{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.8993055555555556,
  "eval_steps": 500,
  "global_step": 4662,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00019290123456790122,
      "grad_norm": 1.1879119873046875,
      "learning_rate": 0.0,
      "loss": 3.905716896057129,
      "step": 1
    },
    {
      "epoch": 0.00038580246913580245,
      "grad_norm": 1.281077265739441,
      "learning_rate": 6.25e-06,
      "loss": 3.675117254257202,
      "step": 2
    },
    {
      "epoch": 0.0005787037037037037,
      "grad_norm": 1.1838959455490112,
      "learning_rate": 1.25e-05,
      "loss": 4.168201923370361,
      "step": 3
    },
    {
      "epoch": 0.0007716049382716049,
      "grad_norm": 1.2642983198165894,
      "learning_rate": 1.875e-05,
      "loss": 3.9727299213409424,
      "step": 4
    },
    {
      "epoch": 0.0009645061728395061,
      "grad_norm": 1.2567980289459229,
      "learning_rate": 2.5e-05,
      "loss": 3.9534950256347656,
      "step": 5
    },
    {
      "epoch": 0.0011574074074074073,
      "grad_norm": 1.1836507320404053,
      "learning_rate": 3.125e-05,
      "loss": 4.162894248962402,
      "step": 6
    },
    {
      "epoch": 0.0013503086419753086,
      "grad_norm": 1.413661241531372,
      "learning_rate": 3.75e-05,
      "loss": 3.7735116481781006,
      "step": 7
    },
    {
      "epoch": 0.0015432098765432098,
      "grad_norm": 1.02397882938385,
      "learning_rate": 4.375e-05,
      "loss": 4.071147918701172,
      "step": 8
    },
    {
      "epoch": 0.001736111111111111,
      "grad_norm": 1.1740243434906006,
      "learning_rate": 5e-05,
      "loss": 3.753986120223999,
      "step": 9
    },
    {
      "epoch": 0.0019290123456790122,
      "grad_norm": 1.1778020858764648,
      "learning_rate": 5.6250000000000005e-05,
      "loss": 4.044284820556641,
      "step": 10
    },
    {
      "epoch": 0.0021219135802469135,
      "grad_norm": 1.2100940942764282,
      "learning_rate": 6.25e-05,
      "loss": 3.6397905349731445,
      "step": 11
    },
    {
      "epoch": 0.0023148148148148147,
      "grad_norm": 1.0850101709365845,
      "learning_rate": 6.875e-05,
      "loss": 4.054936408996582,
      "step": 12
    },
    {
      "epoch": 0.002507716049382716,
      "grad_norm": 1.2088090181350708,
      "learning_rate": 7.5e-05,
      "loss": 3.468900203704834,
      "step": 13
    },
    {
      "epoch": 0.002700617283950617,
      "grad_norm": 1.2167651653289795,
      "learning_rate": 8.125000000000001e-05,
      "loss": 3.671685218811035,
      "step": 14
    },
    {
      "epoch": 0.0028935185185185184,
      "grad_norm": 1.2959723472595215,
      "learning_rate": 8.75e-05,
      "loss": 3.9764201641082764,
      "step": 15
    },
    {
      "epoch": 0.0030864197530864196,
      "grad_norm": 1.152420997619629,
      "learning_rate": 9.375e-05,
      "loss": 3.8697614669799805,
      "step": 16
    },
    {
      "epoch": 0.003279320987654321,
      "grad_norm": 1.3030149936676025,
      "learning_rate": 0.0001,
      "loss": 4.164435386657715,
      "step": 17
    },
    {
      "epoch": 0.003472222222222222,
      "grad_norm": 1.9788873195648193,
      "learning_rate": 0.00010625,
      "loss": 3.968203544616699,
      "step": 18
    },
    {
      "epoch": 0.0036651234567901233,
      "grad_norm": 1.2348462343215942,
      "learning_rate": 0.00011250000000000001,
      "loss": 3.6672310829162598,
      "step": 19
    },
    {
      "epoch": 0.0038580246913580245,
      "grad_norm": 1.5430868864059448,
      "learning_rate": 0.00011875,
      "loss": 3.8145811557769775,
      "step": 20
    },
    {
      "epoch": 0.004050925925925926,
      "grad_norm": 1.0872156620025635,
      "learning_rate": 0.000125,
      "loss": 3.7776124477386475,
      "step": 21
    },
    {
      "epoch": 0.004243827160493827,
      "grad_norm": 1.7292742729187012,
      "learning_rate": 0.00013125000000000002,
      "loss": 3.4879887104034424,
      "step": 22
    },
    {
      "epoch": 0.004436728395061728,
      "grad_norm": 1.0845900774002075,
      "learning_rate": 0.0001375,
      "loss": 3.695199966430664,
      "step": 23
    },
    {
      "epoch": 0.004629629629629629,
      "grad_norm": 1.2406675815582275,
      "learning_rate": 0.00014375,
      "loss": 3.927212953567505,
      "step": 24
    },
    {
      "epoch": 0.004822530864197531,
      "grad_norm": 1.418358325958252,
      "learning_rate": 0.00015,
      "loss": 3.9231297969818115,
      "step": 25
    },
    {
      "epoch": 0.005015432098765432,
      "grad_norm": 2.148092269897461,
      "learning_rate": 0.00015625,
      "loss": 3.6635429859161377,
      "step": 26
    },
    {
      "epoch": 0.005208333333333333,
      "grad_norm": 1.1590771675109863,
      "learning_rate": 0.00016250000000000002,
      "loss": 3.8198790550231934,
      "step": 27
    },
    {
      "epoch": 0.005401234567901234,
      "grad_norm": 1.4394375085830688,
      "learning_rate": 0.00016875,
      "loss": 3.9602131843566895,
      "step": 28
    },
    {
      "epoch": 0.0055941358024691355,
      "grad_norm": 1.0043275356292725,
      "learning_rate": 0.000175,
      "loss": 3.7557246685028076,
      "step": 29
    },
    {
      "epoch": 0.005787037037037037,
      "grad_norm": 1.219319224357605,
      "learning_rate": 0.00018125,
      "loss": 3.774125576019287,
      "step": 30
    },
    {
      "epoch": 0.005979938271604938,
      "grad_norm": 1.2093439102172852,
      "learning_rate": 0.0001875,
      "loss": 3.883024215698242,
      "step": 31
    },
    {
      "epoch": 0.006172839506172839,
      "grad_norm": 1.0561580657958984,
      "learning_rate": 0.00019375000000000002,
      "loss": 4.029836177825928,
      "step": 32
    },
    {
      "epoch": 0.00636574074074074,
      "grad_norm": 1.040693998336792,
      "learning_rate": 0.0002,
      "loss": 3.9616167545318604,
      "step": 33
    },
    {
      "epoch": 0.006558641975308642,
      "grad_norm": 1.1366581916809082,
      "learning_rate": 0.00020625,
      "loss": 3.6981022357940674,
      "step": 34
    },
    {
      "epoch": 0.006751543209876543,
      "grad_norm": 1.404786229133606,
      "learning_rate": 0.0002125,
      "loss": 3.94047212600708,
      "step": 35
    },
    {
      "epoch": 0.006944444444444444,
      "grad_norm": 0.9822430610656738,
      "learning_rate": 0.00021875,
      "loss": 3.7842183113098145,
      "step": 36
    },
    {
      "epoch": 0.007137345679012345,
      "grad_norm": 1.2727572917938232,
      "learning_rate": 0.00022500000000000002,
      "loss": 4.128912925720215,
      "step": 37
    },
    {
      "epoch": 0.0073302469135802465,
      "grad_norm": 1.3097913265228271,
      "learning_rate": 0.00023125,
      "loss": 3.96055269241333,
      "step": 38
    },
    {
      "epoch": 0.007523148148148148,
      "grad_norm": 1.5751665830612183,
      "learning_rate": 0.0002375,
      "loss": 4.306777000427246,
      "step": 39
    },
    {
      "epoch": 0.007716049382716049,
      "grad_norm": 1.077038049697876,
      "learning_rate": 0.00024375,
      "loss": 3.950362205505371,
      "step": 40
    },
    {
      "epoch": 0.007908950617283951,
      "grad_norm": 1.0885233879089355,
      "learning_rate": 0.00025,
      "loss": 3.6393895149230957,
      "step": 41
    },
    {
      "epoch": 0.008101851851851851,
      "grad_norm": 1.4571539163589478,
      "learning_rate": 0.00025624999999999997,
      "loss": 3.7999038696289062,
      "step": 42
    },
    {
      "epoch": 0.008294753086419753,
      "grad_norm": 1.2055562734603882,
      "learning_rate": 0.00026250000000000004,
      "loss": 3.835407018661499,
      "step": 43
    },
    {
      "epoch": 0.008487654320987654,
      "grad_norm": 1.2545666694641113,
      "learning_rate": 0.00026875,
      "loss": 3.6124181747436523,
      "step": 44
    },
    {
      "epoch": 0.008680555555555556,
      "grad_norm": 1.1803745031356812,
      "learning_rate": 0.000275,
      "loss": 3.3365509510040283,
      "step": 45
    },
    {
      "epoch": 0.008873456790123456,
      "grad_norm": 1.3649364709854126,
      "learning_rate": 0.00028125000000000003,
      "loss": 3.6861116886138916,
      "step": 46
    },
    {
      "epoch": 0.009066358024691358,
      "grad_norm": 1.470460295677185,
      "learning_rate": 0.0002875,
      "loss": 3.6258745193481445,
      "step": 47
    },
    {
      "epoch": 0.009259259259259259,
      "grad_norm": 1.3595068454742432,
      "learning_rate": 0.00029375,
      "loss": 3.7867329120635986,
      "step": 48
    },
    {
      "epoch": 0.00945216049382716,
      "grad_norm": 1.1562988758087158,
      "learning_rate": 0.0003,
      "loss": 3.6644954681396484,
      "step": 49
    },
    {
      "epoch": 0.009645061728395061,
      "grad_norm": 1.0972641706466675,
      "learning_rate": 0.00030625000000000004,
      "loss": 3.570897102355957,
      "step": 50
    },
    {
      "epoch": 0.009837962962962963,
      "grad_norm": 1.2119516134262085,
      "learning_rate": 0.0003125,
      "loss": 3.7055177688598633,
      "step": 51
    },
    {
      "epoch": 0.010030864197530864,
      "grad_norm": 1.205535888671875,
      "learning_rate": 0.00031874999999999997,
      "loss": 3.817044734954834,
      "step": 52
    },
    {
      "epoch": 0.010223765432098766,
      "grad_norm": 1.335341215133667,
      "learning_rate": 0.00032500000000000004,
      "loss": 3.812642812728882,
      "step": 53
    },
    {
      "epoch": 0.010416666666666666,
      "grad_norm": 1.383876085281372,
      "learning_rate": 0.00033125,
      "loss": 3.874537229537964,
      "step": 54
    },
    {
      "epoch": 0.010609567901234568,
      "grad_norm": 1.0415407419204712,
      "learning_rate": 0.0003375,
      "loss": 3.6900432109832764,
      "step": 55
    },
    {
      "epoch": 0.010802469135802469,
      "grad_norm": 1.140114188194275,
      "learning_rate": 0.00034375,
      "loss": 3.738718032836914,
      "step": 56
    },
    {
      "epoch": 0.01099537037037037,
      "grad_norm": 1.5068063735961914,
      "learning_rate": 0.00035,
      "loss": 3.656860589981079,
      "step": 57
    },
    {
      "epoch": 0.011188271604938271,
      "grad_norm": 1.2487362623214722,
      "learning_rate": 0.00035625,
      "loss": 3.407489776611328,
      "step": 58
    },
    {
      "epoch": 0.011381172839506173,
      "grad_norm": 1.2244573831558228,
      "learning_rate": 0.0003625,
      "loss": 3.727595806121826,
      "step": 59
    },
    {
      "epoch": 0.011574074074074073,
      "grad_norm": 1.3702818155288696,
      "learning_rate": 0.00036875000000000005,
      "loss": 3.854177474975586,
      "step": 60
    },
    {
      "epoch": 0.011766975308641976,
      "grad_norm": 1.3643356561660767,
      "learning_rate": 0.000375,
      "loss": 3.957091808319092,
      "step": 61
    },
    {
      "epoch": 0.011959876543209876,
      "grad_norm": 1.2439842224121094,
      "learning_rate": 0.00038124999999999997,
      "loss": 3.693758487701416,
      "step": 62
    },
    {
      "epoch": 0.012152777777777778,
      "grad_norm": 1.1346251964569092,
      "learning_rate": 0.00038750000000000004,
      "loss": 3.582406997680664,
      "step": 63
    },
    {
      "epoch": 0.012345679012345678,
      "grad_norm": 1.4942717552185059,
      "learning_rate": 0.00039375,
      "loss": 3.463860511779785,
      "step": 64
    },
    {
      "epoch": 0.01253858024691358,
      "grad_norm": 1.0529946088790894,
      "learning_rate": 0.0004,
      "loss": 3.7295355796813965,
      "step": 65
    },
    {
      "epoch": 0.01273148148148148,
      "grad_norm": 1.2595622539520264,
      "learning_rate": 0.00040625000000000004,
      "loss": 3.463494300842285,
      "step": 66
    },
    {
      "epoch": 0.012924382716049383,
      "grad_norm": 1.4135781526565552,
      "learning_rate": 0.0004125,
      "loss": 3.8828420639038086,
      "step": 67
    },
    {
      "epoch": 0.013117283950617283,
      "grad_norm": 1.118549108505249,
      "learning_rate": 0.00041875,
      "loss": 3.440481424331665,
      "step": 68
    },
    {
      "epoch": 0.013310185185185185,
      "grad_norm": 1.4578617811203003,
      "learning_rate": 0.000425,
      "loss": 3.896862506866455,
      "step": 69
    },
    {
      "epoch": 0.013503086419753086,
      "grad_norm": 1.0785441398620605,
      "learning_rate": 0.00043125000000000005,
      "loss": 3.661886692047119,
      "step": 70
    },
    {
      "epoch": 0.013695987654320988,
      "grad_norm": 1.6024396419525146,
      "learning_rate": 0.0004375,
      "loss": 3.7423009872436523,
      "step": 71
    },
    {
      "epoch": 0.013888888888888888,
      "grad_norm": 1.58858060836792,
      "learning_rate": 0.00044374999999999997,
      "loss": 4.168385982513428,
      "step": 72
    },
    {
      "epoch": 0.01408179012345679,
      "grad_norm": 1.2670390605926514,
      "learning_rate": 0.00045000000000000004,
      "loss": 4.015345573425293,
      "step": 73
    },
    {
      "epoch": 0.01427469135802469,
      "grad_norm": 1.2151397466659546,
      "learning_rate": 0.00045625,
      "loss": 3.8745439052581787,
      "step": 74
    },
    {
      "epoch": 0.014467592592592593,
      "grad_norm": 1.0676912069320679,
      "learning_rate": 0.0004625,
      "loss": 3.607182502746582,
      "step": 75
    },
    {
      "epoch": 0.014660493827160493,
      "grad_norm": 1.4515150785446167,
      "learning_rate": 0.00046875,
      "loss": 3.4360766410827637,
      "step": 76
    },
    {
      "epoch": 0.014853395061728395,
      "grad_norm": 1.657053828239441,
      "learning_rate": 0.000475,
      "loss": 3.647386074066162,
      "step": 77
    },
    {
      "epoch": 0.015046296296296295,
      "grad_norm": 1.1519591808319092,
      "learning_rate": 0.00048125,
      "loss": 3.8803391456604004,
      "step": 78
    },
    {
      "epoch": 0.015239197530864198,
      "grad_norm": 1.3643865585327148,
      "learning_rate": 0.0004875,
      "loss": 3.473296642303467,
      "step": 79
    },
    {
      "epoch": 0.015432098765432098,
      "grad_norm": 1.501682996749878,
      "learning_rate": 0.00049375,
      "loss": 3.5927562713623047,
      "step": 80
    },
    {
      "epoch": 0.015625,
      "grad_norm": 1.5572752952575684,
      "learning_rate": 0.0005,
      "loss": 3.4510440826416016,
      "step": 81
    },
    {
      "epoch": 0.015817901234567902,
      "grad_norm": 1.8052380084991455,
      "learning_rate": 0.0005,
      "loss": 3.769073724746704,
      "step": 82
    },
    {
      "epoch": 0.016010802469135804,
      "grad_norm": 1.3284447193145752,
      "learning_rate": 0.0005,
      "loss": 3.6008520126342773,
      "step": 83
    },
    {
      "epoch": 0.016203703703703703,
      "grad_norm": 1.2453041076660156,
      "learning_rate": 0.0005,
      "loss": 3.9463119506835938,
      "step": 84
    },
    {
      "epoch": 0.016396604938271605,
      "grad_norm": 1.4702776670455933,
      "learning_rate": 0.0005,
      "loss": 3.7285189628601074,
      "step": 85
    },
    {
      "epoch": 0.016589506172839507,
      "grad_norm": 1.342969298362732,
      "learning_rate": 0.0005,
      "loss": 3.698944568634033,
      "step": 86
    },
    {
      "epoch": 0.01678240740740741,
      "grad_norm": 1.3860788345336914,
      "learning_rate": 0.0005,
      "loss": 3.9518280029296875,
      "step": 87
    },
    {
      "epoch": 0.016975308641975308,
      "grad_norm": 1.2387818098068237,
      "learning_rate": 0.0005,
      "loss": 3.8000659942626953,
      "step": 88
    },
    {
      "epoch": 0.01716820987654321,
      "grad_norm": 1.460057258605957,
      "learning_rate": 0.0005,
      "loss": 4.134321212768555,
      "step": 89
    },
    {
      "epoch": 0.017361111111111112,
      "grad_norm": 1.4746754169464111,
      "learning_rate": 0.0005,
      "loss": 3.6365432739257812,
      "step": 90
    },
    {
      "epoch": 0.017554012345679014,
      "grad_norm": 1.4172009229660034,
      "learning_rate": 0.0005,
      "loss": 3.828505516052246,
      "step": 91
    },
    {
      "epoch": 0.017746913580246913,
      "grad_norm": 1.081656813621521,
      "learning_rate": 0.0005,
      "loss": 3.7325501441955566,
      "step": 92
    },
    {
      "epoch": 0.017939814814814815,
      "grad_norm": 1.421101689338684,
      "learning_rate": 0.0005,
      "loss": 3.552325487136841,
      "step": 93
    },
    {
      "epoch": 0.018132716049382717,
      "grad_norm": 1.3186367750167847,
      "learning_rate": 0.0005,
      "loss": 3.7668890953063965,
      "step": 94
    },
    {
      "epoch": 0.01832561728395062,
      "grad_norm": 1.1484084129333496,
      "learning_rate": 0.0005,
      "loss": 3.8461124897003174,
      "step": 95
    },
    {
      "epoch": 0.018518518518518517,
      "grad_norm": 1.513921856880188,
      "learning_rate": 0.0005,
      "loss": 3.8794684410095215,
      "step": 96
    },
    {
      "epoch": 0.01871141975308642,
      "grad_norm": 1.7873536348342896,
      "learning_rate": 0.0005,
      "loss": 3.6786515712738037,
      "step": 97
    },
    {
      "epoch": 0.01890432098765432,
      "grad_norm": 1.3280023336410522,
      "learning_rate": 0.0005,
      "loss": 3.659904956817627,
      "step": 98
    },
    {
      "epoch": 0.019097222222222224,
      "grad_norm": 1.3190447092056274,
      "learning_rate": 0.0005,
      "loss": 4.029374599456787,
      "step": 99
    },
    {
      "epoch": 0.019290123456790122,
      "grad_norm": 1.9318208694458008,
      "learning_rate": 0.0005,
      "loss": 3.735511302947998,
      "step": 100
    },
    {
      "epoch": 0.019483024691358024,
      "grad_norm": 1.2463411092758179,
      "learning_rate": 0.0005,
      "loss": 4.084259033203125,
      "step": 101
    },
    {
      "epoch": 0.019675925925925927,
      "grad_norm": 1.3785924911499023,
      "learning_rate": 0.0005,
      "loss": 3.6822543144226074,
      "step": 102
    },
    {
      "epoch": 0.01986882716049383,
      "grad_norm": 1.1708344221115112,
      "learning_rate": 0.0005,
      "loss": 3.5617332458496094,
      "step": 103
    },
    {
      "epoch": 0.020061728395061727,
      "grad_norm": 1.4579331874847412,
      "learning_rate": 0.0005,
      "loss": 3.5945322513580322,
      "step": 104
    },
    {
      "epoch": 0.02025462962962963,
      "grad_norm": 2.1063106060028076,
      "learning_rate": 0.0005,
      "loss": 3.9631035327911377,
      "step": 105
    },
    {
      "epoch": 0.02044753086419753,
      "grad_norm": 1.3162477016448975,
      "learning_rate": 0.0005,
      "loss": 3.616743564605713,
      "step": 106
    },
    {
      "epoch": 0.020640432098765434,
      "grad_norm": 1.6704325675964355,
      "learning_rate": 0.0005,
      "loss": 3.6599886417388916,
      "step": 107
    },
    {
      "epoch": 0.020833333333333332,
      "grad_norm": 1.3518601655960083,
      "learning_rate": 0.0005,
      "loss": 4.030555248260498,
      "step": 108
    },
    {
      "epoch": 0.021026234567901234,
      "grad_norm": 1.7611132860183716,
      "learning_rate": 0.0005,
      "loss": 3.604458808898926,
      "step": 109
    },
    {
      "epoch": 0.021219135802469136,
      "grad_norm": 1.284462571144104,
      "learning_rate": 0.0005,
      "loss": 3.835451364517212,
      "step": 110
    },
    {
      "epoch": 0.02141203703703704,
      "grad_norm": 1.2162014245986938,
      "learning_rate": 0.0005,
      "loss": 3.618359327316284,
      "step": 111
    },
    {
      "epoch": 0.021604938271604937,
      "grad_norm": 1.5592894554138184,
      "learning_rate": 0.0005,
      "loss": 3.8871707916259766,
      "step": 112
    },
    {
      "epoch": 0.02179783950617284,
      "grad_norm": 1.2892863750457764,
      "learning_rate": 0.0005,
      "loss": 3.5625481605529785,
      "step": 113
    },
    {
      "epoch": 0.02199074074074074,
      "grad_norm": 1.3381227254867554,
      "learning_rate": 0.0005,
      "loss": 3.872616767883301,
      "step": 114
    },
    {
      "epoch": 0.022183641975308643,
      "grad_norm": 1.1598743200302124,
      "learning_rate": 0.0005,
      "loss": 3.753765821456909,
      "step": 115
    },
    {
      "epoch": 0.022376543209876542,
      "grad_norm": 1.3360445499420166,
      "learning_rate": 0.0005,
      "loss": 3.9263358116149902,
      "step": 116
    },
    {
      "epoch": 0.022569444444444444,
      "grad_norm": 1.3400583267211914,
      "learning_rate": 0.0005,
      "loss": 3.5426156520843506,
      "step": 117
    },
    {
      "epoch": 0.022762345679012346,
      "grad_norm": 1.3544574975967407,
      "learning_rate": 0.0005,
      "loss": 3.3801074028015137,
      "step": 118
    },
    {
      "epoch": 0.022955246913580248,
      "grad_norm": 1.3499839305877686,
      "learning_rate": 0.0005,
      "loss": 3.4756247997283936,
      "step": 119
    },
    {
      "epoch": 0.023148148148148147,
      "grad_norm": 1.1695464849472046,
      "learning_rate": 0.0005,
      "loss": 3.47745418548584,
      "step": 120
    },
    {
      "epoch": 0.02334104938271605,
      "grad_norm": 1.2836657762527466,
      "learning_rate": 0.0005,
      "loss": 3.6878838539123535,
      "step": 121
    },
    {
      "epoch": 0.02353395061728395,
      "grad_norm": 1.5552871227264404,
      "learning_rate": 0.0005,
      "loss": 4.145421981811523,
      "step": 122
    },
    {
      "epoch": 0.023726851851851853,
      "grad_norm": 1.1740211248397827,
      "learning_rate": 0.0005,
      "loss": 3.825984001159668,
      "step": 123
    },
    {
      "epoch": 0.023919753086419752,
      "grad_norm": 1.3595014810562134,
      "learning_rate": 0.0005,
      "loss": 3.7620954513549805,
      "step": 124
    },
    {
      "epoch": 0.024112654320987654,
      "grad_norm": 1.4141457080841064,
      "learning_rate": 0.0005,
      "loss": 3.411465644836426,
      "step": 125
    },
    {
      "epoch": 0.024305555555555556,
      "grad_norm": 1.2813143730163574,
      "learning_rate": 0.0005,
      "loss": 3.5657143592834473,
      "step": 126
    },
    {
      "epoch": 0.024498456790123458,
      "grad_norm": 1.4832956790924072,
      "learning_rate": 0.0005,
      "loss": 4.020408630371094,
      "step": 127
    },
    {
      "epoch": 0.024691358024691357,
      "grad_norm": 1.3184418678283691,
      "learning_rate": 0.0005,
      "loss": 3.7670159339904785,
      "step": 128
    },
    {
      "epoch": 0.02488425925925926,
      "grad_norm": 1.153652548789978,
      "learning_rate": 0.0005,
      "loss": 3.8316164016723633,
      "step": 129
    },
    {
      "epoch": 0.02507716049382716,
      "grad_norm": 1.638903260231018,
      "learning_rate": 0.0005,
      "loss": 3.7815632820129395,
      "step": 130
    },
    {
      "epoch": 0.025270061728395063,
      "grad_norm": 1.3114980459213257,
      "learning_rate": 0.0005,
      "loss": 3.938140392303467,
      "step": 131
    },
    {
      "epoch": 0.02546296296296296,
      "grad_norm": 1.0280201435089111,
      "learning_rate": 0.0005,
      "loss": 3.872162342071533,
      "step": 132
    },
    {
      "epoch": 0.025655864197530864,
      "grad_norm": 1.8902846574783325,
      "learning_rate": 0.0005,
      "loss": 3.6973321437835693,
      "step": 133
    },
    {
      "epoch": 0.025848765432098766,
      "grad_norm": 1.2277593612670898,
      "learning_rate": 0.0005,
      "loss": 3.382333278656006,
      "step": 134
    },
    {
      "epoch": 0.026041666666666668,
      "grad_norm": 1.440382957458496,
      "learning_rate": 0.0005,
      "loss": 4.051923751831055,
      "step": 135
    },
    {
      "epoch": 0.026234567901234566,
      "grad_norm": 1.5983177423477173,
      "learning_rate": 0.0005,
      "loss": 4.006111145019531,
      "step": 136
    },
    {
      "epoch": 0.02642746913580247,
      "grad_norm": 1.2373310327529907,
      "learning_rate": 0.0005,
      "loss": 4.0500874519348145,
      "step": 137
    },
    {
      "epoch": 0.02662037037037037,
      "grad_norm": 1.2794286012649536,
      "learning_rate": 0.0005,
      "loss": 3.9101483821868896,
      "step": 138
    },
    {
      "epoch": 0.026813271604938273,
      "grad_norm": 1.31977117061615,
      "learning_rate": 0.0005,
      "loss": 3.5188956260681152,
      "step": 139
    },
    {
      "epoch": 0.02700617283950617,
      "grad_norm": 1.534576416015625,
      "learning_rate": 0.0005,
      "loss": 3.638352155685425,
      "step": 140
    },
    {
      "epoch": 0.027199074074074073,
      "grad_norm": 1.4312539100646973,
      "learning_rate": 0.0005,
      "loss": 3.784294605255127,
      "step": 141
    },
    {
      "epoch": 0.027391975308641976,
      "grad_norm": 1.4572392702102661,
      "learning_rate": 0.0005,
      "loss": 3.783742666244507,
      "step": 142
    },
    {
      "epoch": 0.027584876543209878,
      "grad_norm": 1.1545891761779785,
      "learning_rate": 0.0005,
      "loss": 3.709083080291748,
      "step": 143
    },
    {
      "epoch": 0.027777777777777776,
      "grad_norm": 1.76921808719635,
      "learning_rate": 0.0005,
      "loss": 3.8412742614746094,
      "step": 144
    },
    {
      "epoch": 0.02797067901234568,
      "grad_norm": 1.4666557312011719,
      "learning_rate": 0.0005,
      "loss": 3.6306285858154297,
      "step": 145
    },
    {
      "epoch": 0.02816358024691358,
      "grad_norm": 1.7845485210418701,
      "learning_rate": 0.0005,
      "loss": 3.904366970062256,
      "step": 146
    },
    {
      "epoch": 0.028356481481481483,
      "grad_norm": 2.094230890274048,
      "learning_rate": 0.0005,
      "loss": 3.489095687866211,
      "step": 147
    },
    {
      "epoch": 0.02854938271604938,
      "grad_norm": 1.2497845888137817,
      "learning_rate": 0.0005,
      "loss": 3.777402400970459,
      "step": 148
    },
    {
      "epoch": 0.028742283950617283,
      "grad_norm": 1.2698285579681396,
      "learning_rate": 0.0005,
      "loss": 3.5945043563842773,
      "step": 149
    },
    {
      "epoch": 0.028935185185185185,
      "grad_norm": 1.4772274494171143,
      "learning_rate": 0.0005,
      "loss": 3.5938448905944824,
      "step": 150
    },
    {
      "epoch": 0.029128086419753087,
      "grad_norm": 1.3386154174804688,
      "learning_rate": 0.0005,
      "loss": 3.490586280822754,
      "step": 151
    },
    {
      "epoch": 0.029320987654320986,
      "grad_norm": 1.279863715171814,
      "learning_rate": 0.0005,
      "loss": 3.7381205558776855,
      "step": 152
    },
    {
      "epoch": 0.029513888888888888,
      "grad_norm": 1.9305498600006104,
      "learning_rate": 0.0005,
      "loss": 3.6317875385284424,
      "step": 153
    },
    {
      "epoch": 0.02970679012345679,
      "grad_norm": 1.3103426694869995,
      "learning_rate": 0.0005,
      "loss": 3.9144701957702637,
      "step": 154
    },
    {
      "epoch": 0.029899691358024692,
      "grad_norm": 1.3389663696289062,
      "learning_rate": 0.0005,
      "loss": 3.6266989707946777,
      "step": 155
    },
    {
      "epoch": 0.03009259259259259,
      "grad_norm": 1.8015128374099731,
      "learning_rate": 0.0005,
      "loss": 3.2674074172973633,
      "step": 156
    },
    {
      "epoch": 0.030285493827160493,
      "grad_norm": 1.4903942346572876,
      "learning_rate": 0.0005,
      "loss": 3.7776079177856445,
      "step": 157
    },
    {
      "epoch": 0.030478395061728395,
      "grad_norm": 1.3954755067825317,
      "learning_rate": 0.0005,
      "loss": 3.563872814178467,
      "step": 158
    },
    {
      "epoch": 0.030671296296296297,
      "grad_norm": 1.5027440786361694,
      "learning_rate": 0.0005,
      "loss": 3.437713623046875,
      "step": 159
    },
    {
      "epoch": 0.030864197530864196,
      "grad_norm": 1.513309359550476,
      "learning_rate": 0.0005,
      "loss": 3.8224713802337646,
      "step": 160
    },
    {
      "epoch": 0.031057098765432098,
      "grad_norm": 1.2092558145523071,
      "learning_rate": 0.0005,
      "loss": 3.5412240028381348,
      "step": 161
    },
    {
      "epoch": 0.03125,
      "grad_norm": 1.1557204723358154,
      "learning_rate": 0.0005,
      "loss": 3.8367087841033936,
      "step": 162
    },
    {
      "epoch": 0.0314429012345679,
      "grad_norm": 1.488670825958252,
      "learning_rate": 0.0005,
      "loss": 3.9882254600524902,
      "step": 163
    },
    {
      "epoch": 0.031635802469135804,
      "grad_norm": 1.628855586051941,
      "learning_rate": 0.0005,
      "loss": 4.006572723388672,
      "step": 164
    },
    {
      "epoch": 0.031828703703703706,
      "grad_norm": 1.3451560735702515,
      "learning_rate": 0.0005,
      "loss": 3.45794677734375,
      "step": 165
    },
    {
      "epoch": 0.03202160493827161,
      "grad_norm": 1.6431349515914917,
      "learning_rate": 0.0005,
      "loss": 3.8013768196105957,
      "step": 166
    },
    {
      "epoch": 0.032214506172839504,
      "grad_norm": 1.5793012380599976,
      "learning_rate": 0.0005,
      "loss": 3.5150060653686523,
      "step": 167
    },
    {
      "epoch": 0.032407407407407406,
      "grad_norm": 1.546570897102356,
      "learning_rate": 0.0005,
      "loss": 3.4856133460998535,
      "step": 168
    },
    {
      "epoch": 0.03260030864197531,
      "grad_norm": 1.298073410987854,
      "learning_rate": 0.0005,
      "loss": 4.042288303375244,
      "step": 169
    },
    {
      "epoch": 0.03279320987654321,
      "grad_norm": 1.2446430921554565,
      "learning_rate": 0.0005,
      "loss": 3.6909828186035156,
      "step": 170
    },
    {
      "epoch": 0.03298611111111111,
      "grad_norm": 1.5500397682189941,
      "learning_rate": 0.0005,
      "loss": 3.6376099586486816,
      "step": 171
    },
    {
      "epoch": 0.033179012345679014,
      "grad_norm": 1.5147318840026855,
      "learning_rate": 0.0005,
      "loss": 3.9575319290161133,
      "step": 172
    },
    {
      "epoch": 0.033371913580246916,
      "grad_norm": 1.3858450651168823,
      "learning_rate": 0.0005,
      "loss": 3.6567423343658447,
      "step": 173
    },
    {
      "epoch": 0.03356481481481482,
      "grad_norm": 1.1270774602890015,
      "learning_rate": 0.0005,
      "loss": 3.9291176795959473,
      "step": 174
    },
    {
      "epoch": 0.03375771604938271,
      "grad_norm": 1.2811940908432007,
      "learning_rate": 0.0005,
      "loss": 3.7183971405029297,
      "step": 175
    },
    {
      "epoch": 0.033950617283950615,
      "grad_norm": 1.7076513767242432,
      "learning_rate": 0.0005,
      "loss": 3.845083236694336,
      "step": 176
    },
    {
      "epoch": 0.03414351851851852,
      "grad_norm": 1.2620930671691895,
      "learning_rate": 0.0005,
      "loss": 3.6597037315368652,
      "step": 177
    },
    {
      "epoch": 0.03433641975308642,
      "grad_norm": 1.310682773590088,
      "learning_rate": 0.0005,
      "loss": 3.5478861331939697,
      "step": 178
    },
    {
      "epoch": 0.03452932098765432,
      "grad_norm": 1.3072575330734253,
      "learning_rate": 0.0005,
      "loss": 3.6832706928253174,
      "step": 179
    },
    {
      "epoch": 0.034722222222222224,
      "grad_norm": 1.2117111682891846,
      "learning_rate": 0.0005,
      "loss": 3.8141415119171143,
      "step": 180
    },
    {
      "epoch": 0.034915123456790126,
      "grad_norm": 1.1781898736953735,
      "learning_rate": 0.0005,
      "loss": 3.7138261795043945,
      "step": 181
    },
    {
      "epoch": 0.03510802469135803,
      "grad_norm": 1.0816185474395752,
      "learning_rate": 0.0005,
      "loss": 3.7679271697998047,
      "step": 182
    },
    {
      "epoch": 0.03530092592592592,
      "grad_norm": 1.348868489265442,
      "learning_rate": 0.0005,
      "loss": 3.6297645568847656,
      "step": 183
    },
    {
      "epoch": 0.035493827160493825,
      "grad_norm": 1.921544075012207,
      "learning_rate": 0.0005,
      "loss": 3.479147434234619,
      "step": 184
    },
    {
      "epoch": 0.03568672839506173,
      "grad_norm": 1.3129043579101562,
      "learning_rate": 0.0005,
      "loss": 3.553323745727539,
      "step": 185
    },
    {
      "epoch": 0.03587962962962963,
      "grad_norm": 1.1518778800964355,
      "learning_rate": 0.0005,
      "loss": 3.9791579246520996,
      "step": 186
    },
    {
      "epoch": 0.03607253086419753,
      "grad_norm": 1.2575280666351318,
      "learning_rate": 0.0005,
      "loss": 3.987550973892212,
      "step": 187
    },
    {
      "epoch": 0.036265432098765434,
      "grad_norm": 1.3994395732879639,
      "learning_rate": 0.0005,
      "loss": 3.9356977939605713,
      "step": 188
    },
    {
      "epoch": 0.036458333333333336,
      "grad_norm": 1.5580729246139526,
      "learning_rate": 0.0005,
      "loss": 3.9506328105926514,
      "step": 189
    },
    {
      "epoch": 0.03665123456790124,
      "grad_norm": 1.6496039628982544,
      "learning_rate": 0.0005,
      "loss": 3.9103684425354004,
      "step": 190
    },
    {
      "epoch": 0.03684413580246913,
      "grad_norm": 1.423442006111145,
      "learning_rate": 0.0005,
      "loss": 3.41021990776062,
      "step": 191
    },
    {
      "epoch": 0.037037037037037035,
      "grad_norm": 1.2442165613174438,
      "learning_rate": 0.0005,
      "loss": 3.7414188385009766,
      "step": 192
    },
    {
      "epoch": 0.03722993827160494,
      "grad_norm": 1.3725247383117676,
      "learning_rate": 0.0005,
      "loss": 3.981034994125366,
      "step": 193
    },
    {
      "epoch": 0.03742283950617284,
      "grad_norm": 1.151016116142273,
      "learning_rate": 0.0005,
      "loss": 3.8856234550476074,
      "step": 194
    },
    {
      "epoch": 0.03761574074074074,
      "grad_norm": 1.2752842903137207,
      "learning_rate": 0.0005,
      "loss": 4.090315818786621,
      "step": 195
    },
    {
      "epoch": 0.03780864197530864,
      "grad_norm": 1.4142334461212158,
      "learning_rate": 0.0005,
      "loss": 3.7500412464141846,
      "step": 196
    },
    {
      "epoch": 0.038001543209876545,
      "grad_norm": 1.508241057395935,
      "learning_rate": 0.0005,
      "loss": 3.6497581005096436,
      "step": 197
    },
    {
      "epoch": 0.03819444444444445,
      "grad_norm": 1.3282926082611084,
      "learning_rate": 0.0005,
      "loss": 3.891979455947876,
      "step": 198
    },
    {
      "epoch": 0.03838734567901234,
      "grad_norm": 1.213867425918579,
      "learning_rate": 0.0005,
      "loss": 3.5258572101593018,
      "step": 199
    },
    {
      "epoch": 0.038580246913580245,
      "grad_norm": 1.2745705842971802,
      "learning_rate": 0.0005,
      "loss": 3.7925448417663574,
      "step": 200
    },
    {
      "epoch": 0.03877314814814815,
      "grad_norm": 1.3837401866912842,
      "learning_rate": 0.0005,
      "loss": 3.7137622833251953,
      "step": 201
    },
    {
      "epoch": 0.03896604938271605,
      "grad_norm": 1.2990055084228516,
      "learning_rate": 0.0005,
      "loss": 3.8282501697540283,
      "step": 202
    },
    {
      "epoch": 0.03915895061728395,
      "grad_norm": 1.1232761144638062,
      "learning_rate": 0.0005,
      "loss": 3.7956953048706055,
      "step": 203
    },
    {
      "epoch": 0.03935185185185185,
      "grad_norm": 1.1890406608581543,
      "learning_rate": 0.0005,
      "loss": 3.795719861984253,
      "step": 204
    },
    {
      "epoch": 0.039544753086419755,
      "grad_norm": 1.2189326286315918,
      "learning_rate": 0.0005,
      "loss": 3.495417594909668,
      "step": 205
    },
    {
      "epoch": 0.03973765432098766,
      "grad_norm": 1.1110111474990845,
      "learning_rate": 0.0005,
      "loss": 3.670182704925537,
      "step": 206
    },
    {
      "epoch": 0.03993055555555555,
      "grad_norm": 1.2798066139221191,
      "learning_rate": 0.0005,
      "loss": 3.4889581203460693,
      "step": 207
    },
    {
      "epoch": 0.040123456790123455,
      "grad_norm": 1.3908721208572388,
      "learning_rate": 0.0005,
      "loss": 3.7110114097595215,
      "step": 208
    },
    {
      "epoch": 0.04031635802469136,
      "grad_norm": 1.5076323747634888,
      "learning_rate": 0.0005,
      "loss": 3.7591724395751953,
      "step": 209
    },
    {
      "epoch": 0.04050925925925926,
      "grad_norm": 1.3979840278625488,
      "learning_rate": 0.0005,
      "loss": 3.5506110191345215,
      "step": 210
    },
    {
      "epoch": 0.04070216049382716,
      "grad_norm": 1.5169926881790161,
      "learning_rate": 0.0005,
      "loss": 3.7610023021698,
      "step": 211
    },
    {
      "epoch": 0.04089506172839506,
      "grad_norm": 1.07994544506073,
      "learning_rate": 0.0005,
      "loss": 3.8453238010406494,
      "step": 212
    },
    {
      "epoch": 0.041087962962962965,
      "grad_norm": 1.015008568763733,
      "learning_rate": 0.0005,
      "loss": 3.654393434524536,
      "step": 213
    },
    {
      "epoch": 0.04128086419753087,
      "grad_norm": 1.6805564165115356,
      "learning_rate": 0.0005,
      "loss": 3.984969139099121,
      "step": 214
    },
    {
      "epoch": 0.04147376543209876,
      "grad_norm": 1.6445502042770386,
      "learning_rate": 0.0005,
      "loss": 4.080695152282715,
      "step": 215
    },
    {
      "epoch": 0.041666666666666664,
      "grad_norm": 1.152055025100708,
      "learning_rate": 0.0005,
      "loss": 3.8268885612487793,
      "step": 216
    },
    {
      "epoch": 0.041859567901234566,
      "grad_norm": 1.232177734375,
      "learning_rate": 0.0005,
      "loss": 3.6531262397766113,
      "step": 217
    },
    {
      "epoch": 0.04205246913580247,
      "grad_norm": 1.173888087272644,
      "learning_rate": 0.0005,
      "loss": 3.7033579349517822,
      "step": 218
    },
    {
      "epoch": 0.04224537037037037,
      "grad_norm": 1.3221077919006348,
      "learning_rate": 0.0005,
      "loss": 3.544025182723999,
      "step": 219
    },
    {
      "epoch": 0.04243827160493827,
      "grad_norm": 1.9697234630584717,
      "learning_rate": 0.0005,
      "loss": 3.6551594734191895,
      "step": 220
    },
    {
      "epoch": 0.042631172839506175,
      "grad_norm": 1.2494007349014282,
      "learning_rate": 0.0005,
      "loss": 3.6527905464172363,
      "step": 221
    },
    {
      "epoch": 0.04282407407407408,
      "grad_norm": 1.1801296472549438,
      "learning_rate": 0.0005,
      "loss": 3.576124668121338,
      "step": 222
    },
    {
      "epoch": 0.04301697530864197,
      "grad_norm": 1.1753565073013306,
      "learning_rate": 0.0005,
      "loss": 3.7326486110687256,
      "step": 223
    },
    {
      "epoch": 0.043209876543209874,
      "grad_norm": 1.1077008247375488,
      "learning_rate": 0.0005,
      "loss": 3.592320203781128,
      "step": 224
    },
    {
      "epoch": 0.043402777777777776,
      "grad_norm": 1.3996425867080688,
      "learning_rate": 0.0005,
      "loss": 3.5467309951782227,
      "step": 225
    },
    {
      "epoch": 0.04359567901234568,
      "grad_norm": 1.2290549278259277,
      "learning_rate": 0.0005,
      "loss": 3.7132296562194824,
      "step": 226
    },
    {
      "epoch": 0.04378858024691358,
      "grad_norm": 1.216443657875061,
      "learning_rate": 0.0005,
      "loss": 3.817610263824463,
      "step": 227
    },
    {
      "epoch": 0.04398148148148148,
      "grad_norm": 1.2790571451187134,
      "learning_rate": 0.0005,
      "loss": 3.5374503135681152,
      "step": 228
    },
    {
      "epoch": 0.044174382716049385,
      "grad_norm": 1.2919474840164185,
      "learning_rate": 0.0005,
      "loss": 3.8442702293395996,
      "step": 229
    },
    {
      "epoch": 0.04436728395061729,
      "grad_norm": 1.2768014669418335,
      "learning_rate": 0.0005,
      "loss": 3.77799654006958,
      "step": 230
    },
    {
      "epoch": 0.04456018518518518,
      "grad_norm": 1.1175886392593384,
      "learning_rate": 0.0005,
      "loss": 3.567843437194824,
      "step": 231
    },
    {
      "epoch": 0.044753086419753084,
      "grad_norm": 1.2177129983901978,
      "learning_rate": 0.0005,
      "loss": 3.3076086044311523,
      "step": 232
    },
    {
      "epoch": 0.044945987654320986,
      "grad_norm": 1.2988262176513672,
      "learning_rate": 0.0005,
      "loss": 3.6657874584198,
      "step": 233
    },
    {
      "epoch": 0.04513888888888889,
      "grad_norm": 1.1530272960662842,
      "learning_rate": 0.0005,
      "loss": 3.896561622619629,
      "step": 234
    },
    {
      "epoch": 0.04533179012345679,
      "grad_norm": 1.521980881690979,
      "learning_rate": 0.0005,
      "loss": 3.7518882751464844,
      "step": 235
    },
    {
      "epoch": 0.04552469135802469,
      "grad_norm": 1.3340171575546265,
      "learning_rate": 0.0005,
      "loss": 3.580203056335449,
      "step": 236
    },
    {
      "epoch": 0.045717592592592594,
      "grad_norm": 1.2477669715881348,
      "learning_rate": 0.0005,
      "loss": 3.497690200805664,
      "step": 237
    },
    {
      "epoch": 0.045910493827160496,
      "grad_norm": 1.1120455265045166,
      "learning_rate": 0.0005,
      "loss": 3.706254482269287,
      "step": 238
    },
    {
      "epoch": 0.04610339506172839,
      "grad_norm": 1.3557227849960327,
      "learning_rate": 0.0005,
      "loss": 3.6414618492126465,
      "step": 239
    },
    {
      "epoch": 0.046296296296296294,
      "grad_norm": 1.2988630533218384,
      "learning_rate": 0.0005,
      "loss": 3.703400135040283,
      "step": 240
    },
    {
      "epoch": 0.046489197530864196,
      "grad_norm": 1.4603893756866455,
      "learning_rate": 0.0005,
      "loss": 3.827232837677002,
      "step": 241
    },
    {
      "epoch": 0.0466820987654321,
      "grad_norm": 1.2944862842559814,
      "learning_rate": 0.0005,
      "loss": 3.858074903488159,
      "step": 242
    },
    {
      "epoch": 0.046875,
      "grad_norm": 1.2586787939071655,
      "learning_rate": 0.0005,
      "loss": 3.4835641384124756,
      "step": 243
    },
    {
      "epoch": 0.0470679012345679,
      "grad_norm": 1.1183030605316162,
      "learning_rate": 0.0005,
      "loss": 3.57287859916687,
      "step": 244
    },
    {
      "epoch": 0.047260802469135804,
      "grad_norm": 1.4424525499343872,
      "learning_rate": 0.0005,
      "loss": 3.6359682083129883,
      "step": 245
    },
    {
      "epoch": 0.047453703703703706,
      "grad_norm": 1.6687700748443604,
      "learning_rate": 0.0005,
      "loss": 3.88181209564209,
      "step": 246
    },
    {
      "epoch": 0.04764660493827161,
      "grad_norm": 1.176788091659546,
      "learning_rate": 0.0005,
      "loss": 3.7714085578918457,
      "step": 247
    },
    {
      "epoch": 0.047839506172839504,
      "grad_norm": 1.0965936183929443,
      "learning_rate": 0.0005,
      "loss": 3.8699209690093994,
      "step": 248
    },
    {
      "epoch": 0.048032407407407406,
      "grad_norm": 1.2640173435211182,
      "learning_rate": 0.0005,
      "loss": 3.9561941623687744,
      "step": 249
    },
    {
      "epoch": 0.04822530864197531,
      "grad_norm": 1.2212092876434326,
      "learning_rate": 0.0005,
      "loss": 3.6051220893859863,
      "step": 250
    },
    {
      "epoch": 0.04841820987654321,
      "grad_norm": 1.1415603160858154,
      "learning_rate": 0.0005,
      "loss": 3.75545072555542,
      "step": 251
    },
    {
      "epoch": 0.04861111111111111,
      "grad_norm": 1.2963902950286865,
      "learning_rate": 0.0005,
      "loss": 3.8391199111938477,
      "step": 252
    },
    {
      "epoch": 0.048804012345679014,
      "grad_norm": 1.202605128288269,
      "learning_rate": 0.0005,
      "loss": 4.025400638580322,
      "step": 253
    },
    {
      "epoch": 0.048996913580246916,
      "grad_norm": 1.2146269083023071,
      "learning_rate": 0.0005,
      "loss": 3.3638291358947754,
      "step": 254
    },
    {
      "epoch": 0.04918981481481482,
      "grad_norm": 1.1074600219726562,
      "learning_rate": 0.0005,
      "loss": 3.7058777809143066,
      "step": 255
    },
    {
      "epoch": 0.04938271604938271,
      "grad_norm": 1.4221081733703613,
      "learning_rate": 0.0005,
      "loss": 3.852933883666992,
      "step": 256
    },
    {
      "epoch": 0.049575617283950615,
      "grad_norm": 1.3432855606079102,
      "learning_rate": 0.0005,
      "loss": 3.607649326324463,
      "step": 257
    },
    {
      "epoch": 0.04976851851851852,
      "grad_norm": 1.1616085767745972,
      "learning_rate": 0.0005,
      "loss": 3.7509946823120117,
      "step": 258
    },
    {
      "epoch": 0.04996141975308642,
      "grad_norm": 1.196847677230835,
      "learning_rate": 0.0005,
      "loss": 3.7412755489349365,
      "step": 259
    },
    {
      "epoch": 0.05015432098765432,
      "grad_norm": 1.0041816234588623,
      "learning_rate": 0.0005,
      "loss": 3.8931174278259277,
      "step": 260
    },
    {
      "epoch": 0.050347222222222224,
      "grad_norm": 1.340221643447876,
      "learning_rate": 0.0005,
      "loss": 3.926563262939453,
      "step": 261
    },
    {
      "epoch": 0.050540123456790126,
      "grad_norm": 1.230871319770813,
      "learning_rate": 0.0005,
      "loss": 3.7106854915618896,
      "step": 262
    },
    {
      "epoch": 0.05073302469135803,
      "grad_norm": 1.329300045967102,
      "learning_rate": 0.0005,
      "loss": 3.7967135906219482,
      "step": 263
    },
    {
      "epoch": 0.05092592592592592,
      "grad_norm": 1.2817455530166626,
      "learning_rate": 0.0005,
      "loss": 3.8086562156677246,
      "step": 264
    },
    {
      "epoch": 0.051118827160493825,
      "grad_norm": 1.466762900352478,
      "learning_rate": 0.0005,
      "loss": 3.7692248821258545,
      "step": 265
    },
    {
      "epoch": 0.05131172839506173,
      "grad_norm": 3.355565309524536,
      "learning_rate": 0.0005,
      "loss": 3.6164164543151855,
      "step": 266
    },
    {
      "epoch": 0.05150462962962963,
      "grad_norm": 1.295634150505066,
      "learning_rate": 0.0005,
      "loss": 3.521669387817383,
      "step": 267
    },
    {
      "epoch": 0.05169753086419753,
      "grad_norm": 1.0840654373168945,
      "learning_rate": 0.0005,
      "loss": 3.7489254474639893,
      "step": 268
    },
    {
      "epoch": 0.051890432098765434,
      "grad_norm": 2.3061037063598633,
      "learning_rate": 0.0005,
      "loss": 3.436934232711792,
      "step": 269
    },
    {
      "epoch": 0.052083333333333336,
      "grad_norm": 1.630590796470642,
      "learning_rate": 0.0005,
      "loss": 3.588322639465332,
      "step": 270
    },
    {
      "epoch": 0.05227623456790124,
      "grad_norm": 1.4419217109680176,
      "learning_rate": 0.0005,
      "loss": 3.5664124488830566,
      "step": 271
    },
    {
      "epoch": 0.05246913580246913,
      "grad_norm": 1.1665385961532593,
      "learning_rate": 0.0005,
      "loss": 3.7692313194274902,
      "step": 272
    },
    {
      "epoch": 0.052662037037037035,
      "grad_norm": 1.1241357326507568,
      "learning_rate": 0.0005,
      "loss": 3.856379508972168,
      "step": 273
    },
    {
      "epoch": 0.05285493827160494,
      "grad_norm": 1.6094253063201904,
      "learning_rate": 0.0005,
      "loss": 3.336848735809326,
      "step": 274
    },
    {
      "epoch": 0.05304783950617284,
      "grad_norm": 1.4099920988082886,
      "learning_rate": 0.0005,
      "loss": 3.7986159324645996,
      "step": 275
    },
    {
      "epoch": 0.05324074074074074,
      "grad_norm": 1.1790531873703003,
      "learning_rate": 0.0005,
      "loss": 3.7861688137054443,
      "step": 276
    },
    {
      "epoch": 0.05343364197530864,
      "grad_norm": 1.210378885269165,
      "learning_rate": 0.0005,
      "loss": 3.5102639198303223,
      "step": 277
    },
    {
      "epoch": 0.053626543209876545,
      "grad_norm": 1.2109023332595825,
      "learning_rate": 0.0005,
      "loss": 3.815894842147827,
      "step": 278
    },
    {
      "epoch": 0.05381944444444445,
      "grad_norm": 1.3940558433532715,
      "learning_rate": 0.0005,
      "loss": 3.8155126571655273,
      "step": 279
    },
    {
      "epoch": 0.05401234567901234,
      "grad_norm": 2.823547601699829,
      "learning_rate": 0.0005,
      "loss": 3.5634384155273438,
      "step": 280
    },
    {
      "epoch": 0.054205246913580245,
      "grad_norm": 1.3719772100448608,
      "learning_rate": 0.0005,
      "loss": 3.7513580322265625,
      "step": 281
    },
    {
      "epoch": 0.05439814814814815,
      "grad_norm": 1.1145737171173096,
      "learning_rate": 0.0005,
      "loss": 3.53334641456604,
      "step": 282
    },
    {
      "epoch": 0.05459104938271605,
      "grad_norm": 1.3421119451522827,
      "learning_rate": 0.0005,
      "loss": 3.6254513263702393,
      "step": 283
    },
    {
      "epoch": 0.05478395061728395,
      "grad_norm": 1.0853593349456787,
      "learning_rate": 0.0005,
      "loss": 3.5748538970947266,
      "step": 284
    },
    {
      "epoch": 0.05497685185185185,
      "grad_norm": 1.133806824684143,
      "learning_rate": 0.0005,
      "loss": 3.602917194366455,
      "step": 285
    },
    {
      "epoch": 0.055169753086419755,
      "grad_norm": 1.2822962999343872,
      "learning_rate": 0.0005,
      "loss": 3.3496556282043457,
      "step": 286
    },
    {
      "epoch": 0.05536265432098766,
      "grad_norm": 1.7264524698257446,
      "learning_rate": 0.0005,
      "loss": 3.467275857925415,
      "step": 287
    },
    {
      "epoch": 0.05555555555555555,
      "grad_norm": 1.3050892353057861,
      "learning_rate": 0.0005,
      "loss": 3.488541841506958,
      "step": 288
    },
    {
      "epoch": 0.055748456790123455,
      "grad_norm": 1.5174587965011597,
      "learning_rate": 0.0005,
      "loss": 4.012282848358154,
      "step": 289
    },
    {
      "epoch": 0.05594135802469136,
      "grad_norm": 1.370285153388977,
      "learning_rate": 0.0005,
      "loss": 3.849055051803589,
      "step": 290
    },
    {
      "epoch": 0.05613425925925926,
      "grad_norm": 1.6686738729476929,
      "learning_rate": 0.0005,
      "loss": 3.7102770805358887,
      "step": 291
    },
    {
      "epoch": 0.05632716049382716,
      "grad_norm": 1.5103343725204468,
      "learning_rate": 0.0005,
      "loss": 3.433593273162842,
      "step": 292
    },
    {
      "epoch": 0.05652006172839506,
      "grad_norm": 1.1966673135757446,
      "learning_rate": 0.0005,
      "loss": 3.6435389518737793,
      "step": 293
    },
    {
      "epoch": 0.056712962962962965,
      "grad_norm": 1.6730520725250244,
      "learning_rate": 0.0005,
      "loss": 3.3167364597320557,
      "step": 294
    },
    {
      "epoch": 0.05690586419753087,
      "grad_norm": 1.1460531949996948,
      "learning_rate": 0.0005,
      "loss": 3.8354763984680176,
      "step": 295
    },
    {
      "epoch": 0.05709876543209876,
      "grad_norm": 1.1782690286636353,
      "learning_rate": 0.0005,
      "loss": 3.6417055130004883,
      "step": 296
    },
    {
      "epoch": 0.057291666666666664,
      "grad_norm": 1.281988263130188,
      "learning_rate": 0.0005,
      "loss": 3.779374122619629,
      "step": 297
    },
    {
      "epoch": 0.057484567901234566,
      "grad_norm": 1.2190115451812744,
      "learning_rate": 0.0005,
      "loss": 3.642538070678711,
      "step": 298
    },
    {
      "epoch": 0.05767746913580247,
      "grad_norm": 1.5735485553741455,
      "learning_rate": 0.0005,
      "loss": 3.674450159072876,
      "step": 299
    },
    {
      "epoch": 0.05787037037037037,
      "grad_norm": 1.4354535341262817,
      "learning_rate": 0.0005,
      "loss": 3.6060240268707275,
      "step": 300
    },
    {
      "epoch": 0.05806327160493827,
      "grad_norm": 1.952643871307373,
      "learning_rate": 0.0005,
      "loss": 3.648287534713745,
      "step": 301
    },
    {
      "epoch": 0.058256172839506175,
      "grad_norm": 1.3646862506866455,
      "learning_rate": 0.0005,
      "loss": 3.8044025897979736,
      "step": 302
    },
    {
      "epoch": 0.05844907407407408,
      "grad_norm": 1.647530198097229,
      "learning_rate": 0.0005,
      "loss": 3.286513328552246,
      "step": 303
    },
    {
      "epoch": 0.05864197530864197,
      "grad_norm": 1.1475366353988647,
      "learning_rate": 0.0005,
      "loss": 3.689563751220703,
      "step": 304
    },
    {
      "epoch": 0.058834876543209874,
      "grad_norm": 1.3954098224639893,
      "learning_rate": 0.0005,
      "loss": 3.4699206352233887,
      "step": 305
    },
    {
      "epoch": 0.059027777777777776,
      "grad_norm": 1.340537667274475,
      "learning_rate": 0.0005,
      "loss": 3.759096622467041,
      "step": 306
    },
    {
      "epoch": 0.05922067901234568,
      "grad_norm": 1.842153549194336,
      "learning_rate": 0.0005,
      "loss": 3.8941378593444824,
      "step": 307
    },
    {
      "epoch": 0.05941358024691358,
      "grad_norm": 1.1906794309616089,
      "learning_rate": 0.0005,
      "loss": 3.4608302116394043,
      "step": 308
    },
    {
      "epoch": 0.05960648148148148,
      "grad_norm": 1.2779508829116821,
      "learning_rate": 0.0005,
      "loss": 3.6788060665130615,
      "step": 309
    },
    {
      "epoch": 0.059799382716049385,
      "grad_norm": 1.2016125917434692,
      "learning_rate": 0.0005,
      "loss": 3.539249897003174,
      "step": 310
    },
    {
      "epoch": 0.05999228395061729,
      "grad_norm": 1.3244026899337769,
      "learning_rate": 0.0005,
      "loss": 3.428591728210449,
      "step": 311
    },
    {
      "epoch": 0.06018518518518518,
      "grad_norm": 2.3958308696746826,
      "learning_rate": 0.0005,
      "loss": 3.8364548683166504,
      "step": 312
    },
    {
      "epoch": 0.060378086419753084,
      "grad_norm": 1.210676908493042,
      "learning_rate": 0.0005,
      "loss": 3.5015480518341064,
      "step": 313
    },
    {
      "epoch": 0.060570987654320986,
      "grad_norm": 1.182102918624878,
      "learning_rate": 0.0005,
      "loss": 3.444319725036621,
      "step": 314
    },
    {
      "epoch": 0.06076388888888889,
      "grad_norm": 1.534242033958435,
      "learning_rate": 0.0005,
      "loss": 3.8179917335510254,
      "step": 315
    },
    {
      "epoch": 0.06095679012345679,
      "grad_norm": 1.136762022972107,
      "learning_rate": 0.0005,
      "loss": 3.529285192489624,
      "step": 316
    },
    {
      "epoch": 0.06114969135802469,
      "grad_norm": 1.0861124992370605,
      "learning_rate": 0.0005,
      "loss": 3.518187999725342,
      "step": 317
    },
    {
      "epoch": 0.061342592592592594,
      "grad_norm": 1.1806056499481201,
      "learning_rate": 0.0005,
      "loss": 3.7639527320861816,
      "step": 318
    },
    {
      "epoch": 0.061535493827160496,
      "grad_norm": 1.5713346004486084,
      "learning_rate": 0.0005,
      "loss": 3.512192726135254,
      "step": 319
    },
    {
      "epoch": 0.06172839506172839,
      "grad_norm": 1.380902647972107,
      "learning_rate": 0.0005,
      "loss": 3.7858591079711914,
      "step": 320
    },
    {
      "epoch": 0.061921296296296294,
      "grad_norm": 1.4135419130325317,
      "learning_rate": 0.0005,
      "loss": 3.8143820762634277,
      "step": 321
    },
    {
      "epoch": 0.062114197530864196,
      "grad_norm": 1.4018239974975586,
      "learning_rate": 0.0005,
      "loss": 3.639910936355591,
      "step": 322
    },
    {
      "epoch": 0.0623070987654321,
      "grad_norm": 1.088641881942749,
      "learning_rate": 0.0005,
      "loss": 3.79715633392334,
      "step": 323
    },
    {
      "epoch": 0.0625,
      "grad_norm": 1.885367512702942,
      "learning_rate": 0.0005,
      "loss": 4.003077030181885,
      "step": 324
    },
    {
      "epoch": 0.0626929012345679,
      "grad_norm": 1.371606469154358,
      "learning_rate": 0.0005,
      "loss": 3.7730166912078857,
      "step": 325
    },
    {
      "epoch": 0.0628858024691358,
      "grad_norm": 1.3741323947906494,
      "learning_rate": 0.0005,
      "loss": 3.64087176322937,
      "step": 326
    },
    {
      "epoch": 0.0630787037037037,
      "grad_norm": 1.1313707828521729,
      "learning_rate": 0.0005,
      "loss": 3.7846784591674805,
      "step": 327
    },
    {
      "epoch": 0.06327160493827161,
      "grad_norm": 1.3197203874588013,
      "learning_rate": 0.0005,
      "loss": 3.564077854156494,
      "step": 328
    },
    {
      "epoch": 0.06346450617283951,
      "grad_norm": 1.3068927526474,
      "learning_rate": 0.0005,
      "loss": 3.635511636734009,
      "step": 329
    },
    {
      "epoch": 0.06365740740740741,
      "grad_norm": 1.0828635692596436,
      "learning_rate": 0.0005,
      "loss": 4.038288116455078,
      "step": 330
    },
    {
      "epoch": 0.06385030864197531,
      "grad_norm": 1.4297481775283813,
      "learning_rate": 0.0005,
      "loss": 3.525176525115967,
      "step": 331
    },
    {
      "epoch": 0.06404320987654322,
      "grad_norm": 1.6865818500518799,
      "learning_rate": 0.0005,
      "loss": 3.5704169273376465,
      "step": 332
    },
    {
      "epoch": 0.0642361111111111,
      "grad_norm": 1.0616540908813477,
      "learning_rate": 0.0005,
      "loss": 4.00485372543335,
      "step": 333
    },
    {
      "epoch": 0.06442901234567901,
      "grad_norm": 1.3850042819976807,
      "learning_rate": 0.0005,
      "loss": 3.562979221343994,
      "step": 334
    },
    {
      "epoch": 0.06462191358024691,
      "grad_norm": 1.8505020141601562,
      "learning_rate": 0.0005,
      "loss": 3.663079023361206,
      "step": 335
    },
    {
      "epoch": 0.06481481481481481,
      "grad_norm": 1.298318862915039,
      "learning_rate": 0.0005,
      "loss": 3.7435245513916016,
      "step": 336
    },
    {
      "epoch": 0.06500771604938271,
      "grad_norm": 1.302327275276184,
      "learning_rate": 0.0005,
      "loss": 3.4227442741394043,
      "step": 337
    },
    {
      "epoch": 0.06520061728395062,
      "grad_norm": 1.3631285429000854,
      "learning_rate": 0.0005,
      "loss": 3.7644872665405273,
      "step": 338
    },
    {
      "epoch": 0.06539351851851852,
      "grad_norm": 1.1951576471328735,
      "learning_rate": 0.0005,
      "loss": 3.791361093521118,
      "step": 339
    },
    {
      "epoch": 0.06558641975308642,
      "grad_norm": 1.010726809501648,
      "learning_rate": 0.0005,
      "loss": 3.806070566177368,
      "step": 340
    },
    {
      "epoch": 0.06577932098765432,
      "grad_norm": 1.958269476890564,
      "learning_rate": 0.0005,
      "loss": 3.659554958343506,
      "step": 341
    },
    {
      "epoch": 0.06597222222222222,
      "grad_norm": 2.1599576473236084,
      "learning_rate": 0.0005,
      "loss": 3.777406692504883,
      "step": 342
    },
    {
      "epoch": 0.06616512345679013,
      "grad_norm": 1.2975066900253296,
      "learning_rate": 0.0005,
      "loss": 3.6287641525268555,
      "step": 343
    },
    {
      "epoch": 0.06635802469135803,
      "grad_norm": 1.5331193208694458,
      "learning_rate": 0.0005,
      "loss": 3.9187703132629395,
      "step": 344
    },
    {
      "epoch": 0.06655092592592593,
      "grad_norm": 1.2246979475021362,
      "learning_rate": 0.0005,
      "loss": 3.0362257957458496,
      "step": 345
    },
    {
      "epoch": 0.06674382716049383,
      "grad_norm": 1.0962567329406738,
      "learning_rate": 0.0005,
      "loss": 3.6050612926483154,
      "step": 346
    },
    {
      "epoch": 0.06693672839506173,
      "grad_norm": 1.3687812089920044,
      "learning_rate": 0.0005,
      "loss": 3.8140523433685303,
      "step": 347
    },
    {
      "epoch": 0.06712962962962964,
      "grad_norm": 1.0036567449569702,
      "learning_rate": 0.0005,
      "loss": 3.622077226638794,
      "step": 348
    },
    {
      "epoch": 0.06732253086419752,
      "grad_norm": 1.2300020456314087,
      "learning_rate": 0.0005,
      "loss": 3.8575267791748047,
      "step": 349
    },
    {
      "epoch": 0.06751543209876543,
      "grad_norm": 1.385164737701416,
      "learning_rate": 0.0005,
      "loss": 3.6513800621032715,
      "step": 350
    },
    {
      "epoch": 0.06770833333333333,
      "grad_norm": 1.1817576885223389,
      "learning_rate": 0.0005,
      "loss": 3.4677071571350098,
      "step": 351
    },
    {
      "epoch": 0.06790123456790123,
      "grad_norm": 1.2153571844100952,
      "learning_rate": 0.0005,
      "loss": 4.049473762512207,
      "step": 352
    },
    {
      "epoch": 0.06809413580246913,
      "grad_norm": 1.252871036529541,
      "learning_rate": 0.0005,
      "loss": 3.434054374694824,
      "step": 353
    },
    {
      "epoch": 0.06828703703703703,
      "grad_norm": 1.2587910890579224,
      "learning_rate": 0.0005,
      "loss": 3.6553595066070557,
      "step": 354
    },
    {
      "epoch": 0.06847993827160494,
      "grad_norm": 1.7073911428451538,
      "learning_rate": 0.0005,
      "loss": 3.9832992553710938,
      "step": 355
    },
    {
      "epoch": 0.06867283950617284,
      "grad_norm": 1.148916482925415,
      "learning_rate": 0.0005,
      "loss": 3.7655279636383057,
      "step": 356
    },
    {
      "epoch": 0.06886574074074074,
      "grad_norm": 1.1480268239974976,
      "learning_rate": 0.0005,
      "loss": 3.5605342388153076,
      "step": 357
    },
    {
      "epoch": 0.06905864197530864,
      "grad_norm": 1.3598134517669678,
      "learning_rate": 0.0005,
      "loss": 3.4493765830993652,
      "step": 358
    },
    {
      "epoch": 0.06925154320987655,
      "grad_norm": 1.0921034812927246,
      "learning_rate": 0.0005,
      "loss": 3.3882791996002197,
      "step": 359
    },
    {
      "epoch": 0.06944444444444445,
      "grad_norm": 1.73841392993927,
      "learning_rate": 0.0005,
      "loss": 3.850335121154785,
      "step": 360
    },
    {
      "epoch": 0.06963734567901235,
      "grad_norm": 1.0930064916610718,
      "learning_rate": 0.0005,
      "loss": 3.600897789001465,
      "step": 361
    },
    {
      "epoch": 0.06983024691358025,
      "grad_norm": 1.2556062936782837,
      "learning_rate": 0.0005,
      "loss": 3.6028928756713867,
      "step": 362
    },
    {
      "epoch": 0.07002314814814815,
      "grad_norm": 1.244928240776062,
      "learning_rate": 0.0005,
      "loss": 3.6133041381835938,
      "step": 363
    },
    {
      "epoch": 0.07021604938271606,
      "grad_norm": 1.2315469980239868,
      "learning_rate": 0.0005,
      "loss": 3.6872007846832275,
      "step": 364
    },
    {
      "epoch": 0.07040895061728394,
      "grad_norm": 1.2658771276474,
      "learning_rate": 0.0005,
      "loss": 3.6061387062072754,
      "step": 365
    },
    {
      "epoch": 0.07060185185185185,
      "grad_norm": 1.5925453901290894,
      "learning_rate": 0.0005,
      "loss": 4.019912242889404,
      "step": 366
    },
    {
      "epoch": 0.07079475308641975,
      "grad_norm": 1.2904589176177979,
      "learning_rate": 0.0005,
      "loss": 3.6272201538085938,
      "step": 367
    },
    {
      "epoch": 0.07098765432098765,
      "grad_norm": 1.113761305809021,
      "learning_rate": 0.0005,
      "loss": 3.7014451026916504,
      "step": 368
    },
    {
      "epoch": 0.07118055555555555,
      "grad_norm": 1.172806978225708,
      "learning_rate": 0.0005,
      "loss": 3.520353317260742,
      "step": 369
    },
    {
      "epoch": 0.07137345679012345,
      "grad_norm": 1.4440264701843262,
      "learning_rate": 0.0005,
      "loss": 3.5321383476257324,
      "step": 370
    },
    {
      "epoch": 0.07156635802469136,
      "grad_norm": 1.18381929397583,
      "learning_rate": 0.0005,
      "loss": 3.548255443572998,
      "step": 371
    },
    {
      "epoch": 0.07175925925925926,
      "grad_norm": 1.2736179828643799,
      "learning_rate": 0.0005,
      "loss": 3.4956541061401367,
      "step": 372
    },
    {
      "epoch": 0.07195216049382716,
      "grad_norm": 1.0780549049377441,
      "learning_rate": 0.0005,
      "loss": 3.6641712188720703,
      "step": 373
    },
    {
      "epoch": 0.07214506172839506,
      "grad_norm": 1.2545503377914429,
      "learning_rate": 0.0005,
      "loss": 3.6152150630950928,
      "step": 374
    },
    {
      "epoch": 0.07233796296296297,
      "grad_norm": 1.3015769720077515,
      "learning_rate": 0.0005,
      "loss": 3.845883846282959,
      "step": 375
    },
    {
      "epoch": 0.07253086419753087,
      "grad_norm": 1.3542593717575073,
      "learning_rate": 0.0005,
      "loss": 3.60617995262146,
      "step": 376
    },
    {
      "epoch": 0.07272376543209877,
      "grad_norm": 1.147844672203064,
      "learning_rate": 0.0005,
      "loss": 3.838994026184082,
      "step": 377
    },
    {
      "epoch": 0.07291666666666667,
      "grad_norm": 1.4651304483413696,
      "learning_rate": 0.0005,
      "loss": 3.983508825302124,
      "step": 378
    },
    {
      "epoch": 0.07310956790123457,
      "grad_norm": 1.3039181232452393,
      "learning_rate": 0.0005,
      "loss": 3.7356410026550293,
      "step": 379
    },
    {
      "epoch": 0.07330246913580248,
      "grad_norm": 1.750315546989441,
      "learning_rate": 0.0005,
      "loss": 3.9725418090820312,
      "step": 380
    },
    {
      "epoch": 0.07349537037037036,
      "grad_norm": 1.1458873748779297,
      "learning_rate": 0.0005,
      "loss": 3.824273109436035,
      "step": 381
    },
    {
      "epoch": 0.07368827160493827,
      "grad_norm": 1.3059637546539307,
      "learning_rate": 0.0005,
      "loss": 3.8040266036987305,
      "step": 382
    },
    {
      "epoch": 0.07388117283950617,
      "grad_norm": 1.2448099851608276,
      "learning_rate": 0.0005,
      "loss": 3.5719547271728516,
      "step": 383
    },
    {
      "epoch": 0.07407407407407407,
      "grad_norm": 1.1003605127334595,
      "learning_rate": 0.0005,
      "loss": 3.7712299823760986,
      "step": 384
    },
    {
      "epoch": 0.07426697530864197,
      "grad_norm": 1.5202761888504028,
      "learning_rate": 0.0005,
      "loss": 3.687675952911377,
      "step": 385
    },
    {
      "epoch": 0.07445987654320987,
      "grad_norm": 1.2451661825180054,
      "learning_rate": 0.0005,
      "loss": 3.530005931854248,
      "step": 386
    },
    {
      "epoch": 0.07465277777777778,
      "grad_norm": 1.8608239889144897,
      "learning_rate": 0.0005,
      "loss": 3.661074638366699,
      "step": 387
    },
    {
      "epoch": 0.07484567901234568,
      "grad_norm": 1.3177058696746826,
      "learning_rate": 0.0005,
      "loss": 3.7760682106018066,
      "step": 388
    },
    {
      "epoch": 0.07503858024691358,
      "grad_norm": 1.4145710468292236,
      "learning_rate": 0.0005,
      "loss": 3.532954216003418,
      "step": 389
    },
    {
      "epoch": 0.07523148148148148,
      "grad_norm": 1.2716408967971802,
      "learning_rate": 0.0005,
      "loss": 3.6177005767822266,
      "step": 390
    },
    {
      "epoch": 0.07542438271604938,
      "grad_norm": 1.352758765220642,
      "learning_rate": 0.0005,
      "loss": 3.565272092819214,
      "step": 391
    },
    {
      "epoch": 0.07561728395061729,
      "grad_norm": 1.3513282537460327,
      "learning_rate": 0.0005,
      "loss": 3.515331745147705,
      "step": 392
    },
    {
      "epoch": 0.07581018518518519,
      "grad_norm": 1.3268440961837769,
      "learning_rate": 0.0005,
      "loss": 3.482893943786621,
      "step": 393
    },
    {
      "epoch": 0.07600308641975309,
      "grad_norm": 0.9908361434936523,
      "learning_rate": 0.0005,
      "loss": 3.734281539916992,
      "step": 394
    },
    {
      "epoch": 0.07619598765432099,
      "grad_norm": 1.3419976234436035,
      "learning_rate": 0.0005,
      "loss": 3.303842306137085,
      "step": 395
    },
    {
      "epoch": 0.0763888888888889,
      "grad_norm": 1.2391479015350342,
      "learning_rate": 0.0005,
      "loss": 3.433652877807617,
      "step": 396
    },
    {
      "epoch": 0.07658179012345678,
      "grad_norm": 1.1078441143035889,
      "learning_rate": 0.0005,
      "loss": 3.5523505210876465,
      "step": 397
    },
    {
      "epoch": 0.07677469135802469,
      "grad_norm": 1.1694769859313965,
      "learning_rate": 0.0005,
      "loss": 3.566941022872925,
      "step": 398
    },
    {
      "epoch": 0.07696759259259259,
      "grad_norm": 1.2622727155685425,
      "learning_rate": 0.0005,
      "loss": 3.44706392288208,
      "step": 399
    },
    {
      "epoch": 0.07716049382716049,
      "grad_norm": 1.6326581239700317,
      "learning_rate": 0.0005,
      "loss": 3.4932618141174316,
      "step": 400
    },
    {
      "epoch": 0.07735339506172839,
      "grad_norm": 1.922366976737976,
      "learning_rate": 0.0005,
      "loss": 3.7034201622009277,
      "step": 401
    },
    {
      "epoch": 0.0775462962962963,
      "grad_norm": 1.6329044103622437,
      "learning_rate": 0.0005,
      "loss": 3.77363920211792,
      "step": 402
    },
    {
      "epoch": 0.0777391975308642,
      "grad_norm": 1.2314785718917847,
      "learning_rate": 0.0005,
      "loss": 3.749850273132324,
      "step": 403
    },
    {
      "epoch": 0.0779320987654321,
      "grad_norm": 1.5868536233901978,
      "learning_rate": 0.0005,
      "loss": 3.930659770965576,
      "step": 404
    },
    {
      "epoch": 0.078125,
      "grad_norm": 1.546505331993103,
      "learning_rate": 0.0005,
      "loss": 3.6041717529296875,
      "step": 405
    },
    {
      "epoch": 0.0783179012345679,
      "grad_norm": 1.1810640096664429,
      "learning_rate": 0.0005,
      "loss": 3.5450971126556396,
      "step": 406
    },
    {
      "epoch": 0.0785108024691358,
      "grad_norm": 1.196488857269287,
      "learning_rate": 0.0005,
      "loss": 3.582494020462036,
      "step": 407
    },
    {
      "epoch": 0.0787037037037037,
      "grad_norm": 1.106093406677246,
      "learning_rate": 0.0005,
      "loss": 3.6947598457336426,
      "step": 408
    },
    {
      "epoch": 0.07889660493827161,
      "grad_norm": 1.3387807607650757,
      "learning_rate": 0.0005,
      "loss": 3.702266216278076,
      "step": 409
    },
    {
      "epoch": 0.07908950617283951,
      "grad_norm": 1.205798625946045,
      "learning_rate": 0.0005,
      "loss": 3.4115641117095947,
      "step": 410
    },
    {
      "epoch": 0.07928240740740741,
      "grad_norm": 1.3406968116760254,
      "learning_rate": 0.0005,
      "loss": 3.4552838802337646,
      "step": 411
    },
    {
      "epoch": 0.07947530864197531,
      "grad_norm": 1.2239340543746948,
      "learning_rate": 0.0005,
      "loss": 3.495303153991699,
      "step": 412
    },
    {
      "epoch": 0.07966820987654322,
      "grad_norm": 1.0862449407577515,
      "learning_rate": 0.0005,
      "loss": 3.6800272464752197,
      "step": 413
    },
    {
      "epoch": 0.0798611111111111,
      "grad_norm": 1.2526425123214722,
      "learning_rate": 0.0005,
      "loss": 3.440629243850708,
      "step": 414
    },
    {
      "epoch": 0.08005401234567901,
      "grad_norm": 1.2104847431182861,
      "learning_rate": 0.0005,
      "loss": 3.6726417541503906,
      "step": 415
    },
    {
      "epoch": 0.08024691358024691,
      "grad_norm": 1.1145989894866943,
      "learning_rate": 0.0005,
      "loss": 3.8565330505371094,
      "step": 416
    },
    {
      "epoch": 0.08043981481481481,
      "grad_norm": 1.3068956136703491,
      "learning_rate": 0.0005,
      "loss": 3.231868267059326,
      "step": 417
    },
    {
      "epoch": 0.08063271604938271,
      "grad_norm": 1.3861457109451294,
      "learning_rate": 0.0005,
      "loss": 3.5560927391052246,
      "step": 418
    },
    {
      "epoch": 0.08082561728395062,
      "grad_norm": 1.6336735486984253,
      "learning_rate": 0.0005,
      "loss": 4.009586334228516,
      "step": 419
    },
    {
      "epoch": 0.08101851851851852,
      "grad_norm": 1.4437741041183472,
      "learning_rate": 0.0005,
      "loss": 3.828901529312134,
      "step": 420
    },
    {
      "epoch": 0.08121141975308642,
      "grad_norm": 1.2701531648635864,
      "learning_rate": 0.0005,
      "loss": 3.3267030715942383,
      "step": 421
    },
    {
      "epoch": 0.08140432098765432,
      "grad_norm": 1.3418720960617065,
      "learning_rate": 0.0005,
      "loss": 3.593108654022217,
      "step": 422
    },
    {
      "epoch": 0.08159722222222222,
      "grad_norm": 1.0967100858688354,
      "learning_rate": 0.0005,
      "loss": 3.5215773582458496,
      "step": 423
    },
    {
      "epoch": 0.08179012345679013,
      "grad_norm": 1.1680047512054443,
      "learning_rate": 0.0005,
      "loss": 3.328854560852051,
      "step": 424
    },
    {
      "epoch": 0.08198302469135803,
      "grad_norm": 1.2096976041793823,
      "learning_rate": 0.0005,
      "loss": 3.1990325450897217,
      "step": 425
    },
    {
      "epoch": 0.08217592592592593,
      "grad_norm": 1.1918598413467407,
      "learning_rate": 0.0005,
      "loss": 3.8099169731140137,
      "step": 426
    },
    {
      "epoch": 0.08236882716049383,
      "grad_norm": 1.2880069017410278,
      "learning_rate": 0.0005,
      "loss": 3.573268175125122,
      "step": 427
    },
    {
      "epoch": 0.08256172839506173,
      "grad_norm": 1.2320311069488525,
      "learning_rate": 0.0005,
      "loss": 3.683140277862549,
      "step": 428
    },
    {
      "epoch": 0.08275462962962964,
      "grad_norm": 1.3513623476028442,
      "learning_rate": 0.0005,
      "loss": 3.7024309635162354,
      "step": 429
    },
    {
      "epoch": 0.08294753086419752,
      "grad_norm": 1.2702847719192505,
      "learning_rate": 0.0005,
      "loss": 3.6479058265686035,
      "step": 430
    },
    {
      "epoch": 0.08314043209876543,
      "grad_norm": 1.3983005285263062,
      "learning_rate": 0.0005,
      "loss": 3.6913976669311523,
      "step": 431
    },
    {
      "epoch": 0.08333333333333333,
      "grad_norm": 1.373806357383728,
      "learning_rate": 0.0005,
      "loss": 3.3543829917907715,
      "step": 432
    },
    {
      "epoch": 0.08352623456790123,
      "grad_norm": 1.5676419734954834,
      "learning_rate": 0.0005,
      "loss": 3.9601101875305176,
      "step": 433
    },
    {
      "epoch": 0.08371913580246913,
      "grad_norm": 1.6062133312225342,
      "learning_rate": 0.0005,
      "loss": 3.67623233795166,
      "step": 434
    },
    {
      "epoch": 0.08391203703703703,
      "grad_norm": 1.4886835813522339,
      "learning_rate": 0.0005,
      "loss": 3.8306024074554443,
      "step": 435
    },
    {
      "epoch": 0.08410493827160494,
      "grad_norm": 1.1453768014907837,
      "learning_rate": 0.0005,
      "loss": 3.4680299758911133,
      "step": 436
    },
    {
      "epoch": 0.08429783950617284,
      "grad_norm": 1.5303137302398682,
      "learning_rate": 0.0005,
      "loss": 3.80029034614563,
      "step": 437
    },
    {
      "epoch": 0.08449074074074074,
      "grad_norm": 1.0666272640228271,
      "learning_rate": 0.0005,
      "loss": 3.736417531967163,
      "step": 438
    },
    {
      "epoch": 0.08468364197530864,
      "grad_norm": 1.2510230541229248,
      "learning_rate": 0.0005,
      "loss": 3.1824817657470703,
      "step": 439
    },
    {
      "epoch": 0.08487654320987655,
      "grad_norm": 1.907281517982483,
      "learning_rate": 0.0005,
      "loss": 3.3208394050598145,
      "step": 440
    },
    {
      "epoch": 0.08506944444444445,
      "grad_norm": 1.344112753868103,
      "learning_rate": 0.0005,
      "loss": 3.675384998321533,
      "step": 441
    },
    {
      "epoch": 0.08526234567901235,
      "grad_norm": 1.2153356075286865,
      "learning_rate": 0.0005,
      "loss": 3.2713046073913574,
      "step": 442
    },
    {
      "epoch": 0.08545524691358025,
      "grad_norm": 1.116800308227539,
      "learning_rate": 0.0005,
      "loss": 3.8372325897216797,
      "step": 443
    },
    {
      "epoch": 0.08564814814814815,
      "grad_norm": 1.3750286102294922,
      "learning_rate": 0.0005,
      "loss": 3.6183605194091797,
      "step": 444
    },
    {
      "epoch": 0.08584104938271606,
      "grad_norm": 1.5159918069839478,
      "learning_rate": 0.0005,
      "loss": 3.196272134780884,
      "step": 445
    },
    {
      "epoch": 0.08603395061728394,
      "grad_norm": 1.7818692922592163,
      "learning_rate": 0.0005,
      "loss": 3.455862522125244,
      "step": 446
    },
    {
      "epoch": 0.08622685185185185,
      "grad_norm": 1.3619201183319092,
      "learning_rate": 0.0005,
      "loss": 3.4987359046936035,
      "step": 447
    },
    {
      "epoch": 0.08641975308641975,
      "grad_norm": 1.2619519233703613,
      "learning_rate": 0.0005,
      "loss": 3.531465530395508,
      "step": 448
    },
    {
      "epoch": 0.08661265432098765,
      "grad_norm": 0.9870778322219849,
      "learning_rate": 0.0005,
      "loss": 3.6784090995788574,
      "step": 449
    },
    {
      "epoch": 0.08680555555555555,
      "grad_norm": 1.5232467651367188,
      "learning_rate": 0.0005,
      "loss": 3.3810768127441406,
      "step": 450
    },
    {
      "epoch": 0.08699845679012345,
      "grad_norm": 1.9541943073272705,
      "learning_rate": 0.0005,
      "loss": 3.7687792778015137,
      "step": 451
    },
    {
      "epoch": 0.08719135802469136,
      "grad_norm": 1.4750969409942627,
      "learning_rate": 0.0005,
      "loss": 3.4738216400146484,
      "step": 452
    },
    {
      "epoch": 0.08738425925925926,
      "grad_norm": 1.1606881618499756,
      "learning_rate": 0.0005,
      "loss": 3.604809522628784,
      "step": 453
    },
    {
      "epoch": 0.08757716049382716,
      "grad_norm": 1.1727041006088257,
      "learning_rate": 0.0005,
      "loss": 3.383584976196289,
      "step": 454
    },
    {
      "epoch": 0.08777006172839506,
      "grad_norm": 1.1224572658538818,
      "learning_rate": 0.0005,
      "loss": 3.581493377685547,
      "step": 455
    },
    {
      "epoch": 0.08796296296296297,
      "grad_norm": 3.6764914989471436,
      "learning_rate": 0.0005,
      "loss": 3.821187973022461,
      "step": 456
    },
    {
      "epoch": 0.08815586419753087,
      "grad_norm": 1.5353116989135742,
      "learning_rate": 0.0005,
      "loss": 3.4004440307617188,
      "step": 457
    },
    {
      "epoch": 0.08834876543209877,
      "grad_norm": 1.332394003868103,
      "learning_rate": 0.0005,
      "loss": 3.44679594039917,
      "step": 458
    },
    {
      "epoch": 0.08854166666666667,
      "grad_norm": 1.1422858238220215,
      "learning_rate": 0.0005,
      "loss": 3.63377046585083,
      "step": 459
    },
    {
      "epoch": 0.08873456790123457,
      "grad_norm": 1.1114091873168945,
      "learning_rate": 0.0005,
      "loss": 3.6705570220947266,
      "step": 460
    },
    {
      "epoch": 0.08892746913580248,
      "grad_norm": 1.3809031248092651,
      "learning_rate": 0.0005,
      "loss": 3.491145610809326,
      "step": 461
    },
    {
      "epoch": 0.08912037037037036,
      "grad_norm": 1.3103653192520142,
      "learning_rate": 0.0005,
      "loss": 3.361976146697998,
      "step": 462
    },
    {
      "epoch": 0.08931327160493827,
      "grad_norm": 1.5138816833496094,
      "learning_rate": 0.0005,
      "loss": 3.332667350769043,
      "step": 463
    },
    {
      "epoch": 0.08950617283950617,
      "grad_norm": 1.5852179527282715,
      "learning_rate": 0.0005,
      "loss": 3.657076835632324,
      "step": 464
    },
    {
      "epoch": 0.08969907407407407,
      "grad_norm": 1.3589686155319214,
      "learning_rate": 0.0005,
      "loss": 3.412259101867676,
      "step": 465
    },
    {
      "epoch": 0.08989197530864197,
      "grad_norm": 1.2723749876022339,
      "learning_rate": 0.0005,
      "loss": 3.477722644805908,
      "step": 466
    },
    {
      "epoch": 0.09008487654320987,
      "grad_norm": 1.1056469678878784,
      "learning_rate": 0.0005,
      "loss": 3.556992292404175,
      "step": 467
    },
    {
      "epoch": 0.09027777777777778,
      "grad_norm": 1.0250502824783325,
      "learning_rate": 0.0005,
      "loss": 3.70348858833313,
      "step": 468
    },
    {
      "epoch": 0.09047067901234568,
      "grad_norm": 1.102981448173523,
      "learning_rate": 0.0005,
      "loss": 3.544962167739868,
      "step": 469
    },
    {
      "epoch": 0.09066358024691358,
      "grad_norm": 1.2398360967636108,
      "learning_rate": 0.0005,
      "loss": 3.579246759414673,
      "step": 470
    },
    {
      "epoch": 0.09085648148148148,
      "grad_norm": 1.3386133909225464,
      "learning_rate": 0.0005,
      "loss": 3.2664365768432617,
      "step": 471
    },
    {
      "epoch": 0.09104938271604938,
      "grad_norm": 1.3492672443389893,
      "learning_rate": 0.0005,
      "loss": 3.481532335281372,
      "step": 472
    },
    {
      "epoch": 0.09124228395061729,
      "grad_norm": 1.301705241203308,
      "learning_rate": 0.0005,
      "loss": 3.309187889099121,
      "step": 473
    },
    {
      "epoch": 0.09143518518518519,
      "grad_norm": 1.1598379611968994,
      "learning_rate": 0.0005,
      "loss": 3.6731820106506348,
      "step": 474
    },
    {
      "epoch": 0.09162808641975309,
      "grad_norm": 1.1653698682785034,
      "learning_rate": 0.0005,
      "loss": 3.587973117828369,
      "step": 475
    },
    {
      "epoch": 0.09182098765432099,
      "grad_norm": 1.1496787071228027,
      "learning_rate": 0.0005,
      "loss": 3.363464832305908,
      "step": 476
    },
    {
      "epoch": 0.0920138888888889,
      "grad_norm": 1.0925607681274414,
      "learning_rate": 0.0005,
      "loss": 3.7042934894561768,
      "step": 477
    },
    {
      "epoch": 0.09220679012345678,
      "grad_norm": 1.476737141609192,
      "learning_rate": 0.0005,
      "loss": 3.6533970832824707,
      "step": 478
    },
    {
      "epoch": 0.09239969135802469,
      "grad_norm": 1.2615936994552612,
      "learning_rate": 0.0005,
      "loss": 3.7438762187957764,
      "step": 479
    },
    {
      "epoch": 0.09259259259259259,
      "grad_norm": 1.1540554761886597,
      "learning_rate": 0.0005,
      "loss": 3.372781991958618,
      "step": 480
    },
    {
      "epoch": 0.09278549382716049,
      "grad_norm": 1.4797227382659912,
      "learning_rate": 0.0005,
      "loss": 3.724581241607666,
      "step": 481
    },
    {
      "epoch": 0.09297839506172839,
      "grad_norm": 1.31166672706604,
      "learning_rate": 0.0005,
      "loss": 3.6670033931732178,
      "step": 482
    },
    {
      "epoch": 0.0931712962962963,
      "grad_norm": 1.1213045120239258,
      "learning_rate": 0.0005,
      "loss": 3.71860933303833,
      "step": 483
    },
    {
      "epoch": 0.0933641975308642,
      "grad_norm": 1.2632807493209839,
      "learning_rate": 0.0005,
      "loss": 3.2460432052612305,
      "step": 484
    },
    {
      "epoch": 0.0935570987654321,
      "grad_norm": 1.3811445236206055,
      "learning_rate": 0.0005,
      "loss": 3.4048023223876953,
      "step": 485
    },
    {
      "epoch": 0.09375,
      "grad_norm": 1.637803077697754,
      "learning_rate": 0.0005,
      "loss": 3.8716728687286377,
      "step": 486
    },
    {
      "epoch": 0.0939429012345679,
      "grad_norm": 1.6257916688919067,
      "learning_rate": 0.0005,
      "loss": 3.751460552215576,
      "step": 487
    },
    {
      "epoch": 0.0941358024691358,
      "grad_norm": 1.1820534467697144,
      "learning_rate": 0.0005,
      "loss": 3.438950538635254,
      "step": 488
    },
    {
      "epoch": 0.0943287037037037,
      "grad_norm": 1.2347174882888794,
      "learning_rate": 0.0005,
      "loss": 3.678992986679077,
      "step": 489
    },
    {
      "epoch": 0.09452160493827161,
      "grad_norm": 1.5473805665969849,
      "learning_rate": 0.0005,
      "loss": 3.299762725830078,
      "step": 490
    },
    {
      "epoch": 0.09471450617283951,
      "grad_norm": 1.4582875967025757,
      "learning_rate": 0.0005,
      "loss": 3.669379711151123,
      "step": 491
    },
    {
      "epoch": 0.09490740740740741,
      "grad_norm": 1.433157205581665,
      "learning_rate": 0.0005,
      "loss": 3.54311466217041,
      "step": 492
    },
    {
      "epoch": 0.09510030864197531,
      "grad_norm": 1.7625281810760498,
      "learning_rate": 0.0005,
      "loss": 3.570894241333008,
      "step": 493
    },
    {
      "epoch": 0.09529320987654322,
      "grad_norm": 1.4318376779556274,
      "learning_rate": 0.0005,
      "loss": 3.674067974090576,
      "step": 494
    },
    {
      "epoch": 0.0954861111111111,
      "grad_norm": 1.4747264385223389,
      "learning_rate": 0.0005,
      "loss": 3.8399291038513184,
      "step": 495
    },
    {
      "epoch": 0.09567901234567901,
      "grad_norm": 1.3702301979064941,
      "learning_rate": 0.0005,
      "loss": 3.657477855682373,
      "step": 496
    },
    {
      "epoch": 0.09587191358024691,
      "grad_norm": 1.4646774530410767,
      "learning_rate": 0.0005,
      "loss": 3.4225289821624756,
      "step": 497
    },
    {
      "epoch": 0.09606481481481481,
      "grad_norm": 1.1710195541381836,
      "learning_rate": 0.0005,
      "loss": 3.767091989517212,
      "step": 498
    },
    {
      "epoch": 0.09625771604938271,
      "grad_norm": 1.4643454551696777,
      "learning_rate": 0.0005,
      "loss": 4.037400245666504,
      "step": 499
    },
    {
      "epoch": 0.09645061728395062,
      "grad_norm": 1.2281546592712402,
      "learning_rate": 0.0005,
      "loss": 3.446115255355835,
      "step": 500
    },
    {
      "epoch": 0.09664351851851852,
      "grad_norm": 1.0913312435150146,
      "learning_rate": 0.0005,
      "loss": 3.544445514678955,
      "step": 501
    },
    {
      "epoch": 0.09683641975308642,
      "grad_norm": 1.2342119216918945,
      "learning_rate": 0.0005,
      "loss": 3.2805094718933105,
      "step": 502
    },
    {
      "epoch": 0.09702932098765432,
      "grad_norm": 1.3024507761001587,
      "learning_rate": 0.0005,
      "loss": 3.657787322998047,
      "step": 503
    },
    {
      "epoch": 0.09722222222222222,
      "grad_norm": 1.385835886001587,
      "learning_rate": 0.0005,
      "loss": 3.749901533126831,
      "step": 504
    },
    {
      "epoch": 0.09741512345679013,
      "grad_norm": 1.8197722434997559,
      "learning_rate": 0.0005,
      "loss": 3.5162012577056885,
      "step": 505
    },
    {
      "epoch": 0.09760802469135803,
      "grad_norm": 1.323750615119934,
      "learning_rate": 0.0005,
      "loss": 3.1873600482940674,
      "step": 506
    },
    {
      "epoch": 0.09780092592592593,
      "grad_norm": 1.3018983602523804,
      "learning_rate": 0.0005,
      "loss": 3.4748706817626953,
      "step": 507
    },
    {
      "epoch": 0.09799382716049383,
      "grad_norm": 1.407242774963379,
      "learning_rate": 0.0005,
      "loss": 3.6594197750091553,
      "step": 508
    },
    {
      "epoch": 0.09818672839506173,
      "grad_norm": 1.5374746322631836,
      "learning_rate": 0.0005,
      "loss": 3.4674973487854004,
      "step": 509
    },
    {
      "epoch": 0.09837962962962964,
      "grad_norm": 1.413445234298706,
      "learning_rate": 0.0005,
      "loss": 3.41622257232666,
      "step": 510
    },
    {
      "epoch": 0.09857253086419752,
      "grad_norm": 1.956650972366333,
      "learning_rate": 0.0005,
      "loss": 3.726458787918091,
      "step": 511
    },
    {
      "epoch": 0.09876543209876543,
      "grad_norm": 1.516943335533142,
      "learning_rate": 0.0005,
      "loss": 3.2854549884796143,
      "step": 512
    },
    {
      "epoch": 0.09895833333333333,
      "grad_norm": 1.1098196506500244,
      "learning_rate": 0.0005,
      "loss": 3.3592371940612793,
      "step": 513
    },
    {
      "epoch": 0.09915123456790123,
      "grad_norm": 1.3796637058258057,
      "learning_rate": 0.0005,
      "loss": 3.8580822944641113,
      "step": 514
    },
    {
      "epoch": 0.09934413580246913,
      "grad_norm": 1.736323356628418,
      "learning_rate": 0.0005,
      "loss": 3.6078720092773438,
      "step": 515
    },
    {
      "epoch": 0.09953703703703703,
      "grad_norm": 1.1186288595199585,
      "learning_rate": 0.0005,
      "loss": 3.8750481605529785,
      "step": 516
    },
    {
      "epoch": 0.09972993827160494,
      "grad_norm": 1.2918788194656372,
      "learning_rate": 0.0005,
      "loss": 3.6755082607269287,
      "step": 517
    },
    {
      "epoch": 0.09992283950617284,
      "grad_norm": 2.0960330963134766,
      "learning_rate": 0.0005,
      "loss": 3.409087657928467,
      "step": 518
    },
    {
      "epoch": 0.10011574074074074,
      "grad_norm": 1.3887019157409668,
      "learning_rate": 0.0005,
      "loss": 3.3431432247161865,
      "step": 519
    },
    {
      "epoch": 0.10030864197530864,
      "grad_norm": 1.1522376537322998,
      "learning_rate": 0.0005,
      "loss": 3.3214073181152344,
      "step": 520
    },
    {
      "epoch": 0.10050154320987655,
      "grad_norm": 1.2679548263549805,
      "learning_rate": 0.0005,
      "loss": 3.6279311180114746,
      "step": 521
    },
    {
      "epoch": 0.10069444444444445,
      "grad_norm": 1.7699371576309204,
      "learning_rate": 0.0005,
      "loss": 3.7662148475646973,
      "step": 522
    },
    {
      "epoch": 0.10088734567901235,
      "grad_norm": 1.094833254814148,
      "learning_rate": 0.0005,
      "loss": 3.5336103439331055,
      "step": 523
    },
    {
      "epoch": 0.10108024691358025,
      "grad_norm": 1.2659540176391602,
      "learning_rate": 0.0005,
      "loss": 3.3820383548736572,
      "step": 524
    },
    {
      "epoch": 0.10127314814814815,
      "grad_norm": 1.6098750829696655,
      "learning_rate": 0.0005,
      "loss": 3.6909968852996826,
      "step": 525
    },
    {
      "epoch": 0.10146604938271606,
      "grad_norm": 1.4685800075531006,
      "learning_rate": 0.0005,
      "loss": 3.464573860168457,
      "step": 526
    },
    {
      "epoch": 0.10165895061728394,
      "grad_norm": 1.1567829847335815,
      "learning_rate": 0.0005,
      "loss": 3.653791904449463,
      "step": 527
    },
    {
      "epoch": 0.10185185185185185,
      "grad_norm": 1.3388903141021729,
      "learning_rate": 0.0005,
      "loss": 3.475069522857666,
      "step": 528
    },
    {
      "epoch": 0.10204475308641975,
      "grad_norm": 1.392905592918396,
      "learning_rate": 0.0005,
      "loss": 3.7325942516326904,
      "step": 529
    },
    {
      "epoch": 0.10223765432098765,
      "grad_norm": 1.297948956489563,
      "learning_rate": 0.0005,
      "loss": 3.1753087043762207,
      "step": 530
    },
    {
      "epoch": 0.10243055555555555,
      "grad_norm": 1.7607842683792114,
      "learning_rate": 0.0005,
      "loss": 3.341111660003662,
      "step": 531
    },
    {
      "epoch": 0.10262345679012345,
      "grad_norm": 2.1574742794036865,
      "learning_rate": 0.0005,
      "loss": 3.5739989280700684,
      "step": 532
    },
    {
      "epoch": 0.10281635802469136,
      "grad_norm": 1.7590235471725464,
      "learning_rate": 0.0005,
      "loss": 3.377506732940674,
      "step": 533
    },
    {
      "epoch": 0.10300925925925926,
      "grad_norm": 1.5919744968414307,
      "learning_rate": 0.0005,
      "loss": 3.798572540283203,
      "step": 534
    },
    {
      "epoch": 0.10320216049382716,
      "grad_norm": 1.1754359006881714,
      "learning_rate": 0.0005,
      "loss": 3.606814384460449,
      "step": 535
    },
    {
      "epoch": 0.10339506172839506,
      "grad_norm": 1.1839916706085205,
      "learning_rate": 0.0005,
      "loss": 3.5888428688049316,
      "step": 536
    },
    {
      "epoch": 0.10358796296296297,
      "grad_norm": 0.9830666184425354,
      "learning_rate": 0.0005,
      "loss": 3.3662843704223633,
      "step": 537
    },
    {
      "epoch": 0.10378086419753087,
      "grad_norm": 1.3855167627334595,
      "learning_rate": 0.0005,
      "loss": 3.5342297554016113,
      "step": 538
    },
    {
      "epoch": 0.10397376543209877,
      "grad_norm": 1.2440232038497925,
      "learning_rate": 0.0005,
      "loss": 3.4495692253112793,
      "step": 539
    },
    {
      "epoch": 0.10416666666666667,
      "grad_norm": 1.3799152374267578,
      "learning_rate": 0.0005,
      "loss": 3.6079673767089844,
      "step": 540
    },
    {
      "epoch": 0.10435956790123457,
      "grad_norm": 1.0975626707077026,
      "learning_rate": 0.0005,
      "loss": 3.4278616905212402,
      "step": 541
    },
    {
      "epoch": 0.10455246913580248,
      "grad_norm": 2.0282106399536133,
      "learning_rate": 0.0005,
      "loss": 3.5828065872192383,
      "step": 542
    },
    {
      "epoch": 0.10474537037037036,
      "grad_norm": 1.5207017660140991,
      "learning_rate": 0.0005,
      "loss": 3.311784267425537,
      "step": 543
    },
    {
      "epoch": 0.10493827160493827,
      "grad_norm": 1.1231564283370972,
      "learning_rate": 0.0005,
      "loss": 3.5376410484313965,
      "step": 544
    },
    {
      "epoch": 0.10513117283950617,
      "grad_norm": 1.3744677305221558,
      "learning_rate": 0.0005,
      "loss": 3.430325984954834,
      "step": 545
    },
    {
      "epoch": 0.10532407407407407,
      "grad_norm": 1.6018383502960205,
      "learning_rate": 0.0005,
      "loss": 3.748598575592041,
      "step": 546
    },
    {
      "epoch": 0.10551697530864197,
      "grad_norm": 1.1601258516311646,
      "learning_rate": 0.0005,
      "loss": 3.433321475982666,
      "step": 547
    },
    {
      "epoch": 0.10570987654320987,
      "grad_norm": 1.2177603244781494,
      "learning_rate": 0.0005,
      "loss": 3.358661651611328,
      "step": 548
    },
    {
      "epoch": 0.10590277777777778,
      "grad_norm": 8.880247116088867,
      "learning_rate": 0.0005,
      "loss": 3.536065101623535,
      "step": 549
    },
    {
      "epoch": 0.10609567901234568,
      "grad_norm": 1.532173991203308,
      "learning_rate": 0.0005,
      "loss": 3.612504482269287,
      "step": 550
    },
    {
      "epoch": 0.10628858024691358,
      "grad_norm": 1.1948013305664062,
      "learning_rate": 0.0005,
      "loss": 3.4617631435394287,
      "step": 551
    },
    {
      "epoch": 0.10648148148148148,
      "grad_norm": 1.5016192197799683,
      "learning_rate": 0.0005,
      "loss": 3.198024034500122,
      "step": 552
    },
    {
      "epoch": 0.10667438271604938,
      "grad_norm": 1.0446025133132935,
      "learning_rate": 0.0005,
      "loss": 3.6030611991882324,
      "step": 553
    },
    {
      "epoch": 0.10686728395061729,
      "grad_norm": 1.3206231594085693,
      "learning_rate": 0.0005,
      "loss": 3.7484350204467773,
      "step": 554
    },
    {
      "epoch": 0.10706018518518519,
      "grad_norm": 1.3567694425582886,
      "learning_rate": 0.0005,
      "loss": 3.2235307693481445,
      "step": 555
    },
    {
      "epoch": 0.10725308641975309,
      "grad_norm": 1.2127693891525269,
      "learning_rate": 0.0005,
      "loss": 3.613340139389038,
      "step": 556
    },
    {
      "epoch": 0.10744598765432099,
      "grad_norm": 1.672666072845459,
      "learning_rate": 0.0005,
      "loss": 3.540678024291992,
      "step": 557
    },
    {
      "epoch": 0.1076388888888889,
      "grad_norm": 1.3972340822219849,
      "learning_rate": 0.0005,
      "loss": 3.28371000289917,
      "step": 558
    },
    {
      "epoch": 0.10783179012345678,
      "grad_norm": 1.2128556966781616,
      "learning_rate": 0.0005,
      "loss": 3.3149919509887695,
      "step": 559
    },
    {
      "epoch": 0.10802469135802469,
      "grad_norm": 1.2900439500808716,
      "learning_rate": 0.0005,
      "loss": 3.3502116203308105,
      "step": 560
    },
    {
      "epoch": 0.10821759259259259,
      "grad_norm": 1.0903451442718506,
      "learning_rate": 0.0005,
      "loss": 3.588055372238159,
      "step": 561
    },
    {
      "epoch": 0.10841049382716049,
      "grad_norm": 1.1414294242858887,
      "learning_rate": 0.0005,
      "loss": 3.1839113235473633,
      "step": 562
    },
    {
      "epoch": 0.10860339506172839,
      "grad_norm": 1.3592126369476318,
      "learning_rate": 0.0005,
      "loss": 3.352904796600342,
      "step": 563
    },
    {
      "epoch": 0.1087962962962963,
      "grad_norm": 1.564255952835083,
      "learning_rate": 0.0005,
      "loss": 3.5265097618103027,
      "step": 564
    },
    {
      "epoch": 0.1089891975308642,
      "grad_norm": 1.2471532821655273,
      "learning_rate": 0.0005,
      "loss": 3.597750663757324,
      "step": 565
    },
    {
      "epoch": 0.1091820987654321,
      "grad_norm": 1.294247031211853,
      "learning_rate": 0.0005,
      "loss": 3.4258432388305664,
      "step": 566
    },
    {
      "epoch": 0.109375,
      "grad_norm": 1.7897231578826904,
      "learning_rate": 0.0005,
      "loss": 3.39973783493042,
      "step": 567
    },
    {
      "epoch": 0.1095679012345679,
      "grad_norm": 1.2063484191894531,
      "learning_rate": 0.0005,
      "loss": 3.5417823791503906,
      "step": 568
    },
    {
      "epoch": 0.1097608024691358,
      "grad_norm": 1.0305536985397339,
      "learning_rate": 0.0005,
      "loss": 3.38230562210083,
      "step": 569
    },
    {
      "epoch": 0.1099537037037037,
      "grad_norm": 1.7142891883850098,
      "learning_rate": 0.0005,
      "loss": 3.4101624488830566,
      "step": 570
    },
    {
      "epoch": 0.11014660493827161,
      "grad_norm": 1.1612286567687988,
      "learning_rate": 0.0005,
      "loss": 3.505512237548828,
      "step": 571
    },
    {
      "epoch": 0.11033950617283951,
      "grad_norm": 1.304697036743164,
      "learning_rate": 0.0005,
      "loss": 3.636615514755249,
      "step": 572
    },
    {
      "epoch": 0.11053240740740741,
      "grad_norm": 1.1534284353256226,
      "learning_rate": 0.0005,
      "loss": 3.3105826377868652,
      "step": 573
    },
    {
      "epoch": 0.11072530864197531,
      "grad_norm": 1.3510081768035889,
      "learning_rate": 0.0005,
      "loss": 3.4684529304504395,
      "step": 574
    },
    {
      "epoch": 0.11091820987654322,
      "grad_norm": 1.1888664960861206,
      "learning_rate": 0.0005,
      "loss": 3.44630765914917,
      "step": 575
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 1.201544165611267,
      "learning_rate": 0.0005,
      "loss": 3.6401243209838867,
      "step": 576
    },
    {
      "epoch": 0.11130401234567901,
      "grad_norm": 1.150037407875061,
      "learning_rate": 0.0005,
      "loss": 3.365467071533203,
      "step": 577
    },
    {
      "epoch": 0.11149691358024691,
      "grad_norm": 1.3019529581069946,
      "learning_rate": 0.0005,
      "loss": 3.509678840637207,
      "step": 578
    },
    {
      "epoch": 0.11168981481481481,
      "grad_norm": 1.6328999996185303,
      "learning_rate": 0.0005,
      "loss": 3.8485054969787598,
      "step": 579
    },
    {
      "epoch": 0.11188271604938271,
      "grad_norm": 1.3003342151641846,
      "learning_rate": 0.0005,
      "loss": 3.5101187229156494,
      "step": 580
    },
    {
      "epoch": 0.11207561728395062,
      "grad_norm": 1.1171666383743286,
      "learning_rate": 0.0005,
      "loss": 3.611030101776123,
      "step": 581
    },
    {
      "epoch": 0.11226851851851852,
      "grad_norm": 1.4246530532836914,
      "learning_rate": 0.0005,
      "loss": 3.3652281761169434,
      "step": 582
    },
    {
      "epoch": 0.11246141975308642,
      "grad_norm": 1.1834734678268433,
      "learning_rate": 0.0005,
      "loss": 3.438998222351074,
      "step": 583
    },
    {
      "epoch": 0.11265432098765432,
      "grad_norm": 0.9929169416427612,
      "learning_rate": 0.0005,
      "loss": 3.587484359741211,
      "step": 584
    },
    {
      "epoch": 0.11284722222222222,
      "grad_norm": 1.7786102294921875,
      "learning_rate": 0.0005,
      "loss": 3.6218512058258057,
      "step": 585
    },
    {
      "epoch": 0.11304012345679013,
      "grad_norm": 1.2335901260375977,
      "learning_rate": 0.0005,
      "loss": 3.4116878509521484,
      "step": 586
    },
    {
      "epoch": 0.11323302469135803,
      "grad_norm": 1.0701309442520142,
      "learning_rate": 0.0005,
      "loss": 3.423351287841797,
      "step": 587
    },
    {
      "epoch": 0.11342592592592593,
      "grad_norm": 1.2565034627914429,
      "learning_rate": 0.0005,
      "loss": 3.4585001468658447,
      "step": 588
    },
    {
      "epoch": 0.11361882716049383,
      "grad_norm": 1.1462056636810303,
      "learning_rate": 0.0005,
      "loss": 3.260995388031006,
      "step": 589
    },
    {
      "epoch": 0.11381172839506173,
      "grad_norm": 1.275178074836731,
      "learning_rate": 0.0005,
      "loss": 3.862854242324829,
      "step": 590
    },
    {
      "epoch": 0.11400462962962964,
      "grad_norm": 1.5795482397079468,
      "learning_rate": 0.0005,
      "loss": 3.3268632888793945,
      "step": 591
    },
    {
      "epoch": 0.11419753086419752,
      "grad_norm": 1.1789237260818481,
      "learning_rate": 0.0005,
      "loss": 3.3696494102478027,
      "step": 592
    },
    {
      "epoch": 0.11439043209876543,
      "grad_norm": 1.304183006286621,
      "learning_rate": 0.0005,
      "loss": 3.4128379821777344,
      "step": 593
    },
    {
      "epoch": 0.11458333333333333,
      "grad_norm": 1.1975260972976685,
      "learning_rate": 0.0005,
      "loss": 3.5545811653137207,
      "step": 594
    },
    {
      "epoch": 0.11477623456790123,
      "grad_norm": 1.0304841995239258,
      "learning_rate": 0.0005,
      "loss": 3.7321090698242188,
      "step": 595
    },
    {
      "epoch": 0.11496913580246913,
      "grad_norm": 1.234447717666626,
      "learning_rate": 0.0005,
      "loss": 3.166663885116577,
      "step": 596
    },
    {
      "epoch": 0.11516203703703703,
      "grad_norm": 1.1297004222869873,
      "learning_rate": 0.0005,
      "loss": 3.473278522491455,
      "step": 597
    },
    {
      "epoch": 0.11535493827160494,
      "grad_norm": 1.2614351511001587,
      "learning_rate": 0.0005,
      "loss": 3.5532565116882324,
      "step": 598
    },
    {
      "epoch": 0.11554783950617284,
      "grad_norm": 1.2148919105529785,
      "learning_rate": 0.0005,
      "loss": 3.605142116546631,
      "step": 599
    },
    {
      "epoch": 0.11574074074074074,
      "grad_norm": 1.1464146375656128,
      "learning_rate": 0.0005,
      "loss": 3.480384349822998,
      "step": 600
    },
    {
      "epoch": 0.11593364197530864,
      "grad_norm": 1.2577064037322998,
      "learning_rate": 0.0005,
      "loss": 3.447354793548584,
      "step": 601
    },
    {
      "epoch": 0.11612654320987655,
      "grad_norm": 1.3395886421203613,
      "learning_rate": 0.0005,
      "loss": 3.6224613189697266,
      "step": 602
    },
    {
      "epoch": 0.11631944444444445,
      "grad_norm": 1.289271593093872,
      "learning_rate": 0.0005,
      "loss": 3.744739055633545,
      "step": 603
    },
    {
      "epoch": 0.11651234567901235,
      "grad_norm": 1.422060489654541,
      "learning_rate": 0.0005,
      "loss": 3.3991470336914062,
      "step": 604
    },
    {
      "epoch": 0.11670524691358025,
      "grad_norm": 1.3639577627182007,
      "learning_rate": 0.0005,
      "loss": 3.765862226486206,
      "step": 605
    },
    {
      "epoch": 0.11689814814814815,
      "grad_norm": 1.5927373170852661,
      "learning_rate": 0.0005,
      "loss": 3.739642858505249,
      "step": 606
    },
    {
      "epoch": 0.11709104938271606,
      "grad_norm": 1.3499432802200317,
      "learning_rate": 0.0005,
      "loss": 3.166726589202881,
      "step": 607
    },
    {
      "epoch": 0.11728395061728394,
      "grad_norm": 1.5478922128677368,
      "learning_rate": 0.0005,
      "loss": 3.7656216621398926,
      "step": 608
    },
    {
      "epoch": 0.11747685185185185,
      "grad_norm": 2.1266918182373047,
      "learning_rate": 0.0005,
      "loss": 3.35608172416687,
      "step": 609
    },
    {
      "epoch": 0.11766975308641975,
      "grad_norm": 2.0416078567504883,
      "learning_rate": 0.0005,
      "loss": 3.631129503250122,
      "step": 610
    },
    {
      "epoch": 0.11786265432098765,
      "grad_norm": 1.6562680006027222,
      "learning_rate": 0.0005,
      "loss": 3.4646897315979004,
      "step": 611
    },
    {
      "epoch": 0.11805555555555555,
      "grad_norm": 1.3780306577682495,
      "learning_rate": 0.0005,
      "loss": 3.448427200317383,
      "step": 612
    },
    {
      "epoch": 0.11824845679012345,
      "grad_norm": 1.1480494737625122,
      "learning_rate": 0.0005,
      "loss": 3.2072882652282715,
      "step": 613
    },
    {
      "epoch": 0.11844135802469136,
      "grad_norm": 1.2051708698272705,
      "learning_rate": 0.0005,
      "loss": 3.531014919281006,
      "step": 614
    },
    {
      "epoch": 0.11863425925925926,
      "grad_norm": 1.5139105319976807,
      "learning_rate": 0.0005,
      "loss": 3.7359848022460938,
      "step": 615
    },
    {
      "epoch": 0.11882716049382716,
      "grad_norm": 1.2719051837921143,
      "learning_rate": 0.0005,
      "loss": 3.245485305786133,
      "step": 616
    },
    {
      "epoch": 0.11902006172839506,
      "grad_norm": 1.6115251779556274,
      "learning_rate": 0.0005,
      "loss": 3.53229022026062,
      "step": 617
    },
    {
      "epoch": 0.11921296296296297,
      "grad_norm": 1.5956937074661255,
      "learning_rate": 0.0005,
      "loss": 3.825761318206787,
      "step": 618
    },
    {
      "epoch": 0.11940586419753087,
      "grad_norm": 1.2533913850784302,
      "learning_rate": 0.0005,
      "loss": 3.590665817260742,
      "step": 619
    },
    {
      "epoch": 0.11959876543209877,
      "grad_norm": 1.2315171957015991,
      "learning_rate": 0.0005,
      "loss": 3.2639122009277344,
      "step": 620
    },
    {
      "epoch": 0.11979166666666667,
      "grad_norm": 1.6858584880828857,
      "learning_rate": 0.0005,
      "loss": 3.768361806869507,
      "step": 621
    },
    {
      "epoch": 0.11998456790123457,
      "grad_norm": 1.0826984643936157,
      "learning_rate": 0.0005,
      "loss": 3.6679110527038574,
      "step": 622
    },
    {
      "epoch": 0.12017746913580248,
      "grad_norm": 1.9571703672409058,
      "learning_rate": 0.0005,
      "loss": 3.2555887699127197,
      "step": 623
    },
    {
      "epoch": 0.12037037037037036,
      "grad_norm": 1.324794054031372,
      "learning_rate": 0.0005,
      "loss": 3.533306837081909,
      "step": 624
    },
    {
      "epoch": 0.12056327160493827,
      "grad_norm": 1.092252492904663,
      "learning_rate": 0.0005,
      "loss": 3.206183433532715,
      "step": 625
    },
    {
      "epoch": 0.12075617283950617,
      "grad_norm": 1.3903816938400269,
      "learning_rate": 0.0005,
      "loss": 3.3891355991363525,
      "step": 626
    },
    {
      "epoch": 0.12094907407407407,
      "grad_norm": 1.2059326171875,
      "learning_rate": 0.0005,
      "loss": 3.2376413345336914,
      "step": 627
    },
    {
      "epoch": 0.12114197530864197,
      "grad_norm": 1.0596168041229248,
      "learning_rate": 0.0005,
      "loss": 3.4064278602600098,
      "step": 628
    },
    {
      "epoch": 0.12133487654320987,
      "grad_norm": 2.0079469680786133,
      "learning_rate": 0.0005,
      "loss": 3.248427629470825,
      "step": 629
    },
    {
      "epoch": 0.12152777777777778,
      "grad_norm": 1.297705888748169,
      "learning_rate": 0.0005,
      "loss": 3.6460342407226562,
      "step": 630
    },
    {
      "epoch": 0.12172067901234568,
      "grad_norm": 0.9988003373146057,
      "learning_rate": 0.0005,
      "loss": 3.4143989086151123,
      "step": 631
    },
    {
      "epoch": 0.12191358024691358,
      "grad_norm": 1.2936666011810303,
      "learning_rate": 0.0005,
      "loss": 3.574744701385498,
      "step": 632
    },
    {
      "epoch": 0.12210648148148148,
      "grad_norm": 1.679791808128357,
      "learning_rate": 0.0005,
      "loss": 3.6905181407928467,
      "step": 633
    },
    {
      "epoch": 0.12229938271604938,
      "grad_norm": 1.122711181640625,
      "learning_rate": 0.0005,
      "loss": 3.4268226623535156,
      "step": 634
    },
    {
      "epoch": 0.12249228395061729,
      "grad_norm": 1.3861441612243652,
      "learning_rate": 0.0005,
      "loss": 3.4852449893951416,
      "step": 635
    },
    {
      "epoch": 0.12268518518518519,
      "grad_norm": 1.172989010810852,
      "learning_rate": 0.0005,
      "loss": 3.4740445613861084,
      "step": 636
    },
    {
      "epoch": 0.12287808641975309,
      "grad_norm": 1.5097317695617676,
      "learning_rate": 0.0005,
      "loss": 3.3961896896362305,
      "step": 637
    },
    {
      "epoch": 0.12307098765432099,
      "grad_norm": 1.3097563982009888,
      "learning_rate": 0.0005,
      "loss": 3.485687494277954,
      "step": 638
    },
    {
      "epoch": 0.1232638888888889,
      "grad_norm": 1.7128081321716309,
      "learning_rate": 0.0005,
      "loss": 3.532806158065796,
      "step": 639
    },
    {
      "epoch": 0.12345679012345678,
      "grad_norm": 1.5347670316696167,
      "learning_rate": 0.0005,
      "loss": 3.744718313217163,
      "step": 640
    },
    {
      "epoch": 0.12364969135802469,
      "grad_norm": 1.4066052436828613,
      "learning_rate": 0.0005,
      "loss": 3.320951461791992,
      "step": 641
    },
    {
      "epoch": 0.12384259259259259,
      "grad_norm": 1.5700030326843262,
      "learning_rate": 0.0005,
      "loss": 3.605780601501465,
      "step": 642
    },
    {
      "epoch": 0.12403549382716049,
      "grad_norm": 1.2897077798843384,
      "learning_rate": 0.0005,
      "loss": 3.650787353515625,
      "step": 643
    },
    {
      "epoch": 0.12422839506172839,
      "grad_norm": 1.1705509424209595,
      "learning_rate": 0.0005,
      "loss": 3.5392189025878906,
      "step": 644
    },
    {
      "epoch": 0.1244212962962963,
      "grad_norm": 1.14071524143219,
      "learning_rate": 0.0005,
      "loss": 3.6535511016845703,
      "step": 645
    },
    {
      "epoch": 0.1246141975308642,
      "grad_norm": 1.588837742805481,
      "learning_rate": 0.0005,
      "loss": 3.4731874465942383,
      "step": 646
    },
    {
      "epoch": 0.1248070987654321,
      "grad_norm": 1.1605141162872314,
      "learning_rate": 0.0005,
      "loss": 3.515601634979248,
      "step": 647
    },
    {
      "epoch": 0.125,
      "grad_norm": 1.1023057699203491,
      "learning_rate": 0.0005,
      "loss": 3.1612627506256104,
      "step": 648
    },
    {
      "epoch": 0.1251929012345679,
      "grad_norm": 1.2278584241867065,
      "learning_rate": 0.0005,
      "loss": 3.4631195068359375,
      "step": 649
    },
    {
      "epoch": 0.1253858024691358,
      "grad_norm": 1.273430347442627,
      "learning_rate": 0.0005,
      "loss": 3.3529152870178223,
      "step": 650
    },
    {
      "epoch": 0.1255787037037037,
      "grad_norm": 1.340792179107666,
      "learning_rate": 0.0005,
      "loss": 3.6604135036468506,
      "step": 651
    },
    {
      "epoch": 0.1257716049382716,
      "grad_norm": 1.2687286138534546,
      "learning_rate": 0.0005,
      "loss": 3.259390115737915,
      "step": 652
    },
    {
      "epoch": 0.1259645061728395,
      "grad_norm": 1.1838475465774536,
      "learning_rate": 0.0005,
      "loss": 3.556910991668701,
      "step": 653
    },
    {
      "epoch": 0.1261574074074074,
      "grad_norm": 1.9384517669677734,
      "learning_rate": 0.0005,
      "loss": 3.4541985988616943,
      "step": 654
    },
    {
      "epoch": 0.1263503086419753,
      "grad_norm": 1.2875640392303467,
      "learning_rate": 0.0005,
      "loss": 3.2870326042175293,
      "step": 655
    },
    {
      "epoch": 0.12654320987654322,
      "grad_norm": 1.5170327425003052,
      "learning_rate": 0.0005,
      "loss": 3.7387375831604004,
      "step": 656
    },
    {
      "epoch": 0.1267361111111111,
      "grad_norm": 1.5477646589279175,
      "learning_rate": 0.0005,
      "loss": 3.4933409690856934,
      "step": 657
    },
    {
      "epoch": 0.12692901234567902,
      "grad_norm": 1.2687550783157349,
      "learning_rate": 0.0005,
      "loss": 3.6636290550231934,
      "step": 658
    },
    {
      "epoch": 0.1271219135802469,
      "grad_norm": 1.543900966644287,
      "learning_rate": 0.0005,
      "loss": 3.3884763717651367,
      "step": 659
    },
    {
      "epoch": 0.12731481481481483,
      "grad_norm": 1.141384243965149,
      "learning_rate": 0.0005,
      "loss": 3.530622959136963,
      "step": 660
    },
    {
      "epoch": 0.1275077160493827,
      "grad_norm": 1.4250750541687012,
      "learning_rate": 0.0005,
      "loss": 3.6535773277282715,
      "step": 661
    },
    {
      "epoch": 0.12770061728395063,
      "grad_norm": 1.1011525392532349,
      "learning_rate": 0.0005,
      "loss": 3.3466949462890625,
      "step": 662
    },
    {
      "epoch": 0.12789351851851852,
      "grad_norm": 1.5985902547836304,
      "learning_rate": 0.0005,
      "loss": 3.515746593475342,
      "step": 663
    },
    {
      "epoch": 0.12808641975308643,
      "grad_norm": 1.143218994140625,
      "learning_rate": 0.0005,
      "loss": 3.389608860015869,
      "step": 664
    },
    {
      "epoch": 0.12827932098765432,
      "grad_norm": 1.2273801565170288,
      "learning_rate": 0.0005,
      "loss": 3.82277774810791,
      "step": 665
    },
    {
      "epoch": 0.1284722222222222,
      "grad_norm": 1.3167707920074463,
      "learning_rate": 0.0005,
      "loss": 3.1316375732421875,
      "step": 666
    },
    {
      "epoch": 0.12866512345679013,
      "grad_norm": 1.0857962369918823,
      "learning_rate": 0.0005,
      "loss": 3.481060028076172,
      "step": 667
    },
    {
      "epoch": 0.12885802469135801,
      "grad_norm": 1.4389219284057617,
      "learning_rate": 0.0005,
      "loss": 3.512942314147949,
      "step": 668
    },
    {
      "epoch": 0.12905092592592593,
      "grad_norm": 2.046003818511963,
      "learning_rate": 0.0005,
      "loss": 3.426537036895752,
      "step": 669
    },
    {
      "epoch": 0.12924382716049382,
      "grad_norm": 1.6998050212860107,
      "learning_rate": 0.0005,
      "loss": 3.501582622528076,
      "step": 670
    },
    {
      "epoch": 0.12943672839506173,
      "grad_norm": 1.3560091257095337,
      "learning_rate": 0.0005,
      "loss": 3.575084686279297,
      "step": 671
    },
    {
      "epoch": 0.12962962962962962,
      "grad_norm": 1.0298316478729248,
      "learning_rate": 0.0005,
      "loss": 3.4491820335388184,
      "step": 672
    },
    {
      "epoch": 0.12982253086419754,
      "grad_norm": 0.9638184309005737,
      "learning_rate": 0.0005,
      "loss": 3.2648608684539795,
      "step": 673
    },
    {
      "epoch": 0.13001543209876543,
      "grad_norm": 1.1532272100448608,
      "learning_rate": 0.0005,
      "loss": 3.0521135330200195,
      "step": 674
    },
    {
      "epoch": 0.13020833333333334,
      "grad_norm": 1.364581823348999,
      "learning_rate": 0.0005,
      "loss": 3.7996795177459717,
      "step": 675
    },
    {
      "epoch": 0.13040123456790123,
      "grad_norm": 1.5750479698181152,
      "learning_rate": 0.0005,
      "loss": 3.623044013977051,
      "step": 676
    },
    {
      "epoch": 0.13059413580246915,
      "grad_norm": 1.657015323638916,
      "learning_rate": 0.0005,
      "loss": 3.64475417137146,
      "step": 677
    },
    {
      "epoch": 0.13078703703703703,
      "grad_norm": 1.1378215551376343,
      "learning_rate": 0.0005,
      "loss": 3.623957633972168,
      "step": 678
    },
    {
      "epoch": 0.13097993827160495,
      "grad_norm": 1.5607678890228271,
      "learning_rate": 0.0005,
      "loss": 3.806037425994873,
      "step": 679
    },
    {
      "epoch": 0.13117283950617284,
      "grad_norm": 2.1552493572235107,
      "learning_rate": 0.0005,
      "loss": 3.4351882934570312,
      "step": 680
    },
    {
      "epoch": 0.13136574074074073,
      "grad_norm": 1.7375794649124146,
      "learning_rate": 0.0005,
      "loss": 3.680058002471924,
      "step": 681
    },
    {
      "epoch": 0.13155864197530864,
      "grad_norm": 1.9494624137878418,
      "learning_rate": 0.0005,
      "loss": 3.3511228561401367,
      "step": 682
    },
    {
      "epoch": 0.13175154320987653,
      "grad_norm": 1.9043734073638916,
      "learning_rate": 0.0005,
      "loss": 3.5845301151275635,
      "step": 683
    },
    {
      "epoch": 0.13194444444444445,
      "grad_norm": 1.678818702697754,
      "learning_rate": 0.0005,
      "loss": 3.2590878009796143,
      "step": 684
    },
    {
      "epoch": 0.13213734567901234,
      "grad_norm": 1.311942219734192,
      "learning_rate": 0.0005,
      "loss": 3.4616379737854004,
      "step": 685
    },
    {
      "epoch": 0.13233024691358025,
      "grad_norm": 1.2024143934249878,
      "learning_rate": 0.0005,
      "loss": 3.607743740081787,
      "step": 686
    },
    {
      "epoch": 0.13252314814814814,
      "grad_norm": 1.1210904121398926,
      "learning_rate": 0.0005,
      "loss": 3.588538885116577,
      "step": 687
    },
    {
      "epoch": 0.13271604938271606,
      "grad_norm": 1.889638066291809,
      "learning_rate": 0.0005,
      "loss": 3.7213101387023926,
      "step": 688
    },
    {
      "epoch": 0.13290895061728394,
      "grad_norm": 1.055303692817688,
      "learning_rate": 0.0005,
      "loss": 3.1206564903259277,
      "step": 689
    },
    {
      "epoch": 0.13310185185185186,
      "grad_norm": 1.267528772354126,
      "learning_rate": 0.0005,
      "loss": 3.250030517578125,
      "step": 690
    },
    {
      "epoch": 0.13329475308641975,
      "grad_norm": 1.4954209327697754,
      "learning_rate": 0.0005,
      "loss": 3.560079574584961,
      "step": 691
    },
    {
      "epoch": 0.13348765432098766,
      "grad_norm": 1.3527253866195679,
      "learning_rate": 0.0005,
      "loss": 3.4691123962402344,
      "step": 692
    },
    {
      "epoch": 0.13368055555555555,
      "grad_norm": 1.5166189670562744,
      "learning_rate": 0.0005,
      "loss": 3.3943753242492676,
      "step": 693
    },
    {
      "epoch": 0.13387345679012347,
      "grad_norm": 1.4241434335708618,
      "learning_rate": 0.0005,
      "loss": 3.73844051361084,
      "step": 694
    },
    {
      "epoch": 0.13406635802469136,
      "grad_norm": 1.0770858526229858,
      "learning_rate": 0.0005,
      "loss": 3.3830063343048096,
      "step": 695
    },
    {
      "epoch": 0.13425925925925927,
      "grad_norm": 1.1930732727050781,
      "learning_rate": 0.0005,
      "loss": 3.2448670864105225,
      "step": 696
    },
    {
      "epoch": 0.13445216049382716,
      "grad_norm": 1.1957364082336426,
      "learning_rate": 0.0005,
      "loss": 3.3414416313171387,
      "step": 697
    },
    {
      "epoch": 0.13464506172839505,
      "grad_norm": 1.450555682182312,
      "learning_rate": 0.0005,
      "loss": 3.614260673522949,
      "step": 698
    },
    {
      "epoch": 0.13483796296296297,
      "grad_norm": 1.3581180572509766,
      "learning_rate": 0.0005,
      "loss": 3.3590691089630127,
      "step": 699
    },
    {
      "epoch": 0.13503086419753085,
      "grad_norm": 1.125692367553711,
      "learning_rate": 0.0005,
      "loss": 3.049907922744751,
      "step": 700
    },
    {
      "epoch": 0.13522376543209877,
      "grad_norm": 1.2849427461624146,
      "learning_rate": 0.0005,
      "loss": 3.7860488891601562,
      "step": 701
    },
    {
      "epoch": 0.13541666666666666,
      "grad_norm": 1.2802956104278564,
      "learning_rate": 0.0005,
      "loss": 3.304683208465576,
      "step": 702
    },
    {
      "epoch": 0.13560956790123457,
      "grad_norm": 1.2473196983337402,
      "learning_rate": 0.0005,
      "loss": 3.630114793777466,
      "step": 703
    },
    {
      "epoch": 0.13580246913580246,
      "grad_norm": 1.1371439695358276,
      "learning_rate": 0.0005,
      "loss": 3.2929439544677734,
      "step": 704
    },
    {
      "epoch": 0.13599537037037038,
      "grad_norm": 4.503321170806885,
      "learning_rate": 0.0005,
      "loss": 3.1836581230163574,
      "step": 705
    },
    {
      "epoch": 0.13618827160493827,
      "grad_norm": 1.4052401781082153,
      "learning_rate": 0.0005,
      "loss": 3.127066135406494,
      "step": 706
    },
    {
      "epoch": 0.13638117283950618,
      "grad_norm": 1.2858951091766357,
      "learning_rate": 0.0005,
      "loss": 3.2251033782958984,
      "step": 707
    },
    {
      "epoch": 0.13657407407407407,
      "grad_norm": 4.66586446762085,
      "learning_rate": 0.0005,
      "loss": 3.4737133979797363,
      "step": 708
    },
    {
      "epoch": 0.13676697530864199,
      "grad_norm": 1.3056493997573853,
      "learning_rate": 0.0005,
      "loss": 3.443337917327881,
      "step": 709
    },
    {
      "epoch": 0.13695987654320987,
      "grad_norm": 1.539857268333435,
      "learning_rate": 0.0005,
      "loss": 3.6677186489105225,
      "step": 710
    },
    {
      "epoch": 0.1371527777777778,
      "grad_norm": 1.0732687711715698,
      "learning_rate": 0.0005,
      "loss": 3.5901527404785156,
      "step": 711
    },
    {
      "epoch": 0.13734567901234568,
      "grad_norm": 1.274664282798767,
      "learning_rate": 0.0005,
      "loss": 3.3972723484039307,
      "step": 712
    },
    {
      "epoch": 0.13753858024691357,
      "grad_norm": 1.1421645879745483,
      "learning_rate": 0.0005,
      "loss": 3.2041540145874023,
      "step": 713
    },
    {
      "epoch": 0.13773148148148148,
      "grad_norm": 1.134400486946106,
      "learning_rate": 0.0005,
      "loss": 3.6514766216278076,
      "step": 714
    },
    {
      "epoch": 0.13792438271604937,
      "grad_norm": 1.1513131856918335,
      "learning_rate": 0.0005,
      "loss": 3.505763053894043,
      "step": 715
    },
    {
      "epoch": 0.1381172839506173,
      "grad_norm": 1.4043869972229004,
      "learning_rate": 0.0005,
      "loss": 3.278078556060791,
      "step": 716
    },
    {
      "epoch": 0.13831018518518517,
      "grad_norm": 1.3368698358535767,
      "learning_rate": 0.0005,
      "loss": 3.0938830375671387,
      "step": 717
    },
    {
      "epoch": 0.1385030864197531,
      "grad_norm": 1.3737174272537231,
      "learning_rate": 0.0005,
      "loss": 3.6776185035705566,
      "step": 718
    },
    {
      "epoch": 0.13869598765432098,
      "grad_norm": 2.305671453475952,
      "learning_rate": 0.0005,
      "loss": 3.7734975814819336,
      "step": 719
    },
    {
      "epoch": 0.1388888888888889,
      "grad_norm": 1.1850900650024414,
      "learning_rate": 0.0005,
      "loss": 3.443124771118164,
      "step": 720
    },
    {
      "epoch": 0.13908179012345678,
      "grad_norm": 1.2388434410095215,
      "learning_rate": 0.0005,
      "loss": 3.3962204456329346,
      "step": 721
    },
    {
      "epoch": 0.1392746913580247,
      "grad_norm": 1.3722492456436157,
      "learning_rate": 0.0005,
      "loss": 3.044780731201172,
      "step": 722
    },
    {
      "epoch": 0.1394675925925926,
      "grad_norm": 1.5086047649383545,
      "learning_rate": 0.0005,
      "loss": 3.3853542804718018,
      "step": 723
    },
    {
      "epoch": 0.1396604938271605,
      "grad_norm": 0.9401163458824158,
      "learning_rate": 0.0005,
      "loss": 3.379458427429199,
      "step": 724
    },
    {
      "epoch": 0.1398533950617284,
      "grad_norm": 1.3869901895523071,
      "learning_rate": 0.0005,
      "loss": 3.445977210998535,
      "step": 725
    },
    {
      "epoch": 0.1400462962962963,
      "grad_norm": 1.3870099782943726,
      "learning_rate": 0.0005,
      "loss": 3.205624580383301,
      "step": 726
    },
    {
      "epoch": 0.1402391975308642,
      "grad_norm": 1.6423380374908447,
      "learning_rate": 0.0005,
      "loss": 3.3696258068084717,
      "step": 727
    },
    {
      "epoch": 0.1404320987654321,
      "grad_norm": 1.2613325119018555,
      "learning_rate": 0.0005,
      "loss": 3.094597339630127,
      "step": 728
    },
    {
      "epoch": 0.140625,
      "grad_norm": 1.1994744539260864,
      "learning_rate": 0.0005,
      "loss": 3.552053451538086,
      "step": 729
    },
    {
      "epoch": 0.1408179012345679,
      "grad_norm": 1.1508018970489502,
      "learning_rate": 0.0005,
      "loss": 3.3088302612304688,
      "step": 730
    },
    {
      "epoch": 0.1410108024691358,
      "grad_norm": 1.3247812986373901,
      "learning_rate": 0.0005,
      "loss": 3.117473602294922,
      "step": 731
    },
    {
      "epoch": 0.1412037037037037,
      "grad_norm": 1.131494402885437,
      "learning_rate": 0.0005,
      "loss": 3.58599853515625,
      "step": 732
    },
    {
      "epoch": 0.1413966049382716,
      "grad_norm": 1.2680084705352783,
      "learning_rate": 0.0005,
      "loss": 3.140235424041748,
      "step": 733
    },
    {
      "epoch": 0.1415895061728395,
      "grad_norm": 1.5286396741867065,
      "learning_rate": 0.0005,
      "loss": 3.4294517040252686,
      "step": 734
    },
    {
      "epoch": 0.1417824074074074,
      "grad_norm": 1.5050625801086426,
      "learning_rate": 0.0005,
      "loss": 3.030275344848633,
      "step": 735
    },
    {
      "epoch": 0.1419753086419753,
      "grad_norm": 1.468764305114746,
      "learning_rate": 0.0005,
      "loss": 3.7031161785125732,
      "step": 736
    },
    {
      "epoch": 0.14216820987654322,
      "grad_norm": 1.40556001663208,
      "learning_rate": 0.0005,
      "loss": 3.099886178970337,
      "step": 737
    },
    {
      "epoch": 0.1423611111111111,
      "grad_norm": 1.2693418264389038,
      "learning_rate": 0.0005,
      "loss": 3.2497878074645996,
      "step": 738
    },
    {
      "epoch": 0.14255401234567902,
      "grad_norm": 1.2157824039459229,
      "learning_rate": 0.0005,
      "loss": 3.3915276527404785,
      "step": 739
    },
    {
      "epoch": 0.1427469135802469,
      "grad_norm": 1.2505286931991577,
      "learning_rate": 0.0005,
      "loss": 3.463545322418213,
      "step": 740
    },
    {
      "epoch": 0.14293981481481483,
      "grad_norm": 1.0078980922698975,
      "learning_rate": 0.0005,
      "loss": 3.2447102069854736,
      "step": 741
    },
    {
      "epoch": 0.1431327160493827,
      "grad_norm": 1.149072527885437,
      "learning_rate": 0.0005,
      "loss": 3.4155683517456055,
      "step": 742
    },
    {
      "epoch": 0.14332561728395063,
      "grad_norm": 1.4156914949417114,
      "learning_rate": 0.0005,
      "loss": 3.298419952392578,
      "step": 743
    },
    {
      "epoch": 0.14351851851851852,
      "grad_norm": 1.2257287502288818,
      "learning_rate": 0.0005,
      "loss": 3.1726882457733154,
      "step": 744
    },
    {
      "epoch": 0.14371141975308643,
      "grad_norm": 1.719390630722046,
      "learning_rate": 0.0005,
      "loss": 3.555072546005249,
      "step": 745
    },
    {
      "epoch": 0.14390432098765432,
      "grad_norm": 1.1755965948104858,
      "learning_rate": 0.0005,
      "loss": 3.5185444355010986,
      "step": 746
    },
    {
      "epoch": 0.1440972222222222,
      "grad_norm": 1.247184157371521,
      "learning_rate": 0.0005,
      "loss": 3.2315170764923096,
      "step": 747
    },
    {
      "epoch": 0.14429012345679013,
      "grad_norm": 1.7161874771118164,
      "learning_rate": 0.0005,
      "loss": 3.338794708251953,
      "step": 748
    },
    {
      "epoch": 0.14448302469135801,
      "grad_norm": 1.3936071395874023,
      "learning_rate": 0.0005,
      "loss": 3.327129602432251,
      "step": 749
    },
    {
      "epoch": 0.14467592592592593,
      "grad_norm": 1.8004670143127441,
      "learning_rate": 0.0005,
      "loss": 3.154191017150879,
      "step": 750
    },
    {
      "epoch": 0.14486882716049382,
      "grad_norm": 2.0790092945098877,
      "learning_rate": 0.0005,
      "loss": 3.2558069229125977,
      "step": 751
    },
    {
      "epoch": 0.14506172839506173,
      "grad_norm": 1.38189697265625,
      "learning_rate": 0.0005,
      "loss": 2.998551368713379,
      "step": 752
    },
    {
      "epoch": 0.14525462962962962,
      "grad_norm": 1.0914077758789062,
      "learning_rate": 0.0005,
      "loss": 3.4708640575408936,
      "step": 753
    },
    {
      "epoch": 0.14544753086419754,
      "grad_norm": 1.196287989616394,
      "learning_rate": 0.0005,
      "loss": 3.380042314529419,
      "step": 754
    },
    {
      "epoch": 0.14564043209876543,
      "grad_norm": 0.9742751717567444,
      "learning_rate": 0.0005,
      "loss": 3.521760940551758,
      "step": 755
    },
    {
      "epoch": 0.14583333333333334,
      "grad_norm": 1.7646174430847168,
      "learning_rate": 0.0005,
      "loss": 3.4242141246795654,
      "step": 756
    },
    {
      "epoch": 0.14602623456790123,
      "grad_norm": 1.0953136682510376,
      "learning_rate": 0.0005,
      "loss": 3.505244493484497,
      "step": 757
    },
    {
      "epoch": 0.14621913580246915,
      "grad_norm": 1.5538486242294312,
      "learning_rate": 0.0005,
      "loss": 3.7144546508789062,
      "step": 758
    },
    {
      "epoch": 0.14641203703703703,
      "grad_norm": 1.2666761875152588,
      "learning_rate": 0.0005,
      "loss": 3.5145580768585205,
      "step": 759
    },
    {
      "epoch": 0.14660493827160495,
      "grad_norm": 1.2780039310455322,
      "learning_rate": 0.0005,
      "loss": 3.4992685317993164,
      "step": 760
    },
    {
      "epoch": 0.14679783950617284,
      "grad_norm": 1.5265896320343018,
      "learning_rate": 0.0005,
      "loss": 3.540440320968628,
      "step": 761
    },
    {
      "epoch": 0.14699074074074073,
      "grad_norm": 0.9330888390541077,
      "learning_rate": 0.0005,
      "loss": 3.2767012119293213,
      "step": 762
    },
    {
      "epoch": 0.14718364197530864,
      "grad_norm": 1.1432465314865112,
      "learning_rate": 0.0005,
      "loss": 3.3764147758483887,
      "step": 763
    },
    {
      "epoch": 0.14737654320987653,
      "grad_norm": 1.1555733680725098,
      "learning_rate": 0.0005,
      "loss": 2.875060558319092,
      "step": 764
    },
    {
      "epoch": 0.14756944444444445,
      "grad_norm": 1.6064337491989136,
      "learning_rate": 0.0005,
      "loss": 3.313990354537964,
      "step": 765
    },
    {
      "epoch": 0.14776234567901234,
      "grad_norm": 2.395535469055176,
      "learning_rate": 0.0005,
      "loss": 3.397946834564209,
      "step": 766
    },
    {
      "epoch": 0.14795524691358025,
      "grad_norm": 1.2602503299713135,
      "learning_rate": 0.0005,
      "loss": 3.4790735244750977,
      "step": 767
    },
    {
      "epoch": 0.14814814814814814,
      "grad_norm": 1.07516610622406,
      "learning_rate": 0.0005,
      "loss": 3.288032293319702,
      "step": 768
    },
    {
      "epoch": 0.14834104938271606,
      "grad_norm": 1.3142824172973633,
      "learning_rate": 0.0005,
      "loss": 3.6078076362609863,
      "step": 769
    },
    {
      "epoch": 0.14853395061728394,
      "grad_norm": 1.3241735696792603,
      "learning_rate": 0.0005,
      "loss": 3.2891013622283936,
      "step": 770
    },
    {
      "epoch": 0.14872685185185186,
      "grad_norm": 1.4170715808868408,
      "learning_rate": 0.0005,
      "loss": 3.571732759475708,
      "step": 771
    },
    {
      "epoch": 0.14891975308641975,
      "grad_norm": 1.714975357055664,
      "learning_rate": 0.0005,
      "loss": 3.5987234115600586,
      "step": 772
    },
    {
      "epoch": 0.14911265432098766,
      "grad_norm": 0.9959927201271057,
      "learning_rate": 0.0005,
      "loss": 3.316575527191162,
      "step": 773
    },
    {
      "epoch": 0.14930555555555555,
      "grad_norm": 2.3688244819641113,
      "learning_rate": 0.0005,
      "loss": 3.5866875648498535,
      "step": 774
    },
    {
      "epoch": 0.14949845679012347,
      "grad_norm": 1.163745641708374,
      "learning_rate": 0.0005,
      "loss": 2.9275379180908203,
      "step": 775
    },
    {
      "epoch": 0.14969135802469136,
      "grad_norm": 1.1372026205062866,
      "learning_rate": 0.0005,
      "loss": 3.4403371810913086,
      "step": 776
    },
    {
      "epoch": 0.14988425925925927,
      "grad_norm": 1.4630988836288452,
      "learning_rate": 0.0005,
      "loss": 3.1957056522369385,
      "step": 777
    },
    {
      "epoch": 0.15007716049382716,
      "grad_norm": 1.457415223121643,
      "learning_rate": 0.0005,
      "loss": 3.087283134460449,
      "step": 778
    },
    {
      "epoch": 0.15027006172839505,
      "grad_norm": 1.2011083364486694,
      "learning_rate": 0.0005,
      "loss": 3.0346012115478516,
      "step": 779
    },
    {
      "epoch": 0.15046296296296297,
      "grad_norm": 1.0983169078826904,
      "learning_rate": 0.0005,
      "loss": 3.2572953701019287,
      "step": 780
    },
    {
      "epoch": 0.15065586419753085,
      "grad_norm": 1.3428789377212524,
      "learning_rate": 0.0005,
      "loss": 3.5388450622558594,
      "step": 781
    },
    {
      "epoch": 0.15084876543209877,
      "grad_norm": 1.1256563663482666,
      "learning_rate": 0.0005,
      "loss": 3.283189296722412,
      "step": 782
    },
    {
      "epoch": 0.15104166666666666,
      "grad_norm": 1.1894010305404663,
      "learning_rate": 0.0005,
      "loss": 3.136002540588379,
      "step": 783
    },
    {
      "epoch": 0.15123456790123457,
      "grad_norm": 1.3242759704589844,
      "learning_rate": 0.0005,
      "loss": 3.728269338607788,
      "step": 784
    },
    {
      "epoch": 0.15142746913580246,
      "grad_norm": 1.2382155656814575,
      "learning_rate": 0.0005,
      "loss": 3.5179331302642822,
      "step": 785
    },
    {
      "epoch": 0.15162037037037038,
      "grad_norm": 1.1462693214416504,
      "learning_rate": 0.0005,
      "loss": 3.342766761779785,
      "step": 786
    },
    {
      "epoch": 0.15181327160493827,
      "grad_norm": 1.6539857387542725,
      "learning_rate": 0.0005,
      "loss": 3.377340793609619,
      "step": 787
    },
    {
      "epoch": 0.15200617283950618,
      "grad_norm": 1.2684844732284546,
      "learning_rate": 0.0005,
      "loss": 3.2194619178771973,
      "step": 788
    },
    {
      "epoch": 0.15219907407407407,
      "grad_norm": 1.20352041721344,
      "learning_rate": 0.0005,
      "loss": 3.629943370819092,
      "step": 789
    },
    {
      "epoch": 0.15239197530864199,
      "grad_norm": 1.196768879890442,
      "learning_rate": 0.0005,
      "loss": 3.35768461227417,
      "step": 790
    },
    {
      "epoch": 0.15258487654320987,
      "grad_norm": 1.434312343597412,
      "learning_rate": 0.0005,
      "loss": 3.2064709663391113,
      "step": 791
    },
    {
      "epoch": 0.1527777777777778,
      "grad_norm": 1.1712307929992676,
      "learning_rate": 0.0005,
      "loss": 3.231100082397461,
      "step": 792
    },
    {
      "epoch": 0.15297067901234568,
      "grad_norm": 0.9826122522354126,
      "learning_rate": 0.0005,
      "loss": 3.5673608779907227,
      "step": 793
    },
    {
      "epoch": 0.15316358024691357,
      "grad_norm": 1.2648638486862183,
      "learning_rate": 0.0005,
      "loss": 2.930647850036621,
      "step": 794
    },
    {
      "epoch": 0.15335648148148148,
      "grad_norm": 1.2020974159240723,
      "learning_rate": 0.0005,
      "loss": 3.3211708068847656,
      "step": 795
    },
    {
      "epoch": 0.15354938271604937,
      "grad_norm": 1.2515634298324585,
      "learning_rate": 0.0005,
      "loss": 3.5085766315460205,
      "step": 796
    },
    {
      "epoch": 0.1537422839506173,
      "grad_norm": 1.4619686603546143,
      "learning_rate": 0.0005,
      "loss": 3.7350656986236572,
      "step": 797
    },
    {
      "epoch": 0.15393518518518517,
      "grad_norm": 1.3501040935516357,
      "learning_rate": 0.0005,
      "loss": 3.4571971893310547,
      "step": 798
    },
    {
      "epoch": 0.1541280864197531,
      "grad_norm": 1.1056160926818848,
      "learning_rate": 0.0005,
      "loss": 3.505061626434326,
      "step": 799
    },
    {
      "epoch": 0.15432098765432098,
      "grad_norm": 1.334870457649231,
      "learning_rate": 0.0005,
      "loss": 3.3541717529296875,
      "step": 800
    },
    {
      "epoch": 0.1545138888888889,
      "grad_norm": 1.1468775272369385,
      "learning_rate": 0.0005,
      "loss": 3.245115041732788,
      "step": 801
    },
    {
      "epoch": 0.15470679012345678,
      "grad_norm": 1.2514008283615112,
      "learning_rate": 0.0005,
      "loss": 3.555830478668213,
      "step": 802
    },
    {
      "epoch": 0.1548996913580247,
      "grad_norm": 1.3747822046279907,
      "learning_rate": 0.0005,
      "loss": 3.166449785232544,
      "step": 803
    },
    {
      "epoch": 0.1550925925925926,
      "grad_norm": 1.3374203443527222,
      "learning_rate": 0.0005,
      "loss": 3.175504207611084,
      "step": 804
    },
    {
      "epoch": 0.1552854938271605,
      "grad_norm": 1.1664438247680664,
      "learning_rate": 0.0005,
      "loss": 3.3673782348632812,
      "step": 805
    },
    {
      "epoch": 0.1554783950617284,
      "grad_norm": 1.5782997608184814,
      "learning_rate": 0.0005,
      "loss": 3.0557122230529785,
      "step": 806
    },
    {
      "epoch": 0.1556712962962963,
      "grad_norm": 1.1880592107772827,
      "learning_rate": 0.0005,
      "loss": 3.4078493118286133,
      "step": 807
    },
    {
      "epoch": 0.1558641975308642,
      "grad_norm": 1.1230701208114624,
      "learning_rate": 0.0005,
      "loss": 3.309689521789551,
      "step": 808
    },
    {
      "epoch": 0.1560570987654321,
      "grad_norm": 1.0994583368301392,
      "learning_rate": 0.0005,
      "loss": 3.2158737182617188,
      "step": 809
    },
    {
      "epoch": 0.15625,
      "grad_norm": 1.193199872970581,
      "learning_rate": 0.0005,
      "loss": 3.367051601409912,
      "step": 810
    },
    {
      "epoch": 0.1564429012345679,
      "grad_norm": 1.1527177095413208,
      "learning_rate": 0.0005,
      "loss": 3.421417474746704,
      "step": 811
    },
    {
      "epoch": 0.1566358024691358,
      "grad_norm": 1.2850128412246704,
      "learning_rate": 0.0005,
      "loss": 3.049300193786621,
      "step": 812
    },
    {
      "epoch": 0.1568287037037037,
      "grad_norm": 1.2016037702560425,
      "learning_rate": 0.0005,
      "loss": 3.490446090698242,
      "step": 813
    },
    {
      "epoch": 0.1570216049382716,
      "grad_norm": 1.1377346515655518,
      "learning_rate": 0.0005,
      "loss": 2.9593119621276855,
      "step": 814
    },
    {
      "epoch": 0.1572145061728395,
      "grad_norm": 0.9996798634529114,
      "learning_rate": 0.0005,
      "loss": 3.4983227252960205,
      "step": 815
    },
    {
      "epoch": 0.1574074074074074,
      "grad_norm": 1.2589082717895508,
      "learning_rate": 0.0005,
      "loss": 3.4002952575683594,
      "step": 816
    },
    {
      "epoch": 0.1576003086419753,
      "grad_norm": 1.4035532474517822,
      "learning_rate": 0.0005,
      "loss": 3.8255584239959717,
      "step": 817
    },
    {
      "epoch": 0.15779320987654322,
      "grad_norm": 1.2501684427261353,
      "learning_rate": 0.0005,
      "loss": 3.523439407348633,
      "step": 818
    },
    {
      "epoch": 0.1579861111111111,
      "grad_norm": 1.354841947555542,
      "learning_rate": 0.0005,
      "loss": 3.1466777324676514,
      "step": 819
    },
    {
      "epoch": 0.15817901234567902,
      "grad_norm": 1.6346209049224854,
      "learning_rate": 0.0005,
      "loss": 3.160581350326538,
      "step": 820
    },
    {
      "epoch": 0.1583719135802469,
      "grad_norm": 1.2758642435073853,
      "learning_rate": 0.0005,
      "loss": 3.446786403656006,
      "step": 821
    },
    {
      "epoch": 0.15856481481481483,
      "grad_norm": 1.1655428409576416,
      "learning_rate": 0.0005,
      "loss": 3.3239195346832275,
      "step": 822
    },
    {
      "epoch": 0.1587577160493827,
      "grad_norm": 1.0425509214401245,
      "learning_rate": 0.0005,
      "loss": 3.7353053092956543,
      "step": 823
    },
    {
      "epoch": 0.15895061728395063,
      "grad_norm": 2.0466463565826416,
      "learning_rate": 0.0005,
      "loss": 3.2202649116516113,
      "step": 824
    },
    {
      "epoch": 0.15914351851851852,
      "grad_norm": 1.1932464838027954,
      "learning_rate": 0.0005,
      "loss": 3.2534165382385254,
      "step": 825
    },
    {
      "epoch": 0.15933641975308643,
      "grad_norm": 1.9586917161941528,
      "learning_rate": 0.0005,
      "loss": 3.2152905464172363,
      "step": 826
    },
    {
      "epoch": 0.15952932098765432,
      "grad_norm": 1.2594358921051025,
      "learning_rate": 0.0005,
      "loss": 3.4469237327575684,
      "step": 827
    },
    {
      "epoch": 0.1597222222222222,
      "grad_norm": 1.2640873193740845,
      "learning_rate": 0.0005,
      "loss": 3.223292350769043,
      "step": 828
    },
    {
      "epoch": 0.15991512345679013,
      "grad_norm": 1.0025591850280762,
      "learning_rate": 0.0005,
      "loss": 3.5070459842681885,
      "step": 829
    },
    {
      "epoch": 0.16010802469135801,
      "grad_norm": 1.1853339672088623,
      "learning_rate": 0.0005,
      "loss": 3.297248363494873,
      "step": 830
    },
    {
      "epoch": 0.16030092592592593,
      "grad_norm": 1.4233273267745972,
      "learning_rate": 0.0005,
      "loss": 3.232163906097412,
      "step": 831
    },
    {
      "epoch": 0.16049382716049382,
      "grad_norm": 2.928640365600586,
      "learning_rate": 0.0005,
      "loss": 3.3174729347229004,
      "step": 832
    },
    {
      "epoch": 0.16068672839506173,
      "grad_norm": 1.3288154602050781,
      "learning_rate": 0.0005,
      "loss": 3.5450096130371094,
      "step": 833
    },
    {
      "epoch": 0.16087962962962962,
      "grad_norm": 1.1503957509994507,
      "learning_rate": 0.0005,
      "loss": 3.291593074798584,
      "step": 834
    },
    {
      "epoch": 0.16107253086419754,
      "grad_norm": 1.8578245639801025,
      "learning_rate": 0.0005,
      "loss": 3.5890183448791504,
      "step": 835
    },
    {
      "epoch": 0.16126543209876543,
      "grad_norm": 1.0996836423873901,
      "learning_rate": 0.0005,
      "loss": 3.250011444091797,
      "step": 836
    },
    {
      "epoch": 0.16145833333333334,
      "grad_norm": 1.3537349700927734,
      "learning_rate": 0.0005,
      "loss": 3.123889923095703,
      "step": 837
    },
    {
      "epoch": 0.16165123456790123,
      "grad_norm": 1.0704238414764404,
      "learning_rate": 0.0005,
      "loss": 3.528759241104126,
      "step": 838
    },
    {
      "epoch": 0.16184413580246915,
      "grad_norm": 1.3330707550048828,
      "learning_rate": 0.0005,
      "loss": 2.8792810440063477,
      "step": 839
    },
    {
      "epoch": 0.16203703703703703,
      "grad_norm": 2.39770245552063,
      "learning_rate": 0.0005,
      "loss": 3.5222079753875732,
      "step": 840
    },
    {
      "epoch": 0.16222993827160495,
      "grad_norm": 1.7253334522247314,
      "learning_rate": 0.0005,
      "loss": 3.57205867767334,
      "step": 841
    },
    {
      "epoch": 0.16242283950617284,
      "grad_norm": 1.313813328742981,
      "learning_rate": 0.0005,
      "loss": 3.098656177520752,
      "step": 842
    },
    {
      "epoch": 0.16261574074074073,
      "grad_norm": 1.376001238822937,
      "learning_rate": 0.0005,
      "loss": 3.295229911804199,
      "step": 843
    },
    {
      "epoch": 0.16280864197530864,
      "grad_norm": 1.0871268510818481,
      "learning_rate": 0.0005,
      "loss": 3.605149745941162,
      "step": 844
    },
    {
      "epoch": 0.16300154320987653,
      "grad_norm": 1.303684115409851,
      "learning_rate": 0.0005,
      "loss": 3.336885929107666,
      "step": 845
    },
    {
      "epoch": 0.16319444444444445,
      "grad_norm": 1.2880414724349976,
      "learning_rate": 0.0005,
      "loss": 3.401613712310791,
      "step": 846
    },
    {
      "epoch": 0.16338734567901234,
      "grad_norm": 1.2178229093551636,
      "learning_rate": 0.0005,
      "loss": 3.22833251953125,
      "step": 847
    },
    {
      "epoch": 0.16358024691358025,
      "grad_norm": 1.1216338872909546,
      "learning_rate": 0.0005,
      "loss": 3.3505611419677734,
      "step": 848
    },
    {
      "epoch": 0.16377314814814814,
      "grad_norm": 1.0083426237106323,
      "learning_rate": 0.0005,
      "loss": 3.589815139770508,
      "step": 849
    },
    {
      "epoch": 0.16396604938271606,
      "grad_norm": 1.502934455871582,
      "learning_rate": 0.0005,
      "loss": 3.9343628883361816,
      "step": 850
    },
    {
      "epoch": 0.16415895061728394,
      "grad_norm": 1.1983362436294556,
      "learning_rate": 0.0005,
      "loss": 3.2188315391540527,
      "step": 851
    },
    {
      "epoch": 0.16435185185185186,
      "grad_norm": 1.711596131324768,
      "learning_rate": 0.0005,
      "loss": 3.4222302436828613,
      "step": 852
    },
    {
      "epoch": 0.16454475308641975,
      "grad_norm": 1.1458922624588013,
      "learning_rate": 0.0005,
      "loss": 3.7651617527008057,
      "step": 853
    },
    {
      "epoch": 0.16473765432098766,
      "grad_norm": 1.1364634037017822,
      "learning_rate": 0.0005,
      "loss": 3.0473625659942627,
      "step": 854
    },
    {
      "epoch": 0.16493055555555555,
      "grad_norm": 1.2420090436935425,
      "learning_rate": 0.0005,
      "loss": 3.6208925247192383,
      "step": 855
    },
    {
      "epoch": 0.16512345679012347,
      "grad_norm": 1.5966483354568481,
      "learning_rate": 0.0005,
      "loss": 3.6316871643066406,
      "step": 856
    },
    {
      "epoch": 0.16531635802469136,
      "grad_norm": 1.3465375900268555,
      "learning_rate": 0.0005,
      "loss": 3.410291910171509,
      "step": 857
    },
    {
      "epoch": 0.16550925925925927,
      "grad_norm": 1.5699105262756348,
      "learning_rate": 0.0005,
      "loss": 3.22046160697937,
      "step": 858
    },
    {
      "epoch": 0.16570216049382716,
      "grad_norm": 1.3891017436981201,
      "learning_rate": 0.0005,
      "loss": 3.526592254638672,
      "step": 859
    },
    {
      "epoch": 0.16589506172839505,
      "grad_norm": 1.1718829870224,
      "learning_rate": 0.0005,
      "loss": 3.281425952911377,
      "step": 860
    },
    {
      "epoch": 0.16608796296296297,
      "grad_norm": 1.3261693716049194,
      "learning_rate": 0.0005,
      "loss": 3.4292736053466797,
      "step": 861
    },
    {
      "epoch": 0.16628086419753085,
      "grad_norm": 1.5975664854049683,
      "learning_rate": 0.0005,
      "loss": 3.475996494293213,
      "step": 862
    },
    {
      "epoch": 0.16647376543209877,
      "grad_norm": 1.6766153573989868,
      "learning_rate": 0.0005,
      "loss": 3.268036127090454,
      "step": 863
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 1.4152621030807495,
      "learning_rate": 0.0005,
      "loss": 3.3261499404907227,
      "step": 864
    },
    {
      "epoch": 0.16685956790123457,
      "grad_norm": 1.476380467414856,
      "learning_rate": 0.0005,
      "loss": 3.3849592208862305,
      "step": 865
    },
    {
      "epoch": 0.16705246913580246,
      "grad_norm": 1.1906507015228271,
      "learning_rate": 0.0005,
      "loss": 3.1733264923095703,
      "step": 866
    },
    {
      "epoch": 0.16724537037037038,
      "grad_norm": 1.196179986000061,
      "learning_rate": 0.0005,
      "loss": 3.4221949577331543,
      "step": 867
    },
    {
      "epoch": 0.16743827160493827,
      "grad_norm": 1.1107600927352905,
      "learning_rate": 0.0005,
      "loss": 3.113680839538574,
      "step": 868
    },
    {
      "epoch": 0.16763117283950618,
      "grad_norm": 1.422824501991272,
      "learning_rate": 0.0005,
      "loss": 3.5296072959899902,
      "step": 869
    },
    {
      "epoch": 0.16782407407407407,
      "grad_norm": 1.2267107963562012,
      "learning_rate": 0.0005,
      "loss": 3.4647023677825928,
      "step": 870
    },
    {
      "epoch": 0.16801697530864199,
      "grad_norm": 1.206288456916809,
      "learning_rate": 0.0005,
      "loss": 3.6171443462371826,
      "step": 871
    },
    {
      "epoch": 0.16820987654320987,
      "grad_norm": 1.2737311124801636,
      "learning_rate": 0.0005,
      "loss": 3.2110753059387207,
      "step": 872
    },
    {
      "epoch": 0.1684027777777778,
      "grad_norm": 1.1107686758041382,
      "learning_rate": 0.0005,
      "loss": 3.6240928173065186,
      "step": 873
    },
    {
      "epoch": 0.16859567901234568,
      "grad_norm": 1.7959198951721191,
      "learning_rate": 0.0005,
      "loss": 3.3719630241394043,
      "step": 874
    },
    {
      "epoch": 0.16878858024691357,
      "grad_norm": 1.7297089099884033,
      "learning_rate": 0.0005,
      "loss": 3.308419704437256,
      "step": 875
    },
    {
      "epoch": 0.16898148148148148,
      "grad_norm": 1.3124604225158691,
      "learning_rate": 0.0005,
      "loss": 3.392613649368286,
      "step": 876
    },
    {
      "epoch": 0.16917438271604937,
      "grad_norm": 1.2012425661087036,
      "learning_rate": 0.0005,
      "loss": 3.332406997680664,
      "step": 877
    },
    {
      "epoch": 0.1693672839506173,
      "grad_norm": 1.4161018133163452,
      "learning_rate": 0.0005,
      "loss": 3.3640003204345703,
      "step": 878
    },
    {
      "epoch": 0.16956018518518517,
      "grad_norm": 1.2445167303085327,
      "learning_rate": 0.0005,
      "loss": 3.510707378387451,
      "step": 879
    },
    {
      "epoch": 0.1697530864197531,
      "grad_norm": 1.308907389640808,
      "learning_rate": 0.0005,
      "loss": 3.101559638977051,
      "step": 880
    },
    {
      "epoch": 0.16994598765432098,
      "grad_norm": 1.2668870687484741,
      "learning_rate": 0.0005,
      "loss": 3.073275089263916,
      "step": 881
    },
    {
      "epoch": 0.1701388888888889,
      "grad_norm": 1.1182560920715332,
      "learning_rate": 0.0005,
      "loss": 3.6900761127471924,
      "step": 882
    },
    {
      "epoch": 0.17033179012345678,
      "grad_norm": 1.2988338470458984,
      "learning_rate": 0.0005,
      "loss": 3.8760433197021484,
      "step": 883
    },
    {
      "epoch": 0.1705246913580247,
      "grad_norm": 1.1134073734283447,
      "learning_rate": 0.0005,
      "loss": 3.6526012420654297,
      "step": 884
    },
    {
      "epoch": 0.1707175925925926,
      "grad_norm": 1.9384232759475708,
      "learning_rate": 0.0005,
      "loss": 3.4420628547668457,
      "step": 885
    },
    {
      "epoch": 0.1709104938271605,
      "grad_norm": 1.4820600748062134,
      "learning_rate": 0.0005,
      "loss": 3.2929630279541016,
      "step": 886
    },
    {
      "epoch": 0.1711033950617284,
      "grad_norm": 1.1277427673339844,
      "learning_rate": 0.0005,
      "loss": 3.4197916984558105,
      "step": 887
    },
    {
      "epoch": 0.1712962962962963,
      "grad_norm": 3.2213430404663086,
      "learning_rate": 0.0005,
      "loss": 3.1157593727111816,
      "step": 888
    },
    {
      "epoch": 0.1714891975308642,
      "grad_norm": 1.3912584781646729,
      "learning_rate": 0.0005,
      "loss": 3.206145763397217,
      "step": 889
    },
    {
      "epoch": 0.1716820987654321,
      "grad_norm": 1.287416934967041,
      "learning_rate": 0.0005,
      "loss": 3.178508758544922,
      "step": 890
    },
    {
      "epoch": 0.171875,
      "grad_norm": 1.0386179685592651,
      "learning_rate": 0.0005,
      "loss": 3.4025399684906006,
      "step": 891
    },
    {
      "epoch": 0.1720679012345679,
      "grad_norm": 1.6728167533874512,
      "learning_rate": 0.0005,
      "loss": 3.583883762359619,
      "step": 892
    },
    {
      "epoch": 0.1722608024691358,
      "grad_norm": 1.2906385660171509,
      "learning_rate": 0.0005,
      "loss": 2.988706350326538,
      "step": 893
    },
    {
      "epoch": 0.1724537037037037,
      "grad_norm": 1.0413715839385986,
      "learning_rate": 0.0005,
      "loss": 3.33872127532959,
      "step": 894
    },
    {
      "epoch": 0.1726466049382716,
      "grad_norm": 1.3624870777130127,
      "learning_rate": 0.0005,
      "loss": 3.433851718902588,
      "step": 895
    },
    {
      "epoch": 0.1728395061728395,
      "grad_norm": 1.165010929107666,
      "learning_rate": 0.0005,
      "loss": 3.4489262104034424,
      "step": 896
    },
    {
      "epoch": 0.1730324074074074,
      "grad_norm": 1.8105281591415405,
      "learning_rate": 0.0005,
      "loss": 3.074244737625122,
      "step": 897
    },
    {
      "epoch": 0.1732253086419753,
      "grad_norm": 1.6734081506729126,
      "learning_rate": 0.0005,
      "loss": 3.3021552562713623,
      "step": 898
    },
    {
      "epoch": 0.17341820987654322,
      "grad_norm": 1.5166614055633545,
      "learning_rate": 0.0005,
      "loss": 3.074110507965088,
      "step": 899
    },
    {
      "epoch": 0.1736111111111111,
      "grad_norm": 1.1295934915542603,
      "learning_rate": 0.0005,
      "loss": 3.3670248985290527,
      "step": 900
    },
    {
      "epoch": 0.17380401234567902,
      "grad_norm": 1.715159296989441,
      "learning_rate": 0.0005,
      "loss": 3.8971381187438965,
      "step": 901
    },
    {
      "epoch": 0.1739969135802469,
      "grad_norm": 1.3260111808776855,
      "learning_rate": 0.0005,
      "loss": 3.6411986351013184,
      "step": 902
    },
    {
      "epoch": 0.17418981481481483,
      "grad_norm": 1.3372880220413208,
      "learning_rate": 0.0005,
      "loss": 3.5060787200927734,
      "step": 903
    },
    {
      "epoch": 0.1743827160493827,
      "grad_norm": 1.3664065599441528,
      "learning_rate": 0.0005,
      "loss": 3.204319953918457,
      "step": 904
    },
    {
      "epoch": 0.17457561728395063,
      "grad_norm": 1.4493416547775269,
      "learning_rate": 0.0005,
      "loss": 3.5378594398498535,
      "step": 905
    },
    {
      "epoch": 0.17476851851851852,
      "grad_norm": 1.2324126958847046,
      "learning_rate": 0.0005,
      "loss": 3.147291421890259,
      "step": 906
    },
    {
      "epoch": 0.17496141975308643,
      "grad_norm": 1.0665885210037231,
      "learning_rate": 0.0005,
      "loss": 3.378098487854004,
      "step": 907
    },
    {
      "epoch": 0.17515432098765432,
      "grad_norm": 1.078203558921814,
      "learning_rate": 0.0005,
      "loss": 3.76106595993042,
      "step": 908
    },
    {
      "epoch": 0.1753472222222222,
      "grad_norm": 1.0546363592147827,
      "learning_rate": 0.0005,
      "loss": 2.9920694828033447,
      "step": 909
    },
    {
      "epoch": 0.17554012345679013,
      "grad_norm": 1.2410050630569458,
      "learning_rate": 0.0005,
      "loss": 3.4154276847839355,
      "step": 910
    },
    {
      "epoch": 0.17573302469135801,
      "grad_norm": 1.0361207723617554,
      "learning_rate": 0.0005,
      "loss": 2.9965739250183105,
      "step": 911
    },
    {
      "epoch": 0.17592592592592593,
      "grad_norm": 1.060228943824768,
      "learning_rate": 0.0005,
      "loss": 3.5583882331848145,
      "step": 912
    },
    {
      "epoch": 0.17611882716049382,
      "grad_norm": 1.1804625988006592,
      "learning_rate": 0.0005,
      "loss": 3.585068702697754,
      "step": 913
    },
    {
      "epoch": 0.17631172839506173,
      "grad_norm": 1.406512975692749,
      "learning_rate": 0.0005,
      "loss": 3.370582342147827,
      "step": 914
    },
    {
      "epoch": 0.17650462962962962,
      "grad_norm": 1.14021635055542,
      "learning_rate": 0.0005,
      "loss": 3.0889978408813477,
      "step": 915
    },
    {
      "epoch": 0.17669753086419754,
      "grad_norm": 1.611315131187439,
      "learning_rate": 0.0005,
      "loss": 3.2929587364196777,
      "step": 916
    },
    {
      "epoch": 0.17689043209876543,
      "grad_norm": 1.407925009727478,
      "learning_rate": 0.0005,
      "loss": 3.359138011932373,
      "step": 917
    },
    {
      "epoch": 0.17708333333333334,
      "grad_norm": 1.475178599357605,
      "learning_rate": 0.0005,
      "loss": 3.1570987701416016,
      "step": 918
    },
    {
      "epoch": 0.17727623456790123,
      "grad_norm": 1.3879460096359253,
      "learning_rate": 0.0005,
      "loss": 3.1978721618652344,
      "step": 919
    },
    {
      "epoch": 0.17746913580246915,
      "grad_norm": 1.2455979585647583,
      "learning_rate": 0.0005,
      "loss": 3.2666687965393066,
      "step": 920
    },
    {
      "epoch": 0.17766203703703703,
      "grad_norm": 1.172126054763794,
      "learning_rate": 0.0005,
      "loss": 3.256964683532715,
      "step": 921
    },
    {
      "epoch": 0.17785493827160495,
      "grad_norm": 1.0848896503448486,
      "learning_rate": 0.0005,
      "loss": 3.3211162090301514,
      "step": 922
    },
    {
      "epoch": 0.17804783950617284,
      "grad_norm": 1.6000982522964478,
      "learning_rate": 0.0005,
      "loss": 3.2749838829040527,
      "step": 923
    },
    {
      "epoch": 0.17824074074074073,
      "grad_norm": 2.2100212574005127,
      "learning_rate": 0.0005,
      "loss": 3.403395652770996,
      "step": 924
    },
    {
      "epoch": 0.17843364197530864,
      "grad_norm": 1.9888906478881836,
      "learning_rate": 0.0005,
      "loss": 3.3668909072875977,
      "step": 925
    },
    {
      "epoch": 0.17862654320987653,
      "grad_norm": 1.8041236400604248,
      "learning_rate": 0.0005,
      "loss": 3.3797879219055176,
      "step": 926
    },
    {
      "epoch": 0.17881944444444445,
      "grad_norm": 1.9527931213378906,
      "learning_rate": 0.0005,
      "loss": 3.457714557647705,
      "step": 927
    },
    {
      "epoch": 0.17901234567901234,
      "grad_norm": 1.1940783262252808,
      "learning_rate": 0.0005,
      "loss": 3.3810038566589355,
      "step": 928
    },
    {
      "epoch": 0.17920524691358025,
      "grad_norm": 1.1181375980377197,
      "learning_rate": 0.0005,
      "loss": 3.5727245807647705,
      "step": 929
    },
    {
      "epoch": 0.17939814814814814,
      "grad_norm": 1.299424171447754,
      "learning_rate": 0.0005,
      "loss": 3.487699031829834,
      "step": 930
    },
    {
      "epoch": 0.17959104938271606,
      "grad_norm": 1.1589274406433105,
      "learning_rate": 0.0005,
      "loss": 3.342886447906494,
      "step": 931
    },
    {
      "epoch": 0.17978395061728394,
      "grad_norm": 0.9986428022384644,
      "learning_rate": 0.0005,
      "loss": 3.365297317504883,
      "step": 932
    },
    {
      "epoch": 0.17997685185185186,
      "grad_norm": 1.785826563835144,
      "learning_rate": 0.0005,
      "loss": 3.18237042427063,
      "step": 933
    },
    {
      "epoch": 0.18016975308641975,
      "grad_norm": 1.5919358730316162,
      "learning_rate": 0.0005,
      "loss": 3.3270351886749268,
      "step": 934
    },
    {
      "epoch": 0.18036265432098766,
      "grad_norm": 1.3170785903930664,
      "learning_rate": 0.0005,
      "loss": 3.200207471847534,
      "step": 935
    },
    {
      "epoch": 0.18055555555555555,
      "grad_norm": 1.3487309217453003,
      "learning_rate": 0.0005,
      "loss": 3.2980101108551025,
      "step": 936
    },
    {
      "epoch": 0.18074845679012347,
      "grad_norm": 1.1325939893722534,
      "learning_rate": 0.0005,
      "loss": 3.2706520557403564,
      "step": 937
    },
    {
      "epoch": 0.18094135802469136,
      "grad_norm": 1.117348074913025,
      "learning_rate": 0.0005,
      "loss": 3.5204272270202637,
      "step": 938
    },
    {
      "epoch": 0.18113425925925927,
      "grad_norm": 1.2248374223709106,
      "learning_rate": 0.0005,
      "loss": 3.49800968170166,
      "step": 939
    },
    {
      "epoch": 0.18132716049382716,
      "grad_norm": 1.1287109851837158,
      "learning_rate": 0.0005,
      "loss": 3.5565104484558105,
      "step": 940
    },
    {
      "epoch": 0.18152006172839505,
      "grad_norm": 1.1593806743621826,
      "learning_rate": 0.0005,
      "loss": 3.360276699066162,
      "step": 941
    },
    {
      "epoch": 0.18171296296296297,
      "grad_norm": 1.4642618894577026,
      "learning_rate": 0.0005,
      "loss": 3.1800625324249268,
      "step": 942
    },
    {
      "epoch": 0.18190586419753085,
      "grad_norm": 1.258664846420288,
      "learning_rate": 0.0005,
      "loss": 3.252716064453125,
      "step": 943
    },
    {
      "epoch": 0.18209876543209877,
      "grad_norm": 1.2424207925796509,
      "learning_rate": 0.0005,
      "loss": 3.2662277221679688,
      "step": 944
    },
    {
      "epoch": 0.18229166666666666,
      "grad_norm": 1.261871576309204,
      "learning_rate": 0.0005,
      "loss": 3.414381504058838,
      "step": 945
    },
    {
      "epoch": 0.18248456790123457,
      "grad_norm": 1.2175310850143433,
      "learning_rate": 0.0005,
      "loss": 3.5596201419830322,
      "step": 946
    },
    {
      "epoch": 0.18267746913580246,
      "grad_norm": 2.0341598987579346,
      "learning_rate": 0.0005,
      "loss": 3.35841703414917,
      "step": 947
    },
    {
      "epoch": 0.18287037037037038,
      "grad_norm": 1.5069689750671387,
      "learning_rate": 0.0005,
      "loss": 3.572615146636963,
      "step": 948
    },
    {
      "epoch": 0.18306327160493827,
      "grad_norm": 1.2331359386444092,
      "learning_rate": 0.0005,
      "loss": 3.390082836151123,
      "step": 949
    },
    {
      "epoch": 0.18325617283950618,
      "grad_norm": 1.0868357419967651,
      "learning_rate": 0.0005,
      "loss": 3.1287994384765625,
      "step": 950
    },
    {
      "epoch": 0.18344907407407407,
      "grad_norm": 1.3199561834335327,
      "learning_rate": 0.0005,
      "loss": 3.250819444656372,
      "step": 951
    },
    {
      "epoch": 0.18364197530864199,
      "grad_norm": 1.2340681552886963,
      "learning_rate": 0.0005,
      "loss": 3.2963788509368896,
      "step": 952
    },
    {
      "epoch": 0.18383487654320987,
      "grad_norm": 1.207363486289978,
      "learning_rate": 0.0005,
      "loss": 3.2131550312042236,
      "step": 953
    },
    {
      "epoch": 0.1840277777777778,
      "grad_norm": 1.2569917440414429,
      "learning_rate": 0.0005,
      "loss": 3.337062120437622,
      "step": 954
    },
    {
      "epoch": 0.18422067901234568,
      "grad_norm": 1.5838509798049927,
      "learning_rate": 0.0005,
      "loss": 3.3176612854003906,
      "step": 955
    },
    {
      "epoch": 0.18441358024691357,
      "grad_norm": 0.890523374080658,
      "learning_rate": 0.0005,
      "loss": 3.2932024002075195,
      "step": 956
    },
    {
      "epoch": 0.18460648148148148,
      "grad_norm": 1.2490174770355225,
      "learning_rate": 0.0005,
      "loss": 3.2957916259765625,
      "step": 957
    },
    {
      "epoch": 0.18479938271604937,
      "grad_norm": 1.4013859033584595,
      "learning_rate": 0.0005,
      "loss": 3.166425943374634,
      "step": 958
    },
    {
      "epoch": 0.1849922839506173,
      "grad_norm": 1.078128695487976,
      "learning_rate": 0.0005,
      "loss": 3.292290210723877,
      "step": 959
    },
    {
      "epoch": 0.18518518518518517,
      "grad_norm": 1.1108452081680298,
      "learning_rate": 0.0005,
      "loss": 2.9591991901397705,
      "step": 960
    },
    {
      "epoch": 0.1853780864197531,
      "grad_norm": 1.1784850358963013,
      "learning_rate": 0.0005,
      "loss": 3.286921739578247,
      "step": 961
    },
    {
      "epoch": 0.18557098765432098,
      "grad_norm": 1.2467892169952393,
      "learning_rate": 0.0005,
      "loss": 3.306093692779541,
      "step": 962
    },
    {
      "epoch": 0.1857638888888889,
      "grad_norm": 1.5301681756973267,
      "learning_rate": 0.0005,
      "loss": 3.2669925689697266,
      "step": 963
    },
    {
      "epoch": 0.18595679012345678,
      "grad_norm": 1.0167135000228882,
      "learning_rate": 0.0005,
      "loss": 3.3532145023345947,
      "step": 964
    },
    {
      "epoch": 0.1861496913580247,
      "grad_norm": 0.9865038990974426,
      "learning_rate": 0.0005,
      "loss": 3.4238638877868652,
      "step": 965
    },
    {
      "epoch": 0.1863425925925926,
      "grad_norm": 1.400892972946167,
      "learning_rate": 0.0005,
      "loss": 3.2019121646881104,
      "step": 966
    },
    {
      "epoch": 0.1865354938271605,
      "grad_norm": 1.1249891519546509,
      "learning_rate": 0.0005,
      "loss": 3.1229777336120605,
      "step": 967
    },
    {
      "epoch": 0.1867283950617284,
      "grad_norm": 1.1429269313812256,
      "learning_rate": 0.0005,
      "loss": 3.2696382999420166,
      "step": 968
    },
    {
      "epoch": 0.1869212962962963,
      "grad_norm": 1.0530778169631958,
      "learning_rate": 0.0005,
      "loss": 3.031635284423828,
      "step": 969
    },
    {
      "epoch": 0.1871141975308642,
      "grad_norm": 1.0404366254806519,
      "learning_rate": 0.0005,
      "loss": 3.1488449573516846,
      "step": 970
    },
    {
      "epoch": 0.1873070987654321,
      "grad_norm": 1.5809282064437866,
      "learning_rate": 0.0005,
      "loss": 3.380603790283203,
      "step": 971
    },
    {
      "epoch": 0.1875,
      "grad_norm": 1.135827660560608,
      "learning_rate": 0.0005,
      "loss": 3.228616237640381,
      "step": 972
    },
    {
      "epoch": 0.1876929012345679,
      "grad_norm": 1.289237141609192,
      "learning_rate": 0.0005,
      "loss": 3.2279040813446045,
      "step": 973
    },
    {
      "epoch": 0.1878858024691358,
      "grad_norm": 1.1520103216171265,
      "learning_rate": 0.0005,
      "loss": 3.4567906856536865,
      "step": 974
    },
    {
      "epoch": 0.1880787037037037,
      "grad_norm": 1.4858827590942383,
      "learning_rate": 0.0005,
      "loss": 3.2464990615844727,
      "step": 975
    },
    {
      "epoch": 0.1882716049382716,
      "grad_norm": 1.2899127006530762,
      "learning_rate": 0.0005,
      "loss": 3.4816222190856934,
      "step": 976
    },
    {
      "epoch": 0.1884645061728395,
      "grad_norm": 1.1016758680343628,
      "learning_rate": 0.0005,
      "loss": 3.4943108558654785,
      "step": 977
    },
    {
      "epoch": 0.1886574074074074,
      "grad_norm": 1.1237192153930664,
      "learning_rate": 0.0005,
      "loss": 3.185591220855713,
      "step": 978
    },
    {
      "epoch": 0.1888503086419753,
      "grad_norm": 1.018905520439148,
      "learning_rate": 0.0005,
      "loss": 3.4043805599212646,
      "step": 979
    },
    {
      "epoch": 0.18904320987654322,
      "grad_norm": 1.4134433269500732,
      "learning_rate": 0.0005,
      "loss": 3.3478007316589355,
      "step": 980
    },
    {
      "epoch": 0.1892361111111111,
      "grad_norm": 1.1910099983215332,
      "learning_rate": 0.0005,
      "loss": 3.4046854972839355,
      "step": 981
    },
    {
      "epoch": 0.18942901234567902,
      "grad_norm": 1.2819136381149292,
      "learning_rate": 0.0005,
      "loss": 3.5479280948638916,
      "step": 982
    },
    {
      "epoch": 0.1896219135802469,
      "grad_norm": 1.1353487968444824,
      "learning_rate": 0.0005,
      "loss": 3.209480047225952,
      "step": 983
    },
    {
      "epoch": 0.18981481481481483,
      "grad_norm": 1.4068896770477295,
      "learning_rate": 0.0005,
      "loss": 3.7169032096862793,
      "step": 984
    },
    {
      "epoch": 0.1900077160493827,
      "grad_norm": 1.119163155555725,
      "learning_rate": 0.0005,
      "loss": 3.217240810394287,
      "step": 985
    },
    {
      "epoch": 0.19020061728395063,
      "grad_norm": 1.218953013420105,
      "learning_rate": 0.0005,
      "loss": 3.590144157409668,
      "step": 986
    },
    {
      "epoch": 0.19039351851851852,
      "grad_norm": 1.2800594568252563,
      "learning_rate": 0.0005,
      "loss": 2.8737587928771973,
      "step": 987
    },
    {
      "epoch": 0.19058641975308643,
      "grad_norm": 1.1188737154006958,
      "learning_rate": 0.0005,
      "loss": 3.077705144882202,
      "step": 988
    },
    {
      "epoch": 0.19077932098765432,
      "grad_norm": 1.3743950128555298,
      "learning_rate": 0.0005,
      "loss": 3.293482780456543,
      "step": 989
    },
    {
      "epoch": 0.1909722222222222,
      "grad_norm": 1.2457689046859741,
      "learning_rate": 0.0005,
      "loss": 3.6088318824768066,
      "step": 990
    },
    {
      "epoch": 0.19116512345679013,
      "grad_norm": 1.1580002307891846,
      "learning_rate": 0.0005,
      "loss": 3.0749125480651855,
      "step": 991
    },
    {
      "epoch": 0.19135802469135801,
      "grad_norm": 1.0500259399414062,
      "learning_rate": 0.0005,
      "loss": 3.3573668003082275,
      "step": 992
    },
    {
      "epoch": 0.19155092592592593,
      "grad_norm": 1.4822298288345337,
      "learning_rate": 0.0005,
      "loss": 3.512967586517334,
      "step": 993
    },
    {
      "epoch": 0.19174382716049382,
      "grad_norm": 1.0511717796325684,
      "learning_rate": 0.0005,
      "loss": 3.0235953330993652,
      "step": 994
    },
    {
      "epoch": 0.19193672839506173,
      "grad_norm": 1.2501689195632935,
      "learning_rate": 0.0005,
      "loss": 3.3619284629821777,
      "step": 995
    },
    {
      "epoch": 0.19212962962962962,
      "grad_norm": 1.0996373891830444,
      "learning_rate": 0.0005,
      "loss": 3.04691219329834,
      "step": 996
    },
    {
      "epoch": 0.19232253086419754,
      "grad_norm": 1.2716176509857178,
      "learning_rate": 0.0005,
      "loss": 3.2761611938476562,
      "step": 997
    },
    {
      "epoch": 0.19251543209876543,
      "grad_norm": 1.1524982452392578,
      "learning_rate": 0.0005,
      "loss": 2.9880356788635254,
      "step": 998
    },
    {
      "epoch": 0.19270833333333334,
      "grad_norm": 1.1525171995162964,
      "learning_rate": 0.0005,
      "loss": 2.876267194747925,
      "step": 999
    },
    {
      "epoch": 0.19290123456790123,
      "grad_norm": 1.063979983329773,
      "learning_rate": 0.0005,
      "loss": 3.079962730407715,
      "step": 1000
    },
    {
      "epoch": 0.19309413580246915,
      "grad_norm": 1.0922775268554688,
      "learning_rate": 0.0005,
      "loss": 3.2547988891601562,
      "step": 1001
    },
    {
      "epoch": 0.19328703703703703,
      "grad_norm": 1.0490047931671143,
      "learning_rate": 0.0005,
      "loss": 3.3897862434387207,
      "step": 1002
    },
    {
      "epoch": 0.19347993827160495,
      "grad_norm": 1.075068473815918,
      "learning_rate": 0.0005,
      "loss": 2.946967601776123,
      "step": 1003
    },
    {
      "epoch": 0.19367283950617284,
      "grad_norm": 1.4014204740524292,
      "learning_rate": 0.0005,
      "loss": 3.201106071472168,
      "step": 1004
    },
    {
      "epoch": 0.19386574074074073,
      "grad_norm": 1.231949806213379,
      "learning_rate": 0.0005,
      "loss": 3.344597339630127,
      "step": 1005
    },
    {
      "epoch": 0.19405864197530864,
      "grad_norm": 1.2582616806030273,
      "learning_rate": 0.0005,
      "loss": 3.2150096893310547,
      "step": 1006
    },
    {
      "epoch": 0.19425154320987653,
      "grad_norm": 1.4000601768493652,
      "learning_rate": 0.0005,
      "loss": 2.998565673828125,
      "step": 1007
    },
    {
      "epoch": 0.19444444444444445,
      "grad_norm": 1.5718485116958618,
      "learning_rate": 0.0005,
      "loss": 3.231712818145752,
      "step": 1008
    },
    {
      "epoch": 0.19463734567901234,
      "grad_norm": 1.1060149669647217,
      "learning_rate": 0.0005,
      "loss": 3.0126399993896484,
      "step": 1009
    },
    {
      "epoch": 0.19483024691358025,
      "grad_norm": 1.197762370109558,
      "learning_rate": 0.0005,
      "loss": 3.165226936340332,
      "step": 1010
    },
    {
      "epoch": 0.19502314814814814,
      "grad_norm": 1.2248815298080444,
      "learning_rate": 0.0005,
      "loss": 3.4603655338287354,
      "step": 1011
    },
    {
      "epoch": 0.19521604938271606,
      "grad_norm": 1.5028691291809082,
      "learning_rate": 0.0005,
      "loss": 3.11777400970459,
      "step": 1012
    },
    {
      "epoch": 0.19540895061728394,
      "grad_norm": 1.2531214952468872,
      "learning_rate": 0.0005,
      "loss": 2.994872570037842,
      "step": 1013
    },
    {
      "epoch": 0.19560185185185186,
      "grad_norm": 1.6225529909133911,
      "learning_rate": 0.0005,
      "loss": 3.497903823852539,
      "step": 1014
    },
    {
      "epoch": 0.19579475308641975,
      "grad_norm": 1.7183877229690552,
      "learning_rate": 0.0005,
      "loss": 3.2817986011505127,
      "step": 1015
    },
    {
      "epoch": 0.19598765432098766,
      "grad_norm": 1.1714389324188232,
      "learning_rate": 0.0005,
      "loss": 3.3364412784576416,
      "step": 1016
    },
    {
      "epoch": 0.19618055555555555,
      "grad_norm": 1.1211066246032715,
      "learning_rate": 0.0005,
      "loss": 3.165369987487793,
      "step": 1017
    },
    {
      "epoch": 0.19637345679012347,
      "grad_norm": 1.3419688940048218,
      "learning_rate": 0.0005,
      "loss": 3.1688880920410156,
      "step": 1018
    },
    {
      "epoch": 0.19656635802469136,
      "grad_norm": 1.2756353616714478,
      "learning_rate": 0.0005,
      "loss": 3.3435730934143066,
      "step": 1019
    },
    {
      "epoch": 0.19675925925925927,
      "grad_norm": 1.129713773727417,
      "learning_rate": 0.0005,
      "loss": 3.199223279953003,
      "step": 1020
    },
    {
      "epoch": 0.19695216049382716,
      "grad_norm": 1.2185735702514648,
      "learning_rate": 0.0005,
      "loss": 3.0829286575317383,
      "step": 1021
    },
    {
      "epoch": 0.19714506172839505,
      "grad_norm": 1.4885863065719604,
      "learning_rate": 0.0005,
      "loss": 3.596066474914551,
      "step": 1022
    },
    {
      "epoch": 0.19733796296296297,
      "grad_norm": 1.446761131286621,
      "learning_rate": 0.0005,
      "loss": 3.2404732704162598,
      "step": 1023
    },
    {
      "epoch": 0.19753086419753085,
      "grad_norm": 1.2166497707366943,
      "learning_rate": 0.0005,
      "loss": 3.306499719619751,
      "step": 1024
    },
    {
      "epoch": 0.19772376543209877,
      "grad_norm": 1.2509496212005615,
      "learning_rate": 0.0005,
      "loss": 3.3777196407318115,
      "step": 1025
    },
    {
      "epoch": 0.19791666666666666,
      "grad_norm": 1.161094069480896,
      "learning_rate": 0.0005,
      "loss": 3.384197235107422,
      "step": 1026
    },
    {
      "epoch": 0.19810956790123457,
      "grad_norm": 1.143441915512085,
      "learning_rate": 0.0005,
      "loss": 3.1197643280029297,
      "step": 1027
    },
    {
      "epoch": 0.19830246913580246,
      "grad_norm": 1.291205644607544,
      "learning_rate": 0.0005,
      "loss": 3.2787442207336426,
      "step": 1028
    },
    {
      "epoch": 0.19849537037037038,
      "grad_norm": 1.3114334344863892,
      "learning_rate": 0.0005,
      "loss": 3.701200008392334,
      "step": 1029
    },
    {
      "epoch": 0.19868827160493827,
      "grad_norm": 1.4277044534683228,
      "learning_rate": 0.0005,
      "loss": 3.3144919872283936,
      "step": 1030
    },
    {
      "epoch": 0.19888117283950618,
      "grad_norm": 2.0701632499694824,
      "learning_rate": 0.0005,
      "loss": 3.5577220916748047,
      "step": 1031
    },
    {
      "epoch": 0.19907407407407407,
      "grad_norm": 1.5937782526016235,
      "learning_rate": 0.0005,
      "loss": 3.374060869216919,
      "step": 1032
    },
    {
      "epoch": 0.19926697530864199,
      "grad_norm": 1.0813846588134766,
      "learning_rate": 0.0005,
      "loss": 3.168097496032715,
      "step": 1033
    },
    {
      "epoch": 0.19945987654320987,
      "grad_norm": 1.4003689289093018,
      "learning_rate": 0.0005,
      "loss": 3.7463626861572266,
      "step": 1034
    },
    {
      "epoch": 0.1996527777777778,
      "grad_norm": 1.414156198501587,
      "learning_rate": 0.0005,
      "loss": 3.411179780960083,
      "step": 1035
    },
    {
      "epoch": 0.19984567901234568,
      "grad_norm": 1.3490298986434937,
      "learning_rate": 0.0005,
      "loss": 3.029078960418701,
      "step": 1036
    },
    {
      "epoch": 0.20003858024691357,
      "grad_norm": 1.312694787979126,
      "learning_rate": 0.0005,
      "loss": 3.2928569316864014,
      "step": 1037
    },
    {
      "epoch": 0.20023148148148148,
      "grad_norm": 1.2646923065185547,
      "learning_rate": 0.0005,
      "loss": 3.319643974304199,
      "step": 1038
    },
    {
      "epoch": 0.20042438271604937,
      "grad_norm": 1.3652783632278442,
      "learning_rate": 0.0005,
      "loss": 3.1675496101379395,
      "step": 1039
    },
    {
      "epoch": 0.2006172839506173,
      "grad_norm": 1.1873172521591187,
      "learning_rate": 0.0005,
      "loss": 3.1411008834838867,
      "step": 1040
    },
    {
      "epoch": 0.20081018518518517,
      "grad_norm": 1.6368666887283325,
      "learning_rate": 0.0005,
      "loss": 2.9595065116882324,
      "step": 1041
    },
    {
      "epoch": 0.2010030864197531,
      "grad_norm": 1.5419973134994507,
      "learning_rate": 0.0005,
      "loss": 3.478599786758423,
      "step": 1042
    },
    {
      "epoch": 0.20119598765432098,
      "grad_norm": 1.0857185125350952,
      "learning_rate": 0.0005,
      "loss": 3.1592235565185547,
      "step": 1043
    },
    {
      "epoch": 0.2013888888888889,
      "grad_norm": 1.2233076095581055,
      "learning_rate": 0.0005,
      "loss": 3.14957857131958,
      "step": 1044
    },
    {
      "epoch": 0.20158179012345678,
      "grad_norm": 1.3013355731964111,
      "learning_rate": 0.0005,
      "loss": 3.325678825378418,
      "step": 1045
    },
    {
      "epoch": 0.2017746913580247,
      "grad_norm": 1.2768287658691406,
      "learning_rate": 0.0005,
      "loss": 3.2155656814575195,
      "step": 1046
    },
    {
      "epoch": 0.2019675925925926,
      "grad_norm": 1.1397888660430908,
      "learning_rate": 0.0005,
      "loss": 3.7027087211608887,
      "step": 1047
    },
    {
      "epoch": 0.2021604938271605,
      "grad_norm": 1.1968990564346313,
      "learning_rate": 0.0005,
      "loss": 3.1780314445495605,
      "step": 1048
    },
    {
      "epoch": 0.2023533950617284,
      "grad_norm": 1.170823574066162,
      "learning_rate": 0.0005,
      "loss": 3.120643138885498,
      "step": 1049
    },
    {
      "epoch": 0.2025462962962963,
      "grad_norm": 1.1832911968231201,
      "learning_rate": 0.0005,
      "loss": 3.6276001930236816,
      "step": 1050
    },
    {
      "epoch": 0.2027391975308642,
      "grad_norm": 1.3023420572280884,
      "learning_rate": 0.0005,
      "loss": 3.776827812194824,
      "step": 1051
    },
    {
      "epoch": 0.2029320987654321,
      "grad_norm": 1.5145502090454102,
      "learning_rate": 0.0005,
      "loss": 3.3289084434509277,
      "step": 1052
    },
    {
      "epoch": 0.203125,
      "grad_norm": 1.010790228843689,
      "learning_rate": 0.0005,
      "loss": 3.5123233795166016,
      "step": 1053
    },
    {
      "epoch": 0.2033179012345679,
      "grad_norm": 1.0479371547698975,
      "learning_rate": 0.0005,
      "loss": 3.133657932281494,
      "step": 1054
    },
    {
      "epoch": 0.2035108024691358,
      "grad_norm": 1.4598408937454224,
      "learning_rate": 0.0005,
      "loss": 3.369326114654541,
      "step": 1055
    },
    {
      "epoch": 0.2037037037037037,
      "grad_norm": 1.648645043373108,
      "learning_rate": 0.0005,
      "loss": 3.162513256072998,
      "step": 1056
    },
    {
      "epoch": 0.2038966049382716,
      "grad_norm": 1.0211451053619385,
      "learning_rate": 0.0005,
      "loss": 3.360692024230957,
      "step": 1057
    },
    {
      "epoch": 0.2040895061728395,
      "grad_norm": 1.2158355712890625,
      "learning_rate": 0.0005,
      "loss": 2.8786025047302246,
      "step": 1058
    },
    {
      "epoch": 0.2042824074074074,
      "grad_norm": 1.2202950716018677,
      "learning_rate": 0.0005,
      "loss": 3.2175910472869873,
      "step": 1059
    },
    {
      "epoch": 0.2044753086419753,
      "grad_norm": 1.322141170501709,
      "learning_rate": 0.0005,
      "loss": 3.260125160217285,
      "step": 1060
    },
    {
      "epoch": 0.20466820987654322,
      "grad_norm": 1.2762635946273804,
      "learning_rate": 0.0005,
      "loss": 3.2530159950256348,
      "step": 1061
    },
    {
      "epoch": 0.2048611111111111,
      "grad_norm": 1.042540192604065,
      "learning_rate": 0.0005,
      "loss": 3.2097980976104736,
      "step": 1062
    },
    {
      "epoch": 0.20505401234567902,
      "grad_norm": 1.1133363246917725,
      "learning_rate": 0.0005,
      "loss": 2.711574077606201,
      "step": 1063
    },
    {
      "epoch": 0.2052469135802469,
      "grad_norm": 1.0125757455825806,
      "learning_rate": 0.0005,
      "loss": 3.4549365043640137,
      "step": 1064
    },
    {
      "epoch": 0.20543981481481483,
      "grad_norm": 1.1158356666564941,
      "learning_rate": 0.0005,
      "loss": 3.1606998443603516,
      "step": 1065
    },
    {
      "epoch": 0.2056327160493827,
      "grad_norm": 1.1946762800216675,
      "learning_rate": 0.0005,
      "loss": 2.9930291175842285,
      "step": 1066
    },
    {
      "epoch": 0.20582561728395063,
      "grad_norm": 1.1306159496307373,
      "learning_rate": 0.0005,
      "loss": 3.3693008422851562,
      "step": 1067
    },
    {
      "epoch": 0.20601851851851852,
      "grad_norm": 1.171083927154541,
      "learning_rate": 0.0005,
      "loss": 3.5072951316833496,
      "step": 1068
    },
    {
      "epoch": 0.20621141975308643,
      "grad_norm": 1.3247530460357666,
      "learning_rate": 0.0005,
      "loss": 3.6297175884246826,
      "step": 1069
    },
    {
      "epoch": 0.20640432098765432,
      "grad_norm": 1.3551923036575317,
      "learning_rate": 0.0005,
      "loss": 3.14243221282959,
      "step": 1070
    },
    {
      "epoch": 0.2065972222222222,
      "grad_norm": 1.675675868988037,
      "learning_rate": 0.0005,
      "loss": 3.309018135070801,
      "step": 1071
    },
    {
      "epoch": 0.20679012345679013,
      "grad_norm": 1.0277761220932007,
      "learning_rate": 0.0005,
      "loss": 3.1684603691101074,
      "step": 1072
    },
    {
      "epoch": 0.20698302469135801,
      "grad_norm": 1.200405478477478,
      "learning_rate": 0.0005,
      "loss": 3.0631957054138184,
      "step": 1073
    },
    {
      "epoch": 0.20717592592592593,
      "grad_norm": 1.1563221216201782,
      "learning_rate": 0.0005,
      "loss": 3.120163679122925,
      "step": 1074
    },
    {
      "epoch": 0.20736882716049382,
      "grad_norm": 1.2929681539535522,
      "learning_rate": 0.0005,
      "loss": 3.4808664321899414,
      "step": 1075
    },
    {
      "epoch": 0.20756172839506173,
      "grad_norm": 1.0980279445648193,
      "learning_rate": 0.0005,
      "loss": 3.0505282878875732,
      "step": 1076
    },
    {
      "epoch": 0.20775462962962962,
      "grad_norm": 1.1386038064956665,
      "learning_rate": 0.0005,
      "loss": 3.086277484893799,
      "step": 1077
    },
    {
      "epoch": 0.20794753086419754,
      "grad_norm": 1.1140611171722412,
      "learning_rate": 0.0005,
      "loss": 3.245116949081421,
      "step": 1078
    },
    {
      "epoch": 0.20814043209876543,
      "grad_norm": 1.0110543966293335,
      "learning_rate": 0.0005,
      "loss": 3.273092269897461,
      "step": 1079
    },
    {
      "epoch": 0.20833333333333334,
      "grad_norm": 1.0433850288391113,
      "learning_rate": 0.0005,
      "loss": 3.2159132957458496,
      "step": 1080
    },
    {
      "epoch": 0.20852623456790123,
      "grad_norm": 1.0892943143844604,
      "learning_rate": 0.0005,
      "loss": 3.287719249725342,
      "step": 1081
    },
    {
      "epoch": 0.20871913580246915,
      "grad_norm": 1.3141953945159912,
      "learning_rate": 0.0005,
      "loss": 3.055299758911133,
      "step": 1082
    },
    {
      "epoch": 0.20891203703703703,
      "grad_norm": 1.216133713722229,
      "learning_rate": 0.0005,
      "loss": 3.323164701461792,
      "step": 1083
    },
    {
      "epoch": 0.20910493827160495,
      "grad_norm": 0.9350883960723877,
      "learning_rate": 0.0005,
      "loss": 3.041057825088501,
      "step": 1084
    },
    {
      "epoch": 0.20929783950617284,
      "grad_norm": 1.3338580131530762,
      "learning_rate": 0.0005,
      "loss": 3.1664485931396484,
      "step": 1085
    },
    {
      "epoch": 0.20949074074074073,
      "grad_norm": 1.063892126083374,
      "learning_rate": 0.0005,
      "loss": 2.9295356273651123,
      "step": 1086
    },
    {
      "epoch": 0.20968364197530864,
      "grad_norm": 1.5932011604309082,
      "learning_rate": 0.0005,
      "loss": 3.1057286262512207,
      "step": 1087
    },
    {
      "epoch": 0.20987654320987653,
      "grad_norm": 1.222719430923462,
      "learning_rate": 0.0005,
      "loss": 3.218700885772705,
      "step": 1088
    },
    {
      "epoch": 0.21006944444444445,
      "grad_norm": 1.2460522651672363,
      "learning_rate": 0.0005,
      "loss": 3.243135452270508,
      "step": 1089
    },
    {
      "epoch": 0.21026234567901234,
      "grad_norm": 1.191478967666626,
      "learning_rate": 0.0005,
      "loss": 3.0629782676696777,
      "step": 1090
    },
    {
      "epoch": 0.21045524691358025,
      "grad_norm": 1.1762171983718872,
      "learning_rate": 0.0005,
      "loss": 3.404501438140869,
      "step": 1091
    },
    {
      "epoch": 0.21064814814814814,
      "grad_norm": 1.1480348110198975,
      "learning_rate": 0.0005,
      "loss": 3.4333791732788086,
      "step": 1092
    },
    {
      "epoch": 0.21084104938271606,
      "grad_norm": 1.1525495052337646,
      "learning_rate": 0.0005,
      "loss": 3.3229360580444336,
      "step": 1093
    },
    {
      "epoch": 0.21103395061728394,
      "grad_norm": 1.3105347156524658,
      "learning_rate": 0.0005,
      "loss": 3.1471762657165527,
      "step": 1094
    },
    {
      "epoch": 0.21122685185185186,
      "grad_norm": 1.155412197113037,
      "learning_rate": 0.0005,
      "loss": 3.3786063194274902,
      "step": 1095
    },
    {
      "epoch": 0.21141975308641975,
      "grad_norm": 1.127929925918579,
      "learning_rate": 0.0005,
      "loss": 3.158714532852173,
      "step": 1096
    },
    {
      "epoch": 0.21161265432098766,
      "grad_norm": 1.1567051410675049,
      "learning_rate": 0.0005,
      "loss": 3.2149276733398438,
      "step": 1097
    },
    {
      "epoch": 0.21180555555555555,
      "grad_norm": 2.2255375385284424,
      "learning_rate": 0.0005,
      "loss": 3.2416131496429443,
      "step": 1098
    },
    {
      "epoch": 0.21199845679012347,
      "grad_norm": 1.0785980224609375,
      "learning_rate": 0.0005,
      "loss": 3.196560859680176,
      "step": 1099
    },
    {
      "epoch": 0.21219135802469136,
      "grad_norm": 0.9779481291770935,
      "learning_rate": 0.0005,
      "loss": 3.2560038566589355,
      "step": 1100
    },
    {
      "epoch": 0.21238425925925927,
      "grad_norm": 1.1422432661056519,
      "learning_rate": 0.0005,
      "loss": 3.212639331817627,
      "step": 1101
    },
    {
      "epoch": 0.21257716049382716,
      "grad_norm": 1.253082513809204,
      "learning_rate": 0.0005,
      "loss": 3.1393728256225586,
      "step": 1102
    },
    {
      "epoch": 0.21277006172839505,
      "grad_norm": 1.2418122291564941,
      "learning_rate": 0.0005,
      "loss": 3.113691806793213,
      "step": 1103
    },
    {
      "epoch": 0.21296296296296297,
      "grad_norm": 0.9892879724502563,
      "learning_rate": 0.0005,
      "loss": 2.8011114597320557,
      "step": 1104
    },
    {
      "epoch": 0.21315586419753085,
      "grad_norm": 1.1828877925872803,
      "learning_rate": 0.0005,
      "loss": 3.0321733951568604,
      "step": 1105
    },
    {
      "epoch": 0.21334876543209877,
      "grad_norm": 1.2462985515594482,
      "learning_rate": 0.0005,
      "loss": 2.961839437484741,
      "step": 1106
    },
    {
      "epoch": 0.21354166666666666,
      "grad_norm": 1.3942832946777344,
      "learning_rate": 0.0005,
      "loss": 3.2718634605407715,
      "step": 1107
    },
    {
      "epoch": 0.21373456790123457,
      "grad_norm": 1.3324060440063477,
      "learning_rate": 0.0005,
      "loss": 3.418201446533203,
      "step": 1108
    },
    {
      "epoch": 0.21392746913580246,
      "grad_norm": 1.3392695188522339,
      "learning_rate": 0.0005,
      "loss": 3.64871883392334,
      "step": 1109
    },
    {
      "epoch": 0.21412037037037038,
      "grad_norm": 1.1001994609832764,
      "learning_rate": 0.0005,
      "loss": 3.1158413887023926,
      "step": 1110
    },
    {
      "epoch": 0.21431327160493827,
      "grad_norm": 1.2341011762619019,
      "learning_rate": 0.0005,
      "loss": 3.061631679534912,
      "step": 1111
    },
    {
      "epoch": 0.21450617283950618,
      "grad_norm": 1.466328501701355,
      "learning_rate": 0.0005,
      "loss": 3.1656150817871094,
      "step": 1112
    },
    {
      "epoch": 0.21469907407407407,
      "grad_norm": 1.1995095014572144,
      "learning_rate": 0.0005,
      "loss": 3.3821263313293457,
      "step": 1113
    },
    {
      "epoch": 0.21489197530864199,
      "grad_norm": 1.0427700281143188,
      "learning_rate": 0.0005,
      "loss": 3.345852851867676,
      "step": 1114
    },
    {
      "epoch": 0.21508487654320987,
      "grad_norm": 1.0489330291748047,
      "learning_rate": 0.0005,
      "loss": 3.4766077995300293,
      "step": 1115
    },
    {
      "epoch": 0.2152777777777778,
      "grad_norm": 1.0133702754974365,
      "learning_rate": 0.0005,
      "loss": 3.0381364822387695,
      "step": 1116
    },
    {
      "epoch": 0.21547067901234568,
      "grad_norm": 1.8068865537643433,
      "learning_rate": 0.0005,
      "loss": 3.2660093307495117,
      "step": 1117
    },
    {
      "epoch": 0.21566358024691357,
      "grad_norm": 1.7818894386291504,
      "learning_rate": 0.0005,
      "loss": 2.985841989517212,
      "step": 1118
    },
    {
      "epoch": 0.21585648148148148,
      "grad_norm": 1.6733403205871582,
      "learning_rate": 0.0005,
      "loss": 3.0499377250671387,
      "step": 1119
    },
    {
      "epoch": 0.21604938271604937,
      "grad_norm": 1.5348151922225952,
      "learning_rate": 0.0005,
      "loss": 3.2631354331970215,
      "step": 1120
    },
    {
      "epoch": 0.2162422839506173,
      "grad_norm": 1.445851445198059,
      "learning_rate": 0.0005,
      "loss": 3.1020824909210205,
      "step": 1121
    },
    {
      "epoch": 0.21643518518518517,
      "grad_norm": 1.2497276067733765,
      "learning_rate": 0.0005,
      "loss": 3.1153342723846436,
      "step": 1122
    },
    {
      "epoch": 0.2166280864197531,
      "grad_norm": 1.6426881551742554,
      "learning_rate": 0.0005,
      "loss": 3.2539284229278564,
      "step": 1123
    },
    {
      "epoch": 0.21682098765432098,
      "grad_norm": 1.2617948055267334,
      "learning_rate": 0.0005,
      "loss": 3.2163431644439697,
      "step": 1124
    },
    {
      "epoch": 0.2170138888888889,
      "grad_norm": 1.3876850605010986,
      "learning_rate": 0.0005,
      "loss": 2.7664546966552734,
      "step": 1125
    },
    {
      "epoch": 0.21720679012345678,
      "grad_norm": 1.0398544073104858,
      "learning_rate": 0.0005,
      "loss": 3.232896089553833,
      "step": 1126
    },
    {
      "epoch": 0.2173996913580247,
      "grad_norm": 1.0028421878814697,
      "learning_rate": 0.0005,
      "loss": 3.035579204559326,
      "step": 1127
    },
    {
      "epoch": 0.2175925925925926,
      "grad_norm": 0.9136461019515991,
      "learning_rate": 0.0005,
      "loss": 3.1587717533111572,
      "step": 1128
    },
    {
      "epoch": 0.2177854938271605,
      "grad_norm": 1.3499128818511963,
      "learning_rate": 0.0005,
      "loss": 2.9878299236297607,
      "step": 1129
    },
    {
      "epoch": 0.2179783950617284,
      "grad_norm": 1.214934229850769,
      "learning_rate": 0.0005,
      "loss": 3.431576728820801,
      "step": 1130
    },
    {
      "epoch": 0.2181712962962963,
      "grad_norm": 1.3298282623291016,
      "learning_rate": 0.0005,
      "loss": 3.6404061317443848,
      "step": 1131
    },
    {
      "epoch": 0.2183641975308642,
      "grad_norm": 1.0917558670043945,
      "learning_rate": 0.0005,
      "loss": 3.2444872856140137,
      "step": 1132
    },
    {
      "epoch": 0.2185570987654321,
      "grad_norm": 1.15946364402771,
      "learning_rate": 0.0005,
      "loss": 3.370359182357788,
      "step": 1133
    },
    {
      "epoch": 0.21875,
      "grad_norm": 1.14241361618042,
      "learning_rate": 0.0005,
      "loss": 3.124065399169922,
      "step": 1134
    },
    {
      "epoch": 0.2189429012345679,
      "grad_norm": 1.1438558101654053,
      "learning_rate": 0.0005,
      "loss": 3.3833911418914795,
      "step": 1135
    },
    {
      "epoch": 0.2191358024691358,
      "grad_norm": 1.3074265718460083,
      "learning_rate": 0.0005,
      "loss": 3.1717801094055176,
      "step": 1136
    },
    {
      "epoch": 0.2193287037037037,
      "grad_norm": 1.0762956142425537,
      "learning_rate": 0.0005,
      "loss": 2.8028762340545654,
      "step": 1137
    },
    {
      "epoch": 0.2195216049382716,
      "grad_norm": 1.159356951713562,
      "learning_rate": 0.0005,
      "loss": 2.8611302375793457,
      "step": 1138
    },
    {
      "epoch": 0.2197145061728395,
      "grad_norm": 1.1606115102767944,
      "learning_rate": 0.0005,
      "loss": 3.3934285640716553,
      "step": 1139
    },
    {
      "epoch": 0.2199074074074074,
      "grad_norm": 1.2587398290634155,
      "learning_rate": 0.0005,
      "loss": 3.249131202697754,
      "step": 1140
    },
    {
      "epoch": 0.2201003086419753,
      "grad_norm": 1.5263798236846924,
      "learning_rate": 0.0005,
      "loss": 3.316882610321045,
      "step": 1141
    },
    {
      "epoch": 0.22029320987654322,
      "grad_norm": 1.0362629890441895,
      "learning_rate": 0.0005,
      "loss": 3.1261723041534424,
      "step": 1142
    },
    {
      "epoch": 0.2204861111111111,
      "grad_norm": 1.122950553894043,
      "learning_rate": 0.0005,
      "loss": 3.4278061389923096,
      "step": 1143
    },
    {
      "epoch": 0.22067901234567902,
      "grad_norm": 1.2215384244918823,
      "learning_rate": 0.0005,
      "loss": 3.2424111366271973,
      "step": 1144
    },
    {
      "epoch": 0.2208719135802469,
      "grad_norm": 1.5275861024856567,
      "learning_rate": 0.0005,
      "loss": 3.316068410873413,
      "step": 1145
    },
    {
      "epoch": 0.22106481481481483,
      "grad_norm": 1.1639447212219238,
      "learning_rate": 0.0005,
      "loss": 3.0623741149902344,
      "step": 1146
    },
    {
      "epoch": 0.2212577160493827,
      "grad_norm": 1.3290152549743652,
      "learning_rate": 0.0005,
      "loss": 2.8584978580474854,
      "step": 1147
    },
    {
      "epoch": 0.22145061728395063,
      "grad_norm": 1.081965446472168,
      "learning_rate": 0.0005,
      "loss": 3.1253395080566406,
      "step": 1148
    },
    {
      "epoch": 0.22164351851851852,
      "grad_norm": 1.1353360414505005,
      "learning_rate": 0.0005,
      "loss": 3.085620880126953,
      "step": 1149
    },
    {
      "epoch": 0.22183641975308643,
      "grad_norm": 1.0817134380340576,
      "learning_rate": 0.0005,
      "loss": 3.1916494369506836,
      "step": 1150
    },
    {
      "epoch": 0.22202932098765432,
      "grad_norm": 1.3807525634765625,
      "learning_rate": 0.0005,
      "loss": 3.2224321365356445,
      "step": 1151
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 1.1646698713302612,
      "learning_rate": 0.0005,
      "loss": 3.364759922027588,
      "step": 1152
    },
    {
      "epoch": 0.22241512345679013,
      "grad_norm": 1.4222545623779297,
      "learning_rate": 0.0005,
      "loss": 3.1591269969940186,
      "step": 1153
    },
    {
      "epoch": 0.22260802469135801,
      "grad_norm": 1.1558971405029297,
      "learning_rate": 0.0005,
      "loss": 3.0178487300872803,
      "step": 1154
    },
    {
      "epoch": 0.22280092592592593,
      "grad_norm": 1.083590030670166,
      "learning_rate": 0.0005,
      "loss": 3.153235912322998,
      "step": 1155
    },
    {
      "epoch": 0.22299382716049382,
      "grad_norm": 1.2491824626922607,
      "learning_rate": 0.0005,
      "loss": 3.208542585372925,
      "step": 1156
    },
    {
      "epoch": 0.22318672839506173,
      "grad_norm": 1.1746231317520142,
      "learning_rate": 0.0005,
      "loss": 3.2222185134887695,
      "step": 1157
    },
    {
      "epoch": 0.22337962962962962,
      "grad_norm": 1.1187819242477417,
      "learning_rate": 0.0005,
      "loss": 3.050086259841919,
      "step": 1158
    },
    {
      "epoch": 0.22357253086419754,
      "grad_norm": 1.3732166290283203,
      "learning_rate": 0.0005,
      "loss": 3.230844020843506,
      "step": 1159
    },
    {
      "epoch": 0.22376543209876543,
      "grad_norm": 1.2544374465942383,
      "learning_rate": 0.0005,
      "loss": 3.219630479812622,
      "step": 1160
    },
    {
      "epoch": 0.22395833333333334,
      "grad_norm": 1.0806007385253906,
      "learning_rate": 0.0005,
      "loss": 3.122086524963379,
      "step": 1161
    },
    {
      "epoch": 0.22415123456790123,
      "grad_norm": 1.6098554134368896,
      "learning_rate": 0.0005,
      "loss": 3.333951473236084,
      "step": 1162
    },
    {
      "epoch": 0.22434413580246915,
      "grad_norm": 1.143479824066162,
      "learning_rate": 0.0005,
      "loss": 3.3586721420288086,
      "step": 1163
    },
    {
      "epoch": 0.22453703703703703,
      "grad_norm": 1.514394760131836,
      "learning_rate": 0.0005,
      "loss": 3.2408344745635986,
      "step": 1164
    },
    {
      "epoch": 0.22472993827160495,
      "grad_norm": 1.4589394330978394,
      "learning_rate": 0.0005,
      "loss": 3.510335922241211,
      "step": 1165
    },
    {
      "epoch": 0.22492283950617284,
      "grad_norm": 1.1432428359985352,
      "learning_rate": 0.0005,
      "loss": 2.934203863143921,
      "step": 1166
    },
    {
      "epoch": 0.22511574074074073,
      "grad_norm": 1.1670974493026733,
      "learning_rate": 0.0005,
      "loss": 3.33964204788208,
      "step": 1167
    },
    {
      "epoch": 0.22530864197530864,
      "grad_norm": 1.0750895738601685,
      "learning_rate": 0.0005,
      "loss": 3.409942626953125,
      "step": 1168
    },
    {
      "epoch": 0.22550154320987653,
      "grad_norm": 1.2052689790725708,
      "learning_rate": 0.0005,
      "loss": 2.9955224990844727,
      "step": 1169
    },
    {
      "epoch": 0.22569444444444445,
      "grad_norm": 1.0634727478027344,
      "learning_rate": 0.0005,
      "loss": 3.209463596343994,
      "step": 1170
    },
    {
      "epoch": 0.22588734567901234,
      "grad_norm": 1.0815539360046387,
      "learning_rate": 0.0005,
      "loss": 3.4661827087402344,
      "step": 1171
    },
    {
      "epoch": 0.22608024691358025,
      "grad_norm": 1.7844243049621582,
      "learning_rate": 0.0005,
      "loss": 3.4162025451660156,
      "step": 1172
    },
    {
      "epoch": 0.22627314814814814,
      "grad_norm": 1.1358997821807861,
      "learning_rate": 0.0005,
      "loss": 3.4086732864379883,
      "step": 1173
    },
    {
      "epoch": 0.22646604938271606,
      "grad_norm": 1.1031476259231567,
      "learning_rate": 0.0005,
      "loss": 3.109956741333008,
      "step": 1174
    },
    {
      "epoch": 0.22665895061728394,
      "grad_norm": 1.7434877157211304,
      "learning_rate": 0.0005,
      "loss": 3.2372047901153564,
      "step": 1175
    },
    {
      "epoch": 0.22685185185185186,
      "grad_norm": 1.5098886489868164,
      "learning_rate": 0.0005,
      "loss": 3.240060329437256,
      "step": 1176
    },
    {
      "epoch": 0.22704475308641975,
      "grad_norm": 1.5480265617370605,
      "learning_rate": 0.0005,
      "loss": 3.503444194793701,
      "step": 1177
    },
    {
      "epoch": 0.22723765432098766,
      "grad_norm": 2.316342830657959,
      "learning_rate": 0.0005,
      "loss": 3.230363368988037,
      "step": 1178
    },
    {
      "epoch": 0.22743055555555555,
      "grad_norm": 1.605416178703308,
      "learning_rate": 0.0005,
      "loss": 3.4059343338012695,
      "step": 1179
    },
    {
      "epoch": 0.22762345679012347,
      "grad_norm": 1.3326904773712158,
      "learning_rate": 0.0005,
      "loss": 2.971494436264038,
      "step": 1180
    },
    {
      "epoch": 0.22781635802469136,
      "grad_norm": 1.1135551929473877,
      "learning_rate": 0.0005,
      "loss": 2.9354114532470703,
      "step": 1181
    },
    {
      "epoch": 0.22800925925925927,
      "grad_norm": 1.4506897926330566,
      "learning_rate": 0.0005,
      "loss": 3.4195680618286133,
      "step": 1182
    },
    {
      "epoch": 0.22820216049382716,
      "grad_norm": 1.1851023435592651,
      "learning_rate": 0.0005,
      "loss": 2.877023220062256,
      "step": 1183
    },
    {
      "epoch": 0.22839506172839505,
      "grad_norm": 1.7696037292480469,
      "learning_rate": 0.0005,
      "loss": 3.325911045074463,
      "step": 1184
    },
    {
      "epoch": 0.22858796296296297,
      "grad_norm": 1.0017932653427124,
      "learning_rate": 0.0005,
      "loss": 2.9727494716644287,
      "step": 1185
    },
    {
      "epoch": 0.22878086419753085,
      "grad_norm": 1.0376572608947754,
      "learning_rate": 0.0005,
      "loss": 2.8474020957946777,
      "step": 1186
    },
    {
      "epoch": 0.22897376543209877,
      "grad_norm": 1.165851354598999,
      "learning_rate": 0.0005,
      "loss": 3.213728427886963,
      "step": 1187
    },
    {
      "epoch": 0.22916666666666666,
      "grad_norm": 1.1158924102783203,
      "learning_rate": 0.0005,
      "loss": 3.260072708129883,
      "step": 1188
    },
    {
      "epoch": 0.22935956790123457,
      "grad_norm": 1.105642557144165,
      "learning_rate": 0.0005,
      "loss": 3.172901153564453,
      "step": 1189
    },
    {
      "epoch": 0.22955246913580246,
      "grad_norm": 1.1950089931488037,
      "learning_rate": 0.0005,
      "loss": 3.115774154663086,
      "step": 1190
    },
    {
      "epoch": 0.22974537037037038,
      "grad_norm": 1.0935555696487427,
      "learning_rate": 0.0005,
      "loss": 3.0834851264953613,
      "step": 1191
    },
    {
      "epoch": 0.22993827160493827,
      "grad_norm": 1.2153112888336182,
      "learning_rate": 0.0005,
      "loss": 3.053415060043335,
      "step": 1192
    },
    {
      "epoch": 0.23013117283950618,
      "grad_norm": 1.3526332378387451,
      "learning_rate": 0.0005,
      "loss": 3.3269877433776855,
      "step": 1193
    },
    {
      "epoch": 0.23032407407407407,
      "grad_norm": 1.3531203269958496,
      "learning_rate": 0.0005,
      "loss": 3.244513511657715,
      "step": 1194
    },
    {
      "epoch": 0.23051697530864199,
      "grad_norm": 1.1664690971374512,
      "learning_rate": 0.0005,
      "loss": 3.3224706649780273,
      "step": 1195
    },
    {
      "epoch": 0.23070987654320987,
      "grad_norm": 0.9878566861152649,
      "learning_rate": 0.0005,
      "loss": 3.1604726314544678,
      "step": 1196
    },
    {
      "epoch": 0.2309027777777778,
      "grad_norm": 1.1305325031280518,
      "learning_rate": 0.0005,
      "loss": 2.9156646728515625,
      "step": 1197
    },
    {
      "epoch": 0.23109567901234568,
      "grad_norm": 1.0762407779693604,
      "learning_rate": 0.0005,
      "loss": 3.18211030960083,
      "step": 1198
    },
    {
      "epoch": 0.23128858024691357,
      "grad_norm": 1.2009553909301758,
      "learning_rate": 0.0005,
      "loss": 3.419070243835449,
      "step": 1199
    },
    {
      "epoch": 0.23148148148148148,
      "grad_norm": 1.456424593925476,
      "learning_rate": 0.0005,
      "loss": 3.076817035675049,
      "step": 1200
    },
    {
      "epoch": 0.23167438271604937,
      "grad_norm": 1.2507944107055664,
      "learning_rate": 0.0005,
      "loss": 3.1279468536376953,
      "step": 1201
    },
    {
      "epoch": 0.2318672839506173,
      "grad_norm": 1.1768923997879028,
      "learning_rate": 0.0005,
      "loss": 2.949848175048828,
      "step": 1202
    },
    {
      "epoch": 0.23206018518518517,
      "grad_norm": 1.2228968143463135,
      "learning_rate": 0.0005,
      "loss": 3.144275665283203,
      "step": 1203
    },
    {
      "epoch": 0.2322530864197531,
      "grad_norm": 1.3830986022949219,
      "learning_rate": 0.0005,
      "loss": 3.0533101558685303,
      "step": 1204
    },
    {
      "epoch": 0.23244598765432098,
      "grad_norm": 1.5523402690887451,
      "learning_rate": 0.0005,
      "loss": 3.023268222808838,
      "step": 1205
    },
    {
      "epoch": 0.2326388888888889,
      "grad_norm": 1.2047369480133057,
      "learning_rate": 0.0005,
      "loss": 3.112780809402466,
      "step": 1206
    },
    {
      "epoch": 0.23283179012345678,
      "grad_norm": 1.1276134252548218,
      "learning_rate": 0.0005,
      "loss": 3.57730770111084,
      "step": 1207
    },
    {
      "epoch": 0.2330246913580247,
      "grad_norm": 1.4940801858901978,
      "learning_rate": 0.0005,
      "loss": 3.2162961959838867,
      "step": 1208
    },
    {
      "epoch": 0.2332175925925926,
      "grad_norm": 1.2099101543426514,
      "learning_rate": 0.0005,
      "loss": 3.2194790840148926,
      "step": 1209
    },
    {
      "epoch": 0.2334104938271605,
      "grad_norm": 1.0407192707061768,
      "learning_rate": 0.0005,
      "loss": 3.345205783843994,
      "step": 1210
    },
    {
      "epoch": 0.2336033950617284,
      "grad_norm": 1.180210828781128,
      "learning_rate": 0.0005,
      "loss": 3.1922802925109863,
      "step": 1211
    },
    {
      "epoch": 0.2337962962962963,
      "grad_norm": 1.4632294178009033,
      "learning_rate": 0.0005,
      "loss": 3.217890739440918,
      "step": 1212
    },
    {
      "epoch": 0.2339891975308642,
      "grad_norm": 1.1188876628875732,
      "learning_rate": 0.0005,
      "loss": 3.2257933616638184,
      "step": 1213
    },
    {
      "epoch": 0.2341820987654321,
      "grad_norm": 1.1321454048156738,
      "learning_rate": 0.0005,
      "loss": 3.013334274291992,
      "step": 1214
    },
    {
      "epoch": 0.234375,
      "grad_norm": 0.9891884922981262,
      "learning_rate": 0.0005,
      "loss": 3.1539065837860107,
      "step": 1215
    },
    {
      "epoch": 0.2345679012345679,
      "grad_norm": 1.1257613897323608,
      "learning_rate": 0.0005,
      "loss": 3.1933274269104004,
      "step": 1216
    },
    {
      "epoch": 0.2347608024691358,
      "grad_norm": 1.2468284368515015,
      "learning_rate": 0.0005,
      "loss": 3.2461636066436768,
      "step": 1217
    },
    {
      "epoch": 0.2349537037037037,
      "grad_norm": 1.2066433429718018,
      "learning_rate": 0.0005,
      "loss": 3.2009518146514893,
      "step": 1218
    },
    {
      "epoch": 0.2351466049382716,
      "grad_norm": 1.0884333848953247,
      "learning_rate": 0.0005,
      "loss": 3.348784923553467,
      "step": 1219
    },
    {
      "epoch": 0.2353395061728395,
      "grad_norm": 1.4190400838851929,
      "learning_rate": 0.0005,
      "loss": 3.029418468475342,
      "step": 1220
    },
    {
      "epoch": 0.2355324074074074,
      "grad_norm": 1.4687966108322144,
      "learning_rate": 0.0005,
      "loss": 2.730462074279785,
      "step": 1221
    },
    {
      "epoch": 0.2357253086419753,
      "grad_norm": 1.3050463199615479,
      "learning_rate": 0.0005,
      "loss": 3.3538265228271484,
      "step": 1222
    },
    {
      "epoch": 0.23591820987654322,
      "grad_norm": 1.0784860849380493,
      "learning_rate": 0.0005,
      "loss": 3.1748461723327637,
      "step": 1223
    },
    {
      "epoch": 0.2361111111111111,
      "grad_norm": 1.268298625946045,
      "learning_rate": 0.0005,
      "loss": 3.1567983627319336,
      "step": 1224
    },
    {
      "epoch": 0.23630401234567902,
      "grad_norm": 1.0947843790054321,
      "learning_rate": 0.0005,
      "loss": 3.135288953781128,
      "step": 1225
    },
    {
      "epoch": 0.2364969135802469,
      "grad_norm": 1.0137029886245728,
      "learning_rate": 0.0005,
      "loss": 3.0614309310913086,
      "step": 1226
    },
    {
      "epoch": 0.23668981481481483,
      "grad_norm": 1.0875753164291382,
      "learning_rate": 0.0005,
      "loss": 3.307126045227051,
      "step": 1227
    },
    {
      "epoch": 0.2368827160493827,
      "grad_norm": 1.198175072669983,
      "learning_rate": 0.0005,
      "loss": 3.0849967002868652,
      "step": 1228
    },
    {
      "epoch": 0.23707561728395063,
      "grad_norm": 1.53101646900177,
      "learning_rate": 0.0005,
      "loss": 3.2544002532958984,
      "step": 1229
    },
    {
      "epoch": 0.23726851851851852,
      "grad_norm": 1.1194281578063965,
      "learning_rate": 0.0005,
      "loss": 3.2511849403381348,
      "step": 1230
    },
    {
      "epoch": 0.23746141975308643,
      "grad_norm": 1.162011742591858,
      "learning_rate": 0.0005,
      "loss": 3.4470911026000977,
      "step": 1231
    },
    {
      "epoch": 0.23765432098765432,
      "grad_norm": 0.9444446563720703,
      "learning_rate": 0.0005,
      "loss": 2.973702907562256,
      "step": 1232
    },
    {
      "epoch": 0.2378472222222222,
      "grad_norm": 1.4724972248077393,
      "learning_rate": 0.0005,
      "loss": 2.902574062347412,
      "step": 1233
    },
    {
      "epoch": 0.23804012345679013,
      "grad_norm": 1.2083518505096436,
      "learning_rate": 0.0005,
      "loss": 3.2239534854888916,
      "step": 1234
    },
    {
      "epoch": 0.23823302469135801,
      "grad_norm": 1.9329514503479004,
      "learning_rate": 0.0005,
      "loss": 3.028503656387329,
      "step": 1235
    },
    {
      "epoch": 0.23842592592592593,
      "grad_norm": 1.165121078491211,
      "learning_rate": 0.0005,
      "loss": 3.2502031326293945,
      "step": 1236
    },
    {
      "epoch": 0.23861882716049382,
      "grad_norm": 1.2332971096038818,
      "learning_rate": 0.0005,
      "loss": 3.1086926460266113,
      "step": 1237
    },
    {
      "epoch": 0.23881172839506173,
      "grad_norm": 1.322426676750183,
      "learning_rate": 0.0005,
      "loss": 3.525825023651123,
      "step": 1238
    },
    {
      "epoch": 0.23900462962962962,
      "grad_norm": 1.054040551185608,
      "learning_rate": 0.0005,
      "loss": 3.183349132537842,
      "step": 1239
    },
    {
      "epoch": 0.23919753086419754,
      "grad_norm": 1.386708378791809,
      "learning_rate": 0.0005,
      "loss": 3.42018461227417,
      "step": 1240
    },
    {
      "epoch": 0.23939043209876543,
      "grad_norm": 1.0444529056549072,
      "learning_rate": 0.0005,
      "loss": 2.9527194499969482,
      "step": 1241
    },
    {
      "epoch": 0.23958333333333334,
      "grad_norm": 1.2778369188308716,
      "learning_rate": 0.0005,
      "loss": 3.399580955505371,
      "step": 1242
    },
    {
      "epoch": 0.23977623456790123,
      "grad_norm": 1.2149019241333008,
      "learning_rate": 0.0005,
      "loss": 3.1394381523132324,
      "step": 1243
    },
    {
      "epoch": 0.23996913580246915,
      "grad_norm": 1.0990843772888184,
      "learning_rate": 0.0005,
      "loss": 3.187685489654541,
      "step": 1244
    },
    {
      "epoch": 0.24016203703703703,
      "grad_norm": 0.9557335376739502,
      "learning_rate": 0.0005,
      "loss": 3.235527753829956,
      "step": 1245
    },
    {
      "epoch": 0.24035493827160495,
      "grad_norm": 1.2041339874267578,
      "learning_rate": 0.0005,
      "loss": 3.158607006072998,
      "step": 1246
    },
    {
      "epoch": 0.24054783950617284,
      "grad_norm": 1.3523707389831543,
      "learning_rate": 0.0005,
      "loss": 3.1642889976501465,
      "step": 1247
    },
    {
      "epoch": 0.24074074074074073,
      "grad_norm": 1.1573368310928345,
      "learning_rate": 0.0005,
      "loss": 3.2888216972351074,
      "step": 1248
    },
    {
      "epoch": 0.24093364197530864,
      "grad_norm": 1.212227463722229,
      "learning_rate": 0.0005,
      "loss": 3.5067930221557617,
      "step": 1249
    },
    {
      "epoch": 0.24112654320987653,
      "grad_norm": 1.1416808366775513,
      "learning_rate": 0.0005,
      "loss": 3.200289726257324,
      "step": 1250
    },
    {
      "epoch": 0.24131944444444445,
      "grad_norm": 1.2780585289001465,
      "learning_rate": 0.0005,
      "loss": 3.1198034286499023,
      "step": 1251
    },
    {
      "epoch": 0.24151234567901234,
      "grad_norm": 1.1388648748397827,
      "learning_rate": 0.0005,
      "loss": 2.9118576049804688,
      "step": 1252
    },
    {
      "epoch": 0.24170524691358025,
      "grad_norm": 1.3205623626708984,
      "learning_rate": 0.0005,
      "loss": 3.2032594680786133,
      "step": 1253
    },
    {
      "epoch": 0.24189814814814814,
      "grad_norm": 1.4108232259750366,
      "learning_rate": 0.0005,
      "loss": 3.2178540229797363,
      "step": 1254
    },
    {
      "epoch": 0.24209104938271606,
      "grad_norm": 1.101608395576477,
      "learning_rate": 0.0005,
      "loss": 3.5508384704589844,
      "step": 1255
    },
    {
      "epoch": 0.24228395061728394,
      "grad_norm": 1.1341784000396729,
      "learning_rate": 0.0005,
      "loss": 3.0667848587036133,
      "step": 1256
    },
    {
      "epoch": 0.24247685185185186,
      "grad_norm": 1.8275485038757324,
      "learning_rate": 0.0005,
      "loss": 2.9994823932647705,
      "step": 1257
    },
    {
      "epoch": 0.24266975308641975,
      "grad_norm": 1.5311535596847534,
      "learning_rate": 0.0005,
      "loss": 3.189188241958618,
      "step": 1258
    },
    {
      "epoch": 0.24286265432098766,
      "grad_norm": 1.1312366724014282,
      "learning_rate": 0.0005,
      "loss": 3.118370532989502,
      "step": 1259
    },
    {
      "epoch": 0.24305555555555555,
      "grad_norm": 1.112632155418396,
      "learning_rate": 0.0005,
      "loss": 2.7515134811401367,
      "step": 1260
    },
    {
      "epoch": 0.24324845679012347,
      "grad_norm": 1.114710807800293,
      "learning_rate": 0.0005,
      "loss": 3.176628589630127,
      "step": 1261
    },
    {
      "epoch": 0.24344135802469136,
      "grad_norm": 0.9772851467132568,
      "learning_rate": 0.0005,
      "loss": 3.424440860748291,
      "step": 1262
    },
    {
      "epoch": 0.24363425925925927,
      "grad_norm": 1.0367965698242188,
      "learning_rate": 0.0005,
      "loss": 3.38991379737854,
      "step": 1263
    },
    {
      "epoch": 0.24382716049382716,
      "grad_norm": 1.0514323711395264,
      "learning_rate": 0.0005,
      "loss": 2.9413375854492188,
      "step": 1264
    },
    {
      "epoch": 0.24402006172839505,
      "grad_norm": 1.2497167587280273,
      "learning_rate": 0.0005,
      "loss": 3.1710617542266846,
      "step": 1265
    },
    {
      "epoch": 0.24421296296296297,
      "grad_norm": 1.532235860824585,
      "learning_rate": 0.0005,
      "loss": 3.0098676681518555,
      "step": 1266
    },
    {
      "epoch": 0.24440586419753085,
      "grad_norm": 1.1492140293121338,
      "learning_rate": 0.0005,
      "loss": 3.2026209831237793,
      "step": 1267
    },
    {
      "epoch": 0.24459876543209877,
      "grad_norm": 1.2207633256912231,
      "learning_rate": 0.0005,
      "loss": 3.2455196380615234,
      "step": 1268
    },
    {
      "epoch": 0.24479166666666666,
      "grad_norm": 1.1001873016357422,
      "learning_rate": 0.0005,
      "loss": 3.2205846309661865,
      "step": 1269
    },
    {
      "epoch": 0.24498456790123457,
      "grad_norm": 1.253018856048584,
      "learning_rate": 0.0005,
      "loss": 2.8806207180023193,
      "step": 1270
    },
    {
      "epoch": 0.24517746913580246,
      "grad_norm": 0.987488865852356,
      "learning_rate": 0.0005,
      "loss": 3.1021666526794434,
      "step": 1271
    },
    {
      "epoch": 0.24537037037037038,
      "grad_norm": 1.1358470916748047,
      "learning_rate": 0.0005,
      "loss": 3.0216708183288574,
      "step": 1272
    },
    {
      "epoch": 0.24556327160493827,
      "grad_norm": 1.377466082572937,
      "learning_rate": 0.0005,
      "loss": 2.974902391433716,
      "step": 1273
    },
    {
      "epoch": 0.24575617283950618,
      "grad_norm": 1.1675982475280762,
      "learning_rate": 0.0005,
      "loss": 2.9711270332336426,
      "step": 1274
    },
    {
      "epoch": 0.24594907407407407,
      "grad_norm": 1.3123387098312378,
      "learning_rate": 0.0005,
      "loss": 3.1727778911590576,
      "step": 1275
    },
    {
      "epoch": 0.24614197530864199,
      "grad_norm": 1.4285506010055542,
      "learning_rate": 0.0005,
      "loss": 3.333425998687744,
      "step": 1276
    },
    {
      "epoch": 0.24633487654320987,
      "grad_norm": 1.9822133779525757,
      "learning_rate": 0.0005,
      "loss": 2.640261173248291,
      "step": 1277
    },
    {
      "epoch": 0.2465277777777778,
      "grad_norm": 1.0515432357788086,
      "learning_rate": 0.0005,
      "loss": 2.9865684509277344,
      "step": 1278
    },
    {
      "epoch": 0.24672067901234568,
      "grad_norm": 1.3336446285247803,
      "learning_rate": 0.0005,
      "loss": 3.2500877380371094,
      "step": 1279
    },
    {
      "epoch": 0.24691358024691357,
      "grad_norm": 1.256935477256775,
      "learning_rate": 0.0005,
      "loss": 2.995485782623291,
      "step": 1280
    },
    {
      "epoch": 0.24710648148148148,
      "grad_norm": 1.114327311515808,
      "learning_rate": 0.0005,
      "loss": 3.220986843109131,
      "step": 1281
    },
    {
      "epoch": 0.24729938271604937,
      "grad_norm": 2.0502262115478516,
      "learning_rate": 0.0005,
      "loss": 3.1418890953063965,
      "step": 1282
    },
    {
      "epoch": 0.2474922839506173,
      "grad_norm": 1.4039634466171265,
      "learning_rate": 0.0005,
      "loss": 3.2171790599823,
      "step": 1283
    },
    {
      "epoch": 0.24768518518518517,
      "grad_norm": 1.1835360527038574,
      "learning_rate": 0.0005,
      "loss": 3.124265432357788,
      "step": 1284
    },
    {
      "epoch": 0.2478780864197531,
      "grad_norm": 1.2587207555770874,
      "learning_rate": 0.0005,
      "loss": 3.080225944519043,
      "step": 1285
    },
    {
      "epoch": 0.24807098765432098,
      "grad_norm": 1.4825109243392944,
      "learning_rate": 0.0005,
      "loss": 3.1024668216705322,
      "step": 1286
    },
    {
      "epoch": 0.2482638888888889,
      "grad_norm": 1.492321252822876,
      "learning_rate": 0.0005,
      "loss": 3.258655309677124,
      "step": 1287
    },
    {
      "epoch": 0.24845679012345678,
      "grad_norm": 1.0996559858322144,
      "learning_rate": 0.0005,
      "loss": 3.2091920375823975,
      "step": 1288
    },
    {
      "epoch": 0.2486496913580247,
      "grad_norm": 1.3574262857437134,
      "learning_rate": 0.0005,
      "loss": 3.0061111450195312,
      "step": 1289
    },
    {
      "epoch": 0.2488425925925926,
      "grad_norm": 1.2464364767074585,
      "learning_rate": 0.0005,
      "loss": 3.1789488792419434,
      "step": 1290
    },
    {
      "epoch": 0.2490354938271605,
      "grad_norm": 1.2197762727737427,
      "learning_rate": 0.0005,
      "loss": 3.244077682495117,
      "step": 1291
    },
    {
      "epoch": 0.2492283950617284,
      "grad_norm": 1.0804009437561035,
      "learning_rate": 0.0005,
      "loss": 2.9430222511291504,
      "step": 1292
    },
    {
      "epoch": 0.2494212962962963,
      "grad_norm": 1.711513876914978,
      "learning_rate": 0.0005,
      "loss": 3.210113048553467,
      "step": 1293
    },
    {
      "epoch": 0.2496141975308642,
      "grad_norm": 1.3244335651397705,
      "learning_rate": 0.0005,
      "loss": 3.3260607719421387,
      "step": 1294
    },
    {
      "epoch": 0.2498070987654321,
      "grad_norm": 1.3970402479171753,
      "learning_rate": 0.0005,
      "loss": 3.236534357070923,
      "step": 1295
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.3048523664474487,
      "learning_rate": 0.0005,
      "loss": 3.252958297729492,
      "step": 1296
    },
    {
      "epoch": 0.2501929012345679,
      "grad_norm": 2.0268468856811523,
      "learning_rate": 0.0005,
      "loss": 3.3226094245910645,
      "step": 1297
    },
    {
      "epoch": 0.2503858024691358,
      "grad_norm": 1.2097982168197632,
      "learning_rate": 0.0005,
      "loss": 3.3363876342773438,
      "step": 1298
    },
    {
      "epoch": 0.2505787037037037,
      "grad_norm": 3.3419103622436523,
      "learning_rate": 0.0005,
      "loss": 3.1040663719177246,
      "step": 1299
    },
    {
      "epoch": 0.2507716049382716,
      "grad_norm": 1.3248844146728516,
      "learning_rate": 0.0005,
      "loss": 3.32338809967041,
      "step": 1300
    },
    {
      "epoch": 0.2509645061728395,
      "grad_norm": 1.6041958332061768,
      "learning_rate": 0.0005,
      "loss": 3.116551399230957,
      "step": 1301
    },
    {
      "epoch": 0.2511574074074074,
      "grad_norm": 1.9477965831756592,
      "learning_rate": 0.0005,
      "loss": 3.088604688644409,
      "step": 1302
    },
    {
      "epoch": 0.25135030864197533,
      "grad_norm": 1.2270750999450684,
      "learning_rate": 0.0005,
      "loss": 3.1580517292022705,
      "step": 1303
    },
    {
      "epoch": 0.2515432098765432,
      "grad_norm": 1.1202045679092407,
      "learning_rate": 0.0005,
      "loss": 3.173480987548828,
      "step": 1304
    },
    {
      "epoch": 0.2517361111111111,
      "grad_norm": 1.6609278917312622,
      "learning_rate": 0.0005,
      "loss": 2.997225046157837,
      "step": 1305
    },
    {
      "epoch": 0.251929012345679,
      "grad_norm": 1.1474453210830688,
      "learning_rate": 0.0005,
      "loss": 3.348972797393799,
      "step": 1306
    },
    {
      "epoch": 0.25212191358024694,
      "grad_norm": 0.945825457572937,
      "learning_rate": 0.0005,
      "loss": 2.9480013847351074,
      "step": 1307
    },
    {
      "epoch": 0.2523148148148148,
      "grad_norm": 1.7401955127716064,
      "learning_rate": 0.0005,
      "loss": 3.2018213272094727,
      "step": 1308
    },
    {
      "epoch": 0.2525077160493827,
      "grad_norm": 2.3311970233917236,
      "learning_rate": 0.0005,
      "loss": 3.656003475189209,
      "step": 1309
    },
    {
      "epoch": 0.2527006172839506,
      "grad_norm": 1.2158539295196533,
      "learning_rate": 0.0005,
      "loss": 2.9730935096740723,
      "step": 1310
    },
    {
      "epoch": 0.25289351851851855,
      "grad_norm": 1.1427851915359497,
      "learning_rate": 0.0005,
      "loss": 3.3349380493164062,
      "step": 1311
    },
    {
      "epoch": 0.25308641975308643,
      "grad_norm": 1.47557532787323,
      "learning_rate": 0.0005,
      "loss": 3.0659050941467285,
      "step": 1312
    },
    {
      "epoch": 0.2532793209876543,
      "grad_norm": 1.393542766571045,
      "learning_rate": 0.0005,
      "loss": 3.215304374694824,
      "step": 1313
    },
    {
      "epoch": 0.2534722222222222,
      "grad_norm": 1.1039036512374878,
      "learning_rate": 0.0005,
      "loss": 2.9457225799560547,
      "step": 1314
    },
    {
      "epoch": 0.2536651234567901,
      "grad_norm": 1.244687795639038,
      "learning_rate": 0.0005,
      "loss": 3.032778263092041,
      "step": 1315
    },
    {
      "epoch": 0.25385802469135804,
      "grad_norm": 1.3662551641464233,
      "learning_rate": 0.0005,
      "loss": 3.2821149826049805,
      "step": 1316
    },
    {
      "epoch": 0.25405092592592593,
      "grad_norm": 1.133880376815796,
      "learning_rate": 0.0005,
      "loss": 3.2237963676452637,
      "step": 1317
    },
    {
      "epoch": 0.2542438271604938,
      "grad_norm": 1.0435142517089844,
      "learning_rate": 0.0005,
      "loss": 3.116298198699951,
      "step": 1318
    },
    {
      "epoch": 0.2544367283950617,
      "grad_norm": 1.1695623397827148,
      "learning_rate": 0.0005,
      "loss": 3.2857699394226074,
      "step": 1319
    },
    {
      "epoch": 0.25462962962962965,
      "grad_norm": 1.105344533920288,
      "learning_rate": 0.0005,
      "loss": 2.9807522296905518,
      "step": 1320
    },
    {
      "epoch": 0.25482253086419754,
      "grad_norm": 1.1975263357162476,
      "learning_rate": 0.0005,
      "loss": 3.273414134979248,
      "step": 1321
    },
    {
      "epoch": 0.2550154320987654,
      "grad_norm": 1.017902135848999,
      "learning_rate": 0.0005,
      "loss": 3.300647020339966,
      "step": 1322
    },
    {
      "epoch": 0.2552083333333333,
      "grad_norm": 1.1514610052108765,
      "learning_rate": 0.0005,
      "loss": 3.1852757930755615,
      "step": 1323
    },
    {
      "epoch": 0.25540123456790126,
      "grad_norm": 1.176139235496521,
      "learning_rate": 0.0005,
      "loss": 3.2204980850219727,
      "step": 1324
    },
    {
      "epoch": 0.25559413580246915,
      "grad_norm": 1.0935121774673462,
      "learning_rate": 0.0005,
      "loss": 3.3625495433807373,
      "step": 1325
    },
    {
      "epoch": 0.25578703703703703,
      "grad_norm": 1.2767242193222046,
      "learning_rate": 0.0005,
      "loss": 2.8289437294006348,
      "step": 1326
    },
    {
      "epoch": 0.2559799382716049,
      "grad_norm": 1.0979329347610474,
      "learning_rate": 0.0005,
      "loss": 3.2557404041290283,
      "step": 1327
    },
    {
      "epoch": 0.25617283950617287,
      "grad_norm": 0.9773916602134705,
      "learning_rate": 0.0005,
      "loss": 2.9889354705810547,
      "step": 1328
    },
    {
      "epoch": 0.25636574074074076,
      "grad_norm": 1.5702905654907227,
      "learning_rate": 0.0005,
      "loss": 3.1909542083740234,
      "step": 1329
    },
    {
      "epoch": 0.25655864197530864,
      "grad_norm": 1.10136878490448,
      "learning_rate": 0.0005,
      "loss": 3.011765956878662,
      "step": 1330
    },
    {
      "epoch": 0.25675154320987653,
      "grad_norm": 1.7427232265472412,
      "learning_rate": 0.0005,
      "loss": 2.627765655517578,
      "step": 1331
    },
    {
      "epoch": 0.2569444444444444,
      "grad_norm": 1.1621804237365723,
      "learning_rate": 0.0005,
      "loss": 3.1277785301208496,
      "step": 1332
    },
    {
      "epoch": 0.25713734567901236,
      "grad_norm": 1.138451337814331,
      "learning_rate": 0.0005,
      "loss": 3.1107423305511475,
      "step": 1333
    },
    {
      "epoch": 0.25733024691358025,
      "grad_norm": 1.213721752166748,
      "learning_rate": 0.0005,
      "loss": 3.095836877822876,
      "step": 1334
    },
    {
      "epoch": 0.25752314814814814,
      "grad_norm": 1.2214093208312988,
      "learning_rate": 0.0005,
      "loss": 3.3666000366210938,
      "step": 1335
    },
    {
      "epoch": 0.25771604938271603,
      "grad_norm": 1.2455021142959595,
      "learning_rate": 0.0005,
      "loss": 3.0320916175842285,
      "step": 1336
    },
    {
      "epoch": 0.25790895061728397,
      "grad_norm": 1.3127659559249878,
      "learning_rate": 0.0005,
      "loss": 2.852518320083618,
      "step": 1337
    },
    {
      "epoch": 0.25810185185185186,
      "grad_norm": 1.3739936351776123,
      "learning_rate": 0.0005,
      "loss": 2.7863354682922363,
      "step": 1338
    },
    {
      "epoch": 0.25829475308641975,
      "grad_norm": 1.3191876411437988,
      "learning_rate": 0.0005,
      "loss": 2.9075117111206055,
      "step": 1339
    },
    {
      "epoch": 0.25848765432098764,
      "grad_norm": 1.5768543481826782,
      "learning_rate": 0.0005,
      "loss": 3.1992249488830566,
      "step": 1340
    },
    {
      "epoch": 0.2586805555555556,
      "grad_norm": 1.7083443403244019,
      "learning_rate": 0.0005,
      "loss": 3.248654842376709,
      "step": 1341
    },
    {
      "epoch": 0.25887345679012347,
      "grad_norm": 1.0411288738250732,
      "learning_rate": 0.0005,
      "loss": 3.461453914642334,
      "step": 1342
    },
    {
      "epoch": 0.25906635802469136,
      "grad_norm": 2.1666462421417236,
      "learning_rate": 0.0005,
      "loss": 3.1039462089538574,
      "step": 1343
    },
    {
      "epoch": 0.25925925925925924,
      "grad_norm": 2.6694302558898926,
      "learning_rate": 0.0005,
      "loss": 3.283993721008301,
      "step": 1344
    },
    {
      "epoch": 0.25945216049382713,
      "grad_norm": 1.2278252840042114,
      "learning_rate": 0.0005,
      "loss": 2.9682862758636475,
      "step": 1345
    },
    {
      "epoch": 0.2596450617283951,
      "grad_norm": 1.832724928855896,
      "learning_rate": 0.0005,
      "loss": 3.564518928527832,
      "step": 1346
    },
    {
      "epoch": 0.25983796296296297,
      "grad_norm": 1.2860325574874878,
      "learning_rate": 0.0005,
      "loss": 3.2473037242889404,
      "step": 1347
    },
    {
      "epoch": 0.26003086419753085,
      "grad_norm": 1.4056739807128906,
      "learning_rate": 0.0005,
      "loss": 3.2993083000183105,
      "step": 1348
    },
    {
      "epoch": 0.26022376543209874,
      "grad_norm": 1.1520553827285767,
      "learning_rate": 0.0005,
      "loss": 2.8896732330322266,
      "step": 1349
    },
    {
      "epoch": 0.2604166666666667,
      "grad_norm": 1.7650319337844849,
      "learning_rate": 0.0005,
      "loss": 3.1438770294189453,
      "step": 1350
    },
    {
      "epoch": 0.2606095679012346,
      "grad_norm": 1.197609782218933,
      "learning_rate": 0.0005,
      "loss": 3.05595064163208,
      "step": 1351
    },
    {
      "epoch": 0.26080246913580246,
      "grad_norm": 1.2172296047210693,
      "learning_rate": 0.0005,
      "loss": 2.9846558570861816,
      "step": 1352
    },
    {
      "epoch": 0.26099537037037035,
      "grad_norm": 1.1539822816848755,
      "learning_rate": 0.0005,
      "loss": 3.1790781021118164,
      "step": 1353
    },
    {
      "epoch": 0.2611882716049383,
      "grad_norm": 1.0490851402282715,
      "learning_rate": 0.0005,
      "loss": 3.117785930633545,
      "step": 1354
    },
    {
      "epoch": 0.2613811728395062,
      "grad_norm": 1.2208465337753296,
      "learning_rate": 0.0005,
      "loss": 3.0466227531433105,
      "step": 1355
    },
    {
      "epoch": 0.26157407407407407,
      "grad_norm": 0.9779089689254761,
      "learning_rate": 0.0005,
      "loss": 3.057281732559204,
      "step": 1356
    },
    {
      "epoch": 0.26176697530864196,
      "grad_norm": 1.1905591487884521,
      "learning_rate": 0.0005,
      "loss": 3.3139350414276123,
      "step": 1357
    },
    {
      "epoch": 0.2619598765432099,
      "grad_norm": 1.1475272178649902,
      "learning_rate": 0.0005,
      "loss": 2.8570923805236816,
      "step": 1358
    },
    {
      "epoch": 0.2621527777777778,
      "grad_norm": 1.2281485795974731,
      "learning_rate": 0.0005,
      "loss": 3.0013792514801025,
      "step": 1359
    },
    {
      "epoch": 0.2623456790123457,
      "grad_norm": 1.0548667907714844,
      "learning_rate": 0.0005,
      "loss": 3.1258978843688965,
      "step": 1360
    },
    {
      "epoch": 0.26253858024691357,
      "grad_norm": 1.0842540264129639,
      "learning_rate": 0.0005,
      "loss": 3.1786513328552246,
      "step": 1361
    },
    {
      "epoch": 0.26273148148148145,
      "grad_norm": 1.297906517982483,
      "learning_rate": 0.0005,
      "loss": 3.3154096603393555,
      "step": 1362
    },
    {
      "epoch": 0.2629243827160494,
      "grad_norm": 1.0318034887313843,
      "learning_rate": 0.0005,
      "loss": 2.869579315185547,
      "step": 1363
    },
    {
      "epoch": 0.2631172839506173,
      "grad_norm": 1.5289322137832642,
      "learning_rate": 0.0005,
      "loss": 3.065917491912842,
      "step": 1364
    },
    {
      "epoch": 0.2633101851851852,
      "grad_norm": 1.4675397872924805,
      "learning_rate": 0.0005,
      "loss": 3.283946990966797,
      "step": 1365
    },
    {
      "epoch": 0.26350308641975306,
      "grad_norm": 1.9864387512207031,
      "learning_rate": 0.0005,
      "loss": 3.450622081756592,
      "step": 1366
    },
    {
      "epoch": 0.263695987654321,
      "grad_norm": 1.3815094232559204,
      "learning_rate": 0.0005,
      "loss": 3.288295269012451,
      "step": 1367
    },
    {
      "epoch": 0.2638888888888889,
      "grad_norm": 1.9123926162719727,
      "learning_rate": 0.0005,
      "loss": 3.410879135131836,
      "step": 1368
    },
    {
      "epoch": 0.2640817901234568,
      "grad_norm": 1.2582423686981201,
      "learning_rate": 0.0005,
      "loss": 2.8016462326049805,
      "step": 1369
    },
    {
      "epoch": 0.26427469135802467,
      "grad_norm": 1.5459673404693604,
      "learning_rate": 0.0005,
      "loss": 3.154170036315918,
      "step": 1370
    },
    {
      "epoch": 0.2644675925925926,
      "grad_norm": 2.25335955619812,
      "learning_rate": 0.0005,
      "loss": 3.121142864227295,
      "step": 1371
    },
    {
      "epoch": 0.2646604938271605,
      "grad_norm": 1.26567542552948,
      "learning_rate": 0.0005,
      "loss": 3.0452451705932617,
      "step": 1372
    },
    {
      "epoch": 0.2648533950617284,
      "grad_norm": 1.0601123571395874,
      "learning_rate": 0.0005,
      "loss": 3.0272207260131836,
      "step": 1373
    },
    {
      "epoch": 0.2650462962962963,
      "grad_norm": 1.047709345817566,
      "learning_rate": 0.0005,
      "loss": 3.164313793182373,
      "step": 1374
    },
    {
      "epoch": 0.2652391975308642,
      "grad_norm": 1.2422727346420288,
      "learning_rate": 0.0005,
      "loss": 3.25363826751709,
      "step": 1375
    },
    {
      "epoch": 0.2654320987654321,
      "grad_norm": 1.1408923864364624,
      "learning_rate": 0.0005,
      "loss": 2.6499030590057373,
      "step": 1376
    },
    {
      "epoch": 0.265625,
      "grad_norm": 1.028905987739563,
      "learning_rate": 0.0005,
      "loss": 3.0251431465148926,
      "step": 1377
    },
    {
      "epoch": 0.2658179012345679,
      "grad_norm": 1.2322734594345093,
      "learning_rate": 0.0005,
      "loss": 3.3015503883361816,
      "step": 1378
    },
    {
      "epoch": 0.2660108024691358,
      "grad_norm": 1.327665090560913,
      "learning_rate": 0.0005,
      "loss": 3.0067081451416016,
      "step": 1379
    },
    {
      "epoch": 0.2662037037037037,
      "grad_norm": 1.452215552330017,
      "learning_rate": 0.0005,
      "loss": 3.3260483741760254,
      "step": 1380
    },
    {
      "epoch": 0.2663966049382716,
      "grad_norm": 1.3750605583190918,
      "learning_rate": 0.0005,
      "loss": 3.058302402496338,
      "step": 1381
    },
    {
      "epoch": 0.2665895061728395,
      "grad_norm": 1.2903066873550415,
      "learning_rate": 0.0005,
      "loss": 2.865356922149658,
      "step": 1382
    },
    {
      "epoch": 0.2667824074074074,
      "grad_norm": 1.17560875415802,
      "learning_rate": 0.0005,
      "loss": 3.295301914215088,
      "step": 1383
    },
    {
      "epoch": 0.26697530864197533,
      "grad_norm": 1.4292699098587036,
      "learning_rate": 0.0005,
      "loss": 2.937778949737549,
      "step": 1384
    },
    {
      "epoch": 0.2671682098765432,
      "grad_norm": 1.7447136640548706,
      "learning_rate": 0.0005,
      "loss": 3.0173847675323486,
      "step": 1385
    },
    {
      "epoch": 0.2673611111111111,
      "grad_norm": 14.672932624816895,
      "learning_rate": 0.0005,
      "loss": 2.9759135246276855,
      "step": 1386
    },
    {
      "epoch": 0.267554012345679,
      "grad_norm": 1.3912798166275024,
      "learning_rate": 0.0005,
      "loss": 2.974574089050293,
      "step": 1387
    },
    {
      "epoch": 0.26774691358024694,
      "grad_norm": 1.5064204931259155,
      "learning_rate": 0.0005,
      "loss": 3.0607385635375977,
      "step": 1388
    },
    {
      "epoch": 0.2679398148148148,
      "grad_norm": 1.2243019342422485,
      "learning_rate": 0.0005,
      "loss": 3.25150203704834,
      "step": 1389
    },
    {
      "epoch": 0.2681327160493827,
      "grad_norm": 1.8729947805404663,
      "learning_rate": 0.0005,
      "loss": 3.5017447471618652,
      "step": 1390
    },
    {
      "epoch": 0.2683256172839506,
      "grad_norm": 1.16972815990448,
      "learning_rate": 0.0005,
      "loss": 3.3161208629608154,
      "step": 1391
    },
    {
      "epoch": 0.26851851851851855,
      "grad_norm": 1.1685324907302856,
      "learning_rate": 0.0005,
      "loss": 3.331333637237549,
      "step": 1392
    },
    {
      "epoch": 0.26871141975308643,
      "grad_norm": 1.0211735963821411,
      "learning_rate": 0.0005,
      "loss": 3.407620906829834,
      "step": 1393
    },
    {
      "epoch": 0.2689043209876543,
      "grad_norm": 1.1381590366363525,
      "learning_rate": 0.0005,
      "loss": 3.1225500106811523,
      "step": 1394
    },
    {
      "epoch": 0.2690972222222222,
      "grad_norm": 1.1195327043533325,
      "learning_rate": 0.0005,
      "loss": 3.2464232444763184,
      "step": 1395
    },
    {
      "epoch": 0.2692901234567901,
      "grad_norm": 1.1213867664337158,
      "learning_rate": 0.0005,
      "loss": 2.8454031944274902,
      "step": 1396
    },
    {
      "epoch": 0.26948302469135804,
      "grad_norm": 1.5088833570480347,
      "learning_rate": 0.0005,
      "loss": 3.0573244094848633,
      "step": 1397
    },
    {
      "epoch": 0.26967592592592593,
      "grad_norm": 1.001209020614624,
      "learning_rate": 0.0005,
      "loss": 2.9268782138824463,
      "step": 1398
    },
    {
      "epoch": 0.2698688271604938,
      "grad_norm": 0.9765641093254089,
      "learning_rate": 0.0005,
      "loss": 3.0252182483673096,
      "step": 1399
    },
    {
      "epoch": 0.2700617283950617,
      "grad_norm": 1.0539990663528442,
      "learning_rate": 0.0005,
      "loss": 3.310187816619873,
      "step": 1400
    },
    {
      "epoch": 0.27025462962962965,
      "grad_norm": 0.9781040549278259,
      "learning_rate": 0.0005,
      "loss": 2.8173344135284424,
      "step": 1401
    },
    {
      "epoch": 0.27044753086419754,
      "grad_norm": 1.0781927108764648,
      "learning_rate": 0.0005,
      "loss": 3.1082029342651367,
      "step": 1402
    },
    {
      "epoch": 0.2706404320987654,
      "grad_norm": 1.1271394491195679,
      "learning_rate": 0.0005,
      "loss": 3.2380571365356445,
      "step": 1403
    },
    {
      "epoch": 0.2708333333333333,
      "grad_norm": 0.9748387336730957,
      "learning_rate": 0.0005,
      "loss": 2.9465553760528564,
      "step": 1404
    },
    {
      "epoch": 0.27102623456790126,
      "grad_norm": 1.1975244283676147,
      "learning_rate": 0.0005,
      "loss": 3.0801429748535156,
      "step": 1405
    },
    {
      "epoch": 0.27121913580246915,
      "grad_norm": 1.3642866611480713,
      "learning_rate": 0.0005,
      "loss": 2.9265856742858887,
      "step": 1406
    },
    {
      "epoch": 0.27141203703703703,
      "grad_norm": 1.350411295890808,
      "learning_rate": 0.0005,
      "loss": 2.972567558288574,
      "step": 1407
    },
    {
      "epoch": 0.2716049382716049,
      "grad_norm": 1.2443890571594238,
      "learning_rate": 0.0005,
      "loss": 3.119774580001831,
      "step": 1408
    },
    {
      "epoch": 0.27179783950617287,
      "grad_norm": 1.0580850839614868,
      "learning_rate": 0.0005,
      "loss": 2.9583919048309326,
      "step": 1409
    },
    {
      "epoch": 0.27199074074074076,
      "grad_norm": 1.6327548027038574,
      "learning_rate": 0.0005,
      "loss": 3.199108123779297,
      "step": 1410
    },
    {
      "epoch": 0.27218364197530864,
      "grad_norm": 1.0311663150787354,
      "learning_rate": 0.0005,
      "loss": 3.1445107460021973,
      "step": 1411
    },
    {
      "epoch": 0.27237654320987653,
      "grad_norm": 1.5020384788513184,
      "learning_rate": 0.0005,
      "loss": 2.989419937133789,
      "step": 1412
    },
    {
      "epoch": 0.2725694444444444,
      "grad_norm": 1.6867189407348633,
      "learning_rate": 0.0005,
      "loss": 3.293276071548462,
      "step": 1413
    },
    {
      "epoch": 0.27276234567901236,
      "grad_norm": 1.1877011060714722,
      "learning_rate": 0.0005,
      "loss": 3.253370761871338,
      "step": 1414
    },
    {
      "epoch": 0.27295524691358025,
      "grad_norm": 1.2133152484893799,
      "learning_rate": 0.0005,
      "loss": 3.132547616958618,
      "step": 1415
    },
    {
      "epoch": 0.27314814814814814,
      "grad_norm": 1.1458683013916016,
      "learning_rate": 0.0005,
      "loss": 3.2272958755493164,
      "step": 1416
    },
    {
      "epoch": 0.27334104938271603,
      "grad_norm": 1.326464295387268,
      "learning_rate": 0.0005,
      "loss": 3.4422459602355957,
      "step": 1417
    },
    {
      "epoch": 0.27353395061728397,
      "grad_norm": 1.0794563293457031,
      "learning_rate": 0.0005,
      "loss": 2.931384325027466,
      "step": 1418
    },
    {
      "epoch": 0.27372685185185186,
      "grad_norm": 1.321020245552063,
      "learning_rate": 0.0005,
      "loss": 3.4220495223999023,
      "step": 1419
    },
    {
      "epoch": 0.27391975308641975,
      "grad_norm": 1.1315463781356812,
      "learning_rate": 0.0005,
      "loss": 2.6891281604766846,
      "step": 1420
    },
    {
      "epoch": 0.27411265432098764,
      "grad_norm": 1.1998027563095093,
      "learning_rate": 0.0005,
      "loss": 2.866701126098633,
      "step": 1421
    },
    {
      "epoch": 0.2743055555555556,
      "grad_norm": 1.1826812028884888,
      "learning_rate": 0.0005,
      "loss": 2.9053711891174316,
      "step": 1422
    },
    {
      "epoch": 0.27449845679012347,
      "grad_norm": 0.9316769242286682,
      "learning_rate": 0.0005,
      "loss": 3.0122694969177246,
      "step": 1423
    },
    {
      "epoch": 0.27469135802469136,
      "grad_norm": 1.08944833278656,
      "learning_rate": 0.0005,
      "loss": 3.083057165145874,
      "step": 1424
    },
    {
      "epoch": 0.27488425925925924,
      "grad_norm": 1.1121701002120972,
      "learning_rate": 0.0005,
      "loss": 3.238755226135254,
      "step": 1425
    },
    {
      "epoch": 0.27507716049382713,
      "grad_norm": 1.319972038269043,
      "learning_rate": 0.0005,
      "loss": 2.984229564666748,
      "step": 1426
    },
    {
      "epoch": 0.2752700617283951,
      "grad_norm": 1.2412772178649902,
      "learning_rate": 0.0005,
      "loss": 3.1991491317749023,
      "step": 1427
    },
    {
      "epoch": 0.27546296296296297,
      "grad_norm": 0.9434378743171692,
      "learning_rate": 0.0005,
      "loss": 3.3540496826171875,
      "step": 1428
    },
    {
      "epoch": 0.27565586419753085,
      "grad_norm": 1.2271219491958618,
      "learning_rate": 0.0005,
      "loss": 3.024188995361328,
      "step": 1429
    },
    {
      "epoch": 0.27584876543209874,
      "grad_norm": 0.9993361234664917,
      "learning_rate": 0.0005,
      "loss": 3.4305896759033203,
      "step": 1430
    },
    {
      "epoch": 0.2760416666666667,
      "grad_norm": 1.122151494026184,
      "learning_rate": 0.0005,
      "loss": 3.2006213665008545,
      "step": 1431
    },
    {
      "epoch": 0.2762345679012346,
      "grad_norm": 2.0053296089172363,
      "learning_rate": 0.0005,
      "loss": 2.9976210594177246,
      "step": 1432
    },
    {
      "epoch": 0.27642746913580246,
      "grad_norm": 1.2285410165786743,
      "learning_rate": 0.0005,
      "loss": 3.3181040287017822,
      "step": 1433
    },
    {
      "epoch": 0.27662037037037035,
      "grad_norm": 4.81544828414917,
      "learning_rate": 0.0005,
      "loss": 3.13309907913208,
      "step": 1434
    },
    {
      "epoch": 0.2768132716049383,
      "grad_norm": 1.201432228088379,
      "learning_rate": 0.0005,
      "loss": 2.9484474658966064,
      "step": 1435
    },
    {
      "epoch": 0.2770061728395062,
      "grad_norm": 1.3300158977508545,
      "learning_rate": 0.0005,
      "loss": 3.166450262069702,
      "step": 1436
    },
    {
      "epoch": 0.27719907407407407,
      "grad_norm": 1.2561050653457642,
      "learning_rate": 0.0005,
      "loss": 3.0630831718444824,
      "step": 1437
    },
    {
      "epoch": 0.27739197530864196,
      "grad_norm": 1.2108745574951172,
      "learning_rate": 0.0005,
      "loss": 2.7692184448242188,
      "step": 1438
    },
    {
      "epoch": 0.2775848765432099,
      "grad_norm": 1.2003686428070068,
      "learning_rate": 0.0005,
      "loss": 3.124025344848633,
      "step": 1439
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 1.16705322265625,
      "learning_rate": 0.0005,
      "loss": 3.3811755180358887,
      "step": 1440
    },
    {
      "epoch": 0.2779706790123457,
      "grad_norm": 0.9143825173377991,
      "learning_rate": 0.0005,
      "loss": 3.05228853225708,
      "step": 1441
    },
    {
      "epoch": 0.27816358024691357,
      "grad_norm": 1.1711336374282837,
      "learning_rate": 0.0005,
      "loss": 3.1054329872131348,
      "step": 1442
    },
    {
      "epoch": 0.27835648148148145,
      "grad_norm": 1.352781057357788,
      "learning_rate": 0.0005,
      "loss": 3.4438071250915527,
      "step": 1443
    },
    {
      "epoch": 0.2785493827160494,
      "grad_norm": 1.066315770149231,
      "learning_rate": 0.0005,
      "loss": 3.3088912963867188,
      "step": 1444
    },
    {
      "epoch": 0.2787422839506173,
      "grad_norm": 1.0195189714431763,
      "learning_rate": 0.0005,
      "loss": 3.1205880641937256,
      "step": 1445
    },
    {
      "epoch": 0.2789351851851852,
      "grad_norm": 1.1410871744155884,
      "learning_rate": 0.0005,
      "loss": 3.0841779708862305,
      "step": 1446
    },
    {
      "epoch": 0.27912808641975306,
      "grad_norm": 1.2203128337860107,
      "learning_rate": 0.0005,
      "loss": 2.8398947715759277,
      "step": 1447
    },
    {
      "epoch": 0.279320987654321,
      "grad_norm": 1.1263749599456787,
      "learning_rate": 0.0005,
      "loss": 2.9255433082580566,
      "step": 1448
    },
    {
      "epoch": 0.2795138888888889,
      "grad_norm": 1.1438707113265991,
      "learning_rate": 0.0005,
      "loss": 3.5102505683898926,
      "step": 1449
    },
    {
      "epoch": 0.2797067901234568,
      "grad_norm": 1.1483728885650635,
      "learning_rate": 0.0005,
      "loss": 3.3323922157287598,
      "step": 1450
    },
    {
      "epoch": 0.27989969135802467,
      "grad_norm": 0.9201925992965698,
      "learning_rate": 0.0005,
      "loss": 2.7537589073181152,
      "step": 1451
    },
    {
      "epoch": 0.2800925925925926,
      "grad_norm": 1.0872764587402344,
      "learning_rate": 0.0005,
      "loss": 3.0283355712890625,
      "step": 1452
    },
    {
      "epoch": 0.2802854938271605,
      "grad_norm": 0.9709160327911377,
      "learning_rate": 0.0005,
      "loss": 2.940776824951172,
      "step": 1453
    },
    {
      "epoch": 0.2804783950617284,
      "grad_norm": 1.0380921363830566,
      "learning_rate": 0.0005,
      "loss": 3.07047963142395,
      "step": 1454
    },
    {
      "epoch": 0.2806712962962963,
      "grad_norm": 1.3733843564987183,
      "learning_rate": 0.0005,
      "loss": 2.96885085105896,
      "step": 1455
    },
    {
      "epoch": 0.2808641975308642,
      "grad_norm": 1.3472265005111694,
      "learning_rate": 0.0005,
      "loss": 3.121612548828125,
      "step": 1456
    },
    {
      "epoch": 0.2810570987654321,
      "grad_norm": 1.4170421361923218,
      "learning_rate": 0.0005,
      "loss": 3.0444202423095703,
      "step": 1457
    },
    {
      "epoch": 0.28125,
      "grad_norm": 1.9643678665161133,
      "learning_rate": 0.0005,
      "loss": 2.9749977588653564,
      "step": 1458
    },
    {
      "epoch": 0.2814429012345679,
      "grad_norm": 1.0105599164962769,
      "learning_rate": 0.0005,
      "loss": 2.9008774757385254,
      "step": 1459
    },
    {
      "epoch": 0.2816358024691358,
      "grad_norm": 1.154788851737976,
      "learning_rate": 0.0005,
      "loss": 3.061365842819214,
      "step": 1460
    },
    {
      "epoch": 0.2818287037037037,
      "grad_norm": 1.1740343570709229,
      "learning_rate": 0.0005,
      "loss": 3.398500919342041,
      "step": 1461
    },
    {
      "epoch": 0.2820216049382716,
      "grad_norm": 1.1120940446853638,
      "learning_rate": 0.0005,
      "loss": 2.9307618141174316,
      "step": 1462
    },
    {
      "epoch": 0.2822145061728395,
      "grad_norm": 1.0520544052124023,
      "learning_rate": 0.0005,
      "loss": 3.0179531574249268,
      "step": 1463
    },
    {
      "epoch": 0.2824074074074074,
      "grad_norm": 1.172890067100525,
      "learning_rate": 0.0005,
      "loss": 3.125662088394165,
      "step": 1464
    },
    {
      "epoch": 0.28260030864197533,
      "grad_norm": 1.4665873050689697,
      "learning_rate": 0.0005,
      "loss": 3.010310173034668,
      "step": 1465
    },
    {
      "epoch": 0.2827932098765432,
      "grad_norm": 1.2621527910232544,
      "learning_rate": 0.0005,
      "loss": 3.028557777404785,
      "step": 1466
    },
    {
      "epoch": 0.2829861111111111,
      "grad_norm": 0.9346606731414795,
      "learning_rate": 0.0005,
      "loss": 3.2764317989349365,
      "step": 1467
    },
    {
      "epoch": 0.283179012345679,
      "grad_norm": 1.2040388584136963,
      "learning_rate": 0.0005,
      "loss": 2.6058568954467773,
      "step": 1468
    },
    {
      "epoch": 0.28337191358024694,
      "grad_norm": 0.9911813735961914,
      "learning_rate": 0.0005,
      "loss": 3.2156455516815186,
      "step": 1469
    },
    {
      "epoch": 0.2835648148148148,
      "grad_norm": 1.0673060417175293,
      "learning_rate": 0.0005,
      "loss": 2.838820695877075,
      "step": 1470
    },
    {
      "epoch": 0.2837577160493827,
      "grad_norm": 0.9854177832603455,
      "learning_rate": 0.0005,
      "loss": 2.7635531425476074,
      "step": 1471
    },
    {
      "epoch": 0.2839506172839506,
      "grad_norm": 1.0370179414749146,
      "learning_rate": 0.0005,
      "loss": 3.2496628761291504,
      "step": 1472
    },
    {
      "epoch": 0.28414351851851855,
      "grad_norm": 1.1323490142822266,
      "learning_rate": 0.0005,
      "loss": 3.2404890060424805,
      "step": 1473
    },
    {
      "epoch": 0.28433641975308643,
      "grad_norm": 1.076130747795105,
      "learning_rate": 0.0005,
      "loss": 3.0732648372650146,
      "step": 1474
    },
    {
      "epoch": 0.2845293209876543,
      "grad_norm": 1.0663388967514038,
      "learning_rate": 0.0005,
      "loss": 3.1579532623291016,
      "step": 1475
    },
    {
      "epoch": 0.2847222222222222,
      "grad_norm": 1.0445690155029297,
      "learning_rate": 0.0005,
      "loss": 2.7393336296081543,
      "step": 1476
    },
    {
      "epoch": 0.2849151234567901,
      "grad_norm": 0.9490758776664734,
      "learning_rate": 0.0005,
      "loss": 3.046130418777466,
      "step": 1477
    },
    {
      "epoch": 0.28510802469135804,
      "grad_norm": 1.8869954347610474,
      "learning_rate": 0.0005,
      "loss": 3.246415615081787,
      "step": 1478
    },
    {
      "epoch": 0.28530092592592593,
      "grad_norm": 1.2353553771972656,
      "learning_rate": 0.0005,
      "loss": 3.300375461578369,
      "step": 1479
    },
    {
      "epoch": 0.2854938271604938,
      "grad_norm": 1.2125542163848877,
      "learning_rate": 0.0005,
      "loss": 3.049602746963501,
      "step": 1480
    },
    {
      "epoch": 0.2856867283950617,
      "grad_norm": 1.19862961769104,
      "learning_rate": 0.0005,
      "loss": 2.852203369140625,
      "step": 1481
    },
    {
      "epoch": 0.28587962962962965,
      "grad_norm": 1.0643165111541748,
      "learning_rate": 0.0005,
      "loss": 3.2454802989959717,
      "step": 1482
    },
    {
      "epoch": 0.28607253086419754,
      "grad_norm": 1.0433396100997925,
      "learning_rate": 0.0005,
      "loss": 3.235053062438965,
      "step": 1483
    },
    {
      "epoch": 0.2862654320987654,
      "grad_norm": 1.1440727710723877,
      "learning_rate": 0.0005,
      "loss": 2.7824268341064453,
      "step": 1484
    },
    {
      "epoch": 0.2864583333333333,
      "grad_norm": 1.2212200164794922,
      "learning_rate": 0.0005,
      "loss": 3.363844871520996,
      "step": 1485
    },
    {
      "epoch": 0.28665123456790126,
      "grad_norm": 23.138992309570312,
      "learning_rate": 0.0005,
      "loss": 3.1023168563842773,
      "step": 1486
    },
    {
      "epoch": 0.28684413580246915,
      "grad_norm": 1.9028170108795166,
      "learning_rate": 0.0005,
      "loss": 3.0052366256713867,
      "step": 1487
    },
    {
      "epoch": 0.28703703703703703,
      "grad_norm": 1.1798754930496216,
      "learning_rate": 0.0005,
      "loss": 2.702564239501953,
      "step": 1488
    },
    {
      "epoch": 0.2872299382716049,
      "grad_norm": 1.0081055164337158,
      "learning_rate": 0.0005,
      "loss": 2.9856865406036377,
      "step": 1489
    },
    {
      "epoch": 0.28742283950617287,
      "grad_norm": 0.9386415481567383,
      "learning_rate": 0.0005,
      "loss": 3.2220141887664795,
      "step": 1490
    },
    {
      "epoch": 0.28761574074074076,
      "grad_norm": 0.957796037197113,
      "learning_rate": 0.0005,
      "loss": 2.997018337249756,
      "step": 1491
    },
    {
      "epoch": 0.28780864197530864,
      "grad_norm": 0.9893945455551147,
      "learning_rate": 0.0005,
      "loss": 2.828974962234497,
      "step": 1492
    },
    {
      "epoch": 0.28800154320987653,
      "grad_norm": 0.9837633967399597,
      "learning_rate": 0.0005,
      "loss": 2.9025280475616455,
      "step": 1493
    },
    {
      "epoch": 0.2881944444444444,
      "grad_norm": 1.0399386882781982,
      "learning_rate": 0.0005,
      "loss": 3.0851755142211914,
      "step": 1494
    },
    {
      "epoch": 0.28838734567901236,
      "grad_norm": 1.118652105331421,
      "learning_rate": 0.0005,
      "loss": 3.4446282386779785,
      "step": 1495
    },
    {
      "epoch": 0.28858024691358025,
      "grad_norm": 1.129905343055725,
      "learning_rate": 0.0005,
      "loss": 3.1338894367218018,
      "step": 1496
    },
    {
      "epoch": 0.28877314814814814,
      "grad_norm": 1.2780331373214722,
      "learning_rate": 0.0005,
      "loss": 3.1057910919189453,
      "step": 1497
    },
    {
      "epoch": 0.28896604938271603,
      "grad_norm": 1.0787726640701294,
      "learning_rate": 0.0005,
      "loss": 3.2255749702453613,
      "step": 1498
    },
    {
      "epoch": 0.28915895061728397,
      "grad_norm": 1.0264068841934204,
      "learning_rate": 0.0005,
      "loss": 3.1492085456848145,
      "step": 1499
    },
    {
      "epoch": 0.28935185185185186,
      "grad_norm": 1.061030387878418,
      "learning_rate": 0.0005,
      "loss": 3.3490958213806152,
      "step": 1500
    },
    {
      "epoch": 0.28954475308641975,
      "grad_norm": 1.2560300827026367,
      "learning_rate": 0.0005,
      "loss": 3.2036848068237305,
      "step": 1501
    },
    {
      "epoch": 0.28973765432098764,
      "grad_norm": 1.2295655012130737,
      "learning_rate": 0.0005,
      "loss": 3.0750298500061035,
      "step": 1502
    },
    {
      "epoch": 0.2899305555555556,
      "grad_norm": 1.1920454502105713,
      "learning_rate": 0.0005,
      "loss": 3.023061752319336,
      "step": 1503
    },
    {
      "epoch": 0.29012345679012347,
      "grad_norm": 1.3965004682540894,
      "learning_rate": 0.0005,
      "loss": 2.8715968132019043,
      "step": 1504
    },
    {
      "epoch": 0.29031635802469136,
      "grad_norm": 0.9453427195549011,
      "learning_rate": 0.0005,
      "loss": 2.839097023010254,
      "step": 1505
    },
    {
      "epoch": 0.29050925925925924,
      "grad_norm": 1.0572898387908936,
      "learning_rate": 0.0005,
      "loss": 3.0031685829162598,
      "step": 1506
    },
    {
      "epoch": 0.29070216049382713,
      "grad_norm": 1.000300645828247,
      "learning_rate": 0.0005,
      "loss": 3.154433488845825,
      "step": 1507
    },
    {
      "epoch": 0.2908950617283951,
      "grad_norm": 0.9753072261810303,
      "learning_rate": 0.0005,
      "loss": 3.388913631439209,
      "step": 1508
    },
    {
      "epoch": 0.29108796296296297,
      "grad_norm": 1.2595422267913818,
      "learning_rate": 0.0005,
      "loss": 3.175689220428467,
      "step": 1509
    },
    {
      "epoch": 0.29128086419753085,
      "grad_norm": 1.6582353115081787,
      "learning_rate": 0.0005,
      "loss": 3.0195579528808594,
      "step": 1510
    },
    {
      "epoch": 0.29147376543209874,
      "grad_norm": 1.1948128938674927,
      "learning_rate": 0.0005,
      "loss": 3.115260124206543,
      "step": 1511
    },
    {
      "epoch": 0.2916666666666667,
      "grad_norm": 1.2433794736862183,
      "learning_rate": 0.0005,
      "loss": 3.08060884475708,
      "step": 1512
    },
    {
      "epoch": 0.2918595679012346,
      "grad_norm": 2.272289991378784,
      "learning_rate": 0.0005,
      "loss": 3.2522940635681152,
      "step": 1513
    },
    {
      "epoch": 0.29205246913580246,
      "grad_norm": 0.9873656630516052,
      "learning_rate": 0.0005,
      "loss": 2.9814443588256836,
      "step": 1514
    },
    {
      "epoch": 0.29224537037037035,
      "grad_norm": 1.1507575511932373,
      "learning_rate": 0.0005,
      "loss": 2.7684996128082275,
      "step": 1515
    },
    {
      "epoch": 0.2924382716049383,
      "grad_norm": 0.8420865535736084,
      "learning_rate": 0.0005,
      "loss": 2.9153823852539062,
      "step": 1516
    },
    {
      "epoch": 0.2926311728395062,
      "grad_norm": 0.949357807636261,
      "learning_rate": 0.0005,
      "loss": 3.1091604232788086,
      "step": 1517
    },
    {
      "epoch": 0.29282407407407407,
      "grad_norm": 1.0033303499221802,
      "learning_rate": 0.0005,
      "loss": 3.003309488296509,
      "step": 1518
    },
    {
      "epoch": 0.29301697530864196,
      "grad_norm": 1.2270556688308716,
      "learning_rate": 0.0005,
      "loss": 2.925006866455078,
      "step": 1519
    },
    {
      "epoch": 0.2932098765432099,
      "grad_norm": 0.9790640473365784,
      "learning_rate": 0.0005,
      "loss": 2.8922479152679443,
      "step": 1520
    },
    {
      "epoch": 0.2934027777777778,
      "grad_norm": 1.178783655166626,
      "learning_rate": 0.0005,
      "loss": 3.473832607269287,
      "step": 1521
    },
    {
      "epoch": 0.2935956790123457,
      "grad_norm": 1.2463150024414062,
      "learning_rate": 0.0005,
      "loss": 2.826538324356079,
      "step": 1522
    },
    {
      "epoch": 0.29378858024691357,
      "grad_norm": 1.0489177703857422,
      "learning_rate": 0.0005,
      "loss": 3.3632185459136963,
      "step": 1523
    },
    {
      "epoch": 0.29398148148148145,
      "grad_norm": 1.0059261322021484,
      "learning_rate": 0.0005,
      "loss": 3.019458055496216,
      "step": 1524
    },
    {
      "epoch": 0.2941743827160494,
      "grad_norm": 1.2072103023529053,
      "learning_rate": 0.0005,
      "loss": 2.8891448974609375,
      "step": 1525
    },
    {
      "epoch": 0.2943672839506173,
      "grad_norm": 1.413474440574646,
      "learning_rate": 0.0005,
      "loss": 3.280156373977661,
      "step": 1526
    },
    {
      "epoch": 0.2945601851851852,
      "grad_norm": 0.931283712387085,
      "learning_rate": 0.0005,
      "loss": 3.160447597503662,
      "step": 1527
    },
    {
      "epoch": 0.29475308641975306,
      "grad_norm": 0.996932327747345,
      "learning_rate": 0.0005,
      "loss": 3.085289478302002,
      "step": 1528
    },
    {
      "epoch": 0.294945987654321,
      "grad_norm": 1.1195694208145142,
      "learning_rate": 0.0005,
      "loss": 2.8912854194641113,
      "step": 1529
    },
    {
      "epoch": 0.2951388888888889,
      "grad_norm": 0.9541357159614563,
      "learning_rate": 0.0005,
      "loss": 3.173870801925659,
      "step": 1530
    },
    {
      "epoch": 0.2953317901234568,
      "grad_norm": 1.039275050163269,
      "learning_rate": 0.0005,
      "loss": 2.9076385498046875,
      "step": 1531
    },
    {
      "epoch": 0.29552469135802467,
      "grad_norm": 1.269229769706726,
      "learning_rate": 0.0005,
      "loss": 2.8153305053710938,
      "step": 1532
    },
    {
      "epoch": 0.2957175925925926,
      "grad_norm": 1.1029244661331177,
      "learning_rate": 0.0005,
      "loss": 2.698246955871582,
      "step": 1533
    },
    {
      "epoch": 0.2959104938271605,
      "grad_norm": 1.162664771080017,
      "learning_rate": 0.0005,
      "loss": 3.090254545211792,
      "step": 1534
    },
    {
      "epoch": 0.2961033950617284,
      "grad_norm": 1.175369381904602,
      "learning_rate": 0.0005,
      "loss": 2.975517749786377,
      "step": 1535
    },
    {
      "epoch": 0.2962962962962963,
      "grad_norm": 1.286077618598938,
      "learning_rate": 0.0005,
      "loss": 2.9385733604431152,
      "step": 1536
    },
    {
      "epoch": 0.2964891975308642,
      "grad_norm": 1.0837650299072266,
      "learning_rate": 0.0005,
      "loss": 3.132345676422119,
      "step": 1537
    },
    {
      "epoch": 0.2966820987654321,
      "grad_norm": 0.9717402458190918,
      "learning_rate": 0.0005,
      "loss": 2.808781623840332,
      "step": 1538
    },
    {
      "epoch": 0.296875,
      "grad_norm": 1.08425772190094,
      "learning_rate": 0.0005,
      "loss": 2.7440929412841797,
      "step": 1539
    },
    {
      "epoch": 0.2970679012345679,
      "grad_norm": 1.0594062805175781,
      "learning_rate": 0.0005,
      "loss": 3.0093488693237305,
      "step": 1540
    },
    {
      "epoch": 0.2972608024691358,
      "grad_norm": 1.0251861810684204,
      "learning_rate": 0.0005,
      "loss": 2.9730687141418457,
      "step": 1541
    },
    {
      "epoch": 0.2974537037037037,
      "grad_norm": 1.4592422246932983,
      "learning_rate": 0.0005,
      "loss": 2.8722267150878906,
      "step": 1542
    },
    {
      "epoch": 0.2976466049382716,
      "grad_norm": 1.4315937757492065,
      "learning_rate": 0.0005,
      "loss": 3.2861552238464355,
      "step": 1543
    },
    {
      "epoch": 0.2978395061728395,
      "grad_norm": 2.1497585773468018,
      "learning_rate": 0.0005,
      "loss": 2.943974256515503,
      "step": 1544
    },
    {
      "epoch": 0.2980324074074074,
      "grad_norm": 1.8158682584762573,
      "learning_rate": 0.0005,
      "loss": 3.06166672706604,
      "step": 1545
    },
    {
      "epoch": 0.29822530864197533,
      "grad_norm": 1.3003045320510864,
      "learning_rate": 0.0005,
      "loss": 2.8824987411499023,
      "step": 1546
    },
    {
      "epoch": 0.2984182098765432,
      "grad_norm": 1.1338196992874146,
      "learning_rate": 0.0005,
      "loss": 3.05246639251709,
      "step": 1547
    },
    {
      "epoch": 0.2986111111111111,
      "grad_norm": 1.127642273902893,
      "learning_rate": 0.0005,
      "loss": 2.948775291442871,
      "step": 1548
    },
    {
      "epoch": 0.298804012345679,
      "grad_norm": 1.0240795612335205,
      "learning_rate": 0.0005,
      "loss": 2.8918871879577637,
      "step": 1549
    },
    {
      "epoch": 0.29899691358024694,
      "grad_norm": 1.1385051012039185,
      "learning_rate": 0.0005,
      "loss": 2.8852546215057373,
      "step": 1550
    },
    {
      "epoch": 0.2991898148148148,
      "grad_norm": 1.2859463691711426,
      "learning_rate": 0.0005,
      "loss": 2.9149112701416016,
      "step": 1551
    },
    {
      "epoch": 0.2993827160493827,
      "grad_norm": 0.9776811003684998,
      "learning_rate": 0.0005,
      "loss": 2.761040687561035,
      "step": 1552
    },
    {
      "epoch": 0.2995756172839506,
      "grad_norm": 1.370094656944275,
      "learning_rate": 0.0005,
      "loss": 3.056088924407959,
      "step": 1553
    },
    {
      "epoch": 0.29976851851851855,
      "grad_norm": 1.2351796627044678,
      "learning_rate": 0.0005,
      "loss": 3.069836139678955,
      "step": 1554
    },
    {
      "epoch": 0.29996141975308643,
      "grad_norm": 1.1105061769485474,
      "learning_rate": 0.0005,
      "loss": 2.7503581047058105,
      "step": 1555
    },
    {
      "epoch": 0.3001543209876543,
      "grad_norm": 1.112964391708374,
      "learning_rate": 0.0005,
      "loss": 3.3340659141540527,
      "step": 1556
    },
    {
      "epoch": 0.3003472222222222,
      "grad_norm": 0.9814454317092896,
      "learning_rate": 0.0005,
      "loss": 3.0000524520874023,
      "step": 1557
    },
    {
      "epoch": 0.3005401234567901,
      "grad_norm": 1.2844988107681274,
      "learning_rate": 0.0005,
      "loss": 3.1173932552337646,
      "step": 1558
    },
    {
      "epoch": 0.30073302469135804,
      "grad_norm": 1.2707549333572388,
      "learning_rate": 0.0005,
      "loss": 2.8988635540008545,
      "step": 1559
    },
    {
      "epoch": 0.30092592592592593,
      "grad_norm": 0.9952289462089539,
      "learning_rate": 0.0005,
      "loss": 2.969670295715332,
      "step": 1560
    },
    {
      "epoch": 0.3011188271604938,
      "grad_norm": 1.1566458940505981,
      "learning_rate": 0.0005,
      "loss": 2.9158058166503906,
      "step": 1561
    },
    {
      "epoch": 0.3013117283950617,
      "grad_norm": 1.0937600135803223,
      "learning_rate": 0.0005,
      "loss": 3.075655460357666,
      "step": 1562
    },
    {
      "epoch": 0.30150462962962965,
      "grad_norm": 3.9439024925231934,
      "learning_rate": 0.0005,
      "loss": 2.9592723846435547,
      "step": 1563
    },
    {
      "epoch": 0.30169753086419754,
      "grad_norm": 0.9347966909408569,
      "learning_rate": 0.0005,
      "loss": 2.9173195362091064,
      "step": 1564
    },
    {
      "epoch": 0.3018904320987654,
      "grad_norm": 0.9718467593193054,
      "learning_rate": 0.0005,
      "loss": 3.2784674167633057,
      "step": 1565
    },
    {
      "epoch": 0.3020833333333333,
      "grad_norm": 1.0116664171218872,
      "learning_rate": 0.0005,
      "loss": 3.2284913063049316,
      "step": 1566
    },
    {
      "epoch": 0.30227623456790126,
      "grad_norm": 0.9723894000053406,
      "learning_rate": 0.0005,
      "loss": 2.773686408996582,
      "step": 1567
    },
    {
      "epoch": 0.30246913580246915,
      "grad_norm": 1.6800596714019775,
      "learning_rate": 0.0005,
      "loss": 2.8334264755249023,
      "step": 1568
    },
    {
      "epoch": 0.30266203703703703,
      "grad_norm": 0.9347769021987915,
      "learning_rate": 0.0005,
      "loss": 2.8847556114196777,
      "step": 1569
    },
    {
      "epoch": 0.3028549382716049,
      "grad_norm": 1.0826294422149658,
      "learning_rate": 0.0005,
      "loss": 3.07425594329834,
      "step": 1570
    },
    {
      "epoch": 0.30304783950617287,
      "grad_norm": 1.1107029914855957,
      "learning_rate": 0.0005,
      "loss": 3.077554225921631,
      "step": 1571
    },
    {
      "epoch": 0.30324074074074076,
      "grad_norm": 1.0596611499786377,
      "learning_rate": 0.0005,
      "loss": 3.456753730773926,
      "step": 1572
    },
    {
      "epoch": 0.30343364197530864,
      "grad_norm": 1.0033527612686157,
      "learning_rate": 0.0005,
      "loss": 3.3163187503814697,
      "step": 1573
    },
    {
      "epoch": 0.30362654320987653,
      "grad_norm": 1.1744017601013184,
      "learning_rate": 0.0005,
      "loss": 3.555478572845459,
      "step": 1574
    },
    {
      "epoch": 0.3038194444444444,
      "grad_norm": 1.2162785530090332,
      "learning_rate": 0.0005,
      "loss": 3.022831439971924,
      "step": 1575
    },
    {
      "epoch": 0.30401234567901236,
      "grad_norm": 1.2482140064239502,
      "learning_rate": 0.0005,
      "loss": 2.9981789588928223,
      "step": 1576
    },
    {
      "epoch": 0.30420524691358025,
      "grad_norm": 1.6950832605361938,
      "learning_rate": 0.0005,
      "loss": 3.1372156143188477,
      "step": 1577
    },
    {
      "epoch": 0.30439814814814814,
      "grad_norm": 1.1665459871292114,
      "learning_rate": 0.0005,
      "loss": 3.091085910797119,
      "step": 1578
    },
    {
      "epoch": 0.30459104938271603,
      "grad_norm": 1.118943452835083,
      "learning_rate": 0.0005,
      "loss": 3.1960091590881348,
      "step": 1579
    },
    {
      "epoch": 0.30478395061728397,
      "grad_norm": 1.1394294500350952,
      "learning_rate": 0.0005,
      "loss": 2.8174080848693848,
      "step": 1580
    },
    {
      "epoch": 0.30497685185185186,
      "grad_norm": 1.0220314264297485,
      "learning_rate": 0.0005,
      "loss": 3.361525058746338,
      "step": 1581
    },
    {
      "epoch": 0.30516975308641975,
      "grad_norm": 1.0316832065582275,
      "learning_rate": 0.0005,
      "loss": 2.9034595489501953,
      "step": 1582
    },
    {
      "epoch": 0.30536265432098764,
      "grad_norm": 1.7895100116729736,
      "learning_rate": 0.0005,
      "loss": 2.9370675086975098,
      "step": 1583
    },
    {
      "epoch": 0.3055555555555556,
      "grad_norm": 1.1110780239105225,
      "learning_rate": 0.0005,
      "loss": 2.907743453979492,
      "step": 1584
    },
    {
      "epoch": 0.30574845679012347,
      "grad_norm": 1.0746454000473022,
      "learning_rate": 0.0005,
      "loss": 3.225191354751587,
      "step": 1585
    },
    {
      "epoch": 0.30594135802469136,
      "grad_norm": 1.0891410112380981,
      "learning_rate": 0.0005,
      "loss": 2.936998128890991,
      "step": 1586
    },
    {
      "epoch": 0.30613425925925924,
      "grad_norm": 0.9917362332344055,
      "learning_rate": 0.0005,
      "loss": 3.192096710205078,
      "step": 1587
    },
    {
      "epoch": 0.30632716049382713,
      "grad_norm": 1.1132838726043701,
      "learning_rate": 0.0005,
      "loss": 3.143911361694336,
      "step": 1588
    },
    {
      "epoch": 0.3065200617283951,
      "grad_norm": 1.111167073249817,
      "learning_rate": 0.0005,
      "loss": 3.2190847396850586,
      "step": 1589
    },
    {
      "epoch": 0.30671296296296297,
      "grad_norm": 1.026572585105896,
      "learning_rate": 0.0005,
      "loss": 2.787769079208374,
      "step": 1590
    },
    {
      "epoch": 0.30690586419753085,
      "grad_norm": 1.0788325071334839,
      "learning_rate": 0.0005,
      "loss": 3.157261610031128,
      "step": 1591
    },
    {
      "epoch": 0.30709876543209874,
      "grad_norm": 0.9208263754844666,
      "learning_rate": 0.0005,
      "loss": 3.017590045928955,
      "step": 1592
    },
    {
      "epoch": 0.3072916666666667,
      "grad_norm": 1.3611092567443848,
      "learning_rate": 0.0005,
      "loss": 2.943861246109009,
      "step": 1593
    },
    {
      "epoch": 0.3074845679012346,
      "grad_norm": 1.0296775102615356,
      "learning_rate": 0.0005,
      "loss": 3.10107159614563,
      "step": 1594
    },
    {
      "epoch": 0.30767746913580246,
      "grad_norm": 1.102824091911316,
      "learning_rate": 0.0005,
      "loss": 3.0611202716827393,
      "step": 1595
    },
    {
      "epoch": 0.30787037037037035,
      "grad_norm": 1.431990146636963,
      "learning_rate": 0.0005,
      "loss": 3.158930778503418,
      "step": 1596
    },
    {
      "epoch": 0.3080632716049383,
      "grad_norm": 1.0262755155563354,
      "learning_rate": 0.0005,
      "loss": 3.003206253051758,
      "step": 1597
    },
    {
      "epoch": 0.3082561728395062,
      "grad_norm": 1.589639663696289,
      "learning_rate": 0.0005,
      "loss": 3.0112380981445312,
      "step": 1598
    },
    {
      "epoch": 0.30844907407407407,
      "grad_norm": 1.1525083780288696,
      "learning_rate": 0.0005,
      "loss": 2.7567646503448486,
      "step": 1599
    },
    {
      "epoch": 0.30864197530864196,
      "grad_norm": 1.1607173681259155,
      "learning_rate": 0.0005,
      "loss": 3.4402451515197754,
      "step": 1600
    },
    {
      "epoch": 0.3088348765432099,
      "grad_norm": 1.2707182168960571,
      "learning_rate": 0.0005,
      "loss": 3.157036542892456,
      "step": 1601
    },
    {
      "epoch": 0.3090277777777778,
      "grad_norm": 0.9673983454704285,
      "learning_rate": 0.0005,
      "loss": 2.9517903327941895,
      "step": 1602
    },
    {
      "epoch": 0.3092206790123457,
      "grad_norm": 0.8559006452560425,
      "learning_rate": 0.0005,
      "loss": 3.0044469833374023,
      "step": 1603
    },
    {
      "epoch": 0.30941358024691357,
      "grad_norm": 1.157988429069519,
      "learning_rate": 0.0005,
      "loss": 2.846998691558838,
      "step": 1604
    },
    {
      "epoch": 0.30960648148148145,
      "grad_norm": 0.926384449005127,
      "learning_rate": 0.0005,
      "loss": 2.920346260070801,
      "step": 1605
    },
    {
      "epoch": 0.3097993827160494,
      "grad_norm": 0.9756860733032227,
      "learning_rate": 0.0005,
      "loss": 3.1979758739471436,
      "step": 1606
    },
    {
      "epoch": 0.3099922839506173,
      "grad_norm": 1.2790898084640503,
      "learning_rate": 0.0005,
      "loss": 2.971545457839966,
      "step": 1607
    },
    {
      "epoch": 0.3101851851851852,
      "grad_norm": 0.9731301665306091,
      "learning_rate": 0.0005,
      "loss": 3.1285831928253174,
      "step": 1608
    },
    {
      "epoch": 0.31037808641975306,
      "grad_norm": 1.7664709091186523,
      "learning_rate": 0.0005,
      "loss": 3.074326276779175,
      "step": 1609
    },
    {
      "epoch": 0.310570987654321,
      "grad_norm": 1.167454719543457,
      "learning_rate": 0.0005,
      "loss": 2.935713768005371,
      "step": 1610
    },
    {
      "epoch": 0.3107638888888889,
      "grad_norm": 0.9711547493934631,
      "learning_rate": 0.0005,
      "loss": 2.7732787132263184,
      "step": 1611
    },
    {
      "epoch": 0.3109567901234568,
      "grad_norm": 4.874449253082275,
      "learning_rate": 0.0005,
      "loss": 3.035142660140991,
      "step": 1612
    },
    {
      "epoch": 0.31114969135802467,
      "grad_norm": 1.2665259838104248,
      "learning_rate": 0.0005,
      "loss": 2.899888753890991,
      "step": 1613
    },
    {
      "epoch": 0.3113425925925926,
      "grad_norm": 1.1100025177001953,
      "learning_rate": 0.0005,
      "loss": 2.9239795207977295,
      "step": 1614
    },
    {
      "epoch": 0.3115354938271605,
      "grad_norm": 1.1748125553131104,
      "learning_rate": 0.0005,
      "loss": 3.3437814712524414,
      "step": 1615
    },
    {
      "epoch": 0.3117283950617284,
      "grad_norm": 1.1082388162612915,
      "learning_rate": 0.0005,
      "loss": 2.854513168334961,
      "step": 1616
    },
    {
      "epoch": 0.3119212962962963,
      "grad_norm": 2.136319875717163,
      "learning_rate": 0.0005,
      "loss": 3.095797538757324,
      "step": 1617
    },
    {
      "epoch": 0.3121141975308642,
      "grad_norm": 1.5573621988296509,
      "learning_rate": 0.0005,
      "loss": 2.98348331451416,
      "step": 1618
    },
    {
      "epoch": 0.3123070987654321,
      "grad_norm": 1.249588966369629,
      "learning_rate": 0.0005,
      "loss": 2.7998905181884766,
      "step": 1619
    },
    {
      "epoch": 0.3125,
      "grad_norm": 1.1542643308639526,
      "learning_rate": 0.0005,
      "loss": 2.7980575561523438,
      "step": 1620
    },
    {
      "epoch": 0.3126929012345679,
      "grad_norm": 1.3635066747665405,
      "learning_rate": 0.0005,
      "loss": 3.329413890838623,
      "step": 1621
    },
    {
      "epoch": 0.3128858024691358,
      "grad_norm": 1.1058175563812256,
      "learning_rate": 0.0005,
      "loss": 2.9618377685546875,
      "step": 1622
    },
    {
      "epoch": 0.3130787037037037,
      "grad_norm": 1.1057512760162354,
      "learning_rate": 0.0005,
      "loss": 2.9910407066345215,
      "step": 1623
    },
    {
      "epoch": 0.3132716049382716,
      "grad_norm": 1.3396172523498535,
      "learning_rate": 0.0005,
      "loss": 2.835635185241699,
      "step": 1624
    },
    {
      "epoch": 0.3134645061728395,
      "grad_norm": 1.2156881093978882,
      "learning_rate": 0.0005,
      "loss": 2.7836532592773438,
      "step": 1625
    },
    {
      "epoch": 0.3136574074074074,
      "grad_norm": 1.1299657821655273,
      "learning_rate": 0.0005,
      "loss": 3.3573970794677734,
      "step": 1626
    },
    {
      "epoch": 0.31385030864197533,
      "grad_norm": 1.2500123977661133,
      "learning_rate": 0.0005,
      "loss": 2.7363946437835693,
      "step": 1627
    },
    {
      "epoch": 0.3140432098765432,
      "grad_norm": 1.0957030057907104,
      "learning_rate": 0.0005,
      "loss": 2.8074216842651367,
      "step": 1628
    },
    {
      "epoch": 0.3142361111111111,
      "grad_norm": 1.120901107788086,
      "learning_rate": 0.0005,
      "loss": 2.946542501449585,
      "step": 1629
    },
    {
      "epoch": 0.314429012345679,
      "grad_norm": 1.0509639978408813,
      "learning_rate": 0.0005,
      "loss": 2.8739023208618164,
      "step": 1630
    },
    {
      "epoch": 0.31462191358024694,
      "grad_norm": 1.0083640813827515,
      "learning_rate": 0.0005,
      "loss": 3.3079872131347656,
      "step": 1631
    },
    {
      "epoch": 0.3148148148148148,
      "grad_norm": 1.2283284664154053,
      "learning_rate": 0.0005,
      "loss": 3.1876721382141113,
      "step": 1632
    },
    {
      "epoch": 0.3150077160493827,
      "grad_norm": 1.5427378416061401,
      "learning_rate": 0.0005,
      "loss": 2.80684757232666,
      "step": 1633
    },
    {
      "epoch": 0.3152006172839506,
      "grad_norm": 1.1392532587051392,
      "learning_rate": 0.0005,
      "loss": 2.866835117340088,
      "step": 1634
    },
    {
      "epoch": 0.31539351851851855,
      "grad_norm": 1.0534751415252686,
      "learning_rate": 0.0005,
      "loss": 2.9264204502105713,
      "step": 1635
    },
    {
      "epoch": 0.31558641975308643,
      "grad_norm": 1.1675578355789185,
      "learning_rate": 0.0005,
      "loss": 2.9710521697998047,
      "step": 1636
    },
    {
      "epoch": 0.3157793209876543,
      "grad_norm": 1.068752646446228,
      "learning_rate": 0.0005,
      "loss": 3.1587300300598145,
      "step": 1637
    },
    {
      "epoch": 0.3159722222222222,
      "grad_norm": 1.0235779285430908,
      "learning_rate": 0.0005,
      "loss": 3.1461663246154785,
      "step": 1638
    },
    {
      "epoch": 0.3161651234567901,
      "grad_norm": 1.026876449584961,
      "learning_rate": 0.0005,
      "loss": 3.170508623123169,
      "step": 1639
    },
    {
      "epoch": 0.31635802469135804,
      "grad_norm": 1.095851182937622,
      "learning_rate": 0.0005,
      "loss": 3.1066818237304688,
      "step": 1640
    },
    {
      "epoch": 0.31655092592592593,
      "grad_norm": 1.1235246658325195,
      "learning_rate": 0.0005,
      "loss": 2.9413154125213623,
      "step": 1641
    },
    {
      "epoch": 0.3167438271604938,
      "grad_norm": 1.0722821950912476,
      "learning_rate": 0.0005,
      "loss": 3.320087432861328,
      "step": 1642
    },
    {
      "epoch": 0.3169367283950617,
      "grad_norm": 1.027536153793335,
      "learning_rate": 0.0005,
      "loss": 3.304023265838623,
      "step": 1643
    },
    {
      "epoch": 0.31712962962962965,
      "grad_norm": 1.052401065826416,
      "learning_rate": 0.0005,
      "loss": 2.9436397552490234,
      "step": 1644
    },
    {
      "epoch": 0.31732253086419754,
      "grad_norm": 0.849594235420227,
      "learning_rate": 0.0005,
      "loss": 3.020205497741699,
      "step": 1645
    },
    {
      "epoch": 0.3175154320987654,
      "grad_norm": 0.9554029107093811,
      "learning_rate": 0.0005,
      "loss": 3.1866276264190674,
      "step": 1646
    },
    {
      "epoch": 0.3177083333333333,
      "grad_norm": 0.9977423548698425,
      "learning_rate": 0.0005,
      "loss": 2.8997647762298584,
      "step": 1647
    },
    {
      "epoch": 0.31790123456790126,
      "grad_norm": 1.2209341526031494,
      "learning_rate": 0.0005,
      "loss": 2.7802696228027344,
      "step": 1648
    },
    {
      "epoch": 0.31809413580246915,
      "grad_norm": 1.3176437616348267,
      "learning_rate": 0.0005,
      "loss": 3.141315460205078,
      "step": 1649
    },
    {
      "epoch": 0.31828703703703703,
      "grad_norm": 1.416629433631897,
      "learning_rate": 0.0005,
      "loss": 2.6213648319244385,
      "step": 1650
    },
    {
      "epoch": 0.3184799382716049,
      "grad_norm": 1.308142066001892,
      "learning_rate": 0.0005,
      "loss": 2.885145902633667,
      "step": 1651
    },
    {
      "epoch": 0.31867283950617287,
      "grad_norm": 1.1654293537139893,
      "learning_rate": 0.0005,
      "loss": 3.297149419784546,
      "step": 1652
    },
    {
      "epoch": 0.31886574074074076,
      "grad_norm": 1.300447702407837,
      "learning_rate": 0.0005,
      "loss": 2.8346924781799316,
      "step": 1653
    },
    {
      "epoch": 0.31905864197530864,
      "grad_norm": 0.9426700472831726,
      "learning_rate": 0.0005,
      "loss": 3.176537036895752,
      "step": 1654
    },
    {
      "epoch": 0.31925154320987653,
      "grad_norm": 1.3784924745559692,
      "learning_rate": 0.0005,
      "loss": 2.655485153198242,
      "step": 1655
    },
    {
      "epoch": 0.3194444444444444,
      "grad_norm": 1.1149888038635254,
      "learning_rate": 0.0005,
      "loss": 2.7525501251220703,
      "step": 1656
    },
    {
      "epoch": 0.31963734567901236,
      "grad_norm": 1.176612138748169,
      "learning_rate": 0.0005,
      "loss": 2.945780038833618,
      "step": 1657
    },
    {
      "epoch": 0.31983024691358025,
      "grad_norm": 1.1622158288955688,
      "learning_rate": 0.0005,
      "loss": 2.9338841438293457,
      "step": 1658
    },
    {
      "epoch": 0.32002314814814814,
      "grad_norm": 1.099326491355896,
      "learning_rate": 0.0005,
      "loss": 3.540445327758789,
      "step": 1659
    },
    {
      "epoch": 0.32021604938271603,
      "grad_norm": 1.7168285846710205,
      "learning_rate": 0.0005,
      "loss": 2.8738646507263184,
      "step": 1660
    },
    {
      "epoch": 0.32040895061728397,
      "grad_norm": 1.4835346937179565,
      "learning_rate": 0.0005,
      "loss": 2.7843542098999023,
      "step": 1661
    },
    {
      "epoch": 0.32060185185185186,
      "grad_norm": 1.7009891271591187,
      "learning_rate": 0.0005,
      "loss": 3.0283970832824707,
      "step": 1662
    },
    {
      "epoch": 0.32079475308641975,
      "grad_norm": 1.1003891229629517,
      "learning_rate": 0.0005,
      "loss": 2.904566764831543,
      "step": 1663
    },
    {
      "epoch": 0.32098765432098764,
      "grad_norm": 1.1381734609603882,
      "learning_rate": 0.0005,
      "loss": 2.801577091217041,
      "step": 1664
    },
    {
      "epoch": 0.3211805555555556,
      "grad_norm": 0.9565869569778442,
      "learning_rate": 0.0005,
      "loss": 2.965212821960449,
      "step": 1665
    },
    {
      "epoch": 0.32137345679012347,
      "grad_norm": 1.2768019437789917,
      "learning_rate": 0.0005,
      "loss": 3.0585145950317383,
      "step": 1666
    },
    {
      "epoch": 0.32156635802469136,
      "grad_norm": 1.2815892696380615,
      "learning_rate": 0.0005,
      "loss": 3.029297351837158,
      "step": 1667
    },
    {
      "epoch": 0.32175925925925924,
      "grad_norm": 1.0301200151443481,
      "learning_rate": 0.0005,
      "loss": 3.0879335403442383,
      "step": 1668
    },
    {
      "epoch": 0.32195216049382713,
      "grad_norm": 1.2136260271072388,
      "learning_rate": 0.0005,
      "loss": 2.8012304306030273,
      "step": 1669
    },
    {
      "epoch": 0.3221450617283951,
      "grad_norm": 1.0720369815826416,
      "learning_rate": 0.0005,
      "loss": 3.1398563385009766,
      "step": 1670
    },
    {
      "epoch": 0.32233796296296297,
      "grad_norm": 1.3205171823501587,
      "learning_rate": 0.0005,
      "loss": 3.0114521980285645,
      "step": 1671
    },
    {
      "epoch": 0.32253086419753085,
      "grad_norm": 1.188901424407959,
      "learning_rate": 0.0005,
      "loss": 2.9309654235839844,
      "step": 1672
    },
    {
      "epoch": 0.32272376543209874,
      "grad_norm": 1.145609736442566,
      "learning_rate": 0.0005,
      "loss": 3.118642807006836,
      "step": 1673
    },
    {
      "epoch": 0.3229166666666667,
      "grad_norm": 0.8927665948867798,
      "learning_rate": 0.0005,
      "loss": 3.0635874271392822,
      "step": 1674
    },
    {
      "epoch": 0.3231095679012346,
      "grad_norm": 1.3775444030761719,
      "learning_rate": 0.0005,
      "loss": 3.2481369972229004,
      "step": 1675
    },
    {
      "epoch": 0.32330246913580246,
      "grad_norm": 1.1199214458465576,
      "learning_rate": 0.0005,
      "loss": 3.1083011627197266,
      "step": 1676
    },
    {
      "epoch": 0.32349537037037035,
      "grad_norm": 0.9801725149154663,
      "learning_rate": 0.0005,
      "loss": 3.0737500190734863,
      "step": 1677
    },
    {
      "epoch": 0.3236882716049383,
      "grad_norm": 1.331224799156189,
      "learning_rate": 0.0005,
      "loss": 2.9440505504608154,
      "step": 1678
    },
    {
      "epoch": 0.3238811728395062,
      "grad_norm": 1.0691934823989868,
      "learning_rate": 0.0005,
      "loss": 2.81404972076416,
      "step": 1679
    },
    {
      "epoch": 0.32407407407407407,
      "grad_norm": 0.9698763489723206,
      "learning_rate": 0.0005,
      "loss": 3.0872974395751953,
      "step": 1680
    },
    {
      "epoch": 0.32426697530864196,
      "grad_norm": 1.1651153564453125,
      "learning_rate": 0.0005,
      "loss": 3.2332446575164795,
      "step": 1681
    },
    {
      "epoch": 0.3244598765432099,
      "grad_norm": 1.0909128189086914,
      "learning_rate": 0.0005,
      "loss": 3.2031970024108887,
      "step": 1682
    },
    {
      "epoch": 0.3246527777777778,
      "grad_norm": 1.070007562637329,
      "learning_rate": 0.0005,
      "loss": 2.9510326385498047,
      "step": 1683
    },
    {
      "epoch": 0.3248456790123457,
      "grad_norm": 1.067481517791748,
      "learning_rate": 0.0005,
      "loss": 2.896819591522217,
      "step": 1684
    },
    {
      "epoch": 0.32503858024691357,
      "grad_norm": 0.9923682808876038,
      "learning_rate": 0.0005,
      "loss": 3.0656633377075195,
      "step": 1685
    },
    {
      "epoch": 0.32523148148148145,
      "grad_norm": 1.0897170305252075,
      "learning_rate": 0.0005,
      "loss": 2.714108943939209,
      "step": 1686
    },
    {
      "epoch": 0.3254243827160494,
      "grad_norm": 1.0828784704208374,
      "learning_rate": 0.0005,
      "loss": 3.0779409408569336,
      "step": 1687
    },
    {
      "epoch": 0.3256172839506173,
      "grad_norm": 0.8401271104812622,
      "learning_rate": 0.0005,
      "loss": 2.7838399410247803,
      "step": 1688
    },
    {
      "epoch": 0.3258101851851852,
      "grad_norm": 1.1480897665023804,
      "learning_rate": 0.0005,
      "loss": 2.746670961380005,
      "step": 1689
    },
    {
      "epoch": 0.32600308641975306,
      "grad_norm": 0.986070990562439,
      "learning_rate": 0.0005,
      "loss": 3.0781798362731934,
      "step": 1690
    },
    {
      "epoch": 0.326195987654321,
      "grad_norm": 0.9605697393417358,
      "learning_rate": 0.0005,
      "loss": 2.913037061691284,
      "step": 1691
    },
    {
      "epoch": 0.3263888888888889,
      "grad_norm": 1.1642380952835083,
      "learning_rate": 0.0005,
      "loss": 3.0351791381835938,
      "step": 1692
    },
    {
      "epoch": 0.3265817901234568,
      "grad_norm": 1.1087229251861572,
      "learning_rate": 0.0005,
      "loss": 2.814122200012207,
      "step": 1693
    },
    {
      "epoch": 0.32677469135802467,
      "grad_norm": 1.3995729684829712,
      "learning_rate": 0.0005,
      "loss": 3.0467398166656494,
      "step": 1694
    },
    {
      "epoch": 0.3269675925925926,
      "grad_norm": 1.3072363138198853,
      "learning_rate": 0.0005,
      "loss": 2.8582842350006104,
      "step": 1695
    },
    {
      "epoch": 0.3271604938271605,
      "grad_norm": 1.0853782892227173,
      "learning_rate": 0.0005,
      "loss": 3.1588730812072754,
      "step": 1696
    },
    {
      "epoch": 0.3273533950617284,
      "grad_norm": 1.0398790836334229,
      "learning_rate": 0.0005,
      "loss": 2.954784393310547,
      "step": 1697
    },
    {
      "epoch": 0.3275462962962963,
      "grad_norm": 1.190071940422058,
      "learning_rate": 0.0005,
      "loss": 2.8493094444274902,
      "step": 1698
    },
    {
      "epoch": 0.3277391975308642,
      "grad_norm": 1.17421555519104,
      "learning_rate": 0.0005,
      "loss": 2.886227607727051,
      "step": 1699
    },
    {
      "epoch": 0.3279320987654321,
      "grad_norm": 1.1805647611618042,
      "learning_rate": 0.0005,
      "loss": 2.7171454429626465,
      "step": 1700
    },
    {
      "epoch": 0.328125,
      "grad_norm": 1.087217092514038,
      "learning_rate": 0.0005,
      "loss": 3.215481758117676,
      "step": 1701
    },
    {
      "epoch": 0.3283179012345679,
      "grad_norm": 0.8345917463302612,
      "learning_rate": 0.0005,
      "loss": 3.1549813747406006,
      "step": 1702
    },
    {
      "epoch": 0.3285108024691358,
      "grad_norm": 0.9552971124649048,
      "learning_rate": 0.0005,
      "loss": 3.020746946334839,
      "step": 1703
    },
    {
      "epoch": 0.3287037037037037,
      "grad_norm": 1.0150967836380005,
      "learning_rate": 0.0005,
      "loss": 2.863053798675537,
      "step": 1704
    },
    {
      "epoch": 0.3288966049382716,
      "grad_norm": 0.8384143710136414,
      "learning_rate": 0.0005,
      "loss": 3.25846791267395,
      "step": 1705
    },
    {
      "epoch": 0.3290895061728395,
      "grad_norm": 1.04835045337677,
      "learning_rate": 0.0005,
      "loss": 2.8948068618774414,
      "step": 1706
    },
    {
      "epoch": 0.3292824074074074,
      "grad_norm": 1.0046195983886719,
      "learning_rate": 0.0005,
      "loss": 2.8691859245300293,
      "step": 1707
    },
    {
      "epoch": 0.32947530864197533,
      "grad_norm": 1.3428760766983032,
      "learning_rate": 0.0005,
      "loss": 2.9840540885925293,
      "step": 1708
    },
    {
      "epoch": 0.3296682098765432,
      "grad_norm": 1.8620423078536987,
      "learning_rate": 0.0005,
      "loss": 3.0064568519592285,
      "step": 1709
    },
    {
      "epoch": 0.3298611111111111,
      "grad_norm": 1.2227579355239868,
      "learning_rate": 0.0005,
      "loss": 2.7072863578796387,
      "step": 1710
    },
    {
      "epoch": 0.330054012345679,
      "grad_norm": 1.4436177015304565,
      "learning_rate": 0.0005,
      "loss": 2.971665620803833,
      "step": 1711
    },
    {
      "epoch": 0.33024691358024694,
      "grad_norm": 0.9858396649360657,
      "learning_rate": 0.0005,
      "loss": 3.123358726501465,
      "step": 1712
    },
    {
      "epoch": 0.3304398148148148,
      "grad_norm": 1.029834270477295,
      "learning_rate": 0.0005,
      "loss": 3.1253154277801514,
      "step": 1713
    },
    {
      "epoch": 0.3306327160493827,
      "grad_norm": 0.9656727313995361,
      "learning_rate": 0.0005,
      "loss": 3.1562137603759766,
      "step": 1714
    },
    {
      "epoch": 0.3308256172839506,
      "grad_norm": 1.2183425426483154,
      "learning_rate": 0.0005,
      "loss": 2.817516326904297,
      "step": 1715
    },
    {
      "epoch": 0.33101851851851855,
      "grad_norm": 1.2711708545684814,
      "learning_rate": 0.0005,
      "loss": 3.119426727294922,
      "step": 1716
    },
    {
      "epoch": 0.33121141975308643,
      "grad_norm": 1.3473583459854126,
      "learning_rate": 0.0005,
      "loss": 3.0274574756622314,
      "step": 1717
    },
    {
      "epoch": 0.3314043209876543,
      "grad_norm": 1.0160762071609497,
      "learning_rate": 0.0005,
      "loss": 3.0505990982055664,
      "step": 1718
    },
    {
      "epoch": 0.3315972222222222,
      "grad_norm": 1.0260426998138428,
      "learning_rate": 0.0005,
      "loss": 2.662257432937622,
      "step": 1719
    },
    {
      "epoch": 0.3317901234567901,
      "grad_norm": 1.164322018623352,
      "learning_rate": 0.0005,
      "loss": 2.8466317653656006,
      "step": 1720
    },
    {
      "epoch": 0.33198302469135804,
      "grad_norm": 1.094163417816162,
      "learning_rate": 0.0005,
      "loss": 2.6906981468200684,
      "step": 1721
    },
    {
      "epoch": 0.33217592592592593,
      "grad_norm": 0.9895902276039124,
      "learning_rate": 0.0005,
      "loss": 2.9386672973632812,
      "step": 1722
    },
    {
      "epoch": 0.3323688271604938,
      "grad_norm": 1.0302191972732544,
      "learning_rate": 0.0005,
      "loss": 2.7134201526641846,
      "step": 1723
    },
    {
      "epoch": 0.3325617283950617,
      "grad_norm": 1.2283166646957397,
      "learning_rate": 0.0005,
      "loss": 3.121593952178955,
      "step": 1724
    },
    {
      "epoch": 0.33275462962962965,
      "grad_norm": 5.892521381378174,
      "learning_rate": 0.0005,
      "loss": 2.978428363800049,
      "step": 1725
    },
    {
      "epoch": 0.33294753086419754,
      "grad_norm": 1.287642002105713,
      "learning_rate": 0.0005,
      "loss": 2.8558499813079834,
      "step": 1726
    },
    {
      "epoch": 0.3331404320987654,
      "grad_norm": 1.5772714614868164,
      "learning_rate": 0.0005,
      "loss": 3.0555996894836426,
      "step": 1727
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 1.124414324760437,
      "learning_rate": 0.0005,
      "loss": 2.8363242149353027,
      "step": 1728
    },
    {
      "epoch": 0.33352623456790126,
      "grad_norm": 1.2025573253631592,
      "learning_rate": 0.0005,
      "loss": 3.1870720386505127,
      "step": 1729
    },
    {
      "epoch": 0.33371913580246915,
      "grad_norm": 1.1858551502227783,
      "learning_rate": 0.0005,
      "loss": 3.1253230571746826,
      "step": 1730
    },
    {
      "epoch": 0.33391203703703703,
      "grad_norm": 0.9927303791046143,
      "learning_rate": 0.0005,
      "loss": 2.982785224914551,
      "step": 1731
    },
    {
      "epoch": 0.3341049382716049,
      "grad_norm": 1.5390784740447998,
      "learning_rate": 0.0005,
      "loss": 2.8314290046691895,
      "step": 1732
    },
    {
      "epoch": 0.33429783950617287,
      "grad_norm": 8.196191787719727,
      "learning_rate": 0.0005,
      "loss": 3.002201557159424,
      "step": 1733
    },
    {
      "epoch": 0.33449074074074076,
      "grad_norm": 1.074644923210144,
      "learning_rate": 0.0005,
      "loss": 3.0705204010009766,
      "step": 1734
    },
    {
      "epoch": 0.33468364197530864,
      "grad_norm": 0.9823842644691467,
      "learning_rate": 0.0005,
      "loss": 3.0203018188476562,
      "step": 1735
    },
    {
      "epoch": 0.33487654320987653,
      "grad_norm": 1.11582612991333,
      "learning_rate": 0.0005,
      "loss": 3.15138578414917,
      "step": 1736
    },
    {
      "epoch": 0.3350694444444444,
      "grad_norm": 1.0087093114852905,
      "learning_rate": 0.0005,
      "loss": 2.9125819206237793,
      "step": 1737
    },
    {
      "epoch": 0.33526234567901236,
      "grad_norm": 0.9163546562194824,
      "learning_rate": 0.0005,
      "loss": 2.618929862976074,
      "step": 1738
    },
    {
      "epoch": 0.33545524691358025,
      "grad_norm": 1.4564143419265747,
      "learning_rate": 0.0005,
      "loss": 2.952882766723633,
      "step": 1739
    },
    {
      "epoch": 0.33564814814814814,
      "grad_norm": 1.3389475345611572,
      "learning_rate": 0.0005,
      "loss": 3.200043201446533,
      "step": 1740
    },
    {
      "epoch": 0.33584104938271603,
      "grad_norm": 1.7645174264907837,
      "learning_rate": 0.0005,
      "loss": 2.930140972137451,
      "step": 1741
    },
    {
      "epoch": 0.33603395061728397,
      "grad_norm": 1.0191394090652466,
      "learning_rate": 0.0005,
      "loss": 3.0839624404907227,
      "step": 1742
    },
    {
      "epoch": 0.33622685185185186,
      "grad_norm": 1.041396975517273,
      "learning_rate": 0.0005,
      "loss": 2.9474899768829346,
      "step": 1743
    },
    {
      "epoch": 0.33641975308641975,
      "grad_norm": 1.1022688150405884,
      "learning_rate": 0.0005,
      "loss": 2.8739590644836426,
      "step": 1744
    },
    {
      "epoch": 0.33661265432098764,
      "grad_norm": 0.9928793907165527,
      "learning_rate": 0.0005,
      "loss": 2.6416149139404297,
      "step": 1745
    },
    {
      "epoch": 0.3368055555555556,
      "grad_norm": 0.9685197472572327,
      "learning_rate": 0.0005,
      "loss": 2.8219404220581055,
      "step": 1746
    },
    {
      "epoch": 0.33699845679012347,
      "grad_norm": 1.3046834468841553,
      "learning_rate": 0.0005,
      "loss": 2.976067066192627,
      "step": 1747
    },
    {
      "epoch": 0.33719135802469136,
      "grad_norm": 1.0700308084487915,
      "learning_rate": 0.0005,
      "loss": 3.093863010406494,
      "step": 1748
    },
    {
      "epoch": 0.33738425925925924,
      "grad_norm": 1.0669037103652954,
      "learning_rate": 0.0005,
      "loss": 3.1921117305755615,
      "step": 1749
    },
    {
      "epoch": 0.33757716049382713,
      "grad_norm": 1.147929072380066,
      "learning_rate": 0.0005,
      "loss": 2.644637107849121,
      "step": 1750
    },
    {
      "epoch": 0.3377700617283951,
      "grad_norm": 1.0565563440322876,
      "learning_rate": 0.0005,
      "loss": 3.2551045417785645,
      "step": 1751
    },
    {
      "epoch": 0.33796296296296297,
      "grad_norm": 1.329529881477356,
      "learning_rate": 0.0005,
      "loss": 2.720618724822998,
      "step": 1752
    },
    {
      "epoch": 0.33815586419753085,
      "grad_norm": 0.9451022148132324,
      "learning_rate": 0.0005,
      "loss": 2.8062453269958496,
      "step": 1753
    },
    {
      "epoch": 0.33834876543209874,
      "grad_norm": 1.0331851243972778,
      "learning_rate": 0.0005,
      "loss": 2.784970998764038,
      "step": 1754
    },
    {
      "epoch": 0.3385416666666667,
      "grad_norm": 1.2282851934432983,
      "learning_rate": 0.0005,
      "loss": 3.020956516265869,
      "step": 1755
    },
    {
      "epoch": 0.3387345679012346,
      "grad_norm": 1.2625080347061157,
      "learning_rate": 0.0005,
      "loss": 3.0916008949279785,
      "step": 1756
    },
    {
      "epoch": 0.33892746913580246,
      "grad_norm": 1.272322177886963,
      "learning_rate": 0.0005,
      "loss": 3.0260658264160156,
      "step": 1757
    },
    {
      "epoch": 0.33912037037037035,
      "grad_norm": 1.263224720954895,
      "learning_rate": 0.0005,
      "loss": 2.824657440185547,
      "step": 1758
    },
    {
      "epoch": 0.3393132716049383,
      "grad_norm": 1.1612846851348877,
      "learning_rate": 0.0005,
      "loss": 3.071765661239624,
      "step": 1759
    },
    {
      "epoch": 0.3395061728395062,
      "grad_norm": 1.4451508522033691,
      "learning_rate": 0.0005,
      "loss": 2.8787105083465576,
      "step": 1760
    },
    {
      "epoch": 0.33969907407407407,
      "grad_norm": 1.0832542181015015,
      "learning_rate": 0.0005,
      "loss": 3.0135421752929688,
      "step": 1761
    },
    {
      "epoch": 0.33989197530864196,
      "grad_norm": 0.9900439977645874,
      "learning_rate": 0.0005,
      "loss": 3.097142219543457,
      "step": 1762
    },
    {
      "epoch": 0.3400848765432099,
      "grad_norm": 1.0133986473083496,
      "learning_rate": 0.0005,
      "loss": 2.7488667964935303,
      "step": 1763
    },
    {
      "epoch": 0.3402777777777778,
      "grad_norm": 0.9893466830253601,
      "learning_rate": 0.0005,
      "loss": 2.636204242706299,
      "step": 1764
    },
    {
      "epoch": 0.3404706790123457,
      "grad_norm": 1.1145970821380615,
      "learning_rate": 0.0005,
      "loss": 2.8363656997680664,
      "step": 1765
    },
    {
      "epoch": 0.34066358024691357,
      "grad_norm": 1.1496679782867432,
      "learning_rate": 0.0005,
      "loss": 2.8908352851867676,
      "step": 1766
    },
    {
      "epoch": 0.34085648148148145,
      "grad_norm": 1.091315746307373,
      "learning_rate": 0.0005,
      "loss": 2.8876914978027344,
      "step": 1767
    },
    {
      "epoch": 0.3410493827160494,
      "grad_norm": 0.8847644329071045,
      "learning_rate": 0.0005,
      "loss": 2.890937566757202,
      "step": 1768
    },
    {
      "epoch": 0.3412422839506173,
      "grad_norm": 0.9090626239776611,
      "learning_rate": 0.0005,
      "loss": 2.913113832473755,
      "step": 1769
    },
    {
      "epoch": 0.3414351851851852,
      "grad_norm": 1.1121690273284912,
      "learning_rate": 0.0005,
      "loss": 3.7265560626983643,
      "step": 1770
    },
    {
      "epoch": 0.34162808641975306,
      "grad_norm": 1.3898693323135376,
      "learning_rate": 0.0005,
      "loss": 3.0408666133880615,
      "step": 1771
    },
    {
      "epoch": 0.341820987654321,
      "grad_norm": 1.0450748205184937,
      "learning_rate": 0.0005,
      "loss": 3.0672662258148193,
      "step": 1772
    },
    {
      "epoch": 0.3420138888888889,
      "grad_norm": 1.3404855728149414,
      "learning_rate": 0.0005,
      "loss": 3.133890390396118,
      "step": 1773
    },
    {
      "epoch": 0.3422067901234568,
      "grad_norm": 1.0372672080993652,
      "learning_rate": 0.0005,
      "loss": 3.173314332962036,
      "step": 1774
    },
    {
      "epoch": 0.34239969135802467,
      "grad_norm": 1.016846776008606,
      "learning_rate": 0.0005,
      "loss": 2.8866872787475586,
      "step": 1775
    },
    {
      "epoch": 0.3425925925925926,
      "grad_norm": 1.1108520030975342,
      "learning_rate": 0.0005,
      "loss": 2.9136691093444824,
      "step": 1776
    },
    {
      "epoch": 0.3427854938271605,
      "grad_norm": 1.1731384992599487,
      "learning_rate": 0.0005,
      "loss": 2.84983491897583,
      "step": 1777
    },
    {
      "epoch": 0.3429783950617284,
      "grad_norm": 1.0757440328598022,
      "learning_rate": 0.0005,
      "loss": 2.9199531078338623,
      "step": 1778
    },
    {
      "epoch": 0.3431712962962963,
      "grad_norm": 1.039750099182129,
      "learning_rate": 0.0005,
      "loss": 3.169241428375244,
      "step": 1779
    },
    {
      "epoch": 0.3433641975308642,
      "grad_norm": 0.9726346731185913,
      "learning_rate": 0.0005,
      "loss": 2.7077462673187256,
      "step": 1780
    },
    {
      "epoch": 0.3435570987654321,
      "grad_norm": 0.9721254110336304,
      "learning_rate": 0.0005,
      "loss": 2.9344019889831543,
      "step": 1781
    },
    {
      "epoch": 0.34375,
      "grad_norm": 0.9104084968566895,
      "learning_rate": 0.0005,
      "loss": 2.941011905670166,
      "step": 1782
    },
    {
      "epoch": 0.3439429012345679,
      "grad_norm": 0.8184657692909241,
      "learning_rate": 0.0005,
      "loss": 3.247918128967285,
      "step": 1783
    },
    {
      "epoch": 0.3441358024691358,
      "grad_norm": 1.0904788970947266,
      "learning_rate": 0.0005,
      "loss": 2.843555450439453,
      "step": 1784
    },
    {
      "epoch": 0.3443287037037037,
      "grad_norm": 0.8778148889541626,
      "learning_rate": 0.0005,
      "loss": 3.0134432315826416,
      "step": 1785
    },
    {
      "epoch": 0.3445216049382716,
      "grad_norm": 1.1878602504730225,
      "learning_rate": 0.0005,
      "loss": 2.8189988136291504,
      "step": 1786
    },
    {
      "epoch": 0.3447145061728395,
      "grad_norm": 0.9325968623161316,
      "learning_rate": 0.0005,
      "loss": 3.1753382682800293,
      "step": 1787
    },
    {
      "epoch": 0.3449074074074074,
      "grad_norm": 1.1519818305969238,
      "learning_rate": 0.0005,
      "loss": 2.7887558937072754,
      "step": 1788
    },
    {
      "epoch": 0.34510030864197533,
      "grad_norm": 1.0089974403381348,
      "learning_rate": 0.0005,
      "loss": 2.920506000518799,
      "step": 1789
    },
    {
      "epoch": 0.3452932098765432,
      "grad_norm": 1.1791441440582275,
      "learning_rate": 0.0005,
      "loss": 3.170940399169922,
      "step": 1790
    },
    {
      "epoch": 0.3454861111111111,
      "grad_norm": 1.1173919439315796,
      "learning_rate": 0.0005,
      "loss": 2.797088623046875,
      "step": 1791
    },
    {
      "epoch": 0.345679012345679,
      "grad_norm": 0.9913433790206909,
      "learning_rate": 0.0005,
      "loss": 2.974214792251587,
      "step": 1792
    },
    {
      "epoch": 0.34587191358024694,
      "grad_norm": 1.105783224105835,
      "learning_rate": 0.0005,
      "loss": 2.445699691772461,
      "step": 1793
    },
    {
      "epoch": 0.3460648148148148,
      "grad_norm": 3.2754528522491455,
      "learning_rate": 0.0005,
      "loss": 2.9415159225463867,
      "step": 1794
    },
    {
      "epoch": 0.3462577160493827,
      "grad_norm": 1.0628671646118164,
      "learning_rate": 0.0005,
      "loss": 2.856943368911743,
      "step": 1795
    },
    {
      "epoch": 0.3464506172839506,
      "grad_norm": 1.1280291080474854,
      "learning_rate": 0.0005,
      "loss": 3.145028829574585,
      "step": 1796
    },
    {
      "epoch": 0.34664351851851855,
      "grad_norm": 0.9448619484901428,
      "learning_rate": 0.0005,
      "loss": 3.3262834548950195,
      "step": 1797
    },
    {
      "epoch": 0.34683641975308643,
      "grad_norm": 1.4317046403884888,
      "learning_rate": 0.0005,
      "loss": 2.6870133876800537,
      "step": 1798
    },
    {
      "epoch": 0.3470293209876543,
      "grad_norm": 1.1681407690048218,
      "learning_rate": 0.0005,
      "loss": 3.317636013031006,
      "step": 1799
    },
    {
      "epoch": 0.3472222222222222,
      "grad_norm": 1.349928379058838,
      "learning_rate": 0.0005,
      "loss": 3.175485610961914,
      "step": 1800
    },
    {
      "epoch": 0.3474151234567901,
      "grad_norm": 1.3544358015060425,
      "learning_rate": 0.0005,
      "loss": 2.7223377227783203,
      "step": 1801
    },
    {
      "epoch": 0.34760802469135804,
      "grad_norm": 1.0573428869247437,
      "learning_rate": 0.0005,
      "loss": 2.5442850589752197,
      "step": 1802
    },
    {
      "epoch": 0.34780092592592593,
      "grad_norm": 1.5617552995681763,
      "learning_rate": 0.0005,
      "loss": 2.7013025283813477,
      "step": 1803
    },
    {
      "epoch": 0.3479938271604938,
      "grad_norm": 0.9956454634666443,
      "learning_rate": 0.0005,
      "loss": 2.8768415451049805,
      "step": 1804
    },
    {
      "epoch": 0.3481867283950617,
      "grad_norm": 1.292054533958435,
      "learning_rate": 0.0005,
      "loss": 2.9892163276672363,
      "step": 1805
    },
    {
      "epoch": 0.34837962962962965,
      "grad_norm": 1.196774959564209,
      "learning_rate": 0.0005,
      "loss": 3.072112798690796,
      "step": 1806
    },
    {
      "epoch": 0.34857253086419754,
      "grad_norm": 1.0242851972579956,
      "learning_rate": 0.0005,
      "loss": 2.7532739639282227,
      "step": 1807
    },
    {
      "epoch": 0.3487654320987654,
      "grad_norm": 1.6989951133728027,
      "learning_rate": 0.0005,
      "loss": 2.8700027465820312,
      "step": 1808
    },
    {
      "epoch": 0.3489583333333333,
      "grad_norm": 1.5259538888931274,
      "learning_rate": 0.0005,
      "loss": 3.3334875106811523,
      "step": 1809
    },
    {
      "epoch": 0.34915123456790126,
      "grad_norm": 1.019275426864624,
      "learning_rate": 0.0005,
      "loss": 2.9107351303100586,
      "step": 1810
    },
    {
      "epoch": 0.34934413580246915,
      "grad_norm": 1.243367314338684,
      "learning_rate": 0.0005,
      "loss": 3.053392171859741,
      "step": 1811
    },
    {
      "epoch": 0.34953703703703703,
      "grad_norm": 0.9714746475219727,
      "learning_rate": 0.0005,
      "loss": 2.9070897102355957,
      "step": 1812
    },
    {
      "epoch": 0.3497299382716049,
      "grad_norm": 1.4934803247451782,
      "learning_rate": 0.0005,
      "loss": 3.017167091369629,
      "step": 1813
    },
    {
      "epoch": 0.34992283950617287,
      "grad_norm": 1.0755205154418945,
      "learning_rate": 0.0005,
      "loss": 2.7713844776153564,
      "step": 1814
    },
    {
      "epoch": 0.35011574074074076,
      "grad_norm": 1.1727408170700073,
      "learning_rate": 0.0005,
      "loss": 2.7528676986694336,
      "step": 1815
    },
    {
      "epoch": 0.35030864197530864,
      "grad_norm": 1.1520116329193115,
      "learning_rate": 0.0005,
      "loss": 3.1057076454162598,
      "step": 1816
    },
    {
      "epoch": 0.35050154320987653,
      "grad_norm": 0.9377034306526184,
      "learning_rate": 0.0005,
      "loss": 3.2788445949554443,
      "step": 1817
    },
    {
      "epoch": 0.3506944444444444,
      "grad_norm": 0.9099307656288147,
      "learning_rate": 0.0005,
      "loss": 2.8156962394714355,
      "step": 1818
    },
    {
      "epoch": 0.35088734567901236,
      "grad_norm": 1.2540901899337769,
      "learning_rate": 0.0005,
      "loss": 2.385093927383423,
      "step": 1819
    },
    {
      "epoch": 0.35108024691358025,
      "grad_norm": 2.1355745792388916,
      "learning_rate": 0.0005,
      "loss": 2.9643611907958984,
      "step": 1820
    },
    {
      "epoch": 0.35127314814814814,
      "grad_norm": 1.0395939350128174,
      "learning_rate": 0.0005,
      "loss": 2.9341535568237305,
      "step": 1821
    },
    {
      "epoch": 0.35146604938271603,
      "grad_norm": 1.0558217763900757,
      "learning_rate": 0.0005,
      "loss": 2.8117284774780273,
      "step": 1822
    },
    {
      "epoch": 0.35165895061728397,
      "grad_norm": 1.3646314144134521,
      "learning_rate": 0.0005,
      "loss": 3.1261208057403564,
      "step": 1823
    },
    {
      "epoch": 0.35185185185185186,
      "grad_norm": 0.9892771244049072,
      "learning_rate": 0.0005,
      "loss": 2.7626664638519287,
      "step": 1824
    },
    {
      "epoch": 0.35204475308641975,
      "grad_norm": 1.0881155729293823,
      "learning_rate": 0.0005,
      "loss": 3.1972832679748535,
      "step": 1825
    },
    {
      "epoch": 0.35223765432098764,
      "grad_norm": 0.9284128546714783,
      "learning_rate": 0.0005,
      "loss": 2.6929593086242676,
      "step": 1826
    },
    {
      "epoch": 0.3524305555555556,
      "grad_norm": 1.110364317893982,
      "learning_rate": 0.0005,
      "loss": 2.906217575073242,
      "step": 1827
    },
    {
      "epoch": 0.35262345679012347,
      "grad_norm": 1.100578784942627,
      "learning_rate": 0.0005,
      "loss": 2.95060396194458,
      "step": 1828
    },
    {
      "epoch": 0.35281635802469136,
      "grad_norm": 1.3928951025009155,
      "learning_rate": 0.0005,
      "loss": 2.8296713829040527,
      "step": 1829
    },
    {
      "epoch": 0.35300925925925924,
      "grad_norm": 1.0750648975372314,
      "learning_rate": 0.0005,
      "loss": 2.9896974563598633,
      "step": 1830
    },
    {
      "epoch": 0.35320216049382713,
      "grad_norm": 0.9837684035301208,
      "learning_rate": 0.0005,
      "loss": 2.922480344772339,
      "step": 1831
    },
    {
      "epoch": 0.3533950617283951,
      "grad_norm": 1.1241332292556763,
      "learning_rate": 0.0005,
      "loss": 3.072141647338867,
      "step": 1832
    },
    {
      "epoch": 0.35358796296296297,
      "grad_norm": 1.050372838973999,
      "learning_rate": 0.0005,
      "loss": 3.041221857070923,
      "step": 1833
    },
    {
      "epoch": 0.35378086419753085,
      "grad_norm": 2.0286567211151123,
      "learning_rate": 0.0005,
      "loss": 2.914127826690674,
      "step": 1834
    },
    {
      "epoch": 0.35397376543209874,
      "grad_norm": 1.10655677318573,
      "learning_rate": 0.0005,
      "loss": 2.8150675296783447,
      "step": 1835
    },
    {
      "epoch": 0.3541666666666667,
      "grad_norm": 1.0039434432983398,
      "learning_rate": 0.0005,
      "loss": 2.9360508918762207,
      "step": 1836
    },
    {
      "epoch": 0.3543595679012346,
      "grad_norm": 1.3867230415344238,
      "learning_rate": 0.0005,
      "loss": 2.5483975410461426,
      "step": 1837
    },
    {
      "epoch": 0.35455246913580246,
      "grad_norm": 1.037474274635315,
      "learning_rate": 0.0005,
      "loss": 2.7568063735961914,
      "step": 1838
    },
    {
      "epoch": 0.35474537037037035,
      "grad_norm": 1.1278131008148193,
      "learning_rate": 0.0005,
      "loss": 2.7439687252044678,
      "step": 1839
    },
    {
      "epoch": 0.3549382716049383,
      "grad_norm": 1.2891125679016113,
      "learning_rate": 0.0005,
      "loss": 3.155392646789551,
      "step": 1840
    },
    {
      "epoch": 0.3551311728395062,
      "grad_norm": 1.0209561586380005,
      "learning_rate": 0.0005,
      "loss": 2.949409008026123,
      "step": 1841
    },
    {
      "epoch": 0.35532407407407407,
      "grad_norm": 1.546140193939209,
      "learning_rate": 0.0005,
      "loss": 2.729393482208252,
      "step": 1842
    },
    {
      "epoch": 0.35551697530864196,
      "grad_norm": 0.9463042616844177,
      "learning_rate": 0.0005,
      "loss": 2.853311061859131,
      "step": 1843
    },
    {
      "epoch": 0.3557098765432099,
      "grad_norm": 1.043254017829895,
      "learning_rate": 0.0005,
      "loss": 3.0416102409362793,
      "step": 1844
    },
    {
      "epoch": 0.3559027777777778,
      "grad_norm": 1.3761730194091797,
      "learning_rate": 0.0005,
      "loss": 3.0895233154296875,
      "step": 1845
    },
    {
      "epoch": 0.3560956790123457,
      "grad_norm": 1.1476374864578247,
      "learning_rate": 0.0005,
      "loss": 2.969270706176758,
      "step": 1846
    },
    {
      "epoch": 0.35628858024691357,
      "grad_norm": 1.1709457635879517,
      "learning_rate": 0.0005,
      "loss": 2.9458260536193848,
      "step": 1847
    },
    {
      "epoch": 0.35648148148148145,
      "grad_norm": 1.1076838970184326,
      "learning_rate": 0.0005,
      "loss": 2.7759604454040527,
      "step": 1848
    },
    {
      "epoch": 0.3566743827160494,
      "grad_norm": 1.004628300666809,
      "learning_rate": 0.0005,
      "loss": 2.8387961387634277,
      "step": 1849
    },
    {
      "epoch": 0.3568672839506173,
      "grad_norm": 1.4624437093734741,
      "learning_rate": 0.0005,
      "loss": 3.0611677169799805,
      "step": 1850
    },
    {
      "epoch": 0.3570601851851852,
      "grad_norm": 1.349861979484558,
      "learning_rate": 0.0005,
      "loss": 2.753020763397217,
      "step": 1851
    },
    {
      "epoch": 0.35725308641975306,
      "grad_norm": 1.1914525032043457,
      "learning_rate": 0.0005,
      "loss": 2.6984925270080566,
      "step": 1852
    },
    {
      "epoch": 0.357445987654321,
      "grad_norm": 1.0367826223373413,
      "learning_rate": 0.0005,
      "loss": 2.79571795463562,
      "step": 1853
    },
    {
      "epoch": 0.3576388888888889,
      "grad_norm": 1.1328426599502563,
      "learning_rate": 0.0005,
      "loss": 2.8299248218536377,
      "step": 1854
    },
    {
      "epoch": 0.3578317901234568,
      "grad_norm": 1.9460515975952148,
      "learning_rate": 0.0005,
      "loss": 3.2025344371795654,
      "step": 1855
    },
    {
      "epoch": 0.35802469135802467,
      "grad_norm": 1.2480840682983398,
      "learning_rate": 0.0005,
      "loss": 2.866015911102295,
      "step": 1856
    },
    {
      "epoch": 0.3582175925925926,
      "grad_norm": 1.107469916343689,
      "learning_rate": 0.0005,
      "loss": 2.8705708980560303,
      "step": 1857
    },
    {
      "epoch": 0.3584104938271605,
      "grad_norm": 1.0667515993118286,
      "learning_rate": 0.0005,
      "loss": 2.7045047283172607,
      "step": 1858
    },
    {
      "epoch": 0.3586033950617284,
      "grad_norm": 0.9944905638694763,
      "learning_rate": 0.0005,
      "loss": 2.733550548553467,
      "step": 1859
    },
    {
      "epoch": 0.3587962962962963,
      "grad_norm": 1.002543568611145,
      "learning_rate": 0.0005,
      "loss": 2.7249410152435303,
      "step": 1860
    },
    {
      "epoch": 0.3589891975308642,
      "grad_norm": 1.2080411911010742,
      "learning_rate": 0.0005,
      "loss": 2.880549907684326,
      "step": 1861
    },
    {
      "epoch": 0.3591820987654321,
      "grad_norm": 1.2508925199508667,
      "learning_rate": 0.0005,
      "loss": 3.080502986907959,
      "step": 1862
    },
    {
      "epoch": 0.359375,
      "grad_norm": 1.0574573278427124,
      "learning_rate": 0.0005,
      "loss": 2.774749994277954,
      "step": 1863
    },
    {
      "epoch": 0.3595679012345679,
      "grad_norm": 0.9502508044242859,
      "learning_rate": 0.0005,
      "loss": 3.1576638221740723,
      "step": 1864
    },
    {
      "epoch": 0.3597608024691358,
      "grad_norm": 0.9308778643608093,
      "learning_rate": 0.0005,
      "loss": 2.8290657997131348,
      "step": 1865
    },
    {
      "epoch": 0.3599537037037037,
      "grad_norm": 1.0616929531097412,
      "learning_rate": 0.0005,
      "loss": 2.793269634246826,
      "step": 1866
    },
    {
      "epoch": 0.3601466049382716,
      "grad_norm": 1.313995361328125,
      "learning_rate": 0.0005,
      "loss": 2.780867099761963,
      "step": 1867
    },
    {
      "epoch": 0.3603395061728395,
      "grad_norm": 1.0428333282470703,
      "learning_rate": 0.0005,
      "loss": 2.6854944229125977,
      "step": 1868
    },
    {
      "epoch": 0.3605324074074074,
      "grad_norm": 1.0198346376419067,
      "learning_rate": 0.0005,
      "loss": 3.183627128601074,
      "step": 1869
    },
    {
      "epoch": 0.36072530864197533,
      "grad_norm": 1.024510145187378,
      "learning_rate": 0.0005,
      "loss": 2.942586660385132,
      "step": 1870
    },
    {
      "epoch": 0.3609182098765432,
      "grad_norm": 0.9873897433280945,
      "learning_rate": 0.0005,
      "loss": 3.086729049682617,
      "step": 1871
    },
    {
      "epoch": 0.3611111111111111,
      "grad_norm": 1.1201059818267822,
      "learning_rate": 0.0005,
      "loss": 2.6914942264556885,
      "step": 1872
    },
    {
      "epoch": 0.361304012345679,
      "grad_norm": 0.909243643283844,
      "learning_rate": 0.0005,
      "loss": 3.1842284202575684,
      "step": 1873
    },
    {
      "epoch": 0.36149691358024694,
      "grad_norm": 1.0688844919204712,
      "learning_rate": 0.0005,
      "loss": 2.7189931869506836,
      "step": 1874
    },
    {
      "epoch": 0.3616898148148148,
      "grad_norm": 0.8491013050079346,
      "learning_rate": 0.0005,
      "loss": 3.027265787124634,
      "step": 1875
    },
    {
      "epoch": 0.3618827160493827,
      "grad_norm": 1.1744028329849243,
      "learning_rate": 0.0005,
      "loss": 2.9661428928375244,
      "step": 1876
    },
    {
      "epoch": 0.3620756172839506,
      "grad_norm": 1.622371792793274,
      "learning_rate": 0.0005,
      "loss": 2.8511688709259033,
      "step": 1877
    },
    {
      "epoch": 0.36226851851851855,
      "grad_norm": 1.0052576065063477,
      "learning_rate": 0.0005,
      "loss": 3.2562172412872314,
      "step": 1878
    },
    {
      "epoch": 0.36246141975308643,
      "grad_norm": 1.0646462440490723,
      "learning_rate": 0.0005,
      "loss": 2.4937398433685303,
      "step": 1879
    },
    {
      "epoch": 0.3626543209876543,
      "grad_norm": 1.0271105766296387,
      "learning_rate": 0.0005,
      "loss": 2.6980819702148438,
      "step": 1880
    },
    {
      "epoch": 0.3628472222222222,
      "grad_norm": 0.9273824691772461,
      "learning_rate": 0.0005,
      "loss": 3.1350529193878174,
      "step": 1881
    },
    {
      "epoch": 0.3630401234567901,
      "grad_norm": 1.0527366399765015,
      "learning_rate": 0.0005,
      "loss": 2.5818405151367188,
      "step": 1882
    },
    {
      "epoch": 0.36323302469135804,
      "grad_norm": 1.4585765600204468,
      "learning_rate": 0.0005,
      "loss": 3.307713270187378,
      "step": 1883
    },
    {
      "epoch": 0.36342592592592593,
      "grad_norm": 1.1516088247299194,
      "learning_rate": 0.0005,
      "loss": 3.2744429111480713,
      "step": 1884
    },
    {
      "epoch": 0.3636188271604938,
      "grad_norm": 1.0007749795913696,
      "learning_rate": 0.0005,
      "loss": 2.8413822650909424,
      "step": 1885
    },
    {
      "epoch": 0.3638117283950617,
      "grad_norm": 0.9932330846786499,
      "learning_rate": 0.0005,
      "loss": 3.1911025047302246,
      "step": 1886
    },
    {
      "epoch": 0.36400462962962965,
      "grad_norm": 1.2107396125793457,
      "learning_rate": 0.0005,
      "loss": 3.2141518592834473,
      "step": 1887
    },
    {
      "epoch": 0.36419753086419754,
      "grad_norm": 0.8687191009521484,
      "learning_rate": 0.0005,
      "loss": 2.932096481323242,
      "step": 1888
    },
    {
      "epoch": 0.3643904320987654,
      "grad_norm": 0.9685500264167786,
      "learning_rate": 0.0005,
      "loss": 2.5764613151550293,
      "step": 1889
    },
    {
      "epoch": 0.3645833333333333,
      "grad_norm": 0.9619494080543518,
      "learning_rate": 0.0005,
      "loss": 2.978175163269043,
      "step": 1890
    },
    {
      "epoch": 0.36477623456790126,
      "grad_norm": 1.0653373003005981,
      "learning_rate": 0.0005,
      "loss": 2.688720941543579,
      "step": 1891
    },
    {
      "epoch": 0.36496913580246915,
      "grad_norm": 0.9962583184242249,
      "learning_rate": 0.0005,
      "loss": 2.9139304161071777,
      "step": 1892
    },
    {
      "epoch": 0.36516203703703703,
      "grad_norm": 1.2221657037734985,
      "learning_rate": 0.0005,
      "loss": 3.1390466690063477,
      "step": 1893
    },
    {
      "epoch": 0.3653549382716049,
      "grad_norm": 1.0166919231414795,
      "learning_rate": 0.0005,
      "loss": 3.0222482681274414,
      "step": 1894
    },
    {
      "epoch": 0.36554783950617287,
      "grad_norm": 1.1346269845962524,
      "learning_rate": 0.0005,
      "loss": 3.216400146484375,
      "step": 1895
    },
    {
      "epoch": 0.36574074074074076,
      "grad_norm": 1.337988018989563,
      "learning_rate": 0.0005,
      "loss": 2.8074119091033936,
      "step": 1896
    },
    {
      "epoch": 0.36593364197530864,
      "grad_norm": 27.780370712280273,
      "learning_rate": 0.0005,
      "loss": 2.5120530128479004,
      "step": 1897
    },
    {
      "epoch": 0.36612654320987653,
      "grad_norm": 1.6571828126907349,
      "learning_rate": 0.0005,
      "loss": 3.125947952270508,
      "step": 1898
    },
    {
      "epoch": 0.3663194444444444,
      "grad_norm": 1.3632375001907349,
      "learning_rate": 0.0005,
      "loss": 2.98823881149292,
      "step": 1899
    },
    {
      "epoch": 0.36651234567901236,
      "grad_norm": 2.7468934059143066,
      "learning_rate": 0.0005,
      "loss": 2.86435866355896,
      "step": 1900
    },
    {
      "epoch": 0.36670524691358025,
      "grad_norm": 1.160940408706665,
      "learning_rate": 0.0005,
      "loss": 2.8841278553009033,
      "step": 1901
    },
    {
      "epoch": 0.36689814814814814,
      "grad_norm": 1.0530880689620972,
      "learning_rate": 0.0005,
      "loss": 3.031264305114746,
      "step": 1902
    },
    {
      "epoch": 0.36709104938271603,
      "grad_norm": 1.2317012548446655,
      "learning_rate": 0.0005,
      "loss": 3.1664915084838867,
      "step": 1903
    },
    {
      "epoch": 0.36728395061728397,
      "grad_norm": 1.4449533224105835,
      "learning_rate": 0.0005,
      "loss": 2.853091239929199,
      "step": 1904
    },
    {
      "epoch": 0.36747685185185186,
      "grad_norm": 1.098778247833252,
      "learning_rate": 0.0005,
      "loss": 2.9855997562408447,
      "step": 1905
    },
    {
      "epoch": 0.36766975308641975,
      "grad_norm": 1.4945725202560425,
      "learning_rate": 0.0005,
      "loss": 2.939284324645996,
      "step": 1906
    },
    {
      "epoch": 0.36786265432098764,
      "grad_norm": 1.3518099784851074,
      "learning_rate": 0.0005,
      "loss": 3.323455810546875,
      "step": 1907
    },
    {
      "epoch": 0.3680555555555556,
      "grad_norm": 0.9392699599266052,
      "learning_rate": 0.0005,
      "loss": 2.976696252822876,
      "step": 1908
    },
    {
      "epoch": 0.36824845679012347,
      "grad_norm": 1.1189172267913818,
      "learning_rate": 0.0005,
      "loss": 3.059190273284912,
      "step": 1909
    },
    {
      "epoch": 0.36844135802469136,
      "grad_norm": 1.1103509664535522,
      "learning_rate": 0.0005,
      "loss": 3.1539306640625,
      "step": 1910
    },
    {
      "epoch": 0.36863425925925924,
      "grad_norm": 1.2743295431137085,
      "learning_rate": 0.0005,
      "loss": 3.086869955062866,
      "step": 1911
    },
    {
      "epoch": 0.36882716049382713,
      "grad_norm": 0.9315739870071411,
      "learning_rate": 0.0005,
      "loss": 2.9091110229492188,
      "step": 1912
    },
    {
      "epoch": 0.3690200617283951,
      "grad_norm": 1.1852877140045166,
      "learning_rate": 0.0005,
      "loss": 2.8137705326080322,
      "step": 1913
    },
    {
      "epoch": 0.36921296296296297,
      "grad_norm": 1.0414670705795288,
      "learning_rate": 0.0005,
      "loss": 3.122606039047241,
      "step": 1914
    },
    {
      "epoch": 0.36940586419753085,
      "grad_norm": 1.5098949670791626,
      "learning_rate": 0.0005,
      "loss": 2.883277654647827,
      "step": 1915
    },
    {
      "epoch": 0.36959876543209874,
      "grad_norm": 1.1095370054244995,
      "learning_rate": 0.0005,
      "loss": 2.806992530822754,
      "step": 1916
    },
    {
      "epoch": 0.3697916666666667,
      "grad_norm": 1.0669676065444946,
      "learning_rate": 0.0005,
      "loss": 2.8516342639923096,
      "step": 1917
    },
    {
      "epoch": 0.3699845679012346,
      "grad_norm": 1.0811494588851929,
      "learning_rate": 0.0005,
      "loss": 3.024989604949951,
      "step": 1918
    },
    {
      "epoch": 0.37017746913580246,
      "grad_norm": 1.038159728050232,
      "learning_rate": 0.0005,
      "loss": 2.987344264984131,
      "step": 1919
    },
    {
      "epoch": 0.37037037037037035,
      "grad_norm": 0.9595372676849365,
      "learning_rate": 0.0005,
      "loss": 2.8329713344573975,
      "step": 1920
    },
    {
      "epoch": 0.3705632716049383,
      "grad_norm": 1.1301769018173218,
      "learning_rate": 0.0005,
      "loss": 2.6860227584838867,
      "step": 1921
    },
    {
      "epoch": 0.3707561728395062,
      "grad_norm": 1.066049337387085,
      "learning_rate": 0.0005,
      "loss": 2.6539812088012695,
      "step": 1922
    },
    {
      "epoch": 0.37094907407407407,
      "grad_norm": 0.9354440569877625,
      "learning_rate": 0.0005,
      "loss": 2.6969962120056152,
      "step": 1923
    },
    {
      "epoch": 0.37114197530864196,
      "grad_norm": 1.1510295867919922,
      "learning_rate": 0.0005,
      "loss": 2.9259912967681885,
      "step": 1924
    },
    {
      "epoch": 0.3713348765432099,
      "grad_norm": 0.993327260017395,
      "learning_rate": 0.0005,
      "loss": 2.96394944190979,
      "step": 1925
    },
    {
      "epoch": 0.3715277777777778,
      "grad_norm": 1.0719255208969116,
      "learning_rate": 0.0005,
      "loss": 2.992434501647949,
      "step": 1926
    },
    {
      "epoch": 0.3717206790123457,
      "grad_norm": 1.3848117589950562,
      "learning_rate": 0.0005,
      "loss": 2.499556541442871,
      "step": 1927
    },
    {
      "epoch": 0.37191358024691357,
      "grad_norm": 0.9480854272842407,
      "learning_rate": 0.0005,
      "loss": 2.8432974815368652,
      "step": 1928
    },
    {
      "epoch": 0.37210648148148145,
      "grad_norm": 1.080667495727539,
      "learning_rate": 0.0005,
      "loss": 2.5447235107421875,
      "step": 1929
    },
    {
      "epoch": 0.3722993827160494,
      "grad_norm": 1.3963581323623657,
      "learning_rate": 0.0005,
      "loss": 3.001023769378662,
      "step": 1930
    },
    {
      "epoch": 0.3724922839506173,
      "grad_norm": 1.2027015686035156,
      "learning_rate": 0.0005,
      "loss": 3.098444938659668,
      "step": 1931
    },
    {
      "epoch": 0.3726851851851852,
      "grad_norm": 1.050284743309021,
      "learning_rate": 0.0005,
      "loss": 2.9555163383483887,
      "step": 1932
    },
    {
      "epoch": 0.37287808641975306,
      "grad_norm": 1.2049858570098877,
      "learning_rate": 0.0005,
      "loss": 2.8412294387817383,
      "step": 1933
    },
    {
      "epoch": 0.373070987654321,
      "grad_norm": 0.8481136560440063,
      "learning_rate": 0.0005,
      "loss": 2.8632731437683105,
      "step": 1934
    },
    {
      "epoch": 0.3732638888888889,
      "grad_norm": 1.4162507057189941,
      "learning_rate": 0.0005,
      "loss": 3.0273170471191406,
      "step": 1935
    },
    {
      "epoch": 0.3734567901234568,
      "grad_norm": 1.0347925424575806,
      "learning_rate": 0.0005,
      "loss": 3.0635643005371094,
      "step": 1936
    },
    {
      "epoch": 0.37364969135802467,
      "grad_norm": 1.227612018585205,
      "learning_rate": 0.0005,
      "loss": 3.2677788734436035,
      "step": 1937
    },
    {
      "epoch": 0.3738425925925926,
      "grad_norm": 1.0633823871612549,
      "learning_rate": 0.0005,
      "loss": 3.0227127075195312,
      "step": 1938
    },
    {
      "epoch": 0.3740354938271605,
      "grad_norm": 1.2333145141601562,
      "learning_rate": 0.0005,
      "loss": 2.9023170471191406,
      "step": 1939
    },
    {
      "epoch": 0.3742283950617284,
      "grad_norm": 0.9551900029182434,
      "learning_rate": 0.0005,
      "loss": 2.681368350982666,
      "step": 1940
    },
    {
      "epoch": 0.3744212962962963,
      "grad_norm": 1.084373116493225,
      "learning_rate": 0.0005,
      "loss": 2.8949975967407227,
      "step": 1941
    },
    {
      "epoch": 0.3746141975308642,
      "grad_norm": 1.087235927581787,
      "learning_rate": 0.0005,
      "loss": 2.7494640350341797,
      "step": 1942
    },
    {
      "epoch": 0.3748070987654321,
      "grad_norm": 1.124159812927246,
      "learning_rate": 0.0005,
      "loss": 2.6335856914520264,
      "step": 1943
    },
    {
      "epoch": 0.375,
      "grad_norm": 0.9872838854789734,
      "learning_rate": 0.0005,
      "loss": 2.5818099975585938,
      "step": 1944
    },
    {
      "epoch": 0.3751929012345679,
      "grad_norm": 0.9308594465255737,
      "learning_rate": 0.0005,
      "loss": 3.170116424560547,
      "step": 1945
    },
    {
      "epoch": 0.3753858024691358,
      "grad_norm": 1.0639110803604126,
      "learning_rate": 0.0005,
      "loss": 2.5825958251953125,
      "step": 1946
    },
    {
      "epoch": 0.3755787037037037,
      "grad_norm": 1.0409152507781982,
      "learning_rate": 0.0005,
      "loss": 2.7471208572387695,
      "step": 1947
    },
    {
      "epoch": 0.3757716049382716,
      "grad_norm": 1.3732521533966064,
      "learning_rate": 0.0005,
      "loss": 2.997316360473633,
      "step": 1948
    },
    {
      "epoch": 0.3759645061728395,
      "grad_norm": 1.056031346321106,
      "learning_rate": 0.0005,
      "loss": 2.93198823928833,
      "step": 1949
    },
    {
      "epoch": 0.3761574074074074,
      "grad_norm": 1.218176007270813,
      "learning_rate": 0.0005,
      "loss": 2.802593231201172,
      "step": 1950
    },
    {
      "epoch": 0.37635030864197533,
      "grad_norm": 0.9447706937789917,
      "learning_rate": 0.0005,
      "loss": 3.129380226135254,
      "step": 1951
    },
    {
      "epoch": 0.3765432098765432,
      "grad_norm": 1.0359063148498535,
      "learning_rate": 0.0005,
      "loss": 2.5763931274414062,
      "step": 1952
    },
    {
      "epoch": 0.3767361111111111,
      "grad_norm": 0.9542102217674255,
      "learning_rate": 0.0005,
      "loss": 2.9489104747772217,
      "step": 1953
    },
    {
      "epoch": 0.376929012345679,
      "grad_norm": 0.9696828722953796,
      "learning_rate": 0.0005,
      "loss": 2.819723129272461,
      "step": 1954
    },
    {
      "epoch": 0.37712191358024694,
      "grad_norm": 1.1610702276229858,
      "learning_rate": 0.0005,
      "loss": 2.7218027114868164,
      "step": 1955
    },
    {
      "epoch": 0.3773148148148148,
      "grad_norm": 0.9806154370307922,
      "learning_rate": 0.0005,
      "loss": 2.5241098403930664,
      "step": 1956
    },
    {
      "epoch": 0.3775077160493827,
      "grad_norm": 0.9548748731613159,
      "learning_rate": 0.0005,
      "loss": 3.131108522415161,
      "step": 1957
    },
    {
      "epoch": 0.3777006172839506,
      "grad_norm": 1.0714503526687622,
      "learning_rate": 0.0005,
      "loss": 2.9689154624938965,
      "step": 1958
    },
    {
      "epoch": 0.37789351851851855,
      "grad_norm": 1.3781952857971191,
      "learning_rate": 0.0005,
      "loss": 3.003328800201416,
      "step": 1959
    },
    {
      "epoch": 0.37808641975308643,
      "grad_norm": 1.5054352283477783,
      "learning_rate": 0.0005,
      "loss": 3.1975531578063965,
      "step": 1960
    },
    {
      "epoch": 0.3782793209876543,
      "grad_norm": 1.3634854555130005,
      "learning_rate": 0.0005,
      "loss": 2.7668542861938477,
      "step": 1961
    },
    {
      "epoch": 0.3784722222222222,
      "grad_norm": 1.2348554134368896,
      "learning_rate": 0.0005,
      "loss": 2.805691957473755,
      "step": 1962
    },
    {
      "epoch": 0.3786651234567901,
      "grad_norm": 0.8645180463790894,
      "learning_rate": 0.0005,
      "loss": 2.7012600898742676,
      "step": 1963
    },
    {
      "epoch": 0.37885802469135804,
      "grad_norm": 1.4686152935028076,
      "learning_rate": 0.0005,
      "loss": 2.901340961456299,
      "step": 1964
    },
    {
      "epoch": 0.37905092592592593,
      "grad_norm": 1.137800693511963,
      "learning_rate": 0.0005,
      "loss": 3.253934383392334,
      "step": 1965
    },
    {
      "epoch": 0.3792438271604938,
      "grad_norm": 1.0105557441711426,
      "learning_rate": 0.0005,
      "loss": 2.968698024749756,
      "step": 1966
    },
    {
      "epoch": 0.3794367283950617,
      "grad_norm": 1.323936939239502,
      "learning_rate": 0.0005,
      "loss": 2.8195228576660156,
      "step": 1967
    },
    {
      "epoch": 0.37962962962962965,
      "grad_norm": 1.0486276149749756,
      "learning_rate": 0.0005,
      "loss": 3.1846413612365723,
      "step": 1968
    },
    {
      "epoch": 0.37982253086419754,
      "grad_norm": 1.8890998363494873,
      "learning_rate": 0.0005,
      "loss": 2.6837782859802246,
      "step": 1969
    },
    {
      "epoch": 0.3800154320987654,
      "grad_norm": 1.3045440912246704,
      "learning_rate": 0.0005,
      "loss": 3.1270220279693604,
      "step": 1970
    },
    {
      "epoch": 0.3802083333333333,
      "grad_norm": 0.9404866695404053,
      "learning_rate": 0.0005,
      "loss": 2.786205768585205,
      "step": 1971
    },
    {
      "epoch": 0.38040123456790126,
      "grad_norm": 1.4060357809066772,
      "learning_rate": 0.0005,
      "loss": 2.944959878921509,
      "step": 1972
    },
    {
      "epoch": 0.38059413580246915,
      "grad_norm": 5.0721435546875,
      "learning_rate": 0.0005,
      "loss": 3.001704692840576,
      "step": 1973
    },
    {
      "epoch": 0.38078703703703703,
      "grad_norm": 1.4883365631103516,
      "learning_rate": 0.0005,
      "loss": 3.2271344661712646,
      "step": 1974
    },
    {
      "epoch": 0.3809799382716049,
      "grad_norm": 1.0261949300765991,
      "learning_rate": 0.0005,
      "loss": 2.9799180030822754,
      "step": 1975
    },
    {
      "epoch": 0.38117283950617287,
      "grad_norm": 1.1590077877044678,
      "learning_rate": 0.0005,
      "loss": 2.8552536964416504,
      "step": 1976
    },
    {
      "epoch": 0.38136574074074076,
      "grad_norm": 1.3375060558319092,
      "learning_rate": 0.0005,
      "loss": 2.67509388923645,
      "step": 1977
    },
    {
      "epoch": 0.38155864197530864,
      "grad_norm": 1.0867128372192383,
      "learning_rate": 0.0005,
      "loss": 2.955688953399658,
      "step": 1978
    },
    {
      "epoch": 0.38175154320987653,
      "grad_norm": 0.9399468898773193,
      "learning_rate": 0.0005,
      "loss": 2.823075294494629,
      "step": 1979
    },
    {
      "epoch": 0.3819444444444444,
      "grad_norm": 1.3013381958007812,
      "learning_rate": 0.0005,
      "loss": 2.8182053565979004,
      "step": 1980
    },
    {
      "epoch": 0.38213734567901236,
      "grad_norm": 1.3418012857437134,
      "learning_rate": 0.0005,
      "loss": 3.3765885829925537,
      "step": 1981
    },
    {
      "epoch": 0.38233024691358025,
      "grad_norm": 1.2058593034744263,
      "learning_rate": 0.0005,
      "loss": 2.5894432067871094,
      "step": 1982
    },
    {
      "epoch": 0.38252314814814814,
      "grad_norm": 1.2070367336273193,
      "learning_rate": 0.0005,
      "loss": 2.9426517486572266,
      "step": 1983
    },
    {
      "epoch": 0.38271604938271603,
      "grad_norm": 1.0810426473617554,
      "learning_rate": 0.0005,
      "loss": 2.816197395324707,
      "step": 1984
    },
    {
      "epoch": 0.38290895061728397,
      "grad_norm": 0.9855653643608093,
      "learning_rate": 0.0005,
      "loss": 2.9037132263183594,
      "step": 1985
    },
    {
      "epoch": 0.38310185185185186,
      "grad_norm": 1.3098043203353882,
      "learning_rate": 0.0005,
      "loss": 2.788698434829712,
      "step": 1986
    },
    {
      "epoch": 0.38329475308641975,
      "grad_norm": 1.629482388496399,
      "learning_rate": 0.0005,
      "loss": 2.930814743041992,
      "step": 1987
    },
    {
      "epoch": 0.38348765432098764,
      "grad_norm": 1.3628950119018555,
      "learning_rate": 0.0005,
      "loss": 2.956986904144287,
      "step": 1988
    },
    {
      "epoch": 0.3836805555555556,
      "grad_norm": 1.0424264669418335,
      "learning_rate": 0.0005,
      "loss": 2.7693862915039062,
      "step": 1989
    },
    {
      "epoch": 0.38387345679012347,
      "grad_norm": 1.1343903541564941,
      "learning_rate": 0.0005,
      "loss": 2.7026798725128174,
      "step": 1990
    },
    {
      "epoch": 0.38406635802469136,
      "grad_norm": 0.9883715510368347,
      "learning_rate": 0.0005,
      "loss": 2.814213752746582,
      "step": 1991
    },
    {
      "epoch": 0.38425925925925924,
      "grad_norm": 0.9637265801429749,
      "learning_rate": 0.0005,
      "loss": 2.9766361713409424,
      "step": 1992
    },
    {
      "epoch": 0.38445216049382713,
      "grad_norm": 1.0460253953933716,
      "learning_rate": 0.0005,
      "loss": 2.7881832122802734,
      "step": 1993
    },
    {
      "epoch": 0.3846450617283951,
      "grad_norm": 1.0255895853042603,
      "learning_rate": 0.0005,
      "loss": 2.845305919647217,
      "step": 1994
    },
    {
      "epoch": 0.38483796296296297,
      "grad_norm": 1.041378140449524,
      "learning_rate": 0.0005,
      "loss": 2.972295045852661,
      "step": 1995
    },
    {
      "epoch": 0.38503086419753085,
      "grad_norm": 1.089224100112915,
      "learning_rate": 0.0005,
      "loss": 2.904855251312256,
      "step": 1996
    },
    {
      "epoch": 0.38522376543209874,
      "grad_norm": 0.9838478565216064,
      "learning_rate": 0.0005,
      "loss": 2.9808857440948486,
      "step": 1997
    },
    {
      "epoch": 0.3854166666666667,
      "grad_norm": 1.047728419303894,
      "learning_rate": 0.0005,
      "loss": 2.6077721118927,
      "step": 1998
    },
    {
      "epoch": 0.3856095679012346,
      "grad_norm": 1.376776933670044,
      "learning_rate": 0.0005,
      "loss": 2.836358070373535,
      "step": 1999
    },
    {
      "epoch": 0.38580246913580246,
      "grad_norm": 1.2043671607971191,
      "learning_rate": 0.0005,
      "loss": 3.0084736347198486,
      "step": 2000
    },
    {
      "epoch": 0.38599537037037035,
      "grad_norm": 0.9321793913841248,
      "learning_rate": 0.0005,
      "loss": 2.9599108695983887,
      "step": 2001
    },
    {
      "epoch": 0.3861882716049383,
      "grad_norm": 1.0557023286819458,
      "learning_rate": 0.0005,
      "loss": 2.749427318572998,
      "step": 2002
    },
    {
      "epoch": 0.3863811728395062,
      "grad_norm": 1.2281370162963867,
      "learning_rate": 0.0005,
      "loss": 2.8081912994384766,
      "step": 2003
    },
    {
      "epoch": 0.38657407407407407,
      "grad_norm": 1.1324667930603027,
      "learning_rate": 0.0005,
      "loss": 2.9093332290649414,
      "step": 2004
    },
    {
      "epoch": 0.38676697530864196,
      "grad_norm": 1.1161688566207886,
      "learning_rate": 0.0005,
      "loss": 2.84649658203125,
      "step": 2005
    },
    {
      "epoch": 0.3869598765432099,
      "grad_norm": 1.0920400619506836,
      "learning_rate": 0.0005,
      "loss": 3.043342351913452,
      "step": 2006
    },
    {
      "epoch": 0.3871527777777778,
      "grad_norm": 1.1531368494033813,
      "learning_rate": 0.0005,
      "loss": 3.0296380519866943,
      "step": 2007
    },
    {
      "epoch": 0.3873456790123457,
      "grad_norm": 1.9956390857696533,
      "learning_rate": 0.0005,
      "loss": 2.96517276763916,
      "step": 2008
    },
    {
      "epoch": 0.38753858024691357,
      "grad_norm": 0.9525036215782166,
      "learning_rate": 0.0005,
      "loss": 2.9371635913848877,
      "step": 2009
    },
    {
      "epoch": 0.38773148148148145,
      "grad_norm": 1.055627703666687,
      "learning_rate": 0.0005,
      "loss": 3.0024819374084473,
      "step": 2010
    },
    {
      "epoch": 0.3879243827160494,
      "grad_norm": 0.9347386360168457,
      "learning_rate": 0.0005,
      "loss": 2.819821834564209,
      "step": 2011
    },
    {
      "epoch": 0.3881172839506173,
      "grad_norm": 1.0699706077575684,
      "learning_rate": 0.0005,
      "loss": 2.826547384262085,
      "step": 2012
    },
    {
      "epoch": 0.3883101851851852,
      "grad_norm": 1.1122738122940063,
      "learning_rate": 0.0005,
      "loss": 2.5795583724975586,
      "step": 2013
    },
    {
      "epoch": 0.38850308641975306,
      "grad_norm": 8.977659225463867,
      "learning_rate": 0.0005,
      "loss": 2.6250569820404053,
      "step": 2014
    },
    {
      "epoch": 0.388695987654321,
      "grad_norm": 0.94841068983078,
      "learning_rate": 0.0005,
      "loss": 2.9208621978759766,
      "step": 2015
    },
    {
      "epoch": 0.3888888888888889,
      "grad_norm": 1.0942822694778442,
      "learning_rate": 0.0005,
      "loss": 2.979102849960327,
      "step": 2016
    },
    {
      "epoch": 0.3890817901234568,
      "grad_norm": 1.1148587465286255,
      "learning_rate": 0.0005,
      "loss": 2.8753662109375,
      "step": 2017
    },
    {
      "epoch": 0.38927469135802467,
      "grad_norm": 0.9249332547187805,
      "learning_rate": 0.0005,
      "loss": 2.8840107917785645,
      "step": 2018
    },
    {
      "epoch": 0.3894675925925926,
      "grad_norm": 1.7322481870651245,
      "learning_rate": 0.0005,
      "loss": 2.6811604499816895,
      "step": 2019
    },
    {
      "epoch": 0.3896604938271605,
      "grad_norm": 2.0261390209198,
      "learning_rate": 0.0005,
      "loss": 2.8073878288269043,
      "step": 2020
    },
    {
      "epoch": 0.3898533950617284,
      "grad_norm": 1.524486780166626,
      "learning_rate": 0.0005,
      "loss": 2.8828840255737305,
      "step": 2021
    },
    {
      "epoch": 0.3900462962962963,
      "grad_norm": 1.0235958099365234,
      "learning_rate": 0.0005,
      "loss": 2.9581298828125,
      "step": 2022
    },
    {
      "epoch": 0.3902391975308642,
      "grad_norm": 1.0314463376998901,
      "learning_rate": 0.0005,
      "loss": 2.945526599884033,
      "step": 2023
    },
    {
      "epoch": 0.3904320987654321,
      "grad_norm": 1.2121195793151855,
      "learning_rate": 0.0005,
      "loss": 2.8976876735687256,
      "step": 2024
    },
    {
      "epoch": 0.390625,
      "grad_norm": 1.0583202838897705,
      "learning_rate": 0.0005,
      "loss": 2.582209825515747,
      "step": 2025
    },
    {
      "epoch": 0.3908179012345679,
      "grad_norm": 1.252543568611145,
      "learning_rate": 0.0005,
      "loss": 2.807391405105591,
      "step": 2026
    },
    {
      "epoch": 0.3910108024691358,
      "grad_norm": 1.2127721309661865,
      "learning_rate": 0.0005,
      "loss": 3.0718631744384766,
      "step": 2027
    },
    {
      "epoch": 0.3912037037037037,
      "grad_norm": 1.291550636291504,
      "learning_rate": 0.0005,
      "loss": 2.7149064540863037,
      "step": 2028
    },
    {
      "epoch": 0.3913966049382716,
      "grad_norm": 0.8842282295227051,
      "learning_rate": 0.0005,
      "loss": 2.7012176513671875,
      "step": 2029
    },
    {
      "epoch": 0.3915895061728395,
      "grad_norm": 1.123905897140503,
      "learning_rate": 0.0005,
      "loss": 3.071478843688965,
      "step": 2030
    },
    {
      "epoch": 0.3917824074074074,
      "grad_norm": 1.052509069442749,
      "learning_rate": 0.0005,
      "loss": 3.10776686668396,
      "step": 2031
    },
    {
      "epoch": 0.39197530864197533,
      "grad_norm": 0.9126524925231934,
      "learning_rate": 0.0005,
      "loss": 2.634113311767578,
      "step": 2032
    },
    {
      "epoch": 0.3921682098765432,
      "grad_norm": 1.0364649295806885,
      "learning_rate": 0.0005,
      "loss": 2.791140556335449,
      "step": 2033
    },
    {
      "epoch": 0.3923611111111111,
      "grad_norm": 1.2120144367218018,
      "learning_rate": 0.0005,
      "loss": 2.6160054206848145,
      "step": 2034
    },
    {
      "epoch": 0.392554012345679,
      "grad_norm": 1.0788601636886597,
      "learning_rate": 0.0005,
      "loss": 2.8111534118652344,
      "step": 2035
    },
    {
      "epoch": 0.39274691358024694,
      "grad_norm": 1.0291354656219482,
      "learning_rate": 0.0005,
      "loss": 2.5587196350097656,
      "step": 2036
    },
    {
      "epoch": 0.3929398148148148,
      "grad_norm": 0.9034119844436646,
      "learning_rate": 0.0005,
      "loss": 2.9209489822387695,
      "step": 2037
    },
    {
      "epoch": 0.3931327160493827,
      "grad_norm": 0.752697229385376,
      "learning_rate": 0.0005,
      "loss": 2.967134475708008,
      "step": 2038
    },
    {
      "epoch": 0.3933256172839506,
      "grad_norm": 1.0156341791152954,
      "learning_rate": 0.0005,
      "loss": 2.895606517791748,
      "step": 2039
    },
    {
      "epoch": 0.39351851851851855,
      "grad_norm": 0.9906331896781921,
      "learning_rate": 0.0005,
      "loss": 2.8454408645629883,
      "step": 2040
    },
    {
      "epoch": 0.39371141975308643,
      "grad_norm": 0.9865227341651917,
      "learning_rate": 0.0005,
      "loss": 2.9377081394195557,
      "step": 2041
    },
    {
      "epoch": 0.3939043209876543,
      "grad_norm": 1.1995598077774048,
      "learning_rate": 0.0005,
      "loss": 2.471489191055298,
      "step": 2042
    },
    {
      "epoch": 0.3940972222222222,
      "grad_norm": 1.1729532480239868,
      "learning_rate": 0.0005,
      "loss": 2.965371608734131,
      "step": 2043
    },
    {
      "epoch": 0.3942901234567901,
      "grad_norm": 1.0269639492034912,
      "learning_rate": 0.0005,
      "loss": 2.6324257850646973,
      "step": 2044
    },
    {
      "epoch": 0.39448302469135804,
      "grad_norm": 1.259600043296814,
      "learning_rate": 0.0005,
      "loss": 2.680551528930664,
      "step": 2045
    },
    {
      "epoch": 0.39467592592592593,
      "grad_norm": 1.0171737670898438,
      "learning_rate": 0.0005,
      "loss": 2.6982407569885254,
      "step": 2046
    },
    {
      "epoch": 0.3948688271604938,
      "grad_norm": 1.0706099271774292,
      "learning_rate": 0.0005,
      "loss": 2.7861433029174805,
      "step": 2047
    },
    {
      "epoch": 0.3950617283950617,
      "grad_norm": 1.3991690874099731,
      "learning_rate": 0.0005,
      "loss": 3.0182929039001465,
      "step": 2048
    },
    {
      "epoch": 0.39525462962962965,
      "grad_norm": 1.4050955772399902,
      "learning_rate": 0.0005,
      "loss": 3.1008410453796387,
      "step": 2049
    },
    {
      "epoch": 0.39544753086419754,
      "grad_norm": 0.920244574546814,
      "learning_rate": 0.0005,
      "loss": 2.7473113536834717,
      "step": 2050
    },
    {
      "epoch": 0.3956404320987654,
      "grad_norm": 0.8401632308959961,
      "learning_rate": 0.0005,
      "loss": 2.772373676300049,
      "step": 2051
    },
    {
      "epoch": 0.3958333333333333,
      "grad_norm": 1.0072580575942993,
      "learning_rate": 0.0005,
      "loss": 2.9263391494750977,
      "step": 2052
    },
    {
      "epoch": 0.39602623456790126,
      "grad_norm": 1.0038429498672485,
      "learning_rate": 0.0005,
      "loss": 2.8556694984436035,
      "step": 2053
    },
    {
      "epoch": 0.39621913580246915,
      "grad_norm": 1.0191227197647095,
      "learning_rate": 0.0005,
      "loss": 2.9836347103118896,
      "step": 2054
    },
    {
      "epoch": 0.39641203703703703,
      "grad_norm": 1.4362473487854004,
      "learning_rate": 0.0005,
      "loss": 2.735135316848755,
      "step": 2055
    },
    {
      "epoch": 0.3966049382716049,
      "grad_norm": 1.164077639579773,
      "learning_rate": 0.0005,
      "loss": 2.381136894226074,
      "step": 2056
    },
    {
      "epoch": 0.39679783950617287,
      "grad_norm": 1.2590197324752808,
      "learning_rate": 0.0005,
      "loss": 2.9230449199676514,
      "step": 2057
    },
    {
      "epoch": 0.39699074074074076,
      "grad_norm": 1.1223560571670532,
      "learning_rate": 0.0005,
      "loss": 3.0717201232910156,
      "step": 2058
    },
    {
      "epoch": 0.39718364197530864,
      "grad_norm": 1.079050898551941,
      "learning_rate": 0.0005,
      "loss": 2.4729743003845215,
      "step": 2059
    },
    {
      "epoch": 0.39737654320987653,
      "grad_norm": 1.0445767641067505,
      "learning_rate": 0.0005,
      "loss": 2.990684986114502,
      "step": 2060
    },
    {
      "epoch": 0.3975694444444444,
      "grad_norm": 1.1050992012023926,
      "learning_rate": 0.0005,
      "loss": 3.0531468391418457,
      "step": 2061
    },
    {
      "epoch": 0.39776234567901236,
      "grad_norm": 1.0446953773498535,
      "learning_rate": 0.0005,
      "loss": 2.64328670501709,
      "step": 2062
    },
    {
      "epoch": 0.39795524691358025,
      "grad_norm": 0.8907238245010376,
      "learning_rate": 0.0005,
      "loss": 2.591179609298706,
      "step": 2063
    },
    {
      "epoch": 0.39814814814814814,
      "grad_norm": 1.1589909791946411,
      "learning_rate": 0.0005,
      "loss": 2.826378583908081,
      "step": 2064
    },
    {
      "epoch": 0.39834104938271603,
      "grad_norm": 1.2640798091888428,
      "learning_rate": 0.0005,
      "loss": 3.0582704544067383,
      "step": 2065
    },
    {
      "epoch": 0.39853395061728397,
      "grad_norm": 1.0911563634872437,
      "learning_rate": 0.0005,
      "loss": 2.7632148265838623,
      "step": 2066
    },
    {
      "epoch": 0.39872685185185186,
      "grad_norm": 0.845007061958313,
      "learning_rate": 0.0005,
      "loss": 2.6218645572662354,
      "step": 2067
    },
    {
      "epoch": 0.39891975308641975,
      "grad_norm": 1.5946540832519531,
      "learning_rate": 0.0005,
      "loss": 2.9007277488708496,
      "step": 2068
    },
    {
      "epoch": 0.39911265432098764,
      "grad_norm": 1.141946792602539,
      "learning_rate": 0.0005,
      "loss": 2.9216246604919434,
      "step": 2069
    },
    {
      "epoch": 0.3993055555555556,
      "grad_norm": 1.0192018747329712,
      "learning_rate": 0.0005,
      "loss": 2.8156728744506836,
      "step": 2070
    },
    {
      "epoch": 0.39949845679012347,
      "grad_norm": 0.9978347420692444,
      "learning_rate": 0.0005,
      "loss": 2.748671293258667,
      "step": 2071
    },
    {
      "epoch": 0.39969135802469136,
      "grad_norm": 1.139244556427002,
      "learning_rate": 0.0005,
      "loss": 2.5137276649475098,
      "step": 2072
    },
    {
      "epoch": 0.39988425925925924,
      "grad_norm": 1.06141197681427,
      "learning_rate": 0.0005,
      "loss": 2.9661407470703125,
      "step": 2073
    },
    {
      "epoch": 0.40007716049382713,
      "grad_norm": 2.5499777793884277,
      "learning_rate": 0.0005,
      "loss": 2.8620495796203613,
      "step": 2074
    },
    {
      "epoch": 0.4002700617283951,
      "grad_norm": 1.4293140172958374,
      "learning_rate": 0.0005,
      "loss": 3.086203098297119,
      "step": 2075
    },
    {
      "epoch": 0.40046296296296297,
      "grad_norm": 1.1978298425674438,
      "learning_rate": 0.0005,
      "loss": 2.840430736541748,
      "step": 2076
    },
    {
      "epoch": 0.40065586419753085,
      "grad_norm": 1.4399304389953613,
      "learning_rate": 0.0005,
      "loss": 2.9476048946380615,
      "step": 2077
    },
    {
      "epoch": 0.40084876543209874,
      "grad_norm": 0.9549009203910828,
      "learning_rate": 0.0005,
      "loss": 2.9081838130950928,
      "step": 2078
    },
    {
      "epoch": 0.4010416666666667,
      "grad_norm": 1.0318708419799805,
      "learning_rate": 0.0005,
      "loss": 2.779024600982666,
      "step": 2079
    },
    {
      "epoch": 0.4012345679012346,
      "grad_norm": 1.1916841268539429,
      "learning_rate": 0.0005,
      "loss": 3.0023207664489746,
      "step": 2080
    },
    {
      "epoch": 0.40142746913580246,
      "grad_norm": 1.0943963527679443,
      "learning_rate": 0.0005,
      "loss": 2.7344963550567627,
      "step": 2081
    },
    {
      "epoch": 0.40162037037037035,
      "grad_norm": 1.0541703701019287,
      "learning_rate": 0.0005,
      "loss": 2.76023268699646,
      "step": 2082
    },
    {
      "epoch": 0.4018132716049383,
      "grad_norm": 1.127920389175415,
      "learning_rate": 0.0005,
      "loss": 3.1067986488342285,
      "step": 2083
    },
    {
      "epoch": 0.4020061728395062,
      "grad_norm": 1.276862621307373,
      "learning_rate": 0.0005,
      "loss": 2.6758832931518555,
      "step": 2084
    },
    {
      "epoch": 0.40219907407407407,
      "grad_norm": 1.2502912282943726,
      "learning_rate": 0.0005,
      "loss": 2.899796962738037,
      "step": 2085
    },
    {
      "epoch": 0.40239197530864196,
      "grad_norm": 1.3406175374984741,
      "learning_rate": 0.0005,
      "loss": 2.8879456520080566,
      "step": 2086
    },
    {
      "epoch": 0.4025848765432099,
      "grad_norm": 1.0207059383392334,
      "learning_rate": 0.0005,
      "loss": 2.8293936252593994,
      "step": 2087
    },
    {
      "epoch": 0.4027777777777778,
      "grad_norm": 1.2991563081741333,
      "learning_rate": 0.0005,
      "loss": 2.8218955993652344,
      "step": 2088
    },
    {
      "epoch": 0.4029706790123457,
      "grad_norm": 0.8241744041442871,
      "learning_rate": 0.0005,
      "loss": 2.908201217651367,
      "step": 2089
    },
    {
      "epoch": 0.40316358024691357,
      "grad_norm": 1.11029052734375,
      "learning_rate": 0.0005,
      "loss": 2.9772398471832275,
      "step": 2090
    },
    {
      "epoch": 0.40335648148148145,
      "grad_norm": 0.8666333556175232,
      "learning_rate": 0.0005,
      "loss": 2.762544870376587,
      "step": 2091
    },
    {
      "epoch": 0.4035493827160494,
      "grad_norm": 1.0193818807601929,
      "learning_rate": 0.0005,
      "loss": 2.8334646224975586,
      "step": 2092
    },
    {
      "epoch": 0.4037422839506173,
      "grad_norm": 1.3502488136291504,
      "learning_rate": 0.0005,
      "loss": 2.7408699989318848,
      "step": 2093
    },
    {
      "epoch": 0.4039351851851852,
      "grad_norm": 1.0335867404937744,
      "learning_rate": 0.0005,
      "loss": 2.8490102291107178,
      "step": 2094
    },
    {
      "epoch": 0.40412808641975306,
      "grad_norm": 1.1847959756851196,
      "learning_rate": 0.0005,
      "loss": 2.9781107902526855,
      "step": 2095
    },
    {
      "epoch": 0.404320987654321,
      "grad_norm": 0.9525595903396606,
      "learning_rate": 0.0005,
      "loss": 2.609663963317871,
      "step": 2096
    },
    {
      "epoch": 0.4045138888888889,
      "grad_norm": 1.1576045751571655,
      "learning_rate": 0.0005,
      "loss": 2.861396074295044,
      "step": 2097
    },
    {
      "epoch": 0.4047067901234568,
      "grad_norm": 1.058901071548462,
      "learning_rate": 0.0005,
      "loss": 2.774160385131836,
      "step": 2098
    },
    {
      "epoch": 0.40489969135802467,
      "grad_norm": 1.722320318222046,
      "learning_rate": 0.0005,
      "loss": 2.5326569080352783,
      "step": 2099
    },
    {
      "epoch": 0.4050925925925926,
      "grad_norm": 1.1872422695159912,
      "learning_rate": 0.0005,
      "loss": 2.9172701835632324,
      "step": 2100
    },
    {
      "epoch": 0.4052854938271605,
      "grad_norm": 1.0576393604278564,
      "learning_rate": 0.0005,
      "loss": 2.7122273445129395,
      "step": 2101
    },
    {
      "epoch": 0.4054783950617284,
      "grad_norm": 1.1872286796569824,
      "learning_rate": 0.0005,
      "loss": 2.8663992881774902,
      "step": 2102
    },
    {
      "epoch": 0.4056712962962963,
      "grad_norm": 0.9206895232200623,
      "learning_rate": 0.0005,
      "loss": 2.6467926502227783,
      "step": 2103
    },
    {
      "epoch": 0.4058641975308642,
      "grad_norm": 0.9279674887657166,
      "learning_rate": 0.0005,
      "loss": 3.023409843444824,
      "step": 2104
    },
    {
      "epoch": 0.4060570987654321,
      "grad_norm": 1.2522052526474,
      "learning_rate": 0.0005,
      "loss": 2.8303046226501465,
      "step": 2105
    },
    {
      "epoch": 0.40625,
      "grad_norm": 0.9031372666358948,
      "learning_rate": 0.0005,
      "loss": 2.799515724182129,
      "step": 2106
    },
    {
      "epoch": 0.4064429012345679,
      "grad_norm": 1.173166275024414,
      "learning_rate": 0.0005,
      "loss": 3.0615415573120117,
      "step": 2107
    },
    {
      "epoch": 0.4066358024691358,
      "grad_norm": 1.0542113780975342,
      "learning_rate": 0.0005,
      "loss": 2.8001842498779297,
      "step": 2108
    },
    {
      "epoch": 0.4068287037037037,
      "grad_norm": 1.059507966041565,
      "learning_rate": 0.0005,
      "loss": 2.6748149394989014,
      "step": 2109
    },
    {
      "epoch": 0.4070216049382716,
      "grad_norm": 1.4354017972946167,
      "learning_rate": 0.0005,
      "loss": 2.533303737640381,
      "step": 2110
    },
    {
      "epoch": 0.4072145061728395,
      "grad_norm": 1.4833383560180664,
      "learning_rate": 0.0005,
      "loss": 3.265864849090576,
      "step": 2111
    },
    {
      "epoch": 0.4074074074074074,
      "grad_norm": 1.2930651903152466,
      "learning_rate": 0.0005,
      "loss": 3.0414414405822754,
      "step": 2112
    },
    {
      "epoch": 0.40760030864197533,
      "grad_norm": 1.1749860048294067,
      "learning_rate": 0.0005,
      "loss": 2.7129595279693604,
      "step": 2113
    },
    {
      "epoch": 0.4077932098765432,
      "grad_norm": 1.1219079494476318,
      "learning_rate": 0.0005,
      "loss": 2.9685449600219727,
      "step": 2114
    },
    {
      "epoch": 0.4079861111111111,
      "grad_norm": 0.9581892490386963,
      "learning_rate": 0.0005,
      "loss": 2.877254009246826,
      "step": 2115
    },
    {
      "epoch": 0.408179012345679,
      "grad_norm": 0.9871929883956909,
      "learning_rate": 0.0005,
      "loss": 2.6424121856689453,
      "step": 2116
    },
    {
      "epoch": 0.40837191358024694,
      "grad_norm": 0.968907356262207,
      "learning_rate": 0.0005,
      "loss": 3.01027250289917,
      "step": 2117
    },
    {
      "epoch": 0.4085648148148148,
      "grad_norm": 1.063320517539978,
      "learning_rate": 0.0005,
      "loss": 2.8241193294525146,
      "step": 2118
    },
    {
      "epoch": 0.4087577160493827,
      "grad_norm": 1.309664011001587,
      "learning_rate": 0.0005,
      "loss": 2.8962931632995605,
      "step": 2119
    },
    {
      "epoch": 0.4089506172839506,
      "grad_norm": 1.2340188026428223,
      "learning_rate": 0.0005,
      "loss": 2.993010997772217,
      "step": 2120
    },
    {
      "epoch": 0.40914351851851855,
      "grad_norm": 1.3378002643585205,
      "learning_rate": 0.0005,
      "loss": 2.9596619606018066,
      "step": 2121
    },
    {
      "epoch": 0.40933641975308643,
      "grad_norm": 0.9138287901878357,
      "learning_rate": 0.0005,
      "loss": 2.7183189392089844,
      "step": 2122
    },
    {
      "epoch": 0.4095293209876543,
      "grad_norm": 1.03994619846344,
      "learning_rate": 0.0005,
      "loss": 2.427436351776123,
      "step": 2123
    },
    {
      "epoch": 0.4097222222222222,
      "grad_norm": 0.9887388944625854,
      "learning_rate": 0.0005,
      "loss": 3.223674774169922,
      "step": 2124
    },
    {
      "epoch": 0.4099151234567901,
      "grad_norm": 1.0377761125564575,
      "learning_rate": 0.0005,
      "loss": 2.809298038482666,
      "step": 2125
    },
    {
      "epoch": 0.41010802469135804,
      "grad_norm": 1.0720058679580688,
      "learning_rate": 0.0005,
      "loss": 2.8665878772735596,
      "step": 2126
    },
    {
      "epoch": 0.41030092592592593,
      "grad_norm": 1.176840901374817,
      "learning_rate": 0.0005,
      "loss": 2.995917558670044,
      "step": 2127
    },
    {
      "epoch": 0.4104938271604938,
      "grad_norm": 1.480289340019226,
      "learning_rate": 0.0005,
      "loss": 2.8813061714172363,
      "step": 2128
    },
    {
      "epoch": 0.4106867283950617,
      "grad_norm": 1.1312328577041626,
      "learning_rate": 0.0005,
      "loss": 2.846546173095703,
      "step": 2129
    },
    {
      "epoch": 0.41087962962962965,
      "grad_norm": 0.9022986888885498,
      "learning_rate": 0.0005,
      "loss": 2.7842938899993896,
      "step": 2130
    },
    {
      "epoch": 0.41107253086419754,
      "grad_norm": 1.1534539461135864,
      "learning_rate": 0.0005,
      "loss": 3.1252660751342773,
      "step": 2131
    },
    {
      "epoch": 0.4112654320987654,
      "grad_norm": 1.215195894241333,
      "learning_rate": 0.0005,
      "loss": 2.7979187965393066,
      "step": 2132
    },
    {
      "epoch": 0.4114583333333333,
      "grad_norm": 0.9468469023704529,
      "learning_rate": 0.0005,
      "loss": 2.7190566062927246,
      "step": 2133
    },
    {
      "epoch": 0.41165123456790126,
      "grad_norm": 1.1993099451065063,
      "learning_rate": 0.0005,
      "loss": 2.6864798069000244,
      "step": 2134
    },
    {
      "epoch": 0.41184413580246915,
      "grad_norm": 1.233747959136963,
      "learning_rate": 0.0005,
      "loss": 2.8788113594055176,
      "step": 2135
    },
    {
      "epoch": 0.41203703703703703,
      "grad_norm": 0.9243922829627991,
      "learning_rate": 0.0005,
      "loss": 2.9456305503845215,
      "step": 2136
    },
    {
      "epoch": 0.4122299382716049,
      "grad_norm": 1.1029376983642578,
      "learning_rate": 0.0005,
      "loss": 2.610942840576172,
      "step": 2137
    },
    {
      "epoch": 0.41242283950617287,
      "grad_norm": 1.6234171390533447,
      "learning_rate": 0.0005,
      "loss": 3.135486602783203,
      "step": 2138
    },
    {
      "epoch": 0.41261574074074076,
      "grad_norm": 1.033171534538269,
      "learning_rate": 0.0005,
      "loss": 2.754283905029297,
      "step": 2139
    },
    {
      "epoch": 0.41280864197530864,
      "grad_norm": 1.4820839166641235,
      "learning_rate": 0.0005,
      "loss": 2.755094528198242,
      "step": 2140
    },
    {
      "epoch": 0.41300154320987653,
      "grad_norm": 1.1973789930343628,
      "learning_rate": 0.0005,
      "loss": 3.0207133293151855,
      "step": 2141
    },
    {
      "epoch": 0.4131944444444444,
      "grad_norm": 0.9584393501281738,
      "learning_rate": 0.0005,
      "loss": 2.3866987228393555,
      "step": 2142
    },
    {
      "epoch": 0.41338734567901236,
      "grad_norm": 2.647303581237793,
      "learning_rate": 0.0005,
      "loss": 2.6933579444885254,
      "step": 2143
    },
    {
      "epoch": 0.41358024691358025,
      "grad_norm": 1.4496710300445557,
      "learning_rate": 0.0005,
      "loss": 2.8406105041503906,
      "step": 2144
    },
    {
      "epoch": 0.41377314814814814,
      "grad_norm": 0.986790657043457,
      "learning_rate": 0.0005,
      "loss": 2.599628448486328,
      "step": 2145
    },
    {
      "epoch": 0.41396604938271603,
      "grad_norm": 1.1733201742172241,
      "learning_rate": 0.0005,
      "loss": 2.6556029319763184,
      "step": 2146
    },
    {
      "epoch": 0.41415895061728397,
      "grad_norm": 1.1317232847213745,
      "learning_rate": 0.0005,
      "loss": 2.6065797805786133,
      "step": 2147
    },
    {
      "epoch": 0.41435185185185186,
      "grad_norm": 0.9824839234352112,
      "learning_rate": 0.0005,
      "loss": 2.782144784927368,
      "step": 2148
    },
    {
      "epoch": 0.41454475308641975,
      "grad_norm": 1.032151699066162,
      "learning_rate": 0.0005,
      "loss": 2.8406317234039307,
      "step": 2149
    },
    {
      "epoch": 0.41473765432098764,
      "grad_norm": 0.9709839820861816,
      "learning_rate": 0.0005,
      "loss": 2.9054770469665527,
      "step": 2150
    },
    {
      "epoch": 0.4149305555555556,
      "grad_norm": 1.8968461751937866,
      "learning_rate": 0.0005,
      "loss": 2.7613794803619385,
      "step": 2151
    },
    {
      "epoch": 0.41512345679012347,
      "grad_norm": 1.4253686666488647,
      "learning_rate": 0.0005,
      "loss": 2.9557104110717773,
      "step": 2152
    },
    {
      "epoch": 0.41531635802469136,
      "grad_norm": 1.2594256401062012,
      "learning_rate": 0.0005,
      "loss": 3.6139602661132812,
      "step": 2153
    },
    {
      "epoch": 0.41550925925925924,
      "grad_norm": 0.8922407627105713,
      "learning_rate": 0.0005,
      "loss": 2.6291251182556152,
      "step": 2154
    },
    {
      "epoch": 0.41570216049382713,
      "grad_norm": 1.6898273229599,
      "learning_rate": 0.0005,
      "loss": 2.879197835922241,
      "step": 2155
    },
    {
      "epoch": 0.4158950617283951,
      "grad_norm": 0.9547452330589294,
      "learning_rate": 0.0005,
      "loss": 2.9753189086914062,
      "step": 2156
    },
    {
      "epoch": 0.41608796296296297,
      "grad_norm": 1.0891813039779663,
      "learning_rate": 0.0005,
      "loss": 2.963658332824707,
      "step": 2157
    },
    {
      "epoch": 0.41628086419753085,
      "grad_norm": 1.3470855951309204,
      "learning_rate": 0.0005,
      "loss": 2.6223578453063965,
      "step": 2158
    },
    {
      "epoch": 0.41647376543209874,
      "grad_norm": 0.9538573026657104,
      "learning_rate": 0.0005,
      "loss": 2.724350690841675,
      "step": 2159
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 1.2119953632354736,
      "learning_rate": 0.0005,
      "loss": 2.988276958465576,
      "step": 2160
    },
    {
      "epoch": 0.4168595679012346,
      "grad_norm": 0.9441961646080017,
      "learning_rate": 0.0005,
      "loss": 3.0658833980560303,
      "step": 2161
    },
    {
      "epoch": 0.41705246913580246,
      "grad_norm": 0.8082683682441711,
      "learning_rate": 0.0005,
      "loss": 2.4770445823669434,
      "step": 2162
    },
    {
      "epoch": 0.41724537037037035,
      "grad_norm": 0.9425111413002014,
      "learning_rate": 0.0005,
      "loss": 2.913966655731201,
      "step": 2163
    },
    {
      "epoch": 0.4174382716049383,
      "grad_norm": 0.9855911135673523,
      "learning_rate": 0.0005,
      "loss": 2.6918656826019287,
      "step": 2164
    },
    {
      "epoch": 0.4176311728395062,
      "grad_norm": 1.1337859630584717,
      "learning_rate": 0.0005,
      "loss": 2.8864259719848633,
      "step": 2165
    },
    {
      "epoch": 0.41782407407407407,
      "grad_norm": 0.9661660194396973,
      "learning_rate": 0.0005,
      "loss": 2.6770987510681152,
      "step": 2166
    },
    {
      "epoch": 0.41801697530864196,
      "grad_norm": 1.1800386905670166,
      "learning_rate": 0.0005,
      "loss": 2.551434278488159,
      "step": 2167
    },
    {
      "epoch": 0.4182098765432099,
      "grad_norm": 1.7638169527053833,
      "learning_rate": 0.0005,
      "loss": 2.985585927963257,
      "step": 2168
    },
    {
      "epoch": 0.4184027777777778,
      "grad_norm": 1.0805795192718506,
      "learning_rate": 0.0005,
      "loss": 2.663240909576416,
      "step": 2169
    },
    {
      "epoch": 0.4185956790123457,
      "grad_norm": 1.127482295036316,
      "learning_rate": 0.0005,
      "loss": 2.476938486099243,
      "step": 2170
    },
    {
      "epoch": 0.41878858024691357,
      "grad_norm": 1.2575420141220093,
      "learning_rate": 0.0005,
      "loss": 2.849426746368408,
      "step": 2171
    },
    {
      "epoch": 0.41898148148148145,
      "grad_norm": 1.3482881784439087,
      "learning_rate": 0.0005,
      "loss": 2.802795886993408,
      "step": 2172
    },
    {
      "epoch": 0.4191743827160494,
      "grad_norm": 0.9411700963973999,
      "learning_rate": 0.0005,
      "loss": 2.8278768062591553,
      "step": 2173
    },
    {
      "epoch": 0.4193672839506173,
      "grad_norm": 0.9376500248908997,
      "learning_rate": 0.0005,
      "loss": 3.1084396839141846,
      "step": 2174
    },
    {
      "epoch": 0.4195601851851852,
      "grad_norm": 1.04513418674469,
      "learning_rate": 0.0005,
      "loss": 2.5384116172790527,
      "step": 2175
    },
    {
      "epoch": 0.41975308641975306,
      "grad_norm": 1.1053839921951294,
      "learning_rate": 0.0005,
      "loss": 2.5424716472625732,
      "step": 2176
    },
    {
      "epoch": 0.419945987654321,
      "grad_norm": 1.0812655687332153,
      "learning_rate": 0.0005,
      "loss": 2.7199084758758545,
      "step": 2177
    },
    {
      "epoch": 0.4201388888888889,
      "grad_norm": 1.0101956129074097,
      "learning_rate": 0.0005,
      "loss": 2.7772130966186523,
      "step": 2178
    },
    {
      "epoch": 0.4203317901234568,
      "grad_norm": 1.0949369668960571,
      "learning_rate": 0.0005,
      "loss": 2.954359531402588,
      "step": 2179
    },
    {
      "epoch": 0.42052469135802467,
      "grad_norm": 1.089159369468689,
      "learning_rate": 0.0005,
      "loss": 2.5883591175079346,
      "step": 2180
    },
    {
      "epoch": 0.4207175925925926,
      "grad_norm": 0.9612703323364258,
      "learning_rate": 0.0005,
      "loss": 3.143228054046631,
      "step": 2181
    },
    {
      "epoch": 0.4209104938271605,
      "grad_norm": 1.423957347869873,
      "learning_rate": 0.0005,
      "loss": 2.607935905456543,
      "step": 2182
    },
    {
      "epoch": 0.4211033950617284,
      "grad_norm": 1.0739914178848267,
      "learning_rate": 0.0005,
      "loss": 2.689105987548828,
      "step": 2183
    },
    {
      "epoch": 0.4212962962962963,
      "grad_norm": 0.8885409235954285,
      "learning_rate": 0.0005,
      "loss": 3.00040864944458,
      "step": 2184
    },
    {
      "epoch": 0.4214891975308642,
      "grad_norm": 1.1927995681762695,
      "learning_rate": 0.0005,
      "loss": 2.8787178993225098,
      "step": 2185
    },
    {
      "epoch": 0.4216820987654321,
      "grad_norm": 1.0411367416381836,
      "learning_rate": 0.0005,
      "loss": 2.6448140144348145,
      "step": 2186
    },
    {
      "epoch": 0.421875,
      "grad_norm": 1.1670769453048706,
      "learning_rate": 0.0005,
      "loss": 3.0763444900512695,
      "step": 2187
    },
    {
      "epoch": 0.4220679012345679,
      "grad_norm": 1.9468741416931152,
      "learning_rate": 0.0005,
      "loss": 2.633678913116455,
      "step": 2188
    },
    {
      "epoch": 0.4222608024691358,
      "grad_norm": 1.010142207145691,
      "learning_rate": 0.0005,
      "loss": 2.831838607788086,
      "step": 2189
    },
    {
      "epoch": 0.4224537037037037,
      "grad_norm": 1.4282170534133911,
      "learning_rate": 0.0005,
      "loss": 3.1866347789764404,
      "step": 2190
    },
    {
      "epoch": 0.4226466049382716,
      "grad_norm": 1.0687648057937622,
      "learning_rate": 0.0005,
      "loss": 2.75203275680542,
      "step": 2191
    },
    {
      "epoch": 0.4228395061728395,
      "grad_norm": 1.3350404500961304,
      "learning_rate": 0.0005,
      "loss": 3.0613832473754883,
      "step": 2192
    },
    {
      "epoch": 0.4230324074074074,
      "grad_norm": 1.0426363945007324,
      "learning_rate": 0.0005,
      "loss": 2.685605049133301,
      "step": 2193
    },
    {
      "epoch": 0.42322530864197533,
      "grad_norm": 1.4113754034042358,
      "learning_rate": 0.0005,
      "loss": 3.1666922569274902,
      "step": 2194
    },
    {
      "epoch": 0.4234182098765432,
      "grad_norm": 1.0965588092803955,
      "learning_rate": 0.0005,
      "loss": 2.8262271881103516,
      "step": 2195
    },
    {
      "epoch": 0.4236111111111111,
      "grad_norm": 1.8773521184921265,
      "learning_rate": 0.0005,
      "loss": 3.147054433822632,
      "step": 2196
    },
    {
      "epoch": 0.423804012345679,
      "grad_norm": 0.983902633190155,
      "learning_rate": 0.0005,
      "loss": 2.68467378616333,
      "step": 2197
    },
    {
      "epoch": 0.42399691358024694,
      "grad_norm": 2.509047746658325,
      "learning_rate": 0.0005,
      "loss": 2.6209702491760254,
      "step": 2198
    },
    {
      "epoch": 0.4241898148148148,
      "grad_norm": 1.0194460153579712,
      "learning_rate": 0.0005,
      "loss": 2.8900270462036133,
      "step": 2199
    },
    {
      "epoch": 0.4243827160493827,
      "grad_norm": 1.0130671262741089,
      "learning_rate": 0.0005,
      "loss": 3.114311695098877,
      "step": 2200
    },
    {
      "epoch": 0.4245756172839506,
      "grad_norm": 0.9376267194747925,
      "learning_rate": 0.0005,
      "loss": 2.911271095275879,
      "step": 2201
    },
    {
      "epoch": 0.42476851851851855,
      "grad_norm": 0.9157959222793579,
      "learning_rate": 0.0005,
      "loss": 2.6068713665008545,
      "step": 2202
    },
    {
      "epoch": 0.42496141975308643,
      "grad_norm": 1.0461159944534302,
      "learning_rate": 0.0005,
      "loss": 3.016134738922119,
      "step": 2203
    },
    {
      "epoch": 0.4251543209876543,
      "grad_norm": 2.585955858230591,
      "learning_rate": 0.0005,
      "loss": 2.7239840030670166,
      "step": 2204
    },
    {
      "epoch": 0.4253472222222222,
      "grad_norm": 1.1668611764907837,
      "learning_rate": 0.0005,
      "loss": 2.7095232009887695,
      "step": 2205
    },
    {
      "epoch": 0.4255401234567901,
      "grad_norm": 1.1067570447921753,
      "learning_rate": 0.0005,
      "loss": 2.7675893306732178,
      "step": 2206
    },
    {
      "epoch": 0.42573302469135804,
      "grad_norm": 1.7899507284164429,
      "learning_rate": 0.0005,
      "loss": 2.690746545791626,
      "step": 2207
    },
    {
      "epoch": 0.42592592592592593,
      "grad_norm": 1.2161078453063965,
      "learning_rate": 0.0005,
      "loss": 2.2753849029541016,
      "step": 2208
    },
    {
      "epoch": 0.4261188271604938,
      "grad_norm": 1.2068430185317993,
      "learning_rate": 0.0005,
      "loss": 2.5654048919677734,
      "step": 2209
    },
    {
      "epoch": 0.4263117283950617,
      "grad_norm": 1.0463885068893433,
      "learning_rate": 0.0005,
      "loss": 3.1424667835235596,
      "step": 2210
    },
    {
      "epoch": 0.42650462962962965,
      "grad_norm": 1.1206904649734497,
      "learning_rate": 0.0005,
      "loss": 2.6537668704986572,
      "step": 2211
    },
    {
      "epoch": 0.42669753086419754,
      "grad_norm": 1.0341376066207886,
      "learning_rate": 0.0005,
      "loss": 2.7164218425750732,
      "step": 2212
    },
    {
      "epoch": 0.4268904320987654,
      "grad_norm": 1.0313005447387695,
      "learning_rate": 0.0005,
      "loss": 2.944563865661621,
      "step": 2213
    },
    {
      "epoch": 0.4270833333333333,
      "grad_norm": 1.400488018989563,
      "learning_rate": 0.0005,
      "loss": 3.056245803833008,
      "step": 2214
    },
    {
      "epoch": 0.42727623456790126,
      "grad_norm": 0.9620441794395447,
      "learning_rate": 0.0005,
      "loss": 2.496034622192383,
      "step": 2215
    },
    {
      "epoch": 0.42746913580246915,
      "grad_norm": 1.5832841396331787,
      "learning_rate": 0.0005,
      "loss": 2.9918222427368164,
      "step": 2216
    },
    {
      "epoch": 0.42766203703703703,
      "grad_norm": 1.0982158184051514,
      "learning_rate": 0.0005,
      "loss": 2.7811648845672607,
      "step": 2217
    },
    {
      "epoch": 0.4278549382716049,
      "grad_norm": 1.1333101987838745,
      "learning_rate": 0.0005,
      "loss": 2.8092498779296875,
      "step": 2218
    },
    {
      "epoch": 0.42804783950617287,
      "grad_norm": 0.9747150540351868,
      "learning_rate": 0.0005,
      "loss": 2.7376928329467773,
      "step": 2219
    },
    {
      "epoch": 0.42824074074074076,
      "grad_norm": 1.2374640703201294,
      "learning_rate": 0.0005,
      "loss": 2.7340524196624756,
      "step": 2220
    },
    {
      "epoch": 0.42843364197530864,
      "grad_norm": 1.0566563606262207,
      "learning_rate": 0.0005,
      "loss": 2.8361754417419434,
      "step": 2221
    },
    {
      "epoch": 0.42862654320987653,
      "grad_norm": 1.1327834129333496,
      "learning_rate": 0.0005,
      "loss": 2.8029072284698486,
      "step": 2222
    },
    {
      "epoch": 0.4288194444444444,
      "grad_norm": 0.9526020884513855,
      "learning_rate": 0.0005,
      "loss": 2.927788257598877,
      "step": 2223
    },
    {
      "epoch": 0.42901234567901236,
      "grad_norm": 0.9470782279968262,
      "learning_rate": 0.0005,
      "loss": 2.8421695232391357,
      "step": 2224
    },
    {
      "epoch": 0.42920524691358025,
      "grad_norm": 3.967510461807251,
      "learning_rate": 0.0005,
      "loss": 2.7813072204589844,
      "step": 2225
    },
    {
      "epoch": 0.42939814814814814,
      "grad_norm": 1.3182132244110107,
      "learning_rate": 0.0005,
      "loss": 2.7495341300964355,
      "step": 2226
    },
    {
      "epoch": 0.42959104938271603,
      "grad_norm": 1.0281814336776733,
      "learning_rate": 0.0005,
      "loss": 2.783102512359619,
      "step": 2227
    },
    {
      "epoch": 0.42978395061728397,
      "grad_norm": 1.0481681823730469,
      "learning_rate": 0.0005,
      "loss": 2.9455785751342773,
      "step": 2228
    },
    {
      "epoch": 0.42997685185185186,
      "grad_norm": 1.1064680814743042,
      "learning_rate": 0.0005,
      "loss": 2.8878395557403564,
      "step": 2229
    },
    {
      "epoch": 0.43016975308641975,
      "grad_norm": 1.0495247840881348,
      "learning_rate": 0.0005,
      "loss": 2.798945426940918,
      "step": 2230
    },
    {
      "epoch": 0.43036265432098764,
      "grad_norm": 1.06326162815094,
      "learning_rate": 0.0005,
      "loss": 2.9660489559173584,
      "step": 2231
    },
    {
      "epoch": 0.4305555555555556,
      "grad_norm": 0.8662716150283813,
      "learning_rate": 0.0005,
      "loss": 2.670226573944092,
      "step": 2232
    },
    {
      "epoch": 0.43074845679012347,
      "grad_norm": 0.8796610832214355,
      "learning_rate": 0.0005,
      "loss": 2.6321349143981934,
      "step": 2233
    },
    {
      "epoch": 0.43094135802469136,
      "grad_norm": 1.116765022277832,
      "learning_rate": 0.0005,
      "loss": 3.0810110569000244,
      "step": 2234
    },
    {
      "epoch": 0.43113425925925924,
      "grad_norm": 1.167325496673584,
      "learning_rate": 0.0005,
      "loss": 2.6469452381134033,
      "step": 2235
    },
    {
      "epoch": 0.43132716049382713,
      "grad_norm": 1.0021169185638428,
      "learning_rate": 0.0005,
      "loss": 3.021482467651367,
      "step": 2236
    },
    {
      "epoch": 0.4315200617283951,
      "grad_norm": 0.9675446152687073,
      "learning_rate": 0.0005,
      "loss": 2.773970127105713,
      "step": 2237
    },
    {
      "epoch": 0.43171296296296297,
      "grad_norm": 0.9262160658836365,
      "learning_rate": 0.0005,
      "loss": 2.732689619064331,
      "step": 2238
    },
    {
      "epoch": 0.43190586419753085,
      "grad_norm": 0.9718573093414307,
      "learning_rate": 0.0005,
      "loss": 2.5526022911071777,
      "step": 2239
    },
    {
      "epoch": 0.43209876543209874,
      "grad_norm": 0.9646399617195129,
      "learning_rate": 0.0005,
      "loss": 3.2009668350219727,
      "step": 2240
    },
    {
      "epoch": 0.4322916666666667,
      "grad_norm": 1.103633165359497,
      "learning_rate": 0.0005,
      "loss": 2.7043421268463135,
      "step": 2241
    },
    {
      "epoch": 0.4324845679012346,
      "grad_norm": 1.1819045543670654,
      "learning_rate": 0.0005,
      "loss": 2.7955362796783447,
      "step": 2242
    },
    {
      "epoch": 0.43267746913580246,
      "grad_norm": 1.1135205030441284,
      "learning_rate": 0.0005,
      "loss": 2.3191041946411133,
      "step": 2243
    },
    {
      "epoch": 0.43287037037037035,
      "grad_norm": 1.012429118156433,
      "learning_rate": 0.0005,
      "loss": 2.7456231117248535,
      "step": 2244
    },
    {
      "epoch": 0.4330632716049383,
      "grad_norm": 0.9280381202697754,
      "learning_rate": 0.0005,
      "loss": 2.472320795059204,
      "step": 2245
    },
    {
      "epoch": 0.4332561728395062,
      "grad_norm": 1.102732539176941,
      "learning_rate": 0.0005,
      "loss": 2.66825795173645,
      "step": 2246
    },
    {
      "epoch": 0.43344907407407407,
      "grad_norm": 0.8840869665145874,
      "learning_rate": 0.0005,
      "loss": 2.9902281761169434,
      "step": 2247
    },
    {
      "epoch": 0.43364197530864196,
      "grad_norm": 0.892898678779602,
      "learning_rate": 0.0005,
      "loss": 2.953120708465576,
      "step": 2248
    },
    {
      "epoch": 0.4338348765432099,
      "grad_norm": 1.037732720375061,
      "learning_rate": 0.0005,
      "loss": 2.706658363342285,
      "step": 2249
    },
    {
      "epoch": 0.4340277777777778,
      "grad_norm": 1.7142225503921509,
      "learning_rate": 0.0005,
      "loss": 2.8919639587402344,
      "step": 2250
    },
    {
      "epoch": 0.4342206790123457,
      "grad_norm": 1.0799462795257568,
      "learning_rate": 0.0005,
      "loss": 3.000403881072998,
      "step": 2251
    },
    {
      "epoch": 0.43441358024691357,
      "grad_norm": 0.9110816121101379,
      "learning_rate": 0.0005,
      "loss": 2.894611358642578,
      "step": 2252
    },
    {
      "epoch": 0.43460648148148145,
      "grad_norm": 0.9018135070800781,
      "learning_rate": 0.0005,
      "loss": 2.966155529022217,
      "step": 2253
    },
    {
      "epoch": 0.4347993827160494,
      "grad_norm": 1.304349660873413,
      "learning_rate": 0.0005,
      "loss": 2.553434371948242,
      "step": 2254
    },
    {
      "epoch": 0.4349922839506173,
      "grad_norm": 0.827323317527771,
      "learning_rate": 0.0005,
      "loss": 2.6515026092529297,
      "step": 2255
    },
    {
      "epoch": 0.4351851851851852,
      "grad_norm": 1.0647169351577759,
      "learning_rate": 0.0005,
      "loss": 2.9109535217285156,
      "step": 2256
    },
    {
      "epoch": 0.43537808641975306,
      "grad_norm": 0.8282851576805115,
      "learning_rate": 0.0005,
      "loss": 2.3694727420806885,
      "step": 2257
    },
    {
      "epoch": 0.435570987654321,
      "grad_norm": 0.8782387971878052,
      "learning_rate": 0.0005,
      "loss": 2.9779272079467773,
      "step": 2258
    },
    {
      "epoch": 0.4357638888888889,
      "grad_norm": 1.0292344093322754,
      "learning_rate": 0.0005,
      "loss": 3.099388360977173,
      "step": 2259
    },
    {
      "epoch": 0.4359567901234568,
      "grad_norm": 0.9855065941810608,
      "learning_rate": 0.0005,
      "loss": 2.8709568977355957,
      "step": 2260
    },
    {
      "epoch": 0.43614969135802467,
      "grad_norm": 1.332218885421753,
      "learning_rate": 0.0005,
      "loss": 2.7874350547790527,
      "step": 2261
    },
    {
      "epoch": 0.4363425925925926,
      "grad_norm": 1.0578876733779907,
      "learning_rate": 0.0005,
      "loss": 2.6426753997802734,
      "step": 2262
    },
    {
      "epoch": 0.4365354938271605,
      "grad_norm": 0.8797797560691833,
      "learning_rate": 0.0005,
      "loss": 2.8756322860717773,
      "step": 2263
    },
    {
      "epoch": 0.4367283950617284,
      "grad_norm": 1.0011155605316162,
      "learning_rate": 0.0005,
      "loss": 2.806389808654785,
      "step": 2264
    },
    {
      "epoch": 0.4369212962962963,
      "grad_norm": 1.0449090003967285,
      "learning_rate": 0.0005,
      "loss": 2.587123394012451,
      "step": 2265
    },
    {
      "epoch": 0.4371141975308642,
      "grad_norm": 0.8480119705200195,
      "learning_rate": 0.0005,
      "loss": 2.7110941410064697,
      "step": 2266
    },
    {
      "epoch": 0.4373070987654321,
      "grad_norm": 1.2258059978485107,
      "learning_rate": 0.0005,
      "loss": 2.800673007965088,
      "step": 2267
    },
    {
      "epoch": 0.4375,
      "grad_norm": 1.0854674577713013,
      "learning_rate": 0.0005,
      "loss": 2.6302847862243652,
      "step": 2268
    },
    {
      "epoch": 0.4376929012345679,
      "grad_norm": 0.9183703064918518,
      "learning_rate": 0.0005,
      "loss": 2.824472427368164,
      "step": 2269
    },
    {
      "epoch": 0.4378858024691358,
      "grad_norm": 0.9229046106338501,
      "learning_rate": 0.0005,
      "loss": 2.6423237323760986,
      "step": 2270
    },
    {
      "epoch": 0.4380787037037037,
      "grad_norm": 0.9770431518554688,
      "learning_rate": 0.0005,
      "loss": 2.629876136779785,
      "step": 2271
    },
    {
      "epoch": 0.4382716049382716,
      "grad_norm": 0.9015948176383972,
      "learning_rate": 0.0005,
      "loss": 2.7266643047332764,
      "step": 2272
    },
    {
      "epoch": 0.4384645061728395,
      "grad_norm": 1.0032042264938354,
      "learning_rate": 0.0005,
      "loss": 2.7676286697387695,
      "step": 2273
    },
    {
      "epoch": 0.4386574074074074,
      "grad_norm": 0.8662954568862915,
      "learning_rate": 0.0005,
      "loss": 2.650041103363037,
      "step": 2274
    },
    {
      "epoch": 0.43885030864197533,
      "grad_norm": 1.1364628076553345,
      "learning_rate": 0.0005,
      "loss": 3.11419677734375,
      "step": 2275
    },
    {
      "epoch": 0.4390432098765432,
      "grad_norm": 1.7045077085494995,
      "learning_rate": 0.0005,
      "loss": 2.690176010131836,
      "step": 2276
    },
    {
      "epoch": 0.4392361111111111,
      "grad_norm": 1.0027884244918823,
      "learning_rate": 0.0005,
      "loss": 2.863708734512329,
      "step": 2277
    },
    {
      "epoch": 0.439429012345679,
      "grad_norm": 0.9778143763542175,
      "learning_rate": 0.0005,
      "loss": 2.6826090812683105,
      "step": 2278
    },
    {
      "epoch": 0.43962191358024694,
      "grad_norm": 0.8697158694267273,
      "learning_rate": 0.0005,
      "loss": 2.4936113357543945,
      "step": 2279
    },
    {
      "epoch": 0.4398148148148148,
      "grad_norm": 0.9303709268569946,
      "learning_rate": 0.0005,
      "loss": 2.8072690963745117,
      "step": 2280
    },
    {
      "epoch": 0.4400077160493827,
      "grad_norm": 0.9187392592430115,
      "learning_rate": 0.0005,
      "loss": 2.7927520275115967,
      "step": 2281
    },
    {
      "epoch": 0.4402006172839506,
      "grad_norm": 1.0409859418869019,
      "learning_rate": 0.0005,
      "loss": 3.0875649452209473,
      "step": 2282
    },
    {
      "epoch": 0.44039351851851855,
      "grad_norm": 0.9591084718704224,
      "learning_rate": 0.0005,
      "loss": 2.6045265197753906,
      "step": 2283
    },
    {
      "epoch": 0.44058641975308643,
      "grad_norm": 1.2458511590957642,
      "learning_rate": 0.0005,
      "loss": 2.940573215484619,
      "step": 2284
    },
    {
      "epoch": 0.4407793209876543,
      "grad_norm": 1.1079455614089966,
      "learning_rate": 0.0005,
      "loss": 2.968254804611206,
      "step": 2285
    },
    {
      "epoch": 0.4409722222222222,
      "grad_norm": 1.1544853448867798,
      "learning_rate": 0.0005,
      "loss": 2.712634563446045,
      "step": 2286
    },
    {
      "epoch": 0.4411651234567901,
      "grad_norm": 0.9088228344917297,
      "learning_rate": 0.0005,
      "loss": 2.9657907485961914,
      "step": 2287
    },
    {
      "epoch": 0.44135802469135804,
      "grad_norm": 1.1849900484085083,
      "learning_rate": 0.0005,
      "loss": 2.7798867225646973,
      "step": 2288
    },
    {
      "epoch": 0.44155092592592593,
      "grad_norm": 1.1295303106307983,
      "learning_rate": 0.0005,
      "loss": 2.963667392730713,
      "step": 2289
    },
    {
      "epoch": 0.4417438271604938,
      "grad_norm": 0.9957199096679688,
      "learning_rate": 0.0005,
      "loss": 2.802833318710327,
      "step": 2290
    },
    {
      "epoch": 0.4419367283950617,
      "grad_norm": 1.0952638387680054,
      "learning_rate": 0.0005,
      "loss": 2.938448190689087,
      "step": 2291
    },
    {
      "epoch": 0.44212962962962965,
      "grad_norm": 0.8441085815429688,
      "learning_rate": 0.0005,
      "loss": 2.7509360313415527,
      "step": 2292
    },
    {
      "epoch": 0.44232253086419754,
      "grad_norm": 0.9442469477653503,
      "learning_rate": 0.0005,
      "loss": 2.8094964027404785,
      "step": 2293
    },
    {
      "epoch": 0.4425154320987654,
      "grad_norm": 0.9377251863479614,
      "learning_rate": 0.0005,
      "loss": 2.7323782444000244,
      "step": 2294
    },
    {
      "epoch": 0.4427083333333333,
      "grad_norm": 1.1493581533432007,
      "learning_rate": 0.0005,
      "loss": 2.6403231620788574,
      "step": 2295
    },
    {
      "epoch": 0.44290123456790126,
      "grad_norm": 1.0465320348739624,
      "learning_rate": 0.0005,
      "loss": 3.1126604080200195,
      "step": 2296
    },
    {
      "epoch": 0.44309413580246915,
      "grad_norm": 0.7777162790298462,
      "learning_rate": 0.0005,
      "loss": 2.6748313903808594,
      "step": 2297
    },
    {
      "epoch": 0.44328703703703703,
      "grad_norm": 0.914137601852417,
      "learning_rate": 0.0005,
      "loss": 3.0071535110473633,
      "step": 2298
    },
    {
      "epoch": 0.4434799382716049,
      "grad_norm": 1.219814658164978,
      "learning_rate": 0.0005,
      "loss": 2.922849178314209,
      "step": 2299
    },
    {
      "epoch": 0.44367283950617287,
      "grad_norm": 0.7857640385627747,
      "learning_rate": 0.0005,
      "loss": 2.71505069732666,
      "step": 2300
    },
    {
      "epoch": 0.44386574074074076,
      "grad_norm": 1.143118977546692,
      "learning_rate": 0.0005,
      "loss": 2.9043118953704834,
      "step": 2301
    },
    {
      "epoch": 0.44405864197530864,
      "grad_norm": 1.0760095119476318,
      "learning_rate": 0.0005,
      "loss": 2.6488921642303467,
      "step": 2302
    },
    {
      "epoch": 0.44425154320987653,
      "grad_norm": 1.1320604085922241,
      "learning_rate": 0.0005,
      "loss": 2.788822650909424,
      "step": 2303
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 1.297991156578064,
      "learning_rate": 0.0005,
      "loss": 2.809502363204956,
      "step": 2304
    },
    {
      "epoch": 0.44463734567901236,
      "grad_norm": 1.1291296482086182,
      "learning_rate": 0.0005,
      "loss": 2.793745994567871,
      "step": 2305
    },
    {
      "epoch": 0.44483024691358025,
      "grad_norm": 1.021206021308899,
      "learning_rate": 0.0005,
      "loss": 3.0707383155822754,
      "step": 2306
    },
    {
      "epoch": 0.44502314814814814,
      "grad_norm": 1.006285309791565,
      "learning_rate": 0.0005,
      "loss": 2.5979912281036377,
      "step": 2307
    },
    {
      "epoch": 0.44521604938271603,
      "grad_norm": 1.1222913265228271,
      "learning_rate": 0.0005,
      "loss": 2.546241044998169,
      "step": 2308
    },
    {
      "epoch": 0.44540895061728397,
      "grad_norm": 1.0757912397384644,
      "learning_rate": 0.0005,
      "loss": 2.864044189453125,
      "step": 2309
    },
    {
      "epoch": 0.44560185185185186,
      "grad_norm": 0.9523848295211792,
      "learning_rate": 0.0005,
      "loss": 2.7599005699157715,
      "step": 2310
    },
    {
      "epoch": 0.44579475308641975,
      "grad_norm": 0.8907416462898254,
      "learning_rate": 0.0005,
      "loss": 2.7392053604125977,
      "step": 2311
    },
    {
      "epoch": 0.44598765432098764,
      "grad_norm": 1.3636128902435303,
      "learning_rate": 0.0005,
      "loss": 3.081575393676758,
      "step": 2312
    },
    {
      "epoch": 0.4461805555555556,
      "grad_norm": 1.205923318862915,
      "learning_rate": 0.0005,
      "loss": 2.6003096103668213,
      "step": 2313
    },
    {
      "epoch": 0.44637345679012347,
      "grad_norm": 0.9519714117050171,
      "learning_rate": 0.0005,
      "loss": 2.778048038482666,
      "step": 2314
    },
    {
      "epoch": 0.44656635802469136,
      "grad_norm": 1.0149811506271362,
      "learning_rate": 0.0005,
      "loss": 2.6280364990234375,
      "step": 2315
    },
    {
      "epoch": 0.44675925925925924,
      "grad_norm": 1.309378981590271,
      "learning_rate": 0.0005,
      "loss": 2.8136205673217773,
      "step": 2316
    },
    {
      "epoch": 0.44695216049382713,
      "grad_norm": 1.1249138116836548,
      "learning_rate": 0.0005,
      "loss": 2.774904727935791,
      "step": 2317
    },
    {
      "epoch": 0.4471450617283951,
      "grad_norm": 1.2226718664169312,
      "learning_rate": 0.0005,
      "loss": 2.6661109924316406,
      "step": 2318
    },
    {
      "epoch": 0.44733796296296297,
      "grad_norm": 0.9669070839881897,
      "learning_rate": 0.0005,
      "loss": 2.6386685371398926,
      "step": 2319
    },
    {
      "epoch": 0.44753086419753085,
      "grad_norm": 0.9595359563827515,
      "learning_rate": 0.0005,
      "loss": 2.890230178833008,
      "step": 2320
    },
    {
      "epoch": 0.44772376543209874,
      "grad_norm": 0.8862593770027161,
      "learning_rate": 0.0005,
      "loss": 2.831916332244873,
      "step": 2321
    },
    {
      "epoch": 0.4479166666666667,
      "grad_norm": 1.4633255004882812,
      "learning_rate": 0.0005,
      "loss": 2.7228710651397705,
      "step": 2322
    },
    {
      "epoch": 0.4481095679012346,
      "grad_norm": 1.0320154428482056,
      "learning_rate": 0.0005,
      "loss": 2.692615032196045,
      "step": 2323
    },
    {
      "epoch": 0.44830246913580246,
      "grad_norm": 1.1530041694641113,
      "learning_rate": 0.0005,
      "loss": 2.7352921962738037,
      "step": 2324
    },
    {
      "epoch": 0.44849537037037035,
      "grad_norm": 1.045076608657837,
      "learning_rate": 0.0005,
      "loss": 2.729632616043091,
      "step": 2325
    },
    {
      "epoch": 0.4486882716049383,
      "grad_norm": 1.1475740671157837,
      "learning_rate": 0.0005,
      "loss": 2.871117115020752,
      "step": 2326
    },
    {
      "epoch": 0.4488811728395062,
      "grad_norm": 1.0779026746749878,
      "learning_rate": 0.0005,
      "loss": 2.786309242248535,
      "step": 2327
    },
    {
      "epoch": 0.44907407407407407,
      "grad_norm": 0.917044997215271,
      "learning_rate": 0.0005,
      "loss": 2.817654609680176,
      "step": 2328
    },
    {
      "epoch": 0.44926697530864196,
      "grad_norm": 1.4449005126953125,
      "learning_rate": 0.0005,
      "loss": 2.92498779296875,
      "step": 2329
    },
    {
      "epoch": 0.4494598765432099,
      "grad_norm": 1.2221118211746216,
      "learning_rate": 0.0005,
      "loss": 2.7347257137298584,
      "step": 2330
    },
    {
      "epoch": 0.4496527777777778,
      "grad_norm": 1.2491146326065063,
      "learning_rate": 0.0005,
      "loss": 2.8547792434692383,
      "step": 2331
    },
    {
      "epoch": 0.4498456790123457,
      "grad_norm": 1.4188086986541748,
      "learning_rate": 0.0005,
      "loss": 2.611616611480713,
      "step": 2332
    },
    {
      "epoch": 0.45003858024691357,
      "grad_norm": 1.2013990879058838,
      "learning_rate": 0.0005,
      "loss": 2.8323936462402344,
      "step": 2333
    },
    {
      "epoch": 0.45023148148148145,
      "grad_norm": 1.97731614112854,
      "learning_rate": 0.0005,
      "loss": 2.9220762252807617,
      "step": 2334
    },
    {
      "epoch": 0.4504243827160494,
      "grad_norm": 0.966826319694519,
      "learning_rate": 0.0005,
      "loss": 2.877498149871826,
      "step": 2335
    },
    {
      "epoch": 0.4506172839506173,
      "grad_norm": 0.9617243409156799,
      "learning_rate": 0.0005,
      "loss": 3.086833953857422,
      "step": 2336
    },
    {
      "epoch": 0.4508101851851852,
      "grad_norm": 1.0278342962265015,
      "learning_rate": 0.0005,
      "loss": 2.7973904609680176,
      "step": 2337
    },
    {
      "epoch": 0.45100308641975306,
      "grad_norm": 1.1300044059753418,
      "learning_rate": 0.0005,
      "loss": 2.83260440826416,
      "step": 2338
    },
    {
      "epoch": 0.451195987654321,
      "grad_norm": 0.8997827768325806,
      "learning_rate": 0.0005,
      "loss": 2.591242790222168,
      "step": 2339
    },
    {
      "epoch": 0.4513888888888889,
      "grad_norm": 1.1852127313613892,
      "learning_rate": 0.0005,
      "loss": 3.0024213790893555,
      "step": 2340
    },
    {
      "epoch": 0.4515817901234568,
      "grad_norm": 1.0002694129943848,
      "learning_rate": 0.0005,
      "loss": 2.5516762733459473,
      "step": 2341
    },
    {
      "epoch": 0.45177469135802467,
      "grad_norm": 1.0580235719680786,
      "learning_rate": 0.0005,
      "loss": 2.740039587020874,
      "step": 2342
    },
    {
      "epoch": 0.4519675925925926,
      "grad_norm": 0.9745818972587585,
      "learning_rate": 0.0005,
      "loss": 2.796510696411133,
      "step": 2343
    },
    {
      "epoch": 0.4521604938271605,
      "grad_norm": 0.8815824389457703,
      "learning_rate": 0.0005,
      "loss": 2.857560634613037,
      "step": 2344
    },
    {
      "epoch": 0.4523533950617284,
      "grad_norm": 0.909537136554718,
      "learning_rate": 0.0005,
      "loss": 2.793924331665039,
      "step": 2345
    },
    {
      "epoch": 0.4525462962962963,
      "grad_norm": 1.0240492820739746,
      "learning_rate": 0.0005,
      "loss": 2.551579475402832,
      "step": 2346
    },
    {
      "epoch": 0.4527391975308642,
      "grad_norm": 1.1294986009597778,
      "learning_rate": 0.0005,
      "loss": 2.7679738998413086,
      "step": 2347
    },
    {
      "epoch": 0.4529320987654321,
      "grad_norm": 1.1536246538162231,
      "learning_rate": 0.0005,
      "loss": 2.9047670364379883,
      "step": 2348
    },
    {
      "epoch": 0.453125,
      "grad_norm": 1.0532253980636597,
      "learning_rate": 0.0005,
      "loss": 2.811260223388672,
      "step": 2349
    },
    {
      "epoch": 0.4533179012345679,
      "grad_norm": 1.3276952505111694,
      "learning_rate": 0.0005,
      "loss": 2.88163685798645,
      "step": 2350
    },
    {
      "epoch": 0.4535108024691358,
      "grad_norm": 0.9510443210601807,
      "learning_rate": 0.0005,
      "loss": 2.6497766971588135,
      "step": 2351
    },
    {
      "epoch": 0.4537037037037037,
      "grad_norm": 1.169108510017395,
      "learning_rate": 0.0005,
      "loss": 2.8061389923095703,
      "step": 2352
    },
    {
      "epoch": 0.4538966049382716,
      "grad_norm": 1.3535720109939575,
      "learning_rate": 0.0005,
      "loss": 2.6576948165893555,
      "step": 2353
    },
    {
      "epoch": 0.4540895061728395,
      "grad_norm": 0.9746114015579224,
      "learning_rate": 0.0005,
      "loss": 3.0150699615478516,
      "step": 2354
    },
    {
      "epoch": 0.4542824074074074,
      "grad_norm": 0.9457666277885437,
      "learning_rate": 0.0005,
      "loss": 2.4138259887695312,
      "step": 2355
    },
    {
      "epoch": 0.45447530864197533,
      "grad_norm": 0.970536470413208,
      "learning_rate": 0.0005,
      "loss": 3.068748950958252,
      "step": 2356
    },
    {
      "epoch": 0.4546682098765432,
      "grad_norm": 1.3599309921264648,
      "learning_rate": 0.0005,
      "loss": 2.672116756439209,
      "step": 2357
    },
    {
      "epoch": 0.4548611111111111,
      "grad_norm": 0.855570375919342,
      "learning_rate": 0.0005,
      "loss": 2.6759822368621826,
      "step": 2358
    },
    {
      "epoch": 0.455054012345679,
      "grad_norm": 0.954738974571228,
      "learning_rate": 0.0005,
      "loss": 2.6924376487731934,
      "step": 2359
    },
    {
      "epoch": 0.45524691358024694,
      "grad_norm": 1.2248995304107666,
      "learning_rate": 0.0005,
      "loss": 2.5718002319335938,
      "step": 2360
    },
    {
      "epoch": 0.4554398148148148,
      "grad_norm": 0.8252365589141846,
      "learning_rate": 0.0005,
      "loss": 2.6071746349334717,
      "step": 2361
    },
    {
      "epoch": 0.4556327160493827,
      "grad_norm": 1.1754251718521118,
      "learning_rate": 0.0005,
      "loss": 2.8467016220092773,
      "step": 2362
    },
    {
      "epoch": 0.4558256172839506,
      "grad_norm": 1.0854500532150269,
      "learning_rate": 0.0005,
      "loss": 2.6305348873138428,
      "step": 2363
    },
    {
      "epoch": 0.45601851851851855,
      "grad_norm": 1.1263347864151,
      "learning_rate": 0.0005,
      "loss": 2.822110414505005,
      "step": 2364
    },
    {
      "epoch": 0.45621141975308643,
      "grad_norm": 1.1568304300308228,
      "learning_rate": 0.0005,
      "loss": 2.48759388923645,
      "step": 2365
    },
    {
      "epoch": 0.4564043209876543,
      "grad_norm": 0.9316380023956299,
      "learning_rate": 0.0005,
      "loss": 2.6978111267089844,
      "step": 2366
    },
    {
      "epoch": 0.4565972222222222,
      "grad_norm": 1.0264356136322021,
      "learning_rate": 0.0005,
      "loss": 2.474515676498413,
      "step": 2367
    },
    {
      "epoch": 0.4567901234567901,
      "grad_norm": 1.0856244564056396,
      "learning_rate": 0.0005,
      "loss": 2.6260204315185547,
      "step": 2368
    },
    {
      "epoch": 0.45698302469135804,
      "grad_norm": 1.1010724306106567,
      "learning_rate": 0.0005,
      "loss": 2.932950019836426,
      "step": 2369
    },
    {
      "epoch": 0.45717592592592593,
      "grad_norm": 0.9733690619468689,
      "learning_rate": 0.0005,
      "loss": 2.6817102432250977,
      "step": 2370
    },
    {
      "epoch": 0.4573688271604938,
      "grad_norm": 0.9835867881774902,
      "learning_rate": 0.0005,
      "loss": 2.945415496826172,
      "step": 2371
    },
    {
      "epoch": 0.4575617283950617,
      "grad_norm": 1.0124433040618896,
      "learning_rate": 0.0005,
      "loss": 2.979435443878174,
      "step": 2372
    },
    {
      "epoch": 0.45775462962962965,
      "grad_norm": 0.8630382418632507,
      "learning_rate": 0.0005,
      "loss": 2.617037534713745,
      "step": 2373
    },
    {
      "epoch": 0.45794753086419754,
      "grad_norm": 0.9559342861175537,
      "learning_rate": 0.0005,
      "loss": 2.7961394786834717,
      "step": 2374
    },
    {
      "epoch": 0.4581404320987654,
      "grad_norm": 1.3442776203155518,
      "learning_rate": 0.0005,
      "loss": 2.6424126625061035,
      "step": 2375
    },
    {
      "epoch": 0.4583333333333333,
      "grad_norm": 0.9189745783805847,
      "learning_rate": 0.0005,
      "loss": 2.9086971282958984,
      "step": 2376
    },
    {
      "epoch": 0.45852623456790126,
      "grad_norm": 1.2843660116195679,
      "learning_rate": 0.0005,
      "loss": 2.6923434734344482,
      "step": 2377
    },
    {
      "epoch": 0.45871913580246915,
      "grad_norm": 1.7499113082885742,
      "learning_rate": 0.0005,
      "loss": 2.735750913619995,
      "step": 2378
    },
    {
      "epoch": 0.45891203703703703,
      "grad_norm": 0.9053550958633423,
      "learning_rate": 0.0005,
      "loss": 2.7223868370056152,
      "step": 2379
    },
    {
      "epoch": 0.4591049382716049,
      "grad_norm": 1.01988685131073,
      "learning_rate": 0.0005,
      "loss": 2.636443614959717,
      "step": 2380
    },
    {
      "epoch": 0.45929783950617287,
      "grad_norm": 0.8794488906860352,
      "learning_rate": 0.0005,
      "loss": 2.6001293659210205,
      "step": 2381
    },
    {
      "epoch": 0.45949074074074076,
      "grad_norm": 0.8515363931655884,
      "learning_rate": 0.0005,
      "loss": 2.8567919731140137,
      "step": 2382
    },
    {
      "epoch": 0.45968364197530864,
      "grad_norm": 1.0742669105529785,
      "learning_rate": 0.0005,
      "loss": 2.5471673011779785,
      "step": 2383
    },
    {
      "epoch": 0.45987654320987653,
      "grad_norm": 1.1238077878952026,
      "learning_rate": 0.0005,
      "loss": 2.8511881828308105,
      "step": 2384
    },
    {
      "epoch": 0.4600694444444444,
      "grad_norm": 0.9476693868637085,
      "learning_rate": 0.0005,
      "loss": 2.705345630645752,
      "step": 2385
    },
    {
      "epoch": 0.46026234567901236,
      "grad_norm": 1.5189168453216553,
      "learning_rate": 0.0005,
      "loss": 2.5853614807128906,
      "step": 2386
    },
    {
      "epoch": 0.46045524691358025,
      "grad_norm": 1.3214844465255737,
      "learning_rate": 0.0005,
      "loss": 2.9540624618530273,
      "step": 2387
    },
    {
      "epoch": 0.46064814814814814,
      "grad_norm": 1.2549716234207153,
      "learning_rate": 0.0005,
      "loss": 2.9865317344665527,
      "step": 2388
    },
    {
      "epoch": 0.46084104938271603,
      "grad_norm": 0.8610287308692932,
      "learning_rate": 0.0005,
      "loss": 2.9919567108154297,
      "step": 2389
    },
    {
      "epoch": 0.46103395061728397,
      "grad_norm": 1.170249104499817,
      "learning_rate": 0.0005,
      "loss": 2.732396125793457,
      "step": 2390
    },
    {
      "epoch": 0.46122685185185186,
      "grad_norm": 0.877190887928009,
      "learning_rate": 0.0005,
      "loss": 2.592280626296997,
      "step": 2391
    },
    {
      "epoch": 0.46141975308641975,
      "grad_norm": 1.5075680017471313,
      "learning_rate": 0.0005,
      "loss": 2.7438888549804688,
      "step": 2392
    },
    {
      "epoch": 0.46161265432098764,
      "grad_norm": 1.0035334825515747,
      "learning_rate": 0.0005,
      "loss": 2.9229469299316406,
      "step": 2393
    },
    {
      "epoch": 0.4618055555555556,
      "grad_norm": 1.0379469394683838,
      "learning_rate": 0.0005,
      "loss": 2.854156494140625,
      "step": 2394
    },
    {
      "epoch": 0.46199845679012347,
      "grad_norm": 2.417356014251709,
      "learning_rate": 0.0005,
      "loss": 3.225212335586548,
      "step": 2395
    },
    {
      "epoch": 0.46219135802469136,
      "grad_norm": 0.9414316415786743,
      "learning_rate": 0.0005,
      "loss": 2.69212007522583,
      "step": 2396
    },
    {
      "epoch": 0.46238425925925924,
      "grad_norm": 0.9392073750495911,
      "learning_rate": 0.0005,
      "loss": 2.7283143997192383,
      "step": 2397
    },
    {
      "epoch": 0.46257716049382713,
      "grad_norm": 1.4101554155349731,
      "learning_rate": 0.0005,
      "loss": 2.7472493648529053,
      "step": 2398
    },
    {
      "epoch": 0.4627700617283951,
      "grad_norm": 0.9403383731842041,
      "learning_rate": 0.0005,
      "loss": 2.6866931915283203,
      "step": 2399
    },
    {
      "epoch": 0.46296296296296297,
      "grad_norm": 1.0672167539596558,
      "learning_rate": 0.0005,
      "loss": 2.865185499191284,
      "step": 2400
    },
    {
      "epoch": 0.46315586419753085,
      "grad_norm": 0.8279421329498291,
      "learning_rate": 0.0005,
      "loss": 2.4860453605651855,
      "step": 2401
    },
    {
      "epoch": 0.46334876543209874,
      "grad_norm": 0.8609778881072998,
      "learning_rate": 0.0005,
      "loss": 2.605588436126709,
      "step": 2402
    },
    {
      "epoch": 0.4635416666666667,
      "grad_norm": 0.9213933348655701,
      "learning_rate": 0.0005,
      "loss": 2.743621349334717,
      "step": 2403
    },
    {
      "epoch": 0.4637345679012346,
      "grad_norm": 0.9583457112312317,
      "learning_rate": 0.0005,
      "loss": 2.655815839767456,
      "step": 2404
    },
    {
      "epoch": 0.46392746913580246,
      "grad_norm": 1.0195350646972656,
      "learning_rate": 0.0005,
      "loss": 2.7126643657684326,
      "step": 2405
    },
    {
      "epoch": 0.46412037037037035,
      "grad_norm": 0.8505141735076904,
      "learning_rate": 0.0005,
      "loss": 2.5870280265808105,
      "step": 2406
    },
    {
      "epoch": 0.4643132716049383,
      "grad_norm": 1.3604298830032349,
      "learning_rate": 0.0005,
      "loss": 2.7442188262939453,
      "step": 2407
    },
    {
      "epoch": 0.4645061728395062,
      "grad_norm": 0.9482142329216003,
      "learning_rate": 0.0005,
      "loss": 2.623861312866211,
      "step": 2408
    },
    {
      "epoch": 0.46469907407407407,
      "grad_norm": 0.8727285861968994,
      "learning_rate": 0.0005,
      "loss": 2.3437578678131104,
      "step": 2409
    },
    {
      "epoch": 0.46489197530864196,
      "grad_norm": 1.2055026292800903,
      "learning_rate": 0.0005,
      "loss": 2.873349189758301,
      "step": 2410
    },
    {
      "epoch": 0.4650848765432099,
      "grad_norm": 1.7697619199752808,
      "learning_rate": 0.0005,
      "loss": 2.466168165206909,
      "step": 2411
    },
    {
      "epoch": 0.4652777777777778,
      "grad_norm": 1.0985404253005981,
      "learning_rate": 0.0005,
      "loss": 3.2117981910705566,
      "step": 2412
    },
    {
      "epoch": 0.4654706790123457,
      "grad_norm": 0.7752571702003479,
      "learning_rate": 0.0005,
      "loss": 2.655963897705078,
      "step": 2413
    },
    {
      "epoch": 0.46566358024691357,
      "grad_norm": 0.8121854662895203,
      "learning_rate": 0.0005,
      "loss": 2.588104486465454,
      "step": 2414
    },
    {
      "epoch": 0.46585648148148145,
      "grad_norm": 1.663192629814148,
      "learning_rate": 0.0005,
      "loss": 3.2150802612304688,
      "step": 2415
    },
    {
      "epoch": 0.4660493827160494,
      "grad_norm": 1.0480626821517944,
      "learning_rate": 0.0005,
      "loss": 2.9694032669067383,
      "step": 2416
    },
    {
      "epoch": 0.4662422839506173,
      "grad_norm": 1.0409648418426514,
      "learning_rate": 0.0005,
      "loss": 2.7963995933532715,
      "step": 2417
    },
    {
      "epoch": 0.4664351851851852,
      "grad_norm": 1.1499804258346558,
      "learning_rate": 0.0005,
      "loss": 2.6401610374450684,
      "step": 2418
    },
    {
      "epoch": 0.46662808641975306,
      "grad_norm": 0.9596388339996338,
      "learning_rate": 0.0005,
      "loss": 2.668344497680664,
      "step": 2419
    },
    {
      "epoch": 0.466820987654321,
      "grad_norm": 1.0046603679656982,
      "learning_rate": 0.0005,
      "loss": 2.918771266937256,
      "step": 2420
    },
    {
      "epoch": 0.4670138888888889,
      "grad_norm": 1.057981014251709,
      "learning_rate": 0.0005,
      "loss": 2.6079797744750977,
      "step": 2421
    },
    {
      "epoch": 0.4672067901234568,
      "grad_norm": 0.8877140879631042,
      "learning_rate": 0.0005,
      "loss": 2.6381261348724365,
      "step": 2422
    },
    {
      "epoch": 0.46739969135802467,
      "grad_norm": 1.1974281072616577,
      "learning_rate": 0.0005,
      "loss": 2.370105028152466,
      "step": 2423
    },
    {
      "epoch": 0.4675925925925926,
      "grad_norm": 1.0586743354797363,
      "learning_rate": 0.0005,
      "loss": 2.663236379623413,
      "step": 2424
    },
    {
      "epoch": 0.4677854938271605,
      "grad_norm": 1.0348060131072998,
      "learning_rate": 0.0005,
      "loss": 2.7634289264678955,
      "step": 2425
    },
    {
      "epoch": 0.4679783950617284,
      "grad_norm": 1.0899418592453003,
      "learning_rate": 0.0005,
      "loss": 2.9809908866882324,
      "step": 2426
    },
    {
      "epoch": 0.4681712962962963,
      "grad_norm": 1.0668781995773315,
      "learning_rate": 0.0005,
      "loss": 2.421724796295166,
      "step": 2427
    },
    {
      "epoch": 0.4683641975308642,
      "grad_norm": 1.108742594718933,
      "learning_rate": 0.0005,
      "loss": 2.7695529460906982,
      "step": 2428
    },
    {
      "epoch": 0.4685570987654321,
      "grad_norm": 0.9043829441070557,
      "learning_rate": 0.0005,
      "loss": 2.643526792526245,
      "step": 2429
    },
    {
      "epoch": 0.46875,
      "grad_norm": 1.369551420211792,
      "learning_rate": 0.0005,
      "loss": 2.7845985889434814,
      "step": 2430
    },
    {
      "epoch": 0.4689429012345679,
      "grad_norm": 0.9451095461845398,
      "learning_rate": 0.0005,
      "loss": 2.7207653522491455,
      "step": 2431
    },
    {
      "epoch": 0.4691358024691358,
      "grad_norm": 1.0961410999298096,
      "learning_rate": 0.0005,
      "loss": 2.5860495567321777,
      "step": 2432
    },
    {
      "epoch": 0.4693287037037037,
      "grad_norm": 1.062678337097168,
      "learning_rate": 0.0005,
      "loss": 2.903101682662964,
      "step": 2433
    },
    {
      "epoch": 0.4695216049382716,
      "grad_norm": 1.1599745750427246,
      "learning_rate": 0.0005,
      "loss": 2.508258581161499,
      "step": 2434
    },
    {
      "epoch": 0.4697145061728395,
      "grad_norm": 1.1779913902282715,
      "learning_rate": 0.0005,
      "loss": 2.741611957550049,
      "step": 2435
    },
    {
      "epoch": 0.4699074074074074,
      "grad_norm": 1.4135464429855347,
      "learning_rate": 0.0005,
      "loss": 2.660067558288574,
      "step": 2436
    },
    {
      "epoch": 0.47010030864197533,
      "grad_norm": 1.0421780347824097,
      "learning_rate": 0.0005,
      "loss": 2.8959577083587646,
      "step": 2437
    },
    {
      "epoch": 0.4702932098765432,
      "grad_norm": 0.9533393383026123,
      "learning_rate": 0.0005,
      "loss": 2.8509013652801514,
      "step": 2438
    },
    {
      "epoch": 0.4704861111111111,
      "grad_norm": 0.8419036269187927,
      "learning_rate": 0.0005,
      "loss": 3.092294931411743,
      "step": 2439
    },
    {
      "epoch": 0.470679012345679,
      "grad_norm": 1.3073211908340454,
      "learning_rate": 0.0005,
      "loss": 2.7534613609313965,
      "step": 2440
    },
    {
      "epoch": 0.47087191358024694,
      "grad_norm": 1.179837703704834,
      "learning_rate": 0.0005,
      "loss": 2.8903450965881348,
      "step": 2441
    },
    {
      "epoch": 0.4710648148148148,
      "grad_norm": 1.0427370071411133,
      "learning_rate": 0.0005,
      "loss": 2.5026774406433105,
      "step": 2442
    },
    {
      "epoch": 0.4712577160493827,
      "grad_norm": 1.2700514793395996,
      "learning_rate": 0.0005,
      "loss": 2.920278549194336,
      "step": 2443
    },
    {
      "epoch": 0.4714506172839506,
      "grad_norm": 1.1696163415908813,
      "learning_rate": 0.0005,
      "loss": 2.77987003326416,
      "step": 2444
    },
    {
      "epoch": 0.47164351851851855,
      "grad_norm": 0.9800257086753845,
      "learning_rate": 0.0005,
      "loss": 3.086538314819336,
      "step": 2445
    },
    {
      "epoch": 0.47183641975308643,
      "grad_norm": 1.1978602409362793,
      "learning_rate": 0.0005,
      "loss": 2.7769827842712402,
      "step": 2446
    },
    {
      "epoch": 0.4720293209876543,
      "grad_norm": 0.9120871424674988,
      "learning_rate": 0.0005,
      "loss": 2.7642502784729004,
      "step": 2447
    },
    {
      "epoch": 0.4722222222222222,
      "grad_norm": 1.1650762557983398,
      "learning_rate": 0.0005,
      "loss": 2.5998125076293945,
      "step": 2448
    },
    {
      "epoch": 0.4724151234567901,
      "grad_norm": 0.8436678647994995,
      "learning_rate": 0.0005,
      "loss": 2.660888910293579,
      "step": 2449
    },
    {
      "epoch": 0.47260802469135804,
      "grad_norm": 1.1425628662109375,
      "learning_rate": 0.0005,
      "loss": 2.8877882957458496,
      "step": 2450
    },
    {
      "epoch": 0.47280092592592593,
      "grad_norm": 1.0370200872421265,
      "learning_rate": 0.0005,
      "loss": 3.0236434936523438,
      "step": 2451
    },
    {
      "epoch": 0.4729938271604938,
      "grad_norm": 0.9743171334266663,
      "learning_rate": 0.0005,
      "loss": 2.6910390853881836,
      "step": 2452
    },
    {
      "epoch": 0.4731867283950617,
      "grad_norm": 1.339601755142212,
      "learning_rate": 0.0005,
      "loss": 2.5706398487091064,
      "step": 2453
    },
    {
      "epoch": 0.47337962962962965,
      "grad_norm": 0.9648007750511169,
      "learning_rate": 0.0005,
      "loss": 2.7478699684143066,
      "step": 2454
    },
    {
      "epoch": 0.47357253086419754,
      "grad_norm": 1.0996979475021362,
      "learning_rate": 0.0005,
      "loss": 2.914850950241089,
      "step": 2455
    },
    {
      "epoch": 0.4737654320987654,
      "grad_norm": 1.027542233467102,
      "learning_rate": 0.0005,
      "loss": 2.5749502182006836,
      "step": 2456
    },
    {
      "epoch": 0.4739583333333333,
      "grad_norm": 0.9667055606842041,
      "learning_rate": 0.0005,
      "loss": 2.751211643218994,
      "step": 2457
    },
    {
      "epoch": 0.47415123456790126,
      "grad_norm": 0.8129447102546692,
      "learning_rate": 0.0005,
      "loss": 2.5908761024475098,
      "step": 2458
    },
    {
      "epoch": 0.47434413580246915,
      "grad_norm": 1.1944615840911865,
      "learning_rate": 0.0005,
      "loss": 3.090432643890381,
      "step": 2459
    },
    {
      "epoch": 0.47453703703703703,
      "grad_norm": 1.1014989614486694,
      "learning_rate": 0.0005,
      "loss": 2.7982120513916016,
      "step": 2460
    },
    {
      "epoch": 0.4747299382716049,
      "grad_norm": 2.3558719158172607,
      "learning_rate": 0.0005,
      "loss": 2.768649101257324,
      "step": 2461
    },
    {
      "epoch": 0.47492283950617287,
      "grad_norm": 0.9729190468788147,
      "learning_rate": 0.0005,
      "loss": 2.9052035808563232,
      "step": 2462
    },
    {
      "epoch": 0.47511574074074076,
      "grad_norm": 1.0346605777740479,
      "learning_rate": 0.0005,
      "loss": 2.8030760288238525,
      "step": 2463
    },
    {
      "epoch": 0.47530864197530864,
      "grad_norm": 0.9982994794845581,
      "learning_rate": 0.0005,
      "loss": 2.6721041202545166,
      "step": 2464
    },
    {
      "epoch": 0.47550154320987653,
      "grad_norm": 1.0575019121170044,
      "learning_rate": 0.0005,
      "loss": 2.7604525089263916,
      "step": 2465
    },
    {
      "epoch": 0.4756944444444444,
      "grad_norm": 0.8469808101654053,
      "learning_rate": 0.0005,
      "loss": 2.5195658206939697,
      "step": 2466
    },
    {
      "epoch": 0.47588734567901236,
      "grad_norm": 1.1171798706054688,
      "learning_rate": 0.0005,
      "loss": 2.9471757411956787,
      "step": 2467
    },
    {
      "epoch": 0.47608024691358025,
      "grad_norm": 1.4411083459854126,
      "learning_rate": 0.0005,
      "loss": 3.0175204277038574,
      "step": 2468
    },
    {
      "epoch": 0.47627314814814814,
      "grad_norm": 1.105770230293274,
      "learning_rate": 0.0005,
      "loss": 3.0589795112609863,
      "step": 2469
    },
    {
      "epoch": 0.47646604938271603,
      "grad_norm": 1.0137109756469727,
      "learning_rate": 0.0005,
      "loss": 2.6817727088928223,
      "step": 2470
    },
    {
      "epoch": 0.47665895061728397,
      "grad_norm": 1.1413041353225708,
      "learning_rate": 0.0005,
      "loss": 2.8968498706817627,
      "step": 2471
    },
    {
      "epoch": 0.47685185185185186,
      "grad_norm": 1.100224256515503,
      "learning_rate": 0.0005,
      "loss": 2.6855316162109375,
      "step": 2472
    },
    {
      "epoch": 0.47704475308641975,
      "grad_norm": 0.9389061331748962,
      "learning_rate": 0.0005,
      "loss": 2.533646821975708,
      "step": 2473
    },
    {
      "epoch": 0.47723765432098764,
      "grad_norm": 0.9047874212265015,
      "learning_rate": 0.0005,
      "loss": 2.7762956619262695,
      "step": 2474
    },
    {
      "epoch": 0.4774305555555556,
      "grad_norm": 1.0268542766571045,
      "learning_rate": 0.0005,
      "loss": 2.5937912464141846,
      "step": 2475
    },
    {
      "epoch": 0.47762345679012347,
      "grad_norm": 1.0406244993209839,
      "learning_rate": 0.0005,
      "loss": 2.7320148944854736,
      "step": 2476
    },
    {
      "epoch": 0.47781635802469136,
      "grad_norm": 0.9390333890914917,
      "learning_rate": 0.0005,
      "loss": 2.6314315795898438,
      "step": 2477
    },
    {
      "epoch": 0.47800925925925924,
      "grad_norm": 0.9700362086296082,
      "learning_rate": 0.0005,
      "loss": 2.6161603927612305,
      "step": 2478
    },
    {
      "epoch": 0.47820216049382713,
      "grad_norm": 0.8086886405944824,
      "learning_rate": 0.0005,
      "loss": 2.4899866580963135,
      "step": 2479
    },
    {
      "epoch": 0.4783950617283951,
      "grad_norm": 0.9501393437385559,
      "learning_rate": 0.0005,
      "loss": 2.5489373207092285,
      "step": 2480
    },
    {
      "epoch": 0.47858796296296297,
      "grad_norm": 1.0816712379455566,
      "learning_rate": 0.0005,
      "loss": 3.1206393241882324,
      "step": 2481
    },
    {
      "epoch": 0.47878086419753085,
      "grad_norm": 0.9032701849937439,
      "learning_rate": 0.0005,
      "loss": 2.643157482147217,
      "step": 2482
    },
    {
      "epoch": 0.47897376543209874,
      "grad_norm": 1.0155649185180664,
      "learning_rate": 0.0005,
      "loss": 2.1249825954437256,
      "step": 2483
    },
    {
      "epoch": 0.4791666666666667,
      "grad_norm": 0.9402036666870117,
      "learning_rate": 0.0005,
      "loss": 2.8091697692871094,
      "step": 2484
    },
    {
      "epoch": 0.4793595679012346,
      "grad_norm": 1.0791093111038208,
      "learning_rate": 0.0005,
      "loss": 2.676863670349121,
      "step": 2485
    },
    {
      "epoch": 0.47955246913580246,
      "grad_norm": 0.9040659666061401,
      "learning_rate": 0.0005,
      "loss": 2.7801389694213867,
      "step": 2486
    },
    {
      "epoch": 0.47974537037037035,
      "grad_norm": 1.0024850368499756,
      "learning_rate": 0.0005,
      "loss": 2.868083953857422,
      "step": 2487
    },
    {
      "epoch": 0.4799382716049383,
      "grad_norm": 0.9202367663383484,
      "learning_rate": 0.0005,
      "loss": 2.8639307022094727,
      "step": 2488
    },
    {
      "epoch": 0.4801311728395062,
      "grad_norm": 0.9824749231338501,
      "learning_rate": 0.0005,
      "loss": 2.5826895236968994,
      "step": 2489
    },
    {
      "epoch": 0.48032407407407407,
      "grad_norm": 0.8465791344642639,
      "learning_rate": 0.0005,
      "loss": 2.4279444217681885,
      "step": 2490
    },
    {
      "epoch": 0.48051697530864196,
      "grad_norm": 0.994774580001831,
      "learning_rate": 0.0005,
      "loss": 2.9153285026550293,
      "step": 2491
    },
    {
      "epoch": 0.4807098765432099,
      "grad_norm": 1.2683504819869995,
      "learning_rate": 0.0005,
      "loss": 2.84706449508667,
      "step": 2492
    },
    {
      "epoch": 0.4809027777777778,
      "grad_norm": 0.9831994771957397,
      "learning_rate": 0.0005,
      "loss": 2.627485990524292,
      "step": 2493
    },
    {
      "epoch": 0.4810956790123457,
      "grad_norm": 1.0298120975494385,
      "learning_rate": 0.0005,
      "loss": 2.7139580249786377,
      "step": 2494
    },
    {
      "epoch": 0.48128858024691357,
      "grad_norm": 1.0423063039779663,
      "learning_rate": 0.0005,
      "loss": 2.548640012741089,
      "step": 2495
    },
    {
      "epoch": 0.48148148148148145,
      "grad_norm": 1.3115735054016113,
      "learning_rate": 0.0005,
      "loss": 2.7862155437469482,
      "step": 2496
    },
    {
      "epoch": 0.4816743827160494,
      "grad_norm": 1.0886512994766235,
      "learning_rate": 0.0005,
      "loss": 2.895631790161133,
      "step": 2497
    },
    {
      "epoch": 0.4818672839506173,
      "grad_norm": 1.0870193243026733,
      "learning_rate": 0.0005,
      "loss": 2.7114925384521484,
      "step": 2498
    },
    {
      "epoch": 0.4820601851851852,
      "grad_norm": 0.9316776394844055,
      "learning_rate": 0.0005,
      "loss": 2.756481170654297,
      "step": 2499
    },
    {
      "epoch": 0.48225308641975306,
      "grad_norm": 1.2238572835922241,
      "learning_rate": 0.0005,
      "loss": 2.8762800693511963,
      "step": 2500
    },
    {
      "epoch": 0.482445987654321,
      "grad_norm": 0.8706267476081848,
      "learning_rate": 0.0005,
      "loss": 2.5039916038513184,
      "step": 2501
    },
    {
      "epoch": 0.4826388888888889,
      "grad_norm": 1.0808568000793457,
      "learning_rate": 0.0005,
      "loss": 2.5855932235717773,
      "step": 2502
    },
    {
      "epoch": 0.4828317901234568,
      "grad_norm": 1.0514607429504395,
      "learning_rate": 0.0005,
      "loss": 2.7664668560028076,
      "step": 2503
    },
    {
      "epoch": 0.48302469135802467,
      "grad_norm": 0.9495949149131775,
      "learning_rate": 0.0005,
      "loss": 3.1577887535095215,
      "step": 2504
    },
    {
      "epoch": 0.4832175925925926,
      "grad_norm": 1.027109980583191,
      "learning_rate": 0.0005,
      "loss": 2.438979387283325,
      "step": 2505
    },
    {
      "epoch": 0.4834104938271605,
      "grad_norm": 1.1089708805084229,
      "learning_rate": 0.0005,
      "loss": 2.6870923042297363,
      "step": 2506
    },
    {
      "epoch": 0.4836033950617284,
      "grad_norm": 0.8754459619522095,
      "learning_rate": 0.0005,
      "loss": 2.6680188179016113,
      "step": 2507
    },
    {
      "epoch": 0.4837962962962963,
      "grad_norm": 0.9469267725944519,
      "learning_rate": 0.0005,
      "loss": 2.600203514099121,
      "step": 2508
    },
    {
      "epoch": 0.4839891975308642,
      "grad_norm": 1.0058740377426147,
      "learning_rate": 0.0005,
      "loss": 2.8946003913879395,
      "step": 2509
    },
    {
      "epoch": 0.4841820987654321,
      "grad_norm": 0.8290565609931946,
      "learning_rate": 0.0005,
      "loss": 2.781834125518799,
      "step": 2510
    },
    {
      "epoch": 0.484375,
      "grad_norm": 0.9700128436088562,
      "learning_rate": 0.0005,
      "loss": 2.7705745697021484,
      "step": 2511
    },
    {
      "epoch": 0.4845679012345679,
      "grad_norm": 0.7864044904708862,
      "learning_rate": 0.0005,
      "loss": 2.5052967071533203,
      "step": 2512
    },
    {
      "epoch": 0.4847608024691358,
      "grad_norm": 0.9910566210746765,
      "learning_rate": 0.0005,
      "loss": 2.6039624214172363,
      "step": 2513
    },
    {
      "epoch": 0.4849537037037037,
      "grad_norm": 0.9397326707839966,
      "learning_rate": 0.0005,
      "loss": 2.7481603622436523,
      "step": 2514
    },
    {
      "epoch": 0.4851466049382716,
      "grad_norm": 0.873440146446228,
      "learning_rate": 0.0005,
      "loss": 2.6681628227233887,
      "step": 2515
    },
    {
      "epoch": 0.4853395061728395,
      "grad_norm": 0.8442041873931885,
      "learning_rate": 0.0005,
      "loss": 2.610372543334961,
      "step": 2516
    },
    {
      "epoch": 0.4855324074074074,
      "grad_norm": 1.1140849590301514,
      "learning_rate": 0.0005,
      "loss": 2.695862293243408,
      "step": 2517
    },
    {
      "epoch": 0.48572530864197533,
      "grad_norm": 0.8593665361404419,
      "learning_rate": 0.0005,
      "loss": 2.7729008197784424,
      "step": 2518
    },
    {
      "epoch": 0.4859182098765432,
      "grad_norm": 0.8987506031990051,
      "learning_rate": 0.0005,
      "loss": 2.9644775390625,
      "step": 2519
    },
    {
      "epoch": 0.4861111111111111,
      "grad_norm": 0.8521425724029541,
      "learning_rate": 0.0005,
      "loss": 2.4294004440307617,
      "step": 2520
    },
    {
      "epoch": 0.486304012345679,
      "grad_norm": 1.0405279397964478,
      "learning_rate": 0.0005,
      "loss": 2.7120628356933594,
      "step": 2521
    },
    {
      "epoch": 0.48649691358024694,
      "grad_norm": 0.9702395796775818,
      "learning_rate": 0.0005,
      "loss": 2.6551365852355957,
      "step": 2522
    },
    {
      "epoch": 0.4866898148148148,
      "grad_norm": 0.9465713500976562,
      "learning_rate": 0.0005,
      "loss": 2.8682851791381836,
      "step": 2523
    },
    {
      "epoch": 0.4868827160493827,
      "grad_norm": 1.0674724578857422,
      "learning_rate": 0.0005,
      "loss": 2.7852187156677246,
      "step": 2524
    },
    {
      "epoch": 0.4870756172839506,
      "grad_norm": 1.0635236501693726,
      "learning_rate": 0.0005,
      "loss": 2.6337594985961914,
      "step": 2525
    },
    {
      "epoch": 0.48726851851851855,
      "grad_norm": 2.379758834838867,
      "learning_rate": 0.0005,
      "loss": 2.433763027191162,
      "step": 2526
    },
    {
      "epoch": 0.48746141975308643,
      "grad_norm": 0.9499239921569824,
      "learning_rate": 0.0005,
      "loss": 2.5213863849639893,
      "step": 2527
    },
    {
      "epoch": 0.4876543209876543,
      "grad_norm": 1.4756628274917603,
      "learning_rate": 0.0005,
      "loss": 2.8679378032684326,
      "step": 2528
    },
    {
      "epoch": 0.4878472222222222,
      "grad_norm": 1.0387521982192993,
      "learning_rate": 0.0005,
      "loss": 2.9085183143615723,
      "step": 2529
    },
    {
      "epoch": 0.4880401234567901,
      "grad_norm": 0.9363865256309509,
      "learning_rate": 0.0005,
      "loss": 2.7835240364074707,
      "step": 2530
    },
    {
      "epoch": 0.48823302469135804,
      "grad_norm": 0.8233108520507812,
      "learning_rate": 0.0005,
      "loss": 2.641447067260742,
      "step": 2531
    },
    {
      "epoch": 0.48842592592592593,
      "grad_norm": 1.0047544240951538,
      "learning_rate": 0.0005,
      "loss": 2.7928566932678223,
      "step": 2532
    },
    {
      "epoch": 0.4886188271604938,
      "grad_norm": 1.0277131795883179,
      "learning_rate": 0.0005,
      "loss": 2.4574618339538574,
      "step": 2533
    },
    {
      "epoch": 0.4888117283950617,
      "grad_norm": 1.0812150239944458,
      "learning_rate": 0.0005,
      "loss": 2.547699213027954,
      "step": 2534
    },
    {
      "epoch": 0.48900462962962965,
      "grad_norm": 1.2669482231140137,
      "learning_rate": 0.0005,
      "loss": 2.6120543479919434,
      "step": 2535
    },
    {
      "epoch": 0.48919753086419754,
      "grad_norm": 1.4796525239944458,
      "learning_rate": 0.0005,
      "loss": 2.538367748260498,
      "step": 2536
    },
    {
      "epoch": 0.4893904320987654,
      "grad_norm": 1.5086902379989624,
      "learning_rate": 0.0005,
      "loss": 2.853060722351074,
      "step": 2537
    },
    {
      "epoch": 0.4895833333333333,
      "grad_norm": 0.9665238857269287,
      "learning_rate": 0.0005,
      "loss": 2.782909393310547,
      "step": 2538
    },
    {
      "epoch": 0.48977623456790126,
      "grad_norm": 0.964240550994873,
      "learning_rate": 0.0005,
      "loss": 2.7690725326538086,
      "step": 2539
    },
    {
      "epoch": 0.48996913580246915,
      "grad_norm": 0.996533989906311,
      "learning_rate": 0.0005,
      "loss": 2.718836545944214,
      "step": 2540
    },
    {
      "epoch": 0.49016203703703703,
      "grad_norm": 1.0072367191314697,
      "learning_rate": 0.0005,
      "loss": 2.679051637649536,
      "step": 2541
    },
    {
      "epoch": 0.4903549382716049,
      "grad_norm": 0.8605825304985046,
      "learning_rate": 0.0005,
      "loss": 2.7050132751464844,
      "step": 2542
    },
    {
      "epoch": 0.49054783950617287,
      "grad_norm": 1.25401771068573,
      "learning_rate": 0.0005,
      "loss": 2.9027514457702637,
      "step": 2543
    },
    {
      "epoch": 0.49074074074074076,
      "grad_norm": 0.990777850151062,
      "learning_rate": 0.0005,
      "loss": 2.46633243560791,
      "step": 2544
    },
    {
      "epoch": 0.49093364197530864,
      "grad_norm": 1.1281838417053223,
      "learning_rate": 0.0005,
      "loss": 3.0057291984558105,
      "step": 2545
    },
    {
      "epoch": 0.49112654320987653,
      "grad_norm": 0.9264094829559326,
      "learning_rate": 0.0005,
      "loss": 2.608096122741699,
      "step": 2546
    },
    {
      "epoch": 0.4913194444444444,
      "grad_norm": 1.1260066032409668,
      "learning_rate": 0.0005,
      "loss": 2.9558823108673096,
      "step": 2547
    },
    {
      "epoch": 0.49151234567901236,
      "grad_norm": 0.7742446660995483,
      "learning_rate": 0.0005,
      "loss": 2.734403133392334,
      "step": 2548
    },
    {
      "epoch": 0.49170524691358025,
      "grad_norm": 1.232106328010559,
      "learning_rate": 0.0005,
      "loss": 2.5618388652801514,
      "step": 2549
    },
    {
      "epoch": 0.49189814814814814,
      "grad_norm": 1.0001693964004517,
      "learning_rate": 0.0005,
      "loss": 2.803091049194336,
      "step": 2550
    },
    {
      "epoch": 0.49209104938271603,
      "grad_norm": 0.9184878468513489,
      "learning_rate": 0.0005,
      "loss": 2.4850661754608154,
      "step": 2551
    },
    {
      "epoch": 0.49228395061728397,
      "grad_norm": 1.108848214149475,
      "learning_rate": 0.0005,
      "loss": 2.7986507415771484,
      "step": 2552
    },
    {
      "epoch": 0.49247685185185186,
      "grad_norm": 1.001368761062622,
      "learning_rate": 0.0005,
      "loss": 2.4357805252075195,
      "step": 2553
    },
    {
      "epoch": 0.49266975308641975,
      "grad_norm": 0.7924949526786804,
      "learning_rate": 0.0005,
      "loss": 2.791952610015869,
      "step": 2554
    },
    {
      "epoch": 0.49286265432098764,
      "grad_norm": 1.5141828060150146,
      "learning_rate": 0.0005,
      "loss": 2.4708003997802734,
      "step": 2555
    },
    {
      "epoch": 0.4930555555555556,
      "grad_norm": 1.0704684257507324,
      "learning_rate": 0.0005,
      "loss": 2.5250744819641113,
      "step": 2556
    },
    {
      "epoch": 0.49324845679012347,
      "grad_norm": 1.1652345657348633,
      "learning_rate": 0.0005,
      "loss": 2.34012508392334,
      "step": 2557
    },
    {
      "epoch": 0.49344135802469136,
      "grad_norm": 0.978335976600647,
      "learning_rate": 0.0005,
      "loss": 3.109306573867798,
      "step": 2558
    },
    {
      "epoch": 0.49363425925925924,
      "grad_norm": 0.9362718462944031,
      "learning_rate": 0.0005,
      "loss": 2.623394250869751,
      "step": 2559
    },
    {
      "epoch": 0.49382716049382713,
      "grad_norm": 1.0497750043869019,
      "learning_rate": 0.0005,
      "loss": 2.5587263107299805,
      "step": 2560
    },
    {
      "epoch": 0.4940200617283951,
      "grad_norm": 0.8710463047027588,
      "learning_rate": 0.0005,
      "loss": 3.021824359893799,
      "step": 2561
    },
    {
      "epoch": 0.49421296296296297,
      "grad_norm": 1.082216501235962,
      "learning_rate": 0.0005,
      "loss": 2.4719009399414062,
      "step": 2562
    },
    {
      "epoch": 0.49440586419753085,
      "grad_norm": 1.1408132314682007,
      "learning_rate": 0.0005,
      "loss": 2.4459950923919678,
      "step": 2563
    },
    {
      "epoch": 0.49459876543209874,
      "grad_norm": 1.2745294570922852,
      "learning_rate": 0.0005,
      "loss": 3.1251354217529297,
      "step": 2564
    },
    {
      "epoch": 0.4947916666666667,
      "grad_norm": 1.0603721141815186,
      "learning_rate": 0.0005,
      "loss": 2.631734848022461,
      "step": 2565
    },
    {
      "epoch": 0.4949845679012346,
      "grad_norm": 1.0917565822601318,
      "learning_rate": 0.0005,
      "loss": 2.6239848136901855,
      "step": 2566
    },
    {
      "epoch": 0.49517746913580246,
      "grad_norm": 1.24776029586792,
      "learning_rate": 0.0005,
      "loss": 2.676227569580078,
      "step": 2567
    },
    {
      "epoch": 0.49537037037037035,
      "grad_norm": 1.1912071704864502,
      "learning_rate": 0.0005,
      "loss": 2.933852195739746,
      "step": 2568
    },
    {
      "epoch": 0.4955632716049383,
      "grad_norm": 1.0730986595153809,
      "learning_rate": 0.0005,
      "loss": 3.035813331604004,
      "step": 2569
    },
    {
      "epoch": 0.4957561728395062,
      "grad_norm": 0.9291107058525085,
      "learning_rate": 0.0005,
      "loss": 2.6297824382781982,
      "step": 2570
    },
    {
      "epoch": 0.49594907407407407,
      "grad_norm": 0.8911339044570923,
      "learning_rate": 0.0005,
      "loss": 2.81646728515625,
      "step": 2571
    },
    {
      "epoch": 0.49614197530864196,
      "grad_norm": 1.0587764978408813,
      "learning_rate": 0.0005,
      "loss": 2.5841338634490967,
      "step": 2572
    },
    {
      "epoch": 0.4963348765432099,
      "grad_norm": 1.052833080291748,
      "learning_rate": 0.0005,
      "loss": 2.554751396179199,
      "step": 2573
    },
    {
      "epoch": 0.4965277777777778,
      "grad_norm": 1.0273418426513672,
      "learning_rate": 0.0005,
      "loss": 2.750589370727539,
      "step": 2574
    },
    {
      "epoch": 0.4967206790123457,
      "grad_norm": 1.3887395858764648,
      "learning_rate": 0.0005,
      "loss": 2.660271406173706,
      "step": 2575
    },
    {
      "epoch": 0.49691358024691357,
      "grad_norm": 0.9381862878799438,
      "learning_rate": 0.0005,
      "loss": 2.644256591796875,
      "step": 2576
    },
    {
      "epoch": 0.49710648148148145,
      "grad_norm": 1.0787280797958374,
      "learning_rate": 0.0005,
      "loss": 2.6464133262634277,
      "step": 2577
    },
    {
      "epoch": 0.4972993827160494,
      "grad_norm": 0.9180630445480347,
      "learning_rate": 0.0005,
      "loss": 2.6613399982452393,
      "step": 2578
    },
    {
      "epoch": 0.4974922839506173,
      "grad_norm": 0.8484407067298889,
      "learning_rate": 0.0005,
      "loss": 2.702289581298828,
      "step": 2579
    },
    {
      "epoch": 0.4976851851851852,
      "grad_norm": 0.9638323187828064,
      "learning_rate": 0.0005,
      "loss": 2.8579626083374023,
      "step": 2580
    },
    {
      "epoch": 0.49787808641975306,
      "grad_norm": 1.2265732288360596,
      "learning_rate": 0.0005,
      "loss": 2.8281455039978027,
      "step": 2581
    },
    {
      "epoch": 0.498070987654321,
      "grad_norm": 1.237972617149353,
      "learning_rate": 0.0005,
      "loss": 2.689816474914551,
      "step": 2582
    },
    {
      "epoch": 0.4982638888888889,
      "grad_norm": 1.0481126308441162,
      "learning_rate": 0.0005,
      "loss": 2.6238131523132324,
      "step": 2583
    },
    {
      "epoch": 0.4984567901234568,
      "grad_norm": 0.9428741335868835,
      "learning_rate": 0.0005,
      "loss": 2.525587797164917,
      "step": 2584
    },
    {
      "epoch": 0.49864969135802467,
      "grad_norm": 1.1434558629989624,
      "learning_rate": 0.0005,
      "loss": 2.7599573135375977,
      "step": 2585
    },
    {
      "epoch": 0.4988425925925926,
      "grad_norm": 1.0219429731369019,
      "learning_rate": 0.0005,
      "loss": 2.7436718940734863,
      "step": 2586
    },
    {
      "epoch": 0.4990354938271605,
      "grad_norm": 0.9176177382469177,
      "learning_rate": 0.0005,
      "loss": 2.707857608795166,
      "step": 2587
    },
    {
      "epoch": 0.4992283950617284,
      "grad_norm": 1.509812355041504,
      "learning_rate": 0.0005,
      "loss": 2.5035996437072754,
      "step": 2588
    },
    {
      "epoch": 0.4994212962962963,
      "grad_norm": 1.0650421380996704,
      "learning_rate": 0.0005,
      "loss": 2.6557676792144775,
      "step": 2589
    },
    {
      "epoch": 0.4996141975308642,
      "grad_norm": 0.8638281226158142,
      "learning_rate": 0.0005,
      "loss": 2.612611770629883,
      "step": 2590
    },
    {
      "epoch": 0.4998070987654321,
      "grad_norm": 1.2343685626983643,
      "learning_rate": 0.0005,
      "loss": 2.8283660411834717,
      "step": 2591
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9325932264328003,
      "learning_rate": 0.0005,
      "loss": 2.5783560276031494,
      "step": 2592
    },
    {
      "epoch": 0.5001929012345679,
      "grad_norm": 1.0578433275222778,
      "learning_rate": 0.0005,
      "loss": 2.7748022079467773,
      "step": 2593
    },
    {
      "epoch": 0.5003858024691358,
      "grad_norm": 0.9470496773719788,
      "learning_rate": 0.0005,
      "loss": 2.738271713256836,
      "step": 2594
    },
    {
      "epoch": 0.5005787037037037,
      "grad_norm": 0.9261208176612854,
      "learning_rate": 0.0005,
      "loss": 2.5370850563049316,
      "step": 2595
    },
    {
      "epoch": 0.5007716049382716,
      "grad_norm": 1.660513997077942,
      "learning_rate": 0.0005,
      "loss": 2.655956745147705,
      "step": 2596
    },
    {
      "epoch": 0.5009645061728395,
      "grad_norm": 1.1459946632385254,
      "learning_rate": 0.0005,
      "loss": 2.8298559188842773,
      "step": 2597
    },
    {
      "epoch": 0.5011574074074074,
      "grad_norm": 1.0712239742279053,
      "learning_rate": 0.0005,
      "loss": 2.3371901512145996,
      "step": 2598
    },
    {
      "epoch": 0.5013503086419753,
      "grad_norm": 1.2123863697052002,
      "learning_rate": 0.0005,
      "loss": 2.584980010986328,
      "step": 2599
    },
    {
      "epoch": 0.5015432098765432,
      "grad_norm": 1.1607638597488403,
      "learning_rate": 0.0005,
      "loss": 2.6390609741210938,
      "step": 2600
    },
    {
      "epoch": 0.5017361111111112,
      "grad_norm": 1.188769817352295,
      "learning_rate": 0.0005,
      "loss": 2.593165397644043,
      "step": 2601
    },
    {
      "epoch": 0.501929012345679,
      "grad_norm": 1.1070972681045532,
      "learning_rate": 0.0005,
      "loss": 2.744508743286133,
      "step": 2602
    },
    {
      "epoch": 0.5021219135802469,
      "grad_norm": 1.1875934600830078,
      "learning_rate": 0.0005,
      "loss": 2.602015733718872,
      "step": 2603
    },
    {
      "epoch": 0.5023148148148148,
      "grad_norm": 0.9615129828453064,
      "learning_rate": 0.0005,
      "loss": 2.477797746658325,
      "step": 2604
    },
    {
      "epoch": 0.5025077160493827,
      "grad_norm": 1.0423029661178589,
      "learning_rate": 0.0005,
      "loss": 2.696894645690918,
      "step": 2605
    },
    {
      "epoch": 0.5027006172839507,
      "grad_norm": 1.373232126235962,
      "learning_rate": 0.0005,
      "loss": 2.5920779705047607,
      "step": 2606
    },
    {
      "epoch": 0.5028935185185185,
      "grad_norm": 0.9239135980606079,
      "learning_rate": 0.0005,
      "loss": 2.9200644493103027,
      "step": 2607
    },
    {
      "epoch": 0.5030864197530864,
      "grad_norm": 1.1195327043533325,
      "learning_rate": 0.0005,
      "loss": 2.661426544189453,
      "step": 2608
    },
    {
      "epoch": 0.5032793209876543,
      "grad_norm": 1.7939279079437256,
      "learning_rate": 0.0005,
      "loss": 2.6147217750549316,
      "step": 2609
    },
    {
      "epoch": 0.5034722222222222,
      "grad_norm": 1.214868187904358,
      "learning_rate": 0.0005,
      "loss": 2.5433743000030518,
      "step": 2610
    },
    {
      "epoch": 0.5036651234567902,
      "grad_norm": 0.89569091796875,
      "learning_rate": 0.0005,
      "loss": 2.726041793823242,
      "step": 2611
    },
    {
      "epoch": 0.503858024691358,
      "grad_norm": 1.0027459859848022,
      "learning_rate": 0.0005,
      "loss": 2.893280267715454,
      "step": 2612
    },
    {
      "epoch": 0.5040509259259259,
      "grad_norm": 0.9607511758804321,
      "learning_rate": 0.0005,
      "loss": 3.219651460647583,
      "step": 2613
    },
    {
      "epoch": 0.5042438271604939,
      "grad_norm": 0.8932380676269531,
      "learning_rate": 0.0005,
      "loss": 2.5230441093444824,
      "step": 2614
    },
    {
      "epoch": 0.5044367283950617,
      "grad_norm": 0.8869888186454773,
      "learning_rate": 0.0005,
      "loss": 2.50795316696167,
      "step": 2615
    },
    {
      "epoch": 0.5046296296296297,
      "grad_norm": 1.0237292051315308,
      "learning_rate": 0.0005,
      "loss": 2.4240543842315674,
      "step": 2616
    },
    {
      "epoch": 0.5048225308641975,
      "grad_norm": 1.1754169464111328,
      "learning_rate": 0.0005,
      "loss": 2.6617259979248047,
      "step": 2617
    },
    {
      "epoch": 0.5050154320987654,
      "grad_norm": 0.9214667081832886,
      "learning_rate": 0.0005,
      "loss": 2.586522102355957,
      "step": 2618
    },
    {
      "epoch": 0.5052083333333334,
      "grad_norm": 0.9681439995765686,
      "learning_rate": 0.0005,
      "loss": 2.6925668716430664,
      "step": 2619
    },
    {
      "epoch": 0.5054012345679012,
      "grad_norm": 0.9556257128715515,
      "learning_rate": 0.0005,
      "loss": 2.803438663482666,
      "step": 2620
    },
    {
      "epoch": 0.5055941358024691,
      "grad_norm": 1.1413419246673584,
      "learning_rate": 0.0005,
      "loss": 2.712955951690674,
      "step": 2621
    },
    {
      "epoch": 0.5057870370370371,
      "grad_norm": 0.9335848093032837,
      "learning_rate": 0.0005,
      "loss": 2.6412434577941895,
      "step": 2622
    },
    {
      "epoch": 0.5059799382716049,
      "grad_norm": 1.0807931423187256,
      "learning_rate": 0.0005,
      "loss": 2.7997946739196777,
      "step": 2623
    },
    {
      "epoch": 0.5061728395061729,
      "grad_norm": 1.289844274520874,
      "learning_rate": 0.0005,
      "loss": 2.388495922088623,
      "step": 2624
    },
    {
      "epoch": 0.5063657407407407,
      "grad_norm": 0.8227277398109436,
      "learning_rate": 0.0005,
      "loss": 2.5934135913848877,
      "step": 2625
    },
    {
      "epoch": 0.5065586419753086,
      "grad_norm": 1.0036228895187378,
      "learning_rate": 0.0005,
      "loss": 2.4098358154296875,
      "step": 2626
    },
    {
      "epoch": 0.5067515432098766,
      "grad_norm": 1.246575951576233,
      "learning_rate": 0.0005,
      "loss": 2.4576826095581055,
      "step": 2627
    },
    {
      "epoch": 0.5069444444444444,
      "grad_norm": 1.1494077444076538,
      "learning_rate": 0.0005,
      "loss": 2.8535876274108887,
      "step": 2628
    },
    {
      "epoch": 0.5071373456790124,
      "grad_norm": 0.9682751297950745,
      "learning_rate": 0.0005,
      "loss": 2.8650267124176025,
      "step": 2629
    },
    {
      "epoch": 0.5073302469135802,
      "grad_norm": 0.9134123921394348,
      "learning_rate": 0.0005,
      "loss": 2.657205104827881,
      "step": 2630
    },
    {
      "epoch": 0.5075231481481481,
      "grad_norm": 2.1310653686523438,
      "learning_rate": 0.0005,
      "loss": 3.1818647384643555,
      "step": 2631
    },
    {
      "epoch": 0.5077160493827161,
      "grad_norm": 1.1907165050506592,
      "learning_rate": 0.0005,
      "loss": 2.9913909435272217,
      "step": 2632
    },
    {
      "epoch": 0.5079089506172839,
      "grad_norm": 1.096994161605835,
      "learning_rate": 0.0005,
      "loss": 2.896676778793335,
      "step": 2633
    },
    {
      "epoch": 0.5081018518518519,
      "grad_norm": 0.8893288373947144,
      "learning_rate": 0.0005,
      "loss": 2.884314775466919,
      "step": 2634
    },
    {
      "epoch": 0.5082947530864198,
      "grad_norm": 0.8582719564437866,
      "learning_rate": 0.0005,
      "loss": 2.5775749683380127,
      "step": 2635
    },
    {
      "epoch": 0.5084876543209876,
      "grad_norm": 0.9656723141670227,
      "learning_rate": 0.0005,
      "loss": 2.726569175720215,
      "step": 2636
    },
    {
      "epoch": 0.5086805555555556,
      "grad_norm": 0.9836460947990417,
      "learning_rate": 0.0005,
      "loss": 2.5681939125061035,
      "step": 2637
    },
    {
      "epoch": 0.5088734567901234,
      "grad_norm": 1.0852382183074951,
      "learning_rate": 0.0005,
      "loss": 2.770260810852051,
      "step": 2638
    },
    {
      "epoch": 0.5090663580246914,
      "grad_norm": 0.7895728349685669,
      "learning_rate": 0.0005,
      "loss": 2.4547276496887207,
      "step": 2639
    },
    {
      "epoch": 0.5092592592592593,
      "grad_norm": 1.1373610496520996,
      "learning_rate": 0.0005,
      "loss": 2.5217366218566895,
      "step": 2640
    },
    {
      "epoch": 0.5094521604938271,
      "grad_norm": 0.9813607931137085,
      "learning_rate": 0.0005,
      "loss": 2.9139351844787598,
      "step": 2641
    },
    {
      "epoch": 0.5096450617283951,
      "grad_norm": 0.7505521178245544,
      "learning_rate": 0.0005,
      "loss": 2.6185390949249268,
      "step": 2642
    },
    {
      "epoch": 0.5098379629629629,
      "grad_norm": 1.032511591911316,
      "learning_rate": 0.0005,
      "loss": 2.7301025390625,
      "step": 2643
    },
    {
      "epoch": 0.5100308641975309,
      "grad_norm": 0.8568520545959473,
      "learning_rate": 0.0005,
      "loss": 2.7779600620269775,
      "step": 2644
    },
    {
      "epoch": 0.5102237654320988,
      "grad_norm": 0.9689329862594604,
      "learning_rate": 0.0005,
      "loss": 2.796302318572998,
      "step": 2645
    },
    {
      "epoch": 0.5104166666666666,
      "grad_norm": 0.833884596824646,
      "learning_rate": 0.0005,
      "loss": 2.5778450965881348,
      "step": 2646
    },
    {
      "epoch": 0.5106095679012346,
      "grad_norm": 1.1197234392166138,
      "learning_rate": 0.0005,
      "loss": 2.833120346069336,
      "step": 2647
    },
    {
      "epoch": 0.5108024691358025,
      "grad_norm": 0.7772849202156067,
      "learning_rate": 0.0005,
      "loss": 2.574754476547241,
      "step": 2648
    },
    {
      "epoch": 0.5109953703703703,
      "grad_norm": 0.8663337826728821,
      "learning_rate": 0.0005,
      "loss": 2.574697494506836,
      "step": 2649
    },
    {
      "epoch": 0.5111882716049383,
      "grad_norm": 0.8593010902404785,
      "learning_rate": 0.0005,
      "loss": 2.8009865283966064,
      "step": 2650
    },
    {
      "epoch": 0.5113811728395061,
      "grad_norm": 0.7869443297386169,
      "learning_rate": 0.0005,
      "loss": 2.4349944591522217,
      "step": 2651
    },
    {
      "epoch": 0.5115740740740741,
      "grad_norm": 0.7869073152542114,
      "learning_rate": 0.0005,
      "loss": 2.443479061126709,
      "step": 2652
    },
    {
      "epoch": 0.511766975308642,
      "grad_norm": 0.7973924279212952,
      "learning_rate": 0.0005,
      "loss": 2.456167221069336,
      "step": 2653
    },
    {
      "epoch": 0.5119598765432098,
      "grad_norm": 1.3408918380737305,
      "learning_rate": 0.0005,
      "loss": 2.6786398887634277,
      "step": 2654
    },
    {
      "epoch": 0.5121527777777778,
      "grad_norm": 0.8815799951553345,
      "learning_rate": 0.0005,
      "loss": 2.694584608078003,
      "step": 2655
    },
    {
      "epoch": 0.5123456790123457,
      "grad_norm": 1.2213990688323975,
      "learning_rate": 0.0005,
      "loss": 2.695814609527588,
      "step": 2656
    },
    {
      "epoch": 0.5125385802469136,
      "grad_norm": 1.7630085945129395,
      "learning_rate": 0.0005,
      "loss": 2.8231513500213623,
      "step": 2657
    },
    {
      "epoch": 0.5127314814814815,
      "grad_norm": 1.0734779834747314,
      "learning_rate": 0.0005,
      "loss": 2.754382610321045,
      "step": 2658
    },
    {
      "epoch": 0.5129243827160493,
      "grad_norm": 0.8241650462150574,
      "learning_rate": 0.0005,
      "loss": 2.534731388092041,
      "step": 2659
    },
    {
      "epoch": 0.5131172839506173,
      "grad_norm": 0.8136676549911499,
      "learning_rate": 0.0005,
      "loss": 2.7947192192077637,
      "step": 2660
    },
    {
      "epoch": 0.5133101851851852,
      "grad_norm": 1.1213239431381226,
      "learning_rate": 0.0005,
      "loss": 2.771939277648926,
      "step": 2661
    },
    {
      "epoch": 0.5135030864197531,
      "grad_norm": 1.5558708906173706,
      "learning_rate": 0.0005,
      "loss": 3.1268372535705566,
      "step": 2662
    },
    {
      "epoch": 0.513695987654321,
      "grad_norm": 0.9183586835861206,
      "learning_rate": 0.0005,
      "loss": 2.8179447650909424,
      "step": 2663
    },
    {
      "epoch": 0.5138888888888888,
      "grad_norm": 1.0500361919403076,
      "learning_rate": 0.0005,
      "loss": 2.567690849304199,
      "step": 2664
    },
    {
      "epoch": 0.5140817901234568,
      "grad_norm": 1.4361549615859985,
      "learning_rate": 0.0005,
      "loss": 2.736572265625,
      "step": 2665
    },
    {
      "epoch": 0.5142746913580247,
      "grad_norm": 1.0186270475387573,
      "learning_rate": 0.0005,
      "loss": 2.5957021713256836,
      "step": 2666
    },
    {
      "epoch": 0.5144675925925926,
      "grad_norm": 1.1316128969192505,
      "learning_rate": 0.0005,
      "loss": 2.8323824405670166,
      "step": 2667
    },
    {
      "epoch": 0.5146604938271605,
      "grad_norm": 1.1319947242736816,
      "learning_rate": 0.0005,
      "loss": 3.049736261367798,
      "step": 2668
    },
    {
      "epoch": 0.5148533950617284,
      "grad_norm": 0.9218708872795105,
      "learning_rate": 0.0005,
      "loss": 2.7133679389953613,
      "step": 2669
    },
    {
      "epoch": 0.5150462962962963,
      "grad_norm": 0.8314381837844849,
      "learning_rate": 0.0005,
      "loss": 2.5504794120788574,
      "step": 2670
    },
    {
      "epoch": 0.5152391975308642,
      "grad_norm": 0.9314095973968506,
      "learning_rate": 0.0005,
      "loss": 2.962790012359619,
      "step": 2671
    },
    {
      "epoch": 0.5154320987654321,
      "grad_norm": 1.2696374654769897,
      "learning_rate": 0.0005,
      "loss": 2.7172021865844727,
      "step": 2672
    },
    {
      "epoch": 0.515625,
      "grad_norm": 1.0551631450653076,
      "learning_rate": 0.0005,
      "loss": 2.7806711196899414,
      "step": 2673
    },
    {
      "epoch": 0.5158179012345679,
      "grad_norm": 0.9722585082054138,
      "learning_rate": 0.0005,
      "loss": 2.8612987995147705,
      "step": 2674
    },
    {
      "epoch": 0.5160108024691358,
      "grad_norm": 0.8741613030433655,
      "learning_rate": 0.0005,
      "loss": 2.5617313385009766,
      "step": 2675
    },
    {
      "epoch": 0.5162037037037037,
      "grad_norm": 0.9567078948020935,
      "learning_rate": 0.0005,
      "loss": 2.682217836380005,
      "step": 2676
    },
    {
      "epoch": 0.5163966049382716,
      "grad_norm": 1.3206567764282227,
      "learning_rate": 0.0005,
      "loss": 3.0923233032226562,
      "step": 2677
    },
    {
      "epoch": 0.5165895061728395,
      "grad_norm": 1.3101860284805298,
      "learning_rate": 0.0005,
      "loss": 2.937169075012207,
      "step": 2678
    },
    {
      "epoch": 0.5167824074074074,
      "grad_norm": 1.391945242881775,
      "learning_rate": 0.0005,
      "loss": 2.3384504318237305,
      "step": 2679
    },
    {
      "epoch": 0.5169753086419753,
      "grad_norm": 1.1132102012634277,
      "learning_rate": 0.0005,
      "loss": 2.899339199066162,
      "step": 2680
    },
    {
      "epoch": 0.5171682098765432,
      "grad_norm": 1.3513373136520386,
      "learning_rate": 0.0005,
      "loss": 2.711350917816162,
      "step": 2681
    },
    {
      "epoch": 0.5173611111111112,
      "grad_norm": 0.9098244309425354,
      "learning_rate": 0.0005,
      "loss": 2.8179821968078613,
      "step": 2682
    },
    {
      "epoch": 0.517554012345679,
      "grad_norm": 1.0267701148986816,
      "learning_rate": 0.0005,
      "loss": 2.64351487159729,
      "step": 2683
    },
    {
      "epoch": 0.5177469135802469,
      "grad_norm": 0.836595356464386,
      "learning_rate": 0.0005,
      "loss": 2.647451400756836,
      "step": 2684
    },
    {
      "epoch": 0.5179398148148148,
      "grad_norm": 0.9993710517883301,
      "learning_rate": 0.0005,
      "loss": 2.635268211364746,
      "step": 2685
    },
    {
      "epoch": 0.5181327160493827,
      "grad_norm": 0.9267498254776001,
      "learning_rate": 0.0005,
      "loss": 2.583852529525757,
      "step": 2686
    },
    {
      "epoch": 0.5183256172839507,
      "grad_norm": 0.8461198210716248,
      "learning_rate": 0.0005,
      "loss": 2.6241488456726074,
      "step": 2687
    },
    {
      "epoch": 0.5185185185185185,
      "grad_norm": 0.9551425576210022,
      "learning_rate": 0.0005,
      "loss": 2.5104455947875977,
      "step": 2688
    },
    {
      "epoch": 0.5187114197530864,
      "grad_norm": 0.8995435237884521,
      "learning_rate": 0.0005,
      "loss": 2.5736560821533203,
      "step": 2689
    },
    {
      "epoch": 0.5189043209876543,
      "grad_norm": 0.886172890663147,
      "learning_rate": 0.0005,
      "loss": 2.416830539703369,
      "step": 2690
    },
    {
      "epoch": 0.5190972222222222,
      "grad_norm": 0.7724917531013489,
      "learning_rate": 0.0005,
      "loss": 2.4431204795837402,
      "step": 2691
    },
    {
      "epoch": 0.5192901234567902,
      "grad_norm": 0.9245110154151917,
      "learning_rate": 0.0005,
      "loss": 2.829036235809326,
      "step": 2692
    },
    {
      "epoch": 0.519483024691358,
      "grad_norm": 1.0115001201629639,
      "learning_rate": 0.0005,
      "loss": 2.864873170852661,
      "step": 2693
    },
    {
      "epoch": 0.5196759259259259,
      "grad_norm": 0.8721758723258972,
      "learning_rate": 0.0005,
      "loss": 2.471205234527588,
      "step": 2694
    },
    {
      "epoch": 0.5198688271604939,
      "grad_norm": 0.8937262892723083,
      "learning_rate": 0.0005,
      "loss": 2.8331892490386963,
      "step": 2695
    },
    {
      "epoch": 0.5200617283950617,
      "grad_norm": 1.22829270362854,
      "learning_rate": 0.0005,
      "loss": 2.649834394454956,
      "step": 2696
    },
    {
      "epoch": 0.5202546296296297,
      "grad_norm": 1.3264544010162354,
      "learning_rate": 0.0005,
      "loss": 2.781919479370117,
      "step": 2697
    },
    {
      "epoch": 0.5204475308641975,
      "grad_norm": 1.3136626482009888,
      "learning_rate": 0.0005,
      "loss": 2.608919858932495,
      "step": 2698
    },
    {
      "epoch": 0.5206404320987654,
      "grad_norm": 1.7783770561218262,
      "learning_rate": 0.0005,
      "loss": 2.6761512756347656,
      "step": 2699
    },
    {
      "epoch": 0.5208333333333334,
      "grad_norm": 0.9123928546905518,
      "learning_rate": 0.0005,
      "loss": 2.571798801422119,
      "step": 2700
    },
    {
      "epoch": 0.5210262345679012,
      "grad_norm": 0.9832872152328491,
      "learning_rate": 0.0005,
      "loss": 2.378404378890991,
      "step": 2701
    },
    {
      "epoch": 0.5212191358024691,
      "grad_norm": 0.9709402918815613,
      "learning_rate": 0.0005,
      "loss": 2.4883041381835938,
      "step": 2702
    },
    {
      "epoch": 0.5214120370370371,
      "grad_norm": 0.846027672290802,
      "learning_rate": 0.0005,
      "loss": 2.7293589115142822,
      "step": 2703
    },
    {
      "epoch": 0.5216049382716049,
      "grad_norm": 0.8590870499610901,
      "learning_rate": 0.0005,
      "loss": 2.7244701385498047,
      "step": 2704
    },
    {
      "epoch": 0.5217978395061729,
      "grad_norm": 1.4996544122695923,
      "learning_rate": 0.0005,
      "loss": 2.4741244316101074,
      "step": 2705
    },
    {
      "epoch": 0.5219907407407407,
      "grad_norm": 1.3932757377624512,
      "learning_rate": 0.0005,
      "loss": 2.7671360969543457,
      "step": 2706
    },
    {
      "epoch": 0.5221836419753086,
      "grad_norm": 1.2996923923492432,
      "learning_rate": 0.0005,
      "loss": 2.8475301265716553,
      "step": 2707
    },
    {
      "epoch": 0.5223765432098766,
      "grad_norm": 1.2655022144317627,
      "learning_rate": 0.0005,
      "loss": 2.865172863006592,
      "step": 2708
    },
    {
      "epoch": 0.5225694444444444,
      "grad_norm": 0.9661491513252258,
      "learning_rate": 0.0005,
      "loss": 2.6028261184692383,
      "step": 2709
    },
    {
      "epoch": 0.5227623456790124,
      "grad_norm": 0.8870083093643188,
      "learning_rate": 0.0005,
      "loss": 2.4773285388946533,
      "step": 2710
    },
    {
      "epoch": 0.5229552469135802,
      "grad_norm": 0.9049403071403503,
      "learning_rate": 0.0005,
      "loss": 2.483950614929199,
      "step": 2711
    },
    {
      "epoch": 0.5231481481481481,
      "grad_norm": 0.9683153629302979,
      "learning_rate": 0.0005,
      "loss": 2.427774429321289,
      "step": 2712
    },
    {
      "epoch": 0.5233410493827161,
      "grad_norm": 1.0138287544250488,
      "learning_rate": 0.0005,
      "loss": 2.7985823154449463,
      "step": 2713
    },
    {
      "epoch": 0.5235339506172839,
      "grad_norm": 1.0211596488952637,
      "learning_rate": 0.0005,
      "loss": 2.4440758228302,
      "step": 2714
    },
    {
      "epoch": 0.5237268518518519,
      "grad_norm": 0.8747031688690186,
      "learning_rate": 0.0005,
      "loss": 2.5978269577026367,
      "step": 2715
    },
    {
      "epoch": 0.5239197530864198,
      "grad_norm": 0.9093993902206421,
      "learning_rate": 0.0005,
      "loss": 2.429804801940918,
      "step": 2716
    },
    {
      "epoch": 0.5241126543209876,
      "grad_norm": 0.9912332892417908,
      "learning_rate": 0.0005,
      "loss": 2.6839828491210938,
      "step": 2717
    },
    {
      "epoch": 0.5243055555555556,
      "grad_norm": 0.9965162873268127,
      "learning_rate": 0.0005,
      "loss": 2.695218086242676,
      "step": 2718
    },
    {
      "epoch": 0.5244984567901234,
      "grad_norm": 1.0134668350219727,
      "learning_rate": 0.0005,
      "loss": 2.821805953979492,
      "step": 2719
    },
    {
      "epoch": 0.5246913580246914,
      "grad_norm": 1.0846388339996338,
      "learning_rate": 0.0005,
      "loss": 2.596996307373047,
      "step": 2720
    },
    {
      "epoch": 0.5248842592592593,
      "grad_norm": 1.3753474950790405,
      "learning_rate": 0.0005,
      "loss": 2.6813416481018066,
      "step": 2721
    },
    {
      "epoch": 0.5250771604938271,
      "grad_norm": 1.304664134979248,
      "learning_rate": 0.0005,
      "loss": 2.60012149810791,
      "step": 2722
    },
    {
      "epoch": 0.5252700617283951,
      "grad_norm": 0.8880352973937988,
      "learning_rate": 0.0005,
      "loss": 2.7255115509033203,
      "step": 2723
    },
    {
      "epoch": 0.5254629629629629,
      "grad_norm": 0.9762819409370422,
      "learning_rate": 0.0005,
      "loss": 2.3447518348693848,
      "step": 2724
    },
    {
      "epoch": 0.5256558641975309,
      "grad_norm": 0.9170584678649902,
      "learning_rate": 0.0005,
      "loss": 2.6577577590942383,
      "step": 2725
    },
    {
      "epoch": 0.5258487654320988,
      "grad_norm": 1.0992538928985596,
      "learning_rate": 0.0005,
      "loss": 2.7252020835876465,
      "step": 2726
    },
    {
      "epoch": 0.5260416666666666,
      "grad_norm": 0.9607126116752625,
      "learning_rate": 0.0005,
      "loss": 2.520162582397461,
      "step": 2727
    },
    {
      "epoch": 0.5262345679012346,
      "grad_norm": 0.9207927584648132,
      "learning_rate": 0.0005,
      "loss": 2.311033248901367,
      "step": 2728
    },
    {
      "epoch": 0.5264274691358025,
      "grad_norm": 0.8769688606262207,
      "learning_rate": 0.0005,
      "loss": 2.6028127670288086,
      "step": 2729
    },
    {
      "epoch": 0.5266203703703703,
      "grad_norm": 0.9479060173034668,
      "learning_rate": 0.0005,
      "loss": 2.6518325805664062,
      "step": 2730
    },
    {
      "epoch": 0.5268132716049383,
      "grad_norm": 1.0178492069244385,
      "learning_rate": 0.0005,
      "loss": 2.7053675651550293,
      "step": 2731
    },
    {
      "epoch": 0.5270061728395061,
      "grad_norm": 1.064923882484436,
      "learning_rate": 0.0005,
      "loss": 2.9165267944335938,
      "step": 2732
    },
    {
      "epoch": 0.5271990740740741,
      "grad_norm": 0.9130475521087646,
      "learning_rate": 0.0005,
      "loss": 2.6349871158599854,
      "step": 2733
    },
    {
      "epoch": 0.527391975308642,
      "grad_norm": 0.9593004584312439,
      "learning_rate": 0.0005,
      "loss": 2.5381734371185303,
      "step": 2734
    },
    {
      "epoch": 0.5275848765432098,
      "grad_norm": 1.5832891464233398,
      "learning_rate": 0.0005,
      "loss": 2.5632729530334473,
      "step": 2735
    },
    {
      "epoch": 0.5277777777777778,
      "grad_norm": 1.0324558019638062,
      "learning_rate": 0.0005,
      "loss": 2.8692526817321777,
      "step": 2736
    },
    {
      "epoch": 0.5279706790123457,
      "grad_norm": 0.8274819850921631,
      "learning_rate": 0.0005,
      "loss": 2.670553207397461,
      "step": 2737
    },
    {
      "epoch": 0.5281635802469136,
      "grad_norm": 1.213281273841858,
      "learning_rate": 0.0005,
      "loss": 2.7167811393737793,
      "step": 2738
    },
    {
      "epoch": 0.5283564814814815,
      "grad_norm": 1.0825518369674683,
      "learning_rate": 0.0005,
      "loss": 2.631490707397461,
      "step": 2739
    },
    {
      "epoch": 0.5285493827160493,
      "grad_norm": 1.3302897214889526,
      "learning_rate": 0.0005,
      "loss": 2.8132479190826416,
      "step": 2740
    },
    {
      "epoch": 0.5287422839506173,
      "grad_norm": 1.105326771736145,
      "learning_rate": 0.0005,
      "loss": 2.4857254028320312,
      "step": 2741
    },
    {
      "epoch": 0.5289351851851852,
      "grad_norm": 1.0831555128097534,
      "learning_rate": 0.0005,
      "loss": 2.990117311477661,
      "step": 2742
    },
    {
      "epoch": 0.5291280864197531,
      "grad_norm": 1.2891628742218018,
      "learning_rate": 0.0005,
      "loss": 2.7541985511779785,
      "step": 2743
    },
    {
      "epoch": 0.529320987654321,
      "grad_norm": 1.2976939678192139,
      "learning_rate": 0.0005,
      "loss": 3.0217435359954834,
      "step": 2744
    },
    {
      "epoch": 0.5295138888888888,
      "grad_norm": 0.9716474413871765,
      "learning_rate": 0.0005,
      "loss": 2.510408639907837,
      "step": 2745
    },
    {
      "epoch": 0.5297067901234568,
      "grad_norm": 1.090166687965393,
      "learning_rate": 0.0005,
      "loss": 2.5549097061157227,
      "step": 2746
    },
    {
      "epoch": 0.5298996913580247,
      "grad_norm": 1.22713041305542,
      "learning_rate": 0.0005,
      "loss": 2.8348612785339355,
      "step": 2747
    },
    {
      "epoch": 0.5300925925925926,
      "grad_norm": 1.0868159532546997,
      "learning_rate": 0.0005,
      "loss": 2.9911530017852783,
      "step": 2748
    },
    {
      "epoch": 0.5302854938271605,
      "grad_norm": 0.8694970607757568,
      "learning_rate": 0.0005,
      "loss": 2.6371304988861084,
      "step": 2749
    },
    {
      "epoch": 0.5304783950617284,
      "grad_norm": 1.0529004335403442,
      "learning_rate": 0.0005,
      "loss": 2.8258328437805176,
      "step": 2750
    },
    {
      "epoch": 0.5306712962962963,
      "grad_norm": 0.9896425604820251,
      "learning_rate": 0.0005,
      "loss": 2.1667048931121826,
      "step": 2751
    },
    {
      "epoch": 0.5308641975308642,
      "grad_norm": 0.8914520740509033,
      "learning_rate": 0.0005,
      "loss": 2.941035509109497,
      "step": 2752
    },
    {
      "epoch": 0.5310570987654321,
      "grad_norm": 1.0530465841293335,
      "learning_rate": 0.0005,
      "loss": 2.7375643253326416,
      "step": 2753
    },
    {
      "epoch": 0.53125,
      "grad_norm": 0.9097388982772827,
      "learning_rate": 0.0005,
      "loss": 2.6026864051818848,
      "step": 2754
    },
    {
      "epoch": 0.5314429012345679,
      "grad_norm": 1.0235717296600342,
      "learning_rate": 0.0005,
      "loss": 2.6674795150756836,
      "step": 2755
    },
    {
      "epoch": 0.5316358024691358,
      "grad_norm": 1.2490888833999634,
      "learning_rate": 0.0005,
      "loss": 2.2699782848358154,
      "step": 2756
    },
    {
      "epoch": 0.5318287037037037,
      "grad_norm": 0.8265407681465149,
      "learning_rate": 0.0005,
      "loss": 2.6542654037475586,
      "step": 2757
    },
    {
      "epoch": 0.5320216049382716,
      "grad_norm": 1.0354413986206055,
      "learning_rate": 0.0005,
      "loss": 2.7645750045776367,
      "step": 2758
    },
    {
      "epoch": 0.5322145061728395,
      "grad_norm": 1.0426857471466064,
      "learning_rate": 0.0005,
      "loss": 2.783705472946167,
      "step": 2759
    },
    {
      "epoch": 0.5324074074074074,
      "grad_norm": 1.0057343244552612,
      "learning_rate": 0.0005,
      "loss": 2.2064833641052246,
      "step": 2760
    },
    {
      "epoch": 0.5326003086419753,
      "grad_norm": 1.0876129865646362,
      "learning_rate": 0.0005,
      "loss": 2.6776862144470215,
      "step": 2761
    },
    {
      "epoch": 0.5327932098765432,
      "grad_norm": 1.2098991870880127,
      "learning_rate": 0.0005,
      "loss": 2.6283607482910156,
      "step": 2762
    },
    {
      "epoch": 0.5329861111111112,
      "grad_norm": 1.0574395656585693,
      "learning_rate": 0.0005,
      "loss": 2.673553943634033,
      "step": 2763
    },
    {
      "epoch": 0.533179012345679,
      "grad_norm": 0.9062712788581848,
      "learning_rate": 0.0005,
      "loss": 2.544098138809204,
      "step": 2764
    },
    {
      "epoch": 0.5333719135802469,
      "grad_norm": 0.9388989210128784,
      "learning_rate": 0.0005,
      "loss": 2.4642670154571533,
      "step": 2765
    },
    {
      "epoch": 0.5335648148148148,
      "grad_norm": 1.225743293762207,
      "learning_rate": 0.0005,
      "loss": 2.4348087310791016,
      "step": 2766
    },
    {
      "epoch": 0.5337577160493827,
      "grad_norm": 2.4181101322174072,
      "learning_rate": 0.0005,
      "loss": 2.4504737854003906,
      "step": 2767
    },
    {
      "epoch": 0.5339506172839507,
      "grad_norm": 1.2470062971115112,
      "learning_rate": 0.0005,
      "loss": 2.5573086738586426,
      "step": 2768
    },
    {
      "epoch": 0.5341435185185185,
      "grad_norm": 0.9936566948890686,
      "learning_rate": 0.0005,
      "loss": 2.847041130065918,
      "step": 2769
    },
    {
      "epoch": 0.5343364197530864,
      "grad_norm": 1.1379669904708862,
      "learning_rate": 0.0005,
      "loss": 2.8263096809387207,
      "step": 2770
    },
    {
      "epoch": 0.5345293209876543,
      "grad_norm": 0.928780198097229,
      "learning_rate": 0.0005,
      "loss": 2.589557647705078,
      "step": 2771
    },
    {
      "epoch": 0.5347222222222222,
      "grad_norm": 1.2834675312042236,
      "learning_rate": 0.0005,
      "loss": 2.669341564178467,
      "step": 2772
    },
    {
      "epoch": 0.5349151234567902,
      "grad_norm": 1.0224437713623047,
      "learning_rate": 0.0005,
      "loss": 2.656571865081787,
      "step": 2773
    },
    {
      "epoch": 0.535108024691358,
      "grad_norm": 0.8432079553604126,
      "learning_rate": 0.0005,
      "loss": 2.7256007194519043,
      "step": 2774
    },
    {
      "epoch": 0.5353009259259259,
      "grad_norm": 0.900712788105011,
      "learning_rate": 0.0005,
      "loss": 2.8267323970794678,
      "step": 2775
    },
    {
      "epoch": 0.5354938271604939,
      "grad_norm": 1.0702160596847534,
      "learning_rate": 0.0005,
      "loss": 2.8894147872924805,
      "step": 2776
    },
    {
      "epoch": 0.5356867283950617,
      "grad_norm": 1.0500307083129883,
      "learning_rate": 0.0005,
      "loss": 2.755882978439331,
      "step": 2777
    },
    {
      "epoch": 0.5358796296296297,
      "grad_norm": 1.57016921043396,
      "learning_rate": 0.0005,
      "loss": 2.556756019592285,
      "step": 2778
    },
    {
      "epoch": 0.5360725308641975,
      "grad_norm": 1.1131055355072021,
      "learning_rate": 0.0005,
      "loss": 2.6606032848358154,
      "step": 2779
    },
    {
      "epoch": 0.5362654320987654,
      "grad_norm": 1.0583198070526123,
      "learning_rate": 0.0005,
      "loss": 2.6224255561828613,
      "step": 2780
    },
    {
      "epoch": 0.5364583333333334,
      "grad_norm": 1.042163610458374,
      "learning_rate": 0.0005,
      "loss": 2.9540200233459473,
      "step": 2781
    },
    {
      "epoch": 0.5366512345679012,
      "grad_norm": 0.8490104079246521,
      "learning_rate": 0.0005,
      "loss": 2.766359806060791,
      "step": 2782
    },
    {
      "epoch": 0.5368441358024691,
      "grad_norm": 0.8426582217216492,
      "learning_rate": 0.0005,
      "loss": 2.9103386402130127,
      "step": 2783
    },
    {
      "epoch": 0.5370370370370371,
      "grad_norm": 1.1730788946151733,
      "learning_rate": 0.0005,
      "loss": 2.773902177810669,
      "step": 2784
    },
    {
      "epoch": 0.5372299382716049,
      "grad_norm": 0.9901508688926697,
      "learning_rate": 0.0005,
      "loss": 2.6827714443206787,
      "step": 2785
    },
    {
      "epoch": 0.5374228395061729,
      "grad_norm": 2.586308717727661,
      "learning_rate": 0.0005,
      "loss": 2.545682430267334,
      "step": 2786
    },
    {
      "epoch": 0.5376157407407407,
      "grad_norm": 1.0430729389190674,
      "learning_rate": 0.0005,
      "loss": 2.5574095249176025,
      "step": 2787
    },
    {
      "epoch": 0.5378086419753086,
      "grad_norm": 0.9712969660758972,
      "learning_rate": 0.0005,
      "loss": 2.7000973224639893,
      "step": 2788
    },
    {
      "epoch": 0.5380015432098766,
      "grad_norm": 0.9655336737632751,
      "learning_rate": 0.0005,
      "loss": 2.388092041015625,
      "step": 2789
    },
    {
      "epoch": 0.5381944444444444,
      "grad_norm": 0.9685794711112976,
      "learning_rate": 0.0005,
      "loss": 2.4137730598449707,
      "step": 2790
    },
    {
      "epoch": 0.5383873456790124,
      "grad_norm": 0.8496694564819336,
      "learning_rate": 0.0005,
      "loss": 2.43544602394104,
      "step": 2791
    },
    {
      "epoch": 0.5385802469135802,
      "grad_norm": 0.8931374549865723,
      "learning_rate": 0.0005,
      "loss": 2.627553939819336,
      "step": 2792
    },
    {
      "epoch": 0.5387731481481481,
      "grad_norm": 0.9324349761009216,
      "learning_rate": 0.0005,
      "loss": 2.5255918502807617,
      "step": 2793
    },
    {
      "epoch": 0.5389660493827161,
      "grad_norm": 0.87099289894104,
      "learning_rate": 0.0005,
      "loss": 2.759438991546631,
      "step": 2794
    },
    {
      "epoch": 0.5391589506172839,
      "grad_norm": 0.8626630902290344,
      "learning_rate": 0.0005,
      "loss": 2.4205360412597656,
      "step": 2795
    },
    {
      "epoch": 0.5393518518518519,
      "grad_norm": 0.8745441436767578,
      "learning_rate": 0.0005,
      "loss": 2.644685745239258,
      "step": 2796
    },
    {
      "epoch": 0.5395447530864198,
      "grad_norm": 0.9548233151435852,
      "learning_rate": 0.0005,
      "loss": 2.5699174404144287,
      "step": 2797
    },
    {
      "epoch": 0.5397376543209876,
      "grad_norm": 1.1907957792282104,
      "learning_rate": 0.0005,
      "loss": 2.6339073181152344,
      "step": 2798
    },
    {
      "epoch": 0.5399305555555556,
      "grad_norm": 0.8158191442489624,
      "learning_rate": 0.0005,
      "loss": 2.570683479309082,
      "step": 2799
    },
    {
      "epoch": 0.5401234567901234,
      "grad_norm": 1.3786786794662476,
      "learning_rate": 0.0005,
      "loss": 2.5545578002929688,
      "step": 2800
    },
    {
      "epoch": 0.5403163580246914,
      "grad_norm": 1.030861735343933,
      "learning_rate": 0.0005,
      "loss": 2.6357290744781494,
      "step": 2801
    },
    {
      "epoch": 0.5405092592592593,
      "grad_norm": 1.0594221353530884,
      "learning_rate": 0.0005,
      "loss": 2.502242088317871,
      "step": 2802
    },
    {
      "epoch": 0.5407021604938271,
      "grad_norm": 0.9403478503227234,
      "learning_rate": 0.0005,
      "loss": 2.7333824634552,
      "step": 2803
    },
    {
      "epoch": 0.5408950617283951,
      "grad_norm": 0.9424228072166443,
      "learning_rate": 0.0005,
      "loss": 2.393733501434326,
      "step": 2804
    },
    {
      "epoch": 0.5410879629629629,
      "grad_norm": 1.0457282066345215,
      "learning_rate": 0.0005,
      "loss": 2.9364407062530518,
      "step": 2805
    },
    {
      "epoch": 0.5412808641975309,
      "grad_norm": 1.0228245258331299,
      "learning_rate": 0.0005,
      "loss": 2.8322670459747314,
      "step": 2806
    },
    {
      "epoch": 0.5414737654320988,
      "grad_norm": 0.9208177924156189,
      "learning_rate": 0.0005,
      "loss": 2.8290398120880127,
      "step": 2807
    },
    {
      "epoch": 0.5416666666666666,
      "grad_norm": 0.9547954201698303,
      "learning_rate": 0.0005,
      "loss": 2.695673704147339,
      "step": 2808
    },
    {
      "epoch": 0.5418595679012346,
      "grad_norm": 0.896294355392456,
      "learning_rate": 0.0005,
      "loss": 2.523909568786621,
      "step": 2809
    },
    {
      "epoch": 0.5420524691358025,
      "grad_norm": 0.8737148642539978,
      "learning_rate": 0.0005,
      "loss": 2.646989345550537,
      "step": 2810
    },
    {
      "epoch": 0.5422453703703703,
      "grad_norm": 0.8801800608634949,
      "learning_rate": 0.0005,
      "loss": 2.5937325954437256,
      "step": 2811
    },
    {
      "epoch": 0.5424382716049383,
      "grad_norm": 1.0804433822631836,
      "learning_rate": 0.0005,
      "loss": 2.9118311405181885,
      "step": 2812
    },
    {
      "epoch": 0.5426311728395061,
      "grad_norm": 0.9048448801040649,
      "learning_rate": 0.0005,
      "loss": 2.792558431625366,
      "step": 2813
    },
    {
      "epoch": 0.5428240740740741,
      "grad_norm": 1.153786063194275,
      "learning_rate": 0.0005,
      "loss": 2.531599998474121,
      "step": 2814
    },
    {
      "epoch": 0.543016975308642,
      "grad_norm": 0.9819366335868835,
      "learning_rate": 0.0005,
      "loss": 2.589221954345703,
      "step": 2815
    },
    {
      "epoch": 0.5432098765432098,
      "grad_norm": 0.9496549367904663,
      "learning_rate": 0.0005,
      "loss": 2.452112913131714,
      "step": 2816
    },
    {
      "epoch": 0.5434027777777778,
      "grad_norm": 1.2131134271621704,
      "learning_rate": 0.0005,
      "loss": 2.509402275085449,
      "step": 2817
    },
    {
      "epoch": 0.5435956790123457,
      "grad_norm": 1.2543542385101318,
      "learning_rate": 0.0005,
      "loss": 2.7243874073028564,
      "step": 2818
    },
    {
      "epoch": 0.5437885802469136,
      "grad_norm": 0.9306955337524414,
      "learning_rate": 0.0005,
      "loss": 2.4484198093414307,
      "step": 2819
    },
    {
      "epoch": 0.5439814814814815,
      "grad_norm": 1.00368070602417,
      "learning_rate": 0.0005,
      "loss": 2.800773859024048,
      "step": 2820
    },
    {
      "epoch": 0.5441743827160493,
      "grad_norm": 1.0955363512039185,
      "learning_rate": 0.0005,
      "loss": 2.89418625831604,
      "step": 2821
    },
    {
      "epoch": 0.5443672839506173,
      "grad_norm": 1.3214597702026367,
      "learning_rate": 0.0005,
      "loss": 2.498934268951416,
      "step": 2822
    },
    {
      "epoch": 0.5445601851851852,
      "grad_norm": 1.4082237482070923,
      "learning_rate": 0.0005,
      "loss": 2.622223377227783,
      "step": 2823
    },
    {
      "epoch": 0.5447530864197531,
      "grad_norm": 1.012445330619812,
      "learning_rate": 0.0005,
      "loss": 2.836543560028076,
      "step": 2824
    },
    {
      "epoch": 0.544945987654321,
      "grad_norm": 0.9933595657348633,
      "learning_rate": 0.0005,
      "loss": 2.716006278991699,
      "step": 2825
    },
    {
      "epoch": 0.5451388888888888,
      "grad_norm": 1.5100666284561157,
      "learning_rate": 0.0005,
      "loss": 2.626401424407959,
      "step": 2826
    },
    {
      "epoch": 0.5453317901234568,
      "grad_norm": 1.2796828746795654,
      "learning_rate": 0.0005,
      "loss": 2.3687686920166016,
      "step": 2827
    },
    {
      "epoch": 0.5455246913580247,
      "grad_norm": 1.1431783437728882,
      "learning_rate": 0.0005,
      "loss": 2.5376906394958496,
      "step": 2828
    },
    {
      "epoch": 0.5457175925925926,
      "grad_norm": 1.1273807287216187,
      "learning_rate": 0.0005,
      "loss": 2.566798686981201,
      "step": 2829
    },
    {
      "epoch": 0.5459104938271605,
      "grad_norm": 1.133021354675293,
      "learning_rate": 0.0005,
      "loss": 2.704270362854004,
      "step": 2830
    },
    {
      "epoch": 0.5461033950617284,
      "grad_norm": 0.906532883644104,
      "learning_rate": 0.0005,
      "loss": 2.7640390396118164,
      "step": 2831
    },
    {
      "epoch": 0.5462962962962963,
      "grad_norm": 1.2022407054901123,
      "learning_rate": 0.0005,
      "loss": 2.60975980758667,
      "step": 2832
    },
    {
      "epoch": 0.5464891975308642,
      "grad_norm": 1.0276328325271606,
      "learning_rate": 0.0005,
      "loss": 2.5511093139648438,
      "step": 2833
    },
    {
      "epoch": 0.5466820987654321,
      "grad_norm": 0.9298112392425537,
      "learning_rate": 0.0005,
      "loss": 2.470701217651367,
      "step": 2834
    },
    {
      "epoch": 0.546875,
      "grad_norm": 1.1382502317428589,
      "learning_rate": 0.0005,
      "loss": 2.895285129547119,
      "step": 2835
    },
    {
      "epoch": 0.5470679012345679,
      "grad_norm": 0.8822936415672302,
      "learning_rate": 0.0005,
      "loss": 2.4557077884674072,
      "step": 2836
    },
    {
      "epoch": 0.5472608024691358,
      "grad_norm": 0.8067688941955566,
      "learning_rate": 0.0005,
      "loss": 2.7874739170074463,
      "step": 2837
    },
    {
      "epoch": 0.5474537037037037,
      "grad_norm": 1.066569447517395,
      "learning_rate": 0.0005,
      "loss": 2.4537835121154785,
      "step": 2838
    },
    {
      "epoch": 0.5476466049382716,
      "grad_norm": 0.8556633591651917,
      "learning_rate": 0.0005,
      "loss": 2.5630197525024414,
      "step": 2839
    },
    {
      "epoch": 0.5478395061728395,
      "grad_norm": 0.8526285886764526,
      "learning_rate": 0.0005,
      "loss": 2.316236734390259,
      "step": 2840
    },
    {
      "epoch": 0.5480324074074074,
      "grad_norm": 1.0869247913360596,
      "learning_rate": 0.0005,
      "loss": 2.608694314956665,
      "step": 2841
    },
    {
      "epoch": 0.5482253086419753,
      "grad_norm": 1.1112754344940186,
      "learning_rate": 0.0005,
      "loss": 2.4195213317871094,
      "step": 2842
    },
    {
      "epoch": 0.5484182098765432,
      "grad_norm": 1.0990372896194458,
      "learning_rate": 0.0005,
      "loss": 2.8570549488067627,
      "step": 2843
    },
    {
      "epoch": 0.5486111111111112,
      "grad_norm": 0.9560998678207397,
      "learning_rate": 0.0005,
      "loss": 2.612617254257202,
      "step": 2844
    },
    {
      "epoch": 0.548804012345679,
      "grad_norm": 1.2567368745803833,
      "learning_rate": 0.0005,
      "loss": 2.378286361694336,
      "step": 2845
    },
    {
      "epoch": 0.5489969135802469,
      "grad_norm": 0.8238373398780823,
      "learning_rate": 0.0005,
      "loss": 2.5007948875427246,
      "step": 2846
    },
    {
      "epoch": 0.5491898148148148,
      "grad_norm": 1.0808225870132446,
      "learning_rate": 0.0005,
      "loss": 2.687169313430786,
      "step": 2847
    },
    {
      "epoch": 0.5493827160493827,
      "grad_norm": 0.9260960817337036,
      "learning_rate": 0.0005,
      "loss": 2.6267056465148926,
      "step": 2848
    },
    {
      "epoch": 0.5495756172839507,
      "grad_norm": 0.9405842423439026,
      "learning_rate": 0.0005,
      "loss": 2.663820266723633,
      "step": 2849
    },
    {
      "epoch": 0.5497685185185185,
      "grad_norm": 0.9136174917221069,
      "learning_rate": 0.0005,
      "loss": 2.5622758865356445,
      "step": 2850
    },
    {
      "epoch": 0.5499614197530864,
      "grad_norm": 1.3617236614227295,
      "learning_rate": 0.0005,
      "loss": 2.5037126541137695,
      "step": 2851
    },
    {
      "epoch": 0.5501543209876543,
      "grad_norm": 1.2449063062667847,
      "learning_rate": 0.0005,
      "loss": 2.630495548248291,
      "step": 2852
    },
    {
      "epoch": 0.5503472222222222,
      "grad_norm": 0.9265967011451721,
      "learning_rate": 0.0005,
      "loss": 2.4939937591552734,
      "step": 2853
    },
    {
      "epoch": 0.5505401234567902,
      "grad_norm": 1.1495764255523682,
      "learning_rate": 0.0005,
      "loss": 2.4387431144714355,
      "step": 2854
    },
    {
      "epoch": 0.550733024691358,
      "grad_norm": 1.0106759071350098,
      "learning_rate": 0.0005,
      "loss": 2.4804608821868896,
      "step": 2855
    },
    {
      "epoch": 0.5509259259259259,
      "grad_norm": 0.9403990507125854,
      "learning_rate": 0.0005,
      "loss": 2.679072618484497,
      "step": 2856
    },
    {
      "epoch": 0.5511188271604939,
      "grad_norm": 0.8927901387214661,
      "learning_rate": 0.0005,
      "loss": 2.4056143760681152,
      "step": 2857
    },
    {
      "epoch": 0.5513117283950617,
      "grad_norm": 0.8683069944381714,
      "learning_rate": 0.0005,
      "loss": 2.4418883323669434,
      "step": 2858
    },
    {
      "epoch": 0.5515046296296297,
      "grad_norm": 1.0923259258270264,
      "learning_rate": 0.0005,
      "loss": 2.3893728256225586,
      "step": 2859
    },
    {
      "epoch": 0.5516975308641975,
      "grad_norm": 1.0619606971740723,
      "learning_rate": 0.0005,
      "loss": 2.6043601036071777,
      "step": 2860
    },
    {
      "epoch": 0.5518904320987654,
      "grad_norm": 0.832895040512085,
      "learning_rate": 0.0005,
      "loss": 2.564361810684204,
      "step": 2861
    },
    {
      "epoch": 0.5520833333333334,
      "grad_norm": 0.8672146201133728,
      "learning_rate": 0.0005,
      "loss": 2.885544776916504,
      "step": 2862
    },
    {
      "epoch": 0.5522762345679012,
      "grad_norm": 1.1204534769058228,
      "learning_rate": 0.0005,
      "loss": 2.7729387283325195,
      "step": 2863
    },
    {
      "epoch": 0.5524691358024691,
      "grad_norm": 0.8555780053138733,
      "learning_rate": 0.0005,
      "loss": 2.631282329559326,
      "step": 2864
    },
    {
      "epoch": 0.5526620370370371,
      "grad_norm": 1.2970892190933228,
      "learning_rate": 0.0005,
      "loss": 3.0788540840148926,
      "step": 2865
    },
    {
      "epoch": 0.5528549382716049,
      "grad_norm": 0.9976160526275635,
      "learning_rate": 0.0005,
      "loss": 2.6340880393981934,
      "step": 2866
    },
    {
      "epoch": 0.5530478395061729,
      "grad_norm": 0.8353120684623718,
      "learning_rate": 0.0005,
      "loss": 2.722015619277954,
      "step": 2867
    },
    {
      "epoch": 0.5532407407407407,
      "grad_norm": 0.9335446953773499,
      "learning_rate": 0.0005,
      "loss": 2.3711838722229004,
      "step": 2868
    },
    {
      "epoch": 0.5534336419753086,
      "grad_norm": 0.9167242646217346,
      "learning_rate": 0.0005,
      "loss": 2.3373732566833496,
      "step": 2869
    },
    {
      "epoch": 0.5536265432098766,
      "grad_norm": 0.9949311017990112,
      "learning_rate": 0.0005,
      "loss": 2.834261894226074,
      "step": 2870
    },
    {
      "epoch": 0.5538194444444444,
      "grad_norm": 0.9098691940307617,
      "learning_rate": 0.0005,
      "loss": 2.660040855407715,
      "step": 2871
    },
    {
      "epoch": 0.5540123456790124,
      "grad_norm": 0.9597660303115845,
      "learning_rate": 0.0005,
      "loss": 2.4801106452941895,
      "step": 2872
    },
    {
      "epoch": 0.5542052469135802,
      "grad_norm": 1.0309522151947021,
      "learning_rate": 0.0005,
      "loss": 2.5287303924560547,
      "step": 2873
    },
    {
      "epoch": 0.5543981481481481,
      "grad_norm": 0.9671857953071594,
      "learning_rate": 0.0005,
      "loss": 2.646385669708252,
      "step": 2874
    },
    {
      "epoch": 0.5545910493827161,
      "grad_norm": 0.9499311447143555,
      "learning_rate": 0.0005,
      "loss": 2.5846714973449707,
      "step": 2875
    },
    {
      "epoch": 0.5547839506172839,
      "grad_norm": 0.9226192831993103,
      "learning_rate": 0.0005,
      "loss": 2.6412410736083984,
      "step": 2876
    },
    {
      "epoch": 0.5549768518518519,
      "grad_norm": 1.13937246799469,
      "learning_rate": 0.0005,
      "loss": 2.897493839263916,
      "step": 2877
    },
    {
      "epoch": 0.5551697530864198,
      "grad_norm": 1.007897973060608,
      "learning_rate": 0.0005,
      "loss": 2.471489429473877,
      "step": 2878
    },
    {
      "epoch": 0.5553626543209876,
      "grad_norm": 0.9293507933616638,
      "learning_rate": 0.0005,
      "loss": 2.5389482975006104,
      "step": 2879
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.8675388693809509,
      "learning_rate": 0.0005,
      "loss": 2.750911235809326,
      "step": 2880
    },
    {
      "epoch": 0.5557484567901234,
      "grad_norm": 2.1210386753082275,
      "learning_rate": 0.0005,
      "loss": 2.360957622528076,
      "step": 2881
    },
    {
      "epoch": 0.5559413580246914,
      "grad_norm": 0.8657639026641846,
      "learning_rate": 0.0005,
      "loss": 2.5607314109802246,
      "step": 2882
    },
    {
      "epoch": 0.5561342592592593,
      "grad_norm": 1.0770702362060547,
      "learning_rate": 0.0005,
      "loss": 2.7555062770843506,
      "step": 2883
    },
    {
      "epoch": 0.5563271604938271,
      "grad_norm": 0.9553042054176331,
      "learning_rate": 0.0005,
      "loss": 2.7735719680786133,
      "step": 2884
    },
    {
      "epoch": 0.5565200617283951,
      "grad_norm": 0.8102526068687439,
      "learning_rate": 0.0005,
      "loss": 2.514249563217163,
      "step": 2885
    },
    {
      "epoch": 0.5567129629629629,
      "grad_norm": 1.059950590133667,
      "learning_rate": 0.0005,
      "loss": 2.7863645553588867,
      "step": 2886
    },
    {
      "epoch": 0.5569058641975309,
      "grad_norm": 1.1241967678070068,
      "learning_rate": 0.0005,
      "loss": 2.4788551330566406,
      "step": 2887
    },
    {
      "epoch": 0.5570987654320988,
      "grad_norm": 0.9135527610778809,
      "learning_rate": 0.0005,
      "loss": 2.400501251220703,
      "step": 2888
    },
    {
      "epoch": 0.5572916666666666,
      "grad_norm": 0.8829161524772644,
      "learning_rate": 0.0005,
      "loss": 2.4245567321777344,
      "step": 2889
    },
    {
      "epoch": 0.5574845679012346,
      "grad_norm": 0.8390306234359741,
      "learning_rate": 0.0005,
      "loss": 2.6382458209991455,
      "step": 2890
    },
    {
      "epoch": 0.5576774691358025,
      "grad_norm": 1.3592503070831299,
      "learning_rate": 0.0005,
      "loss": 2.3429784774780273,
      "step": 2891
    },
    {
      "epoch": 0.5578703703703703,
      "grad_norm": 1.237815022468567,
      "learning_rate": 0.0005,
      "loss": 2.7474863529205322,
      "step": 2892
    },
    {
      "epoch": 0.5580632716049383,
      "grad_norm": 1.4043186902999878,
      "learning_rate": 0.0005,
      "loss": 2.5936508178710938,
      "step": 2893
    },
    {
      "epoch": 0.5582561728395061,
      "grad_norm": 0.8964990973472595,
      "learning_rate": 0.0005,
      "loss": 2.307504177093506,
      "step": 2894
    },
    {
      "epoch": 0.5584490740740741,
      "grad_norm": 0.8965327143669128,
      "learning_rate": 0.0005,
      "loss": 2.504253625869751,
      "step": 2895
    },
    {
      "epoch": 0.558641975308642,
      "grad_norm": 0.87591952085495,
      "learning_rate": 0.0005,
      "loss": 2.5575661659240723,
      "step": 2896
    },
    {
      "epoch": 0.5588348765432098,
      "grad_norm": 1.3190513849258423,
      "learning_rate": 0.0005,
      "loss": 2.6252646446228027,
      "step": 2897
    },
    {
      "epoch": 0.5590277777777778,
      "grad_norm": 0.9654766321182251,
      "learning_rate": 0.0005,
      "loss": 2.879244089126587,
      "step": 2898
    },
    {
      "epoch": 0.5592206790123457,
      "grad_norm": 1.072641134262085,
      "learning_rate": 0.0005,
      "loss": 2.570040702819824,
      "step": 2899
    },
    {
      "epoch": 0.5594135802469136,
      "grad_norm": 0.9536446928977966,
      "learning_rate": 0.0005,
      "loss": 2.416785955429077,
      "step": 2900
    },
    {
      "epoch": 0.5596064814814815,
      "grad_norm": 0.8866228461265564,
      "learning_rate": 0.0005,
      "loss": 2.569780111312866,
      "step": 2901
    },
    {
      "epoch": 0.5597993827160493,
      "grad_norm": 3.93562388420105,
      "learning_rate": 0.0005,
      "loss": 2.6298534870147705,
      "step": 2902
    },
    {
      "epoch": 0.5599922839506173,
      "grad_norm": 0.9169299006462097,
      "learning_rate": 0.0005,
      "loss": 2.4759864807128906,
      "step": 2903
    },
    {
      "epoch": 0.5601851851851852,
      "grad_norm": 1.0565873384475708,
      "learning_rate": 0.0005,
      "loss": 2.7607994079589844,
      "step": 2904
    },
    {
      "epoch": 0.5603780864197531,
      "grad_norm": 1.1743874549865723,
      "learning_rate": 0.0005,
      "loss": 2.7193779945373535,
      "step": 2905
    },
    {
      "epoch": 0.560570987654321,
      "grad_norm": 1.0485690832138062,
      "learning_rate": 0.0005,
      "loss": 2.645961284637451,
      "step": 2906
    },
    {
      "epoch": 0.5607638888888888,
      "grad_norm": 1.038757085800171,
      "learning_rate": 0.0005,
      "loss": 2.4445042610168457,
      "step": 2907
    },
    {
      "epoch": 0.5609567901234568,
      "grad_norm": 0.890781819820404,
      "learning_rate": 0.0005,
      "loss": 2.7621402740478516,
      "step": 2908
    },
    {
      "epoch": 0.5611496913580247,
      "grad_norm": 1.1901532411575317,
      "learning_rate": 0.0005,
      "loss": 2.610093116760254,
      "step": 2909
    },
    {
      "epoch": 0.5613425925925926,
      "grad_norm": 1.0317808389663696,
      "learning_rate": 0.0005,
      "loss": 2.099484920501709,
      "step": 2910
    },
    {
      "epoch": 0.5615354938271605,
      "grad_norm": 1.4942501783370972,
      "learning_rate": 0.0005,
      "loss": 2.7886621952056885,
      "step": 2911
    },
    {
      "epoch": 0.5617283950617284,
      "grad_norm": 0.8812448382377625,
      "learning_rate": 0.0005,
      "loss": 2.659228801727295,
      "step": 2912
    },
    {
      "epoch": 0.5619212962962963,
      "grad_norm": 0.8126965165138245,
      "learning_rate": 0.0005,
      "loss": 2.73531436920166,
      "step": 2913
    },
    {
      "epoch": 0.5621141975308642,
      "grad_norm": 0.9162732362747192,
      "learning_rate": 0.0005,
      "loss": 2.47601318359375,
      "step": 2914
    },
    {
      "epoch": 0.5623070987654321,
      "grad_norm": 0.9610905051231384,
      "learning_rate": 0.0005,
      "loss": 2.5258588790893555,
      "step": 2915
    },
    {
      "epoch": 0.5625,
      "grad_norm": 1.2958767414093018,
      "learning_rate": 0.0005,
      "loss": 2.8946189880371094,
      "step": 2916
    },
    {
      "epoch": 0.5626929012345679,
      "grad_norm": 1.01334810256958,
      "learning_rate": 0.0005,
      "loss": 2.7622146606445312,
      "step": 2917
    },
    {
      "epoch": 0.5628858024691358,
      "grad_norm": 1.0222854614257812,
      "learning_rate": 0.0005,
      "loss": 2.649158000946045,
      "step": 2918
    },
    {
      "epoch": 0.5630787037037037,
      "grad_norm": 1.2036411762237549,
      "learning_rate": 0.0005,
      "loss": 2.419396162033081,
      "step": 2919
    },
    {
      "epoch": 0.5632716049382716,
      "grad_norm": 1.1264854669570923,
      "learning_rate": 0.0005,
      "loss": 2.751479148864746,
      "step": 2920
    },
    {
      "epoch": 0.5634645061728395,
      "grad_norm": 0.8845984935760498,
      "learning_rate": 0.0005,
      "loss": 2.4928717613220215,
      "step": 2921
    },
    {
      "epoch": 0.5636574074074074,
      "grad_norm": 1.244949460029602,
      "learning_rate": 0.0005,
      "loss": 2.752347707748413,
      "step": 2922
    },
    {
      "epoch": 0.5638503086419753,
      "grad_norm": 0.9254374504089355,
      "learning_rate": 0.0005,
      "loss": 2.6891672611236572,
      "step": 2923
    },
    {
      "epoch": 0.5640432098765432,
      "grad_norm": 1.006462574005127,
      "learning_rate": 0.0005,
      "loss": 2.533574104309082,
      "step": 2924
    },
    {
      "epoch": 0.5642361111111112,
      "grad_norm": 1.2483757734298706,
      "learning_rate": 0.0005,
      "loss": 2.657712697982788,
      "step": 2925
    },
    {
      "epoch": 0.564429012345679,
      "grad_norm": 1.0010035037994385,
      "learning_rate": 0.0005,
      "loss": 2.3576889038085938,
      "step": 2926
    },
    {
      "epoch": 0.5646219135802469,
      "grad_norm": 0.9618512392044067,
      "learning_rate": 0.0005,
      "loss": 2.6759538650512695,
      "step": 2927
    },
    {
      "epoch": 0.5648148148148148,
      "grad_norm": 0.7596309781074524,
      "learning_rate": 0.0005,
      "loss": 2.42637300491333,
      "step": 2928
    },
    {
      "epoch": 0.5650077160493827,
      "grad_norm": 1.6655088663101196,
      "learning_rate": 0.0005,
      "loss": 2.8589015007019043,
      "step": 2929
    },
    {
      "epoch": 0.5652006172839507,
      "grad_norm": 0.9660887718200684,
      "learning_rate": 0.0005,
      "loss": 2.7586419582366943,
      "step": 2930
    },
    {
      "epoch": 0.5653935185185185,
      "grad_norm": 1.0847678184509277,
      "learning_rate": 0.0005,
      "loss": 2.5346498489379883,
      "step": 2931
    },
    {
      "epoch": 0.5655864197530864,
      "grad_norm": 0.9012241959571838,
      "learning_rate": 0.0005,
      "loss": 2.404254913330078,
      "step": 2932
    },
    {
      "epoch": 0.5657793209876543,
      "grad_norm": 0.9441726207733154,
      "learning_rate": 0.0005,
      "loss": 2.7644662857055664,
      "step": 2933
    },
    {
      "epoch": 0.5659722222222222,
      "grad_norm": 0.9991815686225891,
      "learning_rate": 0.0005,
      "loss": 2.8630666732788086,
      "step": 2934
    },
    {
      "epoch": 0.5661651234567902,
      "grad_norm": 0.9847079515457153,
      "learning_rate": 0.0005,
      "loss": 2.478661060333252,
      "step": 2935
    },
    {
      "epoch": 0.566358024691358,
      "grad_norm": 1.0399528741836548,
      "learning_rate": 0.0005,
      "loss": 2.55792236328125,
      "step": 2936
    },
    {
      "epoch": 0.5665509259259259,
      "grad_norm": 0.849296510219574,
      "learning_rate": 0.0005,
      "loss": 2.7369039058685303,
      "step": 2937
    },
    {
      "epoch": 0.5667438271604939,
      "grad_norm": 1.6402111053466797,
      "learning_rate": 0.0005,
      "loss": 2.7120094299316406,
      "step": 2938
    },
    {
      "epoch": 0.5669367283950617,
      "grad_norm": 0.8918448090553284,
      "learning_rate": 0.0005,
      "loss": 2.3148374557495117,
      "step": 2939
    },
    {
      "epoch": 0.5671296296296297,
      "grad_norm": 1.199957251548767,
      "learning_rate": 0.0005,
      "loss": 2.4543166160583496,
      "step": 2940
    },
    {
      "epoch": 0.5673225308641975,
      "grad_norm": 0.8921889662742615,
      "learning_rate": 0.0005,
      "loss": 2.5088863372802734,
      "step": 2941
    },
    {
      "epoch": 0.5675154320987654,
      "grad_norm": 0.8446884751319885,
      "learning_rate": 0.0005,
      "loss": 2.4555294513702393,
      "step": 2942
    },
    {
      "epoch": 0.5677083333333334,
      "grad_norm": 1.263710856437683,
      "learning_rate": 0.0005,
      "loss": 2.7989885807037354,
      "step": 2943
    },
    {
      "epoch": 0.5679012345679012,
      "grad_norm": 0.9850060939788818,
      "learning_rate": 0.0005,
      "loss": 2.733297824859619,
      "step": 2944
    },
    {
      "epoch": 0.5680941358024691,
      "grad_norm": 1.4712618589401245,
      "learning_rate": 0.0005,
      "loss": 2.5094735622406006,
      "step": 2945
    },
    {
      "epoch": 0.5682870370370371,
      "grad_norm": 0.8442370295524597,
      "learning_rate": 0.0005,
      "loss": 2.584019184112549,
      "step": 2946
    },
    {
      "epoch": 0.5684799382716049,
      "grad_norm": 0.9943140745162964,
      "learning_rate": 0.0005,
      "loss": 2.579361915588379,
      "step": 2947
    },
    {
      "epoch": 0.5686728395061729,
      "grad_norm": 0.8879385590553284,
      "learning_rate": 0.0005,
      "loss": 2.721210479736328,
      "step": 2948
    },
    {
      "epoch": 0.5688657407407407,
      "grad_norm": 1.1772359609603882,
      "learning_rate": 0.0005,
      "loss": 2.822017192840576,
      "step": 2949
    },
    {
      "epoch": 0.5690586419753086,
      "grad_norm": 0.9565602540969849,
      "learning_rate": 0.0005,
      "loss": 2.696796417236328,
      "step": 2950
    },
    {
      "epoch": 0.5692515432098766,
      "grad_norm": 1.2901747226715088,
      "learning_rate": 0.0005,
      "loss": 2.545868396759033,
      "step": 2951
    },
    {
      "epoch": 0.5694444444444444,
      "grad_norm": 0.9160227179527283,
      "learning_rate": 0.0005,
      "loss": 2.5447750091552734,
      "step": 2952
    },
    {
      "epoch": 0.5696373456790124,
      "grad_norm": 0.9313535690307617,
      "learning_rate": 0.0005,
      "loss": 2.8362607955932617,
      "step": 2953
    },
    {
      "epoch": 0.5698302469135802,
      "grad_norm": 0.9187290668487549,
      "learning_rate": 0.0005,
      "loss": 2.4344754219055176,
      "step": 2954
    },
    {
      "epoch": 0.5700231481481481,
      "grad_norm": 0.9274174571037292,
      "learning_rate": 0.0005,
      "loss": 2.7876319885253906,
      "step": 2955
    },
    {
      "epoch": 0.5702160493827161,
      "grad_norm": 0.8396411538124084,
      "learning_rate": 0.0005,
      "loss": 2.62454891204834,
      "step": 2956
    },
    {
      "epoch": 0.5704089506172839,
      "grad_norm": 0.8234838247299194,
      "learning_rate": 0.0005,
      "loss": 2.484341621398926,
      "step": 2957
    },
    {
      "epoch": 0.5706018518518519,
      "grad_norm": 0.853039562702179,
      "learning_rate": 0.0005,
      "loss": 2.567713737487793,
      "step": 2958
    },
    {
      "epoch": 0.5707947530864198,
      "grad_norm": 0.9921619892120361,
      "learning_rate": 0.0005,
      "loss": 2.7806341648101807,
      "step": 2959
    },
    {
      "epoch": 0.5709876543209876,
      "grad_norm": 0.8792877197265625,
      "learning_rate": 0.0005,
      "loss": 2.1867668628692627,
      "step": 2960
    },
    {
      "epoch": 0.5711805555555556,
      "grad_norm": 0.8452260494232178,
      "learning_rate": 0.0005,
      "loss": 2.620001792907715,
      "step": 2961
    },
    {
      "epoch": 0.5713734567901234,
      "grad_norm": 1.5842125415802002,
      "learning_rate": 0.0005,
      "loss": 2.23942494392395,
      "step": 2962
    },
    {
      "epoch": 0.5715663580246914,
      "grad_norm": 0.9609775543212891,
      "learning_rate": 0.0005,
      "loss": 2.2727456092834473,
      "step": 2963
    },
    {
      "epoch": 0.5717592592592593,
      "grad_norm": 1.3443411588668823,
      "learning_rate": 0.0005,
      "loss": 2.304532051086426,
      "step": 2964
    },
    {
      "epoch": 0.5719521604938271,
      "grad_norm": 1.80446457862854,
      "learning_rate": 0.0005,
      "loss": 2.8524680137634277,
      "step": 2965
    },
    {
      "epoch": 0.5721450617283951,
      "grad_norm": 0.9543615579605103,
      "learning_rate": 0.0005,
      "loss": 2.6784417629241943,
      "step": 2966
    },
    {
      "epoch": 0.5723379629629629,
      "grad_norm": 0.9983107447624207,
      "learning_rate": 0.0005,
      "loss": 2.589355707168579,
      "step": 2967
    },
    {
      "epoch": 0.5725308641975309,
      "grad_norm": 1.1836421489715576,
      "learning_rate": 0.0005,
      "loss": 2.7309181690216064,
      "step": 2968
    },
    {
      "epoch": 0.5727237654320988,
      "grad_norm": 0.9131689071655273,
      "learning_rate": 0.0005,
      "loss": 2.409663200378418,
      "step": 2969
    },
    {
      "epoch": 0.5729166666666666,
      "grad_norm": 0.9072160124778748,
      "learning_rate": 0.0005,
      "loss": 2.718496322631836,
      "step": 2970
    },
    {
      "epoch": 0.5731095679012346,
      "grad_norm": 0.8572176098823547,
      "learning_rate": 0.0005,
      "loss": 2.609269618988037,
      "step": 2971
    },
    {
      "epoch": 0.5733024691358025,
      "grad_norm": 0.8713198900222778,
      "learning_rate": 0.0005,
      "loss": 2.431887626647949,
      "step": 2972
    },
    {
      "epoch": 0.5734953703703703,
      "grad_norm": 1.039085030555725,
      "learning_rate": 0.0005,
      "loss": 2.3244850635528564,
      "step": 2973
    },
    {
      "epoch": 0.5736882716049383,
      "grad_norm": 0.8388222455978394,
      "learning_rate": 0.0005,
      "loss": 2.1778440475463867,
      "step": 2974
    },
    {
      "epoch": 0.5738811728395061,
      "grad_norm": 1.1323133707046509,
      "learning_rate": 0.0005,
      "loss": 2.5837836265563965,
      "step": 2975
    },
    {
      "epoch": 0.5740740740740741,
      "grad_norm": 1.7303638458251953,
      "learning_rate": 0.0005,
      "loss": 2.6860013008117676,
      "step": 2976
    },
    {
      "epoch": 0.574266975308642,
      "grad_norm": 0.786433756351471,
      "learning_rate": 0.0005,
      "loss": 2.496966600418091,
      "step": 2977
    },
    {
      "epoch": 0.5744598765432098,
      "grad_norm": 0.9910193681716919,
      "learning_rate": 0.0005,
      "loss": 2.840010643005371,
      "step": 2978
    },
    {
      "epoch": 0.5746527777777778,
      "grad_norm": 1.001837134361267,
      "learning_rate": 0.0005,
      "loss": 2.4967923164367676,
      "step": 2979
    },
    {
      "epoch": 0.5748456790123457,
      "grad_norm": 0.953339159488678,
      "learning_rate": 0.0005,
      "loss": 2.7109925746917725,
      "step": 2980
    },
    {
      "epoch": 0.5750385802469136,
      "grad_norm": 1.6569671630859375,
      "learning_rate": 0.0005,
      "loss": 2.703874111175537,
      "step": 2981
    },
    {
      "epoch": 0.5752314814814815,
      "grad_norm": 0.8305198550224304,
      "learning_rate": 0.0005,
      "loss": 2.341580867767334,
      "step": 2982
    },
    {
      "epoch": 0.5754243827160493,
      "grad_norm": 0.9368364810943604,
      "learning_rate": 0.0005,
      "loss": 2.729891777038574,
      "step": 2983
    },
    {
      "epoch": 0.5756172839506173,
      "grad_norm": 1.2827407121658325,
      "learning_rate": 0.0005,
      "loss": 2.798941135406494,
      "step": 2984
    },
    {
      "epoch": 0.5758101851851852,
      "grad_norm": 0.938123345375061,
      "learning_rate": 0.0005,
      "loss": 2.6416687965393066,
      "step": 2985
    },
    {
      "epoch": 0.5760030864197531,
      "grad_norm": 0.8700226545333862,
      "learning_rate": 0.0005,
      "loss": 2.5228028297424316,
      "step": 2986
    },
    {
      "epoch": 0.576195987654321,
      "grad_norm": 1.2801430225372314,
      "learning_rate": 0.0005,
      "loss": 2.4839956760406494,
      "step": 2987
    },
    {
      "epoch": 0.5763888888888888,
      "grad_norm": 1.2173281908035278,
      "learning_rate": 0.0005,
      "loss": 2.8743391036987305,
      "step": 2988
    },
    {
      "epoch": 0.5765817901234568,
      "grad_norm": 1.1320120096206665,
      "learning_rate": 0.0005,
      "loss": 2.322498321533203,
      "step": 2989
    },
    {
      "epoch": 0.5767746913580247,
      "grad_norm": 1.5253437757492065,
      "learning_rate": 0.0005,
      "loss": 2.4801478385925293,
      "step": 2990
    },
    {
      "epoch": 0.5769675925925926,
      "grad_norm": 0.8948571681976318,
      "learning_rate": 0.0005,
      "loss": 2.6747913360595703,
      "step": 2991
    },
    {
      "epoch": 0.5771604938271605,
      "grad_norm": 1.1314345598220825,
      "learning_rate": 0.0005,
      "loss": 2.7662720680236816,
      "step": 2992
    },
    {
      "epoch": 0.5773533950617284,
      "grad_norm": 1.1102763414382935,
      "learning_rate": 0.0005,
      "loss": 2.673924446105957,
      "step": 2993
    },
    {
      "epoch": 0.5775462962962963,
      "grad_norm": 0.9284820556640625,
      "learning_rate": 0.0005,
      "loss": 2.387240409851074,
      "step": 2994
    },
    {
      "epoch": 0.5777391975308642,
      "grad_norm": 1.0917644500732422,
      "learning_rate": 0.0005,
      "loss": 2.817803382873535,
      "step": 2995
    },
    {
      "epoch": 0.5779320987654321,
      "grad_norm": 1.0070728063583374,
      "learning_rate": 0.0005,
      "loss": 2.7599053382873535,
      "step": 2996
    },
    {
      "epoch": 0.578125,
      "grad_norm": 0.9613296389579773,
      "learning_rate": 0.0005,
      "loss": 2.725370407104492,
      "step": 2997
    },
    {
      "epoch": 0.5783179012345679,
      "grad_norm": 1.0473865270614624,
      "learning_rate": 0.0005,
      "loss": 2.3632097244262695,
      "step": 2998
    },
    {
      "epoch": 0.5785108024691358,
      "grad_norm": 1.1568130254745483,
      "learning_rate": 0.0005,
      "loss": 2.492837905883789,
      "step": 2999
    },
    {
      "epoch": 0.5787037037037037,
      "grad_norm": 0.8513147830963135,
      "learning_rate": 0.0005,
      "loss": 2.5588412284851074,
      "step": 3000
    },
    {
      "epoch": 0.5788966049382716,
      "grad_norm": 1.4105108976364136,
      "learning_rate": 0.0005,
      "loss": 2.7445731163024902,
      "step": 3001
    },
    {
      "epoch": 0.5790895061728395,
      "grad_norm": 0.8537500500679016,
      "learning_rate": 0.0005,
      "loss": 2.5274221897125244,
      "step": 3002
    },
    {
      "epoch": 0.5792824074074074,
      "grad_norm": 0.7541114687919617,
      "learning_rate": 0.0005,
      "loss": 2.629917621612549,
      "step": 3003
    },
    {
      "epoch": 0.5794753086419753,
      "grad_norm": 0.9565783739089966,
      "learning_rate": 0.0005,
      "loss": 2.7188940048217773,
      "step": 3004
    },
    {
      "epoch": 0.5796682098765432,
      "grad_norm": 0.9685673117637634,
      "learning_rate": 0.0005,
      "loss": 2.7297098636627197,
      "step": 3005
    },
    {
      "epoch": 0.5798611111111112,
      "grad_norm": 0.8519525527954102,
      "learning_rate": 0.0005,
      "loss": 2.530366897583008,
      "step": 3006
    },
    {
      "epoch": 0.580054012345679,
      "grad_norm": 1.0190047025680542,
      "learning_rate": 0.0005,
      "loss": 2.4207324981689453,
      "step": 3007
    },
    {
      "epoch": 0.5802469135802469,
      "grad_norm": 0.8921204209327698,
      "learning_rate": 0.0005,
      "loss": 2.454089403152466,
      "step": 3008
    },
    {
      "epoch": 0.5804398148148148,
      "grad_norm": 0.963467001914978,
      "learning_rate": 0.0005,
      "loss": 2.442899227142334,
      "step": 3009
    },
    {
      "epoch": 0.5806327160493827,
      "grad_norm": 0.9508268237113953,
      "learning_rate": 0.0005,
      "loss": 2.500619411468506,
      "step": 3010
    },
    {
      "epoch": 0.5808256172839507,
      "grad_norm": 1.3263740539550781,
      "learning_rate": 0.0005,
      "loss": 2.8401436805725098,
      "step": 3011
    },
    {
      "epoch": 0.5810185185185185,
      "grad_norm": 0.9975385665893555,
      "learning_rate": 0.0005,
      "loss": 2.241323232650757,
      "step": 3012
    },
    {
      "epoch": 0.5812114197530864,
      "grad_norm": 1.1069202423095703,
      "learning_rate": 0.0005,
      "loss": 2.246006965637207,
      "step": 3013
    },
    {
      "epoch": 0.5814043209876543,
      "grad_norm": 0.8420957922935486,
      "learning_rate": 0.0005,
      "loss": 2.5091209411621094,
      "step": 3014
    },
    {
      "epoch": 0.5815972222222222,
      "grad_norm": 0.7883920073509216,
      "learning_rate": 0.0005,
      "loss": 2.5220885276794434,
      "step": 3015
    },
    {
      "epoch": 0.5817901234567902,
      "grad_norm": 0.8776164650917053,
      "learning_rate": 0.0005,
      "loss": 2.6734907627105713,
      "step": 3016
    },
    {
      "epoch": 0.581983024691358,
      "grad_norm": 1.1201817989349365,
      "learning_rate": 0.0005,
      "loss": 2.246371269226074,
      "step": 3017
    },
    {
      "epoch": 0.5821759259259259,
      "grad_norm": 0.861025333404541,
      "learning_rate": 0.0005,
      "loss": 2.5087571144104004,
      "step": 3018
    },
    {
      "epoch": 0.5823688271604939,
      "grad_norm": 1.0142498016357422,
      "learning_rate": 0.0005,
      "loss": 2.7769126892089844,
      "step": 3019
    },
    {
      "epoch": 0.5825617283950617,
      "grad_norm": 1.0209957361221313,
      "learning_rate": 0.0005,
      "loss": 2.3993637561798096,
      "step": 3020
    },
    {
      "epoch": 0.5827546296296297,
      "grad_norm": 0.8673728704452515,
      "learning_rate": 0.0005,
      "loss": 2.7488596439361572,
      "step": 3021
    },
    {
      "epoch": 0.5829475308641975,
      "grad_norm": 0.8916833400726318,
      "learning_rate": 0.0005,
      "loss": 2.3824310302734375,
      "step": 3022
    },
    {
      "epoch": 0.5831404320987654,
      "grad_norm": 1.6118907928466797,
      "learning_rate": 0.0005,
      "loss": 2.591332197189331,
      "step": 3023
    },
    {
      "epoch": 0.5833333333333334,
      "grad_norm": 1.019910216331482,
      "learning_rate": 0.0005,
      "loss": 2.366837501525879,
      "step": 3024
    },
    {
      "epoch": 0.5835262345679012,
      "grad_norm": 0.9533225893974304,
      "learning_rate": 0.0005,
      "loss": 2.577104330062866,
      "step": 3025
    },
    {
      "epoch": 0.5837191358024691,
      "grad_norm": 0.9614185690879822,
      "learning_rate": 0.0005,
      "loss": 2.8320446014404297,
      "step": 3026
    },
    {
      "epoch": 0.5839120370370371,
      "grad_norm": 0.9641319513320923,
      "learning_rate": 0.0005,
      "loss": 2.4657890796661377,
      "step": 3027
    },
    {
      "epoch": 0.5841049382716049,
      "grad_norm": 0.9819316864013672,
      "learning_rate": 0.0005,
      "loss": 2.6850697994232178,
      "step": 3028
    },
    {
      "epoch": 0.5842978395061729,
      "grad_norm": 1.0167906284332275,
      "learning_rate": 0.0005,
      "loss": 3.075186252593994,
      "step": 3029
    },
    {
      "epoch": 0.5844907407407407,
      "grad_norm": 1.6966263055801392,
      "learning_rate": 0.0005,
      "loss": 2.4095497131347656,
      "step": 3030
    },
    {
      "epoch": 0.5846836419753086,
      "grad_norm": 1.0463262796401978,
      "learning_rate": 0.0005,
      "loss": 2.6870977878570557,
      "step": 3031
    },
    {
      "epoch": 0.5848765432098766,
      "grad_norm": 0.9383959770202637,
      "learning_rate": 0.0005,
      "loss": 2.7647666931152344,
      "step": 3032
    },
    {
      "epoch": 0.5850694444444444,
      "grad_norm": 0.7958080768585205,
      "learning_rate": 0.0005,
      "loss": 2.736135482788086,
      "step": 3033
    },
    {
      "epoch": 0.5852623456790124,
      "grad_norm": 0.8643621802330017,
      "learning_rate": 0.0005,
      "loss": 2.6649413108825684,
      "step": 3034
    },
    {
      "epoch": 0.5854552469135802,
      "grad_norm": 1.0870658159255981,
      "learning_rate": 0.0005,
      "loss": 2.378140926361084,
      "step": 3035
    },
    {
      "epoch": 0.5856481481481481,
      "grad_norm": 0.8258690237998962,
      "learning_rate": 0.0005,
      "loss": 2.7057132720947266,
      "step": 3036
    },
    {
      "epoch": 0.5858410493827161,
      "grad_norm": 0.8114559650421143,
      "learning_rate": 0.0005,
      "loss": 2.2579307556152344,
      "step": 3037
    },
    {
      "epoch": 0.5860339506172839,
      "grad_norm": 0.9488770365715027,
      "learning_rate": 0.0005,
      "loss": 2.9725382328033447,
      "step": 3038
    },
    {
      "epoch": 0.5862268518518519,
      "grad_norm": 1.0834629535675049,
      "learning_rate": 0.0005,
      "loss": 2.6250085830688477,
      "step": 3039
    },
    {
      "epoch": 0.5864197530864198,
      "grad_norm": 0.8976984620094299,
      "learning_rate": 0.0005,
      "loss": 2.6041111946105957,
      "step": 3040
    },
    {
      "epoch": 0.5866126543209876,
      "grad_norm": 1.084143877029419,
      "learning_rate": 0.0005,
      "loss": 2.5650219917297363,
      "step": 3041
    },
    {
      "epoch": 0.5868055555555556,
      "grad_norm": 0.9996917843818665,
      "learning_rate": 0.0005,
      "loss": 2.6025233268737793,
      "step": 3042
    },
    {
      "epoch": 0.5869984567901234,
      "grad_norm": 0.865143895149231,
      "learning_rate": 0.0005,
      "loss": 2.354124069213867,
      "step": 3043
    },
    {
      "epoch": 0.5871913580246914,
      "grad_norm": 1.2413913011550903,
      "learning_rate": 0.0005,
      "loss": 2.802997589111328,
      "step": 3044
    },
    {
      "epoch": 0.5873842592592593,
      "grad_norm": 0.9233061075210571,
      "learning_rate": 0.0005,
      "loss": 2.9236207008361816,
      "step": 3045
    },
    {
      "epoch": 0.5875771604938271,
      "grad_norm": 0.8809791803359985,
      "learning_rate": 0.0005,
      "loss": 2.338379383087158,
      "step": 3046
    },
    {
      "epoch": 0.5877700617283951,
      "grad_norm": 0.8451929688453674,
      "learning_rate": 0.0005,
      "loss": 2.7480154037475586,
      "step": 3047
    },
    {
      "epoch": 0.5879629629629629,
      "grad_norm": 0.9945508241653442,
      "learning_rate": 0.0005,
      "loss": 2.4802069664001465,
      "step": 3048
    },
    {
      "epoch": 0.5881558641975309,
      "grad_norm": 0.9292887449264526,
      "learning_rate": 0.0005,
      "loss": 2.7146377563476562,
      "step": 3049
    },
    {
      "epoch": 0.5883487654320988,
      "grad_norm": 0.7008775472640991,
      "learning_rate": 0.0005,
      "loss": 2.6046905517578125,
      "step": 3050
    },
    {
      "epoch": 0.5885416666666666,
      "grad_norm": 3.832425117492676,
      "learning_rate": 0.0005,
      "loss": 2.5534558296203613,
      "step": 3051
    },
    {
      "epoch": 0.5887345679012346,
      "grad_norm": 0.8299717903137207,
      "learning_rate": 0.0005,
      "loss": 2.4916534423828125,
      "step": 3052
    },
    {
      "epoch": 0.5889274691358025,
      "grad_norm": 1.1096410751342773,
      "learning_rate": 0.0005,
      "loss": 2.406484603881836,
      "step": 3053
    },
    {
      "epoch": 0.5891203703703703,
      "grad_norm": 0.9079247117042542,
      "learning_rate": 0.0005,
      "loss": 2.5653281211853027,
      "step": 3054
    },
    {
      "epoch": 0.5893132716049383,
      "grad_norm": 0.8676683902740479,
      "learning_rate": 0.0005,
      "loss": 2.134739398956299,
      "step": 3055
    },
    {
      "epoch": 0.5895061728395061,
      "grad_norm": 0.880120038986206,
      "learning_rate": 0.0005,
      "loss": 2.3980302810668945,
      "step": 3056
    },
    {
      "epoch": 0.5896990740740741,
      "grad_norm": 0.9965208768844604,
      "learning_rate": 0.0005,
      "loss": 2.6090073585510254,
      "step": 3057
    },
    {
      "epoch": 0.589891975308642,
      "grad_norm": 1.007409930229187,
      "learning_rate": 0.0005,
      "loss": 2.566312313079834,
      "step": 3058
    },
    {
      "epoch": 0.5900848765432098,
      "grad_norm": 0.9824986457824707,
      "learning_rate": 0.0005,
      "loss": 2.6280720233917236,
      "step": 3059
    },
    {
      "epoch": 0.5902777777777778,
      "grad_norm": 0.9026748538017273,
      "learning_rate": 0.0005,
      "loss": 2.5260605812072754,
      "step": 3060
    },
    {
      "epoch": 0.5904706790123457,
      "grad_norm": 0.8933073878288269,
      "learning_rate": 0.0005,
      "loss": 2.483463764190674,
      "step": 3061
    },
    {
      "epoch": 0.5906635802469136,
      "grad_norm": 0.9718018174171448,
      "learning_rate": 0.0005,
      "loss": 2.6050286293029785,
      "step": 3062
    },
    {
      "epoch": 0.5908564814814815,
      "grad_norm": 1.0267605781555176,
      "learning_rate": 0.0005,
      "loss": 2.31427001953125,
      "step": 3063
    },
    {
      "epoch": 0.5910493827160493,
      "grad_norm": 1.1448750495910645,
      "learning_rate": 0.0005,
      "loss": 2.323002338409424,
      "step": 3064
    },
    {
      "epoch": 0.5912422839506173,
      "grad_norm": 1.0374051332473755,
      "learning_rate": 0.0005,
      "loss": 2.649801731109619,
      "step": 3065
    },
    {
      "epoch": 0.5914351851851852,
      "grad_norm": 1.2679263353347778,
      "learning_rate": 0.0005,
      "loss": 2.6338610649108887,
      "step": 3066
    },
    {
      "epoch": 0.5916280864197531,
      "grad_norm": 0.8102988004684448,
      "learning_rate": 0.0005,
      "loss": 2.6731653213500977,
      "step": 3067
    },
    {
      "epoch": 0.591820987654321,
      "grad_norm": 1.0765634775161743,
      "learning_rate": 0.0005,
      "loss": 2.9397811889648438,
      "step": 3068
    },
    {
      "epoch": 0.5920138888888888,
      "grad_norm": 1.1004019975662231,
      "learning_rate": 0.0005,
      "loss": 2.6499152183532715,
      "step": 3069
    },
    {
      "epoch": 0.5922067901234568,
      "grad_norm": 0.9479472041130066,
      "learning_rate": 0.0005,
      "loss": 2.429922342300415,
      "step": 3070
    },
    {
      "epoch": 0.5923996913580247,
      "grad_norm": 1.2016078233718872,
      "learning_rate": 0.0005,
      "loss": 2.60282564163208,
      "step": 3071
    },
    {
      "epoch": 0.5925925925925926,
      "grad_norm": 0.862076461315155,
      "learning_rate": 0.0005,
      "loss": 2.3598825931549072,
      "step": 3072
    },
    {
      "epoch": 0.5927854938271605,
      "grad_norm": 0.9882687926292419,
      "learning_rate": 0.0005,
      "loss": 2.597169876098633,
      "step": 3073
    },
    {
      "epoch": 0.5929783950617284,
      "grad_norm": 1.0440512895584106,
      "learning_rate": 0.0005,
      "loss": 2.47430419921875,
      "step": 3074
    },
    {
      "epoch": 0.5931712962962963,
      "grad_norm": 1.0228776931762695,
      "learning_rate": 0.0005,
      "loss": 2.5036816596984863,
      "step": 3075
    },
    {
      "epoch": 0.5933641975308642,
      "grad_norm": 0.989732563495636,
      "learning_rate": 0.0005,
      "loss": 2.500410556793213,
      "step": 3076
    },
    {
      "epoch": 0.5935570987654321,
      "grad_norm": 0.9843257069587708,
      "learning_rate": 0.0005,
      "loss": 2.3501436710357666,
      "step": 3077
    },
    {
      "epoch": 0.59375,
      "grad_norm": 1.281200647354126,
      "learning_rate": 0.0005,
      "loss": 2.645308494567871,
      "step": 3078
    },
    {
      "epoch": 0.5939429012345679,
      "grad_norm": 0.9009104371070862,
      "learning_rate": 0.0005,
      "loss": 2.458557605743408,
      "step": 3079
    },
    {
      "epoch": 0.5941358024691358,
      "grad_norm": 1.156733751296997,
      "learning_rate": 0.0005,
      "loss": 2.681600332260132,
      "step": 3080
    },
    {
      "epoch": 0.5943287037037037,
      "grad_norm": 0.7873286008834839,
      "learning_rate": 0.0005,
      "loss": 2.6489953994750977,
      "step": 3081
    },
    {
      "epoch": 0.5945216049382716,
      "grad_norm": 0.9296462535858154,
      "learning_rate": 0.0005,
      "loss": 2.714571475982666,
      "step": 3082
    },
    {
      "epoch": 0.5947145061728395,
      "grad_norm": 0.9214134216308594,
      "learning_rate": 0.0005,
      "loss": 2.8602566719055176,
      "step": 3083
    },
    {
      "epoch": 0.5949074074074074,
      "grad_norm": 1.0592200756072998,
      "learning_rate": 0.0005,
      "loss": 2.661468267440796,
      "step": 3084
    },
    {
      "epoch": 0.5951003086419753,
      "grad_norm": 0.9128767251968384,
      "learning_rate": 0.0005,
      "loss": 2.6858654022216797,
      "step": 3085
    },
    {
      "epoch": 0.5952932098765432,
      "grad_norm": 1.2884341478347778,
      "learning_rate": 0.0005,
      "loss": 2.6484599113464355,
      "step": 3086
    },
    {
      "epoch": 0.5954861111111112,
      "grad_norm": 0.82351154088974,
      "learning_rate": 0.0005,
      "loss": 2.5117263793945312,
      "step": 3087
    },
    {
      "epoch": 0.595679012345679,
      "grad_norm": 0.8648329973220825,
      "learning_rate": 0.0005,
      "loss": 2.8433475494384766,
      "step": 3088
    },
    {
      "epoch": 0.5958719135802469,
      "grad_norm": 0.8886135816574097,
      "learning_rate": 0.0005,
      "loss": 2.3840625286102295,
      "step": 3089
    },
    {
      "epoch": 0.5960648148148148,
      "grad_norm": 0.9401894807815552,
      "learning_rate": 0.0005,
      "loss": 2.644299030303955,
      "step": 3090
    },
    {
      "epoch": 0.5962577160493827,
      "grad_norm": 1.0998895168304443,
      "learning_rate": 0.0005,
      "loss": 2.5433855056762695,
      "step": 3091
    },
    {
      "epoch": 0.5964506172839507,
      "grad_norm": 1.1793991327285767,
      "learning_rate": 0.0005,
      "loss": 2.3997459411621094,
      "step": 3092
    },
    {
      "epoch": 0.5966435185185185,
      "grad_norm": 1.1988099813461304,
      "learning_rate": 0.0005,
      "loss": 2.7553415298461914,
      "step": 3093
    },
    {
      "epoch": 0.5968364197530864,
      "grad_norm": 1.08368718624115,
      "learning_rate": 0.0005,
      "loss": 2.6205637454986572,
      "step": 3094
    },
    {
      "epoch": 0.5970293209876543,
      "grad_norm": 0.9268056750297546,
      "learning_rate": 0.0005,
      "loss": 2.740802764892578,
      "step": 3095
    },
    {
      "epoch": 0.5972222222222222,
      "grad_norm": 0.8442102670669556,
      "learning_rate": 0.0005,
      "loss": 2.755281925201416,
      "step": 3096
    },
    {
      "epoch": 0.5974151234567902,
      "grad_norm": 0.8285454511642456,
      "learning_rate": 0.0005,
      "loss": 2.529210090637207,
      "step": 3097
    },
    {
      "epoch": 0.597608024691358,
      "grad_norm": 0.7494497299194336,
      "learning_rate": 0.0005,
      "loss": 2.5795884132385254,
      "step": 3098
    },
    {
      "epoch": 0.5978009259259259,
      "grad_norm": 0.942424476146698,
      "learning_rate": 0.0005,
      "loss": 2.550989866256714,
      "step": 3099
    },
    {
      "epoch": 0.5979938271604939,
      "grad_norm": 0.9123691916465759,
      "learning_rate": 0.0005,
      "loss": 2.593510389328003,
      "step": 3100
    },
    {
      "epoch": 0.5981867283950617,
      "grad_norm": 0.9325128197669983,
      "learning_rate": 0.0005,
      "loss": 2.740870237350464,
      "step": 3101
    },
    {
      "epoch": 0.5983796296296297,
      "grad_norm": 1.0145212411880493,
      "learning_rate": 0.0005,
      "loss": 2.291477918624878,
      "step": 3102
    },
    {
      "epoch": 0.5985725308641975,
      "grad_norm": 2.0173416137695312,
      "learning_rate": 0.0005,
      "loss": 2.643598794937134,
      "step": 3103
    },
    {
      "epoch": 0.5987654320987654,
      "grad_norm": 0.9948301911354065,
      "learning_rate": 0.0005,
      "loss": 2.394685745239258,
      "step": 3104
    },
    {
      "epoch": 0.5989583333333334,
      "grad_norm": 1.415831446647644,
      "learning_rate": 0.0005,
      "loss": 2.6582393646240234,
      "step": 3105
    },
    {
      "epoch": 0.5991512345679012,
      "grad_norm": 0.9460482001304626,
      "learning_rate": 0.0005,
      "loss": 2.449709177017212,
      "step": 3106
    },
    {
      "epoch": 0.5993441358024691,
      "grad_norm": 1.5655051469802856,
      "learning_rate": 0.0005,
      "loss": 2.67525577545166,
      "step": 3107
    },
    {
      "epoch": 0.5995370370370371,
      "grad_norm": 0.8798902630805969,
      "learning_rate": 0.0005,
      "loss": 2.598348379135132,
      "step": 3108
    },
    {
      "epoch": 0.5997299382716049,
      "grad_norm": 1.060033917427063,
      "learning_rate": 0.0005,
      "loss": 2.5259876251220703,
      "step": 3109
    },
    {
      "epoch": 0.5999228395061729,
      "grad_norm": 0.9141260385513306,
      "learning_rate": 0.0005,
      "loss": 2.403968572616577,
      "step": 3110
    },
    {
      "epoch": 0.6001157407407407,
      "grad_norm": 1.2321648597717285,
      "learning_rate": 0.0005,
      "loss": 2.6419806480407715,
      "step": 3111
    },
    {
      "epoch": 0.6003086419753086,
      "grad_norm": 1.0657243728637695,
      "learning_rate": 0.0005,
      "loss": 2.3499770164489746,
      "step": 3112
    },
    {
      "epoch": 0.6005015432098766,
      "grad_norm": 1.0635095834732056,
      "learning_rate": 0.0005,
      "loss": 2.244401454925537,
      "step": 3113
    },
    {
      "epoch": 0.6006944444444444,
      "grad_norm": 1.0656328201293945,
      "learning_rate": 0.0005,
      "loss": 2.4548726081848145,
      "step": 3114
    },
    {
      "epoch": 0.6008873456790124,
      "grad_norm": 1.080718994140625,
      "learning_rate": 0.0005,
      "loss": 2.5416126251220703,
      "step": 3115
    },
    {
      "epoch": 0.6010802469135802,
      "grad_norm": 1.065901756286621,
      "learning_rate": 0.0005,
      "loss": 2.602926254272461,
      "step": 3116
    },
    {
      "epoch": 0.6012731481481481,
      "grad_norm": 1.1244171857833862,
      "learning_rate": 0.0005,
      "loss": 2.365874767303467,
      "step": 3117
    },
    {
      "epoch": 0.6014660493827161,
      "grad_norm": 1.2589374780654907,
      "learning_rate": 0.0005,
      "loss": 2.949016571044922,
      "step": 3118
    },
    {
      "epoch": 0.6016589506172839,
      "grad_norm": 1.1468888521194458,
      "learning_rate": 0.0005,
      "loss": 2.5835137367248535,
      "step": 3119
    },
    {
      "epoch": 0.6018518518518519,
      "grad_norm": 1.024679183959961,
      "learning_rate": 0.0005,
      "loss": 2.409762144088745,
      "step": 3120
    },
    {
      "epoch": 0.6020447530864198,
      "grad_norm": 0.9237158298492432,
      "learning_rate": 0.0005,
      "loss": 2.1839985847473145,
      "step": 3121
    },
    {
      "epoch": 0.6022376543209876,
      "grad_norm": 1.0999767780303955,
      "learning_rate": 0.0005,
      "loss": 2.423220157623291,
      "step": 3122
    },
    {
      "epoch": 0.6024305555555556,
      "grad_norm": 1.2920382022857666,
      "learning_rate": 0.0005,
      "loss": 2.652341842651367,
      "step": 3123
    },
    {
      "epoch": 0.6026234567901234,
      "grad_norm": 1.3045142889022827,
      "learning_rate": 0.0005,
      "loss": 2.9039268493652344,
      "step": 3124
    },
    {
      "epoch": 0.6028163580246914,
      "grad_norm": 0.8731772899627686,
      "learning_rate": 0.0005,
      "loss": 2.623671054840088,
      "step": 3125
    },
    {
      "epoch": 0.6030092592592593,
      "grad_norm": 0.8633471131324768,
      "learning_rate": 0.0005,
      "loss": 2.556926727294922,
      "step": 3126
    },
    {
      "epoch": 0.6032021604938271,
      "grad_norm": 1.0543001890182495,
      "learning_rate": 0.0005,
      "loss": 2.342686176300049,
      "step": 3127
    },
    {
      "epoch": 0.6033950617283951,
      "grad_norm": 0.9917211532592773,
      "learning_rate": 0.0005,
      "loss": 2.358912467956543,
      "step": 3128
    },
    {
      "epoch": 0.6035879629629629,
      "grad_norm": 0.9096359014511108,
      "learning_rate": 0.0005,
      "loss": 2.353900909423828,
      "step": 3129
    },
    {
      "epoch": 0.6037808641975309,
      "grad_norm": 1.4991837739944458,
      "learning_rate": 0.0005,
      "loss": 2.9462642669677734,
      "step": 3130
    },
    {
      "epoch": 0.6039737654320988,
      "grad_norm": 1.3728958368301392,
      "learning_rate": 0.0005,
      "loss": 2.8408548831939697,
      "step": 3131
    },
    {
      "epoch": 0.6041666666666666,
      "grad_norm": 1.0970300436019897,
      "learning_rate": 0.0005,
      "loss": 2.54412841796875,
      "step": 3132
    },
    {
      "epoch": 0.6043595679012346,
      "grad_norm": 0.7754305601119995,
      "learning_rate": 0.0005,
      "loss": 2.5678834915161133,
      "step": 3133
    },
    {
      "epoch": 0.6045524691358025,
      "grad_norm": 0.9863206148147583,
      "learning_rate": 0.0005,
      "loss": 2.564021110534668,
      "step": 3134
    },
    {
      "epoch": 0.6047453703703703,
      "grad_norm": 1.0031208992004395,
      "learning_rate": 0.0005,
      "loss": 2.9406073093414307,
      "step": 3135
    },
    {
      "epoch": 0.6049382716049383,
      "grad_norm": 3.1157524585723877,
      "learning_rate": 0.0005,
      "loss": 3.0394206047058105,
      "step": 3136
    },
    {
      "epoch": 0.6051311728395061,
      "grad_norm": 1.4037437438964844,
      "learning_rate": 0.0005,
      "loss": 2.8342514038085938,
      "step": 3137
    },
    {
      "epoch": 0.6053240740740741,
      "grad_norm": 1.0066275596618652,
      "learning_rate": 0.0005,
      "loss": 2.571843147277832,
      "step": 3138
    },
    {
      "epoch": 0.605516975308642,
      "grad_norm": 0.86602783203125,
      "learning_rate": 0.0005,
      "loss": 2.522881031036377,
      "step": 3139
    },
    {
      "epoch": 0.6057098765432098,
      "grad_norm": 0.9290468096733093,
      "learning_rate": 0.0005,
      "loss": 2.5945162773132324,
      "step": 3140
    },
    {
      "epoch": 0.6059027777777778,
      "grad_norm": 0.8485398292541504,
      "learning_rate": 0.0005,
      "loss": 2.505849838256836,
      "step": 3141
    },
    {
      "epoch": 0.6060956790123457,
      "grad_norm": 0.9679067134857178,
      "learning_rate": 0.0005,
      "loss": 2.504855155944824,
      "step": 3142
    },
    {
      "epoch": 0.6062885802469136,
      "grad_norm": 0.8911868929862976,
      "learning_rate": 0.0005,
      "loss": 2.142838954925537,
      "step": 3143
    },
    {
      "epoch": 0.6064814814814815,
      "grad_norm": 1.1928519010543823,
      "learning_rate": 0.0005,
      "loss": 2.6685447692871094,
      "step": 3144
    },
    {
      "epoch": 0.6066743827160493,
      "grad_norm": 0.9613635540008545,
      "learning_rate": 0.0005,
      "loss": 2.5113134384155273,
      "step": 3145
    },
    {
      "epoch": 0.6068672839506173,
      "grad_norm": 0.9878867864608765,
      "learning_rate": 0.0005,
      "loss": 2.7137179374694824,
      "step": 3146
    },
    {
      "epoch": 0.6070601851851852,
      "grad_norm": 0.8602712154388428,
      "learning_rate": 0.0005,
      "loss": 2.427065372467041,
      "step": 3147
    },
    {
      "epoch": 0.6072530864197531,
      "grad_norm": 1.0389480590820312,
      "learning_rate": 0.0005,
      "loss": 2.358949661254883,
      "step": 3148
    },
    {
      "epoch": 0.607445987654321,
      "grad_norm": 1.1401605606079102,
      "learning_rate": 0.0005,
      "loss": 2.6204044818878174,
      "step": 3149
    },
    {
      "epoch": 0.6076388888888888,
      "grad_norm": 1.05484938621521,
      "learning_rate": 0.0005,
      "loss": 2.6347808837890625,
      "step": 3150
    },
    {
      "epoch": 0.6078317901234568,
      "grad_norm": 0.7222098112106323,
      "learning_rate": 0.0005,
      "loss": 2.393235683441162,
      "step": 3151
    },
    {
      "epoch": 0.6080246913580247,
      "grad_norm": 1.2081298828125,
      "learning_rate": 0.0005,
      "loss": 2.293506145477295,
      "step": 3152
    },
    {
      "epoch": 0.6082175925925926,
      "grad_norm": 0.9441677927970886,
      "learning_rate": 0.0005,
      "loss": 2.8002991676330566,
      "step": 3153
    },
    {
      "epoch": 0.6084104938271605,
      "grad_norm": 0.8346012830734253,
      "learning_rate": 0.0005,
      "loss": 2.591662883758545,
      "step": 3154
    },
    {
      "epoch": 0.6086033950617284,
      "grad_norm": 0.870704174041748,
      "learning_rate": 0.0005,
      "loss": 2.345353126525879,
      "step": 3155
    },
    {
      "epoch": 0.6087962962962963,
      "grad_norm": 0.9873252511024475,
      "learning_rate": 0.0005,
      "loss": 2.621262550354004,
      "step": 3156
    },
    {
      "epoch": 0.6089891975308642,
      "grad_norm": 1.0431759357452393,
      "learning_rate": 0.0005,
      "loss": 2.764413356781006,
      "step": 3157
    },
    {
      "epoch": 0.6091820987654321,
      "grad_norm": 0.9615329504013062,
      "learning_rate": 0.0005,
      "loss": 2.3351874351501465,
      "step": 3158
    },
    {
      "epoch": 0.609375,
      "grad_norm": 0.9526799321174622,
      "learning_rate": 0.0005,
      "loss": 2.605301856994629,
      "step": 3159
    },
    {
      "epoch": 0.6095679012345679,
      "grad_norm": 0.9437113404273987,
      "learning_rate": 0.0005,
      "loss": 2.725592613220215,
      "step": 3160
    },
    {
      "epoch": 0.6097608024691358,
      "grad_norm": 0.9254540205001831,
      "learning_rate": 0.0005,
      "loss": 2.4208006858825684,
      "step": 3161
    },
    {
      "epoch": 0.6099537037037037,
      "grad_norm": 0.8511457443237305,
      "learning_rate": 0.0005,
      "loss": 2.5324482917785645,
      "step": 3162
    },
    {
      "epoch": 0.6101466049382716,
      "grad_norm": 0.8364353775978088,
      "learning_rate": 0.0005,
      "loss": 2.704054117202759,
      "step": 3163
    },
    {
      "epoch": 0.6103395061728395,
      "grad_norm": 0.9259713292121887,
      "learning_rate": 0.0005,
      "loss": 2.6020686626434326,
      "step": 3164
    },
    {
      "epoch": 0.6105324074074074,
      "grad_norm": 0.9632112383842468,
      "learning_rate": 0.0005,
      "loss": 2.474432945251465,
      "step": 3165
    },
    {
      "epoch": 0.6107253086419753,
      "grad_norm": 0.9603617787361145,
      "learning_rate": 0.0005,
      "loss": 2.525738477706909,
      "step": 3166
    },
    {
      "epoch": 0.6109182098765432,
      "grad_norm": 0.9075571894645691,
      "learning_rate": 0.0005,
      "loss": 2.800916910171509,
      "step": 3167
    },
    {
      "epoch": 0.6111111111111112,
      "grad_norm": 1.068078875541687,
      "learning_rate": 0.0005,
      "loss": 2.503005027770996,
      "step": 3168
    },
    {
      "epoch": 0.611304012345679,
      "grad_norm": 1.131089210510254,
      "learning_rate": 0.0005,
      "loss": 2.480130910873413,
      "step": 3169
    },
    {
      "epoch": 0.6114969135802469,
      "grad_norm": 1.4725817441940308,
      "learning_rate": 0.0005,
      "loss": 2.590798854827881,
      "step": 3170
    },
    {
      "epoch": 0.6116898148148148,
      "grad_norm": 0.8383456468582153,
      "learning_rate": 0.0005,
      "loss": 2.5091183185577393,
      "step": 3171
    },
    {
      "epoch": 0.6118827160493827,
      "grad_norm": 1.0886870622634888,
      "learning_rate": 0.0005,
      "loss": 2.303436756134033,
      "step": 3172
    },
    {
      "epoch": 0.6120756172839507,
      "grad_norm": 0.9649300575256348,
      "learning_rate": 0.0005,
      "loss": 2.3724889755249023,
      "step": 3173
    },
    {
      "epoch": 0.6122685185185185,
      "grad_norm": 0.9117333292961121,
      "learning_rate": 0.0005,
      "loss": 2.574759006500244,
      "step": 3174
    },
    {
      "epoch": 0.6124614197530864,
      "grad_norm": 0.9268929362297058,
      "learning_rate": 0.0005,
      "loss": 2.4635024070739746,
      "step": 3175
    },
    {
      "epoch": 0.6126543209876543,
      "grad_norm": 0.9543294310569763,
      "learning_rate": 0.0005,
      "loss": 2.7454721927642822,
      "step": 3176
    },
    {
      "epoch": 0.6128472222222222,
      "grad_norm": 0.9927197098731995,
      "learning_rate": 0.0005,
      "loss": 2.361351728439331,
      "step": 3177
    },
    {
      "epoch": 0.6130401234567902,
      "grad_norm": 1.002642273902893,
      "learning_rate": 0.0005,
      "loss": 2.6457438468933105,
      "step": 3178
    },
    {
      "epoch": 0.613233024691358,
      "grad_norm": 0.9556636214256287,
      "learning_rate": 0.0005,
      "loss": 2.838122606277466,
      "step": 3179
    },
    {
      "epoch": 0.6134259259259259,
      "grad_norm": 1.1268442869186401,
      "learning_rate": 0.0005,
      "loss": 2.5726675987243652,
      "step": 3180
    },
    {
      "epoch": 0.6136188271604939,
      "grad_norm": 0.8022611141204834,
      "learning_rate": 0.0005,
      "loss": 2.7939679622650146,
      "step": 3181
    },
    {
      "epoch": 0.6138117283950617,
      "grad_norm": 0.9011068940162659,
      "learning_rate": 0.0005,
      "loss": 2.5127511024475098,
      "step": 3182
    },
    {
      "epoch": 0.6140046296296297,
      "grad_norm": 1.0792242288589478,
      "learning_rate": 0.0005,
      "loss": 2.6915807723999023,
      "step": 3183
    },
    {
      "epoch": 0.6141975308641975,
      "grad_norm": 0.8330584764480591,
      "learning_rate": 0.0005,
      "loss": 2.4488797187805176,
      "step": 3184
    },
    {
      "epoch": 0.6143904320987654,
      "grad_norm": 0.9982746243476868,
      "learning_rate": 0.0005,
      "loss": 2.357673168182373,
      "step": 3185
    },
    {
      "epoch": 0.6145833333333334,
      "grad_norm": 1.0467700958251953,
      "learning_rate": 0.0005,
      "loss": 2.222900867462158,
      "step": 3186
    },
    {
      "epoch": 0.6147762345679012,
      "grad_norm": 1.1191706657409668,
      "learning_rate": 0.0005,
      "loss": 2.788238525390625,
      "step": 3187
    },
    {
      "epoch": 0.6149691358024691,
      "grad_norm": 0.9783146977424622,
      "learning_rate": 0.0005,
      "loss": 2.1874148845672607,
      "step": 3188
    },
    {
      "epoch": 0.6151620370370371,
      "grad_norm": 1.054616093635559,
      "learning_rate": 0.0005,
      "loss": 2.751979351043701,
      "step": 3189
    },
    {
      "epoch": 0.6153549382716049,
      "grad_norm": 1.0292437076568604,
      "learning_rate": 0.0005,
      "loss": 2.431171417236328,
      "step": 3190
    },
    {
      "epoch": 0.6155478395061729,
      "grad_norm": 1.4133189916610718,
      "learning_rate": 0.0005,
      "loss": 2.780601978302002,
      "step": 3191
    },
    {
      "epoch": 0.6157407407407407,
      "grad_norm": 0.9710149168968201,
      "learning_rate": 0.0005,
      "loss": 2.6223738193511963,
      "step": 3192
    },
    {
      "epoch": 0.6159336419753086,
      "grad_norm": 0.9144916534423828,
      "learning_rate": 0.0005,
      "loss": 2.8459537029266357,
      "step": 3193
    },
    {
      "epoch": 0.6161265432098766,
      "grad_norm": 0.9574494957923889,
      "learning_rate": 0.0005,
      "loss": 2.3686275482177734,
      "step": 3194
    },
    {
      "epoch": 0.6163194444444444,
      "grad_norm": 1.050054907798767,
      "learning_rate": 0.0005,
      "loss": 2.566592216491699,
      "step": 3195
    },
    {
      "epoch": 0.6165123456790124,
      "grad_norm": 0.9381875991821289,
      "learning_rate": 0.0005,
      "loss": 2.5065603256225586,
      "step": 3196
    },
    {
      "epoch": 0.6167052469135802,
      "grad_norm": 1.3161442279815674,
      "learning_rate": 0.0005,
      "loss": 2.879770278930664,
      "step": 3197
    },
    {
      "epoch": 0.6168981481481481,
      "grad_norm": 1.0665686130523682,
      "learning_rate": 0.0005,
      "loss": 2.296884298324585,
      "step": 3198
    },
    {
      "epoch": 0.6170910493827161,
      "grad_norm": 0.9169780015945435,
      "learning_rate": 0.0005,
      "loss": 2.5827291011810303,
      "step": 3199
    },
    {
      "epoch": 0.6172839506172839,
      "grad_norm": 0.9684618711471558,
      "learning_rate": 0.0005,
      "loss": 2.331981658935547,
      "step": 3200
    },
    {
      "epoch": 0.6174768518518519,
      "grad_norm": 1.0151481628417969,
      "learning_rate": 0.0005,
      "loss": 2.8826608657836914,
      "step": 3201
    },
    {
      "epoch": 0.6176697530864198,
      "grad_norm": 1.3117631673812866,
      "learning_rate": 0.0005,
      "loss": 2.7109060287475586,
      "step": 3202
    },
    {
      "epoch": 0.6178626543209876,
      "grad_norm": 1.0775617361068726,
      "learning_rate": 0.0005,
      "loss": 2.5503597259521484,
      "step": 3203
    },
    {
      "epoch": 0.6180555555555556,
      "grad_norm": 0.9244205951690674,
      "learning_rate": 0.0005,
      "loss": 2.7474100589752197,
      "step": 3204
    },
    {
      "epoch": 0.6182484567901234,
      "grad_norm": 1.0642344951629639,
      "learning_rate": 0.0005,
      "loss": 2.867084503173828,
      "step": 3205
    },
    {
      "epoch": 0.6184413580246914,
      "grad_norm": 1.031493067741394,
      "learning_rate": 0.0005,
      "loss": 2.3722333908081055,
      "step": 3206
    },
    {
      "epoch": 0.6186342592592593,
      "grad_norm": 1.0260226726531982,
      "learning_rate": 0.0005,
      "loss": 2.5469510555267334,
      "step": 3207
    },
    {
      "epoch": 0.6188271604938271,
      "grad_norm": 0.8604705333709717,
      "learning_rate": 0.0005,
      "loss": 2.5888147354125977,
      "step": 3208
    },
    {
      "epoch": 0.6190200617283951,
      "grad_norm": 0.9273309707641602,
      "learning_rate": 0.0005,
      "loss": 2.410022735595703,
      "step": 3209
    },
    {
      "epoch": 0.6192129629629629,
      "grad_norm": 1.2876570224761963,
      "learning_rate": 0.0005,
      "loss": 2.6291940212249756,
      "step": 3210
    },
    {
      "epoch": 0.6194058641975309,
      "grad_norm": 0.8376878499984741,
      "learning_rate": 0.0005,
      "loss": 2.5160341262817383,
      "step": 3211
    },
    {
      "epoch": 0.6195987654320988,
      "grad_norm": 1.0026054382324219,
      "learning_rate": 0.0005,
      "loss": 2.5839967727661133,
      "step": 3212
    },
    {
      "epoch": 0.6197916666666666,
      "grad_norm": 0.929095983505249,
      "learning_rate": 0.0005,
      "loss": 2.3375420570373535,
      "step": 3213
    },
    {
      "epoch": 0.6199845679012346,
      "grad_norm": 1.3437025547027588,
      "learning_rate": 0.0005,
      "loss": 2.500323534011841,
      "step": 3214
    },
    {
      "epoch": 0.6201774691358025,
      "grad_norm": 1.0352917909622192,
      "learning_rate": 0.0005,
      "loss": 2.399146556854248,
      "step": 3215
    },
    {
      "epoch": 0.6203703703703703,
      "grad_norm": 0.9666303992271423,
      "learning_rate": 0.0005,
      "loss": 2.3024001121520996,
      "step": 3216
    },
    {
      "epoch": 0.6205632716049383,
      "grad_norm": 1.2623027563095093,
      "learning_rate": 0.0005,
      "loss": 2.4808359146118164,
      "step": 3217
    },
    {
      "epoch": 0.6207561728395061,
      "grad_norm": 1.1376432180404663,
      "learning_rate": 0.0005,
      "loss": 2.490548849105835,
      "step": 3218
    },
    {
      "epoch": 0.6209490740740741,
      "grad_norm": 1.0976918935775757,
      "learning_rate": 0.0005,
      "loss": 2.7577404975891113,
      "step": 3219
    },
    {
      "epoch": 0.621141975308642,
      "grad_norm": 0.9828844666481018,
      "learning_rate": 0.0005,
      "loss": 2.6460695266723633,
      "step": 3220
    },
    {
      "epoch": 0.6213348765432098,
      "grad_norm": 0.9449618458747864,
      "learning_rate": 0.0005,
      "loss": 2.543696403503418,
      "step": 3221
    },
    {
      "epoch": 0.6215277777777778,
      "grad_norm": 0.8460173606872559,
      "learning_rate": 0.0005,
      "loss": 2.5460755825042725,
      "step": 3222
    },
    {
      "epoch": 0.6217206790123457,
      "grad_norm": 0.7948652505874634,
      "learning_rate": 0.0005,
      "loss": 2.410203695297241,
      "step": 3223
    },
    {
      "epoch": 0.6219135802469136,
      "grad_norm": 0.9194586873054504,
      "learning_rate": 0.0005,
      "loss": 2.4492273330688477,
      "step": 3224
    },
    {
      "epoch": 0.6221064814814815,
      "grad_norm": 0.978191077709198,
      "learning_rate": 0.0005,
      "loss": 2.4139153957366943,
      "step": 3225
    },
    {
      "epoch": 0.6222993827160493,
      "grad_norm": 0.8118739128112793,
      "learning_rate": 0.0005,
      "loss": 2.256722927093506,
      "step": 3226
    },
    {
      "epoch": 0.6224922839506173,
      "grad_norm": 1.0648419857025146,
      "learning_rate": 0.0005,
      "loss": 2.6065163612365723,
      "step": 3227
    },
    {
      "epoch": 0.6226851851851852,
      "grad_norm": 1.1985114812850952,
      "learning_rate": 0.0005,
      "loss": 2.832988739013672,
      "step": 3228
    },
    {
      "epoch": 0.6228780864197531,
      "grad_norm": 0.8772439360618591,
      "learning_rate": 0.0005,
      "loss": 2.322723388671875,
      "step": 3229
    },
    {
      "epoch": 0.623070987654321,
      "grad_norm": 0.9317454695701599,
      "learning_rate": 0.0005,
      "loss": 2.840799570083618,
      "step": 3230
    },
    {
      "epoch": 0.6232638888888888,
      "grad_norm": 0.8143913149833679,
      "learning_rate": 0.0005,
      "loss": 2.4078190326690674,
      "step": 3231
    },
    {
      "epoch": 0.6234567901234568,
      "grad_norm": 0.7706437706947327,
      "learning_rate": 0.0005,
      "loss": 2.5850138664245605,
      "step": 3232
    },
    {
      "epoch": 0.6236496913580247,
      "grad_norm": 0.7383720278739929,
      "learning_rate": 0.0005,
      "loss": 2.3794422149658203,
      "step": 3233
    },
    {
      "epoch": 0.6238425925925926,
      "grad_norm": 2.4285974502563477,
      "learning_rate": 0.0005,
      "loss": 2.498310089111328,
      "step": 3234
    },
    {
      "epoch": 0.6240354938271605,
      "grad_norm": 0.7175798416137695,
      "learning_rate": 0.0005,
      "loss": 2.4022488594055176,
      "step": 3235
    },
    {
      "epoch": 0.6242283950617284,
      "grad_norm": 0.9867821335792542,
      "learning_rate": 0.0005,
      "loss": 2.354231834411621,
      "step": 3236
    },
    {
      "epoch": 0.6244212962962963,
      "grad_norm": 0.8917577266693115,
      "learning_rate": 0.0005,
      "loss": 2.9437341690063477,
      "step": 3237
    },
    {
      "epoch": 0.6246141975308642,
      "grad_norm": 0.86712646484375,
      "learning_rate": 0.0005,
      "loss": 2.596574068069458,
      "step": 3238
    },
    {
      "epoch": 0.6248070987654321,
      "grad_norm": 1.004791021347046,
      "learning_rate": 0.0005,
      "loss": 2.350536346435547,
      "step": 3239
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.9304865002632141,
      "learning_rate": 0.0005,
      "loss": 2.575040340423584,
      "step": 3240
    },
    {
      "epoch": 0.6251929012345679,
      "grad_norm": 0.8402833938598633,
      "learning_rate": 0.0005,
      "loss": 2.533327102661133,
      "step": 3241
    },
    {
      "epoch": 0.6253858024691358,
      "grad_norm": 6.172116756439209,
      "learning_rate": 0.0005,
      "loss": 2.124115467071533,
      "step": 3242
    },
    {
      "epoch": 0.6255787037037037,
      "grad_norm": 0.9125242233276367,
      "learning_rate": 0.0005,
      "loss": 2.3583929538726807,
      "step": 3243
    },
    {
      "epoch": 0.6257716049382716,
      "grad_norm": 0.9085598587989807,
      "learning_rate": 0.0005,
      "loss": 2.9501333236694336,
      "step": 3244
    },
    {
      "epoch": 0.6259645061728395,
      "grad_norm": 0.9838830828666687,
      "learning_rate": 0.0005,
      "loss": 2.4814510345458984,
      "step": 3245
    },
    {
      "epoch": 0.6261574074074074,
      "grad_norm": 0.7162665724754333,
      "learning_rate": 0.0005,
      "loss": 2.465527057647705,
      "step": 3246
    },
    {
      "epoch": 0.6263503086419753,
      "grad_norm": 1.2466298341751099,
      "learning_rate": 0.0005,
      "loss": 2.7421600818634033,
      "step": 3247
    },
    {
      "epoch": 0.6265432098765432,
      "grad_norm": 0.7437573075294495,
      "learning_rate": 0.0005,
      "loss": 2.6224522590637207,
      "step": 3248
    },
    {
      "epoch": 0.6267361111111112,
      "grad_norm": 0.9808120727539062,
      "learning_rate": 0.0005,
      "loss": 2.4968724250793457,
      "step": 3249
    },
    {
      "epoch": 0.626929012345679,
      "grad_norm": 1.0274622440338135,
      "learning_rate": 0.0005,
      "loss": 2.229506731033325,
      "step": 3250
    },
    {
      "epoch": 0.6271219135802469,
      "grad_norm": 0.9628419876098633,
      "learning_rate": 0.0005,
      "loss": 2.508502960205078,
      "step": 3251
    },
    {
      "epoch": 0.6273148148148148,
      "grad_norm": 1.050758957862854,
      "learning_rate": 0.0005,
      "loss": 2.483400583267212,
      "step": 3252
    },
    {
      "epoch": 0.6275077160493827,
      "grad_norm": 1.1486786603927612,
      "learning_rate": 0.0005,
      "loss": 2.606194496154785,
      "step": 3253
    },
    {
      "epoch": 0.6277006172839507,
      "grad_norm": 0.9252153635025024,
      "learning_rate": 0.0005,
      "loss": 2.3506920337677,
      "step": 3254
    },
    {
      "epoch": 0.6278935185185185,
      "grad_norm": 0.885216474533081,
      "learning_rate": 0.0005,
      "loss": 2.576230049133301,
      "step": 3255
    },
    {
      "epoch": 0.6280864197530864,
      "grad_norm": 0.9247933030128479,
      "learning_rate": 0.0005,
      "loss": 2.489753484725952,
      "step": 3256
    },
    {
      "epoch": 0.6282793209876543,
      "grad_norm": 1.12150239944458,
      "learning_rate": 0.0005,
      "loss": 2.300384998321533,
      "step": 3257
    },
    {
      "epoch": 0.6284722222222222,
      "grad_norm": 1.3168649673461914,
      "learning_rate": 0.0005,
      "loss": 2.4190335273742676,
      "step": 3258
    },
    {
      "epoch": 0.6286651234567902,
      "grad_norm": 0.9230600595474243,
      "learning_rate": 0.0005,
      "loss": 2.2421445846557617,
      "step": 3259
    },
    {
      "epoch": 0.628858024691358,
      "grad_norm": 1.0666850805282593,
      "learning_rate": 0.0005,
      "loss": 2.6231980323791504,
      "step": 3260
    },
    {
      "epoch": 0.6290509259259259,
      "grad_norm": 0.9701528549194336,
      "learning_rate": 0.0005,
      "loss": 2.4009103775024414,
      "step": 3261
    },
    {
      "epoch": 0.6292438271604939,
      "grad_norm": 0.9245839715003967,
      "learning_rate": 0.0005,
      "loss": 2.7147574424743652,
      "step": 3262
    },
    {
      "epoch": 0.6294367283950617,
      "grad_norm": 1.3497035503387451,
      "learning_rate": 0.0005,
      "loss": 2.4509520530700684,
      "step": 3263
    },
    {
      "epoch": 0.6296296296296297,
      "grad_norm": 1.0608885288238525,
      "learning_rate": 0.0005,
      "loss": 2.559971332550049,
      "step": 3264
    },
    {
      "epoch": 0.6298225308641975,
      "grad_norm": 0.8377110362052917,
      "learning_rate": 0.0005,
      "loss": 2.515723705291748,
      "step": 3265
    },
    {
      "epoch": 0.6300154320987654,
      "grad_norm": 1.014129400253296,
      "learning_rate": 0.0005,
      "loss": 2.8914105892181396,
      "step": 3266
    },
    {
      "epoch": 0.6302083333333334,
      "grad_norm": 1.0049947500228882,
      "learning_rate": 0.0005,
      "loss": 2.5779972076416016,
      "step": 3267
    },
    {
      "epoch": 0.6304012345679012,
      "grad_norm": 1.1585136651992798,
      "learning_rate": 0.0005,
      "loss": 2.8793859481811523,
      "step": 3268
    },
    {
      "epoch": 0.6305941358024691,
      "grad_norm": 0.8836054801940918,
      "learning_rate": 0.0005,
      "loss": 2.444190263748169,
      "step": 3269
    },
    {
      "epoch": 0.6307870370370371,
      "grad_norm": 0.9819629192352295,
      "learning_rate": 0.0005,
      "loss": 2.4410319328308105,
      "step": 3270
    },
    {
      "epoch": 0.6309799382716049,
      "grad_norm": 0.8130590915679932,
      "learning_rate": 0.0005,
      "loss": 2.5016028881073,
      "step": 3271
    },
    {
      "epoch": 0.6311728395061729,
      "grad_norm": 1.1877081394195557,
      "learning_rate": 0.0005,
      "loss": 2.596450090408325,
      "step": 3272
    },
    {
      "epoch": 0.6313657407407407,
      "grad_norm": 0.7852888107299805,
      "learning_rate": 0.0005,
      "loss": 2.1815567016601562,
      "step": 3273
    },
    {
      "epoch": 0.6315586419753086,
      "grad_norm": 0.8694871664047241,
      "learning_rate": 0.0005,
      "loss": 2.2635107040405273,
      "step": 3274
    },
    {
      "epoch": 0.6317515432098766,
      "grad_norm": 1.860409140586853,
      "learning_rate": 0.0005,
      "loss": 2.8640377521514893,
      "step": 3275
    },
    {
      "epoch": 0.6319444444444444,
      "grad_norm": 0.9741032123565674,
      "learning_rate": 0.0005,
      "loss": 2.355350971221924,
      "step": 3276
    },
    {
      "epoch": 0.6321373456790124,
      "grad_norm": 0.8734800815582275,
      "learning_rate": 0.0005,
      "loss": 2.5247254371643066,
      "step": 3277
    },
    {
      "epoch": 0.6323302469135802,
      "grad_norm": 0.898216187953949,
      "learning_rate": 0.0005,
      "loss": 2.457047939300537,
      "step": 3278
    },
    {
      "epoch": 0.6325231481481481,
      "grad_norm": 1.0115026235580444,
      "learning_rate": 0.0005,
      "loss": 2.2610669136047363,
      "step": 3279
    },
    {
      "epoch": 0.6327160493827161,
      "grad_norm": 0.8907600045204163,
      "learning_rate": 0.0005,
      "loss": 2.428452253341675,
      "step": 3280
    },
    {
      "epoch": 0.6329089506172839,
      "grad_norm": 0.9316566586494446,
      "learning_rate": 0.0005,
      "loss": 2.3467345237731934,
      "step": 3281
    },
    {
      "epoch": 0.6331018518518519,
      "grad_norm": 0.8313990235328674,
      "learning_rate": 0.0005,
      "loss": 2.6292648315429688,
      "step": 3282
    },
    {
      "epoch": 0.6332947530864198,
      "grad_norm": 1.297088384628296,
      "learning_rate": 0.0005,
      "loss": 2.5700247287750244,
      "step": 3283
    },
    {
      "epoch": 0.6334876543209876,
      "grad_norm": 1.4070854187011719,
      "learning_rate": 0.0005,
      "loss": 2.4519705772399902,
      "step": 3284
    },
    {
      "epoch": 0.6336805555555556,
      "grad_norm": 1.2404242753982544,
      "learning_rate": 0.0005,
      "loss": 2.6977906227111816,
      "step": 3285
    },
    {
      "epoch": 0.6338734567901234,
      "grad_norm": 0.7875744104385376,
      "learning_rate": 0.0005,
      "loss": 2.1547744274139404,
      "step": 3286
    },
    {
      "epoch": 0.6340663580246914,
      "grad_norm": 1.224967360496521,
      "learning_rate": 0.0005,
      "loss": 2.6283302307128906,
      "step": 3287
    },
    {
      "epoch": 0.6342592592592593,
      "grad_norm": 0.9609954357147217,
      "learning_rate": 0.0005,
      "loss": 2.394805908203125,
      "step": 3288
    },
    {
      "epoch": 0.6344521604938271,
      "grad_norm": 0.8626090288162231,
      "learning_rate": 0.0005,
      "loss": 2.4005849361419678,
      "step": 3289
    },
    {
      "epoch": 0.6346450617283951,
      "grad_norm": 0.7950068116188049,
      "learning_rate": 0.0005,
      "loss": 2.6010844707489014,
      "step": 3290
    },
    {
      "epoch": 0.6348379629629629,
      "grad_norm": 0.9726915955543518,
      "learning_rate": 0.0005,
      "loss": 2.47477388381958,
      "step": 3291
    },
    {
      "epoch": 0.6350308641975309,
      "grad_norm": 1.0088231563568115,
      "learning_rate": 0.0005,
      "loss": 2.566007614135742,
      "step": 3292
    },
    {
      "epoch": 0.6352237654320988,
      "grad_norm": 0.9443956017494202,
      "learning_rate": 0.0005,
      "loss": 2.5119071006774902,
      "step": 3293
    },
    {
      "epoch": 0.6354166666666666,
      "grad_norm": 0.9297096729278564,
      "learning_rate": 0.0005,
      "loss": 2.727954387664795,
      "step": 3294
    },
    {
      "epoch": 0.6356095679012346,
      "grad_norm": 0.94853675365448,
      "learning_rate": 0.0005,
      "loss": 2.4636940956115723,
      "step": 3295
    },
    {
      "epoch": 0.6358024691358025,
      "grad_norm": 0.8609843254089355,
      "learning_rate": 0.0005,
      "loss": 2.468750476837158,
      "step": 3296
    },
    {
      "epoch": 0.6359953703703703,
      "grad_norm": 0.9804984331130981,
      "learning_rate": 0.0005,
      "loss": 2.301286458969116,
      "step": 3297
    },
    {
      "epoch": 0.6361882716049383,
      "grad_norm": 0.9897876381874084,
      "learning_rate": 0.0005,
      "loss": 2.626645565032959,
      "step": 3298
    },
    {
      "epoch": 0.6363811728395061,
      "grad_norm": 1.1286207437515259,
      "learning_rate": 0.0005,
      "loss": 2.5425004959106445,
      "step": 3299
    },
    {
      "epoch": 0.6365740740740741,
      "grad_norm": 0.9805856943130493,
      "learning_rate": 0.0005,
      "loss": 2.5657482147216797,
      "step": 3300
    },
    {
      "epoch": 0.636766975308642,
      "grad_norm": 0.9857253432273865,
      "learning_rate": 0.0005,
      "loss": 2.1978354454040527,
      "step": 3301
    },
    {
      "epoch": 0.6369598765432098,
      "grad_norm": 0.9253895878791809,
      "learning_rate": 0.0005,
      "loss": 2.185698986053467,
      "step": 3302
    },
    {
      "epoch": 0.6371527777777778,
      "grad_norm": 0.8612286448478699,
      "learning_rate": 0.0005,
      "loss": 2.589047908782959,
      "step": 3303
    },
    {
      "epoch": 0.6373456790123457,
      "grad_norm": 0.8023710250854492,
      "learning_rate": 0.0005,
      "loss": 2.2657830715179443,
      "step": 3304
    },
    {
      "epoch": 0.6375385802469136,
      "grad_norm": 0.9039671421051025,
      "learning_rate": 0.0005,
      "loss": 2.344027042388916,
      "step": 3305
    },
    {
      "epoch": 0.6377314814814815,
      "grad_norm": 1.1088643074035645,
      "learning_rate": 0.0005,
      "loss": 2.567368268966675,
      "step": 3306
    },
    {
      "epoch": 0.6379243827160493,
      "grad_norm": 1.0181066989898682,
      "learning_rate": 0.0005,
      "loss": 2.553506374359131,
      "step": 3307
    },
    {
      "epoch": 0.6381172839506173,
      "grad_norm": 1.0960928201675415,
      "learning_rate": 0.0005,
      "loss": 2.401348829269409,
      "step": 3308
    },
    {
      "epoch": 0.6383101851851852,
      "grad_norm": 0.9108807444572449,
      "learning_rate": 0.0005,
      "loss": 2.617628335952759,
      "step": 3309
    },
    {
      "epoch": 0.6385030864197531,
      "grad_norm": 1.1979024410247803,
      "learning_rate": 0.0005,
      "loss": 2.4775421619415283,
      "step": 3310
    },
    {
      "epoch": 0.638695987654321,
      "grad_norm": 1.097753643989563,
      "learning_rate": 0.0005,
      "loss": 2.3505423069000244,
      "step": 3311
    },
    {
      "epoch": 0.6388888888888888,
      "grad_norm": 1.0474189519882202,
      "learning_rate": 0.0005,
      "loss": 2.287198066711426,
      "step": 3312
    },
    {
      "epoch": 0.6390817901234568,
      "grad_norm": 0.9272539019584656,
      "learning_rate": 0.0005,
      "loss": 2.39186954498291,
      "step": 3313
    },
    {
      "epoch": 0.6392746913580247,
      "grad_norm": 0.9663146734237671,
      "learning_rate": 0.0005,
      "loss": 2.4775991439819336,
      "step": 3314
    },
    {
      "epoch": 0.6394675925925926,
      "grad_norm": 1.042309045791626,
      "learning_rate": 0.0005,
      "loss": 2.4840517044067383,
      "step": 3315
    },
    {
      "epoch": 0.6396604938271605,
      "grad_norm": 0.8560612797737122,
      "learning_rate": 0.0005,
      "loss": 2.3455686569213867,
      "step": 3316
    },
    {
      "epoch": 0.6398533950617284,
      "grad_norm": 0.9145386219024658,
      "learning_rate": 0.0005,
      "loss": 2.7222719192504883,
      "step": 3317
    },
    {
      "epoch": 0.6400462962962963,
      "grad_norm": 1.2337045669555664,
      "learning_rate": 0.0005,
      "loss": 2.8545725345611572,
      "step": 3318
    },
    {
      "epoch": 0.6402391975308642,
      "grad_norm": 1.155775785446167,
      "learning_rate": 0.0005,
      "loss": 2.4275574684143066,
      "step": 3319
    },
    {
      "epoch": 0.6404320987654321,
      "grad_norm": 0.9480698704719543,
      "learning_rate": 0.0005,
      "loss": 2.090208053588867,
      "step": 3320
    },
    {
      "epoch": 0.640625,
      "grad_norm": 0.8889000415802002,
      "learning_rate": 0.0005,
      "loss": 2.518019199371338,
      "step": 3321
    },
    {
      "epoch": 0.6408179012345679,
      "grad_norm": 0.9961493015289307,
      "learning_rate": 0.0005,
      "loss": 2.3691563606262207,
      "step": 3322
    },
    {
      "epoch": 0.6410108024691358,
      "grad_norm": 1.425723671913147,
      "learning_rate": 0.0005,
      "loss": 2.444187879562378,
      "step": 3323
    },
    {
      "epoch": 0.6412037037037037,
      "grad_norm": 1.0167917013168335,
      "learning_rate": 0.0005,
      "loss": 2.635458469390869,
      "step": 3324
    },
    {
      "epoch": 0.6413966049382716,
      "grad_norm": 1.0222697257995605,
      "learning_rate": 0.0005,
      "loss": 2.4029998779296875,
      "step": 3325
    },
    {
      "epoch": 0.6415895061728395,
      "grad_norm": 1.3106600046157837,
      "learning_rate": 0.0005,
      "loss": 3.0760982036590576,
      "step": 3326
    },
    {
      "epoch": 0.6417824074074074,
      "grad_norm": 0.7622988820075989,
      "learning_rate": 0.0005,
      "loss": 2.6370997428894043,
      "step": 3327
    },
    {
      "epoch": 0.6419753086419753,
      "grad_norm": 1.6448675394058228,
      "learning_rate": 0.0005,
      "loss": 2.6568291187286377,
      "step": 3328
    },
    {
      "epoch": 0.6421682098765432,
      "grad_norm": 0.8644151091575623,
      "learning_rate": 0.0005,
      "loss": 2.606297492980957,
      "step": 3329
    },
    {
      "epoch": 0.6423611111111112,
      "grad_norm": 1.0372774600982666,
      "learning_rate": 0.0005,
      "loss": 2.12113618850708,
      "step": 3330
    },
    {
      "epoch": 0.642554012345679,
      "grad_norm": 1.1236608028411865,
      "learning_rate": 0.0005,
      "loss": 2.394425868988037,
      "step": 3331
    },
    {
      "epoch": 0.6427469135802469,
      "grad_norm": 1.2134835720062256,
      "learning_rate": 0.0005,
      "loss": 2.5052032470703125,
      "step": 3332
    },
    {
      "epoch": 0.6429398148148148,
      "grad_norm": 1.361159324645996,
      "learning_rate": 0.0005,
      "loss": 2.820727825164795,
      "step": 3333
    },
    {
      "epoch": 0.6431327160493827,
      "grad_norm": 1.0521947145462036,
      "learning_rate": 0.0005,
      "loss": 2.3181023597717285,
      "step": 3334
    },
    {
      "epoch": 0.6433256172839507,
      "grad_norm": 0.9624614715576172,
      "learning_rate": 0.0005,
      "loss": 2.643576145172119,
      "step": 3335
    },
    {
      "epoch": 0.6435185185185185,
      "grad_norm": 1.5515050888061523,
      "learning_rate": 0.0005,
      "loss": 2.5506436824798584,
      "step": 3336
    },
    {
      "epoch": 0.6437114197530864,
      "grad_norm": 0.8758318424224854,
      "learning_rate": 0.0005,
      "loss": 2.761343002319336,
      "step": 3337
    },
    {
      "epoch": 0.6439043209876543,
      "grad_norm": 0.936232328414917,
      "learning_rate": 0.0005,
      "loss": 2.7582929134368896,
      "step": 3338
    },
    {
      "epoch": 0.6440972222222222,
      "grad_norm": 0.8349128365516663,
      "learning_rate": 0.0005,
      "loss": 2.139707088470459,
      "step": 3339
    },
    {
      "epoch": 0.6442901234567902,
      "grad_norm": 0.8461217284202576,
      "learning_rate": 0.0005,
      "loss": 2.456818103790283,
      "step": 3340
    },
    {
      "epoch": 0.644483024691358,
      "grad_norm": 0.8925907611846924,
      "learning_rate": 0.0005,
      "loss": 2.6798465251922607,
      "step": 3341
    },
    {
      "epoch": 0.6446759259259259,
      "grad_norm": 1.1689926385879517,
      "learning_rate": 0.0005,
      "loss": 2.7082061767578125,
      "step": 3342
    },
    {
      "epoch": 0.6448688271604939,
      "grad_norm": 2.1491384506225586,
      "learning_rate": 0.0005,
      "loss": 2.5438966751098633,
      "step": 3343
    },
    {
      "epoch": 0.6450617283950617,
      "grad_norm": 0.8379679322242737,
      "learning_rate": 0.0005,
      "loss": 2.3043887615203857,
      "step": 3344
    },
    {
      "epoch": 0.6452546296296297,
      "grad_norm": 1.1682488918304443,
      "learning_rate": 0.0005,
      "loss": 2.534665107727051,
      "step": 3345
    },
    {
      "epoch": 0.6454475308641975,
      "grad_norm": 0.8877894878387451,
      "learning_rate": 0.0005,
      "loss": 2.5290820598602295,
      "step": 3346
    },
    {
      "epoch": 0.6456404320987654,
      "grad_norm": 0.9810742735862732,
      "learning_rate": 0.0005,
      "loss": 2.310762882232666,
      "step": 3347
    },
    {
      "epoch": 0.6458333333333334,
      "grad_norm": 0.9539774656295776,
      "learning_rate": 0.0005,
      "loss": 2.6188690662384033,
      "step": 3348
    },
    {
      "epoch": 0.6460262345679012,
      "grad_norm": 0.8248100280761719,
      "learning_rate": 0.0005,
      "loss": 2.538102149963379,
      "step": 3349
    },
    {
      "epoch": 0.6462191358024691,
      "grad_norm": 0.954107940196991,
      "learning_rate": 0.0005,
      "loss": 2.5134336948394775,
      "step": 3350
    },
    {
      "epoch": 0.6464120370370371,
      "grad_norm": 1.0954664945602417,
      "learning_rate": 0.0005,
      "loss": 2.604271411895752,
      "step": 3351
    },
    {
      "epoch": 0.6466049382716049,
      "grad_norm": 1.0695557594299316,
      "learning_rate": 0.0005,
      "loss": 2.5404632091522217,
      "step": 3352
    },
    {
      "epoch": 0.6467978395061729,
      "grad_norm": 0.96425461769104,
      "learning_rate": 0.0005,
      "loss": 2.459737777709961,
      "step": 3353
    },
    {
      "epoch": 0.6469907407407407,
      "grad_norm": 1.020169734954834,
      "learning_rate": 0.0005,
      "loss": 2.368525981903076,
      "step": 3354
    },
    {
      "epoch": 0.6471836419753086,
      "grad_norm": 1.0571208000183105,
      "learning_rate": 0.0005,
      "loss": 2.370980739593506,
      "step": 3355
    },
    {
      "epoch": 0.6473765432098766,
      "grad_norm": 1.139815330505371,
      "learning_rate": 0.0005,
      "loss": 2.3163459300994873,
      "step": 3356
    },
    {
      "epoch": 0.6475694444444444,
      "grad_norm": 0.8852735161781311,
      "learning_rate": 0.0005,
      "loss": 2.569662094116211,
      "step": 3357
    },
    {
      "epoch": 0.6477623456790124,
      "grad_norm": 1.0312247276306152,
      "learning_rate": 0.0005,
      "loss": 2.6377696990966797,
      "step": 3358
    },
    {
      "epoch": 0.6479552469135802,
      "grad_norm": 0.6795680522918701,
      "learning_rate": 0.0005,
      "loss": 2.204946756362915,
      "step": 3359
    },
    {
      "epoch": 0.6481481481481481,
      "grad_norm": 0.7502221465110779,
      "learning_rate": 0.0005,
      "loss": 2.7407021522521973,
      "step": 3360
    },
    {
      "epoch": 0.6483410493827161,
      "grad_norm": 1.4307422637939453,
      "learning_rate": 0.0005,
      "loss": 2.7021408081054688,
      "step": 3361
    },
    {
      "epoch": 0.6485339506172839,
      "grad_norm": 0.8834598064422607,
      "learning_rate": 0.0005,
      "loss": 2.571549415588379,
      "step": 3362
    },
    {
      "epoch": 0.6487268518518519,
      "grad_norm": 0.9889119267463684,
      "learning_rate": 0.0005,
      "loss": 2.4538681507110596,
      "step": 3363
    },
    {
      "epoch": 0.6489197530864198,
      "grad_norm": 1.2482556104660034,
      "learning_rate": 0.0005,
      "loss": 2.742506980895996,
      "step": 3364
    },
    {
      "epoch": 0.6491126543209876,
      "grad_norm": 0.9997930526733398,
      "learning_rate": 0.0005,
      "loss": 2.439516067504883,
      "step": 3365
    },
    {
      "epoch": 0.6493055555555556,
      "grad_norm": 1.1651371717453003,
      "learning_rate": 0.0005,
      "loss": 2.5979933738708496,
      "step": 3366
    },
    {
      "epoch": 0.6494984567901234,
      "grad_norm": 1.5881969928741455,
      "learning_rate": 0.0005,
      "loss": 2.5462820529937744,
      "step": 3367
    },
    {
      "epoch": 0.6496913580246914,
      "grad_norm": 1.3988213539123535,
      "learning_rate": 0.0005,
      "loss": 2.477271318435669,
      "step": 3368
    },
    {
      "epoch": 0.6498842592592593,
      "grad_norm": 1.0541954040527344,
      "learning_rate": 0.0005,
      "loss": 2.628518581390381,
      "step": 3369
    },
    {
      "epoch": 0.6500771604938271,
      "grad_norm": 0.896094799041748,
      "learning_rate": 0.0005,
      "loss": 2.338677406311035,
      "step": 3370
    },
    {
      "epoch": 0.6502700617283951,
      "grad_norm": 0.9251517057418823,
      "learning_rate": 0.0005,
      "loss": 2.3227405548095703,
      "step": 3371
    },
    {
      "epoch": 0.6504629629629629,
      "grad_norm": 0.9182258248329163,
      "learning_rate": 0.0005,
      "loss": 2.625091075897217,
      "step": 3372
    },
    {
      "epoch": 0.6506558641975309,
      "grad_norm": 0.7779707312583923,
      "learning_rate": 0.0005,
      "loss": 2.1774682998657227,
      "step": 3373
    },
    {
      "epoch": 0.6508487654320988,
      "grad_norm": 0.9437519907951355,
      "learning_rate": 0.0005,
      "loss": 2.544229507446289,
      "step": 3374
    },
    {
      "epoch": 0.6510416666666666,
      "grad_norm": 0.9870719313621521,
      "learning_rate": 0.0005,
      "loss": 2.307572364807129,
      "step": 3375
    },
    {
      "epoch": 0.6512345679012346,
      "grad_norm": 1.0916569232940674,
      "learning_rate": 0.0005,
      "loss": 2.3665480613708496,
      "step": 3376
    },
    {
      "epoch": 0.6514274691358025,
      "grad_norm": 1.085174322128296,
      "learning_rate": 0.0005,
      "loss": 2.4326725006103516,
      "step": 3377
    },
    {
      "epoch": 0.6516203703703703,
      "grad_norm": 1.1112431287765503,
      "learning_rate": 0.0005,
      "loss": 2.4498019218444824,
      "step": 3378
    },
    {
      "epoch": 0.6518132716049383,
      "grad_norm": 0.9397437572479248,
      "learning_rate": 0.0005,
      "loss": 2.492870330810547,
      "step": 3379
    },
    {
      "epoch": 0.6520061728395061,
      "grad_norm": 0.9166593551635742,
      "learning_rate": 0.0005,
      "loss": 2.246826171875,
      "step": 3380
    },
    {
      "epoch": 0.6521990740740741,
      "grad_norm": 0.9024108648300171,
      "learning_rate": 0.0005,
      "loss": 2.7053146362304688,
      "step": 3381
    },
    {
      "epoch": 0.652391975308642,
      "grad_norm": 0.8847792744636536,
      "learning_rate": 0.0005,
      "loss": 2.6206655502319336,
      "step": 3382
    },
    {
      "epoch": 0.6525848765432098,
      "grad_norm": 0.8755400776863098,
      "learning_rate": 0.0005,
      "loss": 2.561455726623535,
      "step": 3383
    },
    {
      "epoch": 0.6527777777777778,
      "grad_norm": 0.9146119952201843,
      "learning_rate": 0.0005,
      "loss": 2.5382041931152344,
      "step": 3384
    },
    {
      "epoch": 0.6529706790123457,
      "grad_norm": 0.876432478427887,
      "learning_rate": 0.0005,
      "loss": 2.6765851974487305,
      "step": 3385
    },
    {
      "epoch": 0.6531635802469136,
      "grad_norm": 0.9236694574356079,
      "learning_rate": 0.0005,
      "loss": 2.7289419174194336,
      "step": 3386
    },
    {
      "epoch": 0.6533564814814815,
      "grad_norm": 0.8628737330436707,
      "learning_rate": 0.0005,
      "loss": 2.6637516021728516,
      "step": 3387
    },
    {
      "epoch": 0.6535493827160493,
      "grad_norm": 1.0210175514221191,
      "learning_rate": 0.0005,
      "loss": 2.3231470584869385,
      "step": 3388
    },
    {
      "epoch": 0.6537422839506173,
      "grad_norm": 0.9491406679153442,
      "learning_rate": 0.0005,
      "loss": 2.1742677688598633,
      "step": 3389
    },
    {
      "epoch": 0.6539351851851852,
      "grad_norm": 0.8317239880561829,
      "learning_rate": 0.0005,
      "loss": 2.336488962173462,
      "step": 3390
    },
    {
      "epoch": 0.6541280864197531,
      "grad_norm": 1.4549037218093872,
      "learning_rate": 0.0005,
      "loss": 2.4063196182250977,
      "step": 3391
    },
    {
      "epoch": 0.654320987654321,
      "grad_norm": 0.694835901260376,
      "learning_rate": 0.0005,
      "loss": 2.6096606254577637,
      "step": 3392
    },
    {
      "epoch": 0.6545138888888888,
      "grad_norm": 1.1961740255355835,
      "learning_rate": 0.0005,
      "loss": 2.3672397136688232,
      "step": 3393
    },
    {
      "epoch": 0.6547067901234568,
      "grad_norm": 1.1885513067245483,
      "learning_rate": 0.0005,
      "loss": 2.465719223022461,
      "step": 3394
    },
    {
      "epoch": 0.6548996913580247,
      "grad_norm": 0.9899243116378784,
      "learning_rate": 0.0005,
      "loss": 2.498311996459961,
      "step": 3395
    },
    {
      "epoch": 0.6550925925925926,
      "grad_norm": 1.223934531211853,
      "learning_rate": 0.0005,
      "loss": 2.542778253555298,
      "step": 3396
    },
    {
      "epoch": 0.6552854938271605,
      "grad_norm": 0.7554685473442078,
      "learning_rate": 0.0005,
      "loss": 2.9395389556884766,
      "step": 3397
    },
    {
      "epoch": 0.6554783950617284,
      "grad_norm": 0.8614925146102905,
      "learning_rate": 0.0005,
      "loss": 2.661740303039551,
      "step": 3398
    },
    {
      "epoch": 0.6556712962962963,
      "grad_norm": 0.8219576478004456,
      "learning_rate": 0.0005,
      "loss": 2.450897455215454,
      "step": 3399
    },
    {
      "epoch": 0.6558641975308642,
      "grad_norm": 0.9707040190696716,
      "learning_rate": 0.0005,
      "loss": 2.2253332138061523,
      "step": 3400
    },
    {
      "epoch": 0.6560570987654321,
      "grad_norm": 0.7446278929710388,
      "learning_rate": 0.0005,
      "loss": 2.650648832321167,
      "step": 3401
    },
    {
      "epoch": 0.65625,
      "grad_norm": 0.8506100177764893,
      "learning_rate": 0.0005,
      "loss": 2.4990930557250977,
      "step": 3402
    },
    {
      "epoch": 0.6564429012345679,
      "grad_norm": 1.1348844766616821,
      "learning_rate": 0.0005,
      "loss": 2.4257187843322754,
      "step": 3403
    },
    {
      "epoch": 0.6566358024691358,
      "grad_norm": 0.9535156488418579,
      "learning_rate": 0.0005,
      "loss": 2.698606014251709,
      "step": 3404
    },
    {
      "epoch": 0.6568287037037037,
      "grad_norm": 1.1171513795852661,
      "learning_rate": 0.0005,
      "loss": 2.662566661834717,
      "step": 3405
    },
    {
      "epoch": 0.6570216049382716,
      "grad_norm": 0.7066720724105835,
      "learning_rate": 0.0005,
      "loss": 2.238511323928833,
      "step": 3406
    },
    {
      "epoch": 0.6572145061728395,
      "grad_norm": 1.3331127166748047,
      "learning_rate": 0.0005,
      "loss": 2.4221787452697754,
      "step": 3407
    },
    {
      "epoch": 0.6574074074074074,
      "grad_norm": 0.7924807071685791,
      "learning_rate": 0.0005,
      "loss": 2.481778144836426,
      "step": 3408
    },
    {
      "epoch": 0.6576003086419753,
      "grad_norm": 0.8476426005363464,
      "learning_rate": 0.0005,
      "loss": 2.24534273147583,
      "step": 3409
    },
    {
      "epoch": 0.6577932098765432,
      "grad_norm": 0.9128028154373169,
      "learning_rate": 0.0005,
      "loss": 2.7168869972229004,
      "step": 3410
    },
    {
      "epoch": 0.6579861111111112,
      "grad_norm": 1.0102169513702393,
      "learning_rate": 0.0005,
      "loss": 2.3986029624938965,
      "step": 3411
    },
    {
      "epoch": 0.658179012345679,
      "grad_norm": 0.9250069856643677,
      "learning_rate": 0.0005,
      "loss": 2.380516529083252,
      "step": 3412
    },
    {
      "epoch": 0.6583719135802469,
      "grad_norm": 1.0473989248275757,
      "learning_rate": 0.0005,
      "loss": 2.7878098487854004,
      "step": 3413
    },
    {
      "epoch": 0.6585648148148148,
      "grad_norm": 0.8264528512954712,
      "learning_rate": 0.0005,
      "loss": 2.66414213180542,
      "step": 3414
    },
    {
      "epoch": 0.6587577160493827,
      "grad_norm": 0.8791531324386597,
      "learning_rate": 0.0005,
      "loss": 1.9455220699310303,
      "step": 3415
    },
    {
      "epoch": 0.6589506172839507,
      "grad_norm": 1.2943921089172363,
      "learning_rate": 0.0005,
      "loss": 2.6924073696136475,
      "step": 3416
    },
    {
      "epoch": 0.6591435185185185,
      "grad_norm": 1.0667084455490112,
      "learning_rate": 0.0005,
      "loss": 2.8597681522369385,
      "step": 3417
    },
    {
      "epoch": 0.6593364197530864,
      "grad_norm": 0.9091923236846924,
      "learning_rate": 0.0005,
      "loss": 2.436413049697876,
      "step": 3418
    },
    {
      "epoch": 0.6595293209876543,
      "grad_norm": 0.7584468126296997,
      "learning_rate": 0.0005,
      "loss": 2.639749526977539,
      "step": 3419
    },
    {
      "epoch": 0.6597222222222222,
      "grad_norm": 0.9781434535980225,
      "learning_rate": 0.0005,
      "loss": 2.4743800163269043,
      "step": 3420
    },
    {
      "epoch": 0.6599151234567902,
      "grad_norm": 0.8973546624183655,
      "learning_rate": 0.0005,
      "loss": 2.7178611755371094,
      "step": 3421
    },
    {
      "epoch": 0.660108024691358,
      "grad_norm": 0.9819142818450928,
      "learning_rate": 0.0005,
      "loss": 2.6025800704956055,
      "step": 3422
    },
    {
      "epoch": 0.6603009259259259,
      "grad_norm": 0.8717110753059387,
      "learning_rate": 0.0005,
      "loss": 2.374041795730591,
      "step": 3423
    },
    {
      "epoch": 0.6604938271604939,
      "grad_norm": 1.010524034500122,
      "learning_rate": 0.0005,
      "loss": 2.5518484115600586,
      "step": 3424
    },
    {
      "epoch": 0.6606867283950617,
      "grad_norm": 0.9407289028167725,
      "learning_rate": 0.0005,
      "loss": 2.66218638420105,
      "step": 3425
    },
    {
      "epoch": 0.6608796296296297,
      "grad_norm": 0.8226357102394104,
      "learning_rate": 0.0005,
      "loss": 2.438870668411255,
      "step": 3426
    },
    {
      "epoch": 0.6610725308641975,
      "grad_norm": 0.873884379863739,
      "learning_rate": 0.0005,
      "loss": 2.3865456581115723,
      "step": 3427
    },
    {
      "epoch": 0.6612654320987654,
      "grad_norm": 2.1321916580200195,
      "learning_rate": 0.0005,
      "loss": 2.5129199028015137,
      "step": 3428
    },
    {
      "epoch": 0.6614583333333334,
      "grad_norm": 0.8269005417823792,
      "learning_rate": 0.0005,
      "loss": 2.595120429992676,
      "step": 3429
    },
    {
      "epoch": 0.6616512345679012,
      "grad_norm": 0.9420583248138428,
      "learning_rate": 0.0005,
      "loss": 2.8209714889526367,
      "step": 3430
    },
    {
      "epoch": 0.6618441358024691,
      "grad_norm": 0.9060837626457214,
      "learning_rate": 0.0005,
      "loss": 2.19975209236145,
      "step": 3431
    },
    {
      "epoch": 0.6620370370370371,
      "grad_norm": 0.7173828482627869,
      "learning_rate": 0.0005,
      "loss": 2.5196373462677,
      "step": 3432
    },
    {
      "epoch": 0.6622299382716049,
      "grad_norm": 0.7611615061759949,
      "learning_rate": 0.0005,
      "loss": 2.4816641807556152,
      "step": 3433
    },
    {
      "epoch": 0.6624228395061729,
      "grad_norm": 0.8862019777297974,
      "learning_rate": 0.0005,
      "loss": 2.0876307487487793,
      "step": 3434
    },
    {
      "epoch": 0.6626157407407407,
      "grad_norm": 1.1156216859817505,
      "learning_rate": 0.0005,
      "loss": 2.3299460411071777,
      "step": 3435
    },
    {
      "epoch": 0.6628086419753086,
      "grad_norm": 0.8920363783836365,
      "learning_rate": 0.0005,
      "loss": 2.6335701942443848,
      "step": 3436
    },
    {
      "epoch": 0.6630015432098766,
      "grad_norm": 0.8463515639305115,
      "learning_rate": 0.0005,
      "loss": 2.569777011871338,
      "step": 3437
    },
    {
      "epoch": 0.6631944444444444,
      "grad_norm": 0.9075930714607239,
      "learning_rate": 0.0005,
      "loss": 2.7799935340881348,
      "step": 3438
    },
    {
      "epoch": 0.6633873456790124,
      "grad_norm": 0.8396744132041931,
      "learning_rate": 0.0005,
      "loss": 2.465579032897949,
      "step": 3439
    },
    {
      "epoch": 0.6635802469135802,
      "grad_norm": 1.065885305404663,
      "learning_rate": 0.0005,
      "loss": 2.286332130432129,
      "step": 3440
    },
    {
      "epoch": 0.6637731481481481,
      "grad_norm": 0.9297690391540527,
      "learning_rate": 0.0005,
      "loss": 2.5898728370666504,
      "step": 3441
    },
    {
      "epoch": 0.6639660493827161,
      "grad_norm": 1.028709888458252,
      "learning_rate": 0.0005,
      "loss": 2.4691097736358643,
      "step": 3442
    },
    {
      "epoch": 0.6641589506172839,
      "grad_norm": 1.0841559171676636,
      "learning_rate": 0.0005,
      "loss": 2.209744930267334,
      "step": 3443
    },
    {
      "epoch": 0.6643518518518519,
      "grad_norm": 0.9929103851318359,
      "learning_rate": 0.0005,
      "loss": 2.8445117473602295,
      "step": 3444
    },
    {
      "epoch": 0.6645447530864198,
      "grad_norm": 1.0137606859207153,
      "learning_rate": 0.0005,
      "loss": 2.3263912200927734,
      "step": 3445
    },
    {
      "epoch": 0.6647376543209876,
      "grad_norm": 1.2616349458694458,
      "learning_rate": 0.0005,
      "loss": 2.6471855640411377,
      "step": 3446
    },
    {
      "epoch": 0.6649305555555556,
      "grad_norm": 1.0120700597763062,
      "learning_rate": 0.0005,
      "loss": 2.4507641792297363,
      "step": 3447
    },
    {
      "epoch": 0.6651234567901234,
      "grad_norm": 0.9556608200073242,
      "learning_rate": 0.0005,
      "loss": 2.3735780715942383,
      "step": 3448
    },
    {
      "epoch": 0.6653163580246914,
      "grad_norm": 0.9615316390991211,
      "learning_rate": 0.0005,
      "loss": 2.819251537322998,
      "step": 3449
    },
    {
      "epoch": 0.6655092592592593,
      "grad_norm": 0.9203129410743713,
      "learning_rate": 0.0005,
      "loss": 2.358473777770996,
      "step": 3450
    },
    {
      "epoch": 0.6657021604938271,
      "grad_norm": 0.8332275748252869,
      "learning_rate": 0.0005,
      "loss": 2.3708698749542236,
      "step": 3451
    },
    {
      "epoch": 0.6658950617283951,
      "grad_norm": 0.9389381408691406,
      "learning_rate": 0.0005,
      "loss": 2.5169289112091064,
      "step": 3452
    },
    {
      "epoch": 0.6660879629629629,
      "grad_norm": 0.9879015684127808,
      "learning_rate": 0.0005,
      "loss": 2.5622148513793945,
      "step": 3453
    },
    {
      "epoch": 0.6662808641975309,
      "grad_norm": 0.7914217710494995,
      "learning_rate": 0.0005,
      "loss": 2.680203437805176,
      "step": 3454
    },
    {
      "epoch": 0.6664737654320988,
      "grad_norm": 0.9238910675048828,
      "learning_rate": 0.0005,
      "loss": 2.3384363651275635,
      "step": 3455
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.8369629383087158,
      "learning_rate": 0.0005,
      "loss": 2.2218358516693115,
      "step": 3456
    },
    {
      "epoch": 0.6668595679012346,
      "grad_norm": 0.8416926860809326,
      "learning_rate": 0.0005,
      "loss": 2.5883235931396484,
      "step": 3457
    },
    {
      "epoch": 0.6670524691358025,
      "grad_norm": 1.0062496662139893,
      "learning_rate": 0.0005,
      "loss": 2.5369513034820557,
      "step": 3458
    },
    {
      "epoch": 0.6672453703703703,
      "grad_norm": 1.0088642835617065,
      "learning_rate": 0.0005,
      "loss": 2.4846606254577637,
      "step": 3459
    },
    {
      "epoch": 0.6674382716049383,
      "grad_norm": 0.8975238800048828,
      "learning_rate": 0.0005,
      "loss": 2.5728769302368164,
      "step": 3460
    },
    {
      "epoch": 0.6676311728395061,
      "grad_norm": 0.8994528651237488,
      "learning_rate": 0.0005,
      "loss": 2.5172219276428223,
      "step": 3461
    },
    {
      "epoch": 0.6678240740740741,
      "grad_norm": 1.0429171323776245,
      "learning_rate": 0.0005,
      "loss": 2.4764339923858643,
      "step": 3462
    },
    {
      "epoch": 0.668016975308642,
      "grad_norm": 1.125678539276123,
      "learning_rate": 0.0005,
      "loss": 2.634761333465576,
      "step": 3463
    },
    {
      "epoch": 0.6682098765432098,
      "grad_norm": 0.9800997972488403,
      "learning_rate": 0.0005,
      "loss": 2.3585078716278076,
      "step": 3464
    },
    {
      "epoch": 0.6684027777777778,
      "grad_norm": 0.9159989953041077,
      "learning_rate": 0.0005,
      "loss": 2.6508610248565674,
      "step": 3465
    },
    {
      "epoch": 0.6685956790123457,
      "grad_norm": 1.0454907417297363,
      "learning_rate": 0.0005,
      "loss": 2.5470941066741943,
      "step": 3466
    },
    {
      "epoch": 0.6687885802469136,
      "grad_norm": 1.0345385074615479,
      "learning_rate": 0.0005,
      "loss": 2.52203369140625,
      "step": 3467
    },
    {
      "epoch": 0.6689814814814815,
      "grad_norm": 1.0390533208847046,
      "learning_rate": 0.0005,
      "loss": 2.661383628845215,
      "step": 3468
    },
    {
      "epoch": 0.6691743827160493,
      "grad_norm": 0.8439213037490845,
      "learning_rate": 0.0005,
      "loss": 2.244771957397461,
      "step": 3469
    },
    {
      "epoch": 0.6693672839506173,
      "grad_norm": 1.566389799118042,
      "learning_rate": 0.0005,
      "loss": 2.5827293395996094,
      "step": 3470
    },
    {
      "epoch": 0.6695601851851852,
      "grad_norm": 0.8250911831855774,
      "learning_rate": 0.0005,
      "loss": 2.6766862869262695,
      "step": 3471
    },
    {
      "epoch": 0.6697530864197531,
      "grad_norm": 0.9454893469810486,
      "learning_rate": 0.0005,
      "loss": 2.201368570327759,
      "step": 3472
    },
    {
      "epoch": 0.669945987654321,
      "grad_norm": 0.9182128310203552,
      "learning_rate": 0.0005,
      "loss": 2.492612361907959,
      "step": 3473
    },
    {
      "epoch": 0.6701388888888888,
      "grad_norm": 0.9643323421478271,
      "learning_rate": 0.0005,
      "loss": 2.7176575660705566,
      "step": 3474
    },
    {
      "epoch": 0.6703317901234568,
      "grad_norm": 0.8798574209213257,
      "learning_rate": 0.0005,
      "loss": 2.3619015216827393,
      "step": 3475
    },
    {
      "epoch": 0.6705246913580247,
      "grad_norm": 1.0796386003494263,
      "learning_rate": 0.0005,
      "loss": 2.8058547973632812,
      "step": 3476
    },
    {
      "epoch": 0.6707175925925926,
      "grad_norm": 1.0487008094787598,
      "learning_rate": 0.0005,
      "loss": 2.6950840950012207,
      "step": 3477
    },
    {
      "epoch": 0.6709104938271605,
      "grad_norm": 1.0762195587158203,
      "learning_rate": 0.0005,
      "loss": 2.651876449584961,
      "step": 3478
    },
    {
      "epoch": 0.6711033950617284,
      "grad_norm": 1.1083506345748901,
      "learning_rate": 0.0005,
      "loss": 2.1878628730773926,
      "step": 3479
    },
    {
      "epoch": 0.6712962962962963,
      "grad_norm": 0.9080002307891846,
      "learning_rate": 0.0005,
      "loss": 2.52097487449646,
      "step": 3480
    },
    {
      "epoch": 0.6714891975308642,
      "grad_norm": 1.0832313299179077,
      "learning_rate": 0.0005,
      "loss": 2.3243255615234375,
      "step": 3481
    },
    {
      "epoch": 0.6716820987654321,
      "grad_norm": 0.9003090858459473,
      "learning_rate": 0.0005,
      "loss": 2.4622397422790527,
      "step": 3482
    },
    {
      "epoch": 0.671875,
      "grad_norm": 0.8375616073608398,
      "learning_rate": 0.0005,
      "loss": 2.5423784255981445,
      "step": 3483
    },
    {
      "epoch": 0.6720679012345679,
      "grad_norm": 0.9000348448753357,
      "learning_rate": 0.0005,
      "loss": 2.7406771183013916,
      "step": 3484
    },
    {
      "epoch": 0.6722608024691358,
      "grad_norm": 0.8797689080238342,
      "learning_rate": 0.0005,
      "loss": 2.4936766624450684,
      "step": 3485
    },
    {
      "epoch": 0.6724537037037037,
      "grad_norm": 1.0055197477340698,
      "learning_rate": 0.0005,
      "loss": 2.705587387084961,
      "step": 3486
    },
    {
      "epoch": 0.6726466049382716,
      "grad_norm": 1.5035340785980225,
      "learning_rate": 0.0005,
      "loss": 2.4729108810424805,
      "step": 3487
    },
    {
      "epoch": 0.6728395061728395,
      "grad_norm": 1.2651820182800293,
      "learning_rate": 0.0005,
      "loss": 2.654261827468872,
      "step": 3488
    },
    {
      "epoch": 0.6730324074074074,
      "grad_norm": 1.4829471111297607,
      "learning_rate": 0.0005,
      "loss": 2.4635584354400635,
      "step": 3489
    },
    {
      "epoch": 0.6732253086419753,
      "grad_norm": 0.8216190934181213,
      "learning_rate": 0.0005,
      "loss": 2.410079002380371,
      "step": 3490
    },
    {
      "epoch": 0.6734182098765432,
      "grad_norm": 0.860927939414978,
      "learning_rate": 0.0005,
      "loss": 2.692833662033081,
      "step": 3491
    },
    {
      "epoch": 0.6736111111111112,
      "grad_norm": 0.8784370422363281,
      "learning_rate": 0.0005,
      "loss": 2.5851831436157227,
      "step": 3492
    },
    {
      "epoch": 0.673804012345679,
      "grad_norm": 1.0116490125656128,
      "learning_rate": 0.0005,
      "loss": 2.508213996887207,
      "step": 3493
    },
    {
      "epoch": 0.6739969135802469,
      "grad_norm": 0.8982697129249573,
      "learning_rate": 0.0005,
      "loss": 2.516599416732788,
      "step": 3494
    },
    {
      "epoch": 0.6741898148148148,
      "grad_norm": 0.7452590465545654,
      "learning_rate": 0.0005,
      "loss": 2.4031333923339844,
      "step": 3495
    },
    {
      "epoch": 0.6743827160493827,
      "grad_norm": 1.2749003171920776,
      "learning_rate": 0.0005,
      "loss": 2.743770122528076,
      "step": 3496
    },
    {
      "epoch": 0.6745756172839507,
      "grad_norm": 1.2322719097137451,
      "learning_rate": 0.0005,
      "loss": 2.315458297729492,
      "step": 3497
    },
    {
      "epoch": 0.6747685185185185,
      "grad_norm": 1.4210439920425415,
      "learning_rate": 0.0005,
      "loss": 2.6475396156311035,
      "step": 3498
    },
    {
      "epoch": 0.6749614197530864,
      "grad_norm": 1.0720142126083374,
      "learning_rate": 0.0005,
      "loss": 2.8980658054351807,
      "step": 3499
    },
    {
      "epoch": 0.6751543209876543,
      "grad_norm": 0.9572576284408569,
      "learning_rate": 0.0005,
      "loss": 2.1613831520080566,
      "step": 3500
    },
    {
      "epoch": 0.6753472222222222,
      "grad_norm": 0.8906975388526917,
      "learning_rate": 0.0005,
      "loss": 2.314560890197754,
      "step": 3501
    },
    {
      "epoch": 0.6755401234567902,
      "grad_norm": 0.8305970430374146,
      "learning_rate": 0.0005,
      "loss": 2.267786979675293,
      "step": 3502
    },
    {
      "epoch": 0.675733024691358,
      "grad_norm": 0.8913347125053406,
      "learning_rate": 0.0005,
      "loss": 2.252768039703369,
      "step": 3503
    },
    {
      "epoch": 0.6759259259259259,
      "grad_norm": 1.4920989274978638,
      "learning_rate": 0.0005,
      "loss": 2.323728561401367,
      "step": 3504
    },
    {
      "epoch": 0.6761188271604939,
      "grad_norm": 1.040428638458252,
      "learning_rate": 0.0005,
      "loss": 2.4040849208831787,
      "step": 3505
    },
    {
      "epoch": 0.6763117283950617,
      "grad_norm": 0.9406843185424805,
      "learning_rate": 0.0005,
      "loss": 2.4854815006256104,
      "step": 3506
    },
    {
      "epoch": 0.6765046296296297,
      "grad_norm": 0.9056935906410217,
      "learning_rate": 0.0005,
      "loss": 2.5483202934265137,
      "step": 3507
    },
    {
      "epoch": 0.6766975308641975,
      "grad_norm": 1.0370373725891113,
      "learning_rate": 0.0005,
      "loss": 2.6909961700439453,
      "step": 3508
    },
    {
      "epoch": 0.6768904320987654,
      "grad_norm": 0.8720027208328247,
      "learning_rate": 0.0005,
      "loss": 2.760012626647949,
      "step": 3509
    },
    {
      "epoch": 0.6770833333333334,
      "grad_norm": 0.8699060678482056,
      "learning_rate": 0.0005,
      "loss": 2.79787015914917,
      "step": 3510
    },
    {
      "epoch": 0.6772762345679012,
      "grad_norm": 1.4459912776947021,
      "learning_rate": 0.0005,
      "loss": 2.427499532699585,
      "step": 3511
    },
    {
      "epoch": 0.6774691358024691,
      "grad_norm": 0.9414815306663513,
      "learning_rate": 0.0005,
      "loss": 2.5879979133605957,
      "step": 3512
    },
    {
      "epoch": 0.6776620370370371,
      "grad_norm": 0.8808560371398926,
      "learning_rate": 0.0005,
      "loss": 2.5737948417663574,
      "step": 3513
    },
    {
      "epoch": 0.6778549382716049,
      "grad_norm": 1.3045051097869873,
      "learning_rate": 0.0005,
      "loss": 2.251650810241699,
      "step": 3514
    },
    {
      "epoch": 0.6780478395061729,
      "grad_norm": 1.161165475845337,
      "learning_rate": 0.0005,
      "loss": 2.5351362228393555,
      "step": 3515
    },
    {
      "epoch": 0.6782407407407407,
      "grad_norm": 0.9927918910980225,
      "learning_rate": 0.0005,
      "loss": 2.4198241233825684,
      "step": 3516
    },
    {
      "epoch": 0.6784336419753086,
      "grad_norm": 1.0183199644088745,
      "learning_rate": 0.0005,
      "loss": 2.391404390335083,
      "step": 3517
    },
    {
      "epoch": 0.6786265432098766,
      "grad_norm": 0.8748757243156433,
      "learning_rate": 0.0005,
      "loss": 2.7380661964416504,
      "step": 3518
    },
    {
      "epoch": 0.6788194444444444,
      "grad_norm": 0.9201221466064453,
      "learning_rate": 0.0005,
      "loss": 2.7029318809509277,
      "step": 3519
    },
    {
      "epoch": 0.6790123456790124,
      "grad_norm": 0.8807194232940674,
      "learning_rate": 0.0005,
      "loss": 2.444398880004883,
      "step": 3520
    },
    {
      "epoch": 0.6792052469135802,
      "grad_norm": 0.9143303632736206,
      "learning_rate": 0.0005,
      "loss": 2.442197799682617,
      "step": 3521
    },
    {
      "epoch": 0.6793981481481481,
      "grad_norm": 0.8685644865036011,
      "learning_rate": 0.0005,
      "loss": 2.3258509635925293,
      "step": 3522
    },
    {
      "epoch": 0.6795910493827161,
      "grad_norm": 0.7995749115943909,
      "learning_rate": 0.0005,
      "loss": 2.2241740226745605,
      "step": 3523
    },
    {
      "epoch": 0.6797839506172839,
      "grad_norm": 0.9240672588348389,
      "learning_rate": 0.0005,
      "loss": 2.49383282661438,
      "step": 3524
    },
    {
      "epoch": 0.6799768518518519,
      "grad_norm": 0.960105836391449,
      "learning_rate": 0.0005,
      "loss": 2.3962526321411133,
      "step": 3525
    },
    {
      "epoch": 0.6801697530864198,
      "grad_norm": 1.8213860988616943,
      "learning_rate": 0.0005,
      "loss": 2.2877843379974365,
      "step": 3526
    },
    {
      "epoch": 0.6803626543209876,
      "grad_norm": 0.9375059008598328,
      "learning_rate": 0.0005,
      "loss": 2.4944510459899902,
      "step": 3527
    },
    {
      "epoch": 0.6805555555555556,
      "grad_norm": 0.9256466031074524,
      "learning_rate": 0.0005,
      "loss": 2.8438069820404053,
      "step": 3528
    },
    {
      "epoch": 0.6807484567901234,
      "grad_norm": 0.973822295665741,
      "learning_rate": 0.0005,
      "loss": 2.4122257232666016,
      "step": 3529
    },
    {
      "epoch": 0.6809413580246914,
      "grad_norm": 0.8750673532485962,
      "learning_rate": 0.0005,
      "loss": 2.496490240097046,
      "step": 3530
    },
    {
      "epoch": 0.6811342592592593,
      "grad_norm": 1.0778626203536987,
      "learning_rate": 0.0005,
      "loss": 2.583956480026245,
      "step": 3531
    },
    {
      "epoch": 0.6813271604938271,
      "grad_norm": 0.8132398724555969,
      "learning_rate": 0.0005,
      "loss": 2.338505506515503,
      "step": 3532
    },
    {
      "epoch": 0.6815200617283951,
      "grad_norm": 0.9471129775047302,
      "learning_rate": 0.0005,
      "loss": 2.3234925270080566,
      "step": 3533
    },
    {
      "epoch": 0.6817129629629629,
      "grad_norm": 0.9136598110198975,
      "learning_rate": 0.0005,
      "loss": 2.6204710006713867,
      "step": 3534
    },
    {
      "epoch": 0.6819058641975309,
      "grad_norm": 0.908458948135376,
      "learning_rate": 0.0005,
      "loss": 2.587636947631836,
      "step": 3535
    },
    {
      "epoch": 0.6820987654320988,
      "grad_norm": 1.2423851490020752,
      "learning_rate": 0.0005,
      "loss": 2.2674667835235596,
      "step": 3536
    },
    {
      "epoch": 0.6822916666666666,
      "grad_norm": 0.9476343393325806,
      "learning_rate": 0.0005,
      "loss": 2.1583919525146484,
      "step": 3537
    },
    {
      "epoch": 0.6824845679012346,
      "grad_norm": 0.7721104621887207,
      "learning_rate": 0.0005,
      "loss": 2.571779251098633,
      "step": 3538
    },
    {
      "epoch": 0.6826774691358025,
      "grad_norm": 0.6909951567649841,
      "learning_rate": 0.0005,
      "loss": 2.3995304107666016,
      "step": 3539
    },
    {
      "epoch": 0.6828703703703703,
      "grad_norm": 0.8911614418029785,
      "learning_rate": 0.0005,
      "loss": 2.265939712524414,
      "step": 3540
    },
    {
      "epoch": 0.6830632716049383,
      "grad_norm": 0.8053275942802429,
      "learning_rate": 0.0005,
      "loss": 2.365652561187744,
      "step": 3541
    },
    {
      "epoch": 0.6832561728395061,
      "grad_norm": 1.1205061674118042,
      "learning_rate": 0.0005,
      "loss": 2.700850009918213,
      "step": 3542
    },
    {
      "epoch": 0.6834490740740741,
      "grad_norm": 0.9038776159286499,
      "learning_rate": 0.0005,
      "loss": 2.3417768478393555,
      "step": 3543
    },
    {
      "epoch": 0.683641975308642,
      "grad_norm": 0.8337164521217346,
      "learning_rate": 0.0005,
      "loss": 2.4079396724700928,
      "step": 3544
    },
    {
      "epoch": 0.6838348765432098,
      "grad_norm": 0.7987378835678101,
      "learning_rate": 0.0005,
      "loss": 2.3163881301879883,
      "step": 3545
    },
    {
      "epoch": 0.6840277777777778,
      "grad_norm": 0.8183894753456116,
      "learning_rate": 0.0005,
      "loss": 2.5193116664886475,
      "step": 3546
    },
    {
      "epoch": 0.6842206790123457,
      "grad_norm": 0.8382671475410461,
      "learning_rate": 0.0005,
      "loss": 2.4204885959625244,
      "step": 3547
    },
    {
      "epoch": 0.6844135802469136,
      "grad_norm": 1.1563053131103516,
      "learning_rate": 0.0005,
      "loss": 2.4808969497680664,
      "step": 3548
    },
    {
      "epoch": 0.6846064814814815,
      "grad_norm": 0.7653918862342834,
      "learning_rate": 0.0005,
      "loss": 2.501317024230957,
      "step": 3549
    },
    {
      "epoch": 0.6847993827160493,
      "grad_norm": 1.0036289691925049,
      "learning_rate": 0.0005,
      "loss": 2.399024724960327,
      "step": 3550
    },
    {
      "epoch": 0.6849922839506173,
      "grad_norm": 1.0526797771453857,
      "learning_rate": 0.0005,
      "loss": 2.326827049255371,
      "step": 3551
    },
    {
      "epoch": 0.6851851851851852,
      "grad_norm": 0.8760165572166443,
      "learning_rate": 0.0005,
      "loss": 2.6248726844787598,
      "step": 3552
    },
    {
      "epoch": 0.6853780864197531,
      "grad_norm": 0.8717449903488159,
      "learning_rate": 0.0005,
      "loss": 2.304683208465576,
      "step": 3553
    },
    {
      "epoch": 0.685570987654321,
      "grad_norm": 1.1364177465438843,
      "learning_rate": 0.0005,
      "loss": 2.5666720867156982,
      "step": 3554
    },
    {
      "epoch": 0.6857638888888888,
      "grad_norm": 0.9629111886024475,
      "learning_rate": 0.0005,
      "loss": 2.270200252532959,
      "step": 3555
    },
    {
      "epoch": 0.6859567901234568,
      "grad_norm": 0.8888624310493469,
      "learning_rate": 0.0005,
      "loss": 2.471597194671631,
      "step": 3556
    },
    {
      "epoch": 0.6861496913580247,
      "grad_norm": 1.1936259269714355,
      "learning_rate": 0.0005,
      "loss": 2.489762783050537,
      "step": 3557
    },
    {
      "epoch": 0.6863425925925926,
      "grad_norm": 1.0357141494750977,
      "learning_rate": 0.0005,
      "loss": 2.861678123474121,
      "step": 3558
    },
    {
      "epoch": 0.6865354938271605,
      "grad_norm": 1.0259877443313599,
      "learning_rate": 0.0005,
      "loss": 3.125169277191162,
      "step": 3559
    },
    {
      "epoch": 0.6867283950617284,
      "grad_norm": 1.0958014726638794,
      "learning_rate": 0.0005,
      "loss": 2.6288247108459473,
      "step": 3560
    },
    {
      "epoch": 0.6869212962962963,
      "grad_norm": 0.9364858269691467,
      "learning_rate": 0.0005,
      "loss": 2.38316011428833,
      "step": 3561
    },
    {
      "epoch": 0.6871141975308642,
      "grad_norm": 0.8554614782333374,
      "learning_rate": 0.0005,
      "loss": 2.4536590576171875,
      "step": 3562
    },
    {
      "epoch": 0.6873070987654321,
      "grad_norm": 2.767972230911255,
      "learning_rate": 0.0005,
      "loss": 2.368964672088623,
      "step": 3563
    },
    {
      "epoch": 0.6875,
      "grad_norm": 1.2713673114776611,
      "learning_rate": 0.0005,
      "loss": 2.38392972946167,
      "step": 3564
    },
    {
      "epoch": 0.6876929012345679,
      "grad_norm": 1.0859785079956055,
      "learning_rate": 0.0005,
      "loss": 2.4190473556518555,
      "step": 3565
    },
    {
      "epoch": 0.6878858024691358,
      "grad_norm": 0.875285267829895,
      "learning_rate": 0.0005,
      "loss": 2.583512783050537,
      "step": 3566
    },
    {
      "epoch": 0.6880787037037037,
      "grad_norm": 0.6855881810188293,
      "learning_rate": 0.0005,
      "loss": 2.2222909927368164,
      "step": 3567
    },
    {
      "epoch": 0.6882716049382716,
      "grad_norm": 1.8856559991836548,
      "learning_rate": 0.0005,
      "loss": 2.7204322814941406,
      "step": 3568
    },
    {
      "epoch": 0.6884645061728395,
      "grad_norm": 1.2125468254089355,
      "learning_rate": 0.0005,
      "loss": 2.225303888320923,
      "step": 3569
    },
    {
      "epoch": 0.6886574074074074,
      "grad_norm": 0.9258739352226257,
      "learning_rate": 0.0005,
      "loss": 2.34149169921875,
      "step": 3570
    },
    {
      "epoch": 0.6888503086419753,
      "grad_norm": 0.908066987991333,
      "learning_rate": 0.0005,
      "loss": 2.3356852531433105,
      "step": 3571
    },
    {
      "epoch": 0.6890432098765432,
      "grad_norm": 1.1121749877929688,
      "learning_rate": 0.0005,
      "loss": 2.4484593868255615,
      "step": 3572
    },
    {
      "epoch": 0.6892361111111112,
      "grad_norm": 1.2196929454803467,
      "learning_rate": 0.0005,
      "loss": 2.402113437652588,
      "step": 3573
    },
    {
      "epoch": 0.689429012345679,
      "grad_norm": 2.4888339042663574,
      "learning_rate": 0.0005,
      "loss": 2.690342426300049,
      "step": 3574
    },
    {
      "epoch": 0.6896219135802469,
      "grad_norm": 0.9753000140190125,
      "learning_rate": 0.0005,
      "loss": 2.1884443759918213,
      "step": 3575
    },
    {
      "epoch": 0.6898148148148148,
      "grad_norm": 0.8446171879768372,
      "learning_rate": 0.0005,
      "loss": 2.325244903564453,
      "step": 3576
    },
    {
      "epoch": 0.6900077160493827,
      "grad_norm": 0.8314900994300842,
      "learning_rate": 0.0005,
      "loss": 2.176635265350342,
      "step": 3577
    },
    {
      "epoch": 0.6902006172839507,
      "grad_norm": 0.8803836703300476,
      "learning_rate": 0.0005,
      "loss": 2.229095458984375,
      "step": 3578
    },
    {
      "epoch": 0.6903935185185185,
      "grad_norm": 1.0658531188964844,
      "learning_rate": 0.0005,
      "loss": 2.270433187484741,
      "step": 3579
    },
    {
      "epoch": 0.6905864197530864,
      "grad_norm": 0.9978479743003845,
      "learning_rate": 0.0005,
      "loss": 2.170340061187744,
      "step": 3580
    },
    {
      "epoch": 0.6907793209876543,
      "grad_norm": 0.9237330555915833,
      "learning_rate": 0.0005,
      "loss": 2.5550174713134766,
      "step": 3581
    },
    {
      "epoch": 0.6909722222222222,
      "grad_norm": 0.9709495306015015,
      "learning_rate": 0.0005,
      "loss": 2.634918451309204,
      "step": 3582
    },
    {
      "epoch": 0.6911651234567902,
      "grad_norm": 1.16819429397583,
      "learning_rate": 0.0005,
      "loss": 2.4495458602905273,
      "step": 3583
    },
    {
      "epoch": 0.691358024691358,
      "grad_norm": 0.9305956959724426,
      "learning_rate": 0.0005,
      "loss": 2.3577678203582764,
      "step": 3584
    },
    {
      "epoch": 0.6915509259259259,
      "grad_norm": 1.0796892642974854,
      "learning_rate": 0.0005,
      "loss": 2.7799196243286133,
      "step": 3585
    },
    {
      "epoch": 0.6917438271604939,
      "grad_norm": 0.7892595529556274,
      "learning_rate": 0.0005,
      "loss": 2.4088478088378906,
      "step": 3586
    },
    {
      "epoch": 0.6919367283950617,
      "grad_norm": 0.8359031081199646,
      "learning_rate": 0.0005,
      "loss": 2.619323253631592,
      "step": 3587
    },
    {
      "epoch": 0.6921296296296297,
      "grad_norm": 1.6856286525726318,
      "learning_rate": 0.0005,
      "loss": 2.5285377502441406,
      "step": 3588
    },
    {
      "epoch": 0.6923225308641975,
      "grad_norm": 0.8554012179374695,
      "learning_rate": 0.0005,
      "loss": 2.460477352142334,
      "step": 3589
    },
    {
      "epoch": 0.6925154320987654,
      "grad_norm": 0.8130179047584534,
      "learning_rate": 0.0005,
      "loss": 2.3872506618499756,
      "step": 3590
    },
    {
      "epoch": 0.6927083333333334,
      "grad_norm": 0.8498758673667908,
      "learning_rate": 0.0005,
      "loss": 2.6494030952453613,
      "step": 3591
    },
    {
      "epoch": 0.6929012345679012,
      "grad_norm": 1.0412579774856567,
      "learning_rate": 0.0005,
      "loss": 2.5124688148498535,
      "step": 3592
    },
    {
      "epoch": 0.6930941358024691,
      "grad_norm": 0.760119616985321,
      "learning_rate": 0.0005,
      "loss": 2.8007144927978516,
      "step": 3593
    },
    {
      "epoch": 0.6932870370370371,
      "grad_norm": 0.8157927989959717,
      "learning_rate": 0.0005,
      "loss": 2.4252490997314453,
      "step": 3594
    },
    {
      "epoch": 0.6934799382716049,
      "grad_norm": 1.0980970859527588,
      "learning_rate": 0.0005,
      "loss": 2.257242202758789,
      "step": 3595
    },
    {
      "epoch": 0.6936728395061729,
      "grad_norm": 1.1453238725662231,
      "learning_rate": 0.0005,
      "loss": 2.5516281127929688,
      "step": 3596
    },
    {
      "epoch": 0.6938657407407407,
      "grad_norm": 1.3154633045196533,
      "learning_rate": 0.0005,
      "loss": 2.368772506713867,
      "step": 3597
    },
    {
      "epoch": 0.6940586419753086,
      "grad_norm": 0.8135665059089661,
      "learning_rate": 0.0005,
      "loss": 2.5730552673339844,
      "step": 3598
    },
    {
      "epoch": 0.6942515432098766,
      "grad_norm": 0.9269306659698486,
      "learning_rate": 0.0005,
      "loss": 2.7534496784210205,
      "step": 3599
    },
    {
      "epoch": 0.6944444444444444,
      "grad_norm": 0.7083647847175598,
      "learning_rate": 0.0005,
      "loss": 2.3324904441833496,
      "step": 3600
    },
    {
      "epoch": 0.6946373456790124,
      "grad_norm": 0.9396036267280579,
      "learning_rate": 0.0005,
      "loss": 2.4089527130126953,
      "step": 3601
    },
    {
      "epoch": 0.6948302469135802,
      "grad_norm": 0.9638203978538513,
      "learning_rate": 0.0005,
      "loss": 2.5764167308807373,
      "step": 3602
    },
    {
      "epoch": 0.6950231481481481,
      "grad_norm": 0.8325454592704773,
      "learning_rate": 0.0005,
      "loss": 2.2612645626068115,
      "step": 3603
    },
    {
      "epoch": 0.6952160493827161,
      "grad_norm": 1.2457983493804932,
      "learning_rate": 0.0005,
      "loss": 2.3200206756591797,
      "step": 3604
    },
    {
      "epoch": 0.6954089506172839,
      "grad_norm": 0.9270808696746826,
      "learning_rate": 0.0005,
      "loss": 2.5762217044830322,
      "step": 3605
    },
    {
      "epoch": 0.6956018518518519,
      "grad_norm": 1.1612268686294556,
      "learning_rate": 0.0005,
      "loss": 2.586918354034424,
      "step": 3606
    },
    {
      "epoch": 0.6957947530864198,
      "grad_norm": 1.00413179397583,
      "learning_rate": 0.0005,
      "loss": 2.3845086097717285,
      "step": 3607
    },
    {
      "epoch": 0.6959876543209876,
      "grad_norm": 2.0982773303985596,
      "learning_rate": 0.0005,
      "loss": 2.6632080078125,
      "step": 3608
    },
    {
      "epoch": 0.6961805555555556,
      "grad_norm": 0.8114655613899231,
      "learning_rate": 0.0005,
      "loss": 2.7952394485473633,
      "step": 3609
    },
    {
      "epoch": 0.6963734567901234,
      "grad_norm": 1.2990516424179077,
      "learning_rate": 0.0005,
      "loss": 2.357680559158325,
      "step": 3610
    },
    {
      "epoch": 0.6965663580246914,
      "grad_norm": 0.78324955701828,
      "learning_rate": 0.0005,
      "loss": 2.252596855163574,
      "step": 3611
    },
    {
      "epoch": 0.6967592592592593,
      "grad_norm": 1.0235880613327026,
      "learning_rate": 0.0005,
      "loss": 2.5001871585845947,
      "step": 3612
    },
    {
      "epoch": 0.6969521604938271,
      "grad_norm": 0.9366657137870789,
      "learning_rate": 0.0005,
      "loss": 2.51578950881958,
      "step": 3613
    },
    {
      "epoch": 0.6971450617283951,
      "grad_norm": 0.6636663675308228,
      "learning_rate": 0.0005,
      "loss": 2.3097102642059326,
      "step": 3614
    },
    {
      "epoch": 0.6973379629629629,
      "grad_norm": 2.0223939418792725,
      "learning_rate": 0.0005,
      "loss": 2.3965554237365723,
      "step": 3615
    },
    {
      "epoch": 0.6975308641975309,
      "grad_norm": 0.948090136051178,
      "learning_rate": 0.0005,
      "loss": 2.270932197570801,
      "step": 3616
    },
    {
      "epoch": 0.6977237654320988,
      "grad_norm": 0.8920234441757202,
      "learning_rate": 0.0005,
      "loss": 2.7619872093200684,
      "step": 3617
    },
    {
      "epoch": 0.6979166666666666,
      "grad_norm": 0.8008525371551514,
      "learning_rate": 0.0005,
      "loss": 2.6064000129699707,
      "step": 3618
    },
    {
      "epoch": 0.6981095679012346,
      "grad_norm": 0.9048497080802917,
      "learning_rate": 0.0005,
      "loss": 2.5042648315429688,
      "step": 3619
    },
    {
      "epoch": 0.6983024691358025,
      "grad_norm": 1.2802094221115112,
      "learning_rate": 0.0005,
      "loss": 2.512990713119507,
      "step": 3620
    },
    {
      "epoch": 0.6984953703703703,
      "grad_norm": 0.8771253228187561,
      "learning_rate": 0.0005,
      "loss": 2.3847923278808594,
      "step": 3621
    },
    {
      "epoch": 0.6986882716049383,
      "grad_norm": 0.9085904955863953,
      "learning_rate": 0.0005,
      "loss": 2.42521071434021,
      "step": 3622
    },
    {
      "epoch": 0.6988811728395061,
      "grad_norm": 1.2045238018035889,
      "learning_rate": 0.0005,
      "loss": 2.2846531867980957,
      "step": 3623
    },
    {
      "epoch": 0.6990740740740741,
      "grad_norm": 1.013070821762085,
      "learning_rate": 0.0005,
      "loss": 2.3431968688964844,
      "step": 3624
    },
    {
      "epoch": 0.699266975308642,
      "grad_norm": 0.7315769195556641,
      "learning_rate": 0.0005,
      "loss": 2.5392794609069824,
      "step": 3625
    },
    {
      "epoch": 0.6994598765432098,
      "grad_norm": 0.7953864336013794,
      "learning_rate": 0.0005,
      "loss": 2.4621968269348145,
      "step": 3626
    },
    {
      "epoch": 0.6996527777777778,
      "grad_norm": 1.25956130027771,
      "learning_rate": 0.0005,
      "loss": 2.4963510036468506,
      "step": 3627
    },
    {
      "epoch": 0.6998456790123457,
      "grad_norm": 0.903225302696228,
      "learning_rate": 0.0005,
      "loss": 2.695584774017334,
      "step": 3628
    },
    {
      "epoch": 0.7000385802469136,
      "grad_norm": 0.8854238390922546,
      "learning_rate": 0.0005,
      "loss": 2.459397792816162,
      "step": 3629
    },
    {
      "epoch": 0.7002314814814815,
      "grad_norm": 0.8380270004272461,
      "learning_rate": 0.0005,
      "loss": 2.596426248550415,
      "step": 3630
    },
    {
      "epoch": 0.7004243827160493,
      "grad_norm": 1.0293370485305786,
      "learning_rate": 0.0005,
      "loss": 2.5030808448791504,
      "step": 3631
    },
    {
      "epoch": 0.7006172839506173,
      "grad_norm": 1.1145778894424438,
      "learning_rate": 0.0005,
      "loss": 2.744584083557129,
      "step": 3632
    },
    {
      "epoch": 0.7008101851851852,
      "grad_norm": 0.7226782441139221,
      "learning_rate": 0.0005,
      "loss": 2.4415154457092285,
      "step": 3633
    },
    {
      "epoch": 0.7010030864197531,
      "grad_norm": 1.0492573976516724,
      "learning_rate": 0.0005,
      "loss": 2.6415772438049316,
      "step": 3634
    },
    {
      "epoch": 0.701195987654321,
      "grad_norm": 0.9992265105247498,
      "learning_rate": 0.0005,
      "loss": 2.90157151222229,
      "step": 3635
    },
    {
      "epoch": 0.7013888888888888,
      "grad_norm": 0.8929638266563416,
      "learning_rate": 0.0005,
      "loss": 2.4235143661499023,
      "step": 3636
    },
    {
      "epoch": 0.7015817901234568,
      "grad_norm": 1.2175328731536865,
      "learning_rate": 0.0005,
      "loss": 2.501962661743164,
      "step": 3637
    },
    {
      "epoch": 0.7017746913580247,
      "grad_norm": 0.8168100118637085,
      "learning_rate": 0.0005,
      "loss": 2.6632189750671387,
      "step": 3638
    },
    {
      "epoch": 0.7019675925925926,
      "grad_norm": 1.0015970468521118,
      "learning_rate": 0.0005,
      "loss": 2.493110179901123,
      "step": 3639
    },
    {
      "epoch": 0.7021604938271605,
      "grad_norm": 0.9743383526802063,
      "learning_rate": 0.0005,
      "loss": 2.588474988937378,
      "step": 3640
    },
    {
      "epoch": 0.7023533950617284,
      "grad_norm": 0.9138283729553223,
      "learning_rate": 0.0005,
      "loss": 2.5404019355773926,
      "step": 3641
    },
    {
      "epoch": 0.7025462962962963,
      "grad_norm": 0.8809142708778381,
      "learning_rate": 0.0005,
      "loss": 2.631730794906616,
      "step": 3642
    },
    {
      "epoch": 0.7027391975308642,
      "grad_norm": 0.9259359836578369,
      "learning_rate": 0.0005,
      "loss": 2.367579221725464,
      "step": 3643
    },
    {
      "epoch": 0.7029320987654321,
      "grad_norm": 0.9951856136322021,
      "learning_rate": 0.0005,
      "loss": 2.180230140686035,
      "step": 3644
    },
    {
      "epoch": 0.703125,
      "grad_norm": 0.8067197799682617,
      "learning_rate": 0.0005,
      "loss": 2.726870536804199,
      "step": 3645
    },
    {
      "epoch": 0.7033179012345679,
      "grad_norm": 0.9978997111320496,
      "learning_rate": 0.0005,
      "loss": 2.430757999420166,
      "step": 3646
    },
    {
      "epoch": 0.7035108024691358,
      "grad_norm": 0.8009635210037231,
      "learning_rate": 0.0005,
      "loss": 2.477015495300293,
      "step": 3647
    },
    {
      "epoch": 0.7037037037037037,
      "grad_norm": 1.0341588258743286,
      "learning_rate": 0.0005,
      "loss": 2.5350501537323,
      "step": 3648
    },
    {
      "epoch": 0.7038966049382716,
      "grad_norm": 0.9191915988922119,
      "learning_rate": 0.0005,
      "loss": 2.3976104259490967,
      "step": 3649
    },
    {
      "epoch": 0.7040895061728395,
      "grad_norm": 0.7319096326828003,
      "learning_rate": 0.0005,
      "loss": 2.1981842517852783,
      "step": 3650
    },
    {
      "epoch": 0.7042824074074074,
      "grad_norm": 0.9228930473327637,
      "learning_rate": 0.0005,
      "loss": 2.5991766452789307,
      "step": 3651
    },
    {
      "epoch": 0.7044753086419753,
      "grad_norm": 0.9303187131881714,
      "learning_rate": 0.0005,
      "loss": 2.401261806488037,
      "step": 3652
    },
    {
      "epoch": 0.7046682098765432,
      "grad_norm": 1.198920726776123,
      "learning_rate": 0.0005,
      "loss": 2.4398674964904785,
      "step": 3653
    },
    {
      "epoch": 0.7048611111111112,
      "grad_norm": 1.00640869140625,
      "learning_rate": 0.0005,
      "loss": 2.550339698791504,
      "step": 3654
    },
    {
      "epoch": 0.705054012345679,
      "grad_norm": 0.9630938172340393,
      "learning_rate": 0.0005,
      "loss": 2.511671304702759,
      "step": 3655
    },
    {
      "epoch": 0.7052469135802469,
      "grad_norm": 0.7734540700912476,
      "learning_rate": 0.0005,
      "loss": 2.2927608489990234,
      "step": 3656
    },
    {
      "epoch": 0.7054398148148148,
      "grad_norm": 0.9632043242454529,
      "learning_rate": 0.0005,
      "loss": 2.3115057945251465,
      "step": 3657
    },
    {
      "epoch": 0.7056327160493827,
      "grad_norm": 0.9520361423492432,
      "learning_rate": 0.0005,
      "loss": 2.389176368713379,
      "step": 3658
    },
    {
      "epoch": 0.7058256172839507,
      "grad_norm": 0.882411003112793,
      "learning_rate": 0.0005,
      "loss": 2.2957606315612793,
      "step": 3659
    },
    {
      "epoch": 0.7060185185185185,
      "grad_norm": 0.9554412961006165,
      "learning_rate": 0.0005,
      "loss": 2.465975522994995,
      "step": 3660
    },
    {
      "epoch": 0.7062114197530864,
      "grad_norm": 0.8923646211624146,
      "learning_rate": 0.0005,
      "loss": 2.6811208724975586,
      "step": 3661
    },
    {
      "epoch": 0.7064043209876543,
      "grad_norm": 0.8769224882125854,
      "learning_rate": 0.0005,
      "loss": 2.3211188316345215,
      "step": 3662
    },
    {
      "epoch": 0.7065972222222222,
      "grad_norm": 1.2249765396118164,
      "learning_rate": 0.0005,
      "loss": 2.8267807960510254,
      "step": 3663
    },
    {
      "epoch": 0.7067901234567902,
      "grad_norm": 1.0580374002456665,
      "learning_rate": 0.0005,
      "loss": 2.7977733612060547,
      "step": 3664
    },
    {
      "epoch": 0.706983024691358,
      "grad_norm": 0.8260430097579956,
      "learning_rate": 0.0005,
      "loss": 2.4933128356933594,
      "step": 3665
    },
    {
      "epoch": 0.7071759259259259,
      "grad_norm": 1.0360075235366821,
      "learning_rate": 0.0005,
      "loss": 2.6079649925231934,
      "step": 3666
    },
    {
      "epoch": 0.7073688271604939,
      "grad_norm": 0.8335801959037781,
      "learning_rate": 0.0005,
      "loss": 2.6550252437591553,
      "step": 3667
    },
    {
      "epoch": 0.7075617283950617,
      "grad_norm": 0.8813936710357666,
      "learning_rate": 0.0005,
      "loss": 2.1647510528564453,
      "step": 3668
    },
    {
      "epoch": 0.7077546296296297,
      "grad_norm": 0.7807191014289856,
      "learning_rate": 0.0005,
      "loss": 2.2389187812805176,
      "step": 3669
    },
    {
      "epoch": 0.7079475308641975,
      "grad_norm": 1.9683469533920288,
      "learning_rate": 0.0005,
      "loss": 2.406705617904663,
      "step": 3670
    },
    {
      "epoch": 0.7081404320987654,
      "grad_norm": 0.8365073800086975,
      "learning_rate": 0.0005,
      "loss": 2.458484172821045,
      "step": 3671
    },
    {
      "epoch": 0.7083333333333334,
      "grad_norm": 0.9418646693229675,
      "learning_rate": 0.0005,
      "loss": 2.8979859352111816,
      "step": 3672
    },
    {
      "epoch": 0.7085262345679012,
      "grad_norm": 1.0187780857086182,
      "learning_rate": 0.0005,
      "loss": 2.4615986347198486,
      "step": 3673
    },
    {
      "epoch": 0.7087191358024691,
      "grad_norm": 0.7670586705207825,
      "learning_rate": 0.0005,
      "loss": 2.189408779144287,
      "step": 3674
    },
    {
      "epoch": 0.7089120370370371,
      "grad_norm": 0.8719812631607056,
      "learning_rate": 0.0005,
      "loss": 2.168779134750366,
      "step": 3675
    },
    {
      "epoch": 0.7091049382716049,
      "grad_norm": 0.7597718238830566,
      "learning_rate": 0.0005,
      "loss": 2.158625364303589,
      "step": 3676
    },
    {
      "epoch": 0.7092978395061729,
      "grad_norm": 0.96282559633255,
      "learning_rate": 0.0005,
      "loss": 2.4963936805725098,
      "step": 3677
    },
    {
      "epoch": 0.7094907407407407,
      "grad_norm": 0.819683849811554,
      "learning_rate": 0.0005,
      "loss": 2.339710235595703,
      "step": 3678
    },
    {
      "epoch": 0.7096836419753086,
      "grad_norm": 0.8883581757545471,
      "learning_rate": 0.0005,
      "loss": 2.4263012409210205,
      "step": 3679
    },
    {
      "epoch": 0.7098765432098766,
      "grad_norm": 0.8610187768936157,
      "learning_rate": 0.0005,
      "loss": 2.396038055419922,
      "step": 3680
    },
    {
      "epoch": 0.7100694444444444,
      "grad_norm": 0.8104565739631653,
      "learning_rate": 0.0005,
      "loss": 2.3762593269348145,
      "step": 3681
    },
    {
      "epoch": 0.7102623456790124,
      "grad_norm": 0.8402350544929504,
      "learning_rate": 0.0005,
      "loss": 2.503263473510742,
      "step": 3682
    },
    {
      "epoch": 0.7104552469135802,
      "grad_norm": 0.8129053711891174,
      "learning_rate": 0.0005,
      "loss": 2.4240074157714844,
      "step": 3683
    },
    {
      "epoch": 0.7106481481481481,
      "grad_norm": 0.7662428617477417,
      "learning_rate": 0.0005,
      "loss": 2.597367286682129,
      "step": 3684
    },
    {
      "epoch": 0.7108410493827161,
      "grad_norm": 0.9198024272918701,
      "learning_rate": 0.0005,
      "loss": 2.675570249557495,
      "step": 3685
    },
    {
      "epoch": 0.7110339506172839,
      "grad_norm": 0.8040941953659058,
      "learning_rate": 0.0005,
      "loss": 2.497117042541504,
      "step": 3686
    },
    {
      "epoch": 0.7112268518518519,
      "grad_norm": 0.8118135333061218,
      "learning_rate": 0.0005,
      "loss": 2.6751699447631836,
      "step": 3687
    },
    {
      "epoch": 0.7114197530864198,
      "grad_norm": 0.9074379801750183,
      "learning_rate": 0.0005,
      "loss": 2.5209779739379883,
      "step": 3688
    },
    {
      "epoch": 0.7116126543209876,
      "grad_norm": 0.8436322808265686,
      "learning_rate": 0.0005,
      "loss": 2.2632508277893066,
      "step": 3689
    },
    {
      "epoch": 0.7118055555555556,
      "grad_norm": 0.874382495880127,
      "learning_rate": 0.0005,
      "loss": 2.4780406951904297,
      "step": 3690
    },
    {
      "epoch": 0.7119984567901234,
      "grad_norm": 0.9695208072662354,
      "learning_rate": 0.0005,
      "loss": 2.759463310241699,
      "step": 3691
    },
    {
      "epoch": 0.7121913580246914,
      "grad_norm": 0.869834303855896,
      "learning_rate": 0.0005,
      "loss": 2.209853410720825,
      "step": 3692
    },
    {
      "epoch": 0.7123842592592593,
      "grad_norm": 0.9138018488883972,
      "learning_rate": 0.0005,
      "loss": 2.6533288955688477,
      "step": 3693
    },
    {
      "epoch": 0.7125771604938271,
      "grad_norm": 0.8024915456771851,
      "learning_rate": 0.0005,
      "loss": 2.3696980476379395,
      "step": 3694
    },
    {
      "epoch": 0.7127700617283951,
      "grad_norm": 0.9456145763397217,
      "learning_rate": 0.0005,
      "loss": 2.331770420074463,
      "step": 3695
    },
    {
      "epoch": 0.7129629629629629,
      "grad_norm": 0.9789454936981201,
      "learning_rate": 0.0005,
      "loss": 2.4384045600891113,
      "step": 3696
    },
    {
      "epoch": 0.7131558641975309,
      "grad_norm": 1.0102145671844482,
      "learning_rate": 0.0005,
      "loss": 2.6180641651153564,
      "step": 3697
    },
    {
      "epoch": 0.7133487654320988,
      "grad_norm": 0.8543821573257446,
      "learning_rate": 0.0005,
      "loss": 2.3020498752593994,
      "step": 3698
    },
    {
      "epoch": 0.7135416666666666,
      "grad_norm": 0.81544029712677,
      "learning_rate": 0.0005,
      "loss": 2.553035020828247,
      "step": 3699
    },
    {
      "epoch": 0.7137345679012346,
      "grad_norm": 0.751787543296814,
      "learning_rate": 0.0005,
      "loss": 2.1827492713928223,
      "step": 3700
    },
    {
      "epoch": 0.7139274691358025,
      "grad_norm": 1.0546284914016724,
      "learning_rate": 0.0005,
      "loss": 2.416123867034912,
      "step": 3701
    },
    {
      "epoch": 0.7141203703703703,
      "grad_norm": 0.9592426419258118,
      "learning_rate": 0.0005,
      "loss": 2.3324761390686035,
      "step": 3702
    },
    {
      "epoch": 0.7143132716049383,
      "grad_norm": 0.8074532747268677,
      "learning_rate": 0.0005,
      "loss": 2.4752368927001953,
      "step": 3703
    },
    {
      "epoch": 0.7145061728395061,
      "grad_norm": 0.8467631340026855,
      "learning_rate": 0.0005,
      "loss": 2.505110263824463,
      "step": 3704
    },
    {
      "epoch": 0.7146990740740741,
      "grad_norm": 0.8680164217948914,
      "learning_rate": 0.0005,
      "loss": 2.39844012260437,
      "step": 3705
    },
    {
      "epoch": 0.714891975308642,
      "grad_norm": 0.8089905381202698,
      "learning_rate": 0.0005,
      "loss": 2.545048952102661,
      "step": 3706
    },
    {
      "epoch": 0.7150848765432098,
      "grad_norm": 1.0086115598678589,
      "learning_rate": 0.0005,
      "loss": 2.3735647201538086,
      "step": 3707
    },
    {
      "epoch": 0.7152777777777778,
      "grad_norm": 0.9836585521697998,
      "learning_rate": 0.0005,
      "loss": 2.3662304878234863,
      "step": 3708
    },
    {
      "epoch": 0.7154706790123457,
      "grad_norm": 0.8409773707389832,
      "learning_rate": 0.0005,
      "loss": 2.2918038368225098,
      "step": 3709
    },
    {
      "epoch": 0.7156635802469136,
      "grad_norm": 1.3055757284164429,
      "learning_rate": 0.0005,
      "loss": 2.4993581771850586,
      "step": 3710
    },
    {
      "epoch": 0.7158564814814815,
      "grad_norm": 0.8927151560783386,
      "learning_rate": 0.0005,
      "loss": 2.4024622440338135,
      "step": 3711
    },
    {
      "epoch": 0.7160493827160493,
      "grad_norm": 0.7709214687347412,
      "learning_rate": 0.0005,
      "loss": 2.3439626693725586,
      "step": 3712
    },
    {
      "epoch": 0.7162422839506173,
      "grad_norm": 1.101965308189392,
      "learning_rate": 0.0005,
      "loss": 2.6021780967712402,
      "step": 3713
    },
    {
      "epoch": 0.7164351851851852,
      "grad_norm": 0.7774822115898132,
      "learning_rate": 0.0005,
      "loss": 2.542609691619873,
      "step": 3714
    },
    {
      "epoch": 0.7166280864197531,
      "grad_norm": 0.8847402930259705,
      "learning_rate": 0.0005,
      "loss": 2.474562168121338,
      "step": 3715
    },
    {
      "epoch": 0.716820987654321,
      "grad_norm": 1.4059540033340454,
      "learning_rate": 0.0005,
      "loss": 2.288973808288574,
      "step": 3716
    },
    {
      "epoch": 0.7170138888888888,
      "grad_norm": 0.8545230031013489,
      "learning_rate": 0.0005,
      "loss": 2.6021780967712402,
      "step": 3717
    },
    {
      "epoch": 0.7172067901234568,
      "grad_norm": 0.8390601277351379,
      "learning_rate": 0.0005,
      "loss": 2.2903130054473877,
      "step": 3718
    },
    {
      "epoch": 0.7173996913580247,
      "grad_norm": 0.8601127862930298,
      "learning_rate": 0.0005,
      "loss": 2.2357864379882812,
      "step": 3719
    },
    {
      "epoch": 0.7175925925925926,
      "grad_norm": 0.9764934778213501,
      "learning_rate": 0.0005,
      "loss": 2.1892213821411133,
      "step": 3720
    },
    {
      "epoch": 0.7177854938271605,
      "grad_norm": 0.8411382436752319,
      "learning_rate": 0.0005,
      "loss": 2.503998279571533,
      "step": 3721
    },
    {
      "epoch": 0.7179783950617284,
      "grad_norm": 0.9928483963012695,
      "learning_rate": 0.0005,
      "loss": 2.3640480041503906,
      "step": 3722
    },
    {
      "epoch": 0.7181712962962963,
      "grad_norm": 0.8669095039367676,
      "learning_rate": 0.0005,
      "loss": 2.2539749145507812,
      "step": 3723
    },
    {
      "epoch": 0.7183641975308642,
      "grad_norm": 2.3581430912017822,
      "learning_rate": 0.0005,
      "loss": 2.4462878704071045,
      "step": 3724
    },
    {
      "epoch": 0.7185570987654321,
      "grad_norm": 0.9202872514724731,
      "learning_rate": 0.0005,
      "loss": 2.560176372528076,
      "step": 3725
    },
    {
      "epoch": 0.71875,
      "grad_norm": 1.3514715433120728,
      "learning_rate": 0.0005,
      "loss": 2.3241281509399414,
      "step": 3726
    },
    {
      "epoch": 0.7189429012345679,
      "grad_norm": 0.9750831127166748,
      "learning_rate": 0.0005,
      "loss": 2.383791923522949,
      "step": 3727
    },
    {
      "epoch": 0.7191358024691358,
      "grad_norm": 0.8095408082008362,
      "learning_rate": 0.0005,
      "loss": 2.5324153900146484,
      "step": 3728
    },
    {
      "epoch": 0.7193287037037037,
      "grad_norm": 1.3729310035705566,
      "learning_rate": 0.0005,
      "loss": 2.300650119781494,
      "step": 3729
    },
    {
      "epoch": 0.7195216049382716,
      "grad_norm": 1.566717505455017,
      "learning_rate": 0.0005,
      "loss": 2.4687819480895996,
      "step": 3730
    },
    {
      "epoch": 0.7197145061728395,
      "grad_norm": 0.7115570306777954,
      "learning_rate": 0.0005,
      "loss": 2.5851616859436035,
      "step": 3731
    },
    {
      "epoch": 0.7199074074074074,
      "grad_norm": 0.8791199326515198,
      "learning_rate": 0.0005,
      "loss": 2.2463338375091553,
      "step": 3732
    },
    {
      "epoch": 0.7201003086419753,
      "grad_norm": 1.203486442565918,
      "learning_rate": 0.0005,
      "loss": 2.31770658493042,
      "step": 3733
    },
    {
      "epoch": 0.7202932098765432,
      "grad_norm": 0.762505054473877,
      "learning_rate": 0.0005,
      "loss": 2.2118029594421387,
      "step": 3734
    },
    {
      "epoch": 0.7204861111111112,
      "grad_norm": 1.2612172365188599,
      "learning_rate": 0.0005,
      "loss": 2.3757095336914062,
      "step": 3735
    },
    {
      "epoch": 0.720679012345679,
      "grad_norm": 1.0834749937057495,
      "learning_rate": 0.0005,
      "loss": 2.36154842376709,
      "step": 3736
    },
    {
      "epoch": 0.7208719135802469,
      "grad_norm": 0.7842273712158203,
      "learning_rate": 0.0005,
      "loss": 2.3565022945404053,
      "step": 3737
    },
    {
      "epoch": 0.7210648148148148,
      "grad_norm": 1.1293256282806396,
      "learning_rate": 0.0005,
      "loss": 2.3152363300323486,
      "step": 3738
    },
    {
      "epoch": 0.7212577160493827,
      "grad_norm": 0.9074370861053467,
      "learning_rate": 0.0005,
      "loss": 2.340117931365967,
      "step": 3739
    },
    {
      "epoch": 0.7214506172839507,
      "grad_norm": 0.9632592797279358,
      "learning_rate": 0.0005,
      "loss": 3.2594711780548096,
      "step": 3740
    },
    {
      "epoch": 0.7216435185185185,
      "grad_norm": 1.1526641845703125,
      "learning_rate": 0.0005,
      "loss": 2.387387752532959,
      "step": 3741
    },
    {
      "epoch": 0.7218364197530864,
      "grad_norm": 0.9002780914306641,
      "learning_rate": 0.0005,
      "loss": 2.115161895751953,
      "step": 3742
    },
    {
      "epoch": 0.7220293209876543,
      "grad_norm": 0.845582902431488,
      "learning_rate": 0.0005,
      "loss": 2.4694433212280273,
      "step": 3743
    },
    {
      "epoch": 0.7222222222222222,
      "grad_norm": 0.7078306078910828,
      "learning_rate": 0.0005,
      "loss": 2.46370530128479,
      "step": 3744
    },
    {
      "epoch": 0.7224151234567902,
      "grad_norm": 0.8888205885887146,
      "learning_rate": 0.0005,
      "loss": 2.487290382385254,
      "step": 3745
    },
    {
      "epoch": 0.722608024691358,
      "grad_norm": 0.8146944642066956,
      "learning_rate": 0.0005,
      "loss": 2.3650572299957275,
      "step": 3746
    },
    {
      "epoch": 0.7228009259259259,
      "grad_norm": 0.7703208923339844,
      "learning_rate": 0.0005,
      "loss": 2.3960208892822266,
      "step": 3747
    },
    {
      "epoch": 0.7229938271604939,
      "grad_norm": 0.8403128385543823,
      "learning_rate": 0.0005,
      "loss": 2.4142653942108154,
      "step": 3748
    },
    {
      "epoch": 0.7231867283950617,
      "grad_norm": 0.9214215278625488,
      "learning_rate": 0.0005,
      "loss": 2.4556429386138916,
      "step": 3749
    },
    {
      "epoch": 0.7233796296296297,
      "grad_norm": 1.682452917098999,
      "learning_rate": 0.0005,
      "loss": 2.4898195266723633,
      "step": 3750
    },
    {
      "epoch": 0.7235725308641975,
      "grad_norm": 1.0163367986679077,
      "learning_rate": 0.0005,
      "loss": 2.433654546737671,
      "step": 3751
    },
    {
      "epoch": 0.7237654320987654,
      "grad_norm": 0.9261779189109802,
      "learning_rate": 0.0005,
      "loss": 2.4977903366088867,
      "step": 3752
    },
    {
      "epoch": 0.7239583333333334,
      "grad_norm": 0.8886110782623291,
      "learning_rate": 0.0005,
      "loss": 2.1663758754730225,
      "step": 3753
    },
    {
      "epoch": 0.7241512345679012,
      "grad_norm": 1.5507409572601318,
      "learning_rate": 0.0005,
      "loss": 2.494863748550415,
      "step": 3754
    },
    {
      "epoch": 0.7243441358024691,
      "grad_norm": 1.3267995119094849,
      "learning_rate": 0.0005,
      "loss": 2.332134485244751,
      "step": 3755
    },
    {
      "epoch": 0.7245370370370371,
      "grad_norm": 0.9683233499526978,
      "learning_rate": 0.0005,
      "loss": 2.377671480178833,
      "step": 3756
    },
    {
      "epoch": 0.7247299382716049,
      "grad_norm": 0.8665006160736084,
      "learning_rate": 0.0005,
      "loss": 2.176572561264038,
      "step": 3757
    },
    {
      "epoch": 0.7249228395061729,
      "grad_norm": 0.9263468384742737,
      "learning_rate": 0.0005,
      "loss": 2.4232852458953857,
      "step": 3758
    },
    {
      "epoch": 0.7251157407407407,
      "grad_norm": 0.8389759659767151,
      "learning_rate": 0.0005,
      "loss": 2.332892417907715,
      "step": 3759
    },
    {
      "epoch": 0.7253086419753086,
      "grad_norm": 1.0399419069290161,
      "learning_rate": 0.0005,
      "loss": 2.5167934894561768,
      "step": 3760
    },
    {
      "epoch": 0.7255015432098766,
      "grad_norm": 1.0230921506881714,
      "learning_rate": 0.0005,
      "loss": 2.496396541595459,
      "step": 3761
    },
    {
      "epoch": 0.7256944444444444,
      "grad_norm": 0.8925200700759888,
      "learning_rate": 0.0005,
      "loss": 2.43924880027771,
      "step": 3762
    },
    {
      "epoch": 0.7258873456790124,
      "grad_norm": 0.8512965440750122,
      "learning_rate": 0.0005,
      "loss": 2.4457526206970215,
      "step": 3763
    },
    {
      "epoch": 0.7260802469135802,
      "grad_norm": 0.7536328434944153,
      "learning_rate": 0.0005,
      "loss": 2.450026035308838,
      "step": 3764
    },
    {
      "epoch": 0.7262731481481481,
      "grad_norm": 0.9155247211456299,
      "learning_rate": 0.0005,
      "loss": 2.2946646213531494,
      "step": 3765
    },
    {
      "epoch": 0.7264660493827161,
      "grad_norm": 0.8756814002990723,
      "learning_rate": 0.0005,
      "loss": 2.425739288330078,
      "step": 3766
    },
    {
      "epoch": 0.7266589506172839,
      "grad_norm": 0.8950318098068237,
      "learning_rate": 0.0005,
      "loss": 2.5365917682647705,
      "step": 3767
    },
    {
      "epoch": 0.7268518518518519,
      "grad_norm": 0.8694406151771545,
      "learning_rate": 0.0005,
      "loss": 2.492999315261841,
      "step": 3768
    },
    {
      "epoch": 0.7270447530864198,
      "grad_norm": 0.8943628072738647,
      "learning_rate": 0.0005,
      "loss": 2.386258125305176,
      "step": 3769
    },
    {
      "epoch": 0.7272376543209876,
      "grad_norm": 0.808721661567688,
      "learning_rate": 0.0005,
      "loss": 2.7528529167175293,
      "step": 3770
    },
    {
      "epoch": 0.7274305555555556,
      "grad_norm": 0.7465051412582397,
      "learning_rate": 0.0005,
      "loss": 2.5610880851745605,
      "step": 3771
    },
    {
      "epoch": 0.7276234567901234,
      "grad_norm": 0.8985435962677002,
      "learning_rate": 0.0005,
      "loss": 2.1549081802368164,
      "step": 3772
    },
    {
      "epoch": 0.7278163580246914,
      "grad_norm": 1.0800234079360962,
      "learning_rate": 0.0005,
      "loss": 2.4083852767944336,
      "step": 3773
    },
    {
      "epoch": 0.7280092592592593,
      "grad_norm": 0.9181095361709595,
      "learning_rate": 0.0005,
      "loss": 2.3832011222839355,
      "step": 3774
    },
    {
      "epoch": 0.7282021604938271,
      "grad_norm": 0.8812609314918518,
      "learning_rate": 0.0005,
      "loss": 2.404906988143921,
      "step": 3775
    },
    {
      "epoch": 0.7283950617283951,
      "grad_norm": 0.8192505836486816,
      "learning_rate": 0.0005,
      "loss": 2.620044708251953,
      "step": 3776
    },
    {
      "epoch": 0.7285879629629629,
      "grad_norm": 0.7946974635124207,
      "learning_rate": 0.0005,
      "loss": 2.4068450927734375,
      "step": 3777
    },
    {
      "epoch": 0.7287808641975309,
      "grad_norm": 0.8566033244132996,
      "learning_rate": 0.0005,
      "loss": 2.1735150814056396,
      "step": 3778
    },
    {
      "epoch": 0.7289737654320988,
      "grad_norm": 0.8123654723167419,
      "learning_rate": 0.0005,
      "loss": 2.148449182510376,
      "step": 3779
    },
    {
      "epoch": 0.7291666666666666,
      "grad_norm": 0.8992065191268921,
      "learning_rate": 0.0005,
      "loss": 2.4054272174835205,
      "step": 3780
    },
    {
      "epoch": 0.7293595679012346,
      "grad_norm": 0.7877509593963623,
      "learning_rate": 0.0005,
      "loss": 2.4301366806030273,
      "step": 3781
    },
    {
      "epoch": 0.7295524691358025,
      "grad_norm": 0.7517518997192383,
      "learning_rate": 0.0005,
      "loss": 2.4798312187194824,
      "step": 3782
    },
    {
      "epoch": 0.7297453703703703,
      "grad_norm": 0.9051979780197144,
      "learning_rate": 0.0005,
      "loss": 2.6239068508148193,
      "step": 3783
    },
    {
      "epoch": 0.7299382716049383,
      "grad_norm": 1.3348783254623413,
      "learning_rate": 0.0005,
      "loss": 2.4168410301208496,
      "step": 3784
    },
    {
      "epoch": 0.7301311728395061,
      "grad_norm": 0.7421735525131226,
      "learning_rate": 0.0005,
      "loss": 2.309828281402588,
      "step": 3785
    },
    {
      "epoch": 0.7303240740740741,
      "grad_norm": 1.3545078039169312,
      "learning_rate": 0.0005,
      "loss": 2.4785540103912354,
      "step": 3786
    },
    {
      "epoch": 0.730516975308642,
      "grad_norm": 0.8045061826705933,
      "learning_rate": 0.0005,
      "loss": 2.4342494010925293,
      "step": 3787
    },
    {
      "epoch": 0.7307098765432098,
      "grad_norm": 1.0680818557739258,
      "learning_rate": 0.0005,
      "loss": 2.583744525909424,
      "step": 3788
    },
    {
      "epoch": 0.7309027777777778,
      "grad_norm": 1.0295944213867188,
      "learning_rate": 0.0005,
      "loss": 2.5924596786499023,
      "step": 3789
    },
    {
      "epoch": 0.7310956790123457,
      "grad_norm": 0.8191244602203369,
      "learning_rate": 0.0005,
      "loss": 2.227973461151123,
      "step": 3790
    },
    {
      "epoch": 0.7312885802469136,
      "grad_norm": 0.7697238326072693,
      "learning_rate": 0.0005,
      "loss": 2.3335318565368652,
      "step": 3791
    },
    {
      "epoch": 0.7314814814814815,
      "grad_norm": 0.7520463466644287,
      "learning_rate": 0.0005,
      "loss": 2.763608694076538,
      "step": 3792
    },
    {
      "epoch": 0.7316743827160493,
      "grad_norm": 0.8911800384521484,
      "learning_rate": 0.0005,
      "loss": 2.242096185684204,
      "step": 3793
    },
    {
      "epoch": 0.7318672839506173,
      "grad_norm": 1.0654127597808838,
      "learning_rate": 0.0005,
      "loss": 2.2311973571777344,
      "step": 3794
    },
    {
      "epoch": 0.7320601851851852,
      "grad_norm": 0.7662272453308105,
      "learning_rate": 0.0005,
      "loss": 2.1829161643981934,
      "step": 3795
    },
    {
      "epoch": 0.7322530864197531,
      "grad_norm": 1.0504969358444214,
      "learning_rate": 0.0005,
      "loss": 2.4232964515686035,
      "step": 3796
    },
    {
      "epoch": 0.732445987654321,
      "grad_norm": 0.7507092952728271,
      "learning_rate": 0.0005,
      "loss": 2.267421245574951,
      "step": 3797
    },
    {
      "epoch": 0.7326388888888888,
      "grad_norm": 1.2207276821136475,
      "learning_rate": 0.0005,
      "loss": 2.6125240325927734,
      "step": 3798
    },
    {
      "epoch": 0.7328317901234568,
      "grad_norm": 0.9888913035392761,
      "learning_rate": 0.0005,
      "loss": 2.6881048679351807,
      "step": 3799
    },
    {
      "epoch": 0.7330246913580247,
      "grad_norm": 0.7948658466339111,
      "learning_rate": 0.0005,
      "loss": 2.2755680084228516,
      "step": 3800
    },
    {
      "epoch": 0.7332175925925926,
      "grad_norm": 1.1799232959747314,
      "learning_rate": 0.0005,
      "loss": 2.141706705093384,
      "step": 3801
    },
    {
      "epoch": 0.7334104938271605,
      "grad_norm": 1.2869489192962646,
      "learning_rate": 0.0005,
      "loss": 2.59027099609375,
      "step": 3802
    },
    {
      "epoch": 0.7336033950617284,
      "grad_norm": 0.9321783781051636,
      "learning_rate": 0.0005,
      "loss": 2.598734140396118,
      "step": 3803
    },
    {
      "epoch": 0.7337962962962963,
      "grad_norm": 0.9714463949203491,
      "learning_rate": 0.0005,
      "loss": 2.235466480255127,
      "step": 3804
    },
    {
      "epoch": 0.7339891975308642,
      "grad_norm": 0.9992873072624207,
      "learning_rate": 0.0005,
      "loss": 2.5331945419311523,
      "step": 3805
    },
    {
      "epoch": 0.7341820987654321,
      "grad_norm": 1.0240544080734253,
      "learning_rate": 0.0005,
      "loss": 2.7208237648010254,
      "step": 3806
    },
    {
      "epoch": 0.734375,
      "grad_norm": 0.7935792207717896,
      "learning_rate": 0.0005,
      "loss": 2.5975823402404785,
      "step": 3807
    },
    {
      "epoch": 0.7345679012345679,
      "grad_norm": 0.9402756094932556,
      "learning_rate": 0.0005,
      "loss": 2.5295281410217285,
      "step": 3808
    },
    {
      "epoch": 0.7347608024691358,
      "grad_norm": 1.0095125436782837,
      "learning_rate": 0.0005,
      "loss": 2.6218981742858887,
      "step": 3809
    },
    {
      "epoch": 0.7349537037037037,
      "grad_norm": 0.9318719506263733,
      "learning_rate": 0.0005,
      "loss": 2.4875779151916504,
      "step": 3810
    },
    {
      "epoch": 0.7351466049382716,
      "grad_norm": 1.0565041303634644,
      "learning_rate": 0.0005,
      "loss": 2.2263617515563965,
      "step": 3811
    },
    {
      "epoch": 0.7353395061728395,
      "grad_norm": 1.4104288816452026,
      "learning_rate": 0.0005,
      "loss": 2.4702181816101074,
      "step": 3812
    },
    {
      "epoch": 0.7355324074074074,
      "grad_norm": 1.096325397491455,
      "learning_rate": 0.0005,
      "loss": 2.2579236030578613,
      "step": 3813
    },
    {
      "epoch": 0.7357253086419753,
      "grad_norm": 0.9410190582275391,
      "learning_rate": 0.0005,
      "loss": 2.1572489738464355,
      "step": 3814
    },
    {
      "epoch": 0.7359182098765432,
      "grad_norm": 0.9377333521842957,
      "learning_rate": 0.0005,
      "loss": 2.3084375858306885,
      "step": 3815
    },
    {
      "epoch": 0.7361111111111112,
      "grad_norm": 0.8905650973320007,
      "learning_rate": 0.0005,
      "loss": 2.195399761199951,
      "step": 3816
    },
    {
      "epoch": 0.736304012345679,
      "grad_norm": 0.7977266907691956,
      "learning_rate": 0.0005,
      "loss": 2.4857325553894043,
      "step": 3817
    },
    {
      "epoch": 0.7364969135802469,
      "grad_norm": 0.9316128492355347,
      "learning_rate": 0.0005,
      "loss": 2.3161425590515137,
      "step": 3818
    },
    {
      "epoch": 0.7366898148148148,
      "grad_norm": 0.9458324909210205,
      "learning_rate": 0.0005,
      "loss": 2.6520814895629883,
      "step": 3819
    },
    {
      "epoch": 0.7368827160493827,
      "grad_norm": 0.9873355031013489,
      "learning_rate": 0.0005,
      "loss": 2.4285831451416016,
      "step": 3820
    },
    {
      "epoch": 0.7370756172839507,
      "grad_norm": 0.979852020740509,
      "learning_rate": 0.0005,
      "loss": 2.370175361633301,
      "step": 3821
    },
    {
      "epoch": 0.7372685185185185,
      "grad_norm": 0.9662837982177734,
      "learning_rate": 0.0005,
      "loss": 2.4519338607788086,
      "step": 3822
    },
    {
      "epoch": 0.7374614197530864,
      "grad_norm": 0.9260045886039734,
      "learning_rate": 0.0005,
      "loss": 2.333951950073242,
      "step": 3823
    },
    {
      "epoch": 0.7376543209876543,
      "grad_norm": 1.0165024995803833,
      "learning_rate": 0.0005,
      "loss": 2.555561065673828,
      "step": 3824
    },
    {
      "epoch": 0.7378472222222222,
      "grad_norm": 1.5074971914291382,
      "learning_rate": 0.0005,
      "loss": 2.4410300254821777,
      "step": 3825
    },
    {
      "epoch": 0.7380401234567902,
      "grad_norm": 0.992114782333374,
      "learning_rate": 0.0005,
      "loss": 2.3615846633911133,
      "step": 3826
    },
    {
      "epoch": 0.738233024691358,
      "grad_norm": 1.319006323814392,
      "learning_rate": 0.0005,
      "loss": 2.297532081604004,
      "step": 3827
    },
    {
      "epoch": 0.7384259259259259,
      "grad_norm": 1.186510443687439,
      "learning_rate": 0.0005,
      "loss": 2.5241775512695312,
      "step": 3828
    },
    {
      "epoch": 0.7386188271604939,
      "grad_norm": 0.9417417049407959,
      "learning_rate": 0.0005,
      "loss": 2.2471842765808105,
      "step": 3829
    },
    {
      "epoch": 0.7388117283950617,
      "grad_norm": 1.2551957368850708,
      "learning_rate": 0.0005,
      "loss": 2.376631736755371,
      "step": 3830
    },
    {
      "epoch": 0.7390046296296297,
      "grad_norm": 1.0244500637054443,
      "learning_rate": 0.0005,
      "loss": 2.310708999633789,
      "step": 3831
    },
    {
      "epoch": 0.7391975308641975,
      "grad_norm": 0.7110412120819092,
      "learning_rate": 0.0005,
      "loss": 2.408975601196289,
      "step": 3832
    },
    {
      "epoch": 0.7393904320987654,
      "grad_norm": 0.7818337082862854,
      "learning_rate": 0.0005,
      "loss": 2.5949058532714844,
      "step": 3833
    },
    {
      "epoch": 0.7395833333333334,
      "grad_norm": 0.7615363001823425,
      "learning_rate": 0.0005,
      "loss": 2.304839611053467,
      "step": 3834
    },
    {
      "epoch": 0.7397762345679012,
      "grad_norm": 0.7794803977012634,
      "learning_rate": 0.0005,
      "loss": 2.248332977294922,
      "step": 3835
    },
    {
      "epoch": 0.7399691358024691,
      "grad_norm": 0.8119602203369141,
      "learning_rate": 0.0005,
      "loss": 2.3014395236968994,
      "step": 3836
    },
    {
      "epoch": 0.7401620370370371,
      "grad_norm": 0.8127692341804504,
      "learning_rate": 0.0005,
      "loss": 2.442157745361328,
      "step": 3837
    },
    {
      "epoch": 0.7403549382716049,
      "grad_norm": 0.9333801865577698,
      "learning_rate": 0.0005,
      "loss": 2.194847345352173,
      "step": 3838
    },
    {
      "epoch": 0.7405478395061729,
      "grad_norm": 0.7465730309486389,
      "learning_rate": 0.0005,
      "loss": 2.593834400177002,
      "step": 3839
    },
    {
      "epoch": 0.7407407407407407,
      "grad_norm": 0.7087646722793579,
      "learning_rate": 0.0005,
      "loss": 2.22231388092041,
      "step": 3840
    },
    {
      "epoch": 0.7409336419753086,
      "grad_norm": 0.894517719745636,
      "learning_rate": 0.0005,
      "loss": 2.137479782104492,
      "step": 3841
    },
    {
      "epoch": 0.7411265432098766,
      "grad_norm": 0.7025262117385864,
      "learning_rate": 0.0005,
      "loss": 2.5922741889953613,
      "step": 3842
    },
    {
      "epoch": 0.7413194444444444,
      "grad_norm": 0.9290765523910522,
      "learning_rate": 0.0005,
      "loss": 2.0133883953094482,
      "step": 3843
    },
    {
      "epoch": 0.7415123456790124,
      "grad_norm": 0.7361077666282654,
      "learning_rate": 0.0005,
      "loss": 2.6089863777160645,
      "step": 3844
    },
    {
      "epoch": 0.7417052469135802,
      "grad_norm": 0.840923011302948,
      "learning_rate": 0.0005,
      "loss": 2.299515962600708,
      "step": 3845
    },
    {
      "epoch": 0.7418981481481481,
      "grad_norm": 1.0207494497299194,
      "learning_rate": 0.0005,
      "loss": 2.314441680908203,
      "step": 3846
    },
    {
      "epoch": 0.7420910493827161,
      "grad_norm": 1.0169216394424438,
      "learning_rate": 0.0005,
      "loss": 2.748303174972534,
      "step": 3847
    },
    {
      "epoch": 0.7422839506172839,
      "grad_norm": 0.769084095954895,
      "learning_rate": 0.0005,
      "loss": 2.224600315093994,
      "step": 3848
    },
    {
      "epoch": 0.7424768518518519,
      "grad_norm": 0.8873063325881958,
      "learning_rate": 0.0005,
      "loss": 2.483884811401367,
      "step": 3849
    },
    {
      "epoch": 0.7426697530864198,
      "grad_norm": 1.178140640258789,
      "learning_rate": 0.0005,
      "loss": 2.62064266204834,
      "step": 3850
    },
    {
      "epoch": 0.7428626543209876,
      "grad_norm": 0.8721909523010254,
      "learning_rate": 0.0005,
      "loss": 2.390906572341919,
      "step": 3851
    },
    {
      "epoch": 0.7430555555555556,
      "grad_norm": 0.9467133283615112,
      "learning_rate": 0.0005,
      "loss": 2.5831503868103027,
      "step": 3852
    },
    {
      "epoch": 0.7432484567901234,
      "grad_norm": 1.0161155462265015,
      "learning_rate": 0.0005,
      "loss": 2.6927077770233154,
      "step": 3853
    },
    {
      "epoch": 0.7434413580246914,
      "grad_norm": 0.8742370009422302,
      "learning_rate": 0.0005,
      "loss": 2.4989967346191406,
      "step": 3854
    },
    {
      "epoch": 0.7436342592592593,
      "grad_norm": 1.0063819885253906,
      "learning_rate": 0.0005,
      "loss": 2.2897889614105225,
      "step": 3855
    },
    {
      "epoch": 0.7438271604938271,
      "grad_norm": 1.1164007186889648,
      "learning_rate": 0.0005,
      "loss": 2.128028631210327,
      "step": 3856
    },
    {
      "epoch": 0.7440200617283951,
      "grad_norm": 0.8305222392082214,
      "learning_rate": 0.0005,
      "loss": 2.2435598373413086,
      "step": 3857
    },
    {
      "epoch": 0.7442129629629629,
      "grad_norm": 0.8364944458007812,
      "learning_rate": 0.0005,
      "loss": 2.5084216594696045,
      "step": 3858
    },
    {
      "epoch": 0.7444058641975309,
      "grad_norm": 1.0792014598846436,
      "learning_rate": 0.0005,
      "loss": 2.39532470703125,
      "step": 3859
    },
    {
      "epoch": 0.7445987654320988,
      "grad_norm": 0.8401211500167847,
      "learning_rate": 0.0005,
      "loss": 2.191394090652466,
      "step": 3860
    },
    {
      "epoch": 0.7447916666666666,
      "grad_norm": 0.9838646650314331,
      "learning_rate": 0.0005,
      "loss": 2.5161657333374023,
      "step": 3861
    },
    {
      "epoch": 0.7449845679012346,
      "grad_norm": 0.8375278115272522,
      "learning_rate": 0.0005,
      "loss": 2.2290902137756348,
      "step": 3862
    },
    {
      "epoch": 0.7451774691358025,
      "grad_norm": 0.78452068567276,
      "learning_rate": 0.0005,
      "loss": 2.3522684574127197,
      "step": 3863
    },
    {
      "epoch": 0.7453703703703703,
      "grad_norm": 0.6754378080368042,
      "learning_rate": 0.0005,
      "loss": 2.341447114944458,
      "step": 3864
    },
    {
      "epoch": 0.7455632716049383,
      "grad_norm": 1.2921468019485474,
      "learning_rate": 0.0005,
      "loss": 2.5862083435058594,
      "step": 3865
    },
    {
      "epoch": 0.7457561728395061,
      "grad_norm": 0.9063507914543152,
      "learning_rate": 0.0005,
      "loss": 2.3717494010925293,
      "step": 3866
    },
    {
      "epoch": 0.7459490740740741,
      "grad_norm": 1.1989684104919434,
      "learning_rate": 0.0005,
      "loss": 2.5028076171875,
      "step": 3867
    },
    {
      "epoch": 0.746141975308642,
      "grad_norm": 0.9136732816696167,
      "learning_rate": 0.0005,
      "loss": 2.4382128715515137,
      "step": 3868
    },
    {
      "epoch": 0.7463348765432098,
      "grad_norm": 0.7209464907646179,
      "learning_rate": 0.0005,
      "loss": 2.2721447944641113,
      "step": 3869
    },
    {
      "epoch": 0.7465277777777778,
      "grad_norm": 0.9960150122642517,
      "learning_rate": 0.0005,
      "loss": 2.659895420074463,
      "step": 3870
    },
    {
      "epoch": 0.7467206790123457,
      "grad_norm": 1.4298051595687866,
      "learning_rate": 0.0005,
      "loss": 2.581979751586914,
      "step": 3871
    },
    {
      "epoch": 0.7469135802469136,
      "grad_norm": 0.9766315221786499,
      "learning_rate": 0.0005,
      "loss": 2.380716323852539,
      "step": 3872
    },
    {
      "epoch": 0.7471064814814815,
      "grad_norm": 1.622687578201294,
      "learning_rate": 0.0005,
      "loss": 2.6326370239257812,
      "step": 3873
    },
    {
      "epoch": 0.7472993827160493,
      "grad_norm": 1.4497500658035278,
      "learning_rate": 0.0005,
      "loss": 2.4899349212646484,
      "step": 3874
    },
    {
      "epoch": 0.7474922839506173,
      "grad_norm": 1.6268843412399292,
      "learning_rate": 0.0005,
      "loss": 2.445100784301758,
      "step": 3875
    },
    {
      "epoch": 0.7476851851851852,
      "grad_norm": 0.9105530977249146,
      "learning_rate": 0.0005,
      "loss": 2.492177963256836,
      "step": 3876
    },
    {
      "epoch": 0.7478780864197531,
      "grad_norm": 0.8493175506591797,
      "learning_rate": 0.0005,
      "loss": 2.4577808380126953,
      "step": 3877
    },
    {
      "epoch": 0.748070987654321,
      "grad_norm": 1.162049412727356,
      "learning_rate": 0.0005,
      "loss": 2.383647918701172,
      "step": 3878
    },
    {
      "epoch": 0.7482638888888888,
      "grad_norm": 1.162028193473816,
      "learning_rate": 0.0005,
      "loss": 2.3409523963928223,
      "step": 3879
    },
    {
      "epoch": 0.7484567901234568,
      "grad_norm": 0.9799649119377136,
      "learning_rate": 0.0005,
      "loss": 2.435936450958252,
      "step": 3880
    },
    {
      "epoch": 0.7486496913580247,
      "grad_norm": 0.8509001731872559,
      "learning_rate": 0.0005,
      "loss": 2.2906458377838135,
      "step": 3881
    },
    {
      "epoch": 0.7488425925925926,
      "grad_norm": 1.1562252044677734,
      "learning_rate": 0.0005,
      "loss": 2.4885101318359375,
      "step": 3882
    },
    {
      "epoch": 0.7490354938271605,
      "grad_norm": 1.090141773223877,
      "learning_rate": 0.0005,
      "loss": 2.4292614459991455,
      "step": 3883
    },
    {
      "epoch": 0.7492283950617284,
      "grad_norm": 1.087666392326355,
      "learning_rate": 0.0005,
      "loss": 2.6404762268066406,
      "step": 3884
    },
    {
      "epoch": 0.7494212962962963,
      "grad_norm": 0.8558295965194702,
      "learning_rate": 0.0005,
      "loss": 2.2948458194732666,
      "step": 3885
    },
    {
      "epoch": 0.7496141975308642,
      "grad_norm": 1.077296257019043,
      "learning_rate": 0.0005,
      "loss": 2.5559446811676025,
      "step": 3886
    },
    {
      "epoch": 0.7498070987654321,
      "grad_norm": 0.8130022883415222,
      "learning_rate": 0.0005,
      "loss": 2.136443614959717,
      "step": 3887
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.7938363552093506,
      "learning_rate": 0.0005,
      "loss": 2.3364710807800293,
      "step": 3888
    },
    {
      "epoch": 0.7501929012345679,
      "grad_norm": 0.779546856880188,
      "learning_rate": 0.0005,
      "loss": 2.223085880279541,
      "step": 3889
    },
    {
      "epoch": 0.7503858024691358,
      "grad_norm": 1.1080012321472168,
      "learning_rate": 0.0005,
      "loss": 2.691929340362549,
      "step": 3890
    },
    {
      "epoch": 0.7505787037037037,
      "grad_norm": 0.6783884763717651,
      "learning_rate": 0.0005,
      "loss": 2.3783721923828125,
      "step": 3891
    },
    {
      "epoch": 0.7507716049382716,
      "grad_norm": 1.0220589637756348,
      "learning_rate": 0.0005,
      "loss": 2.5139713287353516,
      "step": 3892
    },
    {
      "epoch": 0.7509645061728395,
      "grad_norm": 1.0948463678359985,
      "learning_rate": 0.0005,
      "loss": 2.477851152420044,
      "step": 3893
    },
    {
      "epoch": 0.7511574074074074,
      "grad_norm": 0.9309027194976807,
      "learning_rate": 0.0005,
      "loss": 2.2091996669769287,
      "step": 3894
    },
    {
      "epoch": 0.7513503086419753,
      "grad_norm": 0.9030259251594543,
      "learning_rate": 0.0005,
      "loss": 2.3933091163635254,
      "step": 3895
    },
    {
      "epoch": 0.7515432098765432,
      "grad_norm": 0.8052406311035156,
      "learning_rate": 0.0005,
      "loss": 2.619565725326538,
      "step": 3896
    },
    {
      "epoch": 0.7517361111111112,
      "grad_norm": 0.7572227716445923,
      "learning_rate": 0.0005,
      "loss": 2.1347880363464355,
      "step": 3897
    },
    {
      "epoch": 0.751929012345679,
      "grad_norm": 0.8555005192756653,
      "learning_rate": 0.0005,
      "loss": 2.275156021118164,
      "step": 3898
    },
    {
      "epoch": 0.7521219135802469,
      "grad_norm": 0.9835013747215271,
      "learning_rate": 0.0005,
      "loss": 2.3390676975250244,
      "step": 3899
    },
    {
      "epoch": 0.7523148148148148,
      "grad_norm": 1.00340735912323,
      "learning_rate": 0.0005,
      "loss": 2.337296485900879,
      "step": 3900
    },
    {
      "epoch": 0.7525077160493827,
      "grad_norm": 1.31837797164917,
      "learning_rate": 0.0005,
      "loss": 2.5123989582061768,
      "step": 3901
    },
    {
      "epoch": 0.7527006172839507,
      "grad_norm": 1.0357322692871094,
      "learning_rate": 0.0005,
      "loss": 2.221442699432373,
      "step": 3902
    },
    {
      "epoch": 0.7528935185185185,
      "grad_norm": 1.0070263147354126,
      "learning_rate": 0.0005,
      "loss": 2.360778331756592,
      "step": 3903
    },
    {
      "epoch": 0.7530864197530864,
      "grad_norm": 1.0646432638168335,
      "learning_rate": 0.0005,
      "loss": 2.254908800125122,
      "step": 3904
    },
    {
      "epoch": 0.7532793209876543,
      "grad_norm": 0.8078784942626953,
      "learning_rate": 0.0005,
      "loss": 2.412456512451172,
      "step": 3905
    },
    {
      "epoch": 0.7534722222222222,
      "grad_norm": 0.9481543302536011,
      "learning_rate": 0.0005,
      "loss": 2.1023142337799072,
      "step": 3906
    },
    {
      "epoch": 0.7536651234567902,
      "grad_norm": 0.9396046996116638,
      "learning_rate": 0.0005,
      "loss": 2.16507625579834,
      "step": 3907
    },
    {
      "epoch": 0.753858024691358,
      "grad_norm": 0.9909800887107849,
      "learning_rate": 0.0005,
      "loss": 2.252403736114502,
      "step": 3908
    },
    {
      "epoch": 0.7540509259259259,
      "grad_norm": 0.8053315877914429,
      "learning_rate": 0.0005,
      "loss": 2.2906999588012695,
      "step": 3909
    },
    {
      "epoch": 0.7542438271604939,
      "grad_norm": 1.0462552309036255,
      "learning_rate": 0.0005,
      "loss": 2.669482707977295,
      "step": 3910
    },
    {
      "epoch": 0.7544367283950617,
      "grad_norm": 0.7992264628410339,
      "learning_rate": 0.0005,
      "loss": 2.332186222076416,
      "step": 3911
    },
    {
      "epoch": 0.7546296296296297,
      "grad_norm": 0.6707723140716553,
      "learning_rate": 0.0005,
      "loss": 2.2748889923095703,
      "step": 3912
    },
    {
      "epoch": 0.7548225308641975,
      "grad_norm": 0.8143497109413147,
      "learning_rate": 0.0005,
      "loss": 2.3522534370422363,
      "step": 3913
    },
    {
      "epoch": 0.7550154320987654,
      "grad_norm": 0.8599134683609009,
      "learning_rate": 0.0005,
      "loss": 2.2685117721557617,
      "step": 3914
    },
    {
      "epoch": 0.7552083333333334,
      "grad_norm": 0.7393122315406799,
      "learning_rate": 0.0005,
      "loss": 2.3317904472351074,
      "step": 3915
    },
    {
      "epoch": 0.7554012345679012,
      "grad_norm": 0.7878591418266296,
      "learning_rate": 0.0005,
      "loss": 2.0615572929382324,
      "step": 3916
    },
    {
      "epoch": 0.7555941358024691,
      "grad_norm": 0.9246883988380432,
      "learning_rate": 0.0005,
      "loss": 2.4756181240081787,
      "step": 3917
    },
    {
      "epoch": 0.7557870370370371,
      "grad_norm": 0.8190301656723022,
      "learning_rate": 0.0005,
      "loss": 2.5345895290374756,
      "step": 3918
    },
    {
      "epoch": 0.7559799382716049,
      "grad_norm": 2.4433751106262207,
      "learning_rate": 0.0005,
      "loss": 2.869920492172241,
      "step": 3919
    },
    {
      "epoch": 0.7561728395061729,
      "grad_norm": 1.3844934701919556,
      "learning_rate": 0.0005,
      "loss": 2.4293625354766846,
      "step": 3920
    },
    {
      "epoch": 0.7563657407407407,
      "grad_norm": 1.0203297138214111,
      "learning_rate": 0.0005,
      "loss": 2.5262465476989746,
      "step": 3921
    },
    {
      "epoch": 0.7565586419753086,
      "grad_norm": 1.1457542181015015,
      "learning_rate": 0.0005,
      "loss": 2.495331287384033,
      "step": 3922
    },
    {
      "epoch": 0.7567515432098766,
      "grad_norm": 0.845828652381897,
      "learning_rate": 0.0005,
      "loss": 2.335512399673462,
      "step": 3923
    },
    {
      "epoch": 0.7569444444444444,
      "grad_norm": 0.9606698751449585,
      "learning_rate": 0.0005,
      "loss": 2.5956807136535645,
      "step": 3924
    },
    {
      "epoch": 0.7571373456790124,
      "grad_norm": 0.9383800029754639,
      "learning_rate": 0.0005,
      "loss": 2.3773040771484375,
      "step": 3925
    },
    {
      "epoch": 0.7573302469135802,
      "grad_norm": 0.8202351331710815,
      "learning_rate": 0.0005,
      "loss": 2.677915096282959,
      "step": 3926
    },
    {
      "epoch": 0.7575231481481481,
      "grad_norm": 1.1308661699295044,
      "learning_rate": 0.0005,
      "loss": 2.2128210067749023,
      "step": 3927
    },
    {
      "epoch": 0.7577160493827161,
      "grad_norm": 0.9325733184814453,
      "learning_rate": 0.0005,
      "loss": 2.6774888038635254,
      "step": 3928
    },
    {
      "epoch": 0.7579089506172839,
      "grad_norm": 0.7958887815475464,
      "learning_rate": 0.0005,
      "loss": 2.554892063140869,
      "step": 3929
    },
    {
      "epoch": 0.7581018518518519,
      "grad_norm": 1.148095726966858,
      "learning_rate": 0.0005,
      "loss": 2.4859540462493896,
      "step": 3930
    },
    {
      "epoch": 0.7582947530864198,
      "grad_norm": 0.8388424515724182,
      "learning_rate": 0.0005,
      "loss": 2.3321869373321533,
      "step": 3931
    },
    {
      "epoch": 0.7584876543209876,
      "grad_norm": 0.7733467221260071,
      "learning_rate": 0.0005,
      "loss": 2.6936864852905273,
      "step": 3932
    },
    {
      "epoch": 0.7586805555555556,
      "grad_norm": 0.9856963157653809,
      "learning_rate": 0.0005,
      "loss": 2.328181505203247,
      "step": 3933
    },
    {
      "epoch": 0.7588734567901234,
      "grad_norm": 0.8786025643348694,
      "learning_rate": 0.0005,
      "loss": 2.1668827533721924,
      "step": 3934
    },
    {
      "epoch": 0.7590663580246914,
      "grad_norm": 1.1242645978927612,
      "learning_rate": 0.0005,
      "loss": 2.2222819328308105,
      "step": 3935
    },
    {
      "epoch": 0.7592592592592593,
      "grad_norm": 1.158611536026001,
      "learning_rate": 0.0005,
      "loss": 2.617213249206543,
      "step": 3936
    },
    {
      "epoch": 0.7594521604938271,
      "grad_norm": 0.8727080821990967,
      "learning_rate": 0.0005,
      "loss": 2.598644256591797,
      "step": 3937
    },
    {
      "epoch": 0.7596450617283951,
      "grad_norm": 1.0695689916610718,
      "learning_rate": 0.0005,
      "loss": 2.4760870933532715,
      "step": 3938
    },
    {
      "epoch": 0.7598379629629629,
      "grad_norm": 0.8888970613479614,
      "learning_rate": 0.0005,
      "loss": 2.4834320545196533,
      "step": 3939
    },
    {
      "epoch": 0.7600308641975309,
      "grad_norm": 0.7925869822502136,
      "learning_rate": 0.0005,
      "loss": 2.478027820587158,
      "step": 3940
    },
    {
      "epoch": 0.7602237654320988,
      "grad_norm": 0.9552114605903625,
      "learning_rate": 0.0005,
      "loss": 2.3461015224456787,
      "step": 3941
    },
    {
      "epoch": 0.7604166666666666,
      "grad_norm": 1.0410107374191284,
      "learning_rate": 0.0005,
      "loss": 2.659524440765381,
      "step": 3942
    },
    {
      "epoch": 0.7606095679012346,
      "grad_norm": 0.9097924828529358,
      "learning_rate": 0.0005,
      "loss": 2.1649341583251953,
      "step": 3943
    },
    {
      "epoch": 0.7608024691358025,
      "grad_norm": 0.9485329389572144,
      "learning_rate": 0.0005,
      "loss": 2.5660507678985596,
      "step": 3944
    },
    {
      "epoch": 0.7609953703703703,
      "grad_norm": 1.3906539678573608,
      "learning_rate": 0.0005,
      "loss": 2.6074657440185547,
      "step": 3945
    },
    {
      "epoch": 0.7611882716049383,
      "grad_norm": 0.9654066562652588,
      "learning_rate": 0.0005,
      "loss": 2.4483642578125,
      "step": 3946
    },
    {
      "epoch": 0.7613811728395061,
      "grad_norm": 1.0378371477127075,
      "learning_rate": 0.0005,
      "loss": 2.544496774673462,
      "step": 3947
    },
    {
      "epoch": 0.7615740740740741,
      "grad_norm": 0.7246244549751282,
      "learning_rate": 0.0005,
      "loss": 2.2394776344299316,
      "step": 3948
    },
    {
      "epoch": 0.761766975308642,
      "grad_norm": 0.9750453233718872,
      "learning_rate": 0.0005,
      "loss": 2.163175106048584,
      "step": 3949
    },
    {
      "epoch": 0.7619598765432098,
      "grad_norm": 0.906547486782074,
      "learning_rate": 0.0005,
      "loss": 2.480085611343384,
      "step": 3950
    },
    {
      "epoch": 0.7621527777777778,
      "grad_norm": 0.7236847281455994,
      "learning_rate": 0.0005,
      "loss": 2.2031373977661133,
      "step": 3951
    },
    {
      "epoch": 0.7623456790123457,
      "grad_norm": 0.8418940305709839,
      "learning_rate": 0.0005,
      "loss": 2.515340566635132,
      "step": 3952
    },
    {
      "epoch": 0.7625385802469136,
      "grad_norm": 1.0386862754821777,
      "learning_rate": 0.0005,
      "loss": 2.4356658458709717,
      "step": 3953
    },
    {
      "epoch": 0.7627314814814815,
      "grad_norm": 0.8809767961502075,
      "learning_rate": 0.0005,
      "loss": 2.484015464782715,
      "step": 3954
    },
    {
      "epoch": 0.7629243827160493,
      "grad_norm": 1.0682590007781982,
      "learning_rate": 0.0005,
      "loss": 2.515921115875244,
      "step": 3955
    },
    {
      "epoch": 0.7631172839506173,
      "grad_norm": 0.8610051274299622,
      "learning_rate": 0.0005,
      "loss": 2.6550936698913574,
      "step": 3956
    },
    {
      "epoch": 0.7633101851851852,
      "grad_norm": 0.6860175132751465,
      "learning_rate": 0.0005,
      "loss": 2.240034580230713,
      "step": 3957
    },
    {
      "epoch": 0.7635030864197531,
      "grad_norm": 1.0900607109069824,
      "learning_rate": 0.0005,
      "loss": 2.458645820617676,
      "step": 3958
    },
    {
      "epoch": 0.763695987654321,
      "grad_norm": 0.8162661194801331,
      "learning_rate": 0.0005,
      "loss": 2.239633321762085,
      "step": 3959
    },
    {
      "epoch": 0.7638888888888888,
      "grad_norm": 0.9579989910125732,
      "learning_rate": 0.0005,
      "loss": 2.4677133560180664,
      "step": 3960
    },
    {
      "epoch": 0.7640817901234568,
      "grad_norm": 0.9456866979598999,
      "learning_rate": 0.0005,
      "loss": 2.620922327041626,
      "step": 3961
    },
    {
      "epoch": 0.7642746913580247,
      "grad_norm": 0.7483405470848083,
      "learning_rate": 0.0005,
      "loss": 2.639402389526367,
      "step": 3962
    },
    {
      "epoch": 0.7644675925925926,
      "grad_norm": 0.942765474319458,
      "learning_rate": 0.0005,
      "loss": 2.6795310974121094,
      "step": 3963
    },
    {
      "epoch": 0.7646604938271605,
      "grad_norm": 1.78927743434906,
      "learning_rate": 0.0005,
      "loss": 2.586726427078247,
      "step": 3964
    },
    {
      "epoch": 0.7648533950617284,
      "grad_norm": 0.8011084794998169,
      "learning_rate": 0.0005,
      "loss": 2.2644829750061035,
      "step": 3965
    },
    {
      "epoch": 0.7650462962962963,
      "grad_norm": 2.716822624206543,
      "learning_rate": 0.0005,
      "loss": 2.6757707595825195,
      "step": 3966
    },
    {
      "epoch": 0.7652391975308642,
      "grad_norm": 0.8723165988922119,
      "learning_rate": 0.0005,
      "loss": 2.7360262870788574,
      "step": 3967
    },
    {
      "epoch": 0.7654320987654321,
      "grad_norm": 0.932379961013794,
      "learning_rate": 0.0005,
      "loss": 2.3094711303710938,
      "step": 3968
    },
    {
      "epoch": 0.765625,
      "grad_norm": 1.085877776145935,
      "learning_rate": 0.0005,
      "loss": 2.594130039215088,
      "step": 3969
    },
    {
      "epoch": 0.7658179012345679,
      "grad_norm": 0.737618088722229,
      "learning_rate": 0.0005,
      "loss": 2.4395997524261475,
      "step": 3970
    },
    {
      "epoch": 0.7660108024691358,
      "grad_norm": 1.1489170789718628,
      "learning_rate": 0.0005,
      "loss": 2.3174242973327637,
      "step": 3971
    },
    {
      "epoch": 0.7662037037037037,
      "grad_norm": 0.9081822037696838,
      "learning_rate": 0.0005,
      "loss": 2.1619110107421875,
      "step": 3972
    },
    {
      "epoch": 0.7663966049382716,
      "grad_norm": 1.3356738090515137,
      "learning_rate": 0.0005,
      "loss": 2.339945077896118,
      "step": 3973
    },
    {
      "epoch": 0.7665895061728395,
      "grad_norm": 0.9878527522087097,
      "learning_rate": 0.0005,
      "loss": 2.2120916843414307,
      "step": 3974
    },
    {
      "epoch": 0.7667824074074074,
      "grad_norm": 0.8439035415649414,
      "learning_rate": 0.0005,
      "loss": 2.265385150909424,
      "step": 3975
    },
    {
      "epoch": 0.7669753086419753,
      "grad_norm": 1.053658127784729,
      "learning_rate": 0.0005,
      "loss": 2.3886876106262207,
      "step": 3976
    },
    {
      "epoch": 0.7671682098765432,
      "grad_norm": 1.081892967224121,
      "learning_rate": 0.0005,
      "loss": 2.6019465923309326,
      "step": 3977
    },
    {
      "epoch": 0.7673611111111112,
      "grad_norm": 0.8515622615814209,
      "learning_rate": 0.0005,
      "loss": 2.2067604064941406,
      "step": 3978
    },
    {
      "epoch": 0.767554012345679,
      "grad_norm": 0.989128053188324,
      "learning_rate": 0.0005,
      "loss": 2.2100629806518555,
      "step": 3979
    },
    {
      "epoch": 0.7677469135802469,
      "grad_norm": 0.7647455334663391,
      "learning_rate": 0.0005,
      "loss": 2.339670419692993,
      "step": 3980
    },
    {
      "epoch": 0.7679398148148148,
      "grad_norm": 0.8086956143379211,
      "learning_rate": 0.0005,
      "loss": 2.2240407466888428,
      "step": 3981
    },
    {
      "epoch": 0.7681327160493827,
      "grad_norm": 0.8862663507461548,
      "learning_rate": 0.0005,
      "loss": 2.2595503330230713,
      "step": 3982
    },
    {
      "epoch": 0.7683256172839507,
      "grad_norm": 1.511757731437683,
      "learning_rate": 0.0005,
      "loss": 2.471069574356079,
      "step": 3983
    },
    {
      "epoch": 0.7685185185185185,
      "grad_norm": 0.7727442979812622,
      "learning_rate": 0.0005,
      "loss": 2.449225664138794,
      "step": 3984
    },
    {
      "epoch": 0.7687114197530864,
      "grad_norm": 0.7791521549224854,
      "learning_rate": 0.0005,
      "loss": 2.695995807647705,
      "step": 3985
    },
    {
      "epoch": 0.7689043209876543,
      "grad_norm": 0.7527781128883362,
      "learning_rate": 0.0005,
      "loss": 2.201801300048828,
      "step": 3986
    },
    {
      "epoch": 0.7690972222222222,
      "grad_norm": 0.7310953736305237,
      "learning_rate": 0.0005,
      "loss": 2.548746109008789,
      "step": 3987
    },
    {
      "epoch": 0.7692901234567902,
      "grad_norm": 0.7418178915977478,
      "learning_rate": 0.0005,
      "loss": 2.347062826156616,
      "step": 3988
    },
    {
      "epoch": 0.769483024691358,
      "grad_norm": 0.8712979555130005,
      "learning_rate": 0.0005,
      "loss": 2.1597135066986084,
      "step": 3989
    },
    {
      "epoch": 0.7696759259259259,
      "grad_norm": 0.9210782647132874,
      "learning_rate": 0.0005,
      "loss": 2.4289565086364746,
      "step": 3990
    },
    {
      "epoch": 0.7698688271604939,
      "grad_norm": 0.7567746639251709,
      "learning_rate": 0.0005,
      "loss": 2.3339576721191406,
      "step": 3991
    },
    {
      "epoch": 0.7700617283950617,
      "grad_norm": 1.2974834442138672,
      "learning_rate": 0.0005,
      "loss": 2.352954387664795,
      "step": 3992
    },
    {
      "epoch": 0.7702546296296297,
      "grad_norm": 0.7149512767791748,
      "learning_rate": 0.0005,
      "loss": 2.2995572090148926,
      "step": 3993
    },
    {
      "epoch": 0.7704475308641975,
      "grad_norm": 0.8217846155166626,
      "learning_rate": 0.0005,
      "loss": 2.297964334487915,
      "step": 3994
    },
    {
      "epoch": 0.7706404320987654,
      "grad_norm": 0.6223347783088684,
      "learning_rate": 0.0005,
      "loss": 2.124302387237549,
      "step": 3995
    },
    {
      "epoch": 0.7708333333333334,
      "grad_norm": 1.0884701013565063,
      "learning_rate": 0.0005,
      "loss": 2.4449949264526367,
      "step": 3996
    },
    {
      "epoch": 0.7710262345679012,
      "grad_norm": 0.8420057892799377,
      "learning_rate": 0.0005,
      "loss": 2.2909035682678223,
      "step": 3997
    },
    {
      "epoch": 0.7712191358024691,
      "grad_norm": 0.6257821917533875,
      "learning_rate": 0.0005,
      "loss": 2.2956326007843018,
      "step": 3998
    },
    {
      "epoch": 0.7714120370370371,
      "grad_norm": 1.5246082544326782,
      "learning_rate": 0.0005,
      "loss": 2.488067626953125,
      "step": 3999
    },
    {
      "epoch": 0.7716049382716049,
      "grad_norm": 0.7826106548309326,
      "learning_rate": 0.0005,
      "loss": 2.1418137550354004,
      "step": 4000
    },
    {
      "epoch": 0.7717978395061729,
      "grad_norm": 0.8844712376594543,
      "learning_rate": 0.0005,
      "loss": 2.4129796028137207,
      "step": 4001
    },
    {
      "epoch": 0.7719907407407407,
      "grad_norm": 0.8998258113861084,
      "learning_rate": 0.0005,
      "loss": 2.4336934089660645,
      "step": 4002
    },
    {
      "epoch": 0.7721836419753086,
      "grad_norm": 0.8387900590896606,
      "learning_rate": 0.0005,
      "loss": 2.213240146636963,
      "step": 4003
    },
    {
      "epoch": 0.7723765432098766,
      "grad_norm": 0.9509584903717041,
      "learning_rate": 0.0005,
      "loss": 2.132598876953125,
      "step": 4004
    },
    {
      "epoch": 0.7725694444444444,
      "grad_norm": 0.7074143290519714,
      "learning_rate": 0.0005,
      "loss": 2.3624496459960938,
      "step": 4005
    },
    {
      "epoch": 0.7727623456790124,
      "grad_norm": 0.9452943205833435,
      "learning_rate": 0.0005,
      "loss": 2.499594211578369,
      "step": 4006
    },
    {
      "epoch": 0.7729552469135802,
      "grad_norm": 0.9162858724594116,
      "learning_rate": 0.0005,
      "loss": 2.156684637069702,
      "step": 4007
    },
    {
      "epoch": 0.7731481481481481,
      "grad_norm": 1.4112344980239868,
      "learning_rate": 0.0005,
      "loss": 2.3094468116760254,
      "step": 4008
    },
    {
      "epoch": 0.7733410493827161,
      "grad_norm": 0.9268828630447388,
      "learning_rate": 0.0005,
      "loss": 2.4608049392700195,
      "step": 4009
    },
    {
      "epoch": 0.7735339506172839,
      "grad_norm": 1.1239691972732544,
      "learning_rate": 0.0005,
      "loss": 2.7129459381103516,
      "step": 4010
    },
    {
      "epoch": 0.7737268518518519,
      "grad_norm": 0.7099359035491943,
      "learning_rate": 0.0005,
      "loss": 2.446658134460449,
      "step": 4011
    },
    {
      "epoch": 0.7739197530864198,
      "grad_norm": 0.8400730490684509,
      "learning_rate": 0.0005,
      "loss": 2.5276436805725098,
      "step": 4012
    },
    {
      "epoch": 0.7741126543209876,
      "grad_norm": 0.8276575207710266,
      "learning_rate": 0.0005,
      "loss": 2.4473776817321777,
      "step": 4013
    },
    {
      "epoch": 0.7743055555555556,
      "grad_norm": 0.9222487807273865,
      "learning_rate": 0.0005,
      "loss": 2.439368724822998,
      "step": 4014
    },
    {
      "epoch": 0.7744984567901234,
      "grad_norm": 0.9444565176963806,
      "learning_rate": 0.0005,
      "loss": 2.252485990524292,
      "step": 4015
    },
    {
      "epoch": 0.7746913580246914,
      "grad_norm": 0.8934317827224731,
      "learning_rate": 0.0005,
      "loss": 2.5161499977111816,
      "step": 4016
    },
    {
      "epoch": 0.7748842592592593,
      "grad_norm": 0.8747447729110718,
      "learning_rate": 0.0005,
      "loss": 2.3357532024383545,
      "step": 4017
    },
    {
      "epoch": 0.7750771604938271,
      "grad_norm": 0.9775605201721191,
      "learning_rate": 0.0005,
      "loss": 2.3393051624298096,
      "step": 4018
    },
    {
      "epoch": 0.7752700617283951,
      "grad_norm": 0.8720065951347351,
      "learning_rate": 0.0005,
      "loss": 2.4698684215545654,
      "step": 4019
    },
    {
      "epoch": 0.7754629629629629,
      "grad_norm": 0.9905489683151245,
      "learning_rate": 0.0005,
      "loss": 2.552321434020996,
      "step": 4020
    },
    {
      "epoch": 0.7756558641975309,
      "grad_norm": 1.1194870471954346,
      "learning_rate": 0.0005,
      "loss": 2.3011646270751953,
      "step": 4021
    },
    {
      "epoch": 0.7758487654320988,
      "grad_norm": 0.7594485878944397,
      "learning_rate": 0.0005,
      "loss": 2.3731601238250732,
      "step": 4022
    },
    {
      "epoch": 0.7760416666666666,
      "grad_norm": 1.293403148651123,
      "learning_rate": 0.0005,
      "loss": 2.698263645172119,
      "step": 4023
    },
    {
      "epoch": 0.7762345679012346,
      "grad_norm": 1.5036215782165527,
      "learning_rate": 0.0005,
      "loss": 2.474787473678589,
      "step": 4024
    },
    {
      "epoch": 0.7764274691358025,
      "grad_norm": 0.7776609659194946,
      "learning_rate": 0.0005,
      "loss": 2.072199821472168,
      "step": 4025
    },
    {
      "epoch": 0.7766203703703703,
      "grad_norm": 0.8394656181335449,
      "learning_rate": 0.0005,
      "loss": 2.5785861015319824,
      "step": 4026
    },
    {
      "epoch": 0.7768132716049383,
      "grad_norm": 0.9697903394699097,
      "learning_rate": 0.0005,
      "loss": 2.316352367401123,
      "step": 4027
    },
    {
      "epoch": 0.7770061728395061,
      "grad_norm": 0.843604564666748,
      "learning_rate": 0.0005,
      "loss": 2.4543607234954834,
      "step": 4028
    },
    {
      "epoch": 0.7771990740740741,
      "grad_norm": 0.935125470161438,
      "learning_rate": 0.0005,
      "loss": 2.476907253265381,
      "step": 4029
    },
    {
      "epoch": 0.777391975308642,
      "grad_norm": 0.7920745015144348,
      "learning_rate": 0.0005,
      "loss": 2.348292827606201,
      "step": 4030
    },
    {
      "epoch": 0.7775848765432098,
      "grad_norm": 1.2164994478225708,
      "learning_rate": 0.0005,
      "loss": 2.584124803543091,
      "step": 4031
    },
    {
      "epoch": 0.7777777777777778,
      "grad_norm": 0.8090220093727112,
      "learning_rate": 0.0005,
      "loss": 2.46390438079834,
      "step": 4032
    },
    {
      "epoch": 0.7779706790123457,
      "grad_norm": 0.7489145994186401,
      "learning_rate": 0.0005,
      "loss": 2.204920530319214,
      "step": 4033
    },
    {
      "epoch": 0.7781635802469136,
      "grad_norm": 1.9232102632522583,
      "learning_rate": 0.0005,
      "loss": 2.557051181793213,
      "step": 4034
    },
    {
      "epoch": 0.7783564814814815,
      "grad_norm": 0.7654526829719543,
      "learning_rate": 0.0005,
      "loss": 2.343398094177246,
      "step": 4035
    },
    {
      "epoch": 0.7785493827160493,
      "grad_norm": 1.0198501348495483,
      "learning_rate": 0.0005,
      "loss": 2.082815647125244,
      "step": 4036
    },
    {
      "epoch": 0.7787422839506173,
      "grad_norm": 1.1790084838867188,
      "learning_rate": 0.0005,
      "loss": 2.4954323768615723,
      "step": 4037
    },
    {
      "epoch": 0.7789351851851852,
      "grad_norm": 0.7007675766944885,
      "learning_rate": 0.0005,
      "loss": 2.081960678100586,
      "step": 4038
    },
    {
      "epoch": 0.7791280864197531,
      "grad_norm": 0.9582138061523438,
      "learning_rate": 0.0005,
      "loss": 2.4613537788391113,
      "step": 4039
    },
    {
      "epoch": 0.779320987654321,
      "grad_norm": 0.7495672106742859,
      "learning_rate": 0.0005,
      "loss": 2.4526829719543457,
      "step": 4040
    },
    {
      "epoch": 0.7795138888888888,
      "grad_norm": 0.8000105023384094,
      "learning_rate": 0.0005,
      "loss": 2.378077983856201,
      "step": 4041
    },
    {
      "epoch": 0.7797067901234568,
      "grad_norm": 0.9697273969650269,
      "learning_rate": 0.0005,
      "loss": 2.402723550796509,
      "step": 4042
    },
    {
      "epoch": 0.7798996913580247,
      "grad_norm": 0.605275571346283,
      "learning_rate": 0.0005,
      "loss": 2.182840347290039,
      "step": 4043
    },
    {
      "epoch": 0.7800925925925926,
      "grad_norm": 1.0504924058914185,
      "learning_rate": 0.0005,
      "loss": 2.4709181785583496,
      "step": 4044
    },
    {
      "epoch": 0.7802854938271605,
      "grad_norm": 0.7157862782478333,
      "learning_rate": 0.0005,
      "loss": 2.1774308681488037,
      "step": 4045
    },
    {
      "epoch": 0.7804783950617284,
      "grad_norm": 1.0682259798049927,
      "learning_rate": 0.0005,
      "loss": 2.4373159408569336,
      "step": 4046
    },
    {
      "epoch": 0.7806712962962963,
      "grad_norm": 0.7064477801322937,
      "learning_rate": 0.0005,
      "loss": 2.529397487640381,
      "step": 4047
    },
    {
      "epoch": 0.7808641975308642,
      "grad_norm": 0.7985071539878845,
      "learning_rate": 0.0005,
      "loss": 2.6113483905792236,
      "step": 4048
    },
    {
      "epoch": 0.7810570987654321,
      "grad_norm": 0.8346063494682312,
      "learning_rate": 0.0005,
      "loss": 2.2347683906555176,
      "step": 4049
    },
    {
      "epoch": 0.78125,
      "grad_norm": 0.8159800171852112,
      "learning_rate": 0.0005,
      "loss": 2.538214921951294,
      "step": 4050
    },
    {
      "epoch": 0.7814429012345679,
      "grad_norm": 0.8962634205818176,
      "learning_rate": 0.0005,
      "loss": 2.550663948059082,
      "step": 4051
    },
    {
      "epoch": 0.7816358024691358,
      "grad_norm": 0.891882598400116,
      "learning_rate": 0.0005,
      "loss": 2.1414408683776855,
      "step": 4052
    },
    {
      "epoch": 0.7818287037037037,
      "grad_norm": 0.875157356262207,
      "learning_rate": 0.0005,
      "loss": 2.2213892936706543,
      "step": 4053
    },
    {
      "epoch": 0.7820216049382716,
      "grad_norm": 0.9292505383491516,
      "learning_rate": 0.0005,
      "loss": 2.594235897064209,
      "step": 4054
    },
    {
      "epoch": 0.7822145061728395,
      "grad_norm": 0.6782557368278503,
      "learning_rate": 0.0005,
      "loss": 2.2899184226989746,
      "step": 4055
    },
    {
      "epoch": 0.7824074074074074,
      "grad_norm": 0.9250699877738953,
      "learning_rate": 0.0005,
      "loss": 2.5226664543151855,
      "step": 4056
    },
    {
      "epoch": 0.7826003086419753,
      "grad_norm": 0.7693638801574707,
      "learning_rate": 0.0005,
      "loss": 2.097378730773926,
      "step": 4057
    },
    {
      "epoch": 0.7827932098765432,
      "grad_norm": 0.770200252532959,
      "learning_rate": 0.0005,
      "loss": 2.3532955646514893,
      "step": 4058
    },
    {
      "epoch": 0.7829861111111112,
      "grad_norm": 0.9757989645004272,
      "learning_rate": 0.0005,
      "loss": 2.461190700531006,
      "step": 4059
    },
    {
      "epoch": 0.783179012345679,
      "grad_norm": 0.9127476215362549,
      "learning_rate": 0.0005,
      "loss": 2.4384613037109375,
      "step": 4060
    },
    {
      "epoch": 0.7833719135802469,
      "grad_norm": 0.9162417650222778,
      "learning_rate": 0.0005,
      "loss": 2.4460906982421875,
      "step": 4061
    },
    {
      "epoch": 0.7835648148148148,
      "grad_norm": 0.6844830513000488,
      "learning_rate": 0.0005,
      "loss": 2.482504367828369,
      "step": 4062
    },
    {
      "epoch": 0.7837577160493827,
      "grad_norm": 0.9107523560523987,
      "learning_rate": 0.0005,
      "loss": 2.078497886657715,
      "step": 4063
    },
    {
      "epoch": 0.7839506172839507,
      "grad_norm": 1.0750384330749512,
      "learning_rate": 0.0005,
      "loss": 2.671895980834961,
      "step": 4064
    },
    {
      "epoch": 0.7841435185185185,
      "grad_norm": 0.724412202835083,
      "learning_rate": 0.0005,
      "loss": 2.336068630218506,
      "step": 4065
    },
    {
      "epoch": 0.7843364197530864,
      "grad_norm": 0.7198500037193298,
      "learning_rate": 0.0005,
      "loss": 2.212200880050659,
      "step": 4066
    },
    {
      "epoch": 0.7845293209876543,
      "grad_norm": 0.8457812666893005,
      "learning_rate": 0.0005,
      "loss": 2.1299638748168945,
      "step": 4067
    },
    {
      "epoch": 0.7847222222222222,
      "grad_norm": 0.8622943758964539,
      "learning_rate": 0.0005,
      "loss": 2.3139166831970215,
      "step": 4068
    },
    {
      "epoch": 0.7849151234567902,
      "grad_norm": 0.854909360408783,
      "learning_rate": 0.0005,
      "loss": 2.1912832260131836,
      "step": 4069
    },
    {
      "epoch": 0.785108024691358,
      "grad_norm": 0.8137091398239136,
      "learning_rate": 0.0005,
      "loss": 2.3104162216186523,
      "step": 4070
    },
    {
      "epoch": 0.7853009259259259,
      "grad_norm": 0.928063154220581,
      "learning_rate": 0.0005,
      "loss": 2.4828262329101562,
      "step": 4071
    },
    {
      "epoch": 0.7854938271604939,
      "grad_norm": 0.7025105953216553,
      "learning_rate": 0.0005,
      "loss": 2.4227242469787598,
      "step": 4072
    },
    {
      "epoch": 0.7856867283950617,
      "grad_norm": 0.7628340125083923,
      "learning_rate": 0.0005,
      "loss": 2.364288091659546,
      "step": 4073
    },
    {
      "epoch": 0.7858796296296297,
      "grad_norm": 0.7267513871192932,
      "learning_rate": 0.0005,
      "loss": 2.4328927993774414,
      "step": 4074
    },
    {
      "epoch": 0.7860725308641975,
      "grad_norm": 0.8485326766967773,
      "learning_rate": 0.0005,
      "loss": 2.40153169631958,
      "step": 4075
    },
    {
      "epoch": 0.7862654320987654,
      "grad_norm": 0.7791057825088501,
      "learning_rate": 0.0005,
      "loss": 2.626091718673706,
      "step": 4076
    },
    {
      "epoch": 0.7864583333333334,
      "grad_norm": 0.8374156951904297,
      "learning_rate": 0.0005,
      "loss": 2.2526073455810547,
      "step": 4077
    },
    {
      "epoch": 0.7866512345679012,
      "grad_norm": 0.774071455001831,
      "learning_rate": 0.0005,
      "loss": 2.280630111694336,
      "step": 4078
    },
    {
      "epoch": 0.7868441358024691,
      "grad_norm": 0.7552545070648193,
      "learning_rate": 0.0005,
      "loss": 2.632420063018799,
      "step": 4079
    },
    {
      "epoch": 0.7870370370370371,
      "grad_norm": 0.9481733441352844,
      "learning_rate": 0.0005,
      "loss": 2.042095422744751,
      "step": 4080
    },
    {
      "epoch": 0.7872299382716049,
      "grad_norm": 0.8494157791137695,
      "learning_rate": 0.0005,
      "loss": 2.5043983459472656,
      "step": 4081
    },
    {
      "epoch": 0.7874228395061729,
      "grad_norm": 1.5429519414901733,
      "learning_rate": 0.0005,
      "loss": 2.6440091133117676,
      "step": 4082
    },
    {
      "epoch": 0.7876157407407407,
      "grad_norm": 1.0130301713943481,
      "learning_rate": 0.0005,
      "loss": 2.425809621810913,
      "step": 4083
    },
    {
      "epoch": 0.7878086419753086,
      "grad_norm": 0.8762475252151489,
      "learning_rate": 0.0005,
      "loss": 2.655595064163208,
      "step": 4084
    },
    {
      "epoch": 0.7880015432098766,
      "grad_norm": 0.7413312792778015,
      "learning_rate": 0.0005,
      "loss": 2.617795467376709,
      "step": 4085
    },
    {
      "epoch": 0.7881944444444444,
      "grad_norm": 0.9028379917144775,
      "learning_rate": 0.0005,
      "loss": 2.5845301151275635,
      "step": 4086
    },
    {
      "epoch": 0.7883873456790124,
      "grad_norm": 0.807934045791626,
      "learning_rate": 0.0005,
      "loss": 2.273068428039551,
      "step": 4087
    },
    {
      "epoch": 0.7885802469135802,
      "grad_norm": 0.8402209281921387,
      "learning_rate": 0.0005,
      "loss": 2.410771369934082,
      "step": 4088
    },
    {
      "epoch": 0.7887731481481481,
      "grad_norm": 0.8632375597953796,
      "learning_rate": 0.0005,
      "loss": 2.4185657501220703,
      "step": 4089
    },
    {
      "epoch": 0.7889660493827161,
      "grad_norm": 1.0473322868347168,
      "learning_rate": 0.0005,
      "loss": 2.273925542831421,
      "step": 4090
    },
    {
      "epoch": 0.7891589506172839,
      "grad_norm": 0.8406966328620911,
      "learning_rate": 0.0005,
      "loss": 2.207180976867676,
      "step": 4091
    },
    {
      "epoch": 0.7893518518518519,
      "grad_norm": 0.8855248093605042,
      "learning_rate": 0.0005,
      "loss": 2.476670265197754,
      "step": 4092
    },
    {
      "epoch": 0.7895447530864198,
      "grad_norm": 1.0232218503952026,
      "learning_rate": 0.0005,
      "loss": 2.443920850753784,
      "step": 4093
    },
    {
      "epoch": 0.7897376543209876,
      "grad_norm": 0.8633549213409424,
      "learning_rate": 0.0005,
      "loss": 2.595289707183838,
      "step": 4094
    },
    {
      "epoch": 0.7899305555555556,
      "grad_norm": 0.7441900968551636,
      "learning_rate": 0.0005,
      "loss": 2.3890085220336914,
      "step": 4095
    },
    {
      "epoch": 0.7901234567901234,
      "grad_norm": 0.7922549247741699,
      "learning_rate": 0.0005,
      "loss": 2.419116497039795,
      "step": 4096
    },
    {
      "epoch": 0.7903163580246914,
      "grad_norm": 0.8341675400733948,
      "learning_rate": 0.0005,
      "loss": 2.2889575958251953,
      "step": 4097
    },
    {
      "epoch": 0.7905092592592593,
      "grad_norm": 0.6910841464996338,
      "learning_rate": 0.0005,
      "loss": 2.5187370777130127,
      "step": 4098
    },
    {
      "epoch": 0.7907021604938271,
      "grad_norm": 0.6758163571357727,
      "learning_rate": 0.0005,
      "loss": 2.2477314472198486,
      "step": 4099
    },
    {
      "epoch": 0.7908950617283951,
      "grad_norm": 1.3306610584259033,
      "learning_rate": 0.0005,
      "loss": 2.2504987716674805,
      "step": 4100
    },
    {
      "epoch": 0.7910879629629629,
      "grad_norm": 0.7686338424682617,
      "learning_rate": 0.0005,
      "loss": 2.4229207038879395,
      "step": 4101
    },
    {
      "epoch": 0.7912808641975309,
      "grad_norm": 1.7999602556228638,
      "learning_rate": 0.0005,
      "loss": 2.355602264404297,
      "step": 4102
    },
    {
      "epoch": 0.7914737654320988,
      "grad_norm": 0.7976091504096985,
      "learning_rate": 0.0005,
      "loss": 2.423712730407715,
      "step": 4103
    },
    {
      "epoch": 0.7916666666666666,
      "grad_norm": 0.6959619522094727,
      "learning_rate": 0.0005,
      "loss": 2.301720142364502,
      "step": 4104
    },
    {
      "epoch": 0.7918595679012346,
      "grad_norm": 0.8546098470687866,
      "learning_rate": 0.0005,
      "loss": 2.4092226028442383,
      "step": 4105
    },
    {
      "epoch": 0.7920524691358025,
      "grad_norm": 0.8682132959365845,
      "learning_rate": 0.0005,
      "loss": 2.5544466972351074,
      "step": 4106
    },
    {
      "epoch": 0.7922453703703703,
      "grad_norm": 0.8702096939086914,
      "learning_rate": 0.0005,
      "loss": 2.2067573070526123,
      "step": 4107
    },
    {
      "epoch": 0.7924382716049383,
      "grad_norm": 0.7896497249603271,
      "learning_rate": 0.0005,
      "loss": 2.4608354568481445,
      "step": 4108
    },
    {
      "epoch": 0.7926311728395061,
      "grad_norm": 0.7982761859893799,
      "learning_rate": 0.0005,
      "loss": 2.2110328674316406,
      "step": 4109
    },
    {
      "epoch": 0.7928240740740741,
      "grad_norm": 0.8336799740791321,
      "learning_rate": 0.0005,
      "loss": 2.3763322830200195,
      "step": 4110
    },
    {
      "epoch": 0.793016975308642,
      "grad_norm": 0.7510424256324768,
      "learning_rate": 0.0005,
      "loss": 2.391774892807007,
      "step": 4111
    },
    {
      "epoch": 0.7932098765432098,
      "grad_norm": 0.7424915432929993,
      "learning_rate": 0.0005,
      "loss": 2.4596478939056396,
      "step": 4112
    },
    {
      "epoch": 0.7934027777777778,
      "grad_norm": 0.718482494354248,
      "learning_rate": 0.0005,
      "loss": 2.3159477710723877,
      "step": 4113
    },
    {
      "epoch": 0.7935956790123457,
      "grad_norm": 0.8258423805236816,
      "learning_rate": 0.0005,
      "loss": 2.188807249069214,
      "step": 4114
    },
    {
      "epoch": 0.7937885802469136,
      "grad_norm": 1.029729962348938,
      "learning_rate": 0.0005,
      "loss": 2.42038631439209,
      "step": 4115
    },
    {
      "epoch": 0.7939814814814815,
      "grad_norm": 0.6969694495201111,
      "learning_rate": 0.0005,
      "loss": 2.4250693321228027,
      "step": 4116
    },
    {
      "epoch": 0.7941743827160493,
      "grad_norm": 0.7918003797531128,
      "learning_rate": 0.0005,
      "loss": 2.344073534011841,
      "step": 4117
    },
    {
      "epoch": 0.7943672839506173,
      "grad_norm": 0.7701045870780945,
      "learning_rate": 0.0005,
      "loss": 2.5760562419891357,
      "step": 4118
    },
    {
      "epoch": 0.7945601851851852,
      "grad_norm": 0.915513277053833,
      "learning_rate": 0.0005,
      "loss": 2.545067310333252,
      "step": 4119
    },
    {
      "epoch": 0.7947530864197531,
      "grad_norm": 1.0634770393371582,
      "learning_rate": 0.0005,
      "loss": 2.4852793216705322,
      "step": 4120
    },
    {
      "epoch": 0.794945987654321,
      "grad_norm": 0.760886013507843,
      "learning_rate": 0.0005,
      "loss": 2.209291458129883,
      "step": 4121
    },
    {
      "epoch": 0.7951388888888888,
      "grad_norm": 1.1500275135040283,
      "learning_rate": 0.0005,
      "loss": 2.370985984802246,
      "step": 4122
    },
    {
      "epoch": 0.7953317901234568,
      "grad_norm": 0.9748101234436035,
      "learning_rate": 0.0005,
      "loss": 2.5974605083465576,
      "step": 4123
    },
    {
      "epoch": 0.7955246913580247,
      "grad_norm": 1.0193097591400146,
      "learning_rate": 0.0005,
      "loss": 2.4586777687072754,
      "step": 4124
    },
    {
      "epoch": 0.7957175925925926,
      "grad_norm": 0.8007808327674866,
      "learning_rate": 0.0005,
      "loss": 2.178201913833618,
      "step": 4125
    },
    {
      "epoch": 0.7959104938271605,
      "grad_norm": 0.7361099123954773,
      "learning_rate": 0.0005,
      "loss": 2.628812313079834,
      "step": 4126
    },
    {
      "epoch": 0.7961033950617284,
      "grad_norm": 0.7246115803718567,
      "learning_rate": 0.0005,
      "loss": 2.1152472496032715,
      "step": 4127
    },
    {
      "epoch": 0.7962962962962963,
      "grad_norm": 0.9452292323112488,
      "learning_rate": 0.0005,
      "loss": 2.6400012969970703,
      "step": 4128
    },
    {
      "epoch": 0.7964891975308642,
      "grad_norm": 1.0212522745132446,
      "learning_rate": 0.0005,
      "loss": 2.5429842472076416,
      "step": 4129
    },
    {
      "epoch": 0.7966820987654321,
      "grad_norm": 1.0166866779327393,
      "learning_rate": 0.0005,
      "loss": 2.484938144683838,
      "step": 4130
    },
    {
      "epoch": 0.796875,
      "grad_norm": 1.0184763669967651,
      "learning_rate": 0.0005,
      "loss": 2.278028726577759,
      "step": 4131
    },
    {
      "epoch": 0.7970679012345679,
      "grad_norm": 1.0156147480010986,
      "learning_rate": 0.0005,
      "loss": 2.354581594467163,
      "step": 4132
    },
    {
      "epoch": 0.7972608024691358,
      "grad_norm": 1.332249641418457,
      "learning_rate": 0.0005,
      "loss": 2.5259523391723633,
      "step": 4133
    },
    {
      "epoch": 0.7974537037037037,
      "grad_norm": 0.9989661574363708,
      "learning_rate": 0.0005,
      "loss": 2.5243027210235596,
      "step": 4134
    },
    {
      "epoch": 0.7976466049382716,
      "grad_norm": 1.0408153533935547,
      "learning_rate": 0.0005,
      "loss": 2.706892728805542,
      "step": 4135
    },
    {
      "epoch": 0.7978395061728395,
      "grad_norm": 0.808062732219696,
      "learning_rate": 0.0005,
      "loss": 2.3767287731170654,
      "step": 4136
    },
    {
      "epoch": 0.7980324074074074,
      "grad_norm": 0.9538434147834778,
      "learning_rate": 0.0005,
      "loss": 2.4528145790100098,
      "step": 4137
    },
    {
      "epoch": 0.7982253086419753,
      "grad_norm": 1.2047502994537354,
      "learning_rate": 0.0005,
      "loss": 2.236860752105713,
      "step": 4138
    },
    {
      "epoch": 0.7984182098765432,
      "grad_norm": 0.9598162174224854,
      "learning_rate": 0.0005,
      "loss": 2.6154184341430664,
      "step": 4139
    },
    {
      "epoch": 0.7986111111111112,
      "grad_norm": 1.0180914402008057,
      "learning_rate": 0.0005,
      "loss": 2.772827625274658,
      "step": 4140
    },
    {
      "epoch": 0.798804012345679,
      "grad_norm": 1.0095425844192505,
      "learning_rate": 0.0005,
      "loss": 2.5337283611297607,
      "step": 4141
    },
    {
      "epoch": 0.7989969135802469,
      "grad_norm": 0.7676315307617188,
      "learning_rate": 0.0005,
      "loss": 2.2063703536987305,
      "step": 4142
    },
    {
      "epoch": 0.7991898148148148,
      "grad_norm": 0.817319929599762,
      "learning_rate": 0.0005,
      "loss": 1.9537323713302612,
      "step": 4143
    },
    {
      "epoch": 0.7993827160493827,
      "grad_norm": 0.8960694670677185,
      "learning_rate": 0.0005,
      "loss": 2.4206929206848145,
      "step": 4144
    },
    {
      "epoch": 0.7995756172839507,
      "grad_norm": 0.8777639269828796,
      "learning_rate": 0.0005,
      "loss": 2.4044620990753174,
      "step": 4145
    },
    {
      "epoch": 0.7997685185185185,
      "grad_norm": 0.9366826415061951,
      "learning_rate": 0.0005,
      "loss": 2.316382884979248,
      "step": 4146
    },
    {
      "epoch": 0.7999614197530864,
      "grad_norm": 0.762211799621582,
      "learning_rate": 0.0005,
      "loss": 2.4849367141723633,
      "step": 4147
    },
    {
      "epoch": 0.8001543209876543,
      "grad_norm": 0.9688001275062561,
      "learning_rate": 0.0005,
      "loss": 2.3638317584991455,
      "step": 4148
    },
    {
      "epoch": 0.8003472222222222,
      "grad_norm": 0.8036805391311646,
      "learning_rate": 0.0005,
      "loss": 2.4929888248443604,
      "step": 4149
    },
    {
      "epoch": 0.8005401234567902,
      "grad_norm": 0.954524040222168,
      "learning_rate": 0.0004999988505504495,
      "loss": 2.024932622909546,
      "step": 4150
    },
    {
      "epoch": 0.800733024691358,
      "grad_norm": 1.2028392553329468,
      "learning_rate": 0.0004999954022123678,
      "loss": 2.3980607986450195,
      "step": 4151
    },
    {
      "epoch": 0.8009259259259259,
      "grad_norm": 1.0656101703643799,
      "learning_rate": 0.0004999896550174647,
      "loss": 2.324502944946289,
      "step": 4152
    },
    {
      "epoch": 0.8011188271604939,
      "grad_norm": 0.7184222936630249,
      "learning_rate": 0.0004999816090185887,
      "loss": 2.184678554534912,
      "step": 4153
    },
    {
      "epoch": 0.8013117283950617,
      "grad_norm": 0.8506470918655396,
      "learning_rate": 0.0004999712642897278,
      "loss": 2.494540214538574,
      "step": 4154
    },
    {
      "epoch": 0.8015046296296297,
      "grad_norm": 3.795673370361328,
      "learning_rate": 0.000499958620926008,
      "loss": 2.3274805545806885,
      "step": 4155
    },
    {
      "epoch": 0.8016975308641975,
      "grad_norm": 1.0517100095748901,
      "learning_rate": 0.0004999436790436923,
      "loss": 2.602612018585205,
      "step": 4156
    },
    {
      "epoch": 0.8018904320987654,
      "grad_norm": 0.7593478560447693,
      "learning_rate": 0.0004999264387801805,
      "loss": 2.060001850128174,
      "step": 4157
    },
    {
      "epoch": 0.8020833333333334,
      "grad_norm": 0.8149726390838623,
      "learning_rate": 0.000499906900294007,
      "loss": 2.7282330989837646,
      "step": 4158
    },
    {
      "epoch": 0.8022762345679012,
      "grad_norm": 0.8883121609687805,
      "learning_rate": 0.0004998850637648398,
      "loss": 2.2408576011657715,
      "step": 4159
    },
    {
      "epoch": 0.8024691358024691,
      "grad_norm": 1.3402047157287598,
      "learning_rate": 0.0004998609293934788,
      "loss": 2.4989585876464844,
      "step": 4160
    },
    {
      "epoch": 0.8026620370370371,
      "grad_norm": 0.6920716762542725,
      "learning_rate": 0.000499834497401854,
      "loss": 2.5655994415283203,
      "step": 4161
    },
    {
      "epoch": 0.8028549382716049,
      "grad_norm": 0.8917982578277588,
      "learning_rate": 0.0004998057680330233,
      "loss": 2.2766590118408203,
      "step": 4162
    },
    {
      "epoch": 0.8030478395061729,
      "grad_norm": 0.9007627367973328,
      "learning_rate": 0.0004997747415511704,
      "loss": 2.2600247859954834,
      "step": 4163
    },
    {
      "epoch": 0.8032407407407407,
      "grad_norm": 0.9581298828125,
      "learning_rate": 0.0004997414182416022,
      "loss": 2.4484879970550537,
      "step": 4164
    },
    {
      "epoch": 0.8034336419753086,
      "grad_norm": 0.874541163444519,
      "learning_rate": 0.0004997057984107466,
      "loss": 2.388497829437256,
      "step": 4165
    },
    {
      "epoch": 0.8036265432098766,
      "grad_norm": 0.9479749798774719,
      "learning_rate": 0.000499667882386149,
      "loss": 2.7548580169677734,
      "step": 4166
    },
    {
      "epoch": 0.8038194444444444,
      "grad_norm": 0.7654219269752502,
      "learning_rate": 0.00049962767051647,
      "loss": 2.4212636947631836,
      "step": 4167
    },
    {
      "epoch": 0.8040123456790124,
      "grad_norm": 1.0915426015853882,
      "learning_rate": 0.0004995851631714816,
      "loss": 2.148529291152954,
      "step": 4168
    },
    {
      "epoch": 0.8042052469135802,
      "grad_norm": 0.9960481524467468,
      "learning_rate": 0.0004995403607420643,
      "loss": 1.9884190559387207,
      "step": 4169
    },
    {
      "epoch": 0.8043981481481481,
      "grad_norm": 0.9067742228507996,
      "learning_rate": 0.0004994932636402031,
      "loss": 2.122793674468994,
      "step": 4170
    },
    {
      "epoch": 0.8045910493827161,
      "grad_norm": 0.6941888332366943,
      "learning_rate": 0.000499443872298984,
      "loss": 2.1400671005249023,
      "step": 4171
    },
    {
      "epoch": 0.8047839506172839,
      "grad_norm": 0.7285593748092651,
      "learning_rate": 0.0004993921871725897,
      "loss": 2.0411882400512695,
      "step": 4172
    },
    {
      "epoch": 0.8049768518518519,
      "grad_norm": 0.7238715887069702,
      "learning_rate": 0.0004993382087362959,
      "loss": 2.6067373752593994,
      "step": 4173
    },
    {
      "epoch": 0.8051697530864198,
      "grad_norm": 0.7464205026626587,
      "learning_rate": 0.0004992819374864665,
      "loss": 2.0539066791534424,
      "step": 4174
    },
    {
      "epoch": 0.8053626543209876,
      "grad_norm": 0.750238835811615,
      "learning_rate": 0.0004992233739405491,
      "loss": 2.150967836380005,
      "step": 4175
    },
    {
      "epoch": 0.8055555555555556,
      "grad_norm": 1.1631964445114136,
      "learning_rate": 0.0004991625186370706,
      "loss": 2.2481210231781006,
      "step": 4176
    },
    {
      "epoch": 0.8057484567901234,
      "grad_norm": 0.7934996485710144,
      "learning_rate": 0.0004990993721356316,
      "loss": 2.533724784851074,
      "step": 4177
    },
    {
      "epoch": 0.8059413580246914,
      "grad_norm": 1.1923227310180664,
      "learning_rate": 0.0004990339350169021,
      "loss": 2.1868152618408203,
      "step": 4178
    },
    {
      "epoch": 0.8061342592592593,
      "grad_norm": 0.8983781933784485,
      "learning_rate": 0.0004989662078826152,
      "loss": 2.3553810119628906,
      "step": 4179
    },
    {
      "epoch": 0.8063271604938271,
      "grad_norm": 0.9483025670051575,
      "learning_rate": 0.0004988961913555623,
      "loss": 2.2783002853393555,
      "step": 4180
    },
    {
      "epoch": 0.8065200617283951,
      "grad_norm": 0.9404417276382446,
      "learning_rate": 0.0004988238860795873,
      "loss": 2.2413573265075684,
      "step": 4181
    },
    {
      "epoch": 0.8067129629629629,
      "grad_norm": 0.7232756018638611,
      "learning_rate": 0.0004987492927195803,
      "loss": 2.476548194885254,
      "step": 4182
    },
    {
      "epoch": 0.8069058641975309,
      "grad_norm": 0.8728649616241455,
      "learning_rate": 0.0004986724119614715,
      "loss": 2.2621612548828125,
      "step": 4183
    },
    {
      "epoch": 0.8070987654320988,
      "grad_norm": 0.868705689907074,
      "learning_rate": 0.0004985932445122257,
      "loss": 2.1804168224334717,
      "step": 4184
    },
    {
      "epoch": 0.8072916666666666,
      "grad_norm": 1.1848816871643066,
      "learning_rate": 0.0004985117910998345,
      "loss": 2.4312050342559814,
      "step": 4185
    },
    {
      "epoch": 0.8074845679012346,
      "grad_norm": 0.8168928623199463,
      "learning_rate": 0.0004984280524733107,
      "loss": 2.4364686012268066,
      "step": 4186
    },
    {
      "epoch": 0.8076774691358025,
      "grad_norm": 0.7525685429573059,
      "learning_rate": 0.000498342029402681,
      "loss": 2.180500030517578,
      "step": 4187
    },
    {
      "epoch": 0.8078703703703703,
      "grad_norm": 1.013559341430664,
      "learning_rate": 0.0004982537226789787,
      "loss": 2.382143497467041,
      "step": 4188
    },
    {
      "epoch": 0.8080632716049383,
      "grad_norm": 1.4017457962036133,
      "learning_rate": 0.0004981631331142368,
      "loss": 2.3631935119628906,
      "step": 4189
    },
    {
      "epoch": 0.8082561728395061,
      "grad_norm": 0.9139546155929565,
      "learning_rate": 0.0004980702615414805,
      "loss": 2.402909755706787,
      "step": 4190
    },
    {
      "epoch": 0.8084490740740741,
      "grad_norm": 0.6888989210128784,
      "learning_rate": 0.0004979751088147192,
      "loss": 2.0448317527770996,
      "step": 4191
    },
    {
      "epoch": 0.808641975308642,
      "grad_norm": 0.7829864025115967,
      "learning_rate": 0.0004978776758089391,
      "loss": 2.423956871032715,
      "step": 4192
    },
    {
      "epoch": 0.8088348765432098,
      "grad_norm": 0.9188740253448486,
      "learning_rate": 0.0004977779634200946,
      "loss": 2.4399123191833496,
      "step": 4193
    },
    {
      "epoch": 0.8090277777777778,
      "grad_norm": 0.8715519905090332,
      "learning_rate": 0.0004976759725651008,
      "loss": 2.493319034576416,
      "step": 4194
    },
    {
      "epoch": 0.8092206790123457,
      "grad_norm": 0.9550977349281311,
      "learning_rate": 0.0004975717041818242,
      "loss": 2.4751477241516113,
      "step": 4195
    },
    {
      "epoch": 0.8094135802469136,
      "grad_norm": 0.722876787185669,
      "learning_rate": 0.000497465159229075,
      "loss": 2.177001953125,
      "step": 4196
    },
    {
      "epoch": 0.8096064814814815,
      "grad_norm": 0.6902057528495789,
      "learning_rate": 0.0004973563386865974,
      "loss": 2.1002793312072754,
      "step": 4197
    },
    {
      "epoch": 0.8097993827160493,
      "grad_norm": 1.1368229389190674,
      "learning_rate": 0.0004972452435550613,
      "loss": 2.6865930557250977,
      "step": 4198
    },
    {
      "epoch": 0.8099922839506173,
      "grad_norm": 0.8254854083061218,
      "learning_rate": 0.0004971318748560527,
      "loss": 2.4536538124084473,
      "step": 4199
    },
    {
      "epoch": 0.8101851851851852,
      "grad_norm": 0.9757571220397949,
      "learning_rate": 0.0004970162336320644,
      "loss": 2.825188159942627,
      "step": 4200
    },
    {
      "epoch": 0.8103780864197531,
      "grad_norm": 1.0378340482711792,
      "learning_rate": 0.0004968983209464863,
      "loss": 2.3802804946899414,
      "step": 4201
    },
    {
      "epoch": 0.810570987654321,
      "grad_norm": 0.8478212356567383,
      "learning_rate": 0.0004967781378835959,
      "loss": 2.6276891231536865,
      "step": 4202
    },
    {
      "epoch": 0.8107638888888888,
      "grad_norm": 0.794671356678009,
      "learning_rate": 0.0004966556855485483,
      "loss": 2.383371353149414,
      "step": 4203
    },
    {
      "epoch": 0.8109567901234568,
      "grad_norm": 0.8016006946563721,
      "learning_rate": 0.0004965309650673657,
      "loss": 2.3414463996887207,
      "step": 4204
    },
    {
      "epoch": 0.8111496913580247,
      "grad_norm": 0.7997373938560486,
      "learning_rate": 0.0004964039775869272,
      "loss": 2.0966060161590576,
      "step": 4205
    },
    {
      "epoch": 0.8113425925925926,
      "grad_norm": 1.1240710020065308,
      "learning_rate": 0.0004962747242749584,
      "loss": 2.315340042114258,
      "step": 4206
    },
    {
      "epoch": 0.8115354938271605,
      "grad_norm": 1.0288053750991821,
      "learning_rate": 0.0004961432063200207,
      "loss": 2.5171918869018555,
      "step": 4207
    },
    {
      "epoch": 0.8117283950617284,
      "grad_norm": 1.2587889432907104,
      "learning_rate": 0.0004960094249315001,
      "loss": 2.4794795513153076,
      "step": 4208
    },
    {
      "epoch": 0.8119212962962963,
      "grad_norm": 1.023990511894226,
      "learning_rate": 0.0004958733813395963,
      "loss": 2.4167354106903076,
      "step": 4209
    },
    {
      "epoch": 0.8121141975308642,
      "grad_norm": 0.8727887272834778,
      "learning_rate": 0.0004957350767953112,
      "loss": 2.3994646072387695,
      "step": 4210
    },
    {
      "epoch": 0.8123070987654321,
      "grad_norm": 1.1372122764587402,
      "learning_rate": 0.0004955945125704375,
      "loss": 2.2652978897094727,
      "step": 4211
    },
    {
      "epoch": 0.8125,
      "grad_norm": 0.9538111090660095,
      "learning_rate": 0.0004954516899575473,
      "loss": 2.491138458251953,
      "step": 4212
    },
    {
      "epoch": 0.8126929012345679,
      "grad_norm": 0.7650263905525208,
      "learning_rate": 0.0004953066102699795,
      "loss": 2.3862736225128174,
      "step": 4213
    },
    {
      "epoch": 0.8128858024691358,
      "grad_norm": 0.8680559992790222,
      "learning_rate": 0.0004951592748418284,
      "loss": 2.3480443954467773,
      "step": 4214
    },
    {
      "epoch": 0.8130787037037037,
      "grad_norm": 1.0641392469406128,
      "learning_rate": 0.0004950096850279312,
      "loss": 2.8783421516418457,
      "step": 4215
    },
    {
      "epoch": 0.8132716049382716,
      "grad_norm": 1.072219967842102,
      "learning_rate": 0.0004948578422038555,
      "loss": 2.089139938354492,
      "step": 4216
    },
    {
      "epoch": 0.8134645061728395,
      "grad_norm": 0.9452730417251587,
      "learning_rate": 0.0004947037477658864,
      "loss": 2.166377067565918,
      "step": 4217
    },
    {
      "epoch": 0.8136574074074074,
      "grad_norm": 0.8841482996940613,
      "learning_rate": 0.0004945474031310144,
      "loss": 2.534984588623047,
      "step": 4218
    },
    {
      "epoch": 0.8138503086419753,
      "grad_norm": 0.8719977140426636,
      "learning_rate": 0.0004943888097369216,
      "loss": 2.2634456157684326,
      "step": 4219
    },
    {
      "epoch": 0.8140432098765432,
      "grad_norm": 0.8397873044013977,
      "learning_rate": 0.0004942279690419687,
      "loss": 2.699204921722412,
      "step": 4220
    },
    {
      "epoch": 0.8142361111111112,
      "grad_norm": 0.8029166460037231,
      "learning_rate": 0.000494064882525182,
      "loss": 2.447401285171509,
      "step": 4221
    },
    {
      "epoch": 0.814429012345679,
      "grad_norm": 0.9210023283958435,
      "learning_rate": 0.0004938995516862391,
      "loss": 2.9220333099365234,
      "step": 4222
    },
    {
      "epoch": 0.8146219135802469,
      "grad_norm": 1.1459969282150269,
      "learning_rate": 0.0004937319780454559,
      "loss": 2.5518925189971924,
      "step": 4223
    },
    {
      "epoch": 0.8148148148148148,
      "grad_norm": 0.6900899410247803,
      "learning_rate": 0.0004935621631437718,
      "loss": 2.2927608489990234,
      "step": 4224
    },
    {
      "epoch": 0.8150077160493827,
      "grad_norm": 0.8920720219612122,
      "learning_rate": 0.0004933901085427362,
      "loss": 2.065436601638794,
      "step": 4225
    },
    {
      "epoch": 0.8152006172839507,
      "grad_norm": 0.9293947219848633,
      "learning_rate": 0.0004932158158244937,
      "loss": 2.3401122093200684,
      "step": 4226
    },
    {
      "epoch": 0.8153935185185185,
      "grad_norm": 0.8968248963356018,
      "learning_rate": 0.0004930392865917698,
      "loss": 2.238194227218628,
      "step": 4227
    },
    {
      "epoch": 0.8155864197530864,
      "grad_norm": 0.6828739643096924,
      "learning_rate": 0.0004928605224678561,
      "loss": 2.3411688804626465,
      "step": 4228
    },
    {
      "epoch": 0.8157793209876543,
      "grad_norm": 0.8313254117965698,
      "learning_rate": 0.0004926795250965952,
      "loss": 2.075007438659668,
      "step": 4229
    },
    {
      "epoch": 0.8159722222222222,
      "grad_norm": 0.9889801144599915,
      "learning_rate": 0.0004924962961423661,
      "loss": 2.2761454582214355,
      "step": 4230
    },
    {
      "epoch": 0.8161651234567902,
      "grad_norm": 0.7488958239555359,
      "learning_rate": 0.0004923108372900683,
      "loss": 2.0857250690460205,
      "step": 4231
    },
    {
      "epoch": 0.816358024691358,
      "grad_norm": 1.0149108171463013,
      "learning_rate": 0.0004921231502451063,
      "loss": 2.5874414443969727,
      "step": 4232
    },
    {
      "epoch": 0.8165509259259259,
      "grad_norm": 0.911973774433136,
      "learning_rate": 0.0004919332367333747,
      "loss": 2.267219066619873,
      "step": 4233
    },
    {
      "epoch": 0.8167438271604939,
      "grad_norm": 1.2447113990783691,
      "learning_rate": 0.0004917410985012414,
      "loss": 2.3175368309020996,
      "step": 4234
    },
    {
      "epoch": 0.8169367283950617,
      "grad_norm": 1.215767741203308,
      "learning_rate": 0.000491546737315532,
      "loss": 2.33518385887146,
      "step": 4235
    },
    {
      "epoch": 0.8171296296296297,
      "grad_norm": 0.8064855933189392,
      "learning_rate": 0.0004913501549635136,
      "loss": 2.225925922393799,
      "step": 4236
    },
    {
      "epoch": 0.8173225308641975,
      "grad_norm": 0.9042306542396545,
      "learning_rate": 0.000491151353252878,
      "loss": 2.4323158264160156,
      "step": 4237
    },
    {
      "epoch": 0.8175154320987654,
      "grad_norm": 0.9042665958404541,
      "learning_rate": 0.0004909503340117256,
      "loss": 2.340540885925293,
      "step": 4238
    },
    {
      "epoch": 0.8177083333333334,
      "grad_norm": 0.7814400792121887,
      "learning_rate": 0.0004907470990885483,
      "loss": 2.474903106689453,
      "step": 4239
    },
    {
      "epoch": 0.8179012345679012,
      "grad_norm": 0.8088334202766418,
      "learning_rate": 0.0004905416503522123,
      "loss": 2.4025156497955322,
      "step": 4240
    },
    {
      "epoch": 0.8180941358024691,
      "grad_norm": 0.7813072800636292,
      "learning_rate": 0.0004903339896919414,
      "loss": 2.625217914581299,
      "step": 4241
    },
    {
      "epoch": 0.8182870370370371,
      "grad_norm": 0.7929980754852295,
      "learning_rate": 0.000490124119017299,
      "loss": 2.395585536956787,
      "step": 4242
    },
    {
      "epoch": 0.8184799382716049,
      "grad_norm": 0.8747406005859375,
      "learning_rate": 0.0004899120402581715,
      "loss": 2.346165180206299,
      "step": 4243
    },
    {
      "epoch": 0.8186728395061729,
      "grad_norm": 0.925147533416748,
      "learning_rate": 0.0004896977553647491,
      "loss": 2.2360410690307617,
      "step": 4244
    },
    {
      "epoch": 0.8188657407407407,
      "grad_norm": 1.0878984928131104,
      "learning_rate": 0.0004894812663075095,
      "loss": 2.209333896636963,
      "step": 4245
    },
    {
      "epoch": 0.8190586419753086,
      "grad_norm": 1.08043372631073,
      "learning_rate": 0.0004892625750771986,
      "loss": 2.45219087600708,
      "step": 4246
    },
    {
      "epoch": 0.8192515432098766,
      "grad_norm": 0.885403037071228,
      "learning_rate": 0.0004890416836848127,
      "loss": 2.304403305053711,
      "step": 4247
    },
    {
      "epoch": 0.8194444444444444,
      "grad_norm": 1.2689377069473267,
      "learning_rate": 0.00048881859416158,
      "loss": 2.330094814300537,
      "step": 4248
    },
    {
      "epoch": 0.8196373456790124,
      "grad_norm": 1.0296061038970947,
      "learning_rate": 0.0004885933085589416,
      "loss": 2.3945891857147217,
      "step": 4249
    },
    {
      "epoch": 0.8198302469135802,
      "grad_norm": 0.6997410655021667,
      "learning_rate": 0.0004883658289485329,
      "loss": 2.1723451614379883,
      "step": 4250
    },
    {
      "epoch": 0.8200231481481481,
      "grad_norm": 0.8651866912841797,
      "learning_rate": 0.0004881361574221648,
      "loss": 2.3084216117858887,
      "step": 4251
    },
    {
      "epoch": 0.8202160493827161,
      "grad_norm": 0.6476576924324036,
      "learning_rate": 0.00048790429609180374,
      "loss": 2.2792577743530273,
      "step": 4252
    },
    {
      "epoch": 0.8204089506172839,
      "grad_norm": 0.724915087223053,
      "learning_rate": 0.0004876702470895531,
      "loss": 2.386577606201172,
      "step": 4253
    },
    {
      "epoch": 0.8206018518518519,
      "grad_norm": 1.2437185049057007,
      "learning_rate": 0.000487434012567633,
      "loss": 2.3140993118286133,
      "step": 4254
    },
    {
      "epoch": 0.8207947530864198,
      "grad_norm": 0.8570701479911804,
      "learning_rate": 0.0004871955946983607,
      "loss": 2.460369110107422,
      "step": 4255
    },
    {
      "epoch": 0.8209876543209876,
      "grad_norm": 0.6806497573852539,
      "learning_rate": 0.0004869549956741308,
      "loss": 2.180612802505493,
      "step": 4256
    },
    {
      "epoch": 0.8211805555555556,
      "grad_norm": 0.9256541132926941,
      "learning_rate": 0.00048671221770739474,
      "loss": 2.1317343711853027,
      "step": 4257
    },
    {
      "epoch": 0.8213734567901234,
      "grad_norm": 0.7595787048339844,
      "learning_rate": 0.0004864672630306408,
      "loss": 2.4726147651672363,
      "step": 4258
    },
    {
      "epoch": 0.8215663580246914,
      "grad_norm": 0.9882955551147461,
      "learning_rate": 0.00048622013389637325,
      "loss": 2.4139370918273926,
      "step": 4259
    },
    {
      "epoch": 0.8217592592592593,
      "grad_norm": 0.9208806157112122,
      "learning_rate": 0.0004859708325770919,
      "loss": 1.8385791778564453,
      "step": 4260
    },
    {
      "epoch": 0.8219521604938271,
      "grad_norm": 0.8763349056243896,
      "learning_rate": 0.00048571936136527106,
      "loss": 2.445964813232422,
      "step": 4261
    },
    {
      "epoch": 0.8221450617283951,
      "grad_norm": 0.9006690382957458,
      "learning_rate": 0.0004854657225733385,
      "loss": 2.301757574081421,
      "step": 4262
    },
    {
      "epoch": 0.8223379629629629,
      "grad_norm": 0.8023741245269775,
      "learning_rate": 0.00048520991853365415,
      "loss": 2.2783937454223633,
      "step": 4263
    },
    {
      "epoch": 0.8225308641975309,
      "grad_norm": 0.9987475275993347,
      "learning_rate": 0.0004849519515984888,
      "loss": 2.270953416824341,
      "step": 4264
    },
    {
      "epoch": 0.8227237654320988,
      "grad_norm": 1.2020727396011353,
      "learning_rate": 0.00048469182414000216,
      "loss": 2.319864273071289,
      "step": 4265
    },
    {
      "epoch": 0.8229166666666666,
      "grad_norm": 1.3491734266281128,
      "learning_rate": 0.0004844295385502215,
      "loss": 2.555476188659668,
      "step": 4266
    },
    {
      "epoch": 0.8231095679012346,
      "grad_norm": 1.0264090299606323,
      "learning_rate": 0.0004841650972410191,
      "loss": 2.564236640930176,
      "step": 4267
    },
    {
      "epoch": 0.8233024691358025,
      "grad_norm": 0.7996337413787842,
      "learning_rate": 0.00048389850264409054,
      "loss": 2.3640198707580566,
      "step": 4268
    },
    {
      "epoch": 0.8234953703703703,
      "grad_norm": 0.7462021708488464,
      "learning_rate": 0.00048362975721093226,
      "loss": 2.209500789642334,
      "step": 4269
    },
    {
      "epoch": 0.8236882716049383,
      "grad_norm": 1.0072015523910522,
      "learning_rate": 0.00048335886341281865,
      "loss": 2.4763379096984863,
      "step": 4270
    },
    {
      "epoch": 0.8238811728395061,
      "grad_norm": 1.1084001064300537,
      "learning_rate": 0.0004830858237407798,
      "loss": 2.29524302482605,
      "step": 4271
    },
    {
      "epoch": 0.8240740740740741,
      "grad_norm": 1.0697689056396484,
      "learning_rate": 0.00048281064070557835,
      "loss": 2.2846555709838867,
      "step": 4272
    },
    {
      "epoch": 0.824266975308642,
      "grad_norm": 1.0973241329193115,
      "learning_rate": 0.0004825333168376864,
      "loss": 2.3901140689849854,
      "step": 4273
    },
    {
      "epoch": 0.8244598765432098,
      "grad_norm": 0.9868088364601135,
      "learning_rate": 0.00048225385468726235,
      "loss": 2.099226474761963,
      "step": 4274
    },
    {
      "epoch": 0.8246527777777778,
      "grad_norm": 1.0791308879852295,
      "learning_rate": 0.0004819722568241274,
      "loss": 2.5572922229766846,
      "step": 4275
    },
    {
      "epoch": 0.8248456790123457,
      "grad_norm": 1.6777342557907104,
      "learning_rate": 0.00048168852583774165,
      "loss": 2.494964361190796,
      "step": 4276
    },
    {
      "epoch": 0.8250385802469136,
      "grad_norm": 0.8129950165748596,
      "learning_rate": 0.00048140266433718094,
      "loss": 2.416262626647949,
      "step": 4277
    },
    {
      "epoch": 0.8252314814814815,
      "grad_norm": 1.2756813764572144,
      "learning_rate": 0.00048111467495111213,
      "loss": 2.4583756923675537,
      "step": 4278
    },
    {
      "epoch": 0.8254243827160493,
      "grad_norm": 1.1049686670303345,
      "learning_rate": 0.00048082456032776946,
      "loss": 2.1800475120544434,
      "step": 4279
    },
    {
      "epoch": 0.8256172839506173,
      "grad_norm": 0.9218963384628296,
      "learning_rate": 0.0004805323231349299,
      "loss": 2.2758126258850098,
      "step": 4280
    },
    {
      "epoch": 0.8258101851851852,
      "grad_norm": 0.7738329768180847,
      "learning_rate": 0.00048023796605988866,
      "loss": 2.4690961837768555,
      "step": 4281
    },
    {
      "epoch": 0.8260030864197531,
      "grad_norm": 0.8607953190803528,
      "learning_rate": 0.0004799414918094347,
      "loss": 2.575080633163452,
      "step": 4282
    },
    {
      "epoch": 0.826195987654321,
      "grad_norm": 1.113339900970459,
      "learning_rate": 0.0004796429031098255,
      "loss": 2.413130760192871,
      "step": 4283
    },
    {
      "epoch": 0.8263888888888888,
      "grad_norm": 1.486185073852539,
      "learning_rate": 0.0004793422027067623,
      "loss": 2.2649190425872803,
      "step": 4284
    },
    {
      "epoch": 0.8265817901234568,
      "grad_norm": 1.506072759628296,
      "learning_rate": 0.0004790393933653646,
      "loss": 2.0764036178588867,
      "step": 4285
    },
    {
      "epoch": 0.8267746913580247,
      "grad_norm": 0.9712194800376892,
      "learning_rate": 0.0004787344778701449,
      "loss": 2.170299530029297,
      "step": 4286
    },
    {
      "epoch": 0.8269675925925926,
      "grad_norm": 0.8051685690879822,
      "learning_rate": 0.00047842745902498303,
      "loss": 2.1737351417541504,
      "step": 4287
    },
    {
      "epoch": 0.8271604938271605,
      "grad_norm": 0.9015460014343262,
      "learning_rate": 0.0004781183396531004,
      "loss": 2.7120981216430664,
      "step": 4288
    },
    {
      "epoch": 0.8273533950617284,
      "grad_norm": 0.870477557182312,
      "learning_rate": 0.00047780712259703394,
      "loss": 2.1973814964294434,
      "step": 4289
    },
    {
      "epoch": 0.8275462962962963,
      "grad_norm": 0.7461525201797485,
      "learning_rate": 0.00047749381071861014,
      "loss": 2.4445419311523438,
      "step": 4290
    },
    {
      "epoch": 0.8277391975308642,
      "grad_norm": 0.7609556913375854,
      "learning_rate": 0.00047717840689891856,
      "loss": 2.4934184551239014,
      "step": 4291
    },
    {
      "epoch": 0.8279320987654321,
      "grad_norm": 0.8560394644737244,
      "learning_rate": 0.00047686091403828545,
      "loss": 2.380847930908203,
      "step": 4292
    },
    {
      "epoch": 0.828125,
      "grad_norm": 0.8651431202888489,
      "learning_rate": 0.000476541335056247,
      "loss": 2.1616997718811035,
      "step": 4293
    },
    {
      "epoch": 0.8283179012345679,
      "grad_norm": 1.0123590230941772,
      "learning_rate": 0.0004762196728915225,
      "loss": 2.4161577224731445,
      "step": 4294
    },
    {
      "epoch": 0.8285108024691358,
      "grad_norm": 0.8293865323066711,
      "learning_rate": 0.0004758959305019875,
      "loss": 2.4387104511260986,
      "step": 4295
    },
    {
      "epoch": 0.8287037037037037,
      "grad_norm": 0.8286476731300354,
      "learning_rate": 0.0004755701108646463,
      "loss": 2.434634208679199,
      "step": 4296
    },
    {
      "epoch": 0.8288966049382716,
      "grad_norm": 0.7522587180137634,
      "learning_rate": 0.00047524221697560476,
      "loss": 2.3442564010620117,
      "step": 4297
    },
    {
      "epoch": 0.8290895061728395,
      "grad_norm": 0.7827571034431458,
      "learning_rate": 0.00047491225185004285,
      "loss": 2.4657444953918457,
      "step": 4298
    },
    {
      "epoch": 0.8292824074074074,
      "grad_norm": 1.4906474351882935,
      "learning_rate": 0.00047458021852218657,
      "loss": 2.1987087726593018,
      "step": 4299
    },
    {
      "epoch": 0.8294753086419753,
      "grad_norm": 1.065019965171814,
      "learning_rate": 0.0004742461200452804,
      "loss": 2.584010124206543,
      "step": 4300
    },
    {
      "epoch": 0.8296682098765432,
      "grad_norm": 0.6758496165275574,
      "learning_rate": 0.00047390995949155913,
      "loss": 2.1995630264282227,
      "step": 4301
    },
    {
      "epoch": 0.8298611111111112,
      "grad_norm": 0.8581151366233826,
      "learning_rate": 0.00047357173995221956,
      "loss": 2.4383645057678223,
      "step": 4302
    },
    {
      "epoch": 0.830054012345679,
      "grad_norm": 0.8897155523300171,
      "learning_rate": 0.0004732314645373921,
      "loss": 2.2700586318969727,
      "step": 4303
    },
    {
      "epoch": 0.8302469135802469,
      "grad_norm": 0.919867753982544,
      "learning_rate": 0.0004728891363761121,
      "loss": 2.5034732818603516,
      "step": 4304
    },
    {
      "epoch": 0.8304398148148148,
      "grad_norm": 0.9262894988059998,
      "learning_rate": 0.0004725447586162911,
      "loss": 2.2672488689422607,
      "step": 4305
    },
    {
      "epoch": 0.8306327160493827,
      "grad_norm": 0.7599213123321533,
      "learning_rate": 0.0004721983344246881,
      "loss": 2.230861186981201,
      "step": 4306
    },
    {
      "epoch": 0.8308256172839507,
      "grad_norm": 1.2726150751113892,
      "learning_rate": 0.00047184986698688,
      "loss": 2.484114408493042,
      "step": 4307
    },
    {
      "epoch": 0.8310185185185185,
      "grad_norm": 0.816229522228241,
      "learning_rate": 0.0004714993595072329,
      "loss": 2.174160957336426,
      "step": 4308
    },
    {
      "epoch": 0.8312114197530864,
      "grad_norm": 0.8093381524085999,
      "learning_rate": 0.0004711468152088719,
      "loss": 2.2249059677124023,
      "step": 4309
    },
    {
      "epoch": 0.8314043209876543,
      "grad_norm": 0.8408767580986023,
      "learning_rate": 0.00047079223733365234,
      "loss": 2.268172264099121,
      "step": 4310
    },
    {
      "epoch": 0.8315972222222222,
      "grad_norm": 0.9854335784912109,
      "learning_rate": 0.00047043562914212913,
      "loss": 2.389192581176758,
      "step": 4311
    },
    {
      "epoch": 0.8317901234567902,
      "grad_norm": 0.8845558166503906,
      "learning_rate": 0.0004700769939135272,
      "loss": 2.3946447372436523,
      "step": 4312
    },
    {
      "epoch": 0.831983024691358,
      "grad_norm": 0.7163538336753845,
      "learning_rate": 0.0004697163349457114,
      "loss": 2.208831787109375,
      "step": 4313
    },
    {
      "epoch": 0.8321759259259259,
      "grad_norm": 0.930349588394165,
      "learning_rate": 0.0004693536555551562,
      "loss": 2.2043533325195312,
      "step": 4314
    },
    {
      "epoch": 0.8323688271604939,
      "grad_norm": 1.111600399017334,
      "learning_rate": 0.00046898895907691466,
      "loss": 2.3075544834136963,
      "step": 4315
    },
    {
      "epoch": 0.8325617283950617,
      "grad_norm": 0.8377475738525391,
      "learning_rate": 0.0004686222488645886,
      "loss": 2.256256580352783,
      "step": 4316
    },
    {
      "epoch": 0.8327546296296297,
      "grad_norm": 0.9551502466201782,
      "learning_rate": 0.00046825352829029705,
      "loss": 2.3640055656433105,
      "step": 4317
    },
    {
      "epoch": 0.8329475308641975,
      "grad_norm": 1.5507044792175293,
      "learning_rate": 0.0004678828007446456,
      "loss": 2.058969497680664,
      "step": 4318
    },
    {
      "epoch": 0.8331404320987654,
      "grad_norm": 0.9229080677032471,
      "learning_rate": 0.0004675100696366951,
      "loss": 2.590996503829956,
      "step": 4319
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.7178419232368469,
      "learning_rate": 0.0004671353383939304,
      "loss": 2.5995490550994873,
      "step": 4320
    },
    {
      "epoch": 0.8335262345679012,
      "grad_norm": 0.8782360553741455,
      "learning_rate": 0.0004667586104622288,
      "loss": 2.313333511352539,
      "step": 4321
    },
    {
      "epoch": 0.8337191358024691,
      "grad_norm": 0.761301577091217,
      "learning_rate": 0.0004663798893058283,
      "loss": 2.2590396404266357,
      "step": 4322
    },
    {
      "epoch": 0.8339120370370371,
      "grad_norm": 1.0029218196868896,
      "learning_rate": 0.00046599917840729586,
      "loss": 2.4435853958129883,
      "step": 4323
    },
    {
      "epoch": 0.8341049382716049,
      "grad_norm": 0.7815080881118774,
      "learning_rate": 0.0004656164812674951,
      "loss": 2.469480037689209,
      "step": 4324
    },
    {
      "epoch": 0.8342978395061729,
      "grad_norm": 1.0344966650009155,
      "learning_rate": 0.00046523180140555463,
      "loss": 2.441652774810791,
      "step": 4325
    },
    {
      "epoch": 0.8344907407407407,
      "grad_norm": 0.9073982238769531,
      "learning_rate": 0.00046484514235883514,
      "loss": 2.183328866958618,
      "step": 4326
    },
    {
      "epoch": 0.8346836419753086,
      "grad_norm": 0.9534565210342407,
      "learning_rate": 0.00046445650768289716,
      "loss": 2.330497980117798,
      "step": 4327
    },
    {
      "epoch": 0.8348765432098766,
      "grad_norm": 0.80821692943573,
      "learning_rate": 0.00046406590095146827,
      "loss": 2.422351598739624,
      "step": 4328
    },
    {
      "epoch": 0.8350694444444444,
      "grad_norm": 0.9488611221313477,
      "learning_rate": 0.0004636733257564104,
      "loss": 2.2725329399108887,
      "step": 4329
    },
    {
      "epoch": 0.8352623456790124,
      "grad_norm": 0.9683899283409119,
      "learning_rate": 0.00046327878570768654,
      "loss": 2.326889753341675,
      "step": 4330
    },
    {
      "epoch": 0.8354552469135802,
      "grad_norm": 3.4218807220458984,
      "learning_rate": 0.0004628822844333278,
      "loss": 2.2578463554382324,
      "step": 4331
    },
    {
      "epoch": 0.8356481481481481,
      "grad_norm": 0.9912207126617432,
      "learning_rate": 0.00046248382557939995,
      "loss": 2.0256786346435547,
      "step": 4332
    },
    {
      "epoch": 0.8358410493827161,
      "grad_norm": 0.7634944915771484,
      "learning_rate": 0.0004620834128099696,
      "loss": 2.25868558883667,
      "step": 4333
    },
    {
      "epoch": 0.8360339506172839,
      "grad_norm": 0.8015978932380676,
      "learning_rate": 0.00046168104980707104,
      "loss": 2.3756752014160156,
      "step": 4334
    },
    {
      "epoch": 0.8362268518518519,
      "grad_norm": 0.7063593864440918,
      "learning_rate": 0.0004612767402706721,
      "loss": 2.138484001159668,
      "step": 4335
    },
    {
      "epoch": 0.8364197530864198,
      "grad_norm": 1.0130459070205688,
      "learning_rate": 0.00046087048791864017,
      "loss": 2.1450142860412598,
      "step": 4336
    },
    {
      "epoch": 0.8366126543209876,
      "grad_norm": 0.8485569953918457,
      "learning_rate": 0.00046046229648670777,
      "loss": 2.2640936374664307,
      "step": 4337
    },
    {
      "epoch": 0.8368055555555556,
      "grad_norm": 0.9952423572540283,
      "learning_rate": 0.0004600521697284386,
      "loss": 2.4039249420166016,
      "step": 4338
    },
    {
      "epoch": 0.8369984567901234,
      "grad_norm": 0.7892723679542542,
      "learning_rate": 0.00045964011141519286,
      "loss": 2.347684383392334,
      "step": 4339
    },
    {
      "epoch": 0.8371913580246914,
      "grad_norm": 0.7546009421348572,
      "learning_rate": 0.00045922612533609244,
      "loss": 2.50455904006958,
      "step": 4340
    },
    {
      "epoch": 0.8373842592592593,
      "grad_norm": 0.7572872638702393,
      "learning_rate": 0.0004588102152979863,
      "loss": 2.3354105949401855,
      "step": 4341
    },
    {
      "epoch": 0.8375771604938271,
      "grad_norm": 0.7384973168373108,
      "learning_rate": 0.00045839238512541513,
      "loss": 2.2042348384857178,
      "step": 4342
    },
    {
      "epoch": 0.8377700617283951,
      "grad_norm": 0.8486059308052063,
      "learning_rate": 0.0004579726386605768,
      "loss": 2.153439998626709,
      "step": 4343
    },
    {
      "epoch": 0.8379629629629629,
      "grad_norm": 0.8325803279876709,
      "learning_rate": 0.00045755097976329023,
      "loss": 2.3049049377441406,
      "step": 4344
    },
    {
      "epoch": 0.8381558641975309,
      "grad_norm": 0.8914577960968018,
      "learning_rate": 0.00045712741231096054,
      "loss": 2.4953055381774902,
      "step": 4345
    },
    {
      "epoch": 0.8383487654320988,
      "grad_norm": 1.156368613243103,
      "learning_rate": 0.00045670194019854305,
      "loss": 2.21053409576416,
      "step": 4346
    },
    {
      "epoch": 0.8385416666666666,
      "grad_norm": 0.9161633849143982,
      "learning_rate": 0.00045627456733850745,
      "loss": 2.3823351860046387,
      "step": 4347
    },
    {
      "epoch": 0.8387345679012346,
      "grad_norm": 1.8025630712509155,
      "learning_rate": 0.00045584529766080233,
      "loss": 2.6123909950256348,
      "step": 4348
    },
    {
      "epoch": 0.8389274691358025,
      "grad_norm": 0.8952727317810059,
      "learning_rate": 0.00045541413511281825,
      "loss": 2.36330246925354,
      "step": 4349
    },
    {
      "epoch": 0.8391203703703703,
      "grad_norm": 0.8368254899978638,
      "learning_rate": 0.000454981083659352,
      "loss": 2.1939268112182617,
      "step": 4350
    },
    {
      "epoch": 0.8393132716049383,
      "grad_norm": 0.7888182401657104,
      "learning_rate": 0.00045454614728256994,
      "loss": 1.8336836099624634,
      "step": 4351
    },
    {
      "epoch": 0.8395061728395061,
      "grad_norm": 0.7736703753471375,
      "learning_rate": 0.0004541093299819714,
      "loss": 2.445899248123169,
      "step": 4352
    },
    {
      "epoch": 0.8396990740740741,
      "grad_norm": 0.8449828624725342,
      "learning_rate": 0.0004536706357743522,
      "loss": 2.5177361965179443,
      "step": 4353
    },
    {
      "epoch": 0.839891975308642,
      "grad_norm": 0.8112662434577942,
      "learning_rate": 0.000453230068693767,
      "loss": 2.2846121788024902,
      "step": 4354
    },
    {
      "epoch": 0.8400848765432098,
      "grad_norm": 0.7979561686515808,
      "learning_rate": 0.00045278763279149294,
      "loss": 1.9606810808181763,
      "step": 4355
    },
    {
      "epoch": 0.8402777777777778,
      "grad_norm": 0.753654956817627,
      "learning_rate": 0.000452343332135992,
      "loss": 2.414936065673828,
      "step": 4356
    },
    {
      "epoch": 0.8404706790123457,
      "grad_norm": 0.8303817510604858,
      "learning_rate": 0.00045189717081287376,
      "loss": 2.104027509689331,
      "step": 4357
    },
    {
      "epoch": 0.8406635802469136,
      "grad_norm": 0.7519192695617676,
      "learning_rate": 0.00045144915292485754,
      "loss": 2.2604708671569824,
      "step": 4358
    },
    {
      "epoch": 0.8408564814814815,
      "grad_norm": 2.8208436965942383,
      "learning_rate": 0.00045099928259173516,
      "loss": 1.9387545585632324,
      "step": 4359
    },
    {
      "epoch": 0.8410493827160493,
      "grad_norm": 0.8084567785263062,
      "learning_rate": 0.0004505475639503326,
      "loss": 2.4174065589904785,
      "step": 4360
    },
    {
      "epoch": 0.8412422839506173,
      "grad_norm": 0.9199321269989014,
      "learning_rate": 0.0004500940011544721,
      "loss": 2.287814140319824,
      "step": 4361
    },
    {
      "epoch": 0.8414351851851852,
      "grad_norm": 0.8095868229866028,
      "learning_rate": 0.00044963859837493405,
      "loss": 2.446085214614868,
      "step": 4362
    },
    {
      "epoch": 0.8416280864197531,
      "grad_norm": 1.0327123403549194,
      "learning_rate": 0.00044918135979941877,
      "loss": 2.3823013305664062,
      "step": 4363
    },
    {
      "epoch": 0.841820987654321,
      "grad_norm": 0.9072319865226746,
      "learning_rate": 0.0004487222896325076,
      "loss": 2.519277334213257,
      "step": 4364
    },
    {
      "epoch": 0.8420138888888888,
      "grad_norm": 0.9319941401481628,
      "learning_rate": 0.00044826139209562445,
      "loss": 2.184899091720581,
      "step": 4365
    },
    {
      "epoch": 0.8422067901234568,
      "grad_norm": 0.9038587808609009,
      "learning_rate": 0.00044779867142699713,
      "loss": 2.4060263633728027,
      "step": 4366
    },
    {
      "epoch": 0.8423996913580247,
      "grad_norm": 1.0412440299987793,
      "learning_rate": 0.00044733413188161806,
      "loss": 1.9912890195846558,
      "step": 4367
    },
    {
      "epoch": 0.8425925925925926,
      "grad_norm": 0.9273769855499268,
      "learning_rate": 0.0004468677777312056,
      "loss": 2.201397180557251,
      "step": 4368
    },
    {
      "epoch": 0.8427854938271605,
      "grad_norm": 0.8106664419174194,
      "learning_rate": 0.00044639961326416405,
      "loss": 2.5373454093933105,
      "step": 4369
    },
    {
      "epoch": 0.8429783950617284,
      "grad_norm": 1.3007973432540894,
      "learning_rate": 0.000445929642785545,
      "loss": 2.4313712120056152,
      "step": 4370
    },
    {
      "epoch": 0.8431712962962963,
      "grad_norm": 1.1435043811798096,
      "learning_rate": 0.0004454578706170075,
      "loss": 2.267951011657715,
      "step": 4371
    },
    {
      "epoch": 0.8433641975308642,
      "grad_norm": 0.9964225888252258,
      "learning_rate": 0.00044498430109677766,
      "loss": 2.5674405097961426,
      "step": 4372
    },
    {
      "epoch": 0.8435570987654321,
      "grad_norm": 0.7374821901321411,
      "learning_rate": 0.0004445089385796099,
      "loss": 2.335073947906494,
      "step": 4373
    },
    {
      "epoch": 0.84375,
      "grad_norm": 1.0108296871185303,
      "learning_rate": 0.0004440317874367459,
      "loss": 2.121734857559204,
      "step": 4374
    },
    {
      "epoch": 0.8439429012345679,
      "grad_norm": 0.7862834334373474,
      "learning_rate": 0.0004435528520558752,
      "loss": 2.554532527923584,
      "step": 4375
    },
    {
      "epoch": 0.8441358024691358,
      "grad_norm": 0.9278991222381592,
      "learning_rate": 0.0004430721368410941,
      "loss": 2.5028936862945557,
      "step": 4376
    },
    {
      "epoch": 0.8443287037037037,
      "grad_norm": 0.9343278408050537,
      "learning_rate": 0.0004425896462128658,
      "loss": 2.0271501541137695,
      "step": 4377
    },
    {
      "epoch": 0.8445216049382716,
      "grad_norm": 1.0209987163543701,
      "learning_rate": 0.0004421053846079793,
      "loss": 2.6587820053100586,
      "step": 4378
    },
    {
      "epoch": 0.8447145061728395,
      "grad_norm": 0.7824397087097168,
      "learning_rate": 0.0004416193564795089,
      "loss": 2.409214973449707,
      "step": 4379
    },
    {
      "epoch": 0.8449074074074074,
      "grad_norm": 0.8913047313690186,
      "learning_rate": 0.00044113156629677313,
      "loss": 2.300294876098633,
      "step": 4380
    },
    {
      "epoch": 0.8451003086419753,
      "grad_norm": 0.8928208351135254,
      "learning_rate": 0.0004406420185452936,
      "loss": 2.258150339126587,
      "step": 4381
    },
    {
      "epoch": 0.8452932098765432,
      "grad_norm": 1.0385664701461792,
      "learning_rate": 0.00044015071772675394,
      "loss": 2.178175210952759,
      "step": 4382
    },
    {
      "epoch": 0.8454861111111112,
      "grad_norm": 0.9149986505508423,
      "learning_rate": 0.00043965766835895817,
      "loss": 2.4658827781677246,
      "step": 4383
    },
    {
      "epoch": 0.845679012345679,
      "grad_norm": 0.7786886692047119,
      "learning_rate": 0.0004391628749757892,
      "loss": 2.381669521331787,
      "step": 4384
    },
    {
      "epoch": 0.8458719135802469,
      "grad_norm": 0.9823364615440369,
      "learning_rate": 0.0004386663421271674,
      "loss": 2.590035915374756,
      "step": 4385
    },
    {
      "epoch": 0.8460648148148148,
      "grad_norm": 0.7667989134788513,
      "learning_rate": 0.0004381680743790083,
      "loss": 2.54945707321167,
      "step": 4386
    },
    {
      "epoch": 0.8462577160493827,
      "grad_norm": 0.7890955209732056,
      "learning_rate": 0.0004376680763131811,
      "loss": 2.542647361755371,
      "step": 4387
    },
    {
      "epoch": 0.8464506172839507,
      "grad_norm": 1.0327728986740112,
      "learning_rate": 0.0004371663525274662,
      "loss": 2.2442548274993896,
      "step": 4388
    },
    {
      "epoch": 0.8466435185185185,
      "grad_norm": 0.7429035902023315,
      "learning_rate": 0.00043666290763551307,
      "loss": 2.1269373893737793,
      "step": 4389
    },
    {
      "epoch": 0.8468364197530864,
      "grad_norm": 0.8186244964599609,
      "learning_rate": 0.00043615774626679777,
      "loss": 2.2982912063598633,
      "step": 4390
    },
    {
      "epoch": 0.8470293209876543,
      "grad_norm": 0.9581624269485474,
      "learning_rate": 0.00043565087306658034,
      "loss": 2.4220595359802246,
      "step": 4391
    },
    {
      "epoch": 0.8472222222222222,
      "grad_norm": 0.8161712884902954,
      "learning_rate": 0.0004351422926958621,
      "loss": 2.1830782890319824,
      "step": 4392
    },
    {
      "epoch": 0.8474151234567902,
      "grad_norm": 0.7579860091209412,
      "learning_rate": 0.0004346320098313429,
      "loss": 2.1047346591949463,
      "step": 4393
    },
    {
      "epoch": 0.847608024691358,
      "grad_norm": 0.7520943880081177,
      "learning_rate": 0.0004341200291653781,
      "loss": 2.4079995155334473,
      "step": 4394
    },
    {
      "epoch": 0.8478009259259259,
      "grad_norm": 0.7324812412261963,
      "learning_rate": 0.0004336063554059351,
      "loss": 2.1350696086883545,
      "step": 4395
    },
    {
      "epoch": 0.8479938271604939,
      "grad_norm": 0.844523012638092,
      "learning_rate": 0.00043309099327655065,
      "loss": 2.3079891204833984,
      "step": 4396
    },
    {
      "epoch": 0.8481867283950617,
      "grad_norm": 1.0002851486206055,
      "learning_rate": 0.00043257394751628685,
      "loss": 2.336334705352783,
      "step": 4397
    },
    {
      "epoch": 0.8483796296296297,
      "grad_norm": 0.8425604104995728,
      "learning_rate": 0.0004320552228796878,
      "loss": 2.183295249938965,
      "step": 4398
    },
    {
      "epoch": 0.8485725308641975,
      "grad_norm": 0.8839731812477112,
      "learning_rate": 0.0004315348241367358,
      "loss": 2.394740581512451,
      "step": 4399
    },
    {
      "epoch": 0.8487654320987654,
      "grad_norm": 0.9027680158615112,
      "learning_rate": 0.00043101275607280786,
      "loss": 2.5099375247955322,
      "step": 4400
    },
    {
      "epoch": 0.8489583333333334,
      "grad_norm": 0.7222859859466553,
      "learning_rate": 0.0004304890234886311,
      "loss": 1.9138238430023193,
      "step": 4401
    },
    {
      "epoch": 0.8491512345679012,
      "grad_norm": 0.9081523418426514,
      "learning_rate": 0.000429963631200239,
      "loss": 2.343236207962036,
      "step": 4402
    },
    {
      "epoch": 0.8493441358024691,
      "grad_norm": 0.8543051481246948,
      "learning_rate": 0.000429436584038927,
      "loss": 2.279109239578247,
      "step": 4403
    },
    {
      "epoch": 0.8495370370370371,
      "grad_norm": 1.3760536909103394,
      "learning_rate": 0.000428907886851208,
      "loss": 2.394624710083008,
      "step": 4404
    },
    {
      "epoch": 0.8497299382716049,
      "grad_norm": 0.783033013343811,
      "learning_rate": 0.0004283775444987681,
      "loss": 2.2233355045318604,
      "step": 4405
    },
    {
      "epoch": 0.8499228395061729,
      "grad_norm": 0.8479796051979065,
      "learning_rate": 0.0004278455618584214,
      "loss": 2.1030924320220947,
      "step": 4406
    },
    {
      "epoch": 0.8501157407407407,
      "grad_norm": 0.8575901389122009,
      "learning_rate": 0.00042731194382206563,
      "loss": 2.2458128929138184,
      "step": 4407
    },
    {
      "epoch": 0.8503086419753086,
      "grad_norm": 0.7788930535316467,
      "learning_rate": 0.00042677669529663686,
      "loss": 2.2067131996154785,
      "step": 4408
    },
    {
      "epoch": 0.8505015432098766,
      "grad_norm": 0.8605113625526428,
      "learning_rate": 0.00042623982120406457,
      "loss": 2.2365241050720215,
      "step": 4409
    },
    {
      "epoch": 0.8506944444444444,
      "grad_norm": 0.8142244219779968,
      "learning_rate": 0.0004257013264812261,
      "loss": 2.469372272491455,
      "step": 4410
    },
    {
      "epoch": 0.8508873456790124,
      "grad_norm": 0.835762619972229,
      "learning_rate": 0.0004251612160799017,
      "loss": 2.144325017929077,
      "step": 4411
    },
    {
      "epoch": 0.8510802469135802,
      "grad_norm": 0.9783709049224854,
      "learning_rate": 0.00042461949496672857,
      "loss": 2.147855758666992,
      "step": 4412
    },
    {
      "epoch": 0.8512731481481481,
      "grad_norm": 1.0351723432540894,
      "learning_rate": 0.00042407616812315554,
      "loss": 1.92112135887146,
      "step": 4413
    },
    {
      "epoch": 0.8514660493827161,
      "grad_norm": 0.8632723689079285,
      "learning_rate": 0.0004235312405453968,
      "loss": 2.145927906036377,
      "step": 4414
    },
    {
      "epoch": 0.8516589506172839,
      "grad_norm": 0.8559569716453552,
      "learning_rate": 0.0004229847172443866,
      "loss": 2.316648006439209,
      "step": 4415
    },
    {
      "epoch": 0.8518518518518519,
      "grad_norm": 1.0460305213928223,
      "learning_rate": 0.00042243660324573255,
      "loss": 2.5231409072875977,
      "step": 4416
    },
    {
      "epoch": 0.8520447530864198,
      "grad_norm": 0.9026861190795898,
      "learning_rate": 0.0004218869035896697,
      "loss": 2.5163731575012207,
      "step": 4417
    },
    {
      "epoch": 0.8522376543209876,
      "grad_norm": 0.7110809683799744,
      "learning_rate": 0.00042133562333101436,
      "loss": 2.452981948852539,
      "step": 4418
    },
    {
      "epoch": 0.8524305555555556,
      "grad_norm": 0.9129582047462463,
      "learning_rate": 0.00042078276753911725,
      "loss": 2.602818250656128,
      "step": 4419
    },
    {
      "epoch": 0.8526234567901234,
      "grad_norm": 0.782861053943634,
      "learning_rate": 0.0004202283412978171,
      "loss": 2.1811695098876953,
      "step": 4420
    },
    {
      "epoch": 0.8528163580246914,
      "grad_norm": 0.7222602963447571,
      "learning_rate": 0.0004196723497053938,
      "loss": 2.495913505554199,
      "step": 4421
    },
    {
      "epoch": 0.8530092592592593,
      "grad_norm": 0.7599205374717712,
      "learning_rate": 0.00041911479787452177,
      "loss": 2.345876693725586,
      "step": 4422
    },
    {
      "epoch": 0.8532021604938271,
      "grad_norm": 1.3537849187850952,
      "learning_rate": 0.00041855569093222244,
      "loss": 2.0879175662994385,
      "step": 4423
    },
    {
      "epoch": 0.8533950617283951,
      "grad_norm": 0.7820873856544495,
      "learning_rate": 0.00041799503401981783,
      "loss": 2.245609760284424,
      "step": 4424
    },
    {
      "epoch": 0.8535879629629629,
      "grad_norm": 0.8592017889022827,
      "learning_rate": 0.0004174328322928824,
      "loss": 2.4816598892211914,
      "step": 4425
    },
    {
      "epoch": 0.8537808641975309,
      "grad_norm": 0.6740970015525818,
      "learning_rate": 0.0004168690909211965,
      "loss": 2.0867180824279785,
      "step": 4426
    },
    {
      "epoch": 0.8539737654320988,
      "grad_norm": 0.796188473701477,
      "learning_rate": 0.0004163038150886981,
      "loss": 2.0508413314819336,
      "step": 4427
    },
    {
      "epoch": 0.8541666666666666,
      "grad_norm": 0.6464229822158813,
      "learning_rate": 0.0004157370099934358,
      "loss": 2.206763744354248,
      "step": 4428
    },
    {
      "epoch": 0.8543595679012346,
      "grad_norm": 0.978364884853363,
      "learning_rate": 0.00041516868084752034,
      "loss": 2.3159894943237305,
      "step": 4429
    },
    {
      "epoch": 0.8545524691358025,
      "grad_norm": 0.6969420313835144,
      "learning_rate": 0.0004145988328770773,
      "loss": 2.121682643890381,
      "step": 4430
    },
    {
      "epoch": 0.8547453703703703,
      "grad_norm": 0.8344887495040894,
      "learning_rate": 0.00041402747132219847,
      "loss": 2.681483745574951,
      "step": 4431
    },
    {
      "epoch": 0.8549382716049383,
      "grad_norm": 0.7452347874641418,
      "learning_rate": 0.0004134546014368942,
      "loss": 2.268794059753418,
      "step": 4432
    },
    {
      "epoch": 0.8551311728395061,
      "grad_norm": 0.9817368388175964,
      "learning_rate": 0.0004128802284890448,
      "loss": 2.2809391021728516,
      "step": 4433
    },
    {
      "epoch": 0.8553240740740741,
      "grad_norm": 0.7976057529449463,
      "learning_rate": 0.000412304357760352,
      "loss": 2.331509590148926,
      "step": 4434
    },
    {
      "epoch": 0.855516975308642,
      "grad_norm": 0.8193000555038452,
      "learning_rate": 0.0004117269945462906,
      "loss": 2.0636210441589355,
      "step": 4435
    },
    {
      "epoch": 0.8557098765432098,
      "grad_norm": 0.7830603718757629,
      "learning_rate": 0.00041114814415605977,
      "loss": 2.132526397705078,
      "step": 4436
    },
    {
      "epoch": 0.8559027777777778,
      "grad_norm": 0.753635585308075,
      "learning_rate": 0.00041056781191253403,
      "loss": 2.1128745079040527,
      "step": 4437
    },
    {
      "epoch": 0.8560956790123457,
      "grad_norm": 1.1119784116744995,
      "learning_rate": 0.0004099860031522145,
      "loss": 2.723037004470825,
      "step": 4438
    },
    {
      "epoch": 0.8562885802469136,
      "grad_norm": 0.8369061350822449,
      "learning_rate": 0.00040940272322517957,
      "loss": 2.43833065032959,
      "step": 4439
    },
    {
      "epoch": 0.8564814814814815,
      "grad_norm": 0.8192102313041687,
      "learning_rate": 0.00040881797749503626,
      "loss": 2.4980216026306152,
      "step": 4440
    },
    {
      "epoch": 0.8566743827160493,
      "grad_norm": 0.7963478565216064,
      "learning_rate": 0.00040823177133887016,
      "loss": 2.1753859519958496,
      "step": 4441
    },
    {
      "epoch": 0.8568672839506173,
      "grad_norm": 0.6889941096305847,
      "learning_rate": 0.0004076441101471966,
      "loss": 2.3578109741210938,
      "step": 4442
    },
    {
      "epoch": 0.8570601851851852,
      "grad_norm": 1.0013049840927124,
      "learning_rate": 0.0004070549993239106,
      "loss": 2.231081962585449,
      "step": 4443
    },
    {
      "epoch": 0.8572530864197531,
      "grad_norm": 1.608668565750122,
      "learning_rate": 0.0004064644442862376,
      "loss": 2.260166883468628,
      "step": 4444
    },
    {
      "epoch": 0.857445987654321,
      "grad_norm": 0.7016928791999817,
      "learning_rate": 0.0004058724504646834,
      "loss": 2.274103879928589,
      "step": 4445
    },
    {
      "epoch": 0.8576388888888888,
      "grad_norm": 0.8317071795463562,
      "learning_rate": 0.0004052790233029842,
      "loss": 2.1684799194335938,
      "step": 4446
    },
    {
      "epoch": 0.8578317901234568,
      "grad_norm": 0.7947369813919067,
      "learning_rate": 0.0004046841682580568,
      "loss": 2.27060866355896,
      "step": 4447
    },
    {
      "epoch": 0.8580246913580247,
      "grad_norm": 0.6382129788398743,
      "learning_rate": 0.0004040878907999479,
      "loss": 2.223539352416992,
      "step": 4448
    },
    {
      "epoch": 0.8582175925925926,
      "grad_norm": 0.8382622599601746,
      "learning_rate": 0.0004034901964117844,
      "loss": 2.0952835083007812,
      "step": 4449
    },
    {
      "epoch": 0.8584104938271605,
      "grad_norm": 0.7839120030403137,
      "learning_rate": 0.00040289109058972285,
      "loss": 2.3806447982788086,
      "step": 4450
    },
    {
      "epoch": 0.8586033950617284,
      "grad_norm": 0.8339153528213501,
      "learning_rate": 0.00040229057884289844,
      "loss": 2.3718674182891846,
      "step": 4451
    },
    {
      "epoch": 0.8587962962962963,
      "grad_norm": 0.9236778616905212,
      "learning_rate": 0.0004016886666933749,
      "loss": 1.977476716041565,
      "step": 4452
    },
    {
      "epoch": 0.8589891975308642,
      "grad_norm": 1.0554808378219604,
      "learning_rate": 0.0004010853596760933,
      "loss": 1.958768606185913,
      "step": 4453
    },
    {
      "epoch": 0.8591820987654321,
      "grad_norm": 0.6727779507637024,
      "learning_rate": 0.00040048066333882163,
      "loss": 2.021468162536621,
      "step": 4454
    },
    {
      "epoch": 0.859375,
      "grad_norm": 1.0438915491104126,
      "learning_rate": 0.0003998745832421033,
      "loss": 2.2915782928466797,
      "step": 4455
    },
    {
      "epoch": 0.8595679012345679,
      "grad_norm": 0.8267251253128052,
      "learning_rate": 0.00039926712495920626,
      "loss": 2.2504234313964844,
      "step": 4456
    },
    {
      "epoch": 0.8597608024691358,
      "grad_norm": 0.7320423722267151,
      "learning_rate": 0.0003986582940760717,
      "loss": 1.8449982404708862,
      "step": 4457
    },
    {
      "epoch": 0.8599537037037037,
      "grad_norm": 0.770354688167572,
      "learning_rate": 0.0003980480961912627,
      "loss": 2.1540627479553223,
      "step": 4458
    },
    {
      "epoch": 0.8601466049382716,
      "grad_norm": 0.8577663898468018,
      "learning_rate": 0.00039743653691591275,
      "loss": 2.3958473205566406,
      "step": 4459
    },
    {
      "epoch": 0.8603395061728395,
      "grad_norm": 0.7992221713066101,
      "learning_rate": 0.00039682362187367406,
      "loss": 2.405660390853882,
      "step": 4460
    },
    {
      "epoch": 0.8605324074074074,
      "grad_norm": 0.7200354933738708,
      "learning_rate": 0.00039620935670066614,
      "loss": 2.1037192344665527,
      "step": 4461
    },
    {
      "epoch": 0.8607253086419753,
      "grad_norm": 0.9324653744697571,
      "learning_rate": 0.0003955937470454235,
      "loss": 2.111588478088379,
      "step": 4462
    },
    {
      "epoch": 0.8609182098765432,
      "grad_norm": 1.0614620447158813,
      "learning_rate": 0.00039497679856884403,
      "loss": 2.8483738899230957,
      "step": 4463
    },
    {
      "epoch": 0.8611111111111112,
      "grad_norm": 0.7851678729057312,
      "learning_rate": 0.00039435851694413705,
      "loss": 2.0423717498779297,
      "step": 4464
    },
    {
      "epoch": 0.861304012345679,
      "grad_norm": 0.8235148787498474,
      "learning_rate": 0.00039373890785677075,
      "loss": 2.3647942543029785,
      "step": 4465
    },
    {
      "epoch": 0.8614969135802469,
      "grad_norm": 0.7846280336380005,
      "learning_rate": 0.0003931179770044202,
      "loss": 2.2886359691619873,
      "step": 4466
    },
    {
      "epoch": 0.8616898148148148,
      "grad_norm": 0.8326658010482788,
      "learning_rate": 0.0003924957300969151,
      "loss": 2.4450523853302,
      "step": 4467
    },
    {
      "epoch": 0.8618827160493827,
      "grad_norm": 0.7298814654350281,
      "learning_rate": 0.0003918721728561866,
      "loss": 2.1887779235839844,
      "step": 4468
    },
    {
      "epoch": 0.8620756172839507,
      "grad_norm": 0.7994778156280518,
      "learning_rate": 0.0003912473110162156,
      "loss": 1.9696261882781982,
      "step": 4469
    },
    {
      "epoch": 0.8622685185185185,
      "grad_norm": 0.8513348698616028,
      "learning_rate": 0.00039062115032297934,
      "loss": 2.2929911613464355,
      "step": 4470
    },
    {
      "epoch": 0.8624614197530864,
      "grad_norm": 0.7928573489189148,
      "learning_rate": 0.0003899936965343989,
      "loss": 2.034925937652588,
      "step": 4471
    },
    {
      "epoch": 0.8626543209876543,
      "grad_norm": 0.8594978451728821,
      "learning_rate": 0.0003893649554202859,
      "loss": 2.3150882720947266,
      "step": 4472
    },
    {
      "epoch": 0.8628472222222222,
      "grad_norm": 0.8516197800636292,
      "learning_rate": 0.0003887349327622901,
      "loss": 2.2563729286193848,
      "step": 4473
    },
    {
      "epoch": 0.8630401234567902,
      "grad_norm": 0.6696062088012695,
      "learning_rate": 0.0003881036343538455,
      "loss": 2.298635721206665,
      "step": 4474
    },
    {
      "epoch": 0.863233024691358,
      "grad_norm": 0.7367417216300964,
      "learning_rate": 0.0003874710660001174,
      "loss": 2.388029098510742,
      "step": 4475
    },
    {
      "epoch": 0.8634259259259259,
      "grad_norm": 1.0111145973205566,
      "learning_rate": 0.0003868372335179492,
      "loss": 2.3685009479522705,
      "step": 4476
    },
    {
      "epoch": 0.8636188271604939,
      "grad_norm": 0.949461042881012,
      "learning_rate": 0.00038620214273580846,
      "loss": 2.041079521179199,
      "step": 4477
    },
    {
      "epoch": 0.8638117283950617,
      "grad_norm": 0.8826523423194885,
      "learning_rate": 0.00038556579949373384,
      "loss": 2.1052427291870117,
      "step": 4478
    },
    {
      "epoch": 0.8640046296296297,
      "grad_norm": 0.9194908142089844,
      "learning_rate": 0.0003849282096432808,
      "loss": 2.361611843109131,
      "step": 4479
    },
    {
      "epoch": 0.8641975308641975,
      "grad_norm": 0.6709238290786743,
      "learning_rate": 0.0003842893790474684,
      "loss": 2.106813430786133,
      "step": 4480
    },
    {
      "epoch": 0.8643904320987654,
      "grad_norm": 0.839878261089325,
      "learning_rate": 0.000383649313580725,
      "loss": 2.5921432971954346,
      "step": 4481
    },
    {
      "epoch": 0.8645833333333334,
      "grad_norm": 0.7046933174133301,
      "learning_rate": 0.00038300801912883415,
      "loss": 2.1106836795806885,
      "step": 4482
    },
    {
      "epoch": 0.8647762345679012,
      "grad_norm": 0.7898706793785095,
      "learning_rate": 0.00038236550158888096,
      "loss": 2.1555681228637695,
      "step": 4483
    },
    {
      "epoch": 0.8649691358024691,
      "grad_norm": 0.8772872090339661,
      "learning_rate": 0.00038172176686919733,
      "loss": 2.4144299030303955,
      "step": 4484
    },
    {
      "epoch": 0.8651620370370371,
      "grad_norm": 0.6757965087890625,
      "learning_rate": 0.00038107682088930797,
      "loss": 2.5747649669647217,
      "step": 4485
    },
    {
      "epoch": 0.8653549382716049,
      "grad_norm": 0.7828550338745117,
      "learning_rate": 0.0003804306695798757,
      "loss": 2.521523952484131,
      "step": 4486
    },
    {
      "epoch": 0.8655478395061729,
      "grad_norm": 0.9328006505966187,
      "learning_rate": 0.0003797833188826474,
      "loss": 2.335278034210205,
      "step": 4487
    },
    {
      "epoch": 0.8657407407407407,
      "grad_norm": 0.937201976776123,
      "learning_rate": 0.00037913477475039865,
      "loss": 2.1705102920532227,
      "step": 4488
    },
    {
      "epoch": 0.8659336419753086,
      "grad_norm": 0.886305570602417,
      "learning_rate": 0.0003784850431468795,
      "loss": 2.3734097480773926,
      "step": 4489
    },
    {
      "epoch": 0.8661265432098766,
      "grad_norm": 0.8563534617424011,
      "learning_rate": 0.0003778341300467597,
      "loss": 2.0725302696228027,
      "step": 4490
    },
    {
      "epoch": 0.8663194444444444,
      "grad_norm": 1.0131360292434692,
      "learning_rate": 0.00037718204143557334,
      "loss": 2.378586769104004,
      "step": 4491
    },
    {
      "epoch": 0.8665123456790124,
      "grad_norm": 0.7866343259811401,
      "learning_rate": 0.0003765287833096641,
      "loss": 2.1075363159179688,
      "step": 4492
    },
    {
      "epoch": 0.8667052469135802,
      "grad_norm": 0.6561195254325867,
      "learning_rate": 0.00037587436167612996,
      "loss": 2.0144288539886475,
      "step": 4493
    },
    {
      "epoch": 0.8668981481481481,
      "grad_norm": 0.9767882823944092,
      "learning_rate": 0.00037521878255276843,
      "loss": 2.400322198867798,
      "step": 4494
    },
    {
      "epoch": 0.8670910493827161,
      "grad_norm": 1.0052658319473267,
      "learning_rate": 0.00037456205196802016,
      "loss": 2.3075318336486816,
      "step": 4495
    },
    {
      "epoch": 0.8672839506172839,
      "grad_norm": 0.8575477004051208,
      "learning_rate": 0.00037390417596091483,
      "loss": 2.0002543926239014,
      "step": 4496
    },
    {
      "epoch": 0.8674768518518519,
      "grad_norm": 0.7229261994361877,
      "learning_rate": 0.00037324516058101454,
      "loss": 2.326784133911133,
      "step": 4497
    },
    {
      "epoch": 0.8676697530864198,
      "grad_norm": 0.8594565391540527,
      "learning_rate": 0.0003725850118883589,
      "loss": 2.3210628032684326,
      "step": 4498
    },
    {
      "epoch": 0.8678626543209876,
      "grad_norm": 1.021546721458435,
      "learning_rate": 0.00037192373595340864,
      "loss": 2.7256650924682617,
      "step": 4499
    },
    {
      "epoch": 0.8680555555555556,
      "grad_norm": 0.9459084868431091,
      "learning_rate": 0.00037126133885699054,
      "loss": 2.238832950592041,
      "step": 4500
    },
    {
      "epoch": 0.8682484567901234,
      "grad_norm": 0.9374412298202515,
      "learning_rate": 0.00037059782669024096,
      "loss": 2.37666916847229,
      "step": 4501
    },
    {
      "epoch": 0.8684413580246914,
      "grad_norm": 0.9326936602592468,
      "learning_rate": 0.00036993320555454986,
      "loss": 2.0754127502441406,
      "step": 4502
    },
    {
      "epoch": 0.8686342592592593,
      "grad_norm": 0.8777990937232971,
      "learning_rate": 0.0003692674815615051,
      "loss": 2.206611156463623,
      "step": 4503
    },
    {
      "epoch": 0.8688271604938271,
      "grad_norm": 0.7875782251358032,
      "learning_rate": 0.00036860066083283573,
      "loss": 2.1224329471588135,
      "step": 4504
    },
    {
      "epoch": 0.8690200617283951,
      "grad_norm": 0.7453714609146118,
      "learning_rate": 0.0003679327495003561,
      "loss": 2.2584424018859863,
      "step": 4505
    },
    {
      "epoch": 0.8692129629629629,
      "grad_norm": 0.7729862332344055,
      "learning_rate": 0.00036726375370590926,
      "loss": 2.4462106227874756,
      "step": 4506
    },
    {
      "epoch": 0.8694058641975309,
      "grad_norm": 0.9397099018096924,
      "learning_rate": 0.0003665936796013106,
      "loss": 2.405785083770752,
      "step": 4507
    },
    {
      "epoch": 0.8695987654320988,
      "grad_norm": 0.8013811111450195,
      "learning_rate": 0.00036592253334829096,
      "loss": 2.0979745388031006,
      "step": 4508
    },
    {
      "epoch": 0.8697916666666666,
      "grad_norm": 0.8218268752098083,
      "learning_rate": 0.00036525032111844056,
      "loss": 2.1864280700683594,
      "step": 4509
    },
    {
      "epoch": 0.8699845679012346,
      "grad_norm": 0.9055349826812744,
      "learning_rate": 0.0003645770490931517,
      "loss": 2.562889575958252,
      "step": 4510
    },
    {
      "epoch": 0.8701774691358025,
      "grad_norm": 0.8439086675643921,
      "learning_rate": 0.0003639027234635622,
      "loss": 2.1787350177764893,
      "step": 4511
    },
    {
      "epoch": 0.8703703703703703,
      "grad_norm": 0.7876561284065247,
      "learning_rate": 0.00036322735043049846,
      "loss": 2.0281314849853516,
      "step": 4512
    },
    {
      "epoch": 0.8705632716049383,
      "grad_norm": 0.7549490332603455,
      "learning_rate": 0.0003625509362044183,
      "loss": 2.495110511779785,
      "step": 4513
    },
    {
      "epoch": 0.8707561728395061,
      "grad_norm": 0.7722275853157043,
      "learning_rate": 0.0003618734870053539,
      "loss": 2.045380115509033,
      "step": 4514
    },
    {
      "epoch": 0.8709490740740741,
      "grad_norm": 0.7389785647392273,
      "learning_rate": 0.00036119500906285466,
      "loss": 2.190776824951172,
      "step": 4515
    },
    {
      "epoch": 0.871141975308642,
      "grad_norm": 0.7847622036933899,
      "learning_rate": 0.00036051550861592997,
      "loss": 2.337681293487549,
      "step": 4516
    },
    {
      "epoch": 0.8713348765432098,
      "grad_norm": 1.224867820739746,
      "learning_rate": 0.0003598349919129917,
      "loss": 2.4731531143188477,
      "step": 4517
    },
    {
      "epoch": 0.8715277777777778,
      "grad_norm": 0.9563749432563782,
      "learning_rate": 0.0003591534652117967,
      "loss": 2.3619039058685303,
      "step": 4518
    },
    {
      "epoch": 0.8717206790123457,
      "grad_norm": 1.3094151020050049,
      "learning_rate": 0.00035847093477938953,
      "loss": 2.695352077484131,
      "step": 4519
    },
    {
      "epoch": 0.8719135802469136,
      "grad_norm": 0.7677071690559387,
      "learning_rate": 0.00035778740689204456,
      "loss": 2.288857936859131,
      "step": 4520
    },
    {
      "epoch": 0.8721064814814815,
      "grad_norm": 0.8387743830680847,
      "learning_rate": 0.0003571028878352084,
      "loss": 2.462620258331299,
      "step": 4521
    },
    {
      "epoch": 0.8722993827160493,
      "grad_norm": 0.9561936855316162,
      "learning_rate": 0.000356417383903442,
      "loss": 2.00132155418396,
      "step": 4522
    },
    {
      "epoch": 0.8724922839506173,
      "grad_norm": 0.5649753212928772,
      "learning_rate": 0.0003557309014003629,
      "loss": 2.4560437202453613,
      "step": 4523
    },
    {
      "epoch": 0.8726851851851852,
      "grad_norm": 0.7965677976608276,
      "learning_rate": 0.000355043446638587,
      "loss": 2.0003178119659424,
      "step": 4524
    },
    {
      "epoch": 0.8728780864197531,
      "grad_norm": 0.7039920687675476,
      "learning_rate": 0.00035435502593967103,
      "loss": 2.0825533866882324,
      "step": 4525
    },
    {
      "epoch": 0.873070987654321,
      "grad_norm": 0.6512490510940552,
      "learning_rate": 0.0003536656456340537,
      "loss": 2.04719877243042,
      "step": 4526
    },
    {
      "epoch": 0.8732638888888888,
      "grad_norm": 1.1840463876724243,
      "learning_rate": 0.0003529753120609982,
      "loss": 2.6578142642974854,
      "step": 4527
    },
    {
      "epoch": 0.8734567901234568,
      "grad_norm": 0.8846678137779236,
      "learning_rate": 0.0003522840315685334,
      "loss": 2.419741630554199,
      "step": 4528
    },
    {
      "epoch": 0.8736496913580247,
      "grad_norm": 0.853729784488678,
      "learning_rate": 0.0003515918105133957,
      "loss": 2.4146156311035156,
      "step": 4529
    },
    {
      "epoch": 0.8738425925925926,
      "grad_norm": 0.9670975208282471,
      "learning_rate": 0.00035089865526097063,
      "loss": 2.241528034210205,
      "step": 4530
    },
    {
      "epoch": 0.8740354938271605,
      "grad_norm": 1.10213041305542,
      "learning_rate": 0.00035020457218523405,
      "loss": 2.0981650352478027,
      "step": 4531
    },
    {
      "epoch": 0.8742283950617284,
      "grad_norm": 0.7538405656814575,
      "learning_rate": 0.0003495095676686938,
      "loss": 2.378683567047119,
      "step": 4532
    },
    {
      "epoch": 0.8744212962962963,
      "grad_norm": 1.037654161453247,
      "learning_rate": 0.000348813648102331,
      "loss": 2.442664623260498,
      "step": 4533
    },
    {
      "epoch": 0.8746141975308642,
      "grad_norm": 0.8106901049613953,
      "learning_rate": 0.00034811681988554095,
      "loss": 2.3281900882720947,
      "step": 4534
    },
    {
      "epoch": 0.8748070987654321,
      "grad_norm": 0.9910097718238831,
      "learning_rate": 0.00034741908942607485,
      "loss": 2.504253387451172,
      "step": 4535
    },
    {
      "epoch": 0.875,
      "grad_norm": 0.7535983920097351,
      "learning_rate": 0.0003467204631399803,
      "loss": 2.2105538845062256,
      "step": 4536
    },
    {
      "epoch": 0.8751929012345679,
      "grad_norm": 1.0154138803482056,
      "learning_rate": 0.0003460209474515428,
      "loss": 2.081052780151367,
      "step": 4537
    },
    {
      "epoch": 0.8753858024691358,
      "grad_norm": 0.7426328659057617,
      "learning_rate": 0.00034532054879322607,
      "loss": 2.498059034347534,
      "step": 4538
    },
    {
      "epoch": 0.8755787037037037,
      "grad_norm": 0.8300266861915588,
      "learning_rate": 0.00034461927360561375,
      "loss": 2.523817539215088,
      "step": 4539
    },
    {
      "epoch": 0.8757716049382716,
      "grad_norm": 0.7743049263954163,
      "learning_rate": 0.0003439171283373492,
      "loss": 2.3437981605529785,
      "step": 4540
    },
    {
      "epoch": 0.8759645061728395,
      "grad_norm": 0.7992410063743591,
      "learning_rate": 0.0003432141194450772,
      "loss": 2.3727807998657227,
      "step": 4541
    },
    {
      "epoch": 0.8761574074074074,
      "grad_norm": 0.6656820178031921,
      "learning_rate": 0.0003425102533933835,
      "loss": 2.304286003112793,
      "step": 4542
    },
    {
      "epoch": 0.8763503086419753,
      "grad_norm": 0.8043752312660217,
      "learning_rate": 0.0003418055366547364,
      "loss": 2.00764799118042,
      "step": 4543
    },
    {
      "epoch": 0.8765432098765432,
      "grad_norm": 0.7562431693077087,
      "learning_rate": 0.0003410999757094266,
      "loss": 2.1698126792907715,
      "step": 4544
    },
    {
      "epoch": 0.8767361111111112,
      "grad_norm": 0.6445291042327881,
      "learning_rate": 0.0003403935770455077,
      "loss": 2.2369842529296875,
      "step": 4545
    },
    {
      "epoch": 0.876929012345679,
      "grad_norm": 0.847567081451416,
      "learning_rate": 0.00033968634715873683,
      "loss": 2.3395166397094727,
      "step": 4546
    },
    {
      "epoch": 0.8771219135802469,
      "grad_norm": 0.8819823265075684,
      "learning_rate": 0.0003389782925525146,
      "loss": 2.3203911781311035,
      "step": 4547
    },
    {
      "epoch": 0.8773148148148148,
      "grad_norm": 0.8467101454734802,
      "learning_rate": 0.0003382694197378252,
      "loss": 2.2131519317626953,
      "step": 4548
    },
    {
      "epoch": 0.8775077160493827,
      "grad_norm": 0.9196875691413879,
      "learning_rate": 0.00033755973523317715,
      "loss": 2.1824047565460205,
      "step": 4549
    },
    {
      "epoch": 0.8777006172839507,
      "grad_norm": 0.8486588001251221,
      "learning_rate": 0.00033684924556454267,
      "loss": 2.4079537391662598,
      "step": 4550
    },
    {
      "epoch": 0.8778935185185185,
      "grad_norm": 0.9287007451057434,
      "learning_rate": 0.0003361379572652979,
      "loss": 2.307323455810547,
      "step": 4551
    },
    {
      "epoch": 0.8780864197530864,
      "grad_norm": 0.8770464062690735,
      "learning_rate": 0.0003354258768761632,
      "loss": 2.3629655838012695,
      "step": 4552
    },
    {
      "epoch": 0.8782793209876543,
      "grad_norm": 0.8351670503616333,
      "learning_rate": 0.00033471301094514225,
      "loss": 2.322110176086426,
      "step": 4553
    },
    {
      "epoch": 0.8784722222222222,
      "grad_norm": 1.1658130884170532,
      "learning_rate": 0.0003339993660274624,
      "loss": 2.2753868103027344,
      "step": 4554
    },
    {
      "epoch": 0.8786651234567902,
      "grad_norm": 0.8807008266448975,
      "learning_rate": 0.0003332849486855144,
      "loss": 2.258150339126587,
      "step": 4555
    },
    {
      "epoch": 0.878858024691358,
      "grad_norm": 0.7991892099380493,
      "learning_rate": 0.00033256976548879183,
      "loss": 2.6284444332122803,
      "step": 4556
    },
    {
      "epoch": 0.8790509259259259,
      "grad_norm": 0.8214995265007019,
      "learning_rate": 0.0003318538230138306,
      "loss": 2.027841806411743,
      "step": 4557
    },
    {
      "epoch": 0.8792438271604939,
      "grad_norm": 0.7683238387107849,
      "learning_rate": 0.0003311371278441488,
      "loss": 1.8882639408111572,
      "step": 4558
    },
    {
      "epoch": 0.8794367283950617,
      "grad_norm": 0.9896389842033386,
      "learning_rate": 0.000330419686570186,
      "loss": 2.3599019050598145,
      "step": 4559
    },
    {
      "epoch": 0.8796296296296297,
      "grad_norm": 0.7780360579490662,
      "learning_rate": 0.0003297015057892425,
      "loss": 2.1253561973571777,
      "step": 4560
    },
    {
      "epoch": 0.8798225308641975,
      "grad_norm": 0.708859920501709,
      "learning_rate": 0.000328982592105419,
      "loss": 2.241004467010498,
      "step": 4561
    },
    {
      "epoch": 0.8800154320987654,
      "grad_norm": 0.6395586729049683,
      "learning_rate": 0.0003282629521295556,
      "loss": 2.002208709716797,
      "step": 4562
    },
    {
      "epoch": 0.8802083333333334,
      "grad_norm": 0.8250548839569092,
      "learning_rate": 0.0003275425924791711,
      "loss": 2.2599668502807617,
      "step": 4563
    },
    {
      "epoch": 0.8804012345679012,
      "grad_norm": 0.7640634179115295,
      "learning_rate": 0.00032682151977840183,
      "loss": 2.3934860229492188,
      "step": 4564
    },
    {
      "epoch": 0.8805941358024691,
      "grad_norm": 0.7995167970657349,
      "learning_rate": 0.00032609974065794166,
      "loss": 2.0444436073303223,
      "step": 4565
    },
    {
      "epoch": 0.8807870370370371,
      "grad_norm": 0.8252887725830078,
      "learning_rate": 0.00032537726175498,
      "loss": 2.333852767944336,
      "step": 4566
    },
    {
      "epoch": 0.8809799382716049,
      "grad_norm": 0.7681902050971985,
      "learning_rate": 0.0003246540897131411,
      "loss": 2.368431806564331,
      "step": 4567
    },
    {
      "epoch": 0.8811728395061729,
      "grad_norm": 0.986329197883606,
      "learning_rate": 0.00032393023118242335,
      "loss": 2.2167694568634033,
      "step": 4568
    },
    {
      "epoch": 0.8813657407407407,
      "grad_norm": 0.6459740996360779,
      "learning_rate": 0.0003232056928191376,
      "loss": 2.3918535709381104,
      "step": 4569
    },
    {
      "epoch": 0.8815586419753086,
      "grad_norm": 1.0169357061386108,
      "learning_rate": 0.0003224804812858462,
      "loss": 2.3298768997192383,
      "step": 4570
    },
    {
      "epoch": 0.8817515432098766,
      "grad_norm": 0.9447402358055115,
      "learning_rate": 0.0003217546032513017,
      "loss": 2.275602102279663,
      "step": 4571
    },
    {
      "epoch": 0.8819444444444444,
      "grad_norm": 0.8856281638145447,
      "learning_rate": 0.00032102806539038566,
      "loss": 2.360051393508911,
      "step": 4572
    },
    {
      "epoch": 0.8821373456790124,
      "grad_norm": 1.0204685926437378,
      "learning_rate": 0.00032030087438404686,
      "loss": 2.29941987991333,
      "step": 4573
    },
    {
      "epoch": 0.8823302469135802,
      "grad_norm": 0.7413046360015869,
      "learning_rate": 0.00031957303691924037,
      "loss": 1.8944644927978516,
      "step": 4574
    },
    {
      "epoch": 0.8825231481481481,
      "grad_norm": 0.8652514815330505,
      "learning_rate": 0.0003188445596888658,
      "loss": 2.199212074279785,
      "step": 4575
    },
    {
      "epoch": 0.8827160493827161,
      "grad_norm": 0.8189064860343933,
      "learning_rate": 0.0003181154493917057,
      "loss": 2.0403690338134766,
      "step": 4576
    },
    {
      "epoch": 0.8829089506172839,
      "grad_norm": 0.8100828528404236,
      "learning_rate": 0.0003173857127323642,
      "loss": 2.3084053993225098,
      "step": 4577
    },
    {
      "epoch": 0.8831018518518519,
      "grad_norm": 0.6989569067955017,
      "learning_rate": 0.0003166553564212049,
      "loss": 2.0237646102905273,
      "step": 4578
    },
    {
      "epoch": 0.8832947530864198,
      "grad_norm": 1.0631877183914185,
      "learning_rate": 0.00031592438717428984,
      "loss": 2.3211357593536377,
      "step": 4579
    },
    {
      "epoch": 0.8834876543209876,
      "grad_norm": 0.9691649675369263,
      "learning_rate": 0.00031519281171331717,
      "loss": 2.187006711959839,
      "step": 4580
    },
    {
      "epoch": 0.8836805555555556,
      "grad_norm": 1.0575157403945923,
      "learning_rate": 0.00031446063676555946,
      "loss": 1.9019379615783691,
      "step": 4581
    },
    {
      "epoch": 0.8838734567901234,
      "grad_norm": 0.8554700613021851,
      "learning_rate": 0.00031372786906380223,
      "loss": 2.081195831298828,
      "step": 4582
    },
    {
      "epoch": 0.8840663580246914,
      "grad_norm": 0.8454951047897339,
      "learning_rate": 0.0003129945153462813,
      "loss": 2.453037738800049,
      "step": 4583
    },
    {
      "epoch": 0.8842592592592593,
      "grad_norm": 0.7098966836929321,
      "learning_rate": 0.00031226058235662166,
      "loss": 2.1685965061187744,
      "step": 4584
    },
    {
      "epoch": 0.8844521604938271,
      "grad_norm": 0.8605664968490601,
      "learning_rate": 0.0003115260768437747,
      "loss": 2.2758545875549316,
      "step": 4585
    },
    {
      "epoch": 0.8846450617283951,
      "grad_norm": 0.6692126989364624,
      "learning_rate": 0.0003107910055619568,
      "loss": 2.2888588905334473,
      "step": 4586
    },
    {
      "epoch": 0.8848379629629629,
      "grad_norm": 0.6708304286003113,
      "learning_rate": 0.0003100553752705868,
      "loss": 2.255357503890991,
      "step": 4587
    },
    {
      "epoch": 0.8850308641975309,
      "grad_norm": 0.9915811419487,
      "learning_rate": 0.00030931919273422394,
      "loss": 2.341285228729248,
      "step": 4588
    },
    {
      "epoch": 0.8852237654320988,
      "grad_norm": 1.011817216873169,
      "learning_rate": 0.0003085824647225056,
      "loss": 2.065059185028076,
      "step": 4589
    },
    {
      "epoch": 0.8854166666666666,
      "grad_norm": 0.7266179323196411,
      "learning_rate": 0.00030784519801008544,
      "loss": 2.201202154159546,
      "step": 4590
    },
    {
      "epoch": 0.8856095679012346,
      "grad_norm": 0.8039279580116272,
      "learning_rate": 0.0003071073993765703,
      "loss": 2.544863224029541,
      "step": 4591
    },
    {
      "epoch": 0.8858024691358025,
      "grad_norm": 0.8313072323799133,
      "learning_rate": 0.000306369075606459,
      "loss": 2.1491243839263916,
      "step": 4592
    },
    {
      "epoch": 0.8859953703703703,
      "grad_norm": 0.8963239789009094,
      "learning_rate": 0.0003056302334890786,
      "loss": 2.348012924194336,
      "step": 4593
    },
    {
      "epoch": 0.8861882716049383,
      "grad_norm": 0.8898510932922363,
      "learning_rate": 0.00030489087981852326,
      "loss": 2.573726177215576,
      "step": 4594
    },
    {
      "epoch": 0.8863811728395061,
      "grad_norm": 0.9387103319168091,
      "learning_rate": 0.00030415102139359085,
      "loss": 2.206576108932495,
      "step": 4595
    },
    {
      "epoch": 0.8865740740740741,
      "grad_norm": 0.8622972965240479,
      "learning_rate": 0.0003034106650177208,
      "loss": 2.1464061737060547,
      "step": 4596
    },
    {
      "epoch": 0.886766975308642,
      "grad_norm": 0.7085635662078857,
      "learning_rate": 0.0003026698174989316,
      "loss": 2.437865734100342,
      "step": 4597
    },
    {
      "epoch": 0.8869598765432098,
      "grad_norm": 0.934756338596344,
      "learning_rate": 0.000301928485649758,
      "loss": 2.4726717472076416,
      "step": 4598
    },
    {
      "epoch": 0.8871527777777778,
      "grad_norm": 0.8493130803108215,
      "learning_rate": 0.00030118667628718847,
      "loss": 2.3842124938964844,
      "step": 4599
    },
    {
      "epoch": 0.8873456790123457,
      "grad_norm": 1.0455912351608276,
      "learning_rate": 0.00030044439623260264,
      "loss": 2.3878917694091797,
      "step": 4600
    },
    {
      "epoch": 0.8875385802469136,
      "grad_norm": 0.6420663595199585,
      "learning_rate": 0.0002997016523117081,
      "loss": 2.2256619930267334,
      "step": 4601
    },
    {
      "epoch": 0.8877314814814815,
      "grad_norm": 0.6085125803947449,
      "learning_rate": 0.00029895845135447836,
      "loss": 2.1202988624572754,
      "step": 4602
    },
    {
      "epoch": 0.8879243827160493,
      "grad_norm": 0.642828106880188,
      "learning_rate": 0.0002982148001950893,
      "loss": 2.3400630950927734,
      "step": 4603
    },
    {
      "epoch": 0.8881172839506173,
      "grad_norm": 0.6933702826499939,
      "learning_rate": 0.0002974707056718571,
      "loss": 2.390228271484375,
      "step": 4604
    },
    {
      "epoch": 0.8883101851851852,
      "grad_norm": 0.7215062379837036,
      "learning_rate": 0.00029672617462717446,
      "loss": 1.8739956617355347,
      "step": 4605
    },
    {
      "epoch": 0.8885030864197531,
      "grad_norm": 0.7622384428977966,
      "learning_rate": 0.0002959812139074484,
      "loss": 2.276778221130371,
      "step": 4606
    },
    {
      "epoch": 0.888695987654321,
      "grad_norm": 0.7059794664382935,
      "learning_rate": 0.00029523583036303715,
      "loss": 2.1792190074920654,
      "step": 4607
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 1.1496871709823608,
      "learning_rate": 0.00029449003084818686,
      "loss": 2.2390365600585938,
      "step": 4608
    },
    {
      "epoch": 0.8890817901234568,
      "grad_norm": 0.7422599792480469,
      "learning_rate": 0.00029374382222096884,
      "loss": 2.1310219764709473,
      "step": 4609
    },
    {
      "epoch": 0.8892746913580247,
      "grad_norm": 0.7824891805648804,
      "learning_rate": 0.0002929972113432166,
      "loss": 2.1837263107299805,
      "step": 4610
    },
    {
      "epoch": 0.8894675925925926,
      "grad_norm": 0.8498958349227905,
      "learning_rate": 0.0002922502050804623,
      "loss": 2.0867719650268555,
      "step": 4611
    },
    {
      "epoch": 0.8896604938271605,
      "grad_norm": 0.7263019680976868,
      "learning_rate": 0.00029150281030187406,
      "loss": 2.351871967315674,
      "step": 4612
    },
    {
      "epoch": 0.8898533950617284,
      "grad_norm": 0.7867361307144165,
      "learning_rate": 0.0002907550338801927,
      "loss": 2.1022140979766846,
      "step": 4613
    },
    {
      "epoch": 0.8900462962962963,
      "grad_norm": 0.8352975845336914,
      "learning_rate": 0.00029000688269166834,
      "loss": 2.1159701347351074,
      "step": 4614
    },
    {
      "epoch": 0.8902391975308642,
      "grad_norm": 0.9261497259140015,
      "learning_rate": 0.00028925836361599737,
      "loss": 2.284386396408081,
      "step": 4615
    },
    {
      "epoch": 0.8904320987654321,
      "grad_norm": 0.7339146733283997,
      "learning_rate": 0.0002885094835362591,
      "loss": 2.2965874671936035,
      "step": 4616
    },
    {
      "epoch": 0.890625,
      "grad_norm": 0.6676363945007324,
      "learning_rate": 0.0002877602493388525,
      "loss": 2.2008585929870605,
      "step": 4617
    },
    {
      "epoch": 0.8908179012345679,
      "grad_norm": 0.9984210133552551,
      "learning_rate": 0.00028701066791343287,
      "loss": 2.263326644897461,
      "step": 4618
    },
    {
      "epoch": 0.8910108024691358,
      "grad_norm": 0.6634687185287476,
      "learning_rate": 0.0002862607461528485,
      "loss": 2.1731064319610596,
      "step": 4619
    },
    {
      "epoch": 0.8912037037037037,
      "grad_norm": 0.7540804743766785,
      "learning_rate": 0.0002855104909530772,
      "loss": 2.2669429779052734,
      "step": 4620
    },
    {
      "epoch": 0.8913966049382716,
      "grad_norm": 1.0002877712249756,
      "learning_rate": 0.0002847599092131629,
      "loss": 2.3678336143493652,
      "step": 4621
    },
    {
      "epoch": 0.8915895061728395,
      "grad_norm": 0.6719712615013123,
      "learning_rate": 0.00028400900783515255,
      "loss": 2.3743433952331543,
      "step": 4622
    },
    {
      "epoch": 0.8917824074074074,
      "grad_norm": 0.6588225960731506,
      "learning_rate": 0.00028325779372403195,
      "loss": 1.8800773620605469,
      "step": 4623
    },
    {
      "epoch": 0.8919753086419753,
      "grad_norm": 0.8166779279708862,
      "learning_rate": 0.000282506273787663,
      "loss": 2.3505115509033203,
      "step": 4624
    },
    {
      "epoch": 0.8921682098765432,
      "grad_norm": 0.6377955079078674,
      "learning_rate": 0.0002817544549367197,
      "loss": 1.9513225555419922,
      "step": 4625
    },
    {
      "epoch": 0.8923611111111112,
      "grad_norm": 0.5959562659263611,
      "learning_rate": 0.00028100234408462476,
      "loss": 2.0941803455352783,
      "step": 4626
    },
    {
      "epoch": 0.892554012345679,
      "grad_norm": 0.6896045207977295,
      "learning_rate": 0.0002802499481474861,
      "loss": 2.309382438659668,
      "step": 4627
    },
    {
      "epoch": 0.8927469135802469,
      "grad_norm": 0.7685161232948303,
      "learning_rate": 0.00027949727404403304,
      "loss": 2.5852608680725098,
      "step": 4628
    },
    {
      "epoch": 0.8929398148148148,
      "grad_norm": 0.6505386233329773,
      "learning_rate": 0.00027874432869555275,
      "loss": 2.087625503540039,
      "step": 4629
    },
    {
      "epoch": 0.8931327160493827,
      "grad_norm": 0.9097653031349182,
      "learning_rate": 0.00027799111902582696,
      "loss": 2.110858917236328,
      "step": 4630
    },
    {
      "epoch": 0.8933256172839507,
      "grad_norm": 0.7616153955459595,
      "learning_rate": 0.0002772376519610677,
      "loss": 2.1320247650146484,
      "step": 4631
    },
    {
      "epoch": 0.8935185185185185,
      "grad_norm": 0.7666189670562744,
      "learning_rate": 0.000276483934429854,
      "loss": 2.3766345977783203,
      "step": 4632
    },
    {
      "epoch": 0.8937114197530864,
      "grad_norm": 0.886306643486023,
      "learning_rate": 0.0002757299733630681,
      "loss": 2.250570297241211,
      "step": 4633
    },
    {
      "epoch": 0.8939043209876543,
      "grad_norm": 0.8553739786148071,
      "learning_rate": 0.0002749757756938317,
      "loss": 2.1334238052368164,
      "step": 4634
    },
    {
      "epoch": 0.8940972222222222,
      "grad_norm": 0.8607474565505981,
      "learning_rate": 0.0002742213483574421,
      "loss": 2.3990426063537598,
      "step": 4635
    },
    {
      "epoch": 0.8942901234567902,
      "grad_norm": 0.9771364331245422,
      "learning_rate": 0.00027346669829130866,
      "loss": 2.2683403491973877,
      "step": 4636
    },
    {
      "epoch": 0.894483024691358,
      "grad_norm": 0.9929372072219849,
      "learning_rate": 0.0002727118324348888,
      "loss": 2.631392478942871,
      "step": 4637
    },
    {
      "epoch": 0.8946759259259259,
      "grad_norm": 0.7010145783424377,
      "learning_rate": 0.0002719567577296243,
      "loss": 2.238720417022705,
      "step": 4638
    },
    {
      "epoch": 0.8948688271604939,
      "grad_norm": 0.7022841572761536,
      "learning_rate": 0.0002712014811188773,
      "loss": 2.130715847015381,
      "step": 4639
    },
    {
      "epoch": 0.8950617283950617,
      "grad_norm": 0.7485324740409851,
      "learning_rate": 0.00027044600954786687,
      "loss": 2.2142751216888428,
      "step": 4640
    },
    {
      "epoch": 0.8952546296296297,
      "grad_norm": 0.7198985815048218,
      "learning_rate": 0.0002696903499636045,
      "loss": 2.015359878540039,
      "step": 4641
    },
    {
      "epoch": 0.8954475308641975,
      "grad_norm": 0.6248931288719177,
      "learning_rate": 0.00026893450931483084,
      "loss": 2.005542278289795,
      "step": 4642
    },
    {
      "epoch": 0.8956404320987654,
      "grad_norm": 0.7902679443359375,
      "learning_rate": 0.0002681784945519513,
      "loss": 2.1981120109558105,
      "step": 4643
    },
    {
      "epoch": 0.8958333333333334,
      "grad_norm": 0.8462455868721008,
      "learning_rate": 0.00026742231262697283,
      "loss": 1.98976469039917,
      "step": 4644
    },
    {
      "epoch": 0.8960262345679012,
      "grad_norm": 0.7974252700805664,
      "learning_rate": 0.00026666597049343884,
      "loss": 2.160769462585449,
      "step": 4645
    },
    {
      "epoch": 0.8962191358024691,
      "grad_norm": 0.8519898653030396,
      "learning_rate": 0.0002659094751063666,
      "loss": 2.2287216186523438,
      "step": 4646
    },
    {
      "epoch": 0.8964120370370371,
      "grad_norm": 0.7235591411590576,
      "learning_rate": 0.0002651528334221823,
      "loss": 2.330334186553955,
      "step": 4647
    },
    {
      "epoch": 0.8966049382716049,
      "grad_norm": 0.8946005702018738,
      "learning_rate": 0.00026439605239865743,
      "loss": 2.142385482788086,
      "step": 4648
    },
    {
      "epoch": 0.8967978395061729,
      "grad_norm": 0.7570367455482483,
      "learning_rate": 0.00026363913899484485,
      "loss": 2.1771063804626465,
      "step": 4649
    },
    {
      "epoch": 0.8969907407407407,
      "grad_norm": 0.932973325252533,
      "learning_rate": 0.00026288210017101486,
      "loss": 2.435290575027466,
      "step": 4650
    },
    {
      "epoch": 0.8971836419753086,
      "grad_norm": 0.7447062134742737,
      "learning_rate": 0.0002621249428885908,
      "loss": 2.093336582183838,
      "step": 4651
    },
    {
      "epoch": 0.8973765432098766,
      "grad_norm": 0.7993690371513367,
      "learning_rate": 0.0002613676741100855,
      "loss": 2.496166229248047,
      "step": 4652
    },
    {
      "epoch": 0.8975694444444444,
      "grad_norm": 0.9588176012039185,
      "learning_rate": 0.0002606103007990371,
      "loss": 2.425419807434082,
      "step": 4653
    },
    {
      "epoch": 0.8977623456790124,
      "grad_norm": 0.7003004550933838,
      "learning_rate": 0.0002598528299199448,
      "loss": 2.069565773010254,
      "step": 4654
    },
    {
      "epoch": 0.8979552469135802,
      "grad_norm": 0.747869074344635,
      "learning_rate": 0.0002590952684382051,
      "loss": 2.5435664653778076,
      "step": 4655
    },
    {
      "epoch": 0.8981481481481481,
      "grad_norm": 0.847790539264679,
      "learning_rate": 0.0002583376233200477,
      "loss": 1.9799185991287231,
      "step": 4656
    },
    {
      "epoch": 0.8983410493827161,
      "grad_norm": 0.7904678583145142,
      "learning_rate": 0.00025757990153247126,
      "loss": 2.258950710296631,
      "step": 4657
    },
    {
      "epoch": 0.8985339506172839,
      "grad_norm": 0.7372245192527771,
      "learning_rate": 0.0002568221100431795,
      "loss": 2.2634458541870117,
      "step": 4658
    },
    {
      "epoch": 0.8987268518518519,
      "grad_norm": 0.7843319177627563,
      "learning_rate": 0.0002560642558205172,
      "loss": 2.4303104877471924,
      "step": 4659
    },
    {
      "epoch": 0.8989197530864198,
      "grad_norm": 0.788365364074707,
      "learning_rate": 0.00025530634583340587,
      "loss": 2.5910863876342773,
      "step": 4660
    },
    {
      "epoch": 0.8991126543209876,
      "grad_norm": 0.6816380023956299,
      "learning_rate": 0.0002545483870512799,
      "loss": 2.369920253753662,
      "step": 4661
    },
    {
      "epoch": 0.8993055555555556,
      "grad_norm": 0.7813706994056702,
      "learning_rate": 0.00025379038644402234,
      "loss": 2.1013193130493164,
      "step": 4662
    }
  ],
  "logging_steps": 1,
  "max_steps": 5184,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 518,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 7.888943660748767e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}