate": 8.414253211813234e-06,
      "loss": 0.123,
      "step": 970
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.3670414388179779,
      "learning_rate": 8.395648708060187e-06,
      "loss": 0.1274,
      "step": 971
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.34422793984413147,
      "learning_rate": 8.377049902378514e-06,
      "loss": 0.1219,
      "step": 972
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.2934967875480652,
      "learning_rate": 8.358456860824413e-06,
      "loss": 0.1242,
      "step": 973
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.41389068961143494,
      "learning_rate": 8.339869649433582e-06,
      "loss": 0.1112,
      "step": 974
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.3385218381881714,
      "learning_rate": 8.321288334221034e-06,
      "loss": 0.1285,
      "step": 975
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.3838481605052948,
      "learning_rate": 8.302712981180832e-06,
      "loss": 0.1292,
      "step": 976
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.3790236711502075,
      "learning_rate": 8.284143656285858e-06,
      "loss": 0.1249,
      "step": 977
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.3628286123275757,
      "learning_rate": 8.265580425487593e-06,
      "loss": 0.114,
      "step": 978
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.3760746121406555,
      "learning_rate": 8.24702335471587e-06,
      "loss": 0.1244,
      "step": 979
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.397598534822464,
      "learning_rate": 8.228472509878646e-06,
      "loss": 0.125,
      "step": 980
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.3559723496437073,
      "learning_rate": 8.209927956861766e-06,
      "loss": 0.1201,
      "step": 981
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.32869282364845276,
      "learning_rate": 8.191389761528723e-06,
      "loss": 0.1162,
      "step": 982
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.33538904786109924,
      "learning_rate": 8.17285798972044e-06,
      "loss": 0.1179,
      "step": 983
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.4225422143936157,
      "learning_rate": 8.154332707255011e-06,
      "loss": 0.1227,
      "step": 984
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.3333398699760437,
      "learning_rate": 8.135813979927502e-06,
      "loss": 0.1103,
      "step": 985
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.3429222106933594,
      "learning_rate": 8.117301873509675e-06,
      "loss": 0.1185,
      "step": 986
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.3243877589702606,
      "learning_rate": 8.098796453749797e-06,
      "loss": 0.1096,
      "step": 987
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.29434630274772644,
      "learning_rate": 8.08029778637238e-06,
      "loss": 0.1099,
      "step": 988
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.48877862095832825,
      "learning_rate": 8.061805937077943e-06,
      "loss": 0.1253,
      "step": 989
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.39397919178009033,
      "learning_rate": 8.043320971542809e-06,
      "loss": 0.1163,
      "step": 990
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.29214316606521606,
      "learning_rate": 8.024842955418832e-06,
      "loss": 0.1142,
      "step": 991
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.31439605355262756,
      "learning_rate": 8.006371954333201e-06,
      "loss": 0.1143,
      "step": 992
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.3399764895439148,
      "learning_rate": 7.987908033888186e-06,
      "loss": 0.1235,
      "step": 993
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.4030015170574188,
      "learning_rate": 7.9694512596609e-06,
      "loss": 0.1223,
      "step": 994
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.39697757363319397,
      "learning_rate": 7.951001697203085e-06,
      "loss": 0.1314,
      "step": 995
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.3972482979297638,
      "learning_rate": 7.932559412040865e-06,
      "loss": 0.1198,
      "step": 996
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.49958536028862,
      "learning_rate": 7.914124469674524e-06,
      "loss": 0.1206,
      "step": 997
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.34017959237098694,
      "learning_rate": 7.895696935578255e-06,
      "loss": 0.1109,
      "step": 998
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.4304424226284027,
      "learning_rate": 7.877276875199948e-06,
      "loss": 0.1125,
      "step": 999
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.427969753742218,
      "learning_rate": 7.858864353960954e-06,
      "loss": 0.1183,
      "step": 1000
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.4931412637233734,
      "learning_rate": 7.840459437255833e-06,
      "loss": 0.1268,
      "step": 1001
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.44037240743637085,
      "learning_rate": 7.822062190452153e-06,
      "loss": 0.1202,
      "step": 1002
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.43938812613487244,
      "learning_rate": 7.803672678890225e-06,
      "loss": 0.1146,
      "step": 1003
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.41637635231018066,
      "learning_rate": 7.7852909678829e-06,
      "loss": 0.1157,
      "step": 1004
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.48005211353302,
      "learning_rate": 7.766917122715325e-06,
      "loss": 0.1246,
      "step": 1005
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.3882123529911041,
      "learning_rate": 7.748551208644693e-06,
      "loss": 0.1203,
      "step": 1006
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.37888777256011963,
      "learning_rate": 7.73019329090005e-06,
      "loss": 0.1107,
      "step": 1007
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.36659398674964905,
      "learning_rate": 7.711843434682029e-06,
      "loss": 0.1259,
      "step": 1008
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.31240513920783997,
      "learning_rate": 7.693501705162638e-06,
      "loss": 0.1038,
      "step": 1009
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.36880412697792053,
      "learning_rate": 7.675168167485014e-06,
      "loss": 0.1166,
      "step": 1010
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.36817675828933716,
      "learning_rate": 7.65684288676321e-06,
      "loss": 0.1096,
      "step": 1011
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.3085809051990509,
      "learning_rate": 7.638525928081943e-06,
      "loss": 0.1105,
      "step": 1012
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.3951782286167145,
      "learning_rate": 7.62021735649638e-06,
      "loss": 0.1248,
      "step": 1013
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.4194014370441437,
      "learning_rate": 7.601917237031902e-06,
      "loss": 0.1252,
      "step": 1014
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.3867023289203644,
      "learning_rate": 7.583625634683861e-06,
      "loss": 0.1217,
      "step": 1015
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.4165385663509369,
      "learning_rate": 7.56534261441737e-06,
      "loss": 0.1262,
      "step": 1016
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.4824996590614319,
      "learning_rate": 7.547068241167056e-06,
      "loss": 0.1205,
      "step": 1017
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.4766876995563507,
      "learning_rate": 7.528802579836835e-06,
      "loss": 0.1212,
      "step": 1018
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.4633573889732361,
      "learning_rate": 7.510545695299683e-06,
      "loss": 0.1179,
      "step": 1019
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.32987627387046814,
      "learning_rate": 7.492297652397405e-06,
      "loss": 0.1132,
      "step": 1020
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.4370860755443573,
      "learning_rate": 7.474058515940401e-06,
      "loss": 0.1304,
      "step": 1021
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.4461018443107605,
      "learning_rate": 7.455828350707437e-06,
      "loss": 0.1155,
      "step": 1022
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.41251853108406067,
      "learning_rate": 7.4376072214454245e-06,
      "loss": 0.1159,
      "step": 1023
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.4294576346874237,
      "learning_rate": 7.419395192869174e-06,
      "loss": 0.1296,
      "step": 1024
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.3884996473789215,
      "learning_rate": 7.4011923296611795e-06,
      "loss": 0.1198,
      "step": 1025
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.34289804100990295,
      "learning_rate": 7.3829986964713796e-06,
      "loss": 0.1256,
      "step": 1026
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.43898892402648926,
      "learning_rate": 7.364814357916931e-06,
      "loss": 0.1239,
      "step": 1027
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.35503876209259033,
      "learning_rate": 7.346639378581979e-06,
      "loss": 0.123,
      "step": 1028
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.4248340427875519,
      "learning_rate": 7.328473823017434e-06,
      "loss": 0.1232,
      "step": 1029
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.39135444164276123,
      "learning_rate": 7.310317755740728e-06,
      "loss": 0.1111,
      "step": 1030
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.4446059465408325,
      "learning_rate": 7.292171241235603e-06,
      "loss": 0.1243,
      "step": 1031
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.3582511842250824,
      "learning_rate": 7.274034343951863e-06,
      "loss": 0.1096,
      "step": 1032
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.34349268674850464,
      "learning_rate": 7.255907128305166e-06,
      "loss": 0.1235,
      "step": 1033
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.36046260595321655,
      "learning_rate": 7.237789658676775e-06,
      "loss": 0.1139,
      "step": 1034
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.34063178300857544,
      "learning_rate": 7.219681999413341e-06,
      "loss": 0.12,
      "step": 1035
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.3147428035736084,
      "learning_rate": 7.2015842148266825e-06,
      "loss": 0.1135,
      "step": 1036
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.42707544565200806,
      "learning_rate": 7.183496369193528e-06,
      "loss": 0.1206,
      "step": 1037
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.4595549404621124,
      "learning_rate": 7.1654185267553235e-06,
      "loss": 0.1331,
      "step": 1038
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.4516967236995697,
      "learning_rate": 7.1473507517179715e-06,
      "loss": 0.1241,
      "step": 1039
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.3994957506656647,
      "learning_rate": 7.129293108251638e-06,
      "loss": 0.1258,
      "step": 1040
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.43237099051475525,
      "learning_rate": 7.111245660490489e-06,
      "loss": 0.1108,
      "step": 1041
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.4440392851829529,
      "learning_rate": 7.093208472532484e-06,
      "loss": 0.1163,
      "step": 1042
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.4052538573741913,
      "learning_rate": 7.07518160843915e-06,
      "loss": 0.1097,
      "step": 1043
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.38639235496520996,
      "learning_rate": 7.0571651322353355e-06,
      "loss": 0.1122,
      "step": 1044
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.3875058889389038,
      "learning_rate": 7.039159107909006e-06,
      "loss": 0.1202,
      "step": 1045
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.5056825280189514,
      "learning_rate": 7.021163599410995e-06,
      "loss": 0.1245,
      "step": 1046
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.48759153485298157,
      "learning_rate": 7.003178670654798e-06,
      "loss": 0.1224,
      "step": 1047
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.41536685824394226,
      "learning_rate": 6.985204385516333e-06,
      "loss": 0.1217,
      "step": 1048
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.4594440162181854,
      "learning_rate": 6.967240807833707e-06,
      "loss": 0.1259,
      "step": 1049
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.5151720643043518,
      "learning_rate": 6.949288001407013e-06,
      "loss": 0.115,
      "step": 1050
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.46664664149284363,
      "learning_rate": 6.931346029998069e-06,
      "loss": 0.1211,
      "step": 1051
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.43906787037849426,
      "learning_rate": 6.913414957330233e-06,
      "loss": 0.1161,
      "step": 1052
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.401497483253479,
      "learning_rate": 6.895494847088134e-06,
      "loss": 0.1197,
      "step": 1053
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.2786385416984558,
      "learning_rate": 6.877585762917478e-06,
      "loss": 0.1142,
      "step": 1054
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.36774468421936035,
      "learning_rate": 6.859687768424816e-06,
      "loss": 0.1218,
      "step": 1055
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.40317660570144653,
      "learning_rate": 6.841800927177296e-06,
      "loss": 0.121,
      "step": 1056
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.37352508306503296,
      "learning_rate": 6.823925302702469e-06,
      "loss": 0.1179,
      "step": 1057
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.4499046504497528,
      "learning_rate": 6.806060958488034e-06,
      "loss": 0.1139,
      "step": 1058
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.34265848994255066,
      "learning_rate": 6.78820795798164e-06,
      "loss": 0.1122,
      "step": 1059
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.3937309980392456,
      "learning_rate": 6.770366364590644e-06,
      "loss": 0.114,
      "step": 1060
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.35141611099243164,
      "learning_rate": 6.752536241681878e-06,
      "loss": 0.1107,
      "step": 1061
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.4013659954071045,
      "learning_rate": 6.7347176525814536e-06,
      "loss": 0.1177,
      "step": 1062
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.40220212936401367,
      "learning_rate": 6.7169106605745e-06,
      "loss": 0.1202,
      "step": 1063
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.36436206102371216,
      "learning_rate": 6.699115328904975e-06,
      "loss": 0.1223,
      "step": 1064
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.42185303568840027,
      "learning_rate": 6.681331720775403e-06,
      "loss": 0.1174,
      "step": 1065
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.3931892216205597,
      "learning_rate": 6.663559899346692e-06,
      "loss": 0.1223,
      "step": 1066
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.37592729926109314,
      "learning_rate": 6.645799927737877e-06,
      "loss": 0.1057,
      "step": 1067
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.4675831198692322,
      "learning_rate": 6.628051869025902e-06,
      "loss": 0.1236,
      "step": 1068
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.38862234354019165,
      "learning_rate": 6.610315786245416e-06,
      "loss": 0.1111,
      "step": 1069
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.3934907019138336,
      "learning_rate": 6.592591742388513e-06,
      "loss": 0.1182,
      "step": 1070
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.38182657957077026,
      "learning_rate": 6.574879800404548e-06,
      "loss": 0.11,
      "step": 1071
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.4821826219558716,
      "learning_rate": 6.557180023199888e-06,
      "loss": 0.1305,
      "step": 1072
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.398257315158844,
      "learning_rate": 6.539492473637688e-06,
      "loss": 0.1159,
      "step": 1073
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.4881611168384552,
      "learning_rate": 6.521817214537689e-06,
      "loss": 0.1242,
      "step": 1074
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.4210967421531677,
      "learning_rate": 6.5041543086759655e-06,
      "loss": 0.1155,
      "step": 1075
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.45111799240112305,
      "learning_rate": 6.486503818784731e-06,
      "loss": 0.1167,
      "step": 1076
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.39253681898117065,
      "learning_rate": 6.4688658075520895e-06,
      "loss": 0.1177,
      "step": 1077
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.5062733292579651,
      "learning_rate": 6.451240337621837e-06,
      "loss": 0.127,
      "step": 1078
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.47293880581855774,
      "learning_rate": 6.4336274715932245e-06,
      "loss": 0.1123,
      "step": 1079
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.4424663782119751,
      "learning_rate": 6.416027272020731e-06,
      "loss": 0.1099,
      "step": 1080
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.42107245326042175,
      "learning_rate": 6.398439801413858e-06,
      "loss": 0.1125,
      "step": 1081
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.3941892087459564,
      "learning_rate": 6.380865122236892e-06,
      "loss": 0.1153,
      "step": 1082
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.39981457591056824,
      "learning_rate": 6.363303296908692e-06,
      "loss": 0.1222,
      "step": 1083
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.4310297966003418,
      "learning_rate": 6.345754387802467e-06,
      "loss": 0.1165,
      "step": 1084
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.4251634478569031,
      "learning_rate": 6.328218457245547e-06,
      "loss": 0.1176,
      "step": 1085
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.428182452917099,
      "learning_rate": 6.310695567519174e-06,
      "loss": 0.1187,
      "step": 1086
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.410593718290329,
      "learning_rate": 6.293185780858261e-06,
      "loss": 0.1162,
      "step": 1087
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.42972663044929504,
      "learning_rate": 6.275689159451202e-06,
      "loss": 0.1244,
      "step": 1088
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.35343119502067566,
      "learning_rate": 6.258205765439612e-06,
      "loss": 0.1085,
      "step": 1089
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.3476179838180542,
      "learning_rate": 6.240735660918144e-06,
      "loss": 0.105,
      "step": 1090
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.4021570682525635,
      "learning_rate": 6.223278907934251e-06,
      "loss": 0.1198,
      "step": 1091
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.4669889509677887,
      "learning_rate": 6.205835568487952e-06,
      "loss": 0.1192,
      "step": 1092
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.4258028566837311,
      "learning_rate": 6.1884057045316405e-06,
      "loss": 0.1129,
      "step": 1093
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.4519641697406769,
      "learning_rate": 6.170989377969838e-06,
      "loss": 0.1212,
      "step": 1094
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.37600088119506836,
      "learning_rate": 6.1535866506589945e-06,
      "loss": 0.1228,
      "step": 1095
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.4501347839832306,
      "learning_rate": 6.136197584407265e-06,
      "loss": 0.114,
      "step": 1096
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.4128763675689697,
      "learning_rate": 6.118822240974269e-06,
      "loss": 0.1255,
      "step": 1097
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.46924999356269836,
      "learning_rate": 6.101460682070905e-06,
      "loss": 0.1157,
      "step": 1098
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.45293447375297546,
      "learning_rate": 6.0841129693590995e-06,
      "loss": 0.124,
      "step": 1099
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.4573341906070709,
      "learning_rate": 6.0667791644516146e-06,
      "loss": 0.1216,
      "step": 1100
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.549071729183197,
      "learning_rate": 6.049459328911806e-06,
      "loss": 0.1096,
      "step": 1101
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.37328192591667175,
      "learning_rate": 6.032153524253424e-06,
      "loss": 0.1055,
      "step": 1102
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.49592307209968567,
      "learning_rate": 6.014861811940381e-06,
      "loss": 0.1129,
      "step": 1103
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.4542485177516937,
      "learning_rate": 5.997584253386541e-06,
      "loss": 0.1138,
      "step": 1104
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.3650111258029938,
      "learning_rate": 5.980320909955501e-06,
      "loss": 0.1158,
      "step": 1105
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.5523119568824768,
      "learning_rate": 5.963071842960364e-06,
      "loss": 0.1171,
      "step": 1106
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.40748628973960876,
      "learning_rate": 5.945837113663537e-06,
      "loss": 0.1181,
      "step": 1107
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.6059275269508362,
      "learning_rate": 5.928616783276498e-06,
      "loss": 0.1277,
      "step": 1108
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.47895291447639465,
      "learning_rate": 5.911410912959588e-06,
      "loss": 0.1161,
      "step": 1109
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.5759312510490417,
      "learning_rate": 5.894219563821791e-06,
      "loss": 0.123,
      "step": 1110
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.42922481894493103,
      "learning_rate": 5.877042796920514e-06,
      "loss": 0.1118,
      "step": 1111
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.37468940019607544,
      "learning_rate": 5.859880673261384e-06,
      "loss": 0.1096,
      "step": 1112
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.3587656617164612,
      "learning_rate": 5.842733253798005e-06,
      "loss": 0.1099,
      "step": 1113
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.3889211118221283,
      "learning_rate": 5.82560059943177e-06,
      "loss": 0.1142,
      "step": 1114
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.4921272099018097,
      "learning_rate": 5.808482771011625e-06,
      "loss": 0.1154,
      "step": 1115
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.46166226267814636,
      "learning_rate": 5.791379829333856e-06,
      "loss": 0.1308,
      "step": 1116
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.5351108908653259,
      "learning_rate": 5.774291835141892e-06,
      "loss": 0.1186,
      "step": 1117
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.3920232355594635,
      "learning_rate": 5.757218849126053e-06,
      "loss": 0.1118,
      "step": 1118
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.4083298444747925,
      "learning_rate": 5.7401609319233705e-06,
      "loss": 0.1191,
      "step": 1119
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.39702552556991577,
      "learning_rate": 5.723118144117359e-06,
      "loss": 0.1169,
      "step": 1120
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.44372567534446716,
      "learning_rate": 5.706090546237782e-06,
      "loss": 0.1171,
      "step": 1121
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.39780110120773315,
      "learning_rate": 5.689078198760476e-06,
      "loss": 0.1071,
      "step": 1122
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.34773513674736023,
      "learning_rate": 5.672081162107091e-06,
      "loss": 0.1072,
      "step": 1123
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.43540409207344055,
      "learning_rate": 5.655099496644918e-06,
      "loss": 0.1077,
      "step": 1124
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.4789227843284607,
      "learning_rate": 5.638133262686641e-06,
      "loss": 0.1148,
      "step": 1125
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.4218980669975281,
      "learning_rate": 5.621182520490147e-06,
      "loss": 0.1158,
      "step": 1126
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.5167276263237,
      "learning_rate": 5.604247330258299e-06,
      "loss": 0.1108,
      "step": 1127
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.5218992829322815,
      "learning_rate": 5.587327752138718e-06,
      "loss": 0.1115,
      "step": 1128
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.43909481167793274,
      "learning_rate": 5.570423846223589e-06,
      "loss": 0.1138,
      "step": 1129
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.4519633650779724,
      "learning_rate": 5.553535672549421e-06,
      "loss": 0.1052,
      "step": 1130
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.4213640093803406,
      "learning_rate": 5.536663291096864e-06,
      "loss": 0.1156,
      "step": 1131
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.3994320333003998,
      "learning_rate": 5.519806761790462e-06,
      "loss": 0.1178,
      "step": 1132
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.5105064511299133,
      "learning_rate": 5.502966144498471e-06,
      "loss": 0.1096,
      "step": 1133
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.46372565627098083,
      "learning_rate": 5.4861414990326325e-06,
      "loss": 0.1202,
      "step": 1134
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.4745757579803467,
      "learning_rate": 5.469332885147951e-06,
      "loss": 0.1183,
      "step": 1135
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.48656222224235535,
      "learning_rate": 5.452540362542509e-06,
      "loss": 0.1117,
      "step": 1136
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.4049619138240814,
      "learning_rate": 5.435763990857222e-06,
      "loss": 0.1246,
      "step": 1137
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.4251416027545929,
      "learning_rate": 5.419003829675654e-06,
      "loss": 0.117,
      "step": 1138
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.4387785792350769,
      "learning_rate": 5.402259938523798e-06,
      "loss": 0.1135,
      "step": 1139
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.9002676010131836,
      "learning_rate": 5.385532376869845e-06,
      "loss": 0.124,
      "step": 1140
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.499994695186615,
      "learning_rate": 5.368821204124013e-06,
      "loss": 0.1248,
      "step": 1141
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.43259677290916443,
      "learning_rate": 5.3521264796382884e-06,
      "loss": 0.1154,
      "step": 1142
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.5427858233451843,
      "learning_rate": 5.335448262706262e-06,
      "loss": 0.1164,
      "step": 1143
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.48766160011291504,
      "learning_rate": 5.318786612562877e-06,
      "loss": 0.1236,
      "step": 1144
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.4106052815914154,
      "learning_rate": 5.302141588384249e-06,
      "loss": 0.1155,
      "step": 1145
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.4163687229156494,
      "learning_rate": 5.285513249287443e-06,
      "loss": 0.1193,
      "step": 1146
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.40940341353416443,
      "learning_rate": 5.268901654330256e-06,
      "loss": 0.1078,
      "step": 1147
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.36008474230766296,
      "learning_rate": 5.25230686251103e-06,
      "loss": 0.1037,
      "step": 1148
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.392586886882782,
      "learning_rate": 5.235728932768414e-06,
      "loss": 0.1197,
      "step": 1149
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.4692182242870331,
      "learning_rate": 5.2191679239811764e-06,
      "loss": 0.1144,
      "step": 1150
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.40251123905181885,
      "learning_rate": 5.202623894967996e-06,
      "loss": 0.1162,
      "step": 1151
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.5050315260887146,
      "learning_rate": 5.186096904487225e-06,
      "loss": 0.1144,
      "step": 1152
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.42679744958877563,
      "learning_rate": 5.169587011236725e-06,
      "loss": 0.1071,
      "step": 1153
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.5091879963874817,
      "learning_rate": 5.153094273853614e-06,
      "loss": 0.1132,
      "step": 1154
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.49064475297927856,
      "learning_rate": 5.136618750914096e-06,
      "loss": 0.117,
      "step": 1155
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.4914780855178833,
      "learning_rate": 5.120160500933216e-06,
      "loss": 0.1228,
      "step": 1156
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.42524904012680054,
      "learning_rate": 5.10371958236469e-06,
      "loss": 0.1183,
      "step": 1157
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.3876344561576843,
      "learning_rate": 5.087296053600675e-06,
      "loss": 0.1102,
      "step": 1158
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.31190451979637146,
      "learning_rate": 5.070889972971554e-06,
      "loss": 0.1111,
      "step": 1159
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.45548781752586365,
      "learning_rate": 5.054501398745759e-06,
      "loss": 0.1152,
      "step": 1160
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.3743782937526703,
      "learning_rate": 5.038130389129525e-06,
      "loss": 0.1089,
      "step": 1161
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.38007980585098267,
      "learning_rate": 5.021777002266722e-06,
      "loss": 0.1107,
      "step": 1162
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.40839794278144836,
      "learning_rate": 5.005441296238625e-06,
      "loss": 0.1196,
      "step": 1163
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.4734240472316742,
      "learning_rate": 4.989123329063704e-06,
      "loss": 0.1105,
      "step": 1164
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.5531820058822632,
      "learning_rate": 4.972823158697443e-06,
      "loss": 0.1096,
      "step": 1165
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.38676002621650696,
      "learning_rate": 4.956540843032101e-06,
      "loss": 0.1193,
      "step": 1166
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.46789759397506714,
      "learning_rate": 4.940276439896538e-06,
      "loss": 0.1092,
      "step": 1167
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.572537899017334,
      "learning_rate": 4.924030007055985e-06,
      "loss": 0.118,
      "step": 1168
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.5674297213554382,
      "learning_rate": 4.907801602211857e-06,
      "loss": 0.1186,
      "step": 1169
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.5080124139785767,
      "learning_rate": 4.891591283001539e-06,
      "loss": 0.112,
      "step": 1170
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.531479001045227,
      "learning_rate": 4.875399106998172e-06,
      "loss": 0.1244,
      "step": 1171
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.49633359909057617,
      "learning_rate": 4.859225131710476e-06,
      "loss": 0.1171,
      "step": 1172
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.6131958365440369,
      "learning_rate": 4.843069414582511e-06,
      "loss": 0.1299,
      "step": 1173
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.4790622889995575,
      "learning_rate": 4.8269320129935036e-06,
      "loss": 0.1148,
      "step": 1174
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.5745164155960083,
      "learning_rate": 4.810812984257629e-06,
      "loss": 0.1135,
      "step": 1175
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.4696940779685974,
      "learning_rate": 4.7947123856237995e-06,
      "loss": 0.1198,
      "step": 1176
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.5676562190055847,
      "learning_rate": 4.7786302742754845e-06,
      "loss": 0.1185,
      "step": 1177
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.4301071763038635,
      "learning_rate": 4.7625667073304745e-06,
      "loss": 0.1222,
      "step": 1178
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.6371943950653076,
      "learning_rate": 4.746521741840719e-06,
      "loss": 0.117,
      "step": 1179
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.49200645089149475,
      "learning_rate": 4.730495434792082e-06,
      "loss": 0.1139,
      "step": 1180
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.49400895833969116,
      "learning_rate": 4.714487843104172e-06,
      "loss": 0.1225,
      "step": 1181
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.5919182896614075,
      "learning_rate": 4.698499023630128e-06,
      "loss": 0.1131,
      "step": 1182
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.4329252243041992,
      "learning_rate": 4.682529033156403e-06,
      "loss": 0.1153,
      "step": 1183
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.47083157300949097,
      "learning_rate": 4.666577928402597e-06,
      "loss": 0.1144,
      "step": 1184
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.38236671686172485,
      "learning_rate": 4.6506457660212124e-06,
      "loss": 0.1013,
      "step": 1185
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.38545459508895874,
      "learning_rate": 4.634732602597494e-06,
      "loss": 0.1079,
      "step": 1186
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.4455778896808624,
      "learning_rate": 4.6188384946492014e-06,
      "loss": 0.1167,
      "step": 1187
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.39531686902046204,
      "learning_rate": 4.602963498626412e-06,
      "loss": 0.1049,
      "step": 1188
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.47524404525756836,
      "learning_rate": 4.587107670911336e-06,
      "loss": 0.113,
      "step": 1189
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.4057534337043762,
      "learning_rate": 4.57127106781809e-06,
      "loss": 0.1046,
      "step": 1190
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.5006169676780701,
      "learning_rate": 4.555453745592525e-06,
      "loss": 0.1139,
      "step": 1191
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.4547278583049774,
      "learning_rate": 4.539655760412004e-06,
      "loss": 0.1104,
      "step": 1192
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.4942764341831207,
      "learning_rate": 4.523877168385216e-06,
      "loss": 0.1137,
      "step": 1193
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.530991792678833,
      "learning_rate": 4.508118025551979e-06,
      "loss": 0.1078,
      "step": 1194
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.5576902627944946,
      "learning_rate": 4.4923783878830175e-06,
      "loss": 0.1116,
      "step": 1195
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.41624119877815247,
      "learning_rate": 4.476658311279801e-06,
      "loss": 0.1119,
      "step": 1196
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.49815818667411804,
      "learning_rate": 4.460957851574307e-06,
      "loss": 0.1114,
      "step": 1197
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.4246378242969513,
      "learning_rate": 4.445277064528852e-06,
      "loss": 0.1103,
      "step": 1198
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.5607213973999023,
      "learning_rate": 4.429616005835886e-06,
      "loss": 0.1201,
      "step": 1199
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.4459719657897949,
      "learning_rate": 4.413974731117775e-06,
      "loss": 0.111,
      "step": 1200
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.517173707485199,
      "learning_rate": 4.398353295926636e-06,
      "loss": 0.1153,
      "step": 1201
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.5197553634643555,
      "learning_rate": 4.382751755744113e-06,
      "loss": 0.112,
      "step": 1202
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.4612776041030884,
      "learning_rate": 4.367170165981196e-06,
      "loss": 0.1018,
      "step": 1203
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.5459276437759399,
      "learning_rate": 4.351608581978012e-06,
      "loss": 0.1161,
      "step": 1204
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.45408543944358826,
      "learning_rate": 4.336067059003641e-06,
      "loss": 0.1136,
      "step": 1205
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.43536514043807983,
      "learning_rate": 4.320545652255916e-06,
      "loss": 0.1081,
      "step": 1206
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.4147088825702667,
      "learning_rate": 4.305044416861213e-06,
      "loss": 0.1102,
      "step": 1207
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.3780951201915741,
      "learning_rate": 4.289563407874279e-06,
      "loss": 0.1111,
      "step": 1208
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.5387012362480164,
      "learning_rate": 4.274102680278013e-06,
      "loss": 0.1149,
      "step": 1209
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.4820585548877716,
      "learning_rate": 4.258662288983292e-06,
      "loss": 0.115,
      "step": 1210
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.45025435090065,
      "learning_rate": 4.243242288828763e-06,
      "loss": 0.109,
      "step": 1211
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.48669514060020447,
      "learning_rate": 4.2278427345806426e-06,
      "loss": 0.1139,
      "step": 1212
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.491899311542511,
      "learning_rate": 4.212463680932547e-06,
      "loss": 0.1139,
      "step": 1213
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.45830756425857544,
      "learning_rate": 4.197105182505265e-06,
      "loss": 0.0999,
      "step": 1214
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.42406928539276123,
      "learning_rate": 4.1817672938465935e-06,
      "loss": 0.1087,
      "step": 1215
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.4988643527030945,
      "learning_rate": 4.166450069431122e-06,
      "loss": 0.111,
      "step": 1216
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.4968511760234833,
      "learning_rate": 4.151153563660053e-06,
      "loss": 0.1117,
      "step": 1217
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.5506454110145569,
      "learning_rate": 4.135877830861009e-06,
      "loss": 0.1193,
      "step": 1218
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.4745372533798218,
      "learning_rate": 4.120622925287819e-06,
      "loss": 0.1132,
      "step": 1219
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.5285553932189941,
      "learning_rate": 4.105388901120361e-06,
      "loss": 0.1185,
      "step": 1220
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.46258386969566345,
      "learning_rate": 4.090175812464328e-06,
      "loss": 0.1053,
      "step": 1221
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.5369489192962646,
      "learning_rate": 4.074983713351077e-06,
      "loss": 0.1123,
      "step": 1222
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.44612935185432434,
      "learning_rate": 4.059812657737412e-06,
      "loss": 0.1236,
      "step": 1223
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.46147000789642334,
      "learning_rate": 4.04466269950539e-06,
      "loss": 0.1069,
      "step": 1224
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.45773857831954956,
      "learning_rate": 4.0295338924621495e-06,
      "loss": 0.1048,
      "step": 1225
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.4511007070541382,
      "learning_rate": 4.014426290339697e-06,
      "loss": 0.113,
      "step": 1226
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.567457914352417,
      "learning_rate": 3.999339946794737e-06,
      "loss": 0.1144,
      "step": 1227
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.5611966252326965,
      "learning_rate": 3.984274915408461e-06,
      "loss": 0.1139,
      "step": 1228
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.4224526584148407,
      "learning_rate": 3.969231249686377e-06,
      "loss": 0.1084,
      "step": 1229
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.42501670122146606,
      "learning_rate": 3.954209003058108e-06,
      "loss": 0.0979,
      "step": 1230
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.4009574353694916,
      "learning_rate": 3.939208228877195e-06,
      "loss": 0.1173,
      "step": 1231
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.45079198479652405,
      "learning_rate": 3.924228980420931e-06,
      "loss": 0.115,
      "step": 1232
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.45230597257614136,
      "learning_rate": 3.909271310890142e-06,
      "loss": 0.1169,
      "step": 1233
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.5184730887413025,
      "learning_rate": 3.894335273409031e-06,
      "loss": 0.1113,
      "step": 1234
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.4547876715660095,
      "learning_rate": 3.879420921024955e-06,
      "loss": 0.1089,
      "step": 1235
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.6866465210914612,
      "learning_rate": 3.864528306708262e-06,
      "loss": 0.1027,
      "step": 1236
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.49597102403640747,
      "learning_rate": 3.849657483352099e-06,
      "loss": 0.1167,
      "step": 1237
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.5036567449569702,
      "learning_rate": 3.8348085037722045e-06,
      "loss": 0.1165,
      "step": 1238
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.4601985514163971,
      "learning_rate": 3.819981420706756e-06,
      "loss": 0.1038,
      "step": 1239
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.4065699875354767,
      "learning_rate": 3.80517628681614e-06,
      "loss": 0.1066,
      "step": 1240
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.48107531666755676,
      "learning_rate": 3.7903931546828086e-06,
      "loss": 0.1179,
      "step": 1241
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.5847235321998596,
      "learning_rate": 3.775632076811054e-06,
      "loss": 0.1185,
      "step": 1242
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.5021312236785889,
      "learning_rate": 3.7608931056268537e-06,
      "loss": 0.123,
      "step": 1243
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.5660654306411743,
      "learning_rate": 3.7461762934776657e-06,
      "loss": 0.1126,
      "step": 1244
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.5968395471572876,
      "learning_rate": 3.7314816926322396e-06,
      "loss": 0.125,
      "step": 1245
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.5112922191619873,
      "learning_rate": 3.7168093552804516e-06,
      "loss": 0.1249,
      "step": 1246
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.40580469369888306,
      "learning_rate": 3.702159333533093e-06,
      "loss": 0.1173,
      "step": 1247
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.43287238478660583,
      "learning_rate": 3.687531679421712e-06,
      "loss": 0.1051,
      "step": 1248
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.47445225715637207,
      "learning_rate": 3.6729264448983994e-06,
      "loss": 0.117,
      "step": 1249
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.49166321754455566,
      "learning_rate": 3.6583436818356354e-06,
      "loss": 0.1155,
      "step": 1250
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.4318731725215912,
      "learning_rate": 3.6437834420260764e-06,
      "loss": 0.1108,
      "step": 1251
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.40729159116744995,
      "learning_rate": 3.629245777182392e-06,
      "loss": 0.1103,
      "step": 1252
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.5511937737464905,
      "learning_rate": 3.614730738937078e-06,
      "loss": 0.113,
      "step": 1253
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.5207592248916626,
      "learning_rate": 3.6002383788422548e-06,
      "loss": 0.1139,
      "step": 1254
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.4108555018901825,
      "learning_rate": 3.5857687483695124e-06,
      "loss": 0.1182,
      "step": 1255
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.46642324328422546,
      "learning_rate": 3.5713218989097033e-06,
      "loss": 0.1106,
      "step": 1256
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.48571357131004333,
      "learning_rate": 3.5568978817727783e-06,
      "loss": 0.111,
      "step": 1257
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.47130531072616577,
      "learning_rate": 3.542496748187586e-06,
      "loss": 0.0982,
      "step": 1258
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.38559606671333313,
      "learning_rate": 3.5281185493017157e-06,
      "loss": 0.1087,
      "step": 1259
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.5606286525726318,
      "learning_rate": 3.513763336181283e-06,
      "loss": 0.1111,
      "step": 1260
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.6842437982559204,
      "learning_rate": 3.499431159810781e-06,
      "loss": 0.1061,
      "step": 1261
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.4387393891811371,
      "learning_rate": 3.4851220710928835e-06,
      "loss": 0.1084,
      "step": 1262
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.4500303268432617,
      "learning_rate": 3.4708361208482523e-06,
      "loss": 0.1192,
      "step": 1263
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.44867753982543945,
      "learning_rate": 3.456573359815386e-06,
      "loss": 0.1079,
      "step": 1264
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.47414252161979675,
      "learning_rate": 3.4423338386504114e-06,
      "loss": 0.113,
      "step": 1265
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.5146887898445129,
      "learning_rate": 3.4281176079269262e-06,
      "loss": 0.1109,
      "step": 1266
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.49450966715812683,
      "learning_rate": 3.413924718135797e-06,
      "loss": 0.1258,
      "step": 1267
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.5120757818222046,
      "learning_rate": 3.3997552196850048e-06,
      "loss": 0.1172,
      "step": 1268
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.4796096682548523,
      "learning_rate": 3.3856091628994412e-06,
      "loss": 0.1128,
      "step": 1269
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.5177645087242126,
      "learning_rate": 3.371486598020749e-06,
      "loss": 0.1222,
      "step": 1270
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.42458876967430115,
      "learning_rate": 3.357387575207137e-06,
      "loss": 0.1079,
      "step": 1271
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.5077516436576843,
      "learning_rate": 3.343312144533194e-06,
      "loss": 0.1082,
      "step": 1272
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.5280843377113342,
      "learning_rate": 3.3292603559897274e-06,
      "loss": 0.123,
      "step": 1273
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.46623095870018005,
      "learning_rate": 3.3152322594835663e-06,
      "loss": 0.1081,
      "step": 1274
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.504090428352356,
      "learning_rate": 3.301227904837405e-06,
      "loss": 0.1088,
      "step": 1275
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.48985084891319275,
      "learning_rate": 3.2872473417896034e-06,
      "loss": 0.1085,
      "step": 1276
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.5125816464424133,
      "learning_rate": 3.273290619994033e-06,
      "loss": 0.1046,
      "step": 1277
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.3953268826007843,
      "learning_rate": 3.259357789019889e-06,
      "loss": 0.1125,
      "step": 1278
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.42861899733543396,
      "learning_rate": 3.2454488983515043e-06,
      "loss": 0.1113,
      "step": 1279
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.4431779384613037,
      "learning_rate": 3.2315639973881984e-06,
      "loss": 0.1162,
      "step": 1280
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.45942556858062744,
      "learning_rate": 3.2177031354440767e-06,
      "loss": 0.1179,
      "step": 1281
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.4620354175567627,
      "learning_rate": 3.203866361747877e-06,
      "loss": 0.1166,
      "step": 1282
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.46729138493537903,
      "learning_rate": 3.1900537254427745e-06,
      "loss": 0.1154,
      "step": 1283
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.45829087495803833,
      "learning_rate": 3.1762652755862234e-06,
      "loss": 0.1125,
      "step": 1284
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.47549712657928467,
      "learning_rate": 3.1625010611497776e-06,
      "loss": 0.1121,
      "step": 1285
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.4972947835922241,
      "learning_rate": 3.1487611310189083e-06,
      "loss": 0.1137,
      "step": 1286
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.42110076546669006,
      "learning_rate": 3.1350455339928488e-06,
      "loss": 0.1017,
      "step": 1287
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.4348899722099304,
      "learning_rate": 3.1213543187843953e-06,
      "loss": 0.1105,
      "step": 1288
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.5183877944946289,
      "learning_rate": 3.107687534019762e-06,
      "loss": 0.1169,
      "step": 1289
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.5382614731788635,
      "learning_rate": 3.0940452282383937e-06,
      "loss": 0.1225,
      "step": 1290
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.4297749996185303,
      "learning_rate": 3.0804274498927834e-06,
      "loss": 0.1145,
      "step": 1291
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.44973450899124146,
      "learning_rate": 3.0668342473483248e-06,
      "loss": 0.0999,
      "step": 1292
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.4379996955394745,
      "learning_rate": 3.0532656688831174e-06,
      "loss": 0.1095,
      "step": 1293
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.4500684142112732,
      "learning_rate": 3.039721762687816e-06,
      "loss": 0.103,
      "step": 1294
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.4971635937690735,
      "learning_rate": 3.0262025768654334e-06,
      "loss": 0.1127,
      "step": 1295
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.47018080949783325,
      "learning_rate": 3.012708159431199e-06,
      "loss": 0.1195,
      "step": 1296
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.5291146039962769,
      "learning_rate": 2.9992385583123683e-06,
      "loss": 0.111,
      "step": 1297
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.42619314789772034,
      "learning_rate": 2.9857938213480532e-06,
      "loss": 0.105,
      "step": 1298
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.4381212890148163,
      "learning_rate": 2.9723739962890684e-06,
      "loss": 0.1112,
      "step": 1299
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.482317715883255,
      "learning_rate": 2.9589791307977376e-06,
      "loss": 0.1173,
      "step": 1300
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.4490676522254944,
      "learning_rate": 2.945609272447746e-06,
      "loss": 0.1074,
      "step": 1301
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.4494069814682007,
      "learning_rate": 2.9322644687239656e-06,
      "loss": 0.1142,
      "step": 1302
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.4392588436603546,
      "learning_rate": 2.9189447670222693e-06,
      "loss": 0.1159,
      "step": 1303
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.5065415501594543,
      "learning_rate": 2.9056502146493935e-06,
      "loss": 0.1127,
      "step": 1304
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.49169430136680603,
      "learning_rate": 2.89238085882274e-06,
      "loss": 0.1052,
      "step": 1305
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.49049684405326843,
      "learning_rate": 2.8791367466702315e-06,
      "loss": 0.1145,
      "step": 1306
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.5220836997032166,
      "learning_rate": 2.865917925230125e-06,
      "loss": 0.1142,
      "step": 1307
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.46083298325538635,
      "learning_rate": 2.8527244414508615e-06,
      "loss": 0.1055,
      "step": 1308
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.5117243528366089,
      "learning_rate": 2.8395563421908934e-06,
      "loss": 0.1225,
      "step": 1309
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.4779595136642456,
      "learning_rate": 2.8264136742185066e-06,
      "loss": 0.11,
      "step": 1310
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.5158355236053467,
      "learning_rate": 2.813296484211675e-06,
      "loss": 0.1175,
      "step": 1311
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.6859225034713745,
      "learning_rate": 2.8002048187578745e-06,
      "loss": 0.1125,
      "step": 1312
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.538798987865448,
      "learning_rate": 2.787138724353934e-06,
      "loss": 0.1201,
      "step": 1313
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.4592318534851074,
      "learning_rate": 2.7740982474058664e-06,
      "loss": 0.1179,
      "step": 1314
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.5979160070419312,
      "learning_rate": 2.7610834342286874e-06,
      "loss": 0.1181,
      "step": 1315
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.4947759211063385,
      "learning_rate": 2.748094331046279e-06,
      "loss": 0.1103,
      "step": 1316
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.4502725899219513,
      "learning_rate": 2.7351309839911998e-06,
      "loss": 0.1104,
      "step": 1317
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.6392968893051147,
      "learning_rate": 2.722193439104541e-06,
      "loss": 0.1248,
      "step": 1318
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.5100122690200806,
      "learning_rate": 2.709281742335743e-06,
      "loss": 0.1075,
      "step": 1319
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.4282607138156891,
      "learning_rate": 2.696395939542454e-06,
      "loss": 0.1038,
      "step": 1320
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.48218637704849243,
      "learning_rate": 2.683536076490354e-06,
      "loss": 0.1096,
      "step": 1321
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.605644941329956,
      "learning_rate": 2.6707021988529857e-06,
      "loss": 0.1134,
      "step": 1322
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.6176994442939758,
      "learning_rate": 2.657894352211614e-06,
      "loss": 0.1215,
      "step": 1323
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.5815721154212952,
      "learning_rate": 2.6451125820550393e-06,
      "loss": 0.1112,
      "step": 1324
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.4522569477558136,
      "learning_rate": 2.632356933779454e-06,
      "loss": 0.1027,
      "step": 1325
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.5119885802268982,
      "learning_rate": 2.6196274526882803e-06,
      "loss": 0.1273,
      "step": 1326
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.4518607258796692,
      "learning_rate": 2.6069241839919913e-06,
      "loss": 0.1195,
      "step": 1327
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.45767924189567566,
      "learning_rate": 2.5942471728079742e-06,
      "loss": 0.1028,
      "step": 1328
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.5046805739402771,
      "learning_rate": 2.581596464160351e-06,
      "loss": 0.1161,
      "step": 1329
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.4890325963497162,
      "learning_rate": 2.5689721029798364e-06,
      "loss": 0.1163,
      "step": 1330
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.6210603713989258,
      "learning_rate": 2.5563741341035555e-06,
      "loss": 0.1255,
      "step": 1331
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.4806549847126007,
      "learning_rate": 2.5438026022749096e-06,
      "loss": 0.1114,
      "step": 1332
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.6295298933982849,
      "learning_rate": 2.5312575521434014e-06,
      "loss": 0.1104,
      "step": 1333
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.4935235381126404,
      "learning_rate": 2.518739028264474e-06,
      "loss": 0.1105,
      "step": 1334
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.48289936780929565,
      "learning_rate": 2.506247075099372e-06,
      "loss": 0.1104,
      "step": 1335
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.5653237104415894,
      "learning_rate": 2.493781737014953e-06,
      "loss": 0.1125,
      "step": 1336
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.4918563663959503,
      "learning_rate": 2.4813430582835676e-06,
      "loss": 0.1173,
      "step": 1337
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.48344311118125916,
      "learning_rate": 2.468931083082862e-06,
      "loss": 0.1229,
      "step": 1338
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.4403289556503296,
      "learning_rate": 2.4565458554956546e-06,
      "loss": 0.1059,
      "step": 1339
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.4490101933479309,
      "learning_rate": 2.444187419509767e-06,
      "loss": 0.1117,
      "step": 1340
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.39293691515922546,
      "learning_rate": 2.4318558190178553e-06,
      "loss": 0.106,
      "step": 1341
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.49275434017181396,
      "learning_rate": 2.4195510978172776e-06,
      "loss": 0.1098,
      "step": 1342
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.49393993616104126,
      "learning_rate": 2.4072732996099167e-06,
      "loss": 0.1101,
      "step": 1343
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.4499976634979248,
      "learning_rate": 2.395022468002042e-06,
      "loss": 0.1178,
      "step": 1344
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.5163973569869995,
      "learning_rate": 2.3827986465041462e-06,
      "loss": 0.1144,
      "step": 1345
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.4758533835411072,
      "learning_rate": 2.3706018785307873e-06,
      "loss": 0.1132,
      "step": 1346
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.4398626685142517,
      "learning_rate": 2.3584322074004473e-06,
      "loss": 0.1136,
      "step": 1347
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.4442160427570343,
      "learning_rate": 2.3462896763353583e-06,
      "loss": 0.1072,
      "step": 1348
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.45861542224884033,
      "learning_rate": 2.334174328461374e-06,
      "loss": 0.1159,
      "step": 1349
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.3842131793498993,
      "learning_rate": 2.322086206807792e-06,
      "loss": 0.1073,
      "step": 1350
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.5142524838447571,
      "learning_rate": 2.310025354307218e-06,
      "loss": 0.1112,
      "step": 1351
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.4489837884902954,
      "learning_rate": 2.2979918137954115e-06,
      "loss": 0.1095,
      "step": 1352
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.41778650879859924,
      "learning_rate": 2.285985628011119e-06,
      "loss": 0.1087,
      "step": 1353
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.6004676222801208,
      "learning_rate": 2.2740068395959437e-06,
      "loss": 0.112,
      "step": 1354
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.49727168679237366,
      "learning_rate": 2.262055491094174e-06,
      "loss": 0.1015,
      "step": 1355
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.409483939409256,
      "learning_rate": 2.2501316249526484e-06,
      "loss": 0.1146,
      "step": 1356
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.5054729580879211,
      "learning_rate": 2.238235283520599e-06,
      "loss": 0.1163,
      "step": 1357
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.5615144968032837,
      "learning_rate": 2.2263665090494912e-06,
      "loss": 0.1224,
      "step": 1358
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.5416620969772339,
      "learning_rate": 2.214525343692895e-06,
      "loss": 0.1192,
      "step": 1359
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.515235960483551,
      "learning_rate": 2.20271182950631e-06,
      "loss": 0.1152,
      "step": 1360
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.5251460075378418,
      "learning_rate": 2.190926008447042e-06,
      "loss": 0.1085,
      "step": 1361
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.46637800335884094,
      "learning_rate": 2.1791679223740283e-06,
      "loss": 0.1158,
      "step": 1362
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.501814603805542,
      "learning_rate": 2.1674376130477093e-06,
      "loss": 0.1141,
      "step": 1363
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.49686750769615173,
      "learning_rate": 2.155735122129875e-06,
      "loss": 0.1173,
      "step": 1364
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.4420091211795807,
      "learning_rate": 2.144060491183504e-06,
      "loss": 0.1044,
      "step": 1365
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.4773690104484558,
      "learning_rate": 2.1324137616726394e-06,
      "loss": 0.113,
      "step": 1366
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.5471271276473999,
      "learning_rate": 2.120794974962215e-06,
      "loss": 0.1145,
      "step": 1367
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.4789467453956604,
      "learning_rate": 2.1092041723179312e-06,
      "loss": 0.1074,
      "step": 1368
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.5318825840950012,
      "learning_rate": 2.0976413949061025e-06,
      "loss": 0.1201,
      "step": 1369
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.4502650201320648,
      "learning_rate": 2.086106683793494e-06,
      "loss": 0.1057,
      "step": 1370
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.4683941602706909,
      "learning_rate": 2.0746000799472043e-06,
      "loss": 0.108,
      "step": 1371
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.44257885217666626,
      "learning_rate": 2.0631216242344933e-06,
      "loss": 0.1126,
      "step": 1372
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.44339892268180847,
      "learning_rate": 2.051671357422661e-06,
      "loss": 0.104,
      "step": 1373
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.5203191041946411,
      "learning_rate": 2.0402493201788774e-06,
      "loss": 0.1102,
      "step": 1374
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.5002195835113525,
      "learning_rate": 2.0288555530700604e-06,
      "loss": 0.1138,
      "step": 1375
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.7675966024398804,
      "learning_rate": 2.0174900965627253e-06,
      "loss": 0.1103,
      "step": 1376
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.6492874622344971,
      "learning_rate": 2.0061529910228273e-06,
      "loss": 0.112,
      "step": 1377
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.44493889808654785,
      "learning_rate": 1.9948442767156395e-06,
      "loss": 0.1096,
      "step": 1378
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.48615363240242004,
      "learning_rate": 1.983563993805593e-06,
      "loss": 0.1007,
      "step": 1379
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.5373040437698364,
      "learning_rate": 1.972312182356143e-06,
      "loss": 0.1193,
      "step": 1380
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.4781138598918915,
      "learning_rate": 1.9610888823296316e-06,
      "loss": 0.109,
      "step": 1381
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.4730207026004791,
      "learning_rate": 1.9498941335871247e-06,
      "loss": 0.112,
      "step": 1382
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.8251639008522034,
      "learning_rate": 1.938727975888297e-06,
      "loss": 0.118,
      "step": 1383
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.47587111592292786,
      "learning_rate": 1.927590448891269e-06,
      "loss": 0.1127,
      "step": 1384
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.49293115735054016,
      "learning_rate": 1.9164815921524836e-06,
      "loss": 0.1128,
      "step": 1385
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.5393558740615845,
      "learning_rate": 1.9054014451265502e-06,
      "loss": 0.1088,
      "step": 1386
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.5550158023834229,
      "learning_rate": 1.8943500471661158e-06,
      "loss": 0.1095,
      "step": 1387
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.5397169589996338,
      "learning_rate": 1.8833274375217248e-06,
      "loss": 0.1066,
      "step": 1388
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.45565322041511536,
      "learning_rate": 1.8723336553416637e-06,
      "loss": 0.1064,
      "step": 1389
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.5962527394294739,
      "learning_rate": 1.8613687396718482e-06,
      "loss": 0.1191,
      "step": 1390
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.46096280217170715,
      "learning_rate": 1.8504327294556613e-06,
      "loss": 0.104,
      "step": 1391
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.6331014037132263,
      "learning_rate": 1.8395256635338287e-06,
      "loss": 0.117,
      "step": 1392
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.5282611846923828,
      "learning_rate": 1.8286475806442772e-06,
      "loss": 0.1108,
      "step": 1393
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.4994890093803406,
      "learning_rate": 1.8177985194219916e-06,
      "loss": 0.1161,
      "step": 1394
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.5798438191413879,
      "learning_rate": 1.8069785183988896e-06,
      "loss": 0.1245,
      "step": 1395
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.627432644367218,
      "learning_rate": 1.7961876160036685e-06,
      "loss": 0.1169,
      "step": 1396
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.6922875046730042,
      "learning_rate": 1.7854258505616884e-06,
      "loss": 0.1223,
      "step": 1397
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.4853309988975525,
      "learning_rate": 1.774693260294814e-06,
      "loss": 0.1043,
      "step": 1398
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.5511927604675293,
      "learning_rate": 1.7639898833213e-06,
      "loss": 0.0909,
      "step": 1399
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.5001906752586365,
      "learning_rate": 1.7533157576556436e-06,
      "loss": 0.1123,
      "step": 1400
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.503832221031189,
      "learning_rate": 1.742670921208447e-06,
      "loss": 0.0994,
      "step": 1401
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.47644853591918945,
      "learning_rate": 1.7320554117862976e-06,
      "loss": 0.0935,
      "step": 1402
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.4300606846809387,
      "learning_rate": 1.721469267091611e-06,
      "loss": 0.1018,
      "step": 1403
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.5874450206756592,
      "learning_rate": 1.7109125247225211e-06,
      "loss": 0.101,
      "step": 1404
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.5487397313117981,
      "learning_rate": 1.7003852221727335e-06,
      "loss": 0.0963,
      "step": 1405
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.4992470145225525,
      "learning_rate": 1.6898873968313867e-06,
      "loss": 0.113,
      "step": 1406
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.4762040674686432,
      "learning_rate": 1.6794190859829373e-06,
      "loss": 0.1108,
      "step": 1407
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.43763861060142517,
      "learning_rate": 1.6689803268070081e-06,
      "loss": 0.1051,
      "step": 1408
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.5339246988296509,
      "learning_rate": 1.6585711563782736e-06,
      "loss": 0.1014,
      "step": 1409
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.7139047384262085,
      "learning_rate": 1.648191611666311e-06,
      "loss": 0.1193,
      "step": 1410
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.5060892105102539,
      "learning_rate": 1.637841729535483e-06,
      "loss": 0.1101,
      "step": 1411
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.5550698637962341,
      "learning_rate": 1.6275215467448057e-06,
      "loss": 0.1185,
      "step": 1412
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.49698498845100403,
      "learning_rate": 1.6172310999478026e-06,
      "loss": 0.1119,
      "step": 1413
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.4986688792705536,
      "learning_rate": 1.606970425692398e-06,
      "loss": 0.1051,
      "step": 1414
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.5143622756004333,
      "learning_rate": 1.5967395604207626e-06,
      "loss": 0.117,
      "step": 1415
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.5108754634857178,
      "learning_rate": 1.586538540469208e-06,
      "loss": 0.1164,
      "step": 1416
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.4960029721260071,
      "learning_rate": 1.576367402068043e-06,
      "loss": 0.105,
      "step": 1417
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.4742167890071869,
      "learning_rate": 1.5662261813414404e-06,
      "loss": 0.1089,
      "step": 1418
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.5414695739746094,
      "learning_rate": 1.5561149143073284e-06,
      "loss": 0.1155,
      "step": 1419
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.39626675844192505,
      "learning_rate": 1.5460336368772377e-06,
      "loss": 0.1023,
      "step": 1420
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.4288407266139984,
      "learning_rate": 1.5359823848561994e-06,
      "loss": 0.1014,
      "step": 1421
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.4526330530643463,
      "learning_rate": 1.5259611939425957e-06,
      "loss": 0.1154,
      "step": 1422
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.5292221903800964,
      "learning_rate": 1.5159700997280458e-06,
      "loss": 0.1226,
      "step": 1423
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.4773101806640625,
      "learning_rate": 1.5060091376972818e-06,
      "loss": 0.0996,
      "step": 1424
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.631423830986023,
      "learning_rate": 1.4960783432280067e-06,
      "loss": 0.117,
      "step": 1425
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.586578369140625,
      "learning_rate": 1.4861777515907904e-06,
      "loss": 0.1091,
      "step": 1426
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.4156636893749237,
      "learning_rate": 1.476307397948923e-06,
      "loss": 0.1013,
      "step": 1427
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.5883738398551941,
      "learning_rate": 1.46646731735831e-06,
      "loss": 0.1064,
      "step": 1428
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.5149186849594116,
      "learning_rate": 1.456657544767327e-06,
      "loss": 0.1136,
      "step": 1429
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.714019775390625,
      "learning_rate": 1.44687811501672e-06,
      "loss": 0.112,
      "step": 1430
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.4566328525543213,
      "learning_rate": 1.4371290628394551e-06,
      "loss": 0.1187,
      "step": 1431
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.6050567030906677,
      "learning_rate": 1.4274104228606179e-06,
      "loss": 0.1252,
      "step": 1432
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.4701688885688782,
      "learning_rate": 1.4177222295972781e-06,
      "loss": 0.1069,
      "step": 1433
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.49921727180480957,
      "learning_rate": 1.4080645174583652e-06,
      "loss": 0.1099,
      "step": 1434
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.5350949764251709,
      "learning_rate": 1.39843732074456e-06,
      "loss": 0.1074,
      "step": 1435
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.47706282138824463,
      "learning_rate": 1.388840673648154e-06,
      "loss": 0.1095,
      "step": 1436
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.6579143404960632,
      "learning_rate": 1.3792746102529442e-06,
      "loss": 0.1162,
      "step": 1437
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.5335856676101685,
      "learning_rate": 1.3697391645341019e-06,
      "loss": 0.1105,
      "step": 1438
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.6366046071052551,
      "learning_rate": 1.360234370358059e-06,
      "loss": 0.1102,
      "step": 1439
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.4725835621356964,
      "learning_rate": 1.350760261482379e-06,
      "loss": 0.1087,
      "step": 1440
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.5471887588500977,
      "learning_rate": 1.341316871555647e-06,
      "loss": 0.1068,
      "step": 1441
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.4280684292316437,
      "learning_rate": 1.3319042341173471e-06,
      "loss": 0.0997,
      "step": 1442
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.48780158162117004,
      "learning_rate": 1.3225223825977329e-06,
      "loss": 0.1085,
      "step": 1443
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.49698737263679504,
      "learning_rate": 1.3131713503177301e-06,
      "loss": 0.1032,
      "step": 1444
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.5654980540275574,
      "learning_rate": 1.3038511704887935e-06,
      "loss": 0.1016,
      "step": 1445
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.5297404527664185,
      "learning_rate": 1.294561876212811e-06,
      "loss": 0.1144,
      "step": 1446
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.6116273999214172,
      "learning_rate": 1.285303500481968e-06,
      "loss": 0.1202,
      "step": 1447
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.6900144815444946,
      "learning_rate": 1.2760760761786462e-06,
      "loss": 0.1179,
      "step": 1448
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.5525712370872498,
      "learning_rate": 1.2668796360752933e-06,
      "loss": 0.1202,
      "step": 1449
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.5056470632553101,
      "learning_rate": 1.2577142128343155e-06,
      "loss": 0.1117,
      "step": 1450
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.5175157785415649,
      "learning_rate": 1.2485798390079596e-06,
      "loss": 0.1046,
      "step": 1451
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.6875139474868774,
      "learning_rate": 1.2394765470381886e-06,
      "loss": 0.1152,
      "step": 1452
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.5659791231155396,
      "learning_rate": 1.230404369256587e-06,
      "loss": 0.1098,
      "step": 1453
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.5039743185043335,
      "learning_rate": 1.2213633378842183e-06,
      "loss": 0.1028,
      "step": 1454
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.4701502323150635,
      "learning_rate": 1.2123534850315388e-06,
      "loss": 0.1128,
      "step": 1455
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.5661687850952148,
      "learning_rate": 1.2033748426982584e-06,
      "loss": 0.1018,
      "step": 1456
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.601503849029541,
      "learning_rate": 1.1944274427732483e-06,
      "loss": 0.1095,
      "step": 1457
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.4639089107513428,
      "learning_rate": 1.1855113170344113e-06,
      "loss": 0.1206,
      "step": 1458
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.5744399428367615,
      "learning_rate": 1.1766264971485787e-06,
      "loss": 0.1079,
      "step": 1459
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.46826422214508057,
      "learning_rate": 1.1677730146713962e-06,
      "loss": 0.1029,
      "step": 1460
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.5220136642456055,
      "learning_rate": 1.1589509010472055e-06,
      "loss": 0.1137,
      "step": 1461
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.49739062786102295,
      "learning_rate": 1.1501601876089442e-06,
      "loss": 0.0999,
      "step": 1462
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.6703909039497375,
      "learning_rate": 1.1414009055780206e-06,
      "loss": 0.1075,
      "step": 1463
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.5475335717201233,
      "learning_rate": 1.1326730860642178e-06,
      "loss": 0.1036,
      "step": 1464
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.5322338342666626,
      "learning_rate": 1.123976760065566e-06,
      "loss": 0.1091,
      "step": 1465
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.4885544180870056,
      "learning_rate": 1.1153119584682514e-06,
      "loss": 0.1085,
      "step": 1466
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.6464633941650391,
      "learning_rate": 1.1066787120464938e-06,
      "loss": 0.1096,
      "step": 1467
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.5480430722236633,
      "learning_rate": 1.0980770514624361e-06,
      "loss": 0.1095,
      "step": 1468
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.478704571723938,
      "learning_rate": 1.0895070072660473e-06,
      "loss": 0.115,
      "step": 1469
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.550829291343689,
      "learning_rate": 1.080968609895e-06,
      "loss": 0.1114,
      "step": 1470
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.393227756023407,
      "learning_rate": 1.072461889674571e-06,
      "loss": 0.1138,
      "step": 1471
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.584935188293457,
      "learning_rate": 1.0639868768175354e-06,
      "loss": 0.1102,
      "step": 1472
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.4842588007450104,
      "learning_rate": 1.0555436014240471e-06,
      "loss": 0.1086,
      "step": 1473
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.5964325070381165,
      "learning_rate": 1.0471320934815488e-06,
      "loss": 0.1131,
      "step": 1474
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.6417595148086548,
      "learning_rate": 1.0387523828646495e-06,
      "loss": 0.1244,
      "step": 1475
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.5012993812561035,
      "learning_rate": 1.0304044993350314e-06,
      "loss": 0.1102,
      "step": 1476
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.5622046589851379,
      "learning_rate": 1.0220884725413337e-06,
      "loss": 0.1147,
      "step": 1477
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.5318993926048279,
      "learning_rate": 1.0138043320190548e-06,
      "loss": 0.1142,
      "step": 1478
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.4663567841053009,
      "learning_rate": 1.0055521071904473e-06,
      "loss": 0.111,
      "step": 1479
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.513012707233429,
      "learning_rate": 9.973318273644028e-07,
      "loss": 0.1056,
      "step": 1480
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.5155274271965027,
      "learning_rate": 9.891435217363666e-07,
      "loss": 0.1024,
      "step": 1481
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.6125432848930359,
      "learning_rate": 9.80987219388212e-07,
      "loss": 0.1181,
      "step": 1482
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.5376408100128174,
      "learning_rate": 9.728629492881557e-07,
      "loss": 0.1059,
      "step": 1483
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.5973014235496521,
      "learning_rate": 9.647707402906492e-07,
      "loss": 0.1142,
      "step": 1484
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.501593828201294,
      "learning_rate": 9.567106211362664e-07,
      "loss": 0.1024,
      "step": 1485
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.5086384415626526,
      "learning_rate": 9.486826204516175e-07,
      "loss": 0.1054,
      "step": 1486
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.6592398881912231,
      "learning_rate": 9.406867667492347e-07,
      "loss": 0.1138,
      "step": 1487
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.435183584690094,
      "learning_rate": 9.327230884274807e-07,
      "loss": 0.1099,
      "step": 1488
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.5066809058189392,
      "learning_rate": 9.247916137704349e-07,
      "loss": 0.1116,
      "step": 1489
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.507610023021698,
      "learning_rate": 9.168923709478106e-07,
      "loss": 0.1158,
      "step": 1490
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.47743767499923706,
      "learning_rate": 9.090253880148403e-07,
      "loss": 0.1041,
      "step": 1491
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.6116617321968079,
      "learning_rate": 9.011906929121794e-07,
      "loss": 0.1178,
      "step": 1492
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.4776424467563629,
      "learning_rate": 8.93388313465815e-07,
      "loss": 0.0943,
      "step": 1493
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.5685863494873047,
      "learning_rate": 8.856182773869526e-07,
      "loss": 0.121,
      "step": 1494
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.5239565372467041,
      "learning_rate": 8.778806122719308e-07,
      "loss": 0.1216,
      "step": 1495
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.3852762281894684,
      "learning_rate": 8.7017534560212e-07,
      "loss": 0.1107,
      "step": 1496
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.5759314298629761,
      "learning_rate": 8.625025047438185e-07,
      "loss": 0.1072,
      "step": 1497
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.49060699343681335,
      "learning_rate": 8.548621169481641e-07,
      "loss": 0.102,
      "step": 1498
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.5604158043861389,
      "learning_rate": 8.472542093510283e-07,
      "loss": 0.1098,
      "step": 1499
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.48748284578323364,
      "learning_rate": 8.396788089729302e-07,
      "loss": 0.1107,
      "step": 1500
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.5924981832504272,
      "learning_rate": 8.321359427189302e-07,
      "loss": 0.1213,
      "step": 1501
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.5464951992034912,
      "learning_rate": 8.246256373785421e-07,
      "loss": 0.1128,
      "step": 1502
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.6090044379234314,
      "learning_rate": 8.171479196256382e-07,
      "loss": 0.1205,
      "step": 1503
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.5219417810440063,
      "learning_rate": 8.097028160183429e-07,
      "loss": 0.1013,
      "step": 1504
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.541403591632843,
      "learning_rate": 8.022903529989567e-07,
      "loss": 0.1093,
      "step": 1505
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.5827970504760742,
      "learning_rate": 7.949105568938442e-07,
      "loss": 0.1192,
      "step": 1506
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.5167452692985535,
      "learning_rate": 7.875634539133548e-07,
      "loss": 0.1029,
      "step": 1507
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.5737378001213074,
      "learning_rate": 7.80249070151724e-07,
      "loss": 0.1168,
      "step": 1508
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.45662635564804077,
      "learning_rate": 7.729674315869751e-07,
      "loss": 0.1064,
      "step": 1509
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.5497353076934814,
      "learning_rate": 7.657185640808384e-07,
      "loss": 0.1073,
      "step": 1510
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.537789523601532,
      "learning_rate": 7.585024933786489e-07,
      "loss": 0.1007,
      "step": 1511
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.4542001187801361,
      "learning_rate": 7.513192451092633e-07,
      "loss": 0.1209,
      "step": 1512
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.539860725402832,
      "learning_rate": 7.441688447849604e-07,
      "loss": 0.1086,
      "step": 1513
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.5272457599639893,
      "learning_rate": 7.370513178013583e-07,
      "loss": 0.1038,
      "step": 1514
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.4377496838569641,
      "learning_rate": 7.299666894373236e-07,
      "loss": 0.1071,
      "step": 1515
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.43168261647224426,
      "learning_rate": 7.2291498485487e-07,
      "loss": 0.1044,
      "step": 1516
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.5548593997955322,
      "learning_rate": 7.158962290990878e-07,
      "loss": 0.1107,
      "step": 1517
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.5014510154724121,
      "learning_rate": 7.089104470980368e-07,
      "loss": 0.118,
      "step": 1518
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.4984900951385498,
      "learning_rate": 7.019576636626734e-07,
      "loss": 0.1012,
      "step": 1519
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.55745530128479,
      "learning_rate": 6.950379034867516e-07,
      "loss": 0.1108,
      "step": 1520
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.44775208830833435,
      "learning_rate": 6.881511911467365e-07,
      "loss": 0.111,
      "step": 1521
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.5011615753173828,
      "learning_rate": 6.81297551101725e-07,
      "loss": 0.1061,
      "step": 1522
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.5361987352371216,
      "learning_rate": 6.74477007693346e-07,
      "loss": 0.1142,
      "step": 1523
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.6265183091163635,
      "learning_rate": 6.6768958514569e-07,
      "loss": 0.1055,
      "step": 1524
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.4322400391101837,
      "learning_rate": 6.609353075652047e-07,
      "loss": 0.1003,
      "step": 1525
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.5303766131401062,
      "learning_rate": 6.542141989406259e-07,
      "loss": 0.1142,
      "step": 1526
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.45135101675987244,
      "learning_rate": 6.475262831428852e-07,
      "loss": 0.1026,
      "step": 1527
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.5382633805274963,
      "learning_rate": 6.408715839250191e-07,
      "loss": 0.1015,
      "step": 1528
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.5554699897766113,
      "learning_rate": 6.342501249221e-07,
      "loss": 0.1074,
      "step": 1529
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.5529068112373352,
      "learning_rate": 6.276619296511333e-07,
      "loss": 0.1179,
      "step": 1530
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.48695576190948486,
      "learning_rate": 6.211070215109927e-07,
      "loss": 0.1061,
      "step": 1531
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.5391503572463989,
      "learning_rate": 6.145854237823212e-07,
      "loss": 0.0969,
      "step": 1532
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.5016327500343323,
      "learning_rate": 6.08097159627461e-07,
      "loss": 0.1068,
      "step": 1533
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.5264042615890503,
      "learning_rate": 6.01642252090362e-07,
      "loss": 0.112,
      "step": 1534
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.5207239985466003,
      "learning_rate": 5.952207240965036e-07,
      "loss": 0.0979,
      "step": 1535
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.5783727765083313,
      "learning_rate": 5.888325984528154e-07,
      "loss": 0.1136,
      "step": 1536
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.7309806942939758,
      "learning_rate": 5.824778978475909e-07,
      "loss": 0.1073,
      "step": 1537
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.7122464776039124,
      "learning_rate": 5.76156644850413e-07,
      "loss": 0.1079,
      "step": 1538
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.6372246146202087,
      "learning_rate": 5.698688619120718e-07,
      "loss": 0.1174,
      "step": 1539
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.5549096465110779,
      "learning_rate": 5.636145713644781e-07,
      "loss": 0.1132,
      "step": 1540
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.5076556205749512,
      "learning_rate": 5.573937954205977e-07,
      "loss": 0.1104,
      "step": 1541
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.5092944502830505,
      "learning_rate": 5.512065561743563e-07,
      "loss": 0.1136,
      "step": 1542
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.5107739567756653,
      "learning_rate": 5.450528756005779e-07,
      "loss": 0.1192,
      "step": 1543
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.6881799697875977,
      "learning_rate": 5.389327755548901e-07,
      "loss": 0.1169,
      "step": 1544
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.49530842900276184,
      "learning_rate": 5.328462777736631e-07,
      "loss": 0.1028,
      "step": 1545
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.4630642831325531,
      "learning_rate": 5.267934038739186e-07,
      "loss": 0.1174,
      "step": 1546
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.48755156993865967,
      "learning_rate": 5.207741753532591e-07,
      "loss": 0.1147,
      "step": 1547
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.4687301516532898,
      "learning_rate": 5.147886135897929e-07,
      "loss": 0.1026,
      "step": 1548
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.48018935322761536,
      "learning_rate": 5.088367398420524e-07,
      "loss": 0.1085,
      "step": 1549
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.5262064933776855,
      "learning_rate": 5.029185752489285e-07,
      "loss": 0.1105,
      "step": 1550
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.5198432207107544,
      "learning_rate": 4.970341408295842e-07,
      "loss": 0.1143,
      "step": 1551
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.5327163338661194,
      "learning_rate": 4.911834574833874e-07,
      "loss": 0.124,
      "step": 1552
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.5597330927848816,
      "learning_rate": 4.853665459898349e-07,
      "loss": 0.1126,
      "step": 1553
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.5305644273757935,
      "learning_rate": 4.795834270084754e-07,
      "loss": 0.1066,
      "step": 1554
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.44982513785362244,
      "learning_rate": 4.738341210788444e-07,
      "loss": 0.1028,
      "step": 1555
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.5241904854774475,
      "learning_rate": 4.6811864862037816e-07,
      "loss": 0.0963,
      "step": 1556
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.4304783046245575,
      "learning_rate": 4.624370299323544e-07,
      "loss": 0.0978,
      "step": 1557
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.50018310546875,
      "learning_rate": 4.5678928519381607e-07,
      "loss": 0.1061,
      "step": 1558
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.5430971384048462,
      "learning_rate": 4.511754344634911e-07,
      "loss": 0.117,
      "step": 1559
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.5458731651306152,
      "learning_rate": 4.45595497679735e-07,
      "loss": 0.1136,
      "step": 1560
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.4804176688194275,
      "learning_rate": 4.400494946604483e-07,
      "loss": 0.1151,
      "step": 1561
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.5078117251396179,
      "learning_rate": 4.34537445103016e-07,
      "loss": 0.1161,
      "step": 1562
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.5756059288978577,
      "learning_rate": 4.2905936858422926e-07,
      "loss": 0.1049,
      "step": 1563
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.5182216167449951,
      "learning_rate": 4.2361528456021925e-07,
      "loss": 0.105,
      "step": 1564
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.49517199397087097,
      "learning_rate": 4.1820521236639266e-07,
      "loss": 0.1143,
      "step": 1565
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.5841544270515442,
      "learning_rate": 4.1282917121735155e-07,
      "loss": 0.1221,
      "step": 1566
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.5172370672225952,
      "learning_rate": 4.0748718020683697e-07,
      "loss": 0.1052,
      "step": 1567
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.567750871181488,
      "learning_rate": 4.021792583076556e-07,
      "loss": 0.1114,
      "step": 1568
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.6563100218772888,
      "learning_rate": 3.9690542437161193e-07,
      "loss": 0.1078,
      "step": 1569
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.572385847568512,
      "learning_rate": 3.9166569712944524e-07,
      "loss": 0.1027,
      "step": 1570
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.5810696482658386,
      "learning_rate": 3.864600951907538e-07,
      "loss": 0.1125,
      "step": 1571
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.5049096345901489,
      "learning_rate": 3.8128863704394394e-07,
      "loss": 0.1084,
      "step": 1572
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.5063614249229431,
      "learning_rate": 3.7615134105614795e-07,
      "loss": 0.1184,
      "step": 1573
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.5536503791809082,
      "learning_rate": 3.7104822547317065e-07,
      "loss": 0.1151,
      "step": 1574
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.4442182183265686,
      "learning_rate": 3.6597930841942184e-07,
      "loss": 0.1073,
      "step": 1575
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.5786116123199463,
      "learning_rate": 3.609446078978451e-07,
      "loss": 0.1152,
      "step": 1576
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.5246689319610596,
      "learning_rate": 3.559441417898668e-07,
      "loss": 0.1082,
      "step": 1577
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.6363775730133057,
      "learning_rate": 3.50977927855316e-07,
      "loss": 0.1206,
      "step": 1578
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.6120235323905945,
      "learning_rate": 3.460459837323804e-07,
      "loss": 0.1167,
      "step": 1579
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.524994969367981,
      "learning_rate": 3.41148326937526e-07,
      "loss": 0.1008,
      "step": 1580
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.5487150549888611,
      "learning_rate": 3.362849748654451e-07,
      "loss": 0.1094,
      "step": 1581
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.5051952600479126,
      "learning_rate": 3.3145594478899646e-07,
      "loss": 0.109,
      "step": 1582
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.4786909818649292,
      "learning_rate": 3.2666125385913295e-07,
      "loss": 0.1062,
      "step": 1583
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.564490556716919,
      "learning_rate": 3.2190091910485054e-07,
      "loss": 0.1037,
      "step": 1584
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.6258147358894348,
      "learning_rate": 3.171749574331251e-07,
      "loss": 0.1043,
      "step": 1585
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.6567071676254272,
      "learning_rate": 3.1248338562884896e-07,
      "loss": 0.1238,
      "step": 1586
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.47155579924583435,
      "learning_rate": 3.0782622035477993e-07,
      "loss": 0.1073,
      "step": 1587
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.47229817509651184,
      "learning_rate": 3.032034781514681e-07,
      "loss": 0.1015,
      "step": 1588
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.5014959573745728,
      "learning_rate": 2.986151754372146e-07,
      "loss": 0.1128,
      "step": 1589
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.43967363238334656,
      "learning_rate": 2.9406132850799517e-07,
      "loss": 0.1037,
      "step": 1590
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.6219881772994995,
      "learning_rate": 2.895419535374189e-07,
      "loss": 0.1116,
      "step": 1591
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.48437461256980896,
      "learning_rate": 2.85057066576655e-07,
      "loss": 0.1045,
      "step": 1592
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.48167815804481506,
      "learning_rate": 2.806066835543919e-07,
      "loss": 0.1053,
      "step": 1593
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.5501251816749573,
      "learning_rate": 2.7619082027676934e-07,
      "loss": 0.1136,
      "step": 1594
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.6490493416786194,
      "learning_rate": 2.718094924273218e-07,
      "loss": 0.1101,
      "step": 1595
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.5220945477485657,
      "learning_rate": 2.67462715566934e-07,
      "loss": 0.1079,
      "step": 1596
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.5725145936012268,
      "learning_rate": 2.6315050513377017e-07,
      "loss": 0.109,
      "step": 1597
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.4778439700603485,
      "learning_rate": 2.5887287644323243e-07,
      "loss": 0.0981,
      "step": 1598
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.4597511887550354,
      "learning_rate": 2.546298446879003e-07,
      "loss": 0.1045,
      "step": 1599
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.5171710848808289,
      "learning_rate": 2.5042142493747703e-07,
      "loss": 0.1052,
      "step": 1600
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.5707600712776184,
      "learning_rate": 2.462476321387364e-07,
      "loss": 0.1132,
      "step": 1601
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.5540006160736084,
      "learning_rate": 2.4210848111546836e-07,
      "loss": 0.1165,
      "step": 1602
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.4876388907432556,
      "learning_rate": 2.3800398656843248e-07,
      "loss": 0.1054,
      "step": 1603
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.5058432221412659,
      "learning_rate": 2.3393416307529558e-07,
      "loss": 0.0963,
      "step": 1604
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.4933337867259979,
      "learning_rate": 2.298990250905897e-07,
      "loss": 0.1123,
      "step": 1605
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.5581729412078857,
      "learning_rate": 2.2589858694565536e-07,
      "loss": 0.1106,
      "step": 1606
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.5052514672279358,
      "learning_rate": 2.2193286284859062e-07,
      "loss": 0.0937,
      "step": 1607
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.46765267848968506,
      "learning_rate": 2.1800186688420434e-07,
      "loss": 0.1029,
      "step": 1608
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.5132125020027161,
      "learning_rate": 2.1410561301396071e-07,
      "loss": 0.1063,
      "step": 1609
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.4925530254840851,
      "learning_rate": 2.10244115075936e-07,
      "loss": 0.0979,
      "step": 1610
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.5191598534584045,
      "learning_rate": 2.0641738678476188e-07,
      "loss": 0.1076,
      "step": 1611
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.6064251661300659,
      "learning_rate": 2.0262544173158428e-07,
      "loss": 0.0998,
      "step": 1612
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.5160400867462158,
      "learning_rate": 1.9886829338401248e-07,
      "loss": 0.1109,
      "step": 1613
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.4948529005050659,
      "learning_rate": 1.951459550860657e-07,
      "loss": 0.107,
      "step": 1614
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.5270385146141052,
      "learning_rate": 1.9145844005813762e-07,
      "loss": 0.104,
      "step": 1615
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.4943773150444031,
      "learning_rate": 1.8780576139693419e-07,
      "loss": 0.1073,
      "step": 1616
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.5596451163291931,
      "learning_rate": 1.841879320754425e-07,
      "loss": 0.1088,
      "step": 1617
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.45431920886039734,
      "learning_rate": 1.8060496494287205e-07,
      "loss": 0.0998,
      "step": 1618
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.5278708934783936,
      "learning_rate": 1.7705687272462026e-07,
      "loss": 0.1189,
      "step": 1619
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.656265377998352,
      "learning_rate": 1.7354366802221468e-07,
      "loss": 0.1048,
      "step": 1620
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.5217267274856567,
      "learning_rate": 1.7006536331327982e-07,
      "loss": 0.1169,
      "step": 1621
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.664513111114502,
      "learning_rate": 1.6662197095148936e-07,
      "loss": 0.1102,
      "step": 1622
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.5326521992683411,
      "learning_rate": 1.6321350316651497e-07,
      "loss": 0.1074,
      "step": 1623
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.6748685836791992,
      "learning_rate": 1.5983997206399647e-07,
      "loss": 0.0991,
      "step": 1624
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.5503498315811157,
      "learning_rate": 1.56501389625483e-07,
      "loss": 0.1064,
      "step": 1625
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.6422704458236694,
      "learning_rate": 1.5319776770840732e-07,
      "loss": 0.106,
      "step": 1626
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.4623504877090454,
      "learning_rate": 1.4992911804602938e-07,
      "loss": 0.1046,
      "step": 1627
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.5412484407424927,
      "learning_rate": 1.4669545224740178e-07,
      "loss": 0.1102,
      "step": 1628
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.5589399933815002,
      "learning_rate": 1.4349678179732763e-07,
      "loss": 0.1139,
      "step": 1629
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.4704682528972626,
      "learning_rate": 1.4033311805631943e-07,
      "loss": 0.1043,
      "step": 1630
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.5142149925231934,
      "learning_rate": 1.3720447226055923e-07,
      "loss": 0.1058,
      "step": 1631
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.5072668194770813,
      "learning_rate": 1.341108555218562e-07,
      "loss": 0.1143,
      "step": 1632
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.5687965750694275,
      "learning_rate": 1.310522788276103e-07,
      "loss": 0.1106,
      "step": 1633
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.539652943611145,
      "learning_rate": 1.28028753040772e-07,
      "loss": 0.1117,
      "step": 1634
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.5643596649169922,
      "learning_rate": 1.2504028889980368e-07,
      "loss": 0.1115,
      "step": 1635
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.5428621172904968,
      "learning_rate": 1.2208689701864062e-07,
      "loss": 0.1106,
      "step": 1636
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.5219332575798035,
      "learning_rate": 1.1916858788665664e-07,
      "loss": 0.1169,
      "step": 1637
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.5094490051269531,
      "learning_rate": 1.1628537186862077e-07,
      "loss": 0.102,
      "step": 1638
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.7151626348495483,
      "learning_rate": 1.134372592046662e-07,
      "loss": 0.1065,
      "step": 1639
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.5500611066818237,
      "learning_rate": 1.1062426001025361e-07,
      "loss": 0.1076,
      "step": 1640
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.5008072853088379,
      "learning_rate": 1.0784638427612904e-07,
      "loss": 0.1054,
      "step": 1641
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.5121034979820251,
      "learning_rate": 1.0510364186829713e-07,
      "loss": 0.1152,
      "step": 1642
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.702954888343811,
      "learning_rate": 1.0239604252797908e-07,
      "loss": 0.1146,
      "step": 1643
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.5763381123542786,
      "learning_rate": 9.972359587158142e-08,
      "loss": 0.0968,
      "step": 1644
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.5575568079948425,
      "learning_rate": 9.70863113906606e-08,
      "loss": 0.1101,
      "step": 1645
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.7215996384620667,
      "learning_rate": 9.448419845189294e-08,
      "loss": 0.1074,
      "step": 1646
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.5413906574249268,
      "learning_rate": 9.191726629703135e-08,
      "loss": 0.11,
      "step": 1647
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.4970971345901489,
      "learning_rate": 8.938552404288758e-08,
      "loss": 0.1081,
      "step": 1648
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.5927408933639526,
      "learning_rate": 8.68889806812867e-08,
      "loss": 0.1064,
      "step": 1649
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.46663331985473633,
      "learning_rate": 8.442764507904156e-08,
      "loss": 0.1002,
      "step": 1650
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.6554291844367981,
      "learning_rate": 8.200152597791944e-08,
      "loss": 0.1039,
      "step": 1651
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.5500927567481995,
      "learning_rate": 7.961063199461106e-08,
      "loss": 0.1024,
      "step": 1652
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.48627087473869324,
      "learning_rate": 7.72549716207005e-08,
      "loss": 0.1067,
      "step": 1653
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.527597963809967,
      "learning_rate": 7.493455322263754e-08,
      "loss": 0.1056,
      "step": 1654
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.5737511515617371,
      "learning_rate": 7.264938504170205e-08,
      "loss": 0.1181,
      "step": 1655
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.6555765867233276,
      "learning_rate": 7.03994751939785e-08,
      "loss": 0.1216,
      "step": 1656
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.4780556261539459,
      "learning_rate": 6.818483167032597e-08,
      "loss": 0.1037,
      "step": 1657
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.5788078904151917,
      "learning_rate": 6.600546233635485e-08,
      "loss": 0.116,
      "step": 1658
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.5393577814102173,
      "learning_rate": 6.386137493238798e-08,
      "loss": 0.1118,
      "step": 1659
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.5178937911987305,
      "learning_rate": 6.175257707344506e-08,
      "loss": 0.1039,
      "step": 1660
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.5325655341148376,
      "learning_rate": 5.96790762492061e-08,
      "loss": 0.0992,
      "step": 1661
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.5171847343444824,
      "learning_rate": 5.764087982399136e-08,
      "loss": 0.1085,
      "step": 1662
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.45278459787368774,
      "learning_rate": 5.5637995036733615e-08,
      "loss": 0.1098,
      "step": 1663
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.4596705436706543,
      "learning_rate": 5.367042900094821e-08,
      "loss": 0.1111,
      "step": 1664
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.47671201825141907,
      "learning_rate": 5.173818870471303e-08,
      "loss": 0.1011,
      "step": 1665
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.5089001059532166,
      "learning_rate": 4.9841281010641893e-08,
      "loss": 0.1126,
      "step": 1666
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.4517620801925659,
      "learning_rate": 4.797971265586121e-08,
      "loss": 0.1039,
      "step": 1667
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.5216341018676758,
      "learning_rate": 4.615349025198335e-08,
      "loss": 0.0926,
      "step": 1668
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.5143851637840271,
      "learning_rate": 4.4362620285083314e-08,
      "loss": 0.1117,
      "step": 1669
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.5324994325637817,
      "learning_rate": 4.260710911567989e-08,
      "loss": 0.1028,
      "step": 1670
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.5160598158836365,
      "learning_rate": 4.088696297871009e-08,
      "loss": 0.1072,
      "step": 1671
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.4682229161262512,
      "learning_rate": 3.9202187983505836e-08,
      "loss": 0.1087,
      "step": 1672
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.4574357271194458,
      "learning_rate": 3.755279011377399e-08,
      "loss": 0.1035,
      "step": 1673
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.6073874831199646,
      "learning_rate": 3.593877522757527e-08,
      "loss": 0.1064,
      "step": 1674
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.5004748702049255,
      "learning_rate": 3.4360149057302006e-08,
      "loss": 0.1088,
      "step": 1675
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.5684113502502441,
      "learning_rate": 3.2816917209660406e-08,
      "loss": 0.1109,
      "step": 1676
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.6874938607215881,
      "learning_rate": 3.1309085165646124e-08,
      "loss": 0.1114,
      "step": 1677
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.5145574808120728,
      "learning_rate": 2.983665828053206e-08,
      "loss": 0.1135,
      "step": 1678
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.5095924139022827,
      "learning_rate": 2.839964178383947e-08,
      "loss": 0.0986,
      "step": 1679
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.5667029023170471,
      "learning_rate": 2.6998040779329105e-08,
      "loss": 0.1092,
      "step": 1680
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.5607340931892395,
      "learning_rate": 2.563186024497566e-08,
      "loss": 0.1126,
      "step": 1681
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.5800597071647644,
      "learning_rate": 2.4301105032956683e-08,
      "loss": 0.1121,
      "step": 1682
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.641741931438446,
      "learning_rate": 2.3005779869628153e-08,
      "loss": 0.1158,
      "step": 1683
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.8395878076553345,
      "learning_rate": 2.174588935551447e-08,
      "loss": 0.108,
      "step": 1684
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.6455990076065063,
      "learning_rate": 2.0521437965288493e-08,
      "loss": 0.117,
      "step": 1685
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.5158807039260864,
      "learning_rate": 1.933243004775265e-08,
      "loss": 0.1108,
      "step": 1686
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.6876184344291687,
      "learning_rate": 1.817886982583228e-08,
      "loss": 0.1076,
      "step": 1687
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.511794924736023,
      "learning_rate": 1.706076139655344e-08,
      "loss": 0.1031,
      "step": 1688
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.5204125642776489,
      "learning_rate": 1.5978108731028454e-08,
      "loss": 0.1172,
      "step": 1689
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.6092833876609802,
      "learning_rate": 1.4930915674445934e-08,
      "loss": 0.1068,
      "step": 1690
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.4496975541114807,
      "learning_rate": 1.3919185946053016e-08,
      "loss": 0.1026,
      "step": 1691
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.5236508250236511,
      "learning_rate": 1.2942923139147579e-08,
      "loss": 0.1189,
      "step": 1692
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.530936062335968,
      "learning_rate": 1.2002130721057159e-08,
      "loss": 0.1109,
      "step": 1693
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.5151559710502625,
      "learning_rate": 1.1096812033135618e-08,
      "loss": 0.0996,
      "step": 1694
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.6460830569267273,
      "learning_rate": 1.0226970290744264e-08,
      "loss": 0.1108,
      "step": 1695
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.4686736762523651,
      "learning_rate": 9.39260858324298e-09,
      "loss": 0.1008,
      "step": 1696
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.4628830552101135,
      "learning_rate": 8.593729873982437e-09,
      "loss": 0.1019,
      "step": 1697
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.6973066329956055,
      "learning_rate": 7.830337000288568e-09,
      "loss": 0.1123,
      "step": 1698
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.5532433986663818,
      "learning_rate": 7.102432673453674e-09,
      "loss": 0.1138,
      "step": 1699
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.5252135992050171,
      "learning_rate": 6.410019478729767e-09,
      "loss": 0.1044,
      "step": 1700
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.5317535400390625,
      "learning_rate": 5.753099875316359e-09,
      "loss": 0.1093,
      "step": 1701
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.4698743224143982,
      "learning_rate": 5.131676196351576e-09,
      "loss": 0.1076,
      "step": 1702
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.5292864441871643,
      "learning_rate": 4.545750648906611e-09,
      "loss": 0.1048,
      "step": 1703
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.5920404195785522,
      "learning_rate": 3.995325313975728e-09,
      "loss": 0.1101,
      "step": 1704
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.5208712816238403,
      "learning_rate": 3.480402146468498e-09,
      "loss": 0.1057,
      "step": 1705
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.5738739967346191,
      "learning_rate": 3.0009829752053466e-09,
      "loss": 0.1105,
      "step": 1706
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.5885114669799805,
      "learning_rate": 2.557069502908682e-09,
      "loss": 0.1048,
      "step": 1707
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.6356754302978516,
      "learning_rate": 2.14866330619623e-09,
      "loss": 0.1065,
      "step": 1708
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.5434463024139404,
      "learning_rate": 1.7757658355788131e-09,
      "loss": 0.1055,
      "step": 1709
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.4475040137767792,
      "learning_rate": 1.4383784154525793e-09,
      "loss": 0.1076,
      "step": 1710
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.41130390763282776,
      "learning_rate": 1.1365022440945616e-09,
      "loss": 0.1046,
      "step": 1711
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.5430479645729065,
      "learning_rate": 8.701383936582375e-10,
      "loss": 0.1124,
      "step": 1712
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.5993987917900085,
      "learning_rate": 6.392878101724176e-10,
      "loss": 0.0981,
      "step": 1713
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.6595166325569153,
      "learning_rate": 4.4395131353347497e-10,
      "loss": 0.1114,
      "step": 1714
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.611425518989563,
      "learning_rate": 2.841295975053449e-10,
      "loss": 0.1033,
      "step": 1715
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.5242915153503418,
      "learning_rate": 1.598232297161939e-10,
      "loss": 0.1129,
      "step": 1716
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.6320515871047974,
      "learning_rate": 7.103265165842032e-11,
      "loss": 0.119,
      "step": 1717
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.6228506565093994,
      "learning_rate": 1.7758178681992476e-11,
      "loss": 0.1081,
      "step": 1718
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.4629266560077667,
      "learning_rate": 0.0,
      "loss": 0.1196,
      "step": 1719
    },
    {
      "epoch": 3.0,
      "step": 1719,
      "total_flos": 3.550611686681477e+18,
      "train_loss": 0.12722187895824077,
      "train_runtime": 8954.8436,
      "train_samples_per_second": 24.582,
      "train_steps_per_second": 0.192
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 1719,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 700,
  "total_flos": 3.550611686681477e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}