{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.8865248226950354,
  "eval_steps": 226,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0017730496453900709,
      "grad_norm": 6.88893985748291,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 4.048,
      "step": 1
    },
    {
      "epoch": 0.0035460992907801418,
      "grad_norm": 6.882226467132568,
      "learning_rate": 5.9999999999999995e-05,
      "loss": 4.2161,
      "step": 2
    },
    {
      "epoch": 0.005319148936170213,
      "grad_norm": 4.995457649230957,
      "learning_rate": 8.999999999999999e-05,
      "loss": 3.9501,
      "step": 3
    },
    {
      "epoch": 0.0070921985815602835,
      "grad_norm": 3.743339776992798,
      "learning_rate": 0.00011999999999999999,
      "loss": 3.7819,
      "step": 4
    },
    {
      "epoch": 0.008865248226950355,
      "grad_norm": 3.3828792572021484,
      "learning_rate": 0.00015,
      "loss": 3.2351,
      "step": 5
    },
    {
      "epoch": 0.010638297872340425,
      "grad_norm": 3.3940675258636475,
      "learning_rate": 0.00017999999999999998,
      "loss": 2.7343,
      "step": 6
    },
    {
      "epoch": 0.012411347517730497,
      "grad_norm": 3.3435909748077393,
      "learning_rate": 0.00020999999999999998,
      "loss": 2.0898,
      "step": 7
    },
    {
      "epoch": 0.014184397163120567,
      "grad_norm": 4.3948235511779785,
      "learning_rate": 0.00023999999999999998,
      "loss": 1.6608,
      "step": 8
    },
    {
      "epoch": 0.015957446808510637,
      "grad_norm": 4.144540309906006,
      "learning_rate": 0.00027,
      "loss": 0.9104,
      "step": 9
    },
    {
      "epoch": 0.01773049645390071,
      "grad_norm": 5.76711368560791,
      "learning_rate": 0.0003,
      "loss": 0.6078,
      "step": 10
    },
    {
      "epoch": 0.01950354609929078,
      "grad_norm": 11.037919998168945,
      "learning_rate": 0.00029973166368515205,
      "loss": 0.439,
      "step": 11
    },
    {
      "epoch": 0.02127659574468085,
      "grad_norm": 5.379396915435791,
      "learning_rate": 0.00029946332737030407,
      "loss": 0.5497,
      "step": 12
    },
    {
      "epoch": 0.02304964539007092,
      "grad_norm": 6.821712970733643,
      "learning_rate": 0.00029919499105545615,
      "loss": 0.6396,
      "step": 13
    },
    {
      "epoch": 0.024822695035460994,
      "grad_norm": 3.520746946334839,
      "learning_rate": 0.00029892665474060817,
      "loss": 0.4007,
      "step": 14
    },
    {
      "epoch": 0.026595744680851064,
      "grad_norm": 1.7975941896438599,
      "learning_rate": 0.00029865831842576025,
      "loss": 0.2384,
      "step": 15
    },
    {
      "epoch": 0.028368794326241134,
      "grad_norm": 6.044783592224121,
      "learning_rate": 0.0002983899821109123,
      "loss": 0.5119,
      "step": 16
    },
    {
      "epoch": 0.030141843971631204,
      "grad_norm": 0.9617128968238831,
      "learning_rate": 0.0002981216457960644,
      "loss": 0.2032,
      "step": 17
    },
    {
      "epoch": 0.031914893617021274,
      "grad_norm": 1.6461282968521118,
      "learning_rate": 0.0002978533094812164,
      "loss": 0.3118,
      "step": 18
    },
    {
      "epoch": 0.03368794326241135,
      "grad_norm": 1.2195343971252441,
      "learning_rate": 0.0002975849731663685,
      "loss": 0.2522,
      "step": 19
    },
    {
      "epoch": 0.03546099290780142,
      "grad_norm": 1.5621528625488281,
      "learning_rate": 0.00029731663685152057,
      "loss": 0.3118,
      "step": 20
    },
    {
      "epoch": 0.03723404255319149,
      "grad_norm": 1.3373510837554932,
      "learning_rate": 0.0002970483005366726,
      "loss": 0.2101,
      "step": 21
    },
    {
      "epoch": 0.03900709219858156,
      "grad_norm": 0.9622485041618347,
      "learning_rate": 0.00029677996422182467,
      "loss": 0.1797,
      "step": 22
    },
    {
      "epoch": 0.040780141843971635,
      "grad_norm": 1.7572095394134521,
      "learning_rate": 0.0002965116279069767,
      "loss": 0.1338,
      "step": 23
    },
    {
      "epoch": 0.0425531914893617,
      "grad_norm": 1.871632695198059,
      "learning_rate": 0.00029624329159212877,
      "loss": 0.1457,
      "step": 24
    },
    {
      "epoch": 0.044326241134751775,
      "grad_norm": 2.204594612121582,
      "learning_rate": 0.00029597495527728084,
      "loss": 0.1621,
      "step": 25
    },
    {
      "epoch": 0.04609929078014184,
      "grad_norm": 1.5381090641021729,
      "learning_rate": 0.0002957066189624329,
      "loss": 0.1703,
      "step": 26
    },
    {
      "epoch": 0.047872340425531915,
      "grad_norm": 0.8811854124069214,
      "learning_rate": 0.00029543828264758494,
      "loss": 0.0787,
      "step": 27
    },
    {
      "epoch": 0.04964539007092199,
      "grad_norm": 0.8814285397529602,
      "learning_rate": 0.000295169946332737,
      "loss": 0.2091,
      "step": 28
    },
    {
      "epoch": 0.051418439716312055,
      "grad_norm": 1.371406078338623,
      "learning_rate": 0.0002949016100178891,
      "loss": 0.3093,
      "step": 29
    },
    {
      "epoch": 0.05319148936170213,
      "grad_norm": 1.1055055856704712,
      "learning_rate": 0.0002946332737030411,
      "loss": 0.1644,
      "step": 30
    },
    {
      "epoch": 0.0549645390070922,
      "grad_norm": 1.2226752042770386,
      "learning_rate": 0.0002943649373881932,
      "loss": 0.1597,
      "step": 31
    },
    {
      "epoch": 0.05673758865248227,
      "grad_norm": 1.1774011850357056,
      "learning_rate": 0.0002940966010733452,
      "loss": 0.243,
      "step": 32
    },
    {
      "epoch": 0.05851063829787234,
      "grad_norm": 1.8663568496704102,
      "learning_rate": 0.0002938282647584973,
      "loss": 0.3355,
      "step": 33
    },
    {
      "epoch": 0.06028368794326241,
      "grad_norm": 0.7782514095306396,
      "learning_rate": 0.0002935599284436493,
      "loss": 0.1426,
      "step": 34
    },
    {
      "epoch": 0.06205673758865248,
      "grad_norm": 0.6109825968742371,
      "learning_rate": 0.0002932915921288014,
      "loss": 0.1334,
      "step": 35
    },
    {
      "epoch": 0.06382978723404255,
      "grad_norm": 1.297591209411621,
      "learning_rate": 0.00029302325581395347,
      "loss": 0.3717,
      "step": 36
    },
    {
      "epoch": 0.06560283687943262,
      "grad_norm": 0.8587173223495483,
      "learning_rate": 0.00029275491949910554,
      "loss": 0.1469,
      "step": 37
    },
    {
      "epoch": 0.0673758865248227,
      "grad_norm": 0.8923099637031555,
      "learning_rate": 0.00029248658318425756,
      "loss": 0.1838,
      "step": 38
    },
    {
      "epoch": 0.06914893617021277,
      "grad_norm": 0.8410468697547913,
      "learning_rate": 0.00029221824686940964,
      "loss": 0.2471,
      "step": 39
    },
    {
      "epoch": 0.07092198581560284,
      "grad_norm": 0.5034096240997314,
      "learning_rate": 0.0002919499105545617,
      "loss": 0.0616,
      "step": 40
    },
    {
      "epoch": 0.0726950354609929,
      "grad_norm": 0.31462836265563965,
      "learning_rate": 0.00029168157423971374,
      "loss": 0.0519,
      "step": 41
    },
    {
      "epoch": 0.07446808510638298,
      "grad_norm": 0.5840567350387573,
      "learning_rate": 0.0002914132379248658,
      "loss": 0.0814,
      "step": 42
    },
    {
      "epoch": 0.07624113475177305,
      "grad_norm": 0.5441888570785522,
      "learning_rate": 0.00029114490161001784,
      "loss": 0.0731,
      "step": 43
    },
    {
      "epoch": 0.07801418439716312,
      "grad_norm": 1.0477969646453857,
      "learning_rate": 0.0002908765652951699,
      "loss": 0.1842,
      "step": 44
    },
    {
      "epoch": 0.0797872340425532,
      "grad_norm": 1.046530842781067,
      "learning_rate": 0.000290608228980322,
      "loss": 0.2727,
      "step": 45
    },
    {
      "epoch": 0.08156028368794327,
      "grad_norm": 0.721427857875824,
      "learning_rate": 0.00029033989266547406,
      "loss": 0.1343,
      "step": 46
    },
    {
      "epoch": 0.08333333333333333,
      "grad_norm": 2.0451300144195557,
      "learning_rate": 0.0002900715563506261,
      "loss": 0.076,
      "step": 47
    },
    {
      "epoch": 0.0851063829787234,
      "grad_norm": 0.3739748001098633,
      "learning_rate": 0.00028980322003577816,
      "loss": 0.0419,
      "step": 48
    },
    {
      "epoch": 0.08687943262411348,
      "grad_norm": 0.5523413419723511,
      "learning_rate": 0.00028953488372093024,
      "loss": 0.0563,
      "step": 49
    },
    {
      "epoch": 0.08865248226950355,
      "grad_norm": 0.5457972288131714,
      "learning_rate": 0.00028926654740608226,
      "loss": 0.0256,
      "step": 50
    },
    {
      "epoch": 0.09042553191489362,
      "grad_norm": 2.8798046112060547,
      "learning_rate": 0.00028899821109123434,
      "loss": 0.8653,
      "step": 51
    },
    {
      "epoch": 0.09219858156028368,
      "grad_norm": 2.86541748046875,
      "learning_rate": 0.00028872987477638636,
      "loss": 0.5141,
      "step": 52
    },
    {
      "epoch": 0.09397163120567376,
      "grad_norm": 10.113655090332031,
      "learning_rate": 0.00028846153846153843,
      "loss": 0.278,
      "step": 53
    },
    {
      "epoch": 0.09574468085106383,
      "grad_norm": 3.4064385890960693,
      "learning_rate": 0.00028819320214669046,
      "loss": 0.3353,
      "step": 54
    },
    {
      "epoch": 0.0975177304964539,
      "grad_norm": 4.866039752960205,
      "learning_rate": 0.00028792486583184253,
      "loss": 0.4487,
      "step": 55
    },
    {
      "epoch": 0.09929078014184398,
      "grad_norm": 5.0878143310546875,
      "learning_rate": 0.0002876565295169946,
      "loss": 0.2099,
      "step": 56
    },
    {
      "epoch": 0.10106382978723404,
      "grad_norm": 5.835223197937012,
      "learning_rate": 0.0002873881932021467,
      "loss": 0.4639,
      "step": 57
    },
    {
      "epoch": 0.10283687943262411,
      "grad_norm": 3.5590853691101074,
      "learning_rate": 0.00028711985688729876,
      "loss": 0.3146,
      "step": 58
    },
    {
      "epoch": 0.10460992907801418,
      "grad_norm": 5.281091213226318,
      "learning_rate": 0.0002868515205724508,
      "loss": 0.4222,
      "step": 59
    },
    {
      "epoch": 0.10638297872340426,
      "grad_norm": 6.650637149810791,
      "learning_rate": 0.00028658318425760286,
      "loss": 0.4737,
      "step": 60
    },
    {
      "epoch": 0.10815602836879433,
      "grad_norm": 1.2466630935668945,
      "learning_rate": 0.0002863148479427549,
      "loss": 0.2634,
      "step": 61
    },
    {
      "epoch": 0.1099290780141844,
      "grad_norm": 0.73129802942276,
      "learning_rate": 0.00028604651162790696,
      "loss": 0.1581,
      "step": 62
    },
    {
      "epoch": 0.11170212765957446,
      "grad_norm": 0.8522558212280273,
      "learning_rate": 0.000285778175313059,
      "loss": 0.2586,
      "step": 63
    },
    {
      "epoch": 0.11347517730496454,
      "grad_norm": 0.7632489204406738,
      "learning_rate": 0.00028550983899821105,
      "loss": 0.2528,
      "step": 64
    },
    {
      "epoch": 0.11524822695035461,
      "grad_norm": 0.4096511900424957,
      "learning_rate": 0.00028524150268336313,
      "loss": 0.0851,
      "step": 65
    },
    {
      "epoch": 0.11702127659574468,
      "grad_norm": 0.9045661687850952,
      "learning_rate": 0.0002849731663685152,
      "loss": 0.2253,
      "step": 66
    },
    {
      "epoch": 0.11879432624113476,
      "grad_norm": 0.6839063763618469,
      "learning_rate": 0.00028470483005366723,
      "loss": 0.1416,
      "step": 67
    },
    {
      "epoch": 0.12056737588652482,
      "grad_norm": 2.112931728363037,
      "learning_rate": 0.0002844364937388193,
      "loss": 0.2554,
      "step": 68
    },
    {
      "epoch": 0.12234042553191489,
      "grad_norm": 1.187727928161621,
      "learning_rate": 0.0002841681574239714,
      "loss": 0.2891,
      "step": 69
    },
    {
      "epoch": 0.12411347517730496,
      "grad_norm": 0.7449979782104492,
      "learning_rate": 0.0002838998211091234,
      "loss": 0.1743,
      "step": 70
    },
    {
      "epoch": 0.12588652482269502,
      "grad_norm": 0.5655006766319275,
      "learning_rate": 0.0002836314847942755,
      "loss": 0.1247,
      "step": 71
    },
    {
      "epoch": 0.1276595744680851,
      "grad_norm": 0.7953659296035767,
      "learning_rate": 0.0002833631484794275,
      "loss": 0.1277,
      "step": 72
    },
    {
      "epoch": 0.12943262411347517,
      "grad_norm": 0.7307471632957458,
      "learning_rate": 0.0002830948121645796,
      "loss": 0.2575,
      "step": 73
    },
    {
      "epoch": 0.13120567375886524,
      "grad_norm": 0.815560519695282,
      "learning_rate": 0.0002828264758497316,
      "loss": 0.2547,
      "step": 74
    },
    {
      "epoch": 0.13297872340425532,
      "grad_norm": 0.7381014227867126,
      "learning_rate": 0.0002825581395348837,
      "loss": 0.1005,
      "step": 75
    },
    {
      "epoch": 0.1347517730496454,
      "grad_norm": 1.0876243114471436,
      "learning_rate": 0.00028228980322003575,
      "loss": 0.2536,
      "step": 76
    },
    {
      "epoch": 0.13652482269503546,
      "grad_norm": 0.6369747519493103,
      "learning_rate": 0.00028202146690518783,
      "loss": 0.1696,
      "step": 77
    },
    {
      "epoch": 0.13829787234042554,
      "grad_norm": 0.5377818942070007,
      "learning_rate": 0.0002817531305903399,
      "loss": 0.1528,
      "step": 78
    },
    {
      "epoch": 0.1400709219858156,
      "grad_norm": 0.5050985813140869,
      "learning_rate": 0.0002814847942754919,
      "loss": 0.1376,
      "step": 79
    },
    {
      "epoch": 0.14184397163120568,
      "grad_norm": 0.46020108461380005,
      "learning_rate": 0.000281216457960644,
      "loss": 0.1164,
      "step": 80
    },
    {
      "epoch": 0.14361702127659576,
      "grad_norm": 0.4778999090194702,
      "learning_rate": 0.000280948121645796,
      "loss": 0.1078,
      "step": 81
    },
    {
      "epoch": 0.1453900709219858,
      "grad_norm": 0.7942756414413452,
      "learning_rate": 0.0002806797853309481,
      "loss": 0.2139,
      "step": 82
    },
    {
      "epoch": 0.14716312056737588,
      "grad_norm": 0.5343471169471741,
      "learning_rate": 0.0002804114490161001,
      "loss": 0.1331,
      "step": 83
    },
    {
      "epoch": 0.14893617021276595,
      "grad_norm": 0.8038661479949951,
      "learning_rate": 0.0002801431127012522,
      "loss": 0.1801,
      "step": 84
    },
    {
      "epoch": 0.15070921985815602,
      "grad_norm": 0.3083706498146057,
      "learning_rate": 0.0002798747763864043,
      "loss": 0.0437,
      "step": 85
    },
    {
      "epoch": 0.1524822695035461,
      "grad_norm": 0.6344548463821411,
      "learning_rate": 0.00027960644007155635,
      "loss": 0.1676,
      "step": 86
    },
    {
      "epoch": 0.15425531914893617,
      "grad_norm": 0.5747407674789429,
      "learning_rate": 0.00027933810375670837,
      "loss": 0.1079,
      "step": 87
    },
    {
      "epoch": 0.15602836879432624,
      "grad_norm": 0.5270193815231323,
      "learning_rate": 0.00027906976744186045,
      "loss": 0.1263,
      "step": 88
    },
    {
      "epoch": 0.15780141843971632,
      "grad_norm": 0.4451313614845276,
      "learning_rate": 0.0002788014311270125,
      "loss": 0.0511,
      "step": 89
    },
    {
      "epoch": 0.1595744680851064,
      "grad_norm": 0.691502034664154,
      "learning_rate": 0.00027853309481216455,
      "loss": 0.1637,
      "step": 90
    },
    {
      "epoch": 0.16134751773049646,
      "grad_norm": 0.5651169419288635,
      "learning_rate": 0.0002782647584973166,
      "loss": 0.1371,
      "step": 91
    },
    {
      "epoch": 0.16312056737588654,
      "grad_norm": 0.5306984782218933,
      "learning_rate": 0.00027799642218246864,
      "loss": 0.1535,
      "step": 92
    },
    {
      "epoch": 0.16489361702127658,
      "grad_norm": 0.5752344727516174,
      "learning_rate": 0.0002777280858676207,
      "loss": 0.1622,
      "step": 93
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 0.3410884737968445,
      "learning_rate": 0.0002774597495527728,
      "loss": 0.0683,
      "step": 94
    },
    {
      "epoch": 0.16843971631205673,
      "grad_norm": 0.6347271203994751,
      "learning_rate": 0.0002771914132379248,
      "loss": 0.1577,
      "step": 95
    },
    {
      "epoch": 0.1702127659574468,
      "grad_norm": 0.36072492599487305,
      "learning_rate": 0.0002769230769230769,
      "loss": 0.0367,
      "step": 96
    },
    {
      "epoch": 0.17198581560283688,
      "grad_norm": 0.29759401082992554,
      "learning_rate": 0.00027665474060822897,
      "loss": 0.0218,
      "step": 97
    },
    {
      "epoch": 0.17375886524822695,
      "grad_norm": 0.37029051780700684,
      "learning_rate": 0.00027638640429338105,
      "loss": 0.0263,
      "step": 98
    },
    {
      "epoch": 0.17553191489361702,
      "grad_norm": 0.3705526292324066,
      "learning_rate": 0.00027611806797853307,
      "loss": 0.0219,
      "step": 99
    },
    {
      "epoch": 0.1773049645390071,
      "grad_norm": 0.31463027000427246,
      "learning_rate": 0.00027584973166368514,
      "loss": 0.023,
      "step": 100
    },
    {
      "epoch": 0.17907801418439717,
      "grad_norm": 1.7321122884750366,
      "learning_rate": 0.00027558139534883717,
      "loss": 0.6776,
      "step": 101
    },
    {
      "epoch": 0.18085106382978725,
      "grad_norm": 1.03788423538208,
      "learning_rate": 0.00027531305903398924,
      "loss": 0.4406,
      "step": 102
    },
    {
      "epoch": 0.18262411347517732,
      "grad_norm": 1.13809335231781,
      "learning_rate": 0.00027504472271914126,
      "loss": 0.4302,
      "step": 103
    },
    {
      "epoch": 0.18439716312056736,
      "grad_norm": 0.6666789650917053,
      "learning_rate": 0.00027477638640429334,
      "loss": 0.2029,
      "step": 104
    },
    {
      "epoch": 0.18617021276595744,
      "grad_norm": 0.9535943865776062,
      "learning_rate": 0.0002745080500894454,
      "loss": 0.2808,
      "step": 105
    },
    {
      "epoch": 0.1879432624113475,
      "grad_norm": 1.2909356355667114,
      "learning_rate": 0.0002742397137745975,
      "loss": 0.3593,
      "step": 106
    },
    {
      "epoch": 0.18971631205673758,
      "grad_norm": 0.9249064326286316,
      "learning_rate": 0.0002739713774597495,
      "loss": 0.355,
      "step": 107
    },
    {
      "epoch": 0.19148936170212766,
      "grad_norm": 0.5897863507270813,
      "learning_rate": 0.0002737030411449016,
      "loss": 0.2349,
      "step": 108
    },
    {
      "epoch": 0.19326241134751773,
      "grad_norm": 1.060029149055481,
      "learning_rate": 0.00027343470483005367,
      "loss": 0.2517,
      "step": 109
    },
    {
      "epoch": 0.1950354609929078,
      "grad_norm": 0.6262145638465881,
      "learning_rate": 0.0002731663685152057,
      "loss": 0.2119,
      "step": 110
    },
    {
      "epoch": 0.19680851063829788,
      "grad_norm": 0.7881852984428406,
      "learning_rate": 0.00027289803220035777,
      "loss": 0.2792,
      "step": 111
    },
    {
      "epoch": 0.19858156028368795,
      "grad_norm": 0.7693052887916565,
      "learning_rate": 0.0002726296958855098,
      "loss": 0.2824,
      "step": 112
    },
    {
      "epoch": 0.20035460992907803,
      "grad_norm": 0.7235965132713318,
      "learning_rate": 0.00027236135957066186,
      "loss": 0.0918,
      "step": 113
    },
    {
      "epoch": 0.20212765957446807,
      "grad_norm": 1.0118364095687866,
      "learning_rate": 0.00027209302325581394,
      "loss": 0.211,
      "step": 114
    },
    {
      "epoch": 0.20390070921985815,
      "grad_norm": 0.9211074709892273,
      "learning_rate": 0.00027182468694096596,
      "loss": 0.2766,
      "step": 115
    },
    {
      "epoch": 0.20567375886524822,
      "grad_norm": 1.3270055055618286,
      "learning_rate": 0.00027155635062611804,
      "loss": 0.1172,
      "step": 116
    },
    {
      "epoch": 0.2074468085106383,
      "grad_norm": 0.6904440522193909,
      "learning_rate": 0.0002712880143112701,
      "loss": 0.2208,
      "step": 117
    },
    {
      "epoch": 0.20921985815602837,
      "grad_norm": 0.44743916392326355,
      "learning_rate": 0.0002710196779964222,
      "loss": 0.1265,
      "step": 118
    },
    {
      "epoch": 0.21099290780141844,
      "grad_norm": 1.5493639707565308,
      "learning_rate": 0.0002707513416815742,
      "loss": 0.2085,
      "step": 119
    },
    {
      "epoch": 0.2127659574468085,
      "grad_norm": 0.7296533584594727,
      "learning_rate": 0.0002704830053667263,
      "loss": 0.253,
      "step": 120
    },
    {
      "epoch": 0.21453900709219859,
      "grad_norm": 1.300660490989685,
      "learning_rate": 0.0002702146690518783,
      "loss": 0.2164,
      "step": 121
    },
    {
      "epoch": 0.21631205673758866,
      "grad_norm": 0.9226039052009583,
      "learning_rate": 0.0002699463327370304,
      "loss": 0.2533,
      "step": 122
    },
    {
      "epoch": 0.21808510638297873,
      "grad_norm": 0.9466855525970459,
      "learning_rate": 0.0002696779964221824,
      "loss": 0.2528,
      "step": 123
    },
    {
      "epoch": 0.2198581560283688,
      "grad_norm": 0.6484300494194031,
      "learning_rate": 0.0002694096601073345,
      "loss": 0.1598,
      "step": 124
    },
    {
      "epoch": 0.22163120567375885,
      "grad_norm": 0.7165495157241821,
      "learning_rate": 0.00026914132379248656,
      "loss": 0.2448,
      "step": 125
    },
    {
      "epoch": 0.22340425531914893,
      "grad_norm": 1.370329737663269,
      "learning_rate": 0.00026887298747763864,
      "loss": 0.3169,
      "step": 126
    },
    {
      "epoch": 0.225177304964539,
      "grad_norm": 0.41645482182502747,
      "learning_rate": 0.00026860465116279066,
      "loss": 0.1352,
      "step": 127
    },
    {
      "epoch": 0.22695035460992907,
      "grad_norm": 0.4441884458065033,
      "learning_rate": 0.00026833631484794273,
      "loss": 0.1376,
      "step": 128
    },
    {
      "epoch": 0.22872340425531915,
      "grad_norm": 0.3925491273403168,
      "learning_rate": 0.0002680679785330948,
      "loss": 0.1194,
      "step": 129
    },
    {
      "epoch": 0.23049645390070922,
      "grad_norm": 0.3546803295612335,
      "learning_rate": 0.00026779964221824683,
      "loss": 0.0936,
      "step": 130
    },
    {
      "epoch": 0.2322695035460993,
      "grad_norm": 0.683130145072937,
      "learning_rate": 0.0002675313059033989,
      "loss": 0.2044,
      "step": 131
    },
    {
      "epoch": 0.23404255319148937,
      "grad_norm": 0.666155219078064,
      "learning_rate": 0.00026726296958855093,
      "loss": 0.2361,
      "step": 132
    },
    {
      "epoch": 0.23581560283687944,
      "grad_norm": 0.5066004991531372,
      "learning_rate": 0.000266994633273703,
      "loss": 0.1449,
      "step": 133
    },
    {
      "epoch": 0.2375886524822695,
      "grad_norm": 0.745506763458252,
      "learning_rate": 0.0002667262969588551,
      "loss": 0.1987,
      "step": 134
    },
    {
      "epoch": 0.2393617021276596,
      "grad_norm": 0.9249691963195801,
      "learning_rate": 0.0002664579606440071,
      "loss": 0.2485,
      "step": 135
    },
    {
      "epoch": 0.24113475177304963,
      "grad_norm": 0.5891656875610352,
      "learning_rate": 0.0002661896243291592,
      "loss": 0.2172,
      "step": 136
    },
    {
      "epoch": 0.2429078014184397,
      "grad_norm": 0.6771588325500488,
      "learning_rate": 0.00026592128801431126,
      "loss": 0.2293,
      "step": 137
    },
    {
      "epoch": 0.24468085106382978,
      "grad_norm": 1.0185132026672363,
      "learning_rate": 0.00026565295169946333,
      "loss": 0.2819,
      "step": 138
    },
    {
      "epoch": 0.24645390070921985,
      "grad_norm": 0.357791930437088,
      "learning_rate": 0.00026538461538461536,
      "loss": 0.0876,
      "step": 139
    },
    {
      "epoch": 0.24822695035460993,
      "grad_norm": 0.5244898200035095,
      "learning_rate": 0.00026511627906976743,
      "loss": 0.0989,
      "step": 140
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.553326427936554,
      "learning_rate": 0.00026484794275491945,
      "loss": 0.1503,
      "step": 141
    },
    {
      "epoch": 0.25177304964539005,
      "grad_norm": 0.7104494571685791,
      "learning_rate": 0.00026457960644007153,
      "loss": 0.2487,
      "step": 142
    },
    {
      "epoch": 0.25354609929078015,
      "grad_norm": 0.5205265879631042,
      "learning_rate": 0.0002643112701252236,
      "loss": 0.1185,
      "step": 143
    },
    {
      "epoch": 0.2553191489361702,
      "grad_norm": 0.41840335726737976,
      "learning_rate": 0.00026404293381037563,
      "loss": 0.0841,
      "step": 144
    },
    {
      "epoch": 0.2570921985815603,
      "grad_norm": 0.572222888469696,
      "learning_rate": 0.0002637745974955277,
      "loss": 0.1331,
      "step": 145
    },
    {
      "epoch": 0.25886524822695034,
      "grad_norm": 0.8182592391967773,
      "learning_rate": 0.0002635062611806798,
      "loss": 0.2042,
      "step": 146
    },
    {
      "epoch": 0.26063829787234044,
      "grad_norm": 0.5115887522697449,
      "learning_rate": 0.0002632379248658318,
      "loss": 0.1148,
      "step": 147
    },
    {
      "epoch": 0.2624113475177305,
      "grad_norm": 0.5078725218772888,
      "learning_rate": 0.0002629695885509839,
      "loss": 0.0762,
      "step": 148
    },
    {
      "epoch": 0.2641843971631206,
      "grad_norm": 0.22413299977779388,
      "learning_rate": 0.00026270125223613595,
      "loss": 0.0355,
      "step": 149
    },
    {
      "epoch": 0.26595744680851063,
      "grad_norm": 0.25390034914016724,
      "learning_rate": 0.000262432915921288,
      "loss": 0.0319,
      "step": 150
    },
    {
      "epoch": 0.26773049645390073,
      "grad_norm": 3.294741153717041,
      "learning_rate": 0.00026216457960644005,
      "loss": 0.4876,
      "step": 151
    },
    {
      "epoch": 0.2695035460992908,
      "grad_norm": 1.5393906831741333,
      "learning_rate": 0.0002618962432915921,
      "loss": 0.2547,
      "step": 152
    },
    {
      "epoch": 0.2712765957446808,
      "grad_norm": 1.7025604248046875,
      "learning_rate": 0.00026162790697674415,
      "loss": 0.4747,
      "step": 153
    },
    {
      "epoch": 0.2730496453900709,
      "grad_norm": 0.8344413638114929,
      "learning_rate": 0.0002613595706618962,
      "loss": 0.2431,
      "step": 154
    },
    {
      "epoch": 0.274822695035461,
      "grad_norm": 0.9473798871040344,
      "learning_rate": 0.00026109123434704825,
      "loss": 0.2135,
      "step": 155
    },
    {
      "epoch": 0.2765957446808511,
      "grad_norm": 1.0257476568222046,
      "learning_rate": 0.0002608228980322003,
      "loss": 0.1833,
      "step": 156
    },
    {
      "epoch": 0.2783687943262411,
      "grad_norm": 1.2221112251281738,
      "learning_rate": 0.0002605545617173524,
      "loss": 0.3174,
      "step": 157
    },
    {
      "epoch": 0.2801418439716312,
      "grad_norm": 0.5619825124740601,
      "learning_rate": 0.0002602862254025045,
      "loss": 0.1492,
      "step": 158
    },
    {
      "epoch": 0.28191489361702127,
      "grad_norm": 1.1845368146896362,
      "learning_rate": 0.0002600178890876565,
      "loss": 0.4336,
      "step": 159
    },
    {
      "epoch": 0.28368794326241137,
      "grad_norm": 1.2730036973953247,
      "learning_rate": 0.0002597495527728086,
      "loss": 0.3466,
      "step": 160
    },
    {
      "epoch": 0.2854609929078014,
      "grad_norm": 0.4464946985244751,
      "learning_rate": 0.0002594812164579606,
      "loss": 0.1239,
      "step": 161
    },
    {
      "epoch": 0.2872340425531915,
      "grad_norm": 0.5267930030822754,
      "learning_rate": 0.00025921288014311267,
      "loss": 0.1372,
      "step": 162
    },
    {
      "epoch": 0.28900709219858156,
      "grad_norm": 0.4249666929244995,
      "learning_rate": 0.00025894454382826475,
      "loss": 0.1402,
      "step": 163
    },
    {
      "epoch": 0.2907801418439716,
      "grad_norm": 0.7619920372962952,
      "learning_rate": 0.00025867620751341677,
      "loss": 0.2309,
      "step": 164
    },
    {
      "epoch": 0.2925531914893617,
      "grad_norm": 0.9128990769386292,
      "learning_rate": 0.00025840787119856885,
      "loss": 0.2832,
      "step": 165
    },
    {
      "epoch": 0.29432624113475175,
      "grad_norm": 0.4988587200641632,
      "learning_rate": 0.0002581395348837209,
      "loss": 0.1517,
      "step": 166
    },
    {
      "epoch": 0.29609929078014185,
      "grad_norm": 0.7568597793579102,
      "learning_rate": 0.000257871198568873,
      "loss": 0.2449,
      "step": 167
    },
    {
      "epoch": 0.2978723404255319,
      "grad_norm": 0.7556333541870117,
      "learning_rate": 0.000257602862254025,
      "loss": 0.2591,
      "step": 168
    },
    {
      "epoch": 0.299645390070922,
      "grad_norm": 0.9023774266242981,
      "learning_rate": 0.0002573345259391771,
      "loss": 0.239,
      "step": 169
    },
    {
      "epoch": 0.30141843971631205,
      "grad_norm": 0.490642786026001,
      "learning_rate": 0.0002570661896243291,
      "loss": 0.1508,
      "step": 170
    },
    {
      "epoch": 0.30319148936170215,
      "grad_norm": 0.6215077042579651,
      "learning_rate": 0.0002567978533094812,
      "loss": 0.2277,
      "step": 171
    },
    {
      "epoch": 0.3049645390070922,
      "grad_norm": 0.4409044086933136,
      "learning_rate": 0.00025652951699463327,
      "loss": 0.1354,
      "step": 172
    },
    {
      "epoch": 0.3067375886524823,
      "grad_norm": 0.420749694108963,
      "learning_rate": 0.0002562611806797853,
      "loss": 0.068,
      "step": 173
    },
    {
      "epoch": 0.30851063829787234,
      "grad_norm": 0.40215224027633667,
      "learning_rate": 0.00025599284436493737,
      "loss": 0.0707,
      "step": 174
    },
    {
      "epoch": 0.3102836879432624,
      "grad_norm": 0.7107430696487427,
      "learning_rate": 0.0002557245080500894,
      "loss": 0.1413,
      "step": 175
    },
    {
      "epoch": 0.3120567375886525,
      "grad_norm": 0.9405664801597595,
      "learning_rate": 0.00025545617173524147,
      "loss": 0.1555,
      "step": 176
    },
    {
      "epoch": 0.31382978723404253,
      "grad_norm": 0.9159305095672607,
      "learning_rate": 0.00025518783542039354,
      "loss": 0.1827,
      "step": 177
    },
    {
      "epoch": 0.31560283687943264,
      "grad_norm": 1.2095725536346436,
      "learning_rate": 0.0002549194991055456,
      "loss": 0.2713,
      "step": 178
    },
    {
      "epoch": 0.3173758865248227,
      "grad_norm": 0.5971267223358154,
      "learning_rate": 0.00025465116279069764,
      "loss": 0.2094,
      "step": 179
    },
    {
      "epoch": 0.3191489361702128,
      "grad_norm": 0.5078379511833191,
      "learning_rate": 0.0002543828264758497,
      "loss": 0.1102,
      "step": 180
    },
    {
      "epoch": 0.32092198581560283,
      "grad_norm": 0.5540158748626709,
      "learning_rate": 0.0002541144901610018,
      "loss": 0.1408,
      "step": 181
    },
    {
      "epoch": 0.32269503546099293,
      "grad_norm": 0.3938644826412201,
      "learning_rate": 0.0002538461538461538,
      "loss": 0.1044,
      "step": 182
    },
    {
      "epoch": 0.324468085106383,
      "grad_norm": 0.43856045603752136,
      "learning_rate": 0.0002535778175313059,
      "loss": 0.1069,
      "step": 183
    },
    {
      "epoch": 0.3262411347517731,
      "grad_norm": 0.6097897887229919,
      "learning_rate": 0.0002533094812164579,
      "loss": 0.1447,
      "step": 184
    },
    {
      "epoch": 0.3280141843971631,
      "grad_norm": 0.6639567613601685,
      "learning_rate": 0.00025304114490161,
      "loss": 0.1622,
      "step": 185
    },
    {
      "epoch": 0.32978723404255317,
      "grad_norm": 0.4464520812034607,
      "learning_rate": 0.00025277280858676207,
      "loss": 0.1113,
      "step": 186
    },
    {
      "epoch": 0.33156028368794327,
      "grad_norm": 0.45327168703079224,
      "learning_rate": 0.00025250447227191414,
      "loss": 0.1121,
      "step": 187
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.4880326986312866,
      "learning_rate": 0.00025223613595706616,
      "loss": 0.1055,
      "step": 188
    },
    {
      "epoch": 0.3351063829787234,
      "grad_norm": 0.44586634635925293,
      "learning_rate": 0.00025196779964221824,
      "loss": 0.1544,
      "step": 189
    },
    {
      "epoch": 0.33687943262411346,
      "grad_norm": 0.7135107517242432,
      "learning_rate": 0.00025169946332737026,
      "loss": 0.1047,
      "step": 190
    },
    {
      "epoch": 0.33865248226950356,
      "grad_norm": 0.5775635838508606,
      "learning_rate": 0.00025143112701252234,
      "loss": 0.175,
      "step": 191
    },
    {
      "epoch": 0.3404255319148936,
      "grad_norm": 0.6252586841583252,
      "learning_rate": 0.0002511627906976744,
      "loss": 0.1478,
      "step": 192
    },
    {
      "epoch": 0.3421985815602837,
      "grad_norm": 2.061342477798462,
      "learning_rate": 0.00025089445438282644,
      "loss": 0.1931,
      "step": 193
    },
    {
      "epoch": 0.34397163120567376,
      "grad_norm": 0.47072601318359375,
      "learning_rate": 0.0002506261180679785,
      "loss": 0.0928,
      "step": 194
    },
    {
      "epoch": 0.34574468085106386,
      "grad_norm": 0.5164952278137207,
      "learning_rate": 0.00025035778175313053,
      "loss": 0.1287,
      "step": 195
    },
    {
      "epoch": 0.3475177304964539,
      "grad_norm": 0.44344618916511536,
      "learning_rate": 0.0002500894454382826,
      "loss": 0.0924,
      "step": 196
    },
    {
      "epoch": 0.34929078014184395,
      "grad_norm": 0.4380563795566559,
      "learning_rate": 0.0002498211091234347,
      "loss": 0.1393,
      "step": 197
    },
    {
      "epoch": 0.35106382978723405,
      "grad_norm": 0.3074440360069275,
      "learning_rate": 0.00024955277280858676,
      "loss": 0.04,
      "step": 198
    },
    {
      "epoch": 0.3528368794326241,
      "grad_norm": 0.25228482484817505,
      "learning_rate": 0.0002492844364937388,
      "loss": 0.037,
      "step": 199
    },
    {
      "epoch": 0.3546099290780142,
      "grad_norm": 0.4954518973827362,
      "learning_rate": 0.00024901610017889086,
      "loss": 0.1563,
      "step": 200
    },
    {
      "epoch": 0.35638297872340424,
      "grad_norm": 9.530560493469238,
      "learning_rate": 0.00024874776386404294,
      "loss": 0.6406,
      "step": 201
    },
    {
      "epoch": 0.35815602836879434,
      "grad_norm": 1.2426095008850098,
      "learning_rate": 0.00024847942754919496,
      "loss": 0.5336,
      "step": 202
    },
    {
      "epoch": 0.3599290780141844,
      "grad_norm": 1.4315325021743774,
      "learning_rate": 0.00024821109123434703,
      "loss": 0.4691,
      "step": 203
    },
    {
      "epoch": 0.3617021276595745,
      "grad_norm": 1.3048845529556274,
      "learning_rate": 0.00024794275491949906,
      "loss": 0.4098,
      "step": 204
    },
    {
      "epoch": 0.36347517730496454,
      "grad_norm": 1.1803832054138184,
      "learning_rate": 0.00024767441860465113,
      "loss": 0.3023,
      "step": 205
    },
    {
      "epoch": 0.36524822695035464,
      "grad_norm": 0.895564079284668,
      "learning_rate": 0.0002474060822898032,
      "loss": 0.2474,
      "step": 206
    },
    {
      "epoch": 0.3670212765957447,
      "grad_norm": 0.7290531992912292,
      "learning_rate": 0.0002471377459749553,
      "loss": 0.2365,
      "step": 207
    },
    {
      "epoch": 0.36879432624113473,
      "grad_norm": 0.5765687823295593,
      "learning_rate": 0.0002468694096601073,
      "loss": 0.2233,
      "step": 208
    },
    {
      "epoch": 0.37056737588652483,
      "grad_norm": 1.518298625946045,
      "learning_rate": 0.0002466010733452594,
      "loss": 0.3787,
      "step": 209
    },
    {
      "epoch": 0.3723404255319149,
      "grad_norm": 0.9093344211578369,
      "learning_rate": 0.00024633273703041146,
      "loss": 0.2682,
      "step": 210
    },
    {
      "epoch": 0.374113475177305,
      "grad_norm": 0.9133920669555664,
      "learning_rate": 0.0002460644007155635,
      "loss": 0.2236,
      "step": 211
    },
    {
      "epoch": 0.375886524822695,
      "grad_norm": 0.9125071167945862,
      "learning_rate": 0.00024579606440071556,
      "loss": 0.344,
      "step": 212
    },
    {
      "epoch": 0.3776595744680851,
      "grad_norm": 0.9353806972503662,
      "learning_rate": 0.0002455277280858676,
      "loss": 0.3213,
      "step": 213
    },
    {
      "epoch": 0.37943262411347517,
      "grad_norm": 0.32626235485076904,
      "learning_rate": 0.00024525939177101966,
      "loss": 0.089,
      "step": 214
    },
    {
      "epoch": 0.38120567375886527,
      "grad_norm": 0.722865104675293,
      "learning_rate": 0.0002449910554561717,
      "loss": 0.2475,
      "step": 215
    },
    {
      "epoch": 0.3829787234042553,
      "grad_norm": 1.2499586343765259,
      "learning_rate": 0.00024472271914132375,
      "loss": 0.1726,
      "step": 216
    },
    {
      "epoch": 0.38475177304964536,
      "grad_norm": 0.8918619155883789,
      "learning_rate": 0.00024445438282647583,
      "loss": 0.1832,
      "step": 217
    },
    {
      "epoch": 0.38652482269503546,
      "grad_norm": 0.9703326225280762,
      "learning_rate": 0.0002441860465116279,
      "loss": 0.3934,
      "step": 218
    },
    {
      "epoch": 0.3882978723404255,
      "grad_norm": 0.9967647194862366,
      "learning_rate": 0.00024391771019677995,
      "loss": 0.2148,
      "step": 219
    },
    {
      "epoch": 0.3900709219858156,
      "grad_norm": 0.6736385226249695,
      "learning_rate": 0.000243649373881932,
      "loss": 0.2019,
      "step": 220
    },
    {
      "epoch": 0.39184397163120566,
      "grad_norm": 0.6814044117927551,
      "learning_rate": 0.00024338103756708408,
      "loss": 0.3198,
      "step": 221
    },
    {
      "epoch": 0.39361702127659576,
      "grad_norm": 0.6858790516853333,
      "learning_rate": 0.0002431127012522361,
      "loss": 0.2017,
      "step": 222
    },
    {
      "epoch": 0.3953900709219858,
      "grad_norm": 0.4301723837852478,
      "learning_rate": 0.00024284436493738818,
      "loss": 0.1283,
      "step": 223
    },
    {
      "epoch": 0.3971631205673759,
      "grad_norm": 0.47749027609825134,
      "learning_rate": 0.00024257602862254023,
      "loss": 0.1513,
      "step": 224
    },
    {
      "epoch": 0.39893617021276595,
      "grad_norm": 0.5345844626426697,
      "learning_rate": 0.0002423076923076923,
      "loss": 0.0872,
      "step": 225
    },
    {
      "epoch": 0.40070921985815605,
      "grad_norm": 0.5137667059898376,
      "learning_rate": 0.00024203935599284433,
      "loss": 0.0726,
      "step": 226
    },
    {
      "epoch": 0.40070921985815605,
      "eval_loss": 0.08563709259033203,
      "eval_runtime": 47.4226,
      "eval_samples_per_second": 2.657,
      "eval_steps_per_second": 2.657,
      "step": 226
    },
    {
      "epoch": 0.4024822695035461,
      "grad_norm": 0.5459873080253601,
      "learning_rate": 0.0002417710196779964,
      "loss": 0.1341,
      "step": 227
    },
    {
      "epoch": 0.40425531914893614,
      "grad_norm": 0.7590123414993286,
      "learning_rate": 0.00024150268336314848,
      "loss": 0.2418,
      "step": 228
    },
    {
      "epoch": 0.40602836879432624,
      "grad_norm": 0.8846575617790222,
      "learning_rate": 0.00024123434704830053,
      "loss": 0.1606,
      "step": 229
    },
    {
      "epoch": 0.4078014184397163,
      "grad_norm": 0.7346479296684265,
      "learning_rate": 0.00024096601073345258,
      "loss": 0.2458,
      "step": 230
    },
    {
      "epoch": 0.4095744680851064,
      "grad_norm": 0.7012773752212524,
      "learning_rate": 0.00024069767441860462,
      "loss": 0.1353,
      "step": 231
    },
    {
      "epoch": 0.41134751773049644,
      "grad_norm": 0.5895172357559204,
      "learning_rate": 0.0002404293381037567,
      "loss": 0.2138,
      "step": 232
    },
    {
      "epoch": 0.41312056737588654,
      "grad_norm": 0.6732155680656433,
      "learning_rate": 0.00024016100178890872,
      "loss": 0.0906,
      "step": 233
    },
    {
      "epoch": 0.4148936170212766,
      "grad_norm": 0.38057422637939453,
      "learning_rate": 0.0002398926654740608,
      "loss": 0.0973,
      "step": 234
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.4375483989715576,
      "learning_rate": 0.00023962432915921285,
      "loss": 0.0895,
      "step": 235
    },
    {
      "epoch": 0.41843971631205673,
      "grad_norm": 0.45380473136901855,
      "learning_rate": 0.00023935599284436492,
      "loss": 0.1046,
      "step": 236
    },
    {
      "epoch": 0.42021276595744683,
      "grad_norm": 0.38564634323120117,
      "learning_rate": 0.00023908765652951695,
      "loss": 0.0899,
      "step": 237
    },
    {
      "epoch": 0.4219858156028369,
      "grad_norm": 0.837587296962738,
      "learning_rate": 0.00023881932021466902,
      "loss": 0.2611,
      "step": 238
    },
    {
      "epoch": 0.4237588652482269,
      "grad_norm": 0.9749343991279602,
      "learning_rate": 0.0002385509838998211,
      "loss": 0.1351,
      "step": 239
    },
    {
      "epoch": 0.425531914893617,
      "grad_norm": 0.7198343873023987,
      "learning_rate": 0.00023828264758497315,
      "loss": 0.1716,
      "step": 240
    },
    {
      "epoch": 0.42730496453900707,
      "grad_norm": 0.6409445405006409,
      "learning_rate": 0.00023801431127012522,
      "loss": 0.1539,
      "step": 241
    },
    {
      "epoch": 0.42907801418439717,
      "grad_norm": 0.6468048095703125,
      "learning_rate": 0.00023774597495527724,
      "loss": 0.2017,
      "step": 242
    },
    {
      "epoch": 0.4308510638297872,
      "grad_norm": 0.3550443947315216,
      "learning_rate": 0.00023747763864042932,
      "loss": 0.102,
      "step": 243
    },
    {
      "epoch": 0.4326241134751773,
      "grad_norm": 0.5282692313194275,
      "learning_rate": 0.00023720930232558137,
      "loss": 0.0795,
      "step": 244
    },
    {
      "epoch": 0.43439716312056736,
      "grad_norm": 0.4462401568889618,
      "learning_rate": 0.00023694096601073345,
      "loss": 0.1051,
      "step": 245
    },
    {
      "epoch": 0.43617021276595747,
      "grad_norm": 0.33794382214546204,
      "learning_rate": 0.00023667262969588547,
      "loss": 0.0688,
      "step": 246
    },
    {
      "epoch": 0.4379432624113475,
      "grad_norm": 0.3550000786781311,
      "learning_rate": 0.00023640429338103754,
      "loss": 0.0511,
      "step": 247
    },
    {
      "epoch": 0.4397163120567376,
      "grad_norm": 0.2764873206615448,
      "learning_rate": 0.00023613595706618962,
      "loss": 0.0336,
      "step": 248
    },
    {
      "epoch": 0.44148936170212766,
      "grad_norm": 0.2158241719007492,
      "learning_rate": 0.00023586762075134167,
      "loss": 0.0275,
      "step": 249
    },
    {
      "epoch": 0.4432624113475177,
      "grad_norm": 0.41014501452445984,
      "learning_rate": 0.00023559928443649372,
      "loss": 0.1071,
      "step": 250
    },
    {
      "epoch": 0.4450354609929078,
      "grad_norm": 1.6365233659744263,
      "learning_rate": 0.00023533094812164577,
      "loss": 0.5651,
      "step": 251
    },
    {
      "epoch": 0.44680851063829785,
      "grad_norm": 2.2182910442352295,
      "learning_rate": 0.00023506261180679784,
      "loss": 0.4931,
      "step": 252
    },
    {
      "epoch": 0.44858156028368795,
      "grad_norm": 0.7922513484954834,
      "learning_rate": 0.0002347942754919499,
      "loss": 0.165,
      "step": 253
    },
    {
      "epoch": 0.450354609929078,
      "grad_norm": 1.0960763692855835,
      "learning_rate": 0.00023452593917710194,
      "loss": 0.2092,
      "step": 254
    },
    {
      "epoch": 0.4521276595744681,
      "grad_norm": 0.9665970802307129,
      "learning_rate": 0.000234257602862254,
      "loss": 0.3383,
      "step": 255
    },
    {
      "epoch": 0.45390070921985815,
      "grad_norm": 0.9593271613121033,
      "learning_rate": 0.00023398926654740607,
      "loss": 0.2608,
      "step": 256
    },
    {
      "epoch": 0.45567375886524825,
      "grad_norm": 0.9603182673454285,
      "learning_rate": 0.00023372093023255814,
      "loss": 0.3786,
      "step": 257
    },
    {
      "epoch": 0.4574468085106383,
      "grad_norm": 0.6080289483070374,
      "learning_rate": 0.00023345259391771016,
      "loss": 0.1903,
      "step": 258
    },
    {
      "epoch": 0.4592198581560284,
      "grad_norm": 0.4713785946369171,
      "learning_rate": 0.00023318425760286224,
      "loss": 0.1438,
      "step": 259
    },
    {
      "epoch": 0.46099290780141844,
      "grad_norm": 0.7215020656585693,
      "learning_rate": 0.0002329159212880143,
      "loss": 0.1979,
      "step": 260
    },
    {
      "epoch": 0.4627659574468085,
      "grad_norm": 0.7383154630661011,
      "learning_rate": 0.00023264758497316637,
      "loss": 0.3316,
      "step": 261
    },
    {
      "epoch": 0.4645390070921986,
      "grad_norm": 0.8826196789741516,
      "learning_rate": 0.0002323792486583184,
      "loss": 0.3337,
      "step": 262
    },
    {
      "epoch": 0.46631205673758863,
      "grad_norm": 0.6072385907173157,
      "learning_rate": 0.00023211091234347046,
      "loss": 0.2585,
      "step": 263
    },
    {
      "epoch": 0.46808510638297873,
      "grad_norm": 0.5861011743545532,
      "learning_rate": 0.0002318425760286225,
      "loss": 0.1952,
      "step": 264
    },
    {
      "epoch": 0.4698581560283688,
      "grad_norm": 0.8113440871238708,
      "learning_rate": 0.0002315742397137746,
      "loss": 0.1877,
      "step": 265
    },
    {
      "epoch": 0.4716312056737589,
      "grad_norm": 0.3667961359024048,
      "learning_rate": 0.00023130590339892664,
      "loss": 0.0977,
      "step": 266
    },
    {
      "epoch": 0.4734042553191489,
      "grad_norm": 0.4013451933860779,
      "learning_rate": 0.0002310375670840787,
      "loss": 0.0988,
      "step": 267
    },
    {
      "epoch": 0.475177304964539,
      "grad_norm": 0.4402759373188019,
      "learning_rate": 0.00023076923076923076,
      "loss": 0.1747,
      "step": 268
    },
    {
      "epoch": 0.4769503546099291,
      "grad_norm": 0.5780351161956787,
      "learning_rate": 0.0002305008944543828,
      "loss": 0.1446,
      "step": 269
    },
    {
      "epoch": 0.4787234042553192,
      "grad_norm": 1.0400667190551758,
      "learning_rate": 0.00023023255813953486,
      "loss": 0.3404,
      "step": 270
    },
    {
      "epoch": 0.4804964539007092,
      "grad_norm": 0.734530508518219,
      "learning_rate": 0.0002299642218246869,
      "loss": 0.1938,
      "step": 271
    },
    {
      "epoch": 0.48226950354609927,
      "grad_norm": 0.4358377158641815,
      "learning_rate": 0.000229695885509839,
      "loss": 0.1514,
      "step": 272
    },
    {
      "epoch": 0.48404255319148937,
      "grad_norm": 0.766135036945343,
      "learning_rate": 0.00022942754919499104,
      "loss": 0.2335,
      "step": 273
    },
    {
      "epoch": 0.4858156028368794,
      "grad_norm": 0.8013383746147156,
      "learning_rate": 0.00022915921288014308,
      "loss": 0.2433,
      "step": 274
    },
    {
      "epoch": 0.4875886524822695,
      "grad_norm": 0.48911821842193604,
      "learning_rate": 0.00022889087656529513,
      "loss": 0.1466,
      "step": 275
    },
    {
      "epoch": 0.48936170212765956,
      "grad_norm": 0.3731197118759155,
      "learning_rate": 0.0002286225402504472,
      "loss": 0.095,
      "step": 276
    },
    {
      "epoch": 0.49113475177304966,
      "grad_norm": 0.4375539720058441,
      "learning_rate": 0.00022835420393559929,
      "loss": 0.1066,
      "step": 277
    },
    {
      "epoch": 0.4929078014184397,
      "grad_norm": 0.5769888758659363,
      "learning_rate": 0.0002280858676207513,
      "loss": 0.1959,
      "step": 278
    },
    {
      "epoch": 0.4946808510638298,
      "grad_norm": 0.6133276224136353,
      "learning_rate": 0.00022781753130590338,
      "loss": 0.1216,
      "step": 279
    },
    {
      "epoch": 0.49645390070921985,
      "grad_norm": 0.8098652362823486,
      "learning_rate": 0.00022754919499105543,
      "loss": 0.223,
      "step": 280
    },
    {
      "epoch": 0.49822695035460995,
      "grad_norm": 0.38705819845199585,
      "learning_rate": 0.0002272808586762075,
      "loss": 0.0888,
      "step": 281
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.37354519963264465,
      "learning_rate": 0.00022701252236135953,
      "loss": 0.1013,
      "step": 282
    },
    {
      "epoch": 0.50177304964539,
      "grad_norm": 0.6955541968345642,
      "learning_rate": 0.0002267441860465116,
      "loss": 0.1746,
      "step": 283
    },
    {
      "epoch": 0.5035460992907801,
      "grad_norm": 0.44892531633377075,
      "learning_rate": 0.00022647584973166366,
      "loss": 0.123,
      "step": 284
    },
    {
      "epoch": 0.5053191489361702,
      "grad_norm": 0.7901389598846436,
      "learning_rate": 0.00022620751341681573,
      "loss": 0.2863,
      "step": 285
    },
    {
      "epoch": 0.5070921985815603,
      "grad_norm": 0.6648136973381042,
      "learning_rate": 0.00022593917710196778,
      "loss": 0.1577,
      "step": 286
    },
    {
      "epoch": 0.5088652482269503,
      "grad_norm": 0.3378174602985382,
      "learning_rate": 0.00022567084078711983,
      "loss": 0.0916,
      "step": 287
    },
    {
      "epoch": 0.5106382978723404,
      "grad_norm": 0.46413183212280273,
      "learning_rate": 0.0002254025044722719,
      "loss": 0.1292,
      "step": 288
    },
    {
      "epoch": 0.5124113475177305,
      "grad_norm": 0.5988323092460632,
      "learning_rate": 0.00022513416815742396,
      "loss": 0.1586,
      "step": 289
    },
    {
      "epoch": 0.5141843971631206,
      "grad_norm": 0.5859420299530029,
      "learning_rate": 0.000224865831842576,
      "loss": 0.1296,
      "step": 290
    },
    {
      "epoch": 0.5159574468085106,
      "grad_norm": 0.4029462933540344,
      "learning_rate": 0.00022459749552772805,
      "loss": 0.1072,
      "step": 291
    },
    {
      "epoch": 0.5177304964539007,
      "grad_norm": 0.28882545232772827,
      "learning_rate": 0.00022432915921288013,
      "loss": 0.0747,
      "step": 292
    },
    {
      "epoch": 0.5195035460992907,
      "grad_norm": 0.35556840896606445,
      "learning_rate": 0.00022406082289803218,
      "loss": 0.0463,
      "step": 293
    },
    {
      "epoch": 0.5212765957446809,
      "grad_norm": 0.5565528273582458,
      "learning_rate": 0.00022379248658318423,
      "loss": 0.0675,
      "step": 294
    },
    {
      "epoch": 0.5230496453900709,
      "grad_norm": 0.5911768078804016,
      "learning_rate": 0.0002235241502683363,
      "loss": 0.1442,
      "step": 295
    },
    {
      "epoch": 0.524822695035461,
      "grad_norm": 0.4034654498100281,
      "learning_rate": 0.00022325581395348835,
      "loss": 0.1111,
      "step": 296
    },
    {
      "epoch": 0.526595744680851,
      "grad_norm": 0.35688602924346924,
      "learning_rate": 0.00022298747763864043,
      "loss": 0.0828,
      "step": 297
    },
    {
      "epoch": 0.5283687943262412,
      "grad_norm": 0.6449804306030273,
      "learning_rate": 0.00022271914132379245,
      "loss": 0.2104,
      "step": 298
    },
    {
      "epoch": 0.5301418439716312,
      "grad_norm": 0.5375797152519226,
      "learning_rate": 0.00022245080500894453,
      "loss": 0.1546,
      "step": 299
    },
    {
      "epoch": 0.5319148936170213,
      "grad_norm": 0.18811915814876556,
      "learning_rate": 0.00022218246869409658,
      "loss": 0.0246,
      "step": 300
    },
    {
      "epoch": 0.5336879432624113,
      "grad_norm": 1.5281504392623901,
      "learning_rate": 0.00022191413237924865,
      "loss": 0.5446,
      "step": 301
    },
    {
      "epoch": 0.5354609929078015,
      "grad_norm": 1.5097373723983765,
      "learning_rate": 0.00022164579606440067,
      "loss": 0.4033,
      "step": 302
    },
    {
      "epoch": 0.5372340425531915,
      "grad_norm": 1.4294488430023193,
      "learning_rate": 0.00022137745974955275,
      "loss": 0.3835,
      "step": 303
    },
    {
      "epoch": 0.5390070921985816,
      "grad_norm": 0.7668700218200684,
      "learning_rate": 0.00022110912343470483,
      "loss": 0.2103,
      "step": 304
    },
    {
      "epoch": 0.5407801418439716,
      "grad_norm": 0.9126233458518982,
      "learning_rate": 0.00022084078711985688,
      "loss": 0.39,
      "step": 305
    },
    {
      "epoch": 0.5425531914893617,
      "grad_norm": 0.7275278568267822,
      "learning_rate": 0.00022057245080500892,
      "loss": 0.2128,
      "step": 306
    },
    {
      "epoch": 0.5443262411347518,
      "grad_norm": 0.6632657647132874,
      "learning_rate": 0.00022030411449016097,
      "loss": 0.1819,
      "step": 307
    },
    {
      "epoch": 0.5460992907801419,
      "grad_norm": 0.5265213847160339,
      "learning_rate": 0.00022003577817531305,
      "loss": 0.1879,
      "step": 308
    },
    {
      "epoch": 0.5478723404255319,
      "grad_norm": 0.4046447277069092,
      "learning_rate": 0.0002197674418604651,
      "loss": 0.1225,
      "step": 309
    },
    {
      "epoch": 0.549645390070922,
      "grad_norm": 0.7889742851257324,
      "learning_rate": 0.00021949910554561715,
      "loss": 0.3022,
      "step": 310
    },
    {
      "epoch": 0.5514184397163121,
      "grad_norm": 0.739321768283844,
      "learning_rate": 0.0002192307692307692,
      "loss": 0.1621,
      "step": 311
    },
    {
      "epoch": 0.5531914893617021,
      "grad_norm": 0.5458974838256836,
      "learning_rate": 0.00021896243291592127,
      "loss": 0.1505,
      "step": 312
    },
    {
      "epoch": 0.5549645390070922,
      "grad_norm": 0.6029899716377258,
      "learning_rate": 0.00021869409660107332,
      "loss": 0.2114,
      "step": 313
    },
    {
      "epoch": 0.5567375886524822,
      "grad_norm": 0.830324113368988,
      "learning_rate": 0.00021842576028622537,
      "loss": 0.2397,
      "step": 314
    },
    {
      "epoch": 0.5585106382978723,
      "grad_norm": 0.7340556383132935,
      "learning_rate": 0.00021815742397137745,
      "loss": 0.1606,
      "step": 315
    },
    {
      "epoch": 0.5602836879432624,
      "grad_norm": 0.44604039192199707,
      "learning_rate": 0.0002178890876565295,
      "loss": 0.1192,
      "step": 316
    },
    {
      "epoch": 0.5620567375886525,
      "grad_norm": 0.37659141421318054,
      "learning_rate": 0.00021762075134168157,
      "loss": 0.1112,
      "step": 317
    },
    {
      "epoch": 0.5638297872340425,
      "grad_norm": 0.8448424935340881,
      "learning_rate": 0.0002173524150268336,
      "loss": 0.1799,
      "step": 318
    },
    {
      "epoch": 0.5656028368794326,
      "grad_norm": 0.5874999761581421,
      "learning_rate": 0.00021708407871198567,
      "loss": 0.2011,
      "step": 319
    },
    {
      "epoch": 0.5673758865248227,
      "grad_norm": 0.5802026391029358,
      "learning_rate": 0.00021681574239713772,
      "loss": 0.1725,
      "step": 320
    },
    {
      "epoch": 0.5691489361702128,
      "grad_norm": 0.5646753907203674,
      "learning_rate": 0.0002165474060822898,
      "loss": 0.1464,
      "step": 321
    },
    {
      "epoch": 0.5709219858156028,
      "grad_norm": 0.5598042011260986,
      "learning_rate": 0.00021627906976744182,
      "loss": 0.1135,
      "step": 322
    },
    {
      "epoch": 0.5726950354609929,
      "grad_norm": 0.35759246349334717,
      "learning_rate": 0.0002160107334525939,
      "loss": 0.079,
      "step": 323
    },
    {
      "epoch": 0.574468085106383,
      "grad_norm": 0.8581865429878235,
      "learning_rate": 0.00021574239713774597,
      "loss": 0.2422,
      "step": 324
    },
    {
      "epoch": 0.5762411347517731,
      "grad_norm": 0.7060332298278809,
      "learning_rate": 0.00021547406082289802,
      "loss": 0.2283,
      "step": 325
    },
    {
      "epoch": 0.5780141843971631,
      "grad_norm": 0.7340462803840637,
      "learning_rate": 0.00021520572450805007,
      "loss": 0.1915,
      "step": 326
    },
    {
      "epoch": 0.5797872340425532,
      "grad_norm": 0.42532655596733093,
      "learning_rate": 0.00021493738819320212,
      "loss": 0.1125,
      "step": 327
    },
    {
      "epoch": 0.5815602836879432,
      "grad_norm": 0.2977962791919708,
      "learning_rate": 0.0002146690518783542,
      "loss": 0.0812,
      "step": 328
    },
    {
      "epoch": 0.5833333333333334,
      "grad_norm": 0.3650486171245575,
      "learning_rate": 0.00021440071556350624,
      "loss": 0.0999,
      "step": 329
    },
    {
      "epoch": 0.5851063829787234,
      "grad_norm": 0.6662601232528687,
      "learning_rate": 0.0002141323792486583,
      "loss": 0.2518,
      "step": 330
    },
    {
      "epoch": 0.5868794326241135,
      "grad_norm": 0.46315595507621765,
      "learning_rate": 0.00021386404293381034,
      "loss": 0.1164,
      "step": 331
    },
    {
      "epoch": 0.5886524822695035,
      "grad_norm": 0.5522826910018921,
      "learning_rate": 0.00021359570661896242,
      "loss": 0.1563,
      "step": 332
    },
    {
      "epoch": 0.5904255319148937,
      "grad_norm": 0.6960483193397522,
      "learning_rate": 0.0002133273703041145,
      "loss": 0.1998,
      "step": 333
    },
    {
      "epoch": 0.5921985815602837,
      "grad_norm": 0.46060803532600403,
      "learning_rate": 0.00021305903398926651,
      "loss": 0.1235,
      "step": 334
    },
    {
      "epoch": 0.5939716312056738,
      "grad_norm": 0.8451377749443054,
      "learning_rate": 0.0002127906976744186,
      "loss": 0.2462,
      "step": 335
    },
    {
      "epoch": 0.5957446808510638,
      "grad_norm": 0.4865632653236389,
      "learning_rate": 0.00021252236135957064,
      "loss": 0.1759,
      "step": 336
    },
    {
      "epoch": 0.5975177304964538,
      "grad_norm": 0.5368924736976624,
      "learning_rate": 0.00021225402504472272,
      "loss": 0.1632,
      "step": 337
    },
    {
      "epoch": 0.599290780141844,
      "grad_norm": 0.5031744837760925,
      "learning_rate": 0.00021198568872987474,
      "loss": 0.2001,
      "step": 338
    },
    {
      "epoch": 0.601063829787234,
      "grad_norm": 0.35617366433143616,
      "learning_rate": 0.00021171735241502681,
      "loss": 0.0905,
      "step": 339
    },
    {
      "epoch": 0.6028368794326241,
      "grad_norm": 0.4215497076511383,
      "learning_rate": 0.00021144901610017886,
      "loss": 0.1474,
      "step": 340
    },
    {
      "epoch": 0.6046099290780141,
      "grad_norm": 0.3073301911354065,
      "learning_rate": 0.00021118067978533094,
      "loss": 0.053,
      "step": 341
    },
    {
      "epoch": 0.6063829787234043,
      "grad_norm": 0.30299848318099976,
      "learning_rate": 0.00021091234347048301,
      "loss": 0.0715,
      "step": 342
    },
    {
      "epoch": 0.6081560283687943,
      "grad_norm": 0.26558998227119446,
      "learning_rate": 0.00021064400715563504,
      "loss": 0.0573,
      "step": 343
    },
    {
      "epoch": 0.6099290780141844,
      "grad_norm": 0.21946674585342407,
      "learning_rate": 0.0002103756708407871,
      "loss": 0.0502,
      "step": 344
    },
    {
      "epoch": 0.6117021276595744,
      "grad_norm": 0.48567119240760803,
      "learning_rate": 0.00021010733452593916,
      "loss": 0.1224,
      "step": 345
    },
    {
      "epoch": 0.6134751773049646,
      "grad_norm": 0.444938600063324,
      "learning_rate": 0.00020983899821109124,
      "loss": 0.0899,
      "step": 346
    },
    {
      "epoch": 0.6152482269503546,
      "grad_norm": 0.3576812744140625,
      "learning_rate": 0.00020957066189624326,
      "loss": 0.1104,
      "step": 347
    },
    {
      "epoch": 0.6170212765957447,
      "grad_norm": 0.5801045298576355,
      "learning_rate": 0.00020930232558139534,
      "loss": 0.2192,
      "step": 348
    },
    {
      "epoch": 0.6187943262411347,
      "grad_norm": 0.3965771198272705,
      "learning_rate": 0.00020903398926654739,
      "loss": 0.1025,
      "step": 349
    },
    {
      "epoch": 0.6205673758865248,
      "grad_norm": 0.32997211813926697,
      "learning_rate": 0.00020876565295169943,
      "loss": 0.0392,
      "step": 350
    },
    {
      "epoch": 0.6223404255319149,
      "grad_norm": 2.3874707221984863,
      "learning_rate": 0.00020849731663685148,
      "loss": 0.4267,
      "step": 351
    },
    {
      "epoch": 0.624113475177305,
      "grad_norm": 1.0561052560806274,
      "learning_rate": 0.00020822898032200356,
      "loss": 0.3531,
      "step": 352
    },
    {
      "epoch": 0.625886524822695,
      "grad_norm": 1.0714682340621948,
      "learning_rate": 0.00020796064400715564,
      "loss": 0.208,
      "step": 353
    },
    {
      "epoch": 0.6276595744680851,
      "grad_norm": 1.0135624408721924,
      "learning_rate": 0.00020769230769230766,
      "loss": 0.2722,
      "step": 354
    },
    {
      "epoch": 0.6294326241134752,
      "grad_norm": 0.4171317219734192,
      "learning_rate": 0.00020742397137745973,
      "loss": 0.1137,
      "step": 355
    },
    {
      "epoch": 0.6312056737588653,
      "grad_norm": 0.7069351673126221,
      "learning_rate": 0.00020715563506261178,
      "loss": 0.1919,
      "step": 356
    },
    {
      "epoch": 0.6329787234042553,
      "grad_norm": 0.5680619478225708,
      "learning_rate": 0.00020688729874776386,
      "loss": 0.1884,
      "step": 357
    },
    {
      "epoch": 0.6347517730496454,
      "grad_norm": 0.8620048761367798,
      "learning_rate": 0.00020661896243291588,
      "loss": 0.3209,
      "step": 358
    },
    {
      "epoch": 0.6365248226950354,
      "grad_norm": 0.4163174331188202,
      "learning_rate": 0.00020635062611806796,
      "loss": 0.1553,
      "step": 359
    },
    {
      "epoch": 0.6382978723404256,
      "grad_norm": 0.7022566199302673,
      "learning_rate": 0.00020608228980322,
      "loss": 0.2839,
      "step": 360
    },
    {
      "epoch": 0.6400709219858156,
      "grad_norm": 0.35048022866249084,
      "learning_rate": 0.00020581395348837208,
      "loss": 0.1168,
      "step": 361
    },
    {
      "epoch": 0.6418439716312057,
      "grad_norm": 0.6058384776115417,
      "learning_rate": 0.00020554561717352416,
      "loss": 0.2336,
      "step": 362
    },
    {
      "epoch": 0.6436170212765957,
      "grad_norm": 0.5632460117340088,
      "learning_rate": 0.00020527728085867618,
      "loss": 0.1866,
      "step": 363
    },
    {
      "epoch": 0.6453900709219859,
      "grad_norm": 0.8508513569831848,
      "learning_rate": 0.00020500894454382826,
      "loss": 0.2795,
      "step": 364
    },
    {
      "epoch": 0.6471631205673759,
      "grad_norm": 0.3261764645576477,
      "learning_rate": 0.0002047406082289803,
      "loss": 0.0695,
      "step": 365
    },
    {
      "epoch": 0.648936170212766,
      "grad_norm": 0.6185038089752197,
      "learning_rate": 0.00020447227191413238,
      "loss": 0.2125,
      "step": 366
    },
    {
      "epoch": 0.650709219858156,
      "grad_norm": 0.9981937408447266,
      "learning_rate": 0.0002042039355992844,
      "loss": 0.179,
      "step": 367
    },
    {
      "epoch": 0.6524822695035462,
      "grad_norm": 0.625804603099823,
      "learning_rate": 0.00020393559928443648,
      "loss": 0.2315,
      "step": 368
    },
    {
      "epoch": 0.6542553191489362,
      "grad_norm": 0.460195928812027,
      "learning_rate": 0.00020366726296958853,
      "loss": 0.1352,
      "step": 369
    },
    {
      "epoch": 0.6560283687943262,
      "grad_norm": 0.6661739349365234,
      "learning_rate": 0.00020339892665474058,
      "loss": 0.2227,
      "step": 370
    },
    {
      "epoch": 0.6578014184397163,
      "grad_norm": 0.5500937700271606,
      "learning_rate": 0.00020313059033989265,
      "loss": 0.1519,
      "step": 371
    },
    {
      "epoch": 0.6595744680851063,
      "grad_norm": 0.8262995481491089,
      "learning_rate": 0.0002028622540250447,
      "loss": 0.2965,
      "step": 372
    },
    {
      "epoch": 0.6613475177304965,
      "grad_norm": 0.6892026662826538,
      "learning_rate": 0.00020259391771019678,
      "loss": 0.2244,
      "step": 373
    },
    {
      "epoch": 0.6631205673758865,
      "grad_norm": 0.9893831014633179,
      "learning_rate": 0.0002023255813953488,
      "loss": 0.27,
      "step": 374
    },
    {
      "epoch": 0.6648936170212766,
      "grad_norm": 0.26278918981552124,
      "learning_rate": 0.00020205724508050088,
      "loss": 0.0958,
      "step": 375
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.4749915301799774,
      "learning_rate": 0.00020178890876565293,
      "loss": 0.182,
      "step": 376
    },
    {
      "epoch": 0.6684397163120568,
      "grad_norm": 0.36534926295280457,
      "learning_rate": 0.000201520572450805,
      "loss": 0.0955,
      "step": 377
    },
    {
      "epoch": 0.6702127659574468,
      "grad_norm": 0.28374701738357544,
      "learning_rate": 0.00020125223613595702,
      "loss": 0.083,
      "step": 378
    },
    {
      "epoch": 0.6719858156028369,
      "grad_norm": 0.5432383418083191,
      "learning_rate": 0.0002009838998211091,
      "loss": 0.1533,
      "step": 379
    },
    {
      "epoch": 0.6737588652482269,
      "grad_norm": 0.6451116800308228,
      "learning_rate": 0.00020071556350626118,
      "loss": 0.2346,
      "step": 380
    },
    {
      "epoch": 0.675531914893617,
      "grad_norm": 0.3411319851875305,
      "learning_rate": 0.00020044722719141323,
      "loss": 0.0937,
      "step": 381
    },
    {
      "epoch": 0.6773049645390071,
      "grad_norm": 0.40049058198928833,
      "learning_rate": 0.0002001788908765653,
      "loss": 0.1219,
      "step": 382
    },
    {
      "epoch": 0.6790780141843972,
      "grad_norm": 0.3120768070220947,
      "learning_rate": 0.00019991055456171732,
      "loss": 0.1086,
      "step": 383
    },
    {
      "epoch": 0.6808510638297872,
      "grad_norm": 0.3227909207344055,
      "learning_rate": 0.0001996422182468694,
      "loss": 0.0911,
      "step": 384
    },
    {
      "epoch": 0.6826241134751773,
      "grad_norm": 0.22975340485572815,
      "learning_rate": 0.00019937388193202145,
      "loss": 0.0818,
      "step": 385
    },
    {
      "epoch": 0.6843971631205674,
      "grad_norm": 0.3845633268356323,
      "learning_rate": 0.00019910554561717352,
      "loss": 0.1278,
      "step": 386
    },
    {
      "epoch": 0.6861702127659575,
      "grad_norm": 0.4206799864768982,
      "learning_rate": 0.00019883720930232555,
      "loss": 0.1137,
      "step": 387
    },
    {
      "epoch": 0.6879432624113475,
      "grad_norm": 0.4134272634983063,
      "learning_rate": 0.00019856887298747762,
      "loss": 0.1299,
      "step": 388
    },
    {
      "epoch": 0.6897163120567376,
      "grad_norm": 0.44249987602233887,
      "learning_rate": 0.0001983005366726297,
      "loss": 0.1211,
      "step": 389
    },
    {
      "epoch": 0.6914893617021277,
      "grad_norm": 0.3611302971839905,
      "learning_rate": 0.00019803220035778172,
      "loss": 0.114,
      "step": 390
    },
    {
      "epoch": 0.6932624113475178,
      "grad_norm": 0.47269973158836365,
      "learning_rate": 0.0001977638640429338,
      "loss": 0.1405,
      "step": 391
    },
    {
      "epoch": 0.6950354609929078,
      "grad_norm": 0.47546666860580444,
      "learning_rate": 0.00019749552772808585,
      "loss": 0.1447,
      "step": 392
    },
    {
      "epoch": 0.6968085106382979,
      "grad_norm": 0.3153839111328125,
      "learning_rate": 0.00019722719141323792,
      "loss": 0.0972,
      "step": 393
    },
    {
      "epoch": 0.6985815602836879,
      "grad_norm": 0.23356540501117706,
      "learning_rate": 0.00019695885509838994,
      "loss": 0.0429,
      "step": 394
    },
    {
      "epoch": 0.700354609929078,
      "grad_norm": 0.399263858795166,
      "learning_rate": 0.00019669051878354202,
      "loss": 0.1401,
      "step": 395
    },
    {
      "epoch": 0.7021276595744681,
      "grad_norm": 0.36282962560653687,
      "learning_rate": 0.00019642218246869407,
      "loss": 0.11,
      "step": 396
    },
    {
      "epoch": 0.7039007092198581,
      "grad_norm": 0.4403574764728546,
      "learning_rate": 0.00019615384615384615,
      "loss": 0.0706,
      "step": 397
    },
    {
      "epoch": 0.7056737588652482,
      "grad_norm": 0.5901324152946472,
      "learning_rate": 0.00019588550983899817,
      "loss": 0.1191,
      "step": 398
    },
    {
      "epoch": 0.7074468085106383,
      "grad_norm": 0.2563900351524353,
      "learning_rate": 0.00019561717352415024,
      "loss": 0.0731,
      "step": 399
    },
    {
      "epoch": 0.7092198581560284,
      "grad_norm": 0.4444797933101654,
      "learning_rate": 0.00019534883720930232,
      "loss": 0.1425,
      "step": 400
    },
    {
      "epoch": 0.7109929078014184,
      "grad_norm": 1.0342155694961548,
      "learning_rate": 0.00019508050089445437,
      "loss": 0.3125,
      "step": 401
    },
    {
      "epoch": 0.7127659574468085,
      "grad_norm": 0.6980201005935669,
      "learning_rate": 0.00019481216457960644,
      "loss": 0.2188,
      "step": 402
    },
    {
      "epoch": 0.7145390070921985,
      "grad_norm": 0.740844190120697,
      "learning_rate": 0.00019454382826475847,
      "loss": 0.2467,
      "step": 403
    },
    {
      "epoch": 0.7163120567375887,
      "grad_norm": 0.513054370880127,
      "learning_rate": 0.00019427549194991054,
      "loss": 0.1531,
      "step": 404
    },
    {
      "epoch": 0.7180851063829787,
      "grad_norm": 0.4994530975818634,
      "learning_rate": 0.0001940071556350626,
      "loss": 0.1065,
      "step": 405
    },
    {
      "epoch": 0.7198581560283688,
      "grad_norm": 0.9895260334014893,
      "learning_rate": 0.00019373881932021467,
      "loss": 0.4544,
      "step": 406
    },
    {
      "epoch": 0.7216312056737588,
      "grad_norm": 0.422955721616745,
      "learning_rate": 0.0001934704830053667,
      "loss": 0.1079,
      "step": 407
    },
    {
      "epoch": 0.723404255319149,
      "grad_norm": 0.5568337440490723,
      "learning_rate": 0.00019320214669051877,
      "loss": 0.2162,
      "step": 408
    },
    {
      "epoch": 0.725177304964539,
      "grad_norm": 0.4763684868812561,
      "learning_rate": 0.00019293381037567084,
      "loss": 0.1643,
      "step": 409
    },
    {
      "epoch": 0.7269503546099291,
      "grad_norm": 0.4117484390735626,
      "learning_rate": 0.0001926654740608229,
      "loss": 0.1393,
      "step": 410
    },
    {
      "epoch": 0.7287234042553191,
      "grad_norm": 0.4180298447608948,
      "learning_rate": 0.00019239713774597494,
      "loss": 0.183,
      "step": 411
    },
    {
      "epoch": 0.7304964539007093,
      "grad_norm": 0.42761579155921936,
      "learning_rate": 0.000192128801431127,
      "loss": 0.1463,
      "step": 412
    },
    {
      "epoch": 0.7322695035460993,
      "grad_norm": 0.36335864663124084,
      "learning_rate": 0.00019186046511627906,
      "loss": 0.1032,
      "step": 413
    },
    {
      "epoch": 0.7340425531914894,
      "grad_norm": 0.4711301326751709,
      "learning_rate": 0.0001915921288014311,
      "loss": 0.17,
      "step": 414
    },
    {
      "epoch": 0.7358156028368794,
      "grad_norm": 0.38681793212890625,
      "learning_rate": 0.00019132379248658316,
      "loss": 0.1371,
      "step": 415
    },
    {
      "epoch": 0.7375886524822695,
      "grad_norm": 0.38631147146224976,
      "learning_rate": 0.0001910554561717352,
      "loss": 0.1183,
      "step": 416
    },
    {
      "epoch": 0.7393617021276596,
      "grad_norm": 0.4759841561317444,
      "learning_rate": 0.0001907871198568873,
      "loss": 0.1441,
      "step": 417
    },
    {
      "epoch": 0.7411347517730497,
      "grad_norm": 0.42136093974113464,
      "learning_rate": 0.00019051878354203936,
      "loss": 0.1093,
      "step": 418
    },
    {
      "epoch": 0.7429078014184397,
      "grad_norm": 0.5578286051750183,
      "learning_rate": 0.00019025044722719139,
      "loss": 0.1521,
      "step": 419
    },
    {
      "epoch": 0.7446808510638298,
      "grad_norm": 0.32407546043395996,
      "learning_rate": 0.00018998211091234346,
      "loss": 0.0614,
      "step": 420
    },
    {
      "epoch": 0.7464539007092199,
      "grad_norm": 0.621771514415741,
      "learning_rate": 0.0001897137745974955,
      "loss": 0.2111,
      "step": 421
    },
    {
      "epoch": 0.74822695035461,
      "grad_norm": 0.4861868619918823,
      "learning_rate": 0.0001894454382826476,
      "loss": 0.1138,
      "step": 422
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.6632353067398071,
      "learning_rate": 0.0001891771019677996,
      "loss": 0.2492,
      "step": 423
    },
    {
      "epoch": 0.75177304964539,
      "grad_norm": 0.3670256733894348,
      "learning_rate": 0.00018890876565295169,
      "loss": 0.0927,
      "step": 424
    },
    {
      "epoch": 0.7535460992907801,
      "grad_norm": 0.48544085025787354,
      "learning_rate": 0.00018864042933810373,
      "loss": 0.1422,
      "step": 425
    },
    {
      "epoch": 0.7553191489361702,
      "grad_norm": 0.3215571939945221,
      "learning_rate": 0.0001883720930232558,
      "loss": 0.0729,
      "step": 426
    },
    {
      "epoch": 0.7570921985815603,
      "grad_norm": 0.6928660869598389,
      "learning_rate": 0.00018810375670840786,
      "loss": 0.2033,
      "step": 427
    },
    {
      "epoch": 0.7588652482269503,
      "grad_norm": 0.42845651507377625,
      "learning_rate": 0.0001878354203935599,
      "loss": 0.1536,
      "step": 428
    },
    {
      "epoch": 0.7606382978723404,
      "grad_norm": 0.5783790349960327,
      "learning_rate": 0.00018756708407871198,
      "loss": 0.1573,
      "step": 429
    },
    {
      "epoch": 0.7624113475177305,
      "grad_norm": 0.34763625264167786,
      "learning_rate": 0.00018729874776386403,
      "loss": 0.0891,
      "step": 430
    },
    {
      "epoch": 0.7641843971631206,
      "grad_norm": 0.3298535943031311,
      "learning_rate": 0.00018703041144901608,
      "loss": 0.0836,
      "step": 431
    },
    {
      "epoch": 0.7659574468085106,
      "grad_norm": 0.48353129625320435,
      "learning_rate": 0.00018676207513416813,
      "loss": 0.1624,
      "step": 432
    },
    {
      "epoch": 0.7677304964539007,
      "grad_norm": 0.3654095530509949,
      "learning_rate": 0.0001864937388193202,
      "loss": 0.074,
      "step": 433
    },
    {
      "epoch": 0.7695035460992907,
      "grad_norm": 0.6465492248535156,
      "learning_rate": 0.00018622540250447223,
      "loss": 0.2221,
      "step": 434
    },
    {
      "epoch": 0.7712765957446809,
      "grad_norm": 0.29519444704055786,
      "learning_rate": 0.0001859570661896243,
      "loss": 0.073,
      "step": 435
    },
    {
      "epoch": 0.7730496453900709,
      "grad_norm": 0.39645931124687195,
      "learning_rate": 0.00018568872987477636,
      "loss": 0.0799,
      "step": 436
    },
    {
      "epoch": 0.774822695035461,
      "grad_norm": 0.7929151058197021,
      "learning_rate": 0.00018542039355992843,
      "loss": 0.244,
      "step": 437
    },
    {
      "epoch": 0.776595744680851,
      "grad_norm": 0.5418475866317749,
      "learning_rate": 0.0001851520572450805,
      "loss": 0.1439,
      "step": 438
    },
    {
      "epoch": 0.7783687943262412,
      "grad_norm": 0.4036068320274353,
      "learning_rate": 0.00018488372093023253,
      "loss": 0.1158,
      "step": 439
    },
    {
      "epoch": 0.7801418439716312,
      "grad_norm": 0.33123520016670227,
      "learning_rate": 0.0001846153846153846,
      "loss": 0.1182,
      "step": 440
    },
    {
      "epoch": 0.7819148936170213,
      "grad_norm": 0.3179425299167633,
      "learning_rate": 0.00018434704830053665,
      "loss": 0.0763,
      "step": 441
    },
    {
      "epoch": 0.7836879432624113,
      "grad_norm": 0.25326409935951233,
      "learning_rate": 0.00018407871198568873,
      "loss": 0.0388,
      "step": 442
    },
    {
      "epoch": 0.7854609929078015,
      "grad_norm": 0.4816826581954956,
      "learning_rate": 0.00018381037567084075,
      "loss": 0.1308,
      "step": 443
    },
    {
      "epoch": 0.7872340425531915,
      "grad_norm": 0.2538895010948181,
      "learning_rate": 0.00018354203935599283,
      "loss": 0.0607,
      "step": 444
    },
    {
      "epoch": 0.7890070921985816,
      "grad_norm": 0.3773513436317444,
      "learning_rate": 0.00018327370304114488,
      "loss": 0.1168,
      "step": 445
    },
    {
      "epoch": 0.7907801418439716,
      "grad_norm": 0.48236358165740967,
      "learning_rate": 0.00018300536672629695,
      "loss": 0.1078,
      "step": 446
    },
    {
      "epoch": 0.7925531914893617,
      "grad_norm": 0.37481409311294556,
      "learning_rate": 0.000182737030411449,
      "loss": 0.0847,
      "step": 447
    },
    {
      "epoch": 0.7943262411347518,
      "grad_norm": 0.25582361221313477,
      "learning_rate": 0.00018246869409660105,
      "loss": 0.0518,
      "step": 448
    },
    {
      "epoch": 0.7960992907801419,
      "grad_norm": 0.582918643951416,
      "learning_rate": 0.00018220035778175313,
      "loss": 0.199,
      "step": 449
    },
    {
      "epoch": 0.7978723404255319,
      "grad_norm": 0.3821042478084564,
      "learning_rate": 0.00018193202146690518,
      "loss": 0.1308,
      "step": 450
    },
    {
      "epoch": 0.799645390070922,
      "grad_norm": 0.9046875834465027,
      "learning_rate": 0.00018166368515205723,
      "loss": 0.3907,
      "step": 451
    },
    {
      "epoch": 0.8014184397163121,
      "grad_norm": 0.9091282486915588,
      "learning_rate": 0.00018139534883720928,
      "loss": 0.421,
      "step": 452
    },
    {
      "epoch": 0.8014184397163121,
      "eval_loss": 0.07596753537654877,
      "eval_runtime": 47.273,
      "eval_samples_per_second": 2.665,
      "eval_steps_per_second": 2.665,
      "step": 452
    },
    {
      "epoch": 0.8031914893617021,
      "grad_norm": 0.5998420715332031,
      "learning_rate": 0.00018112701252236135,
      "loss": 0.1944,
      "step": 453
    },
    {
      "epoch": 0.8049645390070922,
      "grad_norm": 0.6365135908126831,
      "learning_rate": 0.00018085867620751337,
      "loss": 0.2445,
      "step": 454
    },
    {
      "epoch": 0.8067375886524822,
      "grad_norm": 0.5594285726547241,
      "learning_rate": 0.00018059033989266545,
      "loss": 0.2096,
      "step": 455
    },
    {
      "epoch": 0.8085106382978723,
      "grad_norm": 0.8409189581871033,
      "learning_rate": 0.00018032200357781753,
      "loss": 0.4065,
      "step": 456
    },
    {
      "epoch": 0.8102836879432624,
      "grad_norm": 0.5661482214927673,
      "learning_rate": 0.00018005366726296957,
      "loss": 0.25,
      "step": 457
    },
    {
      "epoch": 0.8120567375886525,
      "grad_norm": 0.5428692102432251,
      "learning_rate": 0.00017978533094812165,
      "loss": 0.2254,
      "step": 458
    },
    {
      "epoch": 0.8138297872340425,
      "grad_norm": 0.4579595923423767,
      "learning_rate": 0.00017951699463327367,
      "loss": 0.1008,
      "step": 459
    },
    {
      "epoch": 0.8156028368794326,
      "grad_norm": 0.4390023946762085,
      "learning_rate": 0.00017924865831842575,
      "loss": 0.1144,
      "step": 460
    },
    {
      "epoch": 0.8173758865248227,
      "grad_norm": 0.5851403474807739,
      "learning_rate": 0.0001789803220035778,
      "loss": 0.2355,
      "step": 461
    },
    {
      "epoch": 0.8191489361702128,
      "grad_norm": 0.31094157695770264,
      "learning_rate": 0.00017871198568872987,
      "loss": 0.1078,
      "step": 462
    },
    {
      "epoch": 0.8209219858156028,
      "grad_norm": 0.5213750600814819,
      "learning_rate": 0.0001784436493738819,
      "loss": 0.1546,
      "step": 463
    },
    {
      "epoch": 0.8226950354609929,
      "grad_norm": 0.3972003161907196,
      "learning_rate": 0.00017817531305903397,
      "loss": 0.1379,
      "step": 464
    },
    {
      "epoch": 0.824468085106383,
      "grad_norm": 0.35936617851257324,
      "learning_rate": 0.00017790697674418605,
      "loss": 0.1238,
      "step": 465
    },
    {
      "epoch": 0.8262411347517731,
      "grad_norm": 0.31084805727005005,
      "learning_rate": 0.0001776386404293381,
      "loss": 0.0898,
      "step": 466
    },
    {
      "epoch": 0.8280141843971631,
      "grad_norm": 0.35060107707977295,
      "learning_rate": 0.00017737030411449015,
      "loss": 0.1444,
      "step": 467
    },
    {
      "epoch": 0.8297872340425532,
      "grad_norm": 0.31308284401893616,
      "learning_rate": 0.0001771019677996422,
      "loss": 0.0722,
      "step": 468
    },
    {
      "epoch": 0.8315602836879432,
      "grad_norm": 0.4981883466243744,
      "learning_rate": 0.00017683363148479427,
      "loss": 0.1764,
      "step": 469
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.5609257817268372,
      "learning_rate": 0.00017656529516994632,
      "loss": 0.175,
      "step": 470
    },
    {
      "epoch": 0.8351063829787234,
      "grad_norm": 0.5611094832420349,
      "learning_rate": 0.00017629695885509837,
      "loss": 0.242,
      "step": 471
    },
    {
      "epoch": 0.8368794326241135,
      "grad_norm": 0.7660516500473022,
      "learning_rate": 0.00017602862254025042,
      "loss": 0.2488,
      "step": 472
    },
    {
      "epoch": 0.8386524822695035,
      "grad_norm": 0.3760780096054077,
      "learning_rate": 0.0001757602862254025,
      "loss": 0.1253,
      "step": 473
    },
    {
      "epoch": 0.8404255319148937,
      "grad_norm": 0.5254774689674377,
      "learning_rate": 0.00017549194991055454,
      "loss": 0.2273,
      "step": 474
    },
    {
      "epoch": 0.8421985815602837,
      "grad_norm": 0.34555211663246155,
      "learning_rate": 0.0001752236135957066,
      "loss": 0.1294,
      "step": 475
    },
    {
      "epoch": 0.8439716312056738,
      "grad_norm": 0.4075317978858948,
      "learning_rate": 0.00017495527728085867,
      "loss": 0.1258,
      "step": 476
    },
    {
      "epoch": 0.8457446808510638,
      "grad_norm": 0.4469994604587555,
      "learning_rate": 0.00017468694096601072,
      "loss": 0.155,
      "step": 477
    },
    {
      "epoch": 0.8475177304964538,
      "grad_norm": 0.3752269744873047,
      "learning_rate": 0.0001744186046511628,
      "loss": 0.1403,
      "step": 478
    },
    {
      "epoch": 0.849290780141844,
      "grad_norm": 0.3446385860443115,
      "learning_rate": 0.00017415026833631482,
      "loss": 0.0925,
      "step": 479
    },
    {
      "epoch": 0.851063829787234,
      "grad_norm": 0.5445675253868103,
      "learning_rate": 0.0001738819320214669,
      "loss": 0.1791,
      "step": 480
    },
    {
      "epoch": 0.8528368794326241,
      "grad_norm": 0.4845312535762787,
      "learning_rate": 0.00017361359570661894,
      "loss": 0.236,
      "step": 481
    },
    {
      "epoch": 0.8546099290780141,
      "grad_norm": 0.21976758539676666,
      "learning_rate": 0.00017334525939177102,
      "loss": 0.0785,
      "step": 482
    },
    {
      "epoch": 0.8563829787234043,
      "grad_norm": 0.3595951497554779,
      "learning_rate": 0.00017307692307692304,
      "loss": 0.1297,
      "step": 483
    },
    {
      "epoch": 0.8581560283687943,
      "grad_norm": 0.34340062737464905,
      "learning_rate": 0.00017280858676207511,
      "loss": 0.1437,
      "step": 484
    },
    {
      "epoch": 0.8599290780141844,
      "grad_norm": 0.40024513006210327,
      "learning_rate": 0.0001725402504472272,
      "loss": 0.111,
      "step": 485
    },
    {
      "epoch": 0.8617021276595744,
      "grad_norm": 0.4997851252555847,
      "learning_rate": 0.00017227191413237924,
      "loss": 0.1406,
      "step": 486
    },
    {
      "epoch": 0.8634751773049646,
      "grad_norm": 0.37670308351516724,
      "learning_rate": 0.0001720035778175313,
      "loss": 0.1653,
      "step": 487
    },
    {
      "epoch": 0.8652482269503546,
      "grad_norm": 0.4352554380893707,
      "learning_rate": 0.00017173524150268334,
      "loss": 0.1275,
      "step": 488
    },
    {
      "epoch": 0.8670212765957447,
      "grad_norm": 0.31643253564834595,
      "learning_rate": 0.00017146690518783541,
      "loss": 0.0982,
      "step": 489
    },
    {
      "epoch": 0.8687943262411347,
      "grad_norm": 0.3775823712348938,
      "learning_rate": 0.00017119856887298746,
      "loss": 0.1087,
      "step": 490
    },
    {
      "epoch": 0.8705673758865248,
      "grad_norm": 0.714429497718811,
      "learning_rate": 0.0001709302325581395,
      "loss": 0.1743,
      "step": 491
    },
    {
      "epoch": 0.8723404255319149,
      "grad_norm": 0.24479807913303375,
      "learning_rate": 0.00017066189624329156,
      "loss": 0.0616,
      "step": 492
    },
    {
      "epoch": 0.874113475177305,
      "grad_norm": 0.4242885708808899,
      "learning_rate": 0.00017039355992844364,
      "loss": 0.1266,
      "step": 493
    },
    {
      "epoch": 0.875886524822695,
      "grad_norm": 0.4247332513332367,
      "learning_rate": 0.00017012522361359571,
      "loss": 0.1287,
      "step": 494
    },
    {
      "epoch": 0.8776595744680851,
      "grad_norm": 0.49524983763694763,
      "learning_rate": 0.00016985688729874774,
      "loss": 0.1957,
      "step": 495
    },
    {
      "epoch": 0.8794326241134752,
      "grad_norm": 0.2352031022310257,
      "learning_rate": 0.0001695885509838998,
      "loss": 0.0585,
      "step": 496
    },
    {
      "epoch": 0.8812056737588653,
      "grad_norm": 0.23934312164783478,
      "learning_rate": 0.00016932021466905186,
      "loss": 0.0477,
      "step": 497
    },
    {
      "epoch": 0.8829787234042553,
      "grad_norm": 0.3493366539478302,
      "learning_rate": 0.00016905187835420394,
      "loss": 0.1074,
      "step": 498
    },
    {
      "epoch": 0.8847517730496454,
      "grad_norm": 0.28466084599494934,
      "learning_rate": 0.00016878354203935596,
      "loss": 0.0575,
      "step": 499
    },
    {
      "epoch": 0.8865248226950354,
      "grad_norm": 0.33539536595344543,
      "learning_rate": 0.00016851520572450803,
      "loss": 0.0637,
      "step": 500
    }
  ],
  "logging_steps": 1,
  "max_steps": 1128,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.208442192384e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}