{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.940092165898617,
  "eval_steps": 500,
  "global_step": 360,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.013824884792626729,
      "grad_norm": 31.00213623046875,
      "learning_rate": 5.0000000000000004e-08,
      "loss": 2.2089,
      "step": 1
    },
    {
      "epoch": 0.027649769585253458,
      "grad_norm": 30.27136993408203,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 2.1536,
      "step": 2
    },
    {
      "epoch": 0.041474654377880185,
      "grad_norm": 30.48703384399414,
      "learning_rate": 1.5000000000000002e-07,
      "loss": 2.1581,
      "step": 3
    },
    {
      "epoch": 0.055299539170506916,
      "grad_norm": 30.779329299926758,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 2.1741,
      "step": 4
    },
    {
      "epoch": 0.06912442396313365,
      "grad_norm": 31.22808837890625,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 2.1864,
      "step": 5
    },
    {
      "epoch": 0.08294930875576037,
      "grad_norm": 30.783327102661133,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 2.1993,
      "step": 6
    },
    {
      "epoch": 0.0967741935483871,
      "grad_norm": 30.57423210144043,
      "learning_rate": 3.5000000000000004e-07,
      "loss": 2.1506,
      "step": 7
    },
    {
      "epoch": 0.11059907834101383,
      "grad_norm": 30.952186584472656,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 2.1599,
      "step": 8
    },
    {
      "epoch": 0.12442396313364056,
      "grad_norm": 30.37245750427246,
      "learning_rate": 4.5000000000000003e-07,
      "loss": 2.1572,
      "step": 9
    },
    {
      "epoch": 0.1382488479262673,
      "grad_norm": 30.930192947387695,
      "learning_rate": 5.000000000000001e-07,
      "loss": 2.1447,
      "step": 10
    },
    {
      "epoch": 0.15207373271889402,
      "grad_norm": 29.735448837280273,
      "learning_rate": 5.5e-07,
      "loss": 2.0742,
      "step": 11
    },
    {
      "epoch": 0.16589861751152074,
      "grad_norm": 29.62826156616211,
      "learning_rate": 6.000000000000001e-07,
      "loss": 2.061,
      "step": 12
    },
    {
      "epoch": 0.17972350230414746,
      "grad_norm": 28.937463760375977,
      "learning_rate": 6.5e-07,
      "loss": 1.9974,
      "step": 13
    },
    {
      "epoch": 0.1935483870967742,
      "grad_norm": 29.24833106994629,
      "learning_rate": 7.000000000000001e-07,
      "loss": 1.9833,
      "step": 14
    },
    {
      "epoch": 0.2073732718894009,
      "grad_norm": 28.122018814086914,
      "learning_rate": 7.5e-07,
      "loss": 1.8934,
      "step": 15
    },
    {
      "epoch": 0.22119815668202766,
      "grad_norm": 28.059659957885742,
      "learning_rate": 8.000000000000001e-07,
      "loss": 1.875,
      "step": 16
    },
    {
      "epoch": 0.2350230414746544,
      "grad_norm": 27.361961364746094,
      "learning_rate": 8.500000000000001e-07,
      "loss": 1.8009,
      "step": 17
    },
    {
      "epoch": 0.2488479262672811,
      "grad_norm": 26.721765518188477,
      "learning_rate": 9.000000000000001e-07,
      "loss": 1.7116,
      "step": 18
    },
    {
      "epoch": 0.2626728110599078,
      "grad_norm": 25.37330436706543,
      "learning_rate": 9.500000000000001e-07,
      "loss": 1.5608,
      "step": 19
    },
    {
      "epoch": 0.2764976958525346,
      "grad_norm": 25.81206703186035,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 1.5043,
      "step": 20
    },
    {
      "epoch": 0.2903225806451613,
      "grad_norm": 25.539344787597656,
      "learning_rate": 1.0500000000000001e-06,
      "loss": 1.3673,
      "step": 21
    },
    {
      "epoch": 0.30414746543778803,
      "grad_norm": 25.097164154052734,
      "learning_rate": 1.1e-06,
      "loss": 1.2029,
      "step": 22
    },
    {
      "epoch": 0.31797235023041476,
      "grad_norm": 24.619497299194336,
      "learning_rate": 1.1500000000000002e-06,
      "loss": 1.0458,
      "step": 23
    },
    {
      "epoch": 0.3317972350230415,
      "grad_norm": 23.820302963256836,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 0.8723,
      "step": 24
    },
    {
      "epoch": 0.3456221198156682,
      "grad_norm": 23.12735939025879,
      "learning_rate": 1.25e-06,
      "loss": 0.7183,
      "step": 25
    },
    {
      "epoch": 0.35944700460829493,
      "grad_norm": 20.127134323120117,
      "learning_rate": 1.3e-06,
      "loss": 0.5248,
      "step": 26
    },
    {
      "epoch": 0.37327188940092165,
      "grad_norm": 15.901495933532715,
      "learning_rate": 1.3500000000000002e-06,
      "loss": 0.3689,
      "step": 27
    },
    {
      "epoch": 0.3870967741935484,
      "grad_norm": 11.053832054138184,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 0.2482,
      "step": 28
    },
    {
      "epoch": 0.4009216589861751,
      "grad_norm": 7.248495578765869,
      "learning_rate": 1.45e-06,
      "loss": 0.1847,
      "step": 29
    },
    {
      "epoch": 0.4147465437788018,
      "grad_norm": 5.378540515899658,
      "learning_rate": 1.5e-06,
      "loss": 0.1423,
      "step": 30
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 3.8371808528900146,
      "learning_rate": 1.5500000000000002e-06,
      "loss": 0.1152,
      "step": 31
    },
    {
      "epoch": 0.4423963133640553,
      "grad_norm": 2.2655274868011475,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.0845,
      "step": 32
    },
    {
      "epoch": 0.45622119815668205,
      "grad_norm": 1.5746861696243286,
      "learning_rate": 1.6500000000000003e-06,
      "loss": 0.0711,
      "step": 33
    },
    {
      "epoch": 0.4700460829493088,
      "grad_norm": 1.3510947227478027,
      "learning_rate": 1.7000000000000002e-06,
      "loss": 0.0734,
      "step": 34
    },
    {
      "epoch": 0.4838709677419355,
      "grad_norm": 0.9737389087677002,
      "learning_rate": 1.75e-06,
      "loss": 0.0651,
      "step": 35
    },
    {
      "epoch": 0.4976958525345622,
      "grad_norm": 0.9815284609794617,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 0.0593,
      "step": 36
    },
    {
      "epoch": 0.511520737327189,
      "grad_norm": 0.8567912578582764,
      "learning_rate": 1.85e-06,
      "loss": 0.0543,
      "step": 37
    },
    {
      "epoch": 0.5253456221198156,
      "grad_norm": 0.6773302555084229,
      "learning_rate": 1.9000000000000002e-06,
      "loss": 0.0622,
      "step": 38
    },
    {
      "epoch": 0.5391705069124424,
      "grad_norm": 0.49936285614967346,
      "learning_rate": 1.9500000000000004e-06,
      "loss": 0.0511,
      "step": 39
    },
    {
      "epoch": 0.5529953917050692,
      "grad_norm": 0.6253588795661926,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.0478,
      "step": 40
    },
    {
      "epoch": 0.5668202764976958,
      "grad_norm": 0.5103089809417725,
      "learning_rate": 2.05e-06,
      "loss": 0.0465,
      "step": 41
    },
    {
      "epoch": 0.5806451612903226,
      "grad_norm": 0.29294702410697937,
      "learning_rate": 2.1000000000000002e-06,
      "loss": 0.0456,
      "step": 42
    },
    {
      "epoch": 0.5944700460829493,
      "grad_norm": 0.4237954616546631,
      "learning_rate": 2.15e-06,
      "loss": 0.0501,
      "step": 43
    },
    {
      "epoch": 0.6082949308755761,
      "grad_norm": 0.42243412137031555,
      "learning_rate": 2.2e-06,
      "loss": 0.0388,
      "step": 44
    },
    {
      "epoch": 0.6221198156682027,
      "grad_norm": 0.37881818413734436,
      "learning_rate": 2.25e-06,
      "loss": 0.0415,
      "step": 45
    },
    {
      "epoch": 0.6359447004608295,
      "grad_norm": 0.4941152036190033,
      "learning_rate": 2.3000000000000004e-06,
      "loss": 0.045,
      "step": 46
    },
    {
      "epoch": 0.6497695852534562,
      "grad_norm": 0.3046450912952423,
      "learning_rate": 2.35e-06,
      "loss": 0.0386,
      "step": 47
    },
    {
      "epoch": 0.663594470046083,
      "grad_norm": 0.39361852407455444,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.0447,
      "step": 48
    },
    {
      "epoch": 0.6774193548387096,
      "grad_norm": 0.5190001130104065,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 0.0364,
      "step": 49
    },
    {
      "epoch": 0.6912442396313364,
      "grad_norm": 0.372072696685791,
      "learning_rate": 2.5e-06,
      "loss": 0.043,
      "step": 50
    },
    {
      "epoch": 0.7050691244239631,
      "grad_norm": 0.3756551146507263,
      "learning_rate": 2.55e-06,
      "loss": 0.0424,
      "step": 51
    },
    {
      "epoch": 0.7188940092165899,
      "grad_norm": 0.4593554437160492,
      "learning_rate": 2.6e-06,
      "loss": 0.0387,
      "step": 52
    },
    {
      "epoch": 0.7327188940092166,
      "grad_norm": 0.2931855618953705,
      "learning_rate": 2.6500000000000005e-06,
      "loss": 0.0396,
      "step": 53
    },
    {
      "epoch": 0.7465437788018433,
      "grad_norm": 0.38429534435272217,
      "learning_rate": 2.7000000000000004e-06,
      "loss": 0.0373,
      "step": 54
    },
    {
      "epoch": 0.7603686635944701,
      "grad_norm": 0.3506857752799988,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 0.04,
      "step": 55
    },
    {
      "epoch": 0.7741935483870968,
      "grad_norm": 0.29847028851509094,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.0369,
      "step": 56
    },
    {
      "epoch": 0.7880184331797235,
      "grad_norm": 0.3653375506401062,
      "learning_rate": 2.85e-06,
      "loss": 0.0396,
      "step": 57
    },
    {
      "epoch": 0.8018433179723502,
      "grad_norm": 0.3163083791732788,
      "learning_rate": 2.9e-06,
      "loss": 0.0337,
      "step": 58
    },
    {
      "epoch": 0.815668202764977,
      "grad_norm": 0.3734363615512848,
      "learning_rate": 2.95e-06,
      "loss": 0.0327,
      "step": 59
    },
    {
      "epoch": 0.8294930875576036,
      "grad_norm": 0.29547712206840515,
      "learning_rate": 3e-06,
      "loss": 0.0365,
      "step": 60
    },
    {
      "epoch": 0.8433179723502304,
      "grad_norm": 0.4041007161140442,
      "learning_rate": 3.05e-06,
      "loss": 0.038,
      "step": 61
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.3602149784564972,
      "learning_rate": 3.1000000000000004e-06,
      "loss": 0.033,
      "step": 62
    },
    {
      "epoch": 0.8709677419354839,
      "grad_norm": 0.2948857545852661,
      "learning_rate": 3.1500000000000003e-06,
      "loss": 0.0386,
      "step": 63
    },
    {
      "epoch": 0.8847926267281107,
      "grad_norm": 0.39098358154296875,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.0323,
      "step": 64
    },
    {
      "epoch": 0.8986175115207373,
      "grad_norm": 0.3692062795162201,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 0.0309,
      "step": 65
    },
    {
      "epoch": 0.9124423963133641,
      "grad_norm": 0.3967229425907135,
      "learning_rate": 3.3000000000000006e-06,
      "loss": 0.0346,
      "step": 66
    },
    {
      "epoch": 0.9262672811059908,
      "grad_norm": 0.47776708006858826,
      "learning_rate": 3.3500000000000005e-06,
      "loss": 0.0355,
      "step": 67
    },
    {
      "epoch": 0.9400921658986175,
      "grad_norm": 0.21545131504535675,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 0.0294,
      "step": 68
    },
    {
      "epoch": 0.9539170506912442,
      "grad_norm": 0.23738539218902588,
      "learning_rate": 3.45e-06,
      "loss": 0.0308,
      "step": 69
    },
    {
      "epoch": 0.967741935483871,
      "grad_norm": 0.29174014925956726,
      "learning_rate": 3.5e-06,
      "loss": 0.0312,
      "step": 70
    },
    {
      "epoch": 0.9815668202764977,
      "grad_norm": 0.38475602865219116,
      "learning_rate": 3.5500000000000003e-06,
      "loss": 0.0324,
      "step": 71
    },
    {
      "epoch": 0.9953917050691244,
      "grad_norm": 0.4077378809452057,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 0.0297,
      "step": 72
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.4077378809452057,
      "learning_rate": 3.65e-06,
      "loss": 0.031,
      "step": 73
    },
    {
      "epoch": 1.0138248847926268,
      "grad_norm": 0.46581539511680603,
      "learning_rate": 3.7e-06,
      "loss": 0.0313,
      "step": 74
    },
    {
      "epoch": 1.0276497695852536,
      "grad_norm": 0.24417200684547424,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.027,
      "step": 75
    },
    {
      "epoch": 1.0414746543778801,
      "grad_norm": 0.20425117015838623,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 0.0307,
      "step": 76
    },
    {
      "epoch": 1.055299539170507,
      "grad_norm": 0.3578161597251892,
      "learning_rate": 3.85e-06,
      "loss": 0.0312,
      "step": 77
    },
    {
      "epoch": 1.0691244239631337,
      "grad_norm": 0.39486679434776306,
      "learning_rate": 3.900000000000001e-06,
      "loss": 0.0294,
      "step": 78
    },
    {
      "epoch": 1.0829493087557605,
      "grad_norm": 0.3932795226573944,
      "learning_rate": 3.95e-06,
      "loss": 0.0307,
      "step": 79
    },
    {
      "epoch": 1.096774193548387,
      "grad_norm": 0.2946235239505768,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.0257,
      "step": 80
    },
    {
      "epoch": 1.1105990783410138,
      "grad_norm": 0.3318672776222229,
      "learning_rate": 4.05e-06,
      "loss": 0.0296,
      "step": 81
    },
    {
      "epoch": 1.1244239631336406,
      "grad_norm": 0.23701588809490204,
      "learning_rate": 4.1e-06,
      "loss": 0.0298,
      "step": 82
    },
    {
      "epoch": 1.1382488479262673,
      "grad_norm": 0.2415941059589386,
      "learning_rate": 4.15e-06,
      "loss": 0.0256,
      "step": 83
    },
    {
      "epoch": 1.1520737327188941,
      "grad_norm": 0.24098087847232819,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.0263,
      "step": 84
    },
    {
      "epoch": 1.1658986175115207,
      "grad_norm": 0.3530103862285614,
      "learning_rate": 4.25e-06,
      "loss": 0.0308,
      "step": 85
    },
    {
      "epoch": 1.1797235023041475,
      "grad_norm": 0.2382838875055313,
      "learning_rate": 4.3e-06,
      "loss": 0.0254,
      "step": 86
    },
    {
      "epoch": 1.1935483870967742,
      "grad_norm": 0.2670588791370392,
      "learning_rate": 4.350000000000001e-06,
      "loss": 0.0255,
      "step": 87
    },
    {
      "epoch": 1.2073732718894008,
      "grad_norm": 0.30723804235458374,
      "learning_rate": 4.4e-06,
      "loss": 0.0263,
      "step": 88
    },
    {
      "epoch": 1.2211981566820276,
      "grad_norm": 0.505890965461731,
      "learning_rate": 4.450000000000001e-06,
      "loss": 0.0265,
      "step": 89
    },
    {
      "epoch": 1.2350230414746544,
      "grad_norm": 0.24307991564273834,
      "learning_rate": 4.5e-06,
      "loss": 0.0227,
      "step": 90
    },
    {
      "epoch": 1.2488479262672811,
      "grad_norm": 0.2198561429977417,
      "learning_rate": 4.5500000000000005e-06,
      "loss": 0.0261,
      "step": 91
    },
    {
      "epoch": 1.262672811059908,
      "grad_norm": 0.2435183823108673,
      "learning_rate": 4.600000000000001e-06,
      "loss": 0.0225,
      "step": 92
    },
    {
      "epoch": 1.2764976958525347,
      "grad_norm": 0.18837811052799225,
      "learning_rate": 4.65e-06,
      "loss": 0.0218,
      "step": 93
    },
    {
      "epoch": 1.2903225806451613,
      "grad_norm": 0.3818771541118622,
      "learning_rate": 4.7e-06,
      "loss": 0.0223,
      "step": 94
    },
    {
      "epoch": 1.304147465437788,
      "grad_norm": 0.2358720600605011,
      "learning_rate": 4.75e-06,
      "loss": 0.0204,
      "step": 95
    },
    {
      "epoch": 1.3179723502304148,
      "grad_norm": 0.25374144315719604,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.022,
      "step": 96
    },
    {
      "epoch": 1.3317972350230414,
      "grad_norm": 0.36181601881980896,
      "learning_rate": 4.85e-06,
      "loss": 0.0244,
      "step": 97
    },
    {
      "epoch": 1.3456221198156681,
      "grad_norm": 0.3156590759754181,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.0233,
      "step": 98
    },
    {
      "epoch": 1.359447004608295,
      "grad_norm": 0.21958638727664948,
      "learning_rate": 4.95e-06,
      "loss": 0.0218,
      "step": 99
    },
    {
      "epoch": 1.3732718894009217,
      "grad_norm": 0.34455621242523193,
      "learning_rate": 5e-06,
      "loss": 0.0267,
      "step": 100
    },
    {
      "epoch": 1.3870967741935485,
      "grad_norm": 0.283086359500885,
      "learning_rate": 4.999888074163108e-06,
      "loss": 0.0238,
      "step": 101
    },
    {
      "epoch": 1.400921658986175,
      "grad_norm": 0.28856486082077026,
      "learning_rate": 4.999552306674345e-06,
      "loss": 0.0186,
      "step": 102
    },
    {
      "epoch": 1.4147465437788018,
      "grad_norm": 0.26721692085266113,
      "learning_rate": 4.998992727598557e-06,
      "loss": 0.0193,
      "step": 103
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.3459971249103546,
      "learning_rate": 4.998209387040829e-06,
      "loss": 0.0218,
      "step": 104
    },
    {
      "epoch": 1.4423963133640554,
      "grad_norm": 0.25979122519493103,
      "learning_rate": 4.9972023551419995e-06,
      "loss": 0.0216,
      "step": 105
    },
    {
      "epoch": 1.456221198156682,
      "grad_norm": 0.19960424304008484,
      "learning_rate": 4.995971722072379e-06,
      "loss": 0.0176,
      "step": 106
    },
    {
      "epoch": 1.4700460829493087,
      "grad_norm": 0.2529441714286804,
      "learning_rate": 4.9945175980236745e-06,
      "loss": 0.0181,
      "step": 107
    },
    {
      "epoch": 1.4838709677419355,
      "grad_norm": 0.2690267264842987,
      "learning_rate": 4.992840113199131e-06,
      "loss": 0.0196,
      "step": 108
    },
    {
      "epoch": 1.4976958525345623,
      "grad_norm": 0.3516470789909363,
      "learning_rate": 4.990939417801859e-06,
      "loss": 0.0182,
      "step": 109
    },
    {
      "epoch": 1.511520737327189,
      "grad_norm": 0.30167508125305176,
      "learning_rate": 4.988815682021398e-06,
      "loss": 0.0205,
      "step": 110
    },
    {
      "epoch": 1.5253456221198156,
      "grad_norm": 0.3920849859714508,
      "learning_rate": 4.986469096018472e-06,
      "loss": 0.0177,
      "step": 111
    },
    {
      "epoch": 1.5391705069124424,
      "grad_norm": 0.3274078369140625,
      "learning_rate": 4.983899869907963e-06,
      "loss": 0.0185,
      "step": 112
    },
    {
      "epoch": 1.5529953917050692,
      "grad_norm": 0.2237282395362854,
      "learning_rate": 4.981108233740096e-06,
      "loss": 0.016,
      "step": 113
    },
    {
      "epoch": 1.5668202764976957,
      "grad_norm": 0.23966379463672638,
      "learning_rate": 4.978094437479843e-06,
      "loss": 0.0183,
      "step": 114
    },
    {
      "epoch": 1.5806451612903225,
      "grad_norm": 0.4027673602104187,
      "learning_rate": 4.97485875098454e-06,
      "loss": 0.0171,
      "step": 115
    },
    {
      "epoch": 1.5944700460829493,
      "grad_norm": 0.24082835018634796,
      "learning_rate": 4.971401463979722e-06,
      "loss": 0.016,
      "step": 116
    },
    {
      "epoch": 1.608294930875576,
      "grad_norm": 0.19387558102607727,
      "learning_rate": 4.967722886033181e-06,
      "loss": 0.0165,
      "step": 117
    },
    {
      "epoch": 1.6221198156682028,
      "grad_norm": 0.33696162700653076,
      "learning_rate": 4.963823346527249e-06,
      "loss": 0.0154,
      "step": 118
    },
    {
      "epoch": 1.6359447004608296,
      "grad_norm": 0.30290740728378296,
      "learning_rate": 4.959703194629304e-06,
      "loss": 0.0175,
      "step": 119
    },
    {
      "epoch": 1.6497695852534562,
      "grad_norm": 0.3781787157058716,
      "learning_rate": 4.955362799260507e-06,
      "loss": 0.0145,
      "step": 120
    },
    {
      "epoch": 1.663594470046083,
      "grad_norm": 0.39995357394218445,
      "learning_rate": 4.950802549062764e-06,
      "loss": 0.015,
      "step": 121
    },
    {
      "epoch": 1.6774193548387095,
      "grad_norm": 0.19926570355892181,
      "learning_rate": 4.946022852363932e-06,
      "loss": 0.0135,
      "step": 122
    },
    {
      "epoch": 1.6912442396313363,
      "grad_norm": 0.22450515627861023,
      "learning_rate": 4.9410241371412525e-06,
      "loss": 0.0135,
      "step": 123
    },
    {
      "epoch": 1.705069124423963,
      "grad_norm": 0.3588384985923767,
      "learning_rate": 4.935806850983034e-06,
      "loss": 0.0125,
      "step": 124
    },
    {
      "epoch": 1.7188940092165899,
      "grad_norm": 0.28571122884750366,
      "learning_rate": 4.9303714610485705e-06,
      "loss": 0.0166,
      "step": 125
    },
    {
      "epoch": 1.7327188940092166,
      "grad_norm": 0.3496967852115631,
      "learning_rate": 4.924718454026318e-06,
      "loss": 0.0139,
      "step": 126
    },
    {
      "epoch": 1.7465437788018434,
      "grad_norm": 0.3279854357242584,
      "learning_rate": 4.918848336090309e-06,
      "loss": 0.0133,
      "step": 127
    },
    {
      "epoch": 1.7603686635944702,
      "grad_norm": 0.19201801717281342,
      "learning_rate": 4.912761632854834e-06,
      "loss": 0.0151,
      "step": 128
    },
    {
      "epoch": 1.7741935483870968,
      "grad_norm": 0.27701929211616516,
      "learning_rate": 4.906458889327375e-06,
      "loss": 0.0148,
      "step": 129
    },
    {
      "epoch": 1.7880184331797235,
      "grad_norm": 0.2757968008518219,
      "learning_rate": 4.899940669859807e-06,
      "loss": 0.0118,
      "step": 130
    },
    {
      "epoch": 1.80184331797235,
      "grad_norm": 0.18373191356658936,
      "learning_rate": 4.893207558097867e-06,
      "loss": 0.0149,
      "step": 131
    },
    {
      "epoch": 1.8156682027649769,
      "grad_norm": 0.2116280496120453,
      "learning_rate": 4.8862601569288885e-06,
      "loss": 0.0129,
      "step": 132
    },
    {
      "epoch": 1.8294930875576036,
      "grad_norm": 0.30384117364883423,
      "learning_rate": 4.879099088427824e-06,
      "loss": 0.0136,
      "step": 133
    },
    {
      "epoch": 1.8433179723502304,
      "grad_norm": 0.3766787052154541,
      "learning_rate": 4.871724993801541e-06,
      "loss": 0.0123,
      "step": 134
    },
    {
      "epoch": 1.8571428571428572,
      "grad_norm": 0.3401263356208801,
      "learning_rate": 4.864138533331411e-06,
      "loss": 0.0122,
      "step": 135
    },
    {
      "epoch": 1.870967741935484,
      "grad_norm": 0.24321958422660828,
      "learning_rate": 4.8563403863141825e-06,
      "loss": 0.0123,
      "step": 136
    },
    {
      "epoch": 1.8847926267281108,
      "grad_norm": 0.16918110847473145,
      "learning_rate": 4.84833125100116e-06,
      "loss": 0.0104,
      "step": 137
    },
    {
      "epoch": 1.8986175115207373,
      "grad_norm": 0.23489230871200562,
      "learning_rate": 4.840111844535682e-06,
      "loss": 0.0122,
      "step": 138
    },
    {
      "epoch": 1.912442396313364,
      "grad_norm": 0.32796236872673035,
      "learning_rate": 4.8316829028889076e-06,
      "loss": 0.0109,
      "step": 139
    },
    {
      "epoch": 1.9262672811059907,
      "grad_norm": 0.24210475385189056,
      "learning_rate": 4.823045180793914e-06,
      "loss": 0.0118,
      "step": 140
    },
    {
      "epoch": 1.9400921658986174,
      "grad_norm": 0.3450548052787781,
      "learning_rate": 4.8141994516781196e-06,
      "loss": 0.0115,
      "step": 141
    },
    {
      "epoch": 1.9539170506912442,
      "grad_norm": 0.23163923621177673,
      "learning_rate": 4.805146507594034e-06,
      "loss": 0.0122,
      "step": 142
    },
    {
      "epoch": 1.967741935483871,
      "grad_norm": 0.8197745084762573,
      "learning_rate": 4.7958871591483305e-06,
      "loss": 0.0101,
      "step": 143
    },
    {
      "epoch": 1.9815668202764978,
      "grad_norm": 0.2917576730251312,
      "learning_rate": 4.786422235429269e-06,
      "loss": 0.0078,
      "step": 144
    },
    {
      "epoch": 1.9953917050691246,
      "grad_norm": 0.24417108297348022,
      "learning_rate": 4.776752583932455e-06,
      "loss": 0.0119,
      "step": 145
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.24417108297348022,
      "learning_rate": 4.766879070484957e-06,
      "loss": 0.0089,
      "step": 146
    },
    {
      "epoch": 2.013824884792627,
      "grad_norm": 0.4215025305747986,
      "learning_rate": 4.756802579167781e-06,
      "loss": 0.007,
      "step": 147
    },
    {
      "epoch": 2.0276497695852536,
      "grad_norm": 0.2002098709344864,
      "learning_rate": 4.746524012236706e-06,
      "loss": 0.0078,
      "step": 148
    },
    {
      "epoch": 2.0414746543778803,
      "grad_norm": 0.16432569921016693,
      "learning_rate": 4.736044290041496e-06,
      "loss": 0.0074,
      "step": 149
    },
    {
      "epoch": 2.055299539170507,
      "grad_norm": 0.2516174018383026,
      "learning_rate": 4.725364350943492e-06,
      "loss": 0.0067,
      "step": 150
    },
    {
      "epoch": 2.0691244239631335,
      "grad_norm": 0.24242427945137024,
      "learning_rate": 4.714485151231593e-06,
      "loss": 0.0083,
      "step": 151
    },
    {
      "epoch": 2.0829493087557602,
      "grad_norm": 0.22929197549819946,
      "learning_rate": 4.703407665036622e-06,
      "loss": 0.0061,
      "step": 152
    },
    {
      "epoch": 2.096774193548387,
      "grad_norm": 0.2929408550262451,
      "learning_rate": 4.692132884244113e-06,
      "loss": 0.0064,
      "step": 153
    },
    {
      "epoch": 2.110599078341014,
      "grad_norm": 0.22497303783893585,
      "learning_rate": 4.680661818405485e-06,
      "loss": 0.0061,
      "step": 154
    },
    {
      "epoch": 2.1244239631336406,
      "grad_norm": 0.13698536157608032,
      "learning_rate": 4.668995494647653e-06,
      "loss": 0.0059,
      "step": 155
    },
    {
      "epoch": 2.1382488479262673,
      "grad_norm": 0.32037150859832764,
      "learning_rate": 4.657134957581057e-06,
      "loss": 0.0067,
      "step": 156
    },
    {
      "epoch": 2.152073732718894,
      "grad_norm": 0.19389067590236664,
      "learning_rate": 4.645081269206128e-06,
      "loss": 0.0062,
      "step": 157
    },
    {
      "epoch": 2.165898617511521,
      "grad_norm": 0.2791127562522888,
      "learning_rate": 4.632835508818192e-06,
      "loss": 0.0058,
      "step": 158
    },
    {
      "epoch": 2.1797235023041477,
      "grad_norm": 0.2178739458322525,
      "learning_rate": 4.620398772910833e-06,
      "loss": 0.0056,
      "step": 159
    },
    {
      "epoch": 2.193548387096774,
      "grad_norm": 0.29685622453689575,
      "learning_rate": 4.607772175077712e-06,
      "loss": 0.0055,
      "step": 160
    },
    {
      "epoch": 2.207373271889401,
      "grad_norm": 0.6792906522750854,
      "learning_rate": 4.59495684591285e-06,
      "loss": 0.0057,
      "step": 161
    },
    {
      "epoch": 2.2211981566820276,
      "grad_norm": 0.17910148203372955,
      "learning_rate": 4.581953932909403e-06,
      "loss": 0.0046,
      "step": 162
    },
    {
      "epoch": 2.2350230414746544,
      "grad_norm": 0.12593543529510498,
      "learning_rate": 4.5687646003569055e-06,
      "loss": 0.0046,
      "step": 163
    },
    {
      "epoch": 2.248847926267281,
      "grad_norm": 0.15383680164813995,
      "learning_rate": 4.555390029237026e-06,
      "loss": 0.0059,
      "step": 164
    },
    {
      "epoch": 2.262672811059908,
      "grad_norm": 0.2324540764093399,
      "learning_rate": 4.541831417117815e-06,
      "loss": 0.0067,
      "step": 165
    },
    {
      "epoch": 2.2764976958525347,
      "grad_norm": 0.21278905868530273,
      "learning_rate": 4.528089978046481e-06,
      "loss": 0.0054,
      "step": 166
    },
    {
      "epoch": 2.2903225806451615,
      "grad_norm": 0.2499057948589325,
      "learning_rate": 4.514166942440679e-06,
      "loss": 0.003,
      "step": 167
    },
    {
      "epoch": 2.3041474654377883,
      "grad_norm": 0.1734611839056015,
      "learning_rate": 4.5000635569783365e-06,
      "loss": 0.0043,
      "step": 168
    },
    {
      "epoch": 2.3179723502304146,
      "grad_norm": 0.17815802991390228,
      "learning_rate": 4.4857810844860325e-06,
      "loss": 0.0048,
      "step": 169
    },
    {
      "epoch": 2.3317972350230414,
      "grad_norm": 0.22731409966945648,
      "learning_rate": 4.471320803825915e-06,
      "loss": 0.0034,
      "step": 170
    },
    {
      "epoch": 2.345622119815668,
      "grad_norm": 0.23811140656471252,
      "learning_rate": 4.4566840097811956e-06,
      "loss": 0.0029,
      "step": 171
    },
    {
      "epoch": 2.359447004608295,
      "grad_norm": 0.17744024097919464,
      "learning_rate": 4.4418720129402145e-06,
      "loss": 0.0029,
      "step": 172
    },
    {
      "epoch": 2.3732718894009217,
      "grad_norm": 0.24912229180335999,
      "learning_rate": 4.426886139579083e-06,
      "loss": 0.0049,
      "step": 173
    },
    {
      "epoch": 2.3870967741935485,
      "grad_norm": 0.17039696872234344,
      "learning_rate": 4.411727731542937e-06,
      "loss": 0.003,
      "step": 174
    },
    {
      "epoch": 2.4009216589861753,
      "grad_norm": 0.3089725375175476,
      "learning_rate": 4.39639814612578e-06,
      "loss": 0.0034,
      "step": 175
    },
    {
      "epoch": 2.4147465437788016,
      "grad_norm": 0.22647598385810852,
      "learning_rate": 4.3808987559489536e-06,
      "loss": 0.0058,
      "step": 176
    },
    {
      "epoch": 2.4285714285714284,
      "grad_norm": 0.19015835225582123,
      "learning_rate": 4.365230948838232e-06,
      "loss": 0.004,
      "step": 177
    },
    {
      "epoch": 2.442396313364055,
      "grad_norm": 0.1825973391532898,
      "learning_rate": 4.349396127699552e-06,
      "loss": 0.0032,
      "step": 178
    },
    {
      "epoch": 2.456221198156682,
      "grad_norm": 0.15705449879169464,
      "learning_rate": 4.3333957103934025e-06,
      "loss": 0.0035,
      "step": 179
    },
    {
      "epoch": 2.4700460829493087,
      "grad_norm": 0.19110225141048431,
      "learning_rate": 4.317231129607859e-06,
      "loss": 0.0019,
      "step": 180
    },
    {
      "epoch": 2.4838709677419355,
      "grad_norm": 0.1481270045042038,
      "learning_rate": 4.30090383273031e-06,
      "loss": 0.0035,
      "step": 181
    },
    {
      "epoch": 2.4976958525345623,
      "grad_norm": 0.19533571600914001,
      "learning_rate": 4.2844152817178476e-06,
      "loss": 0.0023,
      "step": 182
    },
    {
      "epoch": 2.511520737327189,
      "grad_norm": 0.1991293579339981,
      "learning_rate": 4.267766952966369e-06,
      "loss": 0.0025,
      "step": 183
    },
    {
      "epoch": 2.525345622119816,
      "grad_norm": 0.22637878358364105,
      "learning_rate": 4.2509603371783776e-06,
      "loss": 0.0029,
      "step": 184
    },
    {
      "epoch": 2.539170506912442,
      "grad_norm": 0.21984712779521942,
      "learning_rate": 4.233996939229502e-06,
      "loss": 0.0035,
      "step": 185
    },
    {
      "epoch": 2.5529953917050694,
      "grad_norm": 0.25706061720848083,
      "learning_rate": 4.216878278033753e-06,
      "loss": 0.0033,
      "step": 186
    },
    {
      "epoch": 2.5668202764976957,
      "grad_norm": 0.224118173122406,
      "learning_rate": 4.199605886407515e-06,
      "loss": 0.0017,
      "step": 187
    },
    {
      "epoch": 2.5806451612903225,
      "grad_norm": 0.0781751424074173,
      "learning_rate": 4.1821813109322975e-06,
      "loss": 0.002,
      "step": 188
    },
    {
      "epoch": 2.5944700460829493,
      "grad_norm": 0.2209765911102295,
      "learning_rate": 4.164606111816256e-06,
      "loss": 0.0018,
      "step": 189
    },
    {
      "epoch": 2.608294930875576,
      "grad_norm": 0.12815824151039124,
      "learning_rate": 4.146881862754485e-06,
      "loss": 0.003,
      "step": 190
    },
    {
      "epoch": 2.622119815668203,
      "grad_norm": 0.3006991147994995,
      "learning_rate": 4.129010150788112e-06,
      "loss": 0.0022,
      "step": 191
    },
    {
      "epoch": 2.6359447004608296,
      "grad_norm": 0.19085584580898285,
      "learning_rate": 4.110992576162193e-06,
      "loss": 0.0026,
      "step": 192
    },
    {
      "epoch": 2.6497695852534564,
      "grad_norm": 0.13027659058570862,
      "learning_rate": 4.092830752182423e-06,
      "loss": 0.0015,
      "step": 193
    },
    {
      "epoch": 2.6635944700460827,
      "grad_norm": 0.16998590528964996,
      "learning_rate": 4.074526305070679e-06,
      "loss": 0.0018,
      "step": 194
    },
    {
      "epoch": 2.6774193548387095,
      "grad_norm": 0.1743537187576294,
      "learning_rate": 4.056080873819412e-06,
      "loss": 0.0022,
      "step": 195
    },
    {
      "epoch": 2.6912442396313363,
      "grad_norm": 0.3566405177116394,
      "learning_rate": 4.037496110044885e-06,
      "loss": 0.0018,
      "step": 196
    },
    {
      "epoch": 2.705069124423963,
      "grad_norm": 0.274739146232605,
      "learning_rate": 4.018773677839289e-06,
      "loss": 0.0012,
      "step": 197
    },
    {
      "epoch": 2.71889400921659,
      "grad_norm": 0.12038746476173401,
      "learning_rate": 3.999915253621739e-06,
      "loss": 0.0013,
      "step": 198
    },
    {
      "epoch": 2.7327188940092166,
      "grad_norm": 0.12693172693252563,
      "learning_rate": 3.980922525988167e-06,
      "loss": 0.0017,
      "step": 199
    },
    {
      "epoch": 2.7465437788018434,
      "grad_norm": 0.11907753348350525,
      "learning_rate": 3.961797195560118e-06,
      "loss": 0.001,
      "step": 200
    },
    {
      "epoch": 2.76036866359447,
      "grad_norm": 0.1901165395975113,
      "learning_rate": 3.942540974832486e-06,
      "loss": 0.0028,
      "step": 201
    },
    {
      "epoch": 2.774193548387097,
      "grad_norm": 0.2039843052625656,
      "learning_rate": 3.9231555880201655e-06,
      "loss": 0.0011,
      "step": 202
    },
    {
      "epoch": 2.7880184331797233,
      "grad_norm": 0.16181506216526031,
      "learning_rate": 3.903642770903671e-06,
      "loss": 0.003,
      "step": 203
    },
    {
      "epoch": 2.80184331797235,
      "grad_norm": 0.13345211744308472,
      "learning_rate": 3.884004270673711e-06,
      "loss": 0.0023,
      "step": 204
    },
    {
      "epoch": 2.815668202764977,
      "grad_norm": 0.19453725218772888,
      "learning_rate": 3.864241845774746e-06,
      "loss": 0.001,
      "step": 205
    },
    {
      "epoch": 2.8294930875576036,
      "grad_norm": 0.18157535791397095,
      "learning_rate": 3.844357265747531e-06,
      "loss": 0.0029,
      "step": 206
    },
    {
      "epoch": 2.8433179723502304,
      "grad_norm": 0.17876467108726501,
      "learning_rate": 3.8243523110706736e-06,
      "loss": 0.0018,
      "step": 207
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.13000421226024628,
      "learning_rate": 3.8042287730012117e-06,
      "loss": 0.0011,
      "step": 208
    },
    {
      "epoch": 2.870967741935484,
      "grad_norm": 0.08808371424674988,
      "learning_rate": 3.7839884534142157e-06,
      "loss": 0.0018,
      "step": 209
    },
    {
      "epoch": 2.8847926267281108,
      "grad_norm": 0.32318148016929626,
      "learning_rate": 3.7636331646414524e-06,
      "loss": 0.0012,
      "step": 210
    },
    {
      "epoch": 2.8986175115207375,
      "grad_norm": 0.1259954422712326,
      "learning_rate": 3.7431647293091076e-06,
      "loss": 0.0012,
      "step": 211
    },
    {
      "epoch": 2.912442396313364,
      "grad_norm": 0.1344563215970993,
      "learning_rate": 3.7225849801745835e-06,
      "loss": 0.0006,
      "step": 212
    },
    {
      "epoch": 2.9262672811059907,
      "grad_norm": 0.09105626493692398,
      "learning_rate": 3.701895759962397e-06,
      "loss": 0.0007,
      "step": 213
    },
    {
      "epoch": 2.9400921658986174,
      "grad_norm": 0.11718853563070297,
      "learning_rate": 3.6810989211991777e-06,
      "loss": 0.0022,
      "step": 214
    },
    {
      "epoch": 2.953917050691244,
      "grad_norm": 0.10988112539052963,
      "learning_rate": 3.6601963260477923e-06,
      "loss": 0.0007,
      "step": 215
    },
    {
      "epoch": 2.967741935483871,
      "grad_norm": 0.12010538578033447,
      "learning_rate": 3.6391898461406045e-06,
      "loss": 0.0014,
      "step": 216
    },
    {
      "epoch": 2.9815668202764978,
      "grad_norm": 0.12934529781341553,
      "learning_rate": 3.6180813624118898e-06,
      "loss": 0.001,
      "step": 217
    },
    {
      "epoch": 2.9953917050691246,
      "grad_norm": 0.05664035677909851,
      "learning_rate": 3.5968727649294134e-06,
      "loss": 0.0002,
      "step": 218
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.07633747160434723,
      "learning_rate": 3.575565952725193e-06,
      "loss": 0.0002,
      "step": 219
    },
    {
      "epoch": 3.013824884792627,
      "grad_norm": 0.16964735090732574,
      "learning_rate": 3.55416283362546e-06,
      "loss": 0.0005,
      "step": 220
    },
    {
      "epoch": 3.0276497695852536,
      "grad_norm": 0.03826030716300011,
      "learning_rate": 3.5326653240798283e-06,
      "loss": 0.0003,
      "step": 221
    },
    {
      "epoch": 3.0414746543778803,
      "grad_norm": 0.05900357663631439,
      "learning_rate": 3.5110753489896924e-06,
      "loss": 0.0006,
      "step": 222
    },
    {
      "epoch": 3.055299539170507,
      "grad_norm": 0.06874338537454605,
      "learning_rate": 3.4893948415358803e-06,
      "loss": 0.0002,
      "step": 223
    },
    {
      "epoch": 3.0691244239631335,
      "grad_norm": 0.10445930808782578,
      "learning_rate": 3.4676257430055438e-06,
      "loss": 0.0011,
      "step": 224
    },
    {
      "epoch": 3.0829493087557602,
      "grad_norm": 0.03757224604487419,
      "learning_rate": 3.4457700026183378e-06,
      "loss": 0.0002,
      "step": 225
    },
    {
      "epoch": 3.096774193548387,
      "grad_norm": 0.2678232491016388,
      "learning_rate": 3.4238295773518924e-06,
      "loss": 0.0012,
      "step": 226
    },
    {
      "epoch": 3.110599078341014,
      "grad_norm": 0.11278262734413147,
      "learning_rate": 3.4018064317665745e-06,
      "loss": 0.0003,
      "step": 227
    },
    {
      "epoch": 3.1244239631336406,
      "grad_norm": 0.03823389112949371,
      "learning_rate": 3.3797025378295826e-06,
      "loss": 0.0002,
      "step": 228
    },
    {
      "epoch": 3.1382488479262673,
      "grad_norm": 0.015309945680201054,
      "learning_rate": 3.357519874738382e-06,
      "loss": 0.0,
      "step": 229
    },
    {
      "epoch": 3.152073732718894,
      "grad_norm": 0.12372211366891861,
      "learning_rate": 3.3352604287434752e-06,
      "loss": 0.0007,
      "step": 230
    },
    {
      "epoch": 3.165898617511521,
      "grad_norm": 0.062292926013469696,
      "learning_rate": 3.31292619297056e-06,
      "loss": 0.0003,
      "step": 231
    },
    {
      "epoch": 3.1797235023041477,
      "grad_norm": 0.02390543930232525,
      "learning_rate": 3.29051916724206e-06,
      "loss": 0.0001,
      "step": 232
    },
    {
      "epoch": 3.193548387096774,
      "grad_norm": 0.035650208592414856,
      "learning_rate": 3.2680413578980623e-06,
      "loss": 0.0001,
      "step": 233
    },
    {
      "epoch": 3.207373271889401,
      "grad_norm": 0.04304494708776474,
      "learning_rate": 3.245494777616664e-06,
      "loss": 0.0001,
      "step": 234
    },
    {
      "epoch": 3.2211981566820276,
      "grad_norm": 0.07038014382123947,
      "learning_rate": 3.2228814452337587e-06,
      "loss": 0.0005,
      "step": 235
    },
    {
      "epoch": 3.2350230414746544,
      "grad_norm": 0.18309231102466583,
      "learning_rate": 3.2002033855622683e-06,
      "loss": 0.0005,
      "step": 236
    },
    {
      "epoch": 3.248847926267281,
      "grad_norm": 0.04949740692973137,
      "learning_rate": 3.177462629210838e-06,
      "loss": 0.0002,
      "step": 237
    },
    {
      "epoch": 3.262672811059908,
      "grad_norm": 0.0319606214761734,
      "learning_rate": 3.154661212402017e-06,
      "loss": 0.0001,
      "step": 238
    },
    {
      "epoch": 3.2764976958525347,
      "grad_norm": 0.062357429414987564,
      "learning_rate": 3.131801176789934e-06,
      "loss": 0.0003,
      "step": 239
    },
    {
      "epoch": 3.2903225806451615,
      "grad_norm": 0.060603659600019455,
      "learning_rate": 3.1088845692774798e-06,
      "loss": 0.0003,
      "step": 240
    },
    {
      "epoch": 3.3041474654377883,
      "grad_norm": 0.12379086762666702,
      "learning_rate": 3.0859134418330373e-06,
      "loss": 0.0013,
      "step": 241
    },
    {
      "epoch": 3.3179723502304146,
      "grad_norm": 0.028559578582644463,
      "learning_rate": 3.0628898513067357e-06,
      "loss": 0.0002,
      "step": 242
    },
    {
      "epoch": 3.3317972350230414,
      "grad_norm": 0.04983198642730713,
      "learning_rate": 3.0398158592462847e-06,
      "loss": 0.0001,
      "step": 243
    },
    {
      "epoch": 3.345622119815668,
      "grad_norm": 0.07023701816797256,
      "learning_rate": 3.0166935317123824e-06,
      "loss": 0.0009,
      "step": 244
    },
    {
      "epoch": 3.359447004608295,
      "grad_norm": 0.046779777854681015,
      "learning_rate": 2.9935249390937184e-06,
      "loss": 0.0002,
      "step": 245
    },
    {
      "epoch": 3.3732718894009217,
      "grad_norm": 0.07187545299530029,
      "learning_rate": 2.970312155921585e-06,
      "loss": 0.0006,
      "step": 246
    },
    {
      "epoch": 3.3870967741935485,
      "grad_norm": 0.019256649538874626,
      "learning_rate": 2.9470572606841295e-06,
      "loss": 0.0001,
      "step": 247
    },
    {
      "epoch": 3.4009216589861753,
      "grad_norm": 0.0477205291390419,
      "learning_rate": 2.9237623356402423e-06,
      "loss": 0.0002,
      "step": 248
    },
    {
      "epoch": 3.4147465437788016,
      "grad_norm": 0.06807561218738556,
      "learning_rate": 2.900429466633107e-06,
      "loss": 0.0003,
      "step": 249
    },
    {
      "epoch": 3.4285714285714284,
      "grad_norm": 0.1516796499490738,
      "learning_rate": 2.8770607429034352e-06,
      "loss": 0.0006,
      "step": 250
    },
    {
      "epoch": 3.442396313364055,
      "grad_norm": 0.045213282108306885,
      "learning_rate": 2.8536582569023964e-06,
      "loss": 0.0001,
      "step": 251
    },
    {
      "epoch": 3.456221198156682,
      "grad_norm": 0.020110802724957466,
      "learning_rate": 2.8302241041042564e-06,
      "loss": 0.0001,
      "step": 252
    },
    {
      "epoch": 3.4700460829493087,
      "grad_norm": 0.03242102265357971,
      "learning_rate": 2.8067603828187446e-06,
      "loss": 0.0002,
      "step": 253
    },
    {
      "epoch": 3.4838709677419355,
      "grad_norm": 0.039639636874198914,
      "learning_rate": 2.7832691940031755e-06,
      "loss": 0.0001,
      "step": 254
    },
    {
      "epoch": 3.4976958525345623,
      "grad_norm": 0.06943561136722565,
      "learning_rate": 2.759752641074322e-06,
      "loss": 0.0003,
      "step": 255
    },
    {
      "epoch": 3.511520737327189,
      "grad_norm": 0.02593497559428215,
      "learning_rate": 2.7362128297200784e-06,
      "loss": 0.0001,
      "step": 256
    },
    {
      "epoch": 3.525345622119816,
      "grad_norm": 0.02811415307223797,
      "learning_rate": 2.712651867710914e-06,
      "loss": 0.0002,
      "step": 257
    },
    {
      "epoch": 3.539170506912442,
      "grad_norm": 0.07381757348775864,
      "learning_rate": 2.6890718647111424e-06,
      "loss": 0.0003,
      "step": 258
    },
    {
      "epoch": 3.5529953917050694,
      "grad_norm": 0.014391073025763035,
      "learning_rate": 2.665474932090017e-06,
      "loss": 0.0001,
      "step": 259
    },
    {
      "epoch": 3.5668202764976957,
      "grad_norm": 0.027200503274798393,
      "learning_rate": 2.6418631827326857e-06,
      "loss": 0.0001,
      "step": 260
    },
    {
      "epoch": 3.5806451612903225,
      "grad_norm": 0.02720312774181366,
      "learning_rate": 2.6182387308509927e-06,
      "loss": 0.0001,
      "step": 261
    },
    {
      "epoch": 3.5944700460829493,
      "grad_norm": 0.04352420195937157,
      "learning_rate": 2.5946036917941765e-06,
      "loss": 0.0004,
      "step": 262
    },
    {
      "epoch": 3.608294930875576,
      "grad_norm": 0.03459783270955086,
      "learning_rate": 2.570960181859458e-06,
      "loss": 0.0001,
      "step": 263
    },
    {
      "epoch": 3.622119815668203,
      "grad_norm": 0.03097033128142357,
      "learning_rate": 2.547310318102548e-06,
      "loss": 0.0001,
      "step": 264
    },
    {
      "epoch": 3.6359447004608296,
      "grad_norm": 0.0076720695942640305,
      "learning_rate": 2.5236562181480794e-06,
      "loss": 0.0,
      "step": 265
    },
    {
      "epoch": 3.6497695852534564,
      "grad_norm": 0.023994900286197662,
      "learning_rate": 2.5e-06,
      "loss": 0.0001,
      "step": 266
    },
    {
      "epoch": 3.6635944700460827,
      "grad_norm": 0.005682840943336487,
      "learning_rate": 2.4763437818519205e-06,
      "loss": 0.0,
      "step": 267
    },
    {
      "epoch": 3.6774193548387095,
      "grad_norm": 0.030443254858255386,
      "learning_rate": 2.4526896818974534e-06,
      "loss": 0.0002,
      "step": 268
    },
    {
      "epoch": 3.6912442396313363,
      "grad_norm": 0.008863283321261406,
      "learning_rate": 2.429039818140543e-06,
      "loss": 0.0,
      "step": 269
    },
    {
      "epoch": 3.705069124423963,
      "grad_norm": 0.009775679558515549,
      "learning_rate": 2.405396308205825e-06,
      "loss": 0.0001,
      "step": 270
    },
    {
      "epoch": 3.71889400921659,
      "grad_norm": 0.019227130338549614,
      "learning_rate": 2.381761269149009e-06,
      "loss": 0.0001,
      "step": 271
    },
    {
      "epoch": 3.7327188940092166,
      "grad_norm": 0.037880923599004745,
      "learning_rate": 2.358136817267315e-06,
      "loss": 0.0002,
      "step": 272
    },
    {
      "epoch": 3.7465437788018434,
      "grad_norm": 0.006014773156493902,
      "learning_rate": 2.334525067909983e-06,
      "loss": 0.0,
      "step": 273
    },
    {
      "epoch": 3.76036866359447,
      "grad_norm": 0.024770596995949745,
      "learning_rate": 2.3109281352888593e-06,
      "loss": 0.0001,
      "step": 274
    },
    {
      "epoch": 3.774193548387097,
      "grad_norm": 0.008392867632210255,
      "learning_rate": 2.2873481322890866e-06,
      "loss": 0.0,
      "step": 275
    },
    {
      "epoch": 3.7880184331797233,
      "grad_norm": 0.030915284529328346,
      "learning_rate": 2.263787170279922e-06,
      "loss": 0.0002,
      "step": 276
    },
    {
      "epoch": 3.80184331797235,
      "grad_norm": 0.04161324352025986,
      "learning_rate": 2.2402473589256793e-06,
      "loss": 0.0002,
      "step": 277
    },
    {
      "epoch": 3.815668202764977,
      "grad_norm": 0.04104781523346901,
      "learning_rate": 2.2167308059968258e-06,
      "loss": 0.0001,
      "step": 278
    },
    {
      "epoch": 3.8294930875576036,
      "grad_norm": 0.02981358952820301,
      "learning_rate": 2.193239617181256e-06,
      "loss": 0.0002,
      "step": 279
    },
    {
      "epoch": 3.8433179723502304,
      "grad_norm": 0.03616194799542427,
      "learning_rate": 2.169775895895745e-06,
      "loss": 0.0002,
      "step": 280
    },
    {
      "epoch": 3.857142857142857,
      "grad_norm": 0.003307241713628173,
      "learning_rate": 2.146341743097604e-06,
      "loss": 0.0,
      "step": 281
    },
    {
      "epoch": 3.870967741935484,
      "grad_norm": 0.023682212457060814,
      "learning_rate": 2.1229392570965656e-06,
      "loss": 0.0002,
      "step": 282
    },
    {
      "epoch": 3.8847926267281108,
      "grad_norm": 0.08077914267778397,
      "learning_rate": 2.0995705333668948e-06,
      "loss": 0.0006,
      "step": 283
    },
    {
      "epoch": 3.8986175115207375,
      "grad_norm": 0.012258109636604786,
      "learning_rate": 2.0762376643597586e-06,
      "loss": 0.0001,
      "step": 284
    },
    {
      "epoch": 3.912442396313364,
      "grad_norm": 0.012420260347425938,
      "learning_rate": 2.0529427393158704e-06,
      "loss": 0.0001,
      "step": 285
    },
    {
      "epoch": 3.9262672811059907,
      "grad_norm": 0.03773212060332298,
      "learning_rate": 2.0296878440784164e-06,
      "loss": 0.0001,
      "step": 286
    },
    {
      "epoch": 3.9400921658986174,
      "grad_norm": 0.03834720700979233,
      "learning_rate": 2.006475060906283e-06,
      "loss": 0.0001,
      "step": 287
    },
    {
      "epoch": 3.953917050691244,
      "grad_norm": 0.06677021831274033,
      "learning_rate": 1.9833064682876175e-06,
      "loss": 0.0001,
      "step": 288
    },
    {
      "epoch": 3.967741935483871,
      "grad_norm": 0.010336378589272499,
      "learning_rate": 1.9601841407537157e-06,
      "loss": 0.0,
      "step": 289
    },
    {
      "epoch": 3.9815668202764978,
      "grad_norm": 0.003101527225226164,
      "learning_rate": 1.937110148693265e-06,
      "loss": 0.0,
      "step": 290
    },
    {
      "epoch": 3.9953917050691246,
      "grad_norm": 0.033553846180438995,
      "learning_rate": 1.9140865581669627e-06,
      "loss": 0.0001,
      "step": 291
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.033553846180438995,
      "learning_rate": 1.8911154307225204e-06,
      "loss": 0.0001,
      "step": 292
    },
    {
      "epoch": 4.013824884792626,
      "grad_norm": 0.050998423248529434,
      "learning_rate": 1.8681988232100674e-06,
      "loss": 0.0,
      "step": 293
    },
    {
      "epoch": 4.027649769585254,
      "grad_norm": 0.016056543216109276,
      "learning_rate": 1.8453387875979834e-06,
      "loss": 0.0,
      "step": 294
    },
    {
      "epoch": 4.04147465437788,
      "grad_norm": 0.003037769114598632,
      "learning_rate": 1.822537370789163e-06,
      "loss": 0.0,
      "step": 295
    },
    {
      "epoch": 4.055299539170507,
      "grad_norm": 0.01739078015089035,
      "learning_rate": 1.7997966144377328e-06,
      "loss": 0.0001,
      "step": 296
    },
    {
      "epoch": 4.0691244239631335,
      "grad_norm": 0.012232556939125061,
      "learning_rate": 1.7771185547662417e-06,
      "loss": 0.0,
      "step": 297
    },
    {
      "epoch": 4.082949308755761,
      "grad_norm": 0.0021200303453952074,
      "learning_rate": 1.754505222383337e-06,
      "loss": 0.0,
      "step": 298
    },
    {
      "epoch": 4.096774193548387,
      "grad_norm": 0.0028847770299762487,
      "learning_rate": 1.7319586421019383e-06,
      "loss": 0.0,
      "step": 299
    },
    {
      "epoch": 4.110599078341014,
      "grad_norm": 0.02263806015253067,
      "learning_rate": 1.7094808327579401e-06,
      "loss": 0.0001,
      "step": 300
    },
    {
      "epoch": 4.124423963133641,
      "grad_norm": 0.008537916466593742,
      "learning_rate": 1.6870738070294412e-06,
      "loss": 0.0,
      "step": 301
    },
    {
      "epoch": 4.138248847926267,
      "grad_norm": 0.0025778906419873238,
      "learning_rate": 1.6647395712565256e-06,
      "loss": 0.0,
      "step": 302
    },
    {
      "epoch": 4.152073732718894,
      "grad_norm": 0.008922383189201355,
      "learning_rate": 1.6424801252616186e-06,
      "loss": 0.0,
      "step": 303
    },
    {
      "epoch": 4.1658986175115205,
      "grad_norm": 0.008659109473228455,
      "learning_rate": 1.6202974621704176e-06,
      "loss": 0.0,
      "step": 304
    },
    {
      "epoch": 4.179723502304148,
      "grad_norm": 0.006877637468278408,
      "learning_rate": 1.5981935682334266e-06,
      "loss": 0.0001,
      "step": 305
    },
    {
      "epoch": 4.193548387096774,
      "grad_norm": 0.00891530979424715,
      "learning_rate": 1.5761704226481078e-06,
      "loss": 0.0001,
      "step": 306
    },
    {
      "epoch": 4.207373271889401,
      "grad_norm": 0.005250364542007446,
      "learning_rate": 1.5542299973816626e-06,
      "loss": 0.0,
      "step": 307
    },
    {
      "epoch": 4.221198156682028,
      "grad_norm": 0.005915890447795391,
      "learning_rate": 1.5323742569944573e-06,
      "loss": 0.0,
      "step": 308
    },
    {
      "epoch": 4.235023041474655,
      "grad_norm": 0.008372033014893532,
      "learning_rate": 1.5106051584641208e-06,
      "loss": 0.0,
      "step": 309
    },
    {
      "epoch": 4.248847926267281,
      "grad_norm": 0.0033532341476529837,
      "learning_rate": 1.4889246510103078e-06,
      "loss": 0.0,
      "step": 310
    },
    {
      "epoch": 4.2626728110599075,
      "grad_norm": 0.017146175727248192,
      "learning_rate": 1.4673346759201728e-06,
      "loss": 0.0001,
      "step": 311
    },
    {
      "epoch": 4.276497695852535,
      "grad_norm": 0.010326577350497246,
      "learning_rate": 1.44583716637454e-06,
      "loss": 0.0,
      "step": 312
    },
    {
      "epoch": 4.290322580645161,
      "grad_norm": 0.0025458873715251684,
      "learning_rate": 1.4244340472748076e-06,
      "loss": 0.0,
      "step": 313
    },
    {
      "epoch": 4.304147465437788,
      "grad_norm": 0.0022526225075125694,
      "learning_rate": 1.403127235070587e-06,
      "loss": 0.0,
      "step": 314
    },
    {
      "epoch": 4.317972350230415,
      "grad_norm": 0.029883896932005882,
      "learning_rate": 1.381918637588112e-06,
      "loss": 0.0001,
      "step": 315
    },
    {
      "epoch": 4.331797235023042,
      "grad_norm": 0.005568996071815491,
      "learning_rate": 1.3608101538593965e-06,
      "loss": 0.0,
      "step": 316
    },
    {
      "epoch": 4.345622119815668,
      "grad_norm": 0.005829329136759043,
      "learning_rate": 1.3398036739522088e-06,
      "loss": 0.0,
      "step": 317
    },
    {
      "epoch": 4.359447004608295,
      "grad_norm": 0.010301432572305202,
      "learning_rate": 1.3189010788008234e-06,
      "loss": 0.0,
      "step": 318
    },
    {
      "epoch": 4.373271889400922,
      "grad_norm": 0.0017158942064270377,
      "learning_rate": 1.2981042400376032e-06,
      "loss": 0.0,
      "step": 319
    },
    {
      "epoch": 4.387096774193548,
      "grad_norm": 0.00932268425822258,
      "learning_rate": 1.277415019825417e-06,
      "loss": 0.0,
      "step": 320
    },
    {
      "epoch": 4.400921658986175,
      "grad_norm": 0.002834129147231579,
      "learning_rate": 1.2568352706908937e-06,
      "loss": 0.0,
      "step": 321
    },
    {
      "epoch": 4.414746543778802,
      "grad_norm": 0.002881410764530301,
      "learning_rate": 1.2363668353585486e-06,
      "loss": 0.0,
      "step": 322
    },
    {
      "epoch": 4.428571428571429,
      "grad_norm": 0.01270334329456091,
      "learning_rate": 1.216011546585785e-06,
      "loss": 0.0,
      "step": 323
    },
    {
      "epoch": 4.442396313364055,
      "grad_norm": 0.010466398671269417,
      "learning_rate": 1.195771226998789e-06,
      "loss": 0.0,
      "step": 324
    },
    {
      "epoch": 4.456221198156682,
      "grad_norm": 0.00680310744792223,
      "learning_rate": 1.1756476889293269e-06,
      "loss": 0.0001,
      "step": 325
    },
    {
      "epoch": 4.470046082949309,
      "grad_norm": 0.012853645719587803,
      "learning_rate": 1.1556427342524698e-06,
      "loss": 0.0,
      "step": 326
    },
    {
      "epoch": 4.483870967741936,
      "grad_norm": 0.009427334181964397,
      "learning_rate": 1.1357581542252555e-06,
      "loss": 0.0,
      "step": 327
    },
    {
      "epoch": 4.497695852534562,
      "grad_norm": 0.002769877901300788,
      "learning_rate": 1.1159957293262888e-06,
      "loss": 0.0,
      "step": 328
    },
    {
      "epoch": 4.511520737327189,
      "grad_norm": 0.003061062190681696,
      "learning_rate": 1.0963572290963298e-06,
      "loss": 0.0,
      "step": 329
    },
    {
      "epoch": 4.525345622119816,
      "grad_norm": 0.02316022664308548,
      "learning_rate": 1.0768444119798357e-06,
      "loss": 0.0001,
      "step": 330
    },
    {
      "epoch": 4.539170506912442,
      "grad_norm": 0.005801917053759098,
      "learning_rate": 1.0574590251675145e-06,
      "loss": 0.0,
      "step": 331
    },
    {
      "epoch": 4.552995391705069,
      "grad_norm": 0.006601040717214346,
      "learning_rate": 1.0382028044398823e-06,
      "loss": 0.0,
      "step": 332
    },
    {
      "epoch": 4.566820276497696,
      "grad_norm": 0.0029323461931198835,
      "learning_rate": 1.0190774740118343e-06,
      "loss": 0.0,
      "step": 333
    },
    {
      "epoch": 4.580645161290323,
      "grad_norm": 0.02404218353331089,
      "learning_rate": 1.0000847463782615e-06,
      "loss": 0.0,
      "step": 334
    },
    {
      "epoch": 4.594470046082949,
      "grad_norm": 0.0017514690989628434,
      "learning_rate": 9.812263221607114e-07,
      "loss": 0.0,
      "step": 335
    },
    {
      "epoch": 4.6082949308755765,
      "grad_norm": 0.006374394986778498,
      "learning_rate": 9.625038899551162e-07,
      "loss": 0.0,
      "step": 336
    },
    {
      "epoch": 4.622119815668203,
      "grad_norm": 0.009410557337105274,
      "learning_rate": 9.439191261805894e-07,
      "loss": 0.0001,
      "step": 337
    },
    {
      "epoch": 4.635944700460829,
      "grad_norm": 0.006760374642908573,
      "learning_rate": 9.254736949293216e-07,
      "loss": 0.0,
      "step": 338
    },
    {
      "epoch": 4.649769585253456,
      "grad_norm": 0.002450983738526702,
      "learning_rate": 9.07169247817579e-07,
      "loss": 0.0,
      "step": 339
    },
    {
      "epoch": 4.663594470046083,
      "grad_norm": 0.006942540407180786,
      "learning_rate": 8.890074238378074e-07,
      "loss": 0.0,
      "step": 340
    },
    {
      "epoch": 4.67741935483871,
      "grad_norm": 0.0011738522443920374,
      "learning_rate": 8.709898492118885e-07,
      "loss": 0.0,
      "step": 341
    },
    {
      "epoch": 4.691244239631336,
      "grad_norm": 0.002247450640425086,
      "learning_rate": 8.531181372455161e-07,
      "loss": 0.0,
      "step": 342
    },
    {
      "epoch": 4.705069124423963,
      "grad_norm": 0.007639207877218723,
      "learning_rate": 8.353938881837445e-07,
      "loss": 0.0,
      "step": 343
    },
    {
      "epoch": 4.71889400921659,
      "grad_norm": 0.03200659900903702,
      "learning_rate": 8.178186890677029e-07,
      "loss": 0.0001,
      "step": 344
    },
    {
      "epoch": 4.732718894009217,
      "grad_norm": 0.006035828962922096,
      "learning_rate": 8.003941135924859e-07,
      "loss": 0.0,
      "step": 345
    },
    {
      "epoch": 4.746543778801843,
      "grad_norm": 0.008785420097410679,
      "learning_rate": 7.83121721966248e-07,
      "loss": 0.0,
      "step": 346
    },
    {
      "epoch": 4.76036866359447,
      "grad_norm": 0.0018518904689699411,
      "learning_rate": 7.66003060770498e-07,
      "loss": 0.0,
      "step": 347
    },
    {
      "epoch": 4.774193548387097,
      "grad_norm": 0.0026455053593963385,
      "learning_rate": 7.490396628216237e-07,
      "loss": 0.0,
      "step": 348
    },
    {
      "epoch": 4.788018433179723,
      "grad_norm": 0.011131388135254383,
      "learning_rate": 7.322330470336314e-07,
      "loss": 0.0,
      "step": 349
    },
    {
      "epoch": 4.8018433179723505,
      "grad_norm": 0.00781510304659605,
      "learning_rate": 7.155847182821524e-07,
      "loss": 0.0,
      "step": 350
    },
    {
      "epoch": 4.815668202764977,
      "grad_norm": 0.007307104766368866,
      "learning_rate": 6.990961672696908e-07,
      "loss": 0.0,
      "step": 351
    },
    {
      "epoch": 4.829493087557603,
      "grad_norm": 0.007883993908762932,
      "learning_rate": 6.827688703921407e-07,
      "loss": 0.0,
      "step": 352
    },
    {
      "epoch": 4.84331797235023,
      "grad_norm": 0.029551049694418907,
      "learning_rate": 6.666042896065983e-07,
      "loss": 0.0,
      "step": 353
    },
    {
      "epoch": 4.857142857142857,
      "grad_norm": 0.00934284646064043,
      "learning_rate": 6.506038723004484e-07,
      "loss": 0.0,
      "step": 354
    },
    {
      "epoch": 4.870967741935484,
      "grad_norm": 0.003877162467688322,
      "learning_rate": 6.347690511617693e-07,
      "loss": 0.0,
      "step": 355
    },
    {
      "epoch": 4.88479262672811,
      "grad_norm": 0.005847269669175148,
      "learning_rate": 6.191012440510469e-07,
      "loss": 0.0,
      "step": 356
    },
    {
      "epoch": 4.8986175115207375,
      "grad_norm": 0.001081401132978499,
      "learning_rate": 6.036018538742208e-07,
      "loss": 0.0,
      "step": 357
    },
    {
      "epoch": 4.912442396313364,
      "grad_norm": 0.002679765224456787,
      "learning_rate": 5.882722684570638e-07,
      "loss": 0.0,
      "step": 358
    },
    {
      "epoch": 4.926267281105991,
      "grad_norm": 0.011573799885809422,
      "learning_rate": 5.731138604209169e-07,
      "loss": 0.0,
      "step": 359
    },
    {
      "epoch": 4.940092165898617,
      "grad_norm": 0.0016941127832978964,
      "learning_rate": 5.581279870597866e-07,
      "loss": 0.0,
      "step": 360
    }
  ],
  "logging_steps": 1,
  "max_steps": 432,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 72,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.298790820660537e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}